THEBRYGGENPAPERS
MainSeriesNo6
THEBRYGGENPAPERS
giveascholarlypresentationofthearchaeologicalfindsfromtheexcavationsatBryggenandother
medievalandearlymodernsitesinBergen.
Thepapersconsistoftwoseries,theMainSeriesandtheSupplementarySeries.
EditorialboardappointedbytheUniversityofBergen:
ProfessorIngvildØye
SeniorExecutiveOfficerAnnChristensson
ProfessorElseMundal
SenioradvisorAnneÅgotnes
IngvildØyeactsasChiefEditorforbothseries.
PublishedintheMainSeries
Vol1 (1984)AsbjørnE.Herteig:TheArchaeologicalExcavationsatBryggen.‘TheGermanWharf ’
inBergen1955-68.ArneEmilChristensen:BoatFindsfromBryggen.
Vol2 (1988)IngvildØye:TextileEquipmentanditsWorkingEnvironment,BryggeninBergen
c1150-1500.
Vol3 Part1(1990)AsbjørnE.Herteig:TheBuildingsatBryggen,theirTopograhicalandChronologicalDevelopment.
Vol3 Part2(1991)AsbjørnE.Herteig:TheBuildingsatBryggen,theirTopographicalandChronologicalDevelopment.
Vol4 (1992)ArneJ.Larsen:FootwearfromtheGullskoenAreaofBryggen.
Vol5 (2004)OleMikalOlsen.MedievalFishingTacklefromBergen.HelgeSørheim:Borgundand
BorgundfjordFishing.
PublishedintheSupplementarySeries
No1 (1984)Studiesontheearliestfarmsettlement,thefirstbuilt-upareaalongtheshore,animal
hairproducts,coins,andsealjugs.
No2 (1988)PresentationofrunicinscriptionsfoundatBryggen.
No3(1988)Brewing,cordageproducts,soundtollsandmusic.
No4(1989)TheBryggenPottery1.
No5 (1994)TheBryggenPottery2.
No6 (1998)MedievalFiresinBergen–Revisited.
No7 (2000)ShipsandCommodities.
THEBRYGGENPAPERS
MainSeries
No6
BERGENc800-c1170
THEEMERGENCEOFATOWN
GitteHansen
Fagbokforlagetas
©2005by
FagbokforlagetVigmostad&BjørkeAS
AllRightsReserved
Printedinxxxbyxxx
ISBN82-xxx-xxxx-x
Published with a grant from Skolebestyrer B.E. Bendixens legat, University of Bergen, Bergen
UniversityMuseumandThefacultyofArts,UniversityofBergen
Distributionoffice
Fagbokforlaget,Vigmostad&BjørkeAS
Layout:Media&CommunicationCentre,UniversityofBergen
Inquiriesaboutthistextcanbedirectedto
Fagbokforlaget
Kanalveien51
N-5068Bergen
P.O.Box6050Postterminalen
N-5892Bergen
Telephone +4755388800
Telefax +4755388801
www.fagbokforlaget.no
fagbokforlaget@fagbokforlaget.no
Allrightsreserved.Nopartofthispublicationcanbereproduced,storedinretrievalsystem,or
transmitted,inanyformorbyanymeans,electronic,mechanical,photocopying,recording,or
otherwise,withoutthepriorwrittenpermissionofthepublisher.
FOREWORD
InthisvolumeoftheBryggenPaperswepresentastudyoftheprocessesoftheurbandevelopment
ofBergen,howthetownemergedanddevelopedintoanimportanturbancommunitybytheearly
MiddleAges.Thestudyisprimarilybasedoncontemporaryarchaeologicalsourcematerialfromc.
800toc.1170-acomplexandcompositematerial,comprisingtracesofcultivation,culturelayers,
buildings, plots and artefacts. Its main aim is to investigate the town’s structure, plots and plots
systems,andthedifferentactivities,craftsandproductionaswellasthecharacteroftheurbansettlementanditsdevelopmentuntilaround1170.Themainquestionsthatareaddressedarewhen,how,
whyandontheinitiativeofwhomBergenmergedasatown
ThepublicationofthisvolumehasbeenfinancedbyTheFacultyofArts,UniversityofBergen,
BergenUniversityMuseum,andskolebestyrerB.E.Bendixen’slegateattheUniversityofBergen.
TheeditorialboardresponsibleforthepublicationoftheseriesconsistsofSeniorExecutiveOfficer
AnnChristensson,DirectorateforCulturalHeritage,DistrictOfficeWest,Bergen,ProfessorElse
Mundal,CentreofMedievalStudies,UniversityofBergen,SeniorAdvisorAnneÅgotnes,Bryggens
Museum,andProfessorIngvildØye,DepartmentofArchaeology,UniversityofBergen.
Bergen,November2005
IngvildØye
ChiefEditor
5
6
CONTENTS
Foreword........................................................................... 5
Acknowledgements .................................................................. 13
PARTI
17
AIMS,BACKGROUND,THEORETICAL,METHODOLOGICAL
APPROACHESANDSOURCES
17
1Introduction...................................................................... 17
2Thebackground................................................................... 19
Whatisatown? .....................................................................19
Geographicalsetting.................................................................20
Historicalsetting....................................................................20
EarlyBergen,stateofresearch..........................................................23
TheoriginofBergen ...............................................................23
Thephysicallayoutoftheearliesttown.................................................27
Artefactstudies...................................................................29
3Thepresentstudy,theoreticalapproachesanddemarcations................................. 30
Thepart-studies.....................................................................33
Horizon1(c800-c1020/30),abackdrop ...............................................33
Plotsandplotsystems ..............................................................33
TowhatextentwastheBergenarea‘occupied’? ...........................................34
Craftsandproduction ..............................................................34
Trade...........................................................................34
Thecharacterofthesettlementontheplots.............................................35
4Generalpresentationofthearchaeological,botanicalandtopographicalsources.................. 35
Investigationsbefore1899...........................................................36
Investigationsfromthelate1800suntilc1920 ...........................................37
Investigationscarriedoutbetween1929and1955........................................37
Investigationscarriedoutbetween1955and1979........................................38
Investigationscarriedoutfrom1980until1998..........................................39
Recentstudiesofthemedievalchurches................................................40
Botanicalinvestigations .............................................................40
Randomobservations..............................................................42
Geo-technicalinvestigations.........................................................42
Maps ...........................................................................42
5Generalmethodologicalapproaches,definitionsanddemarcations............................ 42
TheBergenarea..................................................................42
Adiachronicapproach..............................................................42
Spatialanalysis....................................................................43
Classificationofthematerialintobasic,supplementaryorgeneralbackgroundsources .............43
Theplotasananalyticunit ..........................................................48
Levelofinquiry ...................................................................50
Landuse,terminology..............................................................50
Approachestotheartefactmaterial....................................................50
7
6Reconstructionofthenaturaltopographyabout1000 ...................................... 53
Stateofresearch .....................................................................53
Methodologicalapproachesandpremisesforthereconstructionofthenaturaltopographyabout1000...54
Majorfeaturesofthereconstructednaturaltopographyabout1000.............................55
7Evaluationofthearchaeologicalandbotanicalsources ...................................... 56
TheHolmenarea....................................................................56
Site1,Koengen(BotanicalinvestigationinVeisanbyKariLoeHjelle)(1986)...................56
Site2,TheChristchurchCathedral(StoreKristkirke) ......................................57
Site3,Christchurchminor(LilleKristkirke) .............................................57
Site4,TheChurchoftheApostles(Apostelkirken) ........................................57
Site5,ØysteinMagnusson’shallatHolmen.............................................58
Thenortherntownarea ...............................................................58
Site6,Bryggen(1955-1979)BRM0...................................................58
Site7,ØvreDreggsalmenningen(1989)BRM298.......................................67
Site8,Dreggsalmenningen14-16(1986and1990)BRM237...............................70
Site9,Sandbrugaten5(1967)BRM3 .................................................72
Site10,Sandbrugaten3(1953).......................................................77
Site11,Dreggsalmenningen20(1967)BRM4...........................................77
Site12,Dreggsalmenningen10-12(1972)BRM42......................................80
Site13,Dreggsalmenningen10-16(1986)BRM242.....................................80
Site14,Dreggsalmenningen(1979)BRM83............................................81
Site15,Stallen,Svensgården(1980/82)BRM90 .........................................82
Site16,Bryggeparken(1989)BRM287...............................................83
Site17,Nikolaikirkealmenningen(1985)BRM202......................................83
Site18,Koren-WibergsPlass(1980)BRM143..........................................84
Site19,Wesenbergsmauet(1989)BRM297............................................84
Site20,Øvregaten39(1981)BRM94.................................................84
Site21,Klingesmauet(1989)BRM299................................................87
Site22,Kroken3(1984)BRM223 ...................................................88
Site23,TheChurchofStMary(Mariakirken)...........................................89
Site24,TheChurchofStPeter(Peterskirken)............................................90
Site25,TheChurchofStOlavontheHill(OlavskirkenpåBakkene).........................91
Themiddletownarea................................................................91
Site26,Finnegården6a(1981)BRM104...............................................91
Site27,Finnegården3a(1982)BRM110..............................................92
Site28,Rosenkrantzgaten4(1978/79and1981)BRM76.................................94
Site29,Vetrlidsalmenningen2,Kjøttbasaren(1996and1997)(BRM490).....................95
Site30,Vetrlidsalmenningen(1991/92)BRM342 ........................................95
Site32,TheChurchofStNicholas(Nikolaikirken).......................................98
Site33,TheChurchofStColumba(Steinkirken) .........................................99
Thesoutherntownarea ...............................................................99
Site34,LilleØvregatenfriområde(1994)BRM465.......................................99
Site35,Korskirken(1984)BRM200..................................................99
Site36,Skostredet10(1992)BRM346................................................99
Site37,NedreKorskirkealmenning/Vågsalmenning(1998)BRM544........................100
Site38,Domkirkegaten6(1987)BRM245...........................................101
Site39,TheChurchofStOlavinVågsbotn(OlavskirkeniVågsbotn)........................102
Site40,TheChurchofStCross(Korskirken) ...........................................102
8
TheNordnesandNonneseterareas.....................................................103
Site41,Rådstuplass2-3,‘Vestlandsbanken’(1963)BRM20................................103
Site42,Nygaten2(1991)BRM333..................................................103
Site43,TheMunkelivBenedictineAbbeywiththeChurchofStMichael(Munkelivkloster).......103
Site44,StJohn’sAugustinianAbbey(Jonskloster)........................................103
Site45,TheChurchofAllSaints(AlleHelgensKirke) ....................................103
Site46,TheNonneseterconvent(NonneseterKloster) ....................................105
Thespatialandtemporaldistributionofthesources........................................105
Thetemporaldistribution ..........................................................105
Thespatialdistribution ............................................................106
Therepresentativityoftheartefactandecofactmaterial...................................108
PARTII
127
MAJORINITIATIVESANDDAILYACTIVITIESINEARLYBERGEN
127
8Horizon1(c800-c1020/30),abackdrop.............................................. 127
Locationandgenerallanduse.........................................................127
Urbanornon-urban? ................................................................128
The‘Holmensettlement’...........................................................128
ThePieratsite30 ................................................................130
Conclusions .......................................................................131
9Plotsandplotsystemsinthetownarea................................................ 131
Plotboundaries....................................................................131
Horizon2(c1020/30-c1070)......................................................132
Horizon3(c1070-c1100).........................................................134
Horizon4(c1100-1120s)..........................................................134
Horizon5(1120s-c1170)..........................................................134
Oneorseveralplotsystems?..........................................................138
Evaluationoftheplotsystemsdiscernedandcentraldatesofsourcesassignedtohorizons2and3.....140
Theextentofthetwoplotsystems......................................................141
Thehorizon3system.............................................................141
Thehorizon2system.............................................................143
Conclusions .......................................................................144
10TowhatextentwastheBergenarea‘occupied’?......................................... 145
Horizon2(c1020/30-c1070)......................................................145
Horizon3(c1070-c1100).........................................................147
Horizon4(c1100-1120s)..........................................................150
Horizon5(1120s-c1170)..........................................................152
Conclusions .......................................................................156
11CraftsandproductioninearlyBergen................................................ 157
Placesofproduction................................................................159
Combmakingandmiscellaneousantler,bone,hornandwhale/walrusboneworking.............159
Shoemakingandotherleatherworking................................................162
Metalworking...................................................................165
Stoneworking...................................................................168
Woodworking ...................................................................170
Skinning.......................................................................172
Textileproduction ................................................................172
9
Fishing........................................................................174
Huntingandwar.................................................................176
AgricultureinearlyBergen.........................................................176
Basiccooking,foodandbeverageprocessing ............................................177
Summary.......................................................................180
Whatwasthenatureoftheproductiveactivitiesandhowweretheyorganised? ....................180
Combmaking...................................................................180
Miscellaneousantler,bone,hornandwhale/walrusboneworking ............................185
Shoemaking....................................................................186
‘Otherleatherwork’ ...............................................................189
Metalworking...................................................................191
Stoneworking...................................................................194
Woodworking ...................................................................196
Skinning.......................................................................199
Textileproduction ................................................................199
Fishing,huntingand,farming .......................................................200
Basiccookingandfoodandbeverageprocessing.........................................200
Summary.......................................................................203
WereanyoftheproductiveactivitiesfundamentalfortheemergenceofBergen? ...................203
12Trade......................................................................... 205
Horizon2(c1020/30-c1070)......................................................207
Horizon3(c1070-c1100).........................................................209
Horizon4(c1100-1120s)..........................................................210
Horizon5(1120s-c1170)..........................................................212
Conclusions .......................................................................217
13Thecharacterofthesettlementinthetownarea........................................ 218
Horizon3(c1070-c1100).........................................................219
Horizon4(c1100-1120s)..........................................................219
Horizon5(1120s-c1170)..........................................................219
Conclusions .......................................................................220
PARTIII
THESYNTHESIS
221
221
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?................ 221
How,when–andbywhom?..........................................................221
Anewmajorinitiative,horizon3(c1070-c1100).......................................224
Horizon4(c1100-1120s)..........................................................225
Horizon5(1120s-c1170)..........................................................226
Conclusions....................................................................228
WhywasBergenfounded?Andhowdidthetowndevelop?...................................228
ThenewtownintheBergenarea,horizon2(1020/30-c1070)..............................230
OlavKyrre’sBergen,horizon3(c1070-c1100).........................................231
Bergenduringhorizon4(c1100-1120s)...............................................233
Bergenduringhorizon5(1120s-c1170)...............................................235
Conclusions....................................................................237
15Conclusions .................................................................... 237
10
APPENDIXES
241
Appendix1....................................................................... 241
Sourcesforthenaturaltopographyabouttheyear1000anddiscussionofthecourseofthecontourlinesin
thereconstruction..................................................................241
Appendix2....................................................................... 257
DateddendrochronologicalsamplesfromearlyBergen......................................257
Appendix3....................................................................... 261
Elevenartefactassemblagesfromsite9,Sandbrugaten5(1967)BRM3.........................261
Appendix4....................................................................... 265
Sevenartefactassemblagesfromsite11,Dreggsalmenningen20BRM4(1967)...................265
FOOTNOTES
268
LISTOFFIGURES
274
LISTOFTABLES
277
Maps............................................................................279
Publishedandunpublishedtitles.......................................................279
REFERENCES
279
INDEX
295
11
12
ACKNOWLEDGEMENTS
FromDecember1996untilJune2001,IwasemployedasajuniorresearchfellowattheDepartment
ofArchaeology,UniversityofBergen,financedbyascholarshipfromtheFacultyofArtsattheUniversityofBergen.FromJune2001,IhavebeenemployedasacuratorattheMedievalCollectionsof
theBergenUniversityMuseum.Ithanktheseinstitutionsforprovidingmewithfinancialsupport
andexcellentworkingconditionswhilecarryingoutthepresentstudy.
Manypeoplehavebeenengagedinandsupportedmeincarryingoutthisproject.Iamgratefulto
mysupervisor,ProfessorIngvildØye,whocarefullyandeffectivelyreadandcommentedondrafts
timeandagain.Shehasprobablyexhaustedmanygreycellsandcertainlymanygreypencilsonmy
project,butmostimportantlyshehasgivenmevaluablefeedbackonalllevels.
Several researchers from various disciplines and institutions have contributed with analyses of
thematerial:DendrochronologicalanalyseshavebeencarriedoutbyTerjeThunattheInstituteof
Botany,attheUniversityofScienseandTechnologyTrondheim.Petrologicalanalysesofstoneitems
werecarriedoutbyHelgeAskvikandHaraldFurnesattheDepartmentofEarthScience,University
ofBergen,andØysteinJJansenatBergenUniversityMuseum,Geology.KirstiRisøenattheSection
ofConservationatBergenUniversityMuseumhasbeenresponsibleformetallurgicalanalysesand
AnneKarinHufthammeratBergenUniversityMuseum,Zoologycarriedoutosteologicalanalyses.
Iamgratefulfortheirwillingassistance.
I also thank colleagues who kindly provided me with illustrations and contextual information
onfindsfromtheircollections:JesperHjermindandTuriThomsenatViborgStiftsmuseum(Denmark),SoniaJefferyatLödöseMuseum,GunillaGardelinatKultureninLund(Sweden),Ingrid
UlbrichtSchlossGottorf(Germany),TinaWibergattheNorwegianInstituteforCulturalHeritage
Research(NIKU)DistrictOfficeOslo,andJosteinBergstølattheUniversityMuseumofCultural
History,UniversityofOslo.
In the early phase of the project Mona Mortensen in the ‘Dokumentasjonsprosjektet’ gave me
accesstothepreliminarydigitisedversionoftheBergenUniversityMuseum’saccessioncatalogue
(‘tilvekst’).LynBlackmore,PerKristianMadsen,IanReedandAlanGVincegavemeadviceand
referencesforpotterydates.JanBillansweredmyquestionsonboats.Ithankthemforgivingme
theirtime.
ManfredThalleroftheformerHITcentre(nowAKSIS),SørenDiinhofandDavidSimpsonat
BergenUniversityMuseumhelpedmewithGeographicalInformationSystems.Iwouldalsoliketo
thankthestaffattheITdepartmentoftheFacultyofArtsfortheircompetentandexpedienthelp
onthetechnicalaspectsofIT.
EllinorHoffattheCulturalHistoryCollectionsofBergenUniversityMuseumdrewartefactsand
SveinSkareatthePhotoSectionofBergenUniversityMuseumphotographedartefactsfromBergen
University Museum. Melanie Wriggelsworth has proof-read the preliminary manuscript. I thank
themfortheirpatience.
ThroughouttheprojectIhavehadmyofficeatBryggen’sMuseumthathousesafriendlyandgenerouslotofmediaevalistsandotherbeings.Ihavediscussedideasandproblemswithmostofthem.
ThankstoallfriendsandcolleaguesattheDepartmentofArchaeologyUniversityofBergen,atthe
NorwegianInstituteforCulturalHeritageResearch(NIKU)DistrictOfficeBergen,attheDirectorateforCulturalHeritageDistrictOfficeWest,andattheBryggen’sMuseumFoundation.
IwouldliketoextendspecialthankstoArneLarsenandEgillReimerswhoguidedmethrough
‘their’storeroomsandarchivesattheMedievalCollectionsofBergenUniversityMuseum.Theyhave
givenmegreathelpwiththeirprofoundinsightinthematerialandhavebeenreadywithahelping
handandfordiscussionsatalltimes.RoryDunlop,hasbeenmyhouse-expertonpottery,myonline
dictionaryoftheEnglishlanguageandagenerousdiscussionpartner.ThanksalsotoOleMagne
NøttveitandOleMikalOlsenwholetmeusedatafromtheirthesesandtoSigridSamsetwhohas
contributedwithadviceonliteratureonchildrenintheMiddleAgesandwithasteadyhandwhen
13
calculatingshoesizesformedievalchildren.HanneMereteRosseidMoldungvolunteeredherselfas
myassistantfortwoweeks.Sheworkedinthestorerooms,puncheddata,madesketchesofartefacts
andwasagooddiscussionpartner.IamalsoindebtedtoFrodeIversenforhiseverinspiredattitude
towardsourmétierandforhisgeneroushelpin‘emergencysituations’intheITbusiness.Frodealso
readandcommentedondraftsofpartofmymanuscriptasdidKnutHøiås,ArneLarsen,Janicke
LarsenandVidarTrædal.
IalsoacknowledgeStewartClarkeattheNorwegianUniversityofScienceandTechnologyforhis
finalproof-readingofthisthesis.
LastbutnotleastIwanttothankmychildrenFrida,SigurdandPerViggo,forsimplybeingthere.
–AndthankstoyouKnutAndreas;inbetweenrunningourfamilywehavehadinspiringandilluminatingdiscussions,youaremybestfriendandcolleague.
ThestudywasdefendedfortheDr.Art.degreeonMay282004.Professorem.HansAndersson
UniversityofLund(S)andProfessorElseRoesdalÅrhusUniversity(DK)wereopponents.Ithank
themfortheircommentsandforinterestingdiscussions.
Thethesiswasfirstpublishedin2003.Minoralterationshavebeenaddedtothepresentedition.
Bergen,September2005
GitteHansen
14
Bergenc800-c1170
TheEmergenceofaTown
15
16
PARTI
AIMS,BACKGROUND,THEORETICAL,
METHODOLOGICALAPPROACHESANDSOURCES
1INTRODUCTION
In the Middle Ages Bergen appeared as the
mostimportanttowninNorway.Fromtheend
ofthethirteenthcenturyBergenwasknownas
thecountry’slargesttradingcentreandfromthe
end of the twelfth century it was the ecclesiasticcentreofwesternNorway.Accordingtosaga
traditions, King Olav Kyrre (the Gentle, ’the
Peaceful’)(1066-1093)foundedthetown,probablyabout1070,andbasedondifferentsources
andmethodologicalapproaches,researchershave
studiedearlyBergenandtheking’srolethrough
thecenturies.Todayalargebodyofarchaeologicalmaterialcanbedrawnintothediscussionand
formsthebasisfornewapproaches.Thetheme
formystudyistheemergenceofBergenandthe
developmentofthetownuntilc1170.Thiscase
study of urban development in Scandinavia in
theearlyMiddleAgesisbaseduponheterogeneous source material comprising archaeological,
botanical,topographicalandwrittensources.
Myoverallaimistostudytheprocessesofhow
a place developed into a living urban communityintheinterplaybetweenpeoplefromdifferentlevelsofthesocialhierarchyandtheirwider
historicalcontext.Themainquestionstobeaddressedarehow,when,bytheinitiativesofwhom
and why did Bergen emerge. These basic questions are approached through six studies of majorinitiativesanddailyactivitiesreflectedinthe
availablesourcesoftheearlytownanditspeople.
Thestudiescompriseaninvestigationofactivity
intheBergenareabetweentheninthcenturyand
c1020/30,aswellasinvestigationsofplotsand
plotsystems,settlementdevelopment,craftsand
production,trade,andthecharacterofthesettlementinBergenbetweenc1020/30andc1170.
1Introducion
The archaeological remains, spanning from
tracesofcultivation,plots,buildings,culture-layers,toartefactsreflecthowmajorinitiativesand
daily activities in time shaped the urban community. My aim is to understand some of the
strategies behind these initiatives and activities
inordertoelucidatethequestionsofwhyandby
theinitiativeofwhomthetownemerged.
The period from the ninth century to about
1170isinvestigatedwithamainfocusonactivitiesbetweenc1020/30andc1170.Inorderto
obtainavariedandmorenuancedunderstanding of the processes of the urban development
duringthisperiodthesourcesareanalysedwithin a chronological framework of five horizons.
The time spans of the horizons are defined on
thebasisofthebeginningandendofphasesin
thearchaeologicalmaterial.Somehorizonsalso
coincide with events mentioned in the written
sources(horizons1to5,cfp55).Ihavechosen
c 1170 as the upper chronological limit for my
studyforratherpragmaticreasons.Iwantedto
study the early history of Bergen, with a focus
ontheeleventhandtwelfthcenturies.FiresdestroyedBergenin1170/71andin1198andleft
firelayersthatmarkthe‘endofphase’atmanyarchaeologicalsites.Fromapracticalpointofview
c1170or1198wouldthusbeconvenientplaces
to stop. The amount of archaeological data to
be analysed would become too large to handle
withinthepresentprojecthadIchosen1198as
theuppertimelimit,Ithereforechoosec1170.
The area around the Bay of Vågen denoted
as‘theBergenarea’(Figure1)iscoveredinthe
study.TheBergenareaisdividedintosixareas.
Thedivisionservesasananalytictoolandasa
reference when orientating oneself geographically.Thesixareascomprise(1)Holmen,(2)the
17
northern town area, (3) the middle town area,
(4)thesoutherntownarea,(5)theNonneseter
area, and (6) the Nordnes peninsula. An inlet,
in the High Middle Ages known as a swampy
area called Veisan, separated Holmen from the
northern town area. The stretch of land along
theVågenBayinthenorthernandmiddletown
areas is known today as Bryggen, the southern
townareaisknownasVågsbotn.Thenaturaltopographyabout1000willbereconstructedand
providesthespatialframeworkwithinwhichthe
sourcesareanalysedandinterpreted.
Contemporaryarchaeological,botanical,and
written material, as well as the reconstructed
natural topography form the empirical basis of
thisstudy.Howeveryoungerwrittenrecordsand
later patterns in the archaeological material are
drawn upon when relevant. The archaeological
andbotanicalmaterialcomprisesbothpublished
and unpublished data from investigations and
masonrystudiesonbuildingsandruinsfromaltogether46sitescoveringabout14924m2and
149 profiles in trenches. These sites have been
investigated from the nineteenth century until
Figure1.BergenonthewestcoastofNorway.TheBergenarea
18
1998. To simplify references the sites are numberedfrom1to46(Table21,p105).
I have regarded it as a methodological challenge to activate and thus be able to make use
ofasmuchofthematerialaspossible,whether
retrievedduringthenineteenthorthetwentieth
century.Inherentinthemethodsappliedisthat
the Bergen area is considered as one site where
data from the various sources will be analysed
spatially in relation to one another and to the
naturaltopography.Materialfromindividualarchaeological and botanical sites and their close
vicinitieswillbeinterpreteddrawinguponpatterns and main tendencies in the material discerned when the sources are considered on a
broaderbasis.Methodologicallytheproduction
of maps is used as an important analytic tool
for the visualisation and interpretation of the
sources.Thesourceswillalsobeevaluatedand
divided into different categories according to
theirreliabilityasevidence.Inthiswaysources
that are poorly dated or located may be drawn
into the study, while inherent uncertainties of
thematerialarekeptinmind.
Thearchaeological‘rawdata’iswithafewexceptionsdestroyedwhenexcavated.Asresearchersweareleftwithmaterialthathasbeendocumented to a varying degree and the excavator’s
interpretation. The Bergen material has come
to light through different methodological approaches and the questions posed in this study
have rarely been considered in the reports. Archaeological and other data do not tell a story
initself;wehavetoaskquestionsinordertoget
answersthatmaybeusedfurtheroninanalyses
anddiscussions.Inordertousethematerialas
sourcesformystudy,anumberofquestionsare
posed,rangingfrombasicquestionsofchronology and localisation of the single sites to questions on a higher level of abstraction involving
the study of patterns across the sites and in a
widerhistoricalcontext.Insomecasesthematerial consists of ‘hard facts’, it poses resistance
andthereisastraightandnarrowanswertothe
questions.Often,however,theanswersarecomplexandaninterpretationofthematerialisdependent on ‘circumstantial evidence’, chains of
indicationsandconvincingarguments.
Thethreepartdivisionofthethesisreflectsthe
complexprocessofanalysingthesources.PartI
presentstheaimsandsourcesandabackground
forthestudy.Italsogivesanoutlineofoverall
theoreticalandmethodologicalapproachestothe
sources.Thenaturaltopographyisreconstructed
hereandthesourcesareclassifiedsotheycanbe
usedinbroadanalysesacrossthesites.PartIIof
thethesiscomprisessixpart-studiesofthesources across the sites, addressing different themes
thataredrawnuponinpartIIIinthesynthesisingdiscussionsofhow,when,bytheinitiativeof
whomandwhyBergenemerged.
2THEBACKGROUND
Whatisatown?
Urbanisation displays great differences chronologicallyandgeographicallyanddifferentsociopolitical settings produce urban communities
with various functions and characteristics (eg
Andrén 1989; Hodges (1982) 1989). Through
thehistoryofresearchmanyattemptshavebeen
made to define the medieval town. Narrow legalistic approaches were in time abandoned for
2TheBackground
broader Kriterienbündel approaches; a ‘bundle’
ofcriteriawerelistedandsettlementsqualifiedas
townsiftheyfulfilledoneormoreofthecriteria
(forfurtherreferencesegSchück1926;Hodges
(1982)1989,20ff).
The Norwegian historians Knut Helle and
ArnvedNedkvitne’s(1977)‘bundle’ofstructural
andfunctionalcriteriaarequiterepresentativefor
historical, geographical and social criteria suggestedintheliteratureandarecommonlyused
inrelationtoVikingageandmedievaltownsin
modern Scandinavian research (eg Ambrosiani
andClarke1995(1991),3).AccordingtoHelle
andNedkvitneasettlementmaybedefinedasa
townifitispermanentanddenserinstructure
thansettlementsinitshinterland.Furthermore,
thesettlementshouldhavespecialisedfunctions
comparedtothesurroundingruralarea.These
specialisedfunctionsmaybeeconomic,jurisdictional,administrative,religious,and/orcultural
andthetownspeopleshouldpredominantlylive
off such activities. If a place was considered as
‘urban’intheeyesofcontemporarypeoplethis
isconsideredsufficientforthatplacetoqualify
asatowneveniftheplacewasapparentlysmall
and insignificant (Helle and Nedkvitne 1977,
190-191). Recently, ‘mental criteria’ such as an
urbanlifestylehasalsobeensuggestedasacriterionforaplacetoqualifyasa‘real’townthat
isfundamentallydifferentfromthesurrounding
ruralcommunity(Carelli2001,99).
The legalistic, functional, structural and
mentalcriteriacharacterisethepermanentlysettled, urban community but not seasonal marketplaces, like eigth century Ribe in Denmark
(Frandsen, Madsen, and Mikkelsen 1988, 8;
Jensen1992;Ferveile1994)ortheninthcentury
Löddeköpinge in Sweden (Ohlsson 1973). The
criteriarelatetoalivingurbancommunityand
presupposethatthiscommunityhasbeenunder
developmentforsometime.Thusthecriteriado
notrelatetowhatmaybedesignatedembryonic
stagesofanurbancommunity,neitherdothey
covertheplannedtownthatdidnotdevelopinto
a living urban community as for instance the
eighthcenturyplannedtownofAnjar(Hodges
2000,49ff).
ThroughthepresentcasestudyIaimtoelucidatehowalivingurbancommunitydeveloped
in the Bergen area and how this development
19
took place in the interplay between various actorsandthewidersociety.TheaimisnottodeterminewhenearlyBergenfulfilledasufficient
numberofcriteriatoqualifyasa‘real’medieval
townbutrathertopresentacasestudyofprocessestowardsapermanentlysettledlivingurban
communityinScandinavia.
Analytic tools are necessary when approachingthemyriadofsourcesavailable.Thebundle
offunctionalandstructuralcriteriasuggestedby
Helle and Nedkvitne apply to central themes,
some of which can be aptly discussed on the
basisofapredominatelyarchaeologicalbodyof
sources.Theymayserveasapointofdeparture
whendecidingwhichmajorinitiativesanddaily
activities to study and also as a loose frame of
referencewhendiscussingthestructuralfeatures
and different functions and activities discerned
intheBergenmaterial.
Geographicalsetting
Bergen grew around the Vågen Bay located on
theinnercoastofwesternNorway.IntheMiddleAges,Bergen’shinterlandwasrelativelyrich
inarablelandcomparedtolocalstandards,and
agriculture could be supplemented by fishing
andhunting.TheBergenareacouldbereached
fromthemainlandbyhorseoronfoot,butboat
wasnodoubtthebestmeansoftransportwhen
carryingaheavyload.BergenhadacentrallocationforseagoingtransportbetweenLofotenand
Vesteråleninthenorthandcontinentalnorthern
Europeanharbours.Thetownalsohadacentral
locationfortrafficacrosstheAtlanticheadingfor
Iceland,Greenland,theFaroeIslands,Shetland,
theOrkneysandtheBritishIslands(Helle1982,
53-70,withreferences).
Historicalsetting
TheemergenceofBergenshouldbeseeninthe
context of the considerable changes that took
placeinmanyimportantaspectsoflifebetween
theninthcenturyandc1170withinthemedievalboundariesofNorway.Thepoliticalsystem
inNorwaychangedtowardsacentralmonarchy.
Previously,powerhadbeencentredinthehands
ofpettykingsormagnates(Andersen1977,185).
Thepoliticalcentralisationwasaresultofalong
process that first comes to light in the written
recordswithHaraldHårfagre’s(HaraldFairhair)
20
efforts to win recognition as a king in the last
halfoftheninthcentury.Strongerconnections
toWesternEuropethroughraids,tradeandcolonisationhavebeenseenasthebackgroundfor
thiscentralisationprocess.Intheyearstocome
Harald’sdescendantsaimedtogainroyalpower
overthewholeorpartsofNorwayinopposition
to local magnates - especially the Lade earls of
Trøndelag - and in periods between 960 and
1034alsoinoppositiontotherulersofDenmark
who were distant overlords (Andersen 1977,
84ff). From 1034 and in the following three
centuriesroyalpowerwasinthehandsofNorwegiankings.Between1130and1240rivalling
jointkingsandpretenderstothethronefought
eachotherandcivilwarsravagedthecountry.
As part of the centralisation policy, Harald
established royal estates through land confiscation, at least in western Norway. Establishing
the royal estates has been seen as a strategy to
secure an economic foundation for the central
kingdom.Thecollectionoflandrent(landskyld)
-taxonlandpaidtolandowners-theking(and
lateralsotootherlaylandownersandecclesiastic
landowners)andveitslerageneraltaxpaidtothe
kingbyallpersonsliabletotaxationwereintroducedandperhapsadministeredfromtheroyal
estates (Andersen 1977, 88-99, 295ff). Both
landskyldandveitslerwerepaidinkind(KLNM
X277ff,XIX632).Of13possibleroyalestates
dated to before 1100, four were located in the
close vicinity of Bergen. Such concentration of
royalestatesisuniqueinwesternNorway.Alrekstad,about2kmsoutheastoftheBayofVågen
wasoneoftheroyalestatesalreadyfromthedays
of Harald. The others were Herdla, Seim, and
Lygra(Iversen2004).
During the reign of King Olav Haraldsson
(later Saint Olaf) (1015-1028) Christianity was
introduced as the official religion. Researchers
haveseentheofficialconversionasameansfor
centralkings-firstOlavTryggvason(994/995999/1000), later Olav Haraldsson and his successors throughout the eleventh century - to
strengthen royal territorial control over Norway.Thecentralkingwastherealleaderofthe
ChurchandprobablyusedtheChurchtoadministratetheland.Fromthelasthalfoftheeleventh
century churches were built (Skre 1995), the
kingsareknownasdonatorsoflandforchurches
andmonasteriesandfoundersofmanychurches
throughoutthecountry(Krag1995,191).Bishopswerechosenandappointedbythekings(Andersen1977,103,124,153,289-90).Aspartof
theking’sattendantguard(hird)thefirstbishops
travelledwiththeking.ThecountrywasnotdividedintodiocesesuntilthereignofOlavKyrre
(1066-1193).Theepiscopalresidenceofwestern
Norway was located at Selja, a small island on
the coast in the northern part of western Norway.Thebishop,however,mayhaveresidedona
regularbasisinBergenbeforetheepiscopalresidencewasformallyattachedtoBergen,probably
about 1170 (Helle 1982, 92, 146; Lidén 1993,
10). Tithe was introduced after 1111, during
thereignofthejointkingsØysteinMagnusson
(EysteinMagnusson)(1103-1123),SigurdMagnussonJorsalfar(SigurthJerusalemfarer)(11031130) and Olav Magnusson (1003-1115) (Andersen 1977, 181). In 1152/53 the Church was
formally given the right to administer its own
propertyandincomeandappointchurchleaders
and other clergy. It is uncertain to what extent
these rights were immediately carried into life,
butitseemsclearthattheChurchnowtookan
important step towards independence from the
Crown(Helle1995,31).
Towns were also introduced in Norway as a
newfeatureintheperiodstudiedhere.Theterm
town or urban is used here in accordance with
the wide bundle of town criteria suggested by
Helle and Nedkvitne (1977). The Viking Age
town Kaupang in Tjølling, Vestfold, or Sciringshealismentionedincontemporarysources
about890(HelleandNedkvitne1977,192)and
archaeologicallydatedtobetweenthelateeighth
centuryandthelateninthcenturies(Ambrosiani
andClarke1995(1991),65ff;Blindheim,Heyerdal-Larsen,andIngstad1999,162).OtherIron
Age towns may have existed in Norway. Toponymic evidence suggests that places where the
exchangeandtranshipmentofgoodstookplace
existedthroughouttheland,manyoftheseplaces
werelocalisedclosetotheseatsoflocalmagnates
orroyalestates(Andersen1977,222ff;Christophersen1991).Sofar,however,nonehavebeen
directly located and investigated archaeologically.Consequently,thedate,structure,function
andcharacteroftheseplacesareinthedark.
In the tenth, eleventh, and twelfth centuries
2TheBackground
a number of towns emerged. Ordericus Vitalis
mentions six civitates on the Norwegian coast
whenwritingaboutNorwayabout1135.These
have been identified as Trondheim (Nidaros),
Oslo, Tønsberg, Konghelle, Sarpsborg (Borg),
andBergen.Sincetheseweretheonlyonesmentioned by Ordericus they may have been the
largestormostimportant(HelleandNedkvitne
1977, 206). In addition eight other places referredtoinurbantermsarerelatedtotheperiod
before1200indocumentaryrecords(Helleand
Nedkvitne1977,206ff),Figure2presentsthese
14 places. Starting with Ordericus’ towns - excludingBergen-excavationsinTrondheimhave
dated the first non-agrarian phase tentatively
intothefirsthalfoftenthcentury;thearea,however,wasnotpermanentlysettleduntiltheend
of the century (Christophersen and Nordeide
1994,266,274).Thefirst‘townphase’inOslois
datedthrougharchaeologytoabout1000(Schia
1991, 116ff; Schia 1992, 46). Tønsberg may,
basedonarchaeologicalsources,datebacktothe
first quarter of the tenth century (Brendalsmo
1994,113).WrittensourcesmentionKonghelle
from the reign of Olav Tryggvason (994/995999/1000) but not as a town until the days of
Olav Kyrre in the late eleventh century (Helle
andNedkvitne1977,214).Accordingtowritten
sources Borg (Sarpsborg) was founded by Olav
Haraldsson about 1016 (Helle and Nedkvitne
1977,212),thistownhasnotbeeninvestigated
archaeologically(Schia1992,32).TheurbanlocalitiesnotmentionedbyOrdericusare:Vågan,
Steinkjer,Veøy,Borgund,
Kaupanger,Stavanger,Skien,andHamar.The
datingoftheoriginofVåganinLofotenasacentralplacehasbeenconsideredobscure(Bertelsen
andUrbanczyk1988,98).Howeveraccordingto
severalsagas,ØysteinMagnussonbuiltachurch
andlodgesforfishermeninVågan,andonthis
basis Vågan has been considered as a church
centreandacentreforstockfishtradefromthe
beginningofthetwelfthcentury.Accordingto
latersagatradition,Steinkjerwasfoundedinthe
earlyeleventhcentury,butthereisnomentionof
theplacelateronandthereisnoarchaeological
record of a central place here (Helle and Nedkvitne1977,214ff).Veøymayhavefunctioned
asacentrefromthemiddleofthetwelfthcentury, according to the archaeological material
21
Figure2.Fourteenmedievaltownsrelatedtotheperiodbefore1200inthedocumentaryrecords.(ModifiedfromHelle1992,8)
22
andothersources(Solli1996,206).InBorgund,
settlement has been dated to the early eleventh
century through archaeological investigations
(Herteig 1957, 462), however, the character of
thissettlementisnotclear.KaupangerinSogn
wasfirstmentionedinconnectionwitheventsin
1183-84, an actual settlement area has yet not
been identified. Based on toponymical, written
and archaeological evidence, Øye suggests that
Kaupanger may have functioned as a central
placewithanurbancharacterasearlyasthelast
halfoftheeleventhcentury(Øye1989,149ff).
Stavanger became an episcopal seat shortly after1120,buttheplacewasnotmentionedasa
town until the last half of the twelfth century
(Helle 1992, 15). Through archaeological investigationstheoldestnon-ruralphaseinSkien
has been dated to the second half of the tenth
century.However,theareawasnotpermanently
settleduntiltheeleventhcentury(Myrvoll1992,
249ff).AtHamar,HaraldSigurdssonHardråde
(Harald Hardruler) (1046-1066) struck coins,
andtheplaceisknownasatownin1154,when
a letter refers to the newly established episcopalseatthere(HelleandNedkvitne1977,216).
FromthisaccountitshouldbeclearthatBergen
didnotemergeinanurbanvacuum,townswere
aknownphenomenoninNorwayintheperiod
underinvestigation,andthetenthandeleventh
centuries seem to have been a very productive
period of urbanisation. The same trend is seen
incontemporaryDenmarkandSweden(Andrén
1989).
EarlyBergen,stateofresearch
A scholarly interest in early Bergen goes far
back, to the socalled Bergen humanists in the
sixteenth century (Edvardsen 1951 (163095); Edvardsen 1952 (1630-95); Sørlie 1957
(1559/60))TheoriginofBergenhasbeenacentralquestion;wasBergenanorganicallygrown
townorafoundedtown,andhowfarbackcan
the town be dated? Studies on the oldest Bergen are numerous, and only the most relevant
tomystudywillbepresentedhere.Ihavechosen to emphasise the character of the sources
studiedandthemethodologicalapproachesand
explanationsprovidedinordertosingleoutand
compareimportantelementsofrelevancetomy
own analysis and trying to work out new ap2TheBackground
proachesandnewinputstothetown’searliest
phases.
According to the Kings sagas: Morkinskinna, Fagrskinna and Heimskringla, written in
the 1220-30s, a town was founded in Bergen
duringthereignofOlavKyrre.Morkinskinna
and Fagrskinna tell that the town was founded and Heimskringla tells explicitly that Olav
Kyrrefoundedthetown(setjakaupstad)(Helle
1982,86-87).TheOldNorseverbsetjaisused
in differing ways in the written sources. It is
used in the sense that something is founded
juridically:anestablishedsettlementwasgiven
jurisdictionorwasdemarcatedtopographically.
Buttheverbisalsousedwhensomethingwas
actually founded on a virgin site like a building,achurchoratown(Bjørgo1971b,69-73;
Helle1982,87-90).TheKingssagasareinother words somewhat ambiguous when describingthecharacterofthefoundationofthetown
andthequestionofwhatactuallyhappenedin
Olav’sdayshasbeenacentralthemethroughoutthehistoryofresearch.
TheoriginofBergen
As early as in the seventeenth and eighteenth
centuries different historians and topographers
arguedthattheareabyVågenhadbeensettled
before Olav Kyrre; the settlement had its roots
inrichherringfisheriesnearVågen.OlavKyrre
foundedthetownbygivingprivileges.Thisview
wasbasedonaseriesofinterrelatedsources:local tradition, the place name Bjorgvin, and sagas, the convenient location for trade was also
stressed (Meyer 1904 (1764), 16-17; Edvardsen
1951 (1630-95), 32-34). Yngvar Nielsen added
closeness of the sheltered Bay of Vågen to the
royalestateatAlrekstadasanadditionalfactor
that may have triggered the growth of a small
settlement by Vågen. According to Nielsen the
townwasfoundedwhengivenprivilegesbyOlav
Kyrre. The area around Vågen belonged to the
royalestateAlrekstadandthekingdonatedland
tothetown,laidoutplots,wharvesandstreets
andalsopointedoutBergenastheepiscopalseat
forwesternNorway,theareajurisdictionallydescribedasGulatingslagen(Nielsen1877,1-7).
Fromthenineteenthcentury,researchonthe
early history of Bergen has generally followed
two main lines of thinking, known as ‘the or23
ganic town tradition’ and the ‘founded town
tradition’.PAMunch,ahistorianclaimedthat
the major towns of Norway, including Bergen,
were‘organictowns’grownoutofearlymarket-,
trading-orfishingplacesandonlylaterregulated
andgiventownprivilegesbytheking.Thetowns
were thus neither founded nor planned by the
kingfromthebeginning(Munch1849,27-30).
This hypothesis was rooted in local traditions
(HelleandNedkvitne1977,207)andinoneof
theSagasoftheIcelanderswrittenfromthemiddle of the thirteenth century and later (Storm
1899; KLNM VII 496-513). In 1899, the historian Gustav Storm rejected the Sagas of the
Icelandersashistoricalevidencetourbanhistory
astheycontaintoomanyanachronisms.InoppositiontoMunch’s‘organictownhypothesis’he
claimedthattownsofNorway,werefoundedon
virginsitesasmarkettowns,laidoutandstructuredbytheking.Stormbasedthishypothesison
theKingssagaSnorreSturlason’sHeimskringla
andananticipateduniformbuildingtopography
in Trondheim, Oslo and Bergen. According to
Storm, Bergen’s original name, Bjorgvin, shows
that Bergen originated from a farm, the farm
wasroyalpropertythatthekingchosetodevelop
into a town (Storm 1899, 433-36). Storm’s hypothesisofhowthetownswerefoundedandorganisedimpliedastrongcentralpower,theking.
Hishypothesis,however,didnotreceivesupport
untilthe1950s.
Inthebeginningofthetwentiethcenturythe
localhistorian,painteraswellasdirectorofthe
HanseaticMuseum(HanseatiskMuseum)ChristianKoren-Wibergfoundsupportforthe‘organic tradition’ through a new category of source
material:seculararchaeologicalmaterial,which
hedocumentedinthemiddletownarea(Figure
3). He suggested that the town had grown out
ofanumberoffarmsteadslocatedalonganold
roadatthefootofFløyfjellet.Theoldbuilding
pattern structured the layout of the new settlementthatwasgivenlawsandanadministration
byOlavKyrre.Thekingalsobuiltchurches,and
inthissensefoundedthetown.Thesettlement
prior to Olav Kyrre emerged because the topographicallocationofBergenattractedmerchants
and fishermen (Koren-Wiberg 1908a, 149; Koren-Wiberg1921,14-22,45-51).
ThehistorianBerntLorentzen,alsodirectorof
24
theHanseaticMuseum,succeedingKoren-Wiberg,wasthefirstwhosupportedStorm’sfounded
townhypothesisforBergen.Inhisdoctoralthesisof1952hismainaimwastoreconstructthe
medievalsecularbuildingtopographybasedon
relevantwrittensourcesincludinglatemedieval
documentary evidence, used retrospectively to
illuminatetheearliestphases(Lorentzen1952).
This approach in many respects represented a
newapproachtothesourcematerialandopened
formoredetailedstudiesofthelocaltopography
intheMiddleAges.Lorentzenfoundsupportfor
the old hypothesis that the area around Vågen
wasoriginallyroyalpropertyandarguedthatthe
riseofBergenmusthavehaditsbackgroundina
royalinitiative.Lorentzen,however,foundlittle
evidencetosupportthetheorythatOlavKyrre
plannedthebuildingtopographicallayoutofthe
town.Thelayout,asreconstructedbyLorentzen,
was rather a result of natural gradual growth
spreadingoutfromthenortherntownarea(Figure 4). He did not elaborate on why the king
foundedBergen.ThecentrallocationofBergen
inwesternNorwaywas,however,consideredfavourable as a religious centre (Lorentzen 1952,
38-42,75-77).
The historian Johan Schreiner, acting as opponentatLorentzen’sthesis,lentfullsupportto
Lorentzen’s thesis. Based on general considerationsoftheneedsofthelateViking-earlymedieval elite, he added that the town must have
been founded as an institution on a national
level, a commercial centre for the exchange of
local, national and international goods to serve
theinterestsofboththesecularandecclesiastic
land owning aristocracy (Schreiner 1953, 43637).Inthe1950scentralresearchersthusagreed
thatBergenwasfoundedbyKingOlavKyrreon
asitenotpreviouslyoccupiedbyanurbancommunity.However,thekingdidnotplanthetown
physically.
Intheyearsfrom1955to1969(andwithseveralcampaignsintheyearsuntil1979)thefirst
modern excavation of urban secular medieval
remains was carried out in Bergen at the Bryggensiteinthenortherntownarea.Theearliest
building topographical layout along the Vågen
waterfront was exposed, with the oldest structures tentatively dated to the 1130-50s and accordinglyyoungerthanOlavKyrre’sreign.The
Figure3.Koren-Wiberg’sreconstructionofthesettlementinBergenbeforeOlavKyrre.(Koren-Wiberg,1921,48PlanIII)
preliminaryresultsfromtheexcavation(Herteig
1969) gave way to a renewed debate about the
origins of Bergen. The excavation supervisor
and archaeologist Asbjørn E Herteig argued
that Bergen was founded, organised and given
itsphysicallayoutbyOlavKyrre.Thisdidnot,
however,excludethepossibilityofasmallersettlementintheBergenareapriortoOlavKyrre,
whetherpermanentorseasonalorconnectedto
alanding-placefortheroyalestateatAlrekstad.
Hehadnodirectevidencetosupportthistheory.
Still,hemadethecasethatindirectlythebuildingtopographicallayoutwithdoubletenements
asbasicunitsfromthestart,indicatedaregular
townplanorganisedonahighlevelbytheking.1
Bergenwasthusfoundedinordertocoordinate
andcontroltradealongthecoast.Herteig’sargu2TheBackground
mentswerebasedontheregulatedlayoutofthe
oldestrecordedstructuresattheBryggensite(althoughdatedtothemiddleofthetwelfthcentury),andtheywerebasedonLorentzen’stheories
thattheareaaroundVågenwasoriginallyroyal
property, and that the oldest tenements were
located towards Holmen (west of the Bryggen
site)(Herteig1969,139-46,210;Herteig1970;
Herteig1985,11).Withthisinterpretationofthe
sourcesHerteiglandedona‘mild’versionofthe
foundedtowntradition;thetownwasfounded
physically,butnotnecessarilyonavirginsite.
ThehistorianNarveBjørgo,whoparticipated
inthediscussion,criticisedbothLorentzen’sand
Herteig’s arguments for being loosely founded
and weak from a critical point of view regardingthesourcematerial.Inthiswayhereduced
25
Figure4.Lorentzen’sreconstructionofBergenc1200.(Lorentzen1952,75)
Lorentzen’stheoryofthelocalisationoftheoldest settlement to a mere chain of indications.
Furthermore, since the oldest structures at the
Bryggen site were not older than c 1130 they
could not directly elucidate Olav Kyrre’s Bergen.Andthedoubletenementbuildingpattern
couldbeexplainedinotherways,suchasanarchitectonicanswertothespecialtopographyand
orientation towards the harbour and therefore
couldnotbeusedasaconclusiveargumentfora
royalinitiative.Bjørgolookedfurtherintophilologicalexpressionsandinterpretationsofthesaga
textsandshowedthattheOldNorwegianterm
‘setja’, could mean that Olav Kyrre regulated
an already existing settlement administratively.
(Bjørgo1971b;Bjørgo1971c,69ff,106ff,126).
In the rather heated discussion that followed,
Lorentzen, Herteig and Bjørgo contributed
throughanumberofchroniclesinalocalnewspaperBergensTidende(SeeegBjørgo1971a;Bjørgo
1971c;Herteig1971a;Herteig1971b;Lorentzen
1971a;Lorentzen1971b).Noconcludingargumentscouldbeadvancedsothequestionabout
theoriginofBergenwasnotsettled.
In1982,inthefirstvolumeofBergen’stown
history,Hellegaveathoroughdiscussionofthe
26
available sources for the oldest Bergen and an
evaluation of earlier arguments. Helle’s discussion of the period prior to 1130 was based on
thewrittensources,andtopographicaldata,and
he also considered the general conditions for
the location of a town in the Bergen area. He
agreedwithBjørgothatinascientificsensethe
earlyhistoryofBergenwasstillinthedarkand
onlyfurtherarchaeologicalinvestigationscould
elucidate the theme. Helle gave a thorough accountanddiscussionofsourcesthatmayelucidatetheoriginalownershipoflandintheBergen
area,andfoundevidencetosupportthetheory
thattheareaaroundVågenwasmostlikelyroyal
propertybeforeatownemergedhere.Hefound
itlikelythatthetownwasprecededbyanearlier
undefinedsettlementpriortoOlavKyrre;asthe
royalestateatAlrekstadprobablyhaditsharbour
orlanding-placeforboatsbytheVågenbay,this
could involve a small year-round settlement by
Vågen.Hereferredtocontemporarysourcesto
supportthisview.Accordingtoecclesiasticrules,
bishopsshouldhavetheirseatinatown.However, Olav Kyrre did not place the first bishop
inwesternNorwayinBergen,butratheratSelja
furtherupthecoast.Furthermore,whenwriting
thehistoryoftheHamburg-Bremenarchiepiscopalinthe1070s,AdamofBremendidnotmentionBergen,thismayimplythattherewasnot
amajorsettlementbyVågenatthistime.Thus,
accordingtoHelle,BergenwasprobablyfoundedbyOlavKyrreinthesensethatasmallharbourcentrewashelpedalongthewaytoachieve
thestatusofatownintheeyesofthecontemporaries. Olav Kyrre’s initiative to build churches
atHolmenandinitiatetheestablishmentofan
episcopalresidenceheremayhavebeensufficient
forhisnametobeassociatedwiththefoundation
ofthetown.However,stillaccordingtoHelle,it
isalsoreasonabletothinkthatthetownwasgiven its own judicial rights, a special administrationandwasfoundedinthisway.Hesuggested
thatOlavKyrremayhavegivenlandtopeople
thatwishedtobuildhere(Helle1982,86-113).
Helle thus concluded with a position between
theorganictowntheoryandthefoundedtown
theory: the town grew up in a place previously
occupied by a small permanent settlement, but
wasfoundedjurisdictionallybyOlav.
In the 1980s, botanical investigations added
yet new source material to the history of early
Bergen.Withabasisin14Cdatedecofacts,botanistsKnutKrzywinski,PeterEmilKalandand
Kari Loe Hjelle found arguments for a denser
non-rural settlement prior to the days of Olav
Kyrre,inthenortherntownareaandinthearea
by the Veisan inlet, (Krzywinski and Kaland
1984;Hjelle1986).Theirresultshavebeenused
asanindicationofanon-ruraldensersettlement
intheBergenareapriortoOlavKyrre(cfHelle
1992;Ersland1994;Herteig2000).
Inhisdoctoralthesisfrom1994,thehistorian
GeirAtleErslanddiscussedtheearlyhistoryof
Bergenusingbothcomparativeandretrospective
methods. Several north European towns were
studiedand‘thetypicalprocessoftownfoundation’identified.Elementsfromthetypicaltown
foundationprocesswerethencomparedwiththe
Bergenmaterial.Throughtheownershipofland
in the later Middle Ages, Ersland showed that
medieval Bergen may have consisted of several
plan-units.BasedontheaxiomthatOlavKyrre
founded Bergen and a plot layout for the Bryggen area, reconstructed among others on sixteenth century sources and the eighteenth centurytenementlayoutofthisarea,Erslandargued
2TheBackground
that the Bryggen area formed a plan-unit with
a plot pattern resembling that of eleventh and
twelfthcenturyplannedtownsinnorthernEurope.ThiswouldindicatethatBergenwasalsoa
plannedtown,wherelandwasinitiallyparcelled
intoplots.Thestudycouldnotdetermineifthe
suggested Bryggen plan-unit was the oldest of
such.
The comparative study did not exclude the
possibility that Bergen also had an organically
grown pre-urban phase, and Ersland suggested
thatKrzywinski,KalandandHjelle’sbotanical
materialindicatessuchasettlementintheVeisan
area,betweenHolmenandthetownarea.The
Vågsbotn area at the mouth of the Vågen Bay
wasalsosuggestedasalocationforapre-urban
settlementcentre.ErslandthusarguedthatBergen was planned and founded physically, but
mayhavehadanorganicorigin(Ersland1994,
30,44,72ff).
Bythemiddleofthe1990stherewasgeneral consensus as far as the origin of Bergen was
concerned. The town was most likely founded
byOlavKyrre,perhapsjuridically(Helle,1982,
1992)perhapsphysicallybyparcellingtheBryggenareaintoplots(Ersland1994)butprobably
on a site previously occupied by a denser nonruralsettlement.
Inthe1980sand1990sarchaeologicalinvestigationswerecarriedoutinvariouspartsofBergen.None,however,produceddatathatcouldbe
datedtotheperiodbeforethe1120s.Researchers
with material culture as a point of outset took
thestateofresearchasrepresentedbyHelle1982
and1992asapointofdeparturewhendealing
withearlyBergenandotherthemesthantheoriginofBergenwereaddressed.
Thephysicallayoutoftheearliesttown
Thephysicallayoutoftheearliesttownwasone
suchtheme(seeHansen1994forahistoryofresearchpriortothe1980s).ThearchaeologistSiri
Myrvollgaveashortpresentationofexcavations
carriedoutinthetownareaunderherdirection
from1980to1987.Onthebasisofmaterialdated
tothesecondquarteroftwelfthcentury,Myrvoll
suggested that the oldest town originally could
haveconsistedoftwosettlementcentres,onein
thesoutherntownareaandoneinthenorthern
town area (Figure 5) (Myrvoll 1987) (cf also
27
28
Figure5.Thedoublenucleussituation.(Myrvoll1993,87)
Dunlop1985a,Plan12).Thepresentationgave
nodiscussionorsuggestionsaboutthehistorical
backgroundforthisdoublenucleussituation.
Helle criticised Myrvoll’s presentation as
weakly founded empirically and argued that
large parts of the available building land in
the town area was settled through most of the
twelfthcentury.AccordingtoHellethelocation
oftheChurchofStNicholasbetweenMyrvoll’s
twonucleishowsthatthisareawasoccupiedby
secularsettlementalreadyatanearlystageinthe
townhistory(Helle1992,26).
The art historian Hans Emil Lidén on the
other hand supported Myrvoll’s double nuclei
hypothesis. Based on a discussion of the initial
function of the twelfth century churches in
Bergen he argued that the Church of St Mary
(oractuallyapossiblepredecessortothestanding church) in the northern town area and the
Church of St Cross in the southern town area
mayreflecttwosettlementnuclei(Lidén1993).
My own master’s thesis (Hansen 1994) may
also be seen as a contribution to the theme of
building topography, as it represented the first
attempttodiscussthecompletespanofarchaeological, botanical, written and topographical
material from early Bergen under one theme.
Throughthemethodsofmapproduction,divisionofthesourcesintocategoriesandacritical
surveyofthecontemporaryavailablesources,a
‘time-picture’ of Bergen around 1190 was produced.Thiswasasfarbackintimeasonecould
gointhearchaeologicalsourceswithanacceptablelevelofsecurity;archaeologicalmaterialolder
thanthelatetwelfthcenturywasgenerallydated
onweakpremises.Structuresandthelocationof
culture-layerswereusedassourcesforthebuilding topography, while artefacts were only used
as a means of dating. Through the new methodological approaches I showed that the town
areawasnotasdenselybuiltinthelatetwelfth
centuryasassumedinearlierresearch(Hansen
1994b,134).
SigurdGrieg’s1933publicationonurbanmedievalfindsfromBergenandOslowasforalong
timethemainreferenceforurbanmedievalfinds
in Norway, supplemented by data from a few
published excavations from the 1960s in DenmarkandSweden.
In the late 1970s and 1980s a number of
projects were initiated in Norway based on the
vastamountsofarchaeologicalmaterialfromurbanexcavationsundertakenfrom1955andafter.
Theprojectshaddifferentresearchstrategiesand
aims.Inthe‘Gamlebyenproject’artefactsfromexcavationsinOslowerestudiedingroupsconsideringtheneedforbasictypologicalandchronologicalstudiesofthematerialandaccordingtothe
specialinterestoftheindividualresearchers(Molaug1991,93).InTrondheim,thepublicationof
the large body of material from the Folkebibliotekstomten site (the Library site) started with
theresearchandpublicationproject‘Trondheims
FortidiBygrunnen’in1985.Theresearchstrategy
wastostudyselectedgroupsofartefactsthatwere
considered especially important as sources for
themaingoaloftheproject(Christophersenand
Nordeide1994,25).Theartefactstudieswereintegratedinthesynthesisingpublicationfromthe
FolkebibliotekstomtensiteinTrondheim(ChristophersenandNordeide1994).
From the 1980s several scholarly studies of
artefact material from Bergen were also undertaken. Those of relevance here are: boat finds
(Christensen 1985), textile equipment (Øye
1988), footwear (Larsen 1992), different kinds
ofpottery(Lüdtke1989;BlackmoreandVince
1994), runic inscriptions (Dyvik 1988; Seim
1988b; Seim 1988a), tools of trade (Grandell
1988),coins(Skaare1984),andselectedosteologicalmaterials(Hufthammer1987;Hufthammer1994)allfromtheBryggensite.Theseartefactgroupswerestudiedinconnectionwiththe
BryggenProjectandpublishedfromthemiddle
ofthe1980s.ThestudiesofpotteryhadthespecificaimtodatethematerialfromtheBryggen
site.Fromthelate1990sartefactgroupsfromthe
Artefactstudies
wholetownareainBergenhavebeenstudiedin
At the Bryggen excavations (1955-79) all arte- several master’s theses. Those of relevance here
facts were collected systematically for the first are: fishing tackle (Olsen 1998) and weapons
timeinNorway.Artefactsfromwerefromthen (Nøttveit2000).
on collected as a routine for all excavations in
Basic identification and classification of the
Bergen.
artefactshavebeenatimeconsumingaimofall
2TheBackground
29
these studies. In addition the finds have been
discussedthematicallythroughspaceandtime.
However,inmoststudiesthematerialhasbeen
consideredwithinwidetimerangesandmostoften all finds from the period before 1170 have
beenstudiedinoneunit.Theartefactmaterial
fromBergenhasthereforenotbeendiscussedin
connection with the earliest history of Bergen.
Onereasonforthisisthatnoartefactcontexts
havesofarbeendatedtobeforethefirstquarter
ofthetwelfthcentury,andcontextsfrombefore
c1170havebeenconsideredasdatedonarather
uncertainbasis.
3THEPRESENTSTUDY,
THEORETICALAPPROACHES
ANDDEMARCATIONS
As shown in the preceding chapter the present
study builds on a history of research with long
traditionswhenevaluatingthewrittenevidence,
thegeneralconditionsforthelocationofatown
in the Bergen area, and the natural scientific
sources.Alsolongtraditionsofresearchintothe
archaeological sources including the churches
mayprovideaplatforminmystudy.Sofarresearch has not been able to throw much light
upontheearliesthistoryofBergen.
InmystudyoftheearliesthistoryofBergen
different theoretical and methodological approachestothesourceswillbeapplied.
The earliest urbanisation of Bergen can be
seenaspartofamoregeneraldiscussiononthe
riseoftownsinearlymedievalScandinavia.The
theoryofatowncommunity,‘organicallygrown’
outofthemoreorlessspontaneousinitiativesof
merchants,artisansandotherindividualsonone
hand,andthetheoryofastrongcentralpower-
thekingorthechurch-asafounderandplanner
of towns on the other hand, may be argued to
havetheirrootsintwobasicallydifferenttheoriesofaction;twofundamentallydivergingunderstandings of the ways of social change. The
classicaldiscussionbetweenthemethodological
individualist approach to social change on the
one hand and the methodological collectivist
approach on the other is thus also reflected in
urbanhistory.Thesepositionsareoftenreferred
30
toassubjectivism/anactor-basedperspectiveand
objectivism/a structure-based perspective. Max
Weberwasanearlyadvocateofthefirst,Emile
Durkheim supported the latter position (Gilje
andGrimen1992,202;Bugge1999,1;Hansen
2000).
According to the organic town theory, trade
carriedoutbyindependentmerchants,wasconsideredthemostimportantfactorforthedevelopmentoftowns.InNorwaythistradition,ina
broadview,goesbackintothenineteenthcenturywithMunch(Munch1849).Inthetwentieth
centurytheideasbecameestablishedinSweden
andinDenmarkthroughtheworkofASchück
andHMatthiessen(Schück1926;Matthiessen
1927).HPirenne’swork(1925and1939)onthe
riseoftownsinEuropealsoinspiredthislineof
thinking(Nielsen1997,181,183).Bettermeans
of transport, an increase in international and
nationaltradeandasurplusofgoodsserveasa
backdropfortheemergenceoftownswithinthis
tradition.Themainactorsweremerchants-individualswho:‘sawtheadvantagesofpermanent
trading places on appropriate localities, where
they could meet their customers’ (my translation)(Olsen1975,248).Later,asthetowndeveloped, professional and independent artisans
wereattractedtothesettlementthatprovideda
suitableplacefortheproductionandsaleofcrafts
totownspeopleandawiderlocalmarket(Olsen
1975,250ff).Theroleofthekingwastoprovide
protectionandsecuremarketpeaceforthetowns
(see eg Olsen 1975; Skovgaard-Petersen 1977).
According to this line of thinking, which was
clearlyinfluencedbyamethodologicalindividualistapproachtoaction,theriseoftownswasexplainedastheproductofenterprisingindividuals
thatfollowedtheirowninterestsandseemingly
didnothavetotakeintoconsiderationthesocietytheywerepartof(Hansen2000,5).Theking
played a secondary role, merchants and craftsmenaprimaryrolefortheriseoftownswithin
thistradition(Christophersen1982,104).
Withintheothermaintraditionofthought,
scholars have emphasised the institution of the
kingand/orthechurchasimportantfactorsfor
theriseoftowns.Fromabroadviewthisdirection also has a long history. In Norway it goes
back into the end of the nineteenth century
with Storm’s founded town theory as well as E
Bull’s (1918) theory that the establishment of
church centres was the main determinant for
the development of towns (Storm 1899; Helle
and Nedkvitne 1977, 207-208). The advocates
of this tradition claim that the medieval town
was too complex to have evolved by itself, and
consequentlystronginstitutionslikethekingor
thechurchmusthaveplayedadecisiverole.In
early studies, towns were perceived as founded
bythekinginanattempttocentralisetradeand
craftstoplaceswithacentrallocationinrelation
totransportandcommunication(Herteig1969).
MedievalarchaeologistshaveinmorerecentresearchconnectedtheriseoftownstoabasicreorganisationofsocietyintheearlyMiddleAges
(Andersson 1977; Andersson 1990, 84). Trade
andcraftshavebeengivenlessweightandtowns
are perceived as regional centres of administrationandpower,foundedbythecentralkingdom
as an instrument of physical control over a region and as centres of administration and concentration of taxes (Christophersen 1982, 118;
Andrén1985,119-120).Tocitethearchaeologist
AxelChristophersen,thetownisexplainedas‘a
functionalelementinasocio-economicsystem...
wheretheemergenceofamonarchyfoundedon
feudalrelationsisstressedasaprimarydynamic
factorintheoldestphaseoftheprocessofurbanisation’ (my translation) (Christophersen 1982,
120). According to Christophersen, the church
didnotcountasanindependentdynamicfactor
intheearlyMiddleAgesbutgaveideologicaland
spiritualsupporttotheemergingcentralmonarchy.Tradeisprimarilyseenasaresponsetothe
needsarisingfromthetowns’functionaspoliticalandadministrativecentres.CraftsintheearlyScandinaviantownareprimarilycharacterised
by the production of luxury items by artisans
who were part of the king’s household (Christophersen1982;Christophersen1989,130,144;
Christophersen 1994). This approach is clearly
based on a methodological collectivist view of
social change; towns are founded by the royal
institutionasaresponsetotheinherentneedsof
thecentralmonarchy;thetownsserveascentres
ofcontrolandadministration,thechurch,trade
andcraftarenotindependentfactorsforthefirst
riseoftowns(Hansen2000).
From what may be characterised as an intermediate position, Helle and Nedkvitne have
suggestedthattheriseoftownsinNorwaybetween1000and1135haditsmainbasisinthe
developmentofacentralmonarchyandtheestablishmentofaChristianchurchorganisation.
Based on political, military and administrative
considerations these two institutions initiated
and developed centres. A denser settlement developed as a direct or indirect result of activitiesconnectedtothekingandthechurch.The
centres then attracted trade and crafts, and especially trade became a deciding factor for the
developmentofthetowns(HelleandNedkvitne
1977,225).Thustownswereinitiatedaspolitical, military and administrative centres for the
kingandthechurch.Individualmerchantsand
craftsmen are, however, also seen as important
determinants that operate independently of the
kingandthechurch.Anincreaseinthepopulationandinagrarianproductionisseenasageneralbackgroundfortheriseoftowns.
Theclassical‘organictowntradition’hasbeen
criticisedforreducingtheurbanoriginintobeinganelementinthehistoryoftradewherethe
town is regarded mainly as a medium for the
growthoftradeandcapitalism(Christophersen
1989, 113). The ‘founded town’ tradition may
likewisebecriticisedforreducingtheearlytowns
intomereinstrumentsfortheemergingcentral
monarchy. Being too narrow in scope, the two
traditions both offer a somewhat one-dimensionalview.Withinthe‘organictowntradition’
individualsact,apparentlywithoutconsideration
ofthewiderframeworkofsocietytheyareapart
of. Within the founded town tradition, towns
areseenasaproductofthesystemofsocietyand
the individuals as marionettes of the ‘system’
(Hansen2000,6).
In the present study, social change is understoodasaproductoftheinterplaybetweenpeople from different levels of the social hierarchy
andtheirwiderhistoricalcontext.Notaseither
theproductofthefreewillofindividualsorasthe
productofinherent‘laws’ofthesystemofsociety.ThesociologistAnthonyGiddens’theoryof
structurationinspiresthislineofthinking(Giddens1979;Giddens1984;Giddens1995(1981)).
Accordingtothetheoryofstructuration,human
actioncreatesandinfluencesstructuresthatconstitutesocialsystemsandviceversa:‘Thestructured properties of social systems are both the
3Thepresentstudy,theoreticalapproachesanddemarcations
31
medium and the outcome of the practices that
constitute those systems’ (Giddens 1979, 69).
Structuresmayberulesandresources(material
orauthoritative)drawnuponwhenacting(Giddens1984,xxxi).Actionisnotalwaysrestricted
byrules,becausepeopleareknowledgeable,creative and conscious actors and find solutions to
problemsalongtheway,thismayresultinnew
rules.Furthermorethoughactorsareknowledgeable, action may not always have the intended
outcome,becausetheactor’sscopeofcontrolis
usuallylimitedtotheimmediatecontextsofactionandinteraction(Giddens1984,8-10;Cassel
1993,10-11).
Somewhat simplified and schematically one
may say that in a stratified society, like that of
theearlyNorwegiancentralkingdom,initiatives
were taken from the ‘top and down’ and from
‘the bottom and up’. In this model, top-down
initiativesweretakenbyresourcefulactorswitha
centralpositioninsociety,theseindividualshad
anopiniononhowsocietywastobeformedand
did perhaps have the resources to realise their
ideas.Thekingorhisrepresentativesbelongedto
thisgroupofactors,whilethechurchwashardly
anindependentfactorpriortothemiddleofthe
twelfth century (cf Andersen 1977, 311). Bottom-upinitiativesweretakenbylessresourceful
actorspositionedatlowerlevelsofthesocialhierarchy.ThetownspeopleandvisitorsinBergen
mayberegardedasrepresentativesforsuch.Both
typesofactorscouldcarryoutmajorinitiatives
anddailyactivitiesininterplaybetweenoneanotherandinawiderhistoricalcontext.
Actiontakesplaceinaphysicallocation.The
actorsinfluenceorcreatethephysicalsetting:the
landscape,thebuilding,butthephysicalsetting
also structures activities (Giddens 1984, 118;
Cassell1993,19).Accordingly,thearchaeologicalmaterialmaybeseennotonlyasareflection
of events and decisions that took place during
Bergen’searlyperiod,thephysicalmaterial,such
asthebuildingtopographyortheinfrastructure
ofthetown,alsorepresentstheassetsandconstraintsthattheactorsoftheearlytowncouldor
hadtoactinrelationto.Thesourcesthusreflect
boththeintendedandtheactual(andsometimes
random)outcomeofmajorinitiativesanddaily
activitiescarriedoutbythesepeopleininterplay
withintheirwiderhistoricalcontext.
32
Today,itiscommonlyacceptedthatmaterial
culturemayhavemeaningsthatarenotexhausted by their physical attributes (eg Tilley 1989,
185).Materialculturemayconveyitsmeanings
in a wide diversity of ways, these ‘meanings’
being culture-specific (Ucko 1989, XIV). The
sourcesofearlyBergencomprisealargebodyof
data spanning from humble production waste
toimpressiveecclesiasticandsecularmonumentalbuildings,nottomentionthatthetownasa
wholeconstitutesaphysicalmaterialmanifestation.Theconceptualmeaningcontainedwithin
the material remains probably spans the whole
scalebetweenthefunctionalandsymbolic.
In order to elucidate why Bergen was initiatedIwilldiscusstheintendedandactualfunctionsofthetown.Thesefunctionsmaybemore
complex than is practical and they may have
changedasthehistoricalcontextchanged.The
intendedfunctionsarethosethattheinitiator(s)
ofthetownhadplannedforthetown,whereas
theactualfunctionsarethosethatwerecarried
intolifebytheusersofthetown.Iwillmainly
address the functional aspects involved. I hold
asapremisethatactivityintheverybeginning
reflectstheintendedfunctionofthetown.The
town’s physical layout as well as ‘rules’ for the
use of the town probably constrained and limitedtheusersbutalsopresentedassetsandpossibilities.Intime,newfunctionsmayhavebeen
introducedandthetownmayhavebeguntolive
alifeofitsown.
Remains of buildings and constructions are
likely to reflect conscious strategies and intentionalactions.Structuresindicatingboundaries
andmonumentalandsecularbuildingsmaythus
reflect sets of major initiatives that had a sustainedimpactonthedevelopmentofalivingurbancommunityintheBergenarea.Twosetsof
majorinitiativesdiscernedinthesourcesarethe
establishmentofplotsinthedifferenttownareas
andtheoccupationoftheplotsandotherparts
oftheBergenarea.Thesemajorinitiativeswillbe
addressedintwosub-studiesofthesourcesfrom
horizon2tohorizon5(Chapters9and10).
Accumulatedlayers,wasteandartefactseither
lostorthrownawayreflectallkindsofdailyactivities, some of which are relevant for the understanding of why Bergen emerged. In order
to elucidate the actual function of Bergen it is
central to discuss traces of productive activities
and trade discerned in the sources and discuss
thecharacterofthesettlementsinearlyBergen.
Thesedailyactivitiesarediscussedinthreesubstudiescoveringhorizon2tohorizon5(Chapters11,12and13).
AsIseeit,majorinitiativesthatcanbeassociatedwithtop-downinitiativesreflecttheintended functions of the town, whereas sources that
signifybottom-upinitiativescarriedoutbythe
usersoftheplotsanddailyactivitiesundertaken
in the town reflect the actual functions of the
town.Iwilllinkthetwogroupsofactorstothe
majorinitiativesanddailyactivitiesandelucidate
howthetownemergedintheinterplaybetween
theactorsandtheirwiderhistoricalcontext.By
thisapproachmyaimistoachieveamorevariedandnuancedunderstandingoftheoriginof
Bergen.
Thepart-studies
Horizon1(c800-c1020/30),abackdrop
Inthefirstpart-studythelocalisationofactivity
in the Bergen area in the oldest horizon (horizon1)isstudied.Thestudyelucidateswhether
ornotBergengreworganicallyoutofanearlier
non-ruralsettlement.Theactivityisgoingtobe
characterised in terms of general land use and
intermsofurbanversusnon-urbansettlement.
This serves as a backdrop for the study of the
youngerhorizons(horizon2tohorizon5).
Plotsandplotsystems
Theexistenceofaregulatedtopographicallayout
withdistinctplotsorinfrastructureiscommonly
seenasasignifierofatownormarketplacedeliberatelyfoundedbyacentralauthority(egSkovgaard-Petersen 1977; Tesch 1992; Ambrosiani
andClarke1995(1991),137).Fencesindicating
boundariesareamongtheoldeststructuresdiscernedintheBergenmaterial.Theinvestigation
of plots and plot systems in the Bergen area is
therefore relevant for understanding how and
bytheinitiativeofwhomBergencameabout.A
plotisheredefinedasapieceoflandparcelled
outfromalargerestate.Theplotmayreflectan
ownershipunitorauser’sunit.Theoriginalplot
may in time have been subdivided into smaller
propertiesorpartsmayhavebeenrentedouton
a more or less permanent basis. King Magnus’
town code of 1276 refers to such a subdivision
(Bl1923,48).
WAvanEs’large-scaleexcavationsinDorestadhavedemonstratedthatthisearlytownhada
regulatedlayout(Hodges1999).Theseinvestigationsgavewaytotherecognitionofplots,boundariesandregulatedinfrastructureinplaceslike
theeighthcenturymarketplaceofRibe(Frandsen,Madsen,andMikkelsen1988;Jensen1992;
Ferveile 1994), and the early medieval town of
Sigtuna(Tesch1990;Tesch1992).RecentstudiesofVikingAgeKaupanginTjøllingandmedievalOslo,TønsbergandTrondheimhavealso
shown that plots were parcelled out at an early
stage in these towns (Schia 1987a; Brendalsmo
1994;ChristophersenandNordeide1994;Skre,
Pilø, and Pedersen 2001). In medieval Lund,
plots were laid out about 1020 in an area that
until then was occupied by perhaps a political
and ecclesiastic central place. These plots were
re-regulatedintosmallerplotsinthelasthalfof
the century or in the beginning of the twelfth
century(Carelli2001,106-109).
Until Herteig’s first publication of the Bryggensitein1969,therewas,asmentionedearlier,
ageneralconsensusthatthelayoutinBergenwas
‘sonatural’(Lorentzen1952),or‘aresultofnatural conditions and local needs’ (Bjørgo 1971),
that it could not have been actually founded
physically.Themedievalbuildingtopographyin
thenorthernandmiddletownareaswas,however, synonymous with the layout of tenements
formed by long rows of buildings that ran perpendiculartotheVågenshoreline.Herteigwas
thefirsttoclaimthatBergenwasfoundedwith
a planned building topographical layout. But
Herteig also referred to the concrete tenement
layoutratherthantotheplotsunderthebuildingsandheldthatthedoubletenementwasthe
centralunitintheplannedtownscape(Herteig
1969).InErsland’sthesisfrom1994theplotunderthebuildingrowswasintroducedasacentral
unitforthefirsttime(Ersland1994).Ersland’s
reconstruction of plots was, as already mentioned,basedonamongothersixteenthcentury
sourcesandnineteenthcenturymapsofBergen,
notonarchaeologicallydocumentedboundaries
fromearlymedievalBergen.Whenstudyingthe
3Thepresentstudy,theoreticalapproachesanddemarcations
33
Gullskogården area at the Bryggen site in her
master’s thesis the archaeologist Hanne M R
Moldungtookacloserlookattheplots.Hereshe
foundthatthroughthemiddleagesthebuilding
pattern within the Gullskogården area varied,
the location of eaves drops that mark the plot
boundarywas,however,stable(Moldung2000,
116-7).Theplotasacentralunitonamoregenerallevelwasnotdiscussed.HerteighasalsosuggestedinarecentarticlethatOlavKyrrelaidout
plotsinBergen(Herteig2000).Thequestionof
plotsizesandsystemswasnotaddressed.
Whenstudyingplotsandplotsystemsinearly
BergenIwillfocusontheplotsunderthebuilding rows, not the tenements and buildings as
such.Themethodologicalapproachisaccounted
forinChapter9.
TowhatextentwastheBergenarea
‘occupied’?
Thisthememayelucidatehowandtowhatextendactorsfromdifferentlevelsofthesocialhierarchyinvestedintheearlytown.Wereplotsin
thetownareaoccupiedfromthebeginningand
wastherepressureonbuildingland?Theextent
towhichtheBergenareawasoccupiedorbuilt
upisanoldquestion.Asmentionedearlier,severalattemptshavebeenmadetoreconstructthe
built-upareaoftheearlytown,butonarather
general level. In my master’s study the lowest
level of inquiry was the excavated site, individualvariationsbetweenplotswerenotdiscussed
(Hansen 1994b). In the present study smaller
entities,suchasthesingleplotormonumentwill
constitutethemainanalyticunit(cfp65ff)and
amorein-depthpictureofthescopeofoccupied
and vacant areas is given. The methodological
approachisaccountedforinChapter10.
Craftsandproduction
In earlier research on town history productive
activities have been seen both as a fundamental economic basis for the rise of towns and as
subordinate to this. Activities carried out on a
household basis and those aimed at in ‘interurban market’ did not ‘add value’ to the town
community and therefore may not have made
upafundamentaleconomicbasisfortheinitial
riseofthetown.Asopposedtothis,productive
activitiesthatservedawidermarketbeyondthe
34
townmayhaveplayedanimportantpartasan
independent economic factor in the early town
(Christophersen 1982, 108). I will look closer
intothesourcematerialtryingtosingleoutproductive activities of different categories accordingtothisperspective.
Combproduction,textileproduction,metalworking,andshoemakinghavebeenconsidered
andanalysedinseveralstudiesofproductiveactivitiesintheearlymedievalScandinaviantowns
(eg Christophersen 1980; Øye 1988; Bergquist
1989;Flodin1989;Ulriksen1996;Rytter1997;
Tørhaug 1998; Carelli 2001; Hagen (1988)
1994). I will focus on two questions. The first
being,thenatureofproductiveactivitiesidentified in the Bergen material; were the products
manufactured and used within the household,
or were they made by professionals? I am goingtouseKarinGjølHagen’sdefinitionofthe
termprofessional,whereprofessionalproduction
isproductionforsale,asopposedtoproduction
for consumption within the household (Hagen
(1988)1994,29-31).Saleinthiscontextisused
in a broad sense, meaning distribution beyond
thehousehold(cfTradebelow).Thetermhouseholdisalsousedinabroadsense,includingfamily members, free and possible un-free servants
(cf KLNM XVII 230ff). Guests or visitors are
peoplethatwerenotmembersofthesettlement
butvisitedforashorterperiodoftime(KLNM
V689ff).Thesecondquestionaddressedis:how
weretheactivitiesorganised-weretheartisans/
producers sedentary residents of the site where
production took place, or were they visitors of
thetownandtravelledbetweenseveralplacesof
production?
Iaimtoidentifyplaceswhereproductiveactivities were carried out and discuss the nature
andorganisationoftheseactivitiesandthuselucidatewhethertheproductiveactivitiesfoundin
earlyBergenmayhaveprovidedafundamental
economicbasisfortheriseofBergen.Theconcrete methodological approach is accounted for
inChapter11.
Trade
WhenapproachingthethemeoftradeIwillfocusononequestion:Waslong-distancetradean
importantfactorfortheriseofBergen?Tradeis
hereusedasawidetermforbuying,sellingorex-
changeofcommoditiesforprofit.Theprofitmay
bematerialorofasocialcharacter(Carelli2001,
178).Thesocialmechanismsinvolvedwhenexchanginggoods(cfegChristophersen1989)will
notbedealtwithinmystudy.
InthelateVikingAge,long-distancetradein
heavybulkcommoditieswasintroduced(Jensen
1990;Näsman1990;CarelliandKresten1997).
In the earlier periods, long-distance trade had
mainly been directed towards the acquisition
ofprestigegoodsandscarcitiesofsmallvolume
and weight (Hodges (1982) 1989, 53). Trade
was probably mainly carried out within the
social sphere, by members of the leading class
or representatives of such. With the new bulk
goods trade a new social group of specialised
salesmen evolved. Whether, in the early part
oftheperiodstudiedhere,theytradedfortheir
ownprofitorasrepresentativesforothersisnot
clear(Näsman1990,112ff).Thestatusofthe
salesmen, and the social context within which
tradewasconductedhasbeenacentralquestion
in modern urban archaeology, but will not be
exploredhere.
HonesfromEidsborgineasternNorwayand
dark grey schist hones from western Norway,
quernstonesfromHyllestadinwesternNorway
andsoapstonevesselsaresomeofthenon-perishable products that were exported from Norway
fromthelateVikingAgeandonwards(Mitchell,
Askvik,andResi1984;Myrvoll1986;Christophersen1989;Jensen1990;Carelli2001;Baug
2002).Anincreaseintheproductionofstockfish
inLofotenandVesteråleninnorthernNorwayis
knownfromthewrittenrecordsfromtheearly
twelfth century (Helle 1982, 116). In 1177, 40
or50boatswereontheirwaysouthwardsfrom
Vågan,probablywithstockfish(egSs16;Helle
1982, 162) if this is correct, it is clear that the
production was extensive and must have been
directedatexports.Theproductionofbogiron
alsoincreasedintheearlyMiddleAgesandwas
probablydirectedatmarketproduction(Narmo
1997, 133, 187ff; Narmo (1991) 1996, 195ff).
Along with the increase in the weight and volumeofgoodstraded,newboattypesweredeveloped.ThelongandslenderVikingAgewarships
weresupplementedbybothseagoingandcoastal
goingfreightcarrierswithhighcargocapacities
(Crumlin-Pedersen1991).
InearlyBergen,trademayhavetakenplacein
variousforms.Rawmaterialsanditems,notproducedinBergen,showthatgoodsfromnearand
farwerebroughtintotownandusedhere,and
someofthesematerialsmayreflecttrade.When
elucidatingtheexistenceoftradeasadailyactivityandtheimportanceoftradefortheusersof
Bergen,onlytracesofregionalandinternational
long-distancetradearegoingtobeinvestigated.
Thustheenteringanddepartureofgoodsinand
outofBergenisconsiderednottheredistribution
ofgoodswithinthetown.Thebroadtermlongdistancetradecoversbothtradethatwaspartof
an international network and trade limited to
Norway.Ishalldiscusstheimportanceoflongdistance trade to the initiators of early Bergen
andthefunctionofthetownasaplaceoftrade.
Themethodologicalapproachisaccountedforin
Chapter12.
Thecharacterofthesettlementontheplots
Werethesettlementsonthetownplotsofearly
Bergenpermanentandwellestablishedorwere
they seasonal or only occasional? Neither of
these questions have been discussed earlier for
Bergen.Theyrefertoclassicalcriteriaforaplace
toqualifyasatownandareaddressedinChapter13wherethemethodologicalapproachisalso
accounted.
4GENERALPRESENTATION
OFTHEARCHAEOLOGICAL,
BOTANICALAND
TOPOGRAPHICALSOURCES
Inthischapterthearchaeological,botanicaland
natural topographical sources are presented on
agenerallevel.Themainexcavationanddating
methodsappliedthroughthehistoryofresearch
areaccountedforandIwilldeterminehowthe
sources can be used in the study, on a general
level.
Until 1899, archaeological investigations in
Norwayweremainlycarriedoutby‘Foreningen
til Norske Fortidsminders Bevaring’ (hereafter:
Fortidsminneforeningen), (The Society for the
Preservation of Norwegian Antiquities) established in 1844. Medieval archaeology was then
4Generalpresentationofthearchaeological,botanicalandtopographicalsources
35
moreorlesssynonymouswiththestudyofmonumental architecture, as the interest for more
seculartypesofobjectshadstillnotbeenevoked.
In1920theCentralOffice(nowDirectorate)for
Monuments and Sites (hereafter: Riksantikvaren),whichhaditsseatinKristiania(nowOslo)
became the authority for protecting medieval
archaeological remains in addition to medieval
standingbuildings(Myhre1985,180).
Until1981Riksantikvarendidnothavealocal
excavationofficeinBergen.Therefore,although
Riksantikvaren has been formally in charge of
all excavations in the medieval parts of Bergen
since1920,theprofessionalresponsibilityforthe
actualinvestigationswasdelegatedthroughthe
yearstovariouspersonswhocametoBergenon
specialassignments,usuallyinconnectionwith
therestorationofmonumentalbuildings.From
1955,however,archaeologistswithresponsibility
forthemedievaltownwereattachedtoBergen
on a more permanent basis. Through Historisk
Museum, The University Museum of Bergen
(now Bergen University Museum) on behalf of
Riksantikvaren, archaeologists connected to the
milieuaroundtheBryggenexcavationstookcare
ofthefieldarchaeologyinmedievalBergenfrom
1955to1979.From1981to1994,Riksantikvaren
establishedapermanentexcavationunitinBergen.Inthefewyearsbetween1979and1981the
centralofficeofRiksantikvareninOslorecruited
excavation supervisors for Bergen projects from
outsidetheBryggenmilieu.Riksantikvaren,Utgravningskontoret for Bergen (The Excavation
Unit of Bergen under the Central Office for
MonumentsandSites)(hereafter:TheExcavation
Unit)wasestablishedin1980andcarriedoutall
investigations in the medieval town of Bergen
until 1994. In 1995 Norsk Institutt for KulturminneforskningNorwegianInstituteforCultural
HeritageResearch(NIKU)wasestablishedand
NIKU has carried out all investigations in medievalBergensincethen.Themethodsusedbetween1979-1994and1994-1998havebasically
been the same and it is natural to consider the
methodological approaches applied in these investigationsasone.
ThemethodsusedinBergenhavedepended
ontheprofessionalsupervisorsoftheinvestigations. In accordance with the development of
medievalarchaeologyasadiscipline,thefollow36
ing groups of investigations provided valuable
sourcesforthequestionsdealtwithhere:
• Investigationsbefore1899
• Investigations from the late 1800s until c
1920
• Investigationscarriedoutbetween1929and
1955
• Investigationscarriedoutbetween1955and
1979
• Investigations carried out from 1980 until
1998
In addition the archaeological, botanical and
topographicalsourcesstemfrom:
• Recentstudiesofthemedievalchurches
• Independentbotanicalinvestigationscarried
outinBergen
• Informationfromprobedrillinginconnectionwithbuildingprojects
Investigationsbefore1899
Between1844and1899anumberofregulararchaeological investigations in Bergen were carried out on the initiative and responsibility of
Fortidsminneforeningen. In 1860 the antiquarian Nicolay Nicolaysen excavated the ruins of
theMunkelivAbbeyatNordnes,andcollected
information from random observations in the
vicinity of the abbey (Nicolaysen 1861). In the
1870s and 1880s Peter Blix, an architect, investigated the foundations of the Church of St
OlavinVågsbunnen,locatedunderthepresent
Cathedral in the southern town area. Blix also
carriedoutinvestigationsatBergenhus,themedieval royal and ecclesiastic centre at Holmen,
andattheNonneseterconventintheNonneseter
area.Inthe1890sanotherarchitectSchakBull
also excavated at Nonneseter (Øye 1997, 443).
B E Bendixen, a local school headmaster, was
also connected to Fortidsminneforeningen. He
wasanactiveobserverandinvestigatorofmedievalphysicalremainsfoundduringconstruction
work. Bendixen documented the foundations
for the Church of St Nicholas in the middle
townareaatthefootofFløyfjellet.Healsoobservedandcollectedrandominformationonthe
churchandchurchyardfromotherconstruction
works in the vicinity (Bendixen 1896). All the
regularinvestigationsfrombefore1899comprise
churches or church buildings at monasteries. I
have not studied the original documentation
fromthesitesasthematerialhasbeenthoroughly discussed and included in more recent studiesofthechurchesofBergeninthebookseries
NorgesKirker(ChurchesofNorway)(Lidénand
Magerøy1980;LidénandMagerøy1983;Lidén
andMagerøy1990)(cfp51).
Investigationsfromthelate1800suntilc1920
Betweenthelate1800sandc1920Koren-Wiberg worked as a town archaeologist. When the
eighteenthcenturytenementsalongtheBryggen
harbourfront,inthemiddletownareaweretorn
down, Koren-Wiberg documented the exposed
medievalremains.Inhissearchforthemedieval
townhallandwinecellar,andforthechurches
ofStPeterandStColumbahealsocarriedout
smallerinvestigationsonsiteswithinthenorthern and middle town areas. Koren-Wiberg was
active collecting information from random observations made by construction workers (Koren-Wiberg 1900; Koren-Wiberg 1908b; Koren-Wiberg 1908a, 150; Koren-Wiberg 1921;
Ersland1988,54-59).Healsocollectedartefacts
during his observations, however, only special
andmoreorlesscompleteobjectswerecollected.
The artefact material is thus not representative
forwhatwasreallytobefoundthere.Inthemuseumcataloguesthelocationofthefindsisonly
specifiedonatenementlevelanditisnotpossible
to localise or date the activities represented by
Koren-Wiberg’sartefactsmoreprecisely.
Koren-Wiberg was the first to show interest
in thesecularbuildingsfrommedievalBergen.
Hewasalsothefirsttousethemethodofdatingbycorrelatingfire-layersfoundunderground
with fires known through written sources, a
methodwhichideallycouldalsodatestructures
unknown and undated in written sources. By
counting the number of fire-layers and dates
forhistoricallyrecordedfiresinBergen,KorenWibergsuggestedthatthenumberoffire-layers
couldbeusedtodatetheactualfire-layersfound
atthesite.Thedatesprovidedshould,however,
becheckedbyartefactsfoundinthelayers(Koren-Wiberg1921,15).
Koren-Wiberg did not take into account the
manyuncertaintiesinvolvedbyusingthismethod;localorotherundocumentedfiresmayhave
struck and the fires known from the written
sourcesmayhaveravagedpartsofthetownonly.
Furthermore,inKoren-Wiberg’sdaysthetypologyandchronologyofartefactswerenotrefined
andcouldnotbeusedtodatethefire-layersindependently.Consequentlyitisdifficulttousehis
datesforstructuresthatarenotknownanddated
throughthewrittensources.Excludingthismaterialwearestillleftwithvaluableobservations
thatthrowlightonthenaturaltopography,the
churchesandchurchyardtopography.
Koren-Wibergpublishedsomeofhisresultsin
surveysofBergenandasillustrationsandmodels
(Koren-Wiberg1908a;Koren-Wiberg1921).But
muchofhismaterialremainedunpublishedand
istodayfoundatByarkivet,(theCityArchives)
inBergenandattheUniversityLibraryinBergen.Hismaterialwillbeusedassourcesbothfor
thechurchtopography(twosites)andassources
ofthenaturaltopography(fivesites).
I have examined artefacts delivered to the
Bergen University Museum in the days of Koren-Wiberginordertoidentifyactivitiesinthe
period investigated here. The artefacts have no
detailedinformationoncontextandhavethereforebeendatedtypologicallybyanalogy.None
oftheartefactscouldbedatedsafelytotheperiodbeforec1170,andcanthusnotbeusedas
sourcesfortheperiodunderinvestigation.
Investigationscarriedoutbetween
1929and1955
Between1929and1957GerhardFischer,anarchitect,andcolleaguescarriedoutseveralmajor
investigationsatthecastleofBergenhusonHolmen,thecastleofSverresborgnorthofHolmen
andtheArchbishop’spalaceatNordnes.Fischer
also investigated and documented minor sites
inthenortherntownareaaroundBradbenken,
Sandbrugaten, and around the Church of St
Mary’s.Inthesoutherntownareaobservations
weremadeatSkostredet,andbytheChurchof
StCross.AtNordnesobservationsweremadeat
‘Vestlandsbanken’ and Østre Holbergsalmenning.DorothFisher,CatoEnger,anengineer,
and a student (later architect) Håkon Christie
assisted Fischer. The documentation material
from regular excavations carried out in medievalBergenbefore1955isfoundatRiksantikvaren’s Archives in Oslo. I have gone through
4Generalpresentationofthearchaeological,botanicalandtopographicalsources
37
the material sorted under Bergen in search of
A hand-written note dated 22 May 1953,
informationthatcouldbeofuseinmystudy.
probablywrittenbyCatoEnger,givesinformationaboutobservationsmadeatthe‘Hotelsite’
Themajorinvestigations
atSandbrugaten,theseobservationsareusedas
Amongthemajorinvestigationsonlythosefrom asourceforearlyBergenandasasourceforthe
the Holmen area are relevant. This material reconstructionofthenaturaltopography.
compriseschurches,secularmonumentalbuildAswiththematerialfromthelate1800sunings and botanical sources. The archaeological tillc1920,Ihavealsostudiedartefactsdelivered
material was published in 1980 (Fischer and totheBergenUniversityMuseumfrom1927to
Fischer1980).Dataconcerningthechurchesat 1955.TheartefactshavebeendatedtypologicalHolmenhasalsobeenincludedinTheChurches ly,andnonecouldbedatedwithanycertaintyto
ofNorway(LidénandMagerøy1980).Ihavenot beforec1170.
gone through the original documentation but
basedmystudyonthepublishedmaterial.The Investigationscarriedoutbetween
botanicalmaterialwaspublishedin1979(Fægri 1955and1979
1979). Since the material is not dated it is not Excavationanddatingmethods
includedhere.
Theexcavationanddatingmethodsappliedbetween 1955 and 1979 were developed through
Theminorsites
theBryggenexcavation.BasedonfieldmethodThe minor sites in the town area and at Nor- ology in prehistoric archaeology and being the
dneswereinvestigatedinconnectionwithcon- first large-scale excavation of secular remains
structionwork.Mostoftenthetrenchlocations from the Middle Ages, the Bryggen excavation
were described according to buildings, street represented methodological pioneering work.
cornersetcandcanonlybetentativelylocated The method applied was stratigraphical excatoday. In general, structures were described in vation,withconstructionsandfire-layersasthe
somedetail,layersonlyinvaguetermsandob- leading strata. Where culture-layers exceeded
servations of the natural subsoil were merely about 15 cm in thickness, they were excavated
commented upon. Structures were often lev- inarbitraryabout15cmthicklayers.Fire-layelled and photo documented. Sometimes ar- ers were documented and given numbers as
tefacts were collected, however, without a de- separate defined layers; other layers were only
scriptionofcontext.Noattemptsweremadeto documented according to their main features.
date structures, which could not be identified Most constructions were numbered according
as buildings known from the written sources. to their excavation unit and artefacts were reThedocumentationofthismaterialisfoundin cordedinrelationtoconstructionsandfire-layers
hand-writtendiariesalongwithafewdrawings (Herteig1985,33).Thefire-layerswerecentral
(FischerUndated).Theseformthebasisformy in the documentation of the relative as well as
evaluationofthematerial.
theabsolutechronology(Herteig1985,22)and
Having identified the approximate location theuseofpermanentandintermediatestanding
of the observations made in the town area and baulks aided the stratigraphical analysis of the
at Nordnes, it seems clear that Fischer and his site(Herteig1985,18).
colleagues did not reach eleventh and twelfth
When the Bryggen excavation was carried
century layers in the trenches at Bradbenken/ out, the material could not be dated through
Sandbrugaten, in Skostredet and at ‘Vestlands- the archaeological finds alone; important artebanken’.Allthesetrencheswerelocatedinareas fact groups such as pottery, shoes and combs,
thatwereopenwaterintheVeisaninletorinthe were still not sufficiently studied and dated at
VågenBayintheeleventhandtwelfthcenturies. thisearlystateofmedievalprofessionalarchaeThetrenchesaroundStMary’s,StCrossandat ology.Consequentlythedatingmethodapplied
ØstreHolbergsalmenninghavebeendifficultto atthesitewasbasedupontheassumptionthat
localiseandtheobservationsaresovaguethatI thestratigraphicallyrecordedfire-layerscouldbe
hadtodisregardthem.
identifiedwithaseriesoffiresknownfromme38
dieval written sources (Herteig 1985, 22). The
methodimpliedthatitwaspossibletofindphysicaltracesofthesefiresandthattheycouldserve
asastratigraphicalandchronologicalframework
forthewholesite.Theframeworkwasgivenan
absolutedate,mainlybycorrelatingandcounting fire-layers and relating them to the historic
firesmentionedinwrittenrecords.Afewrunic
inscriptions and the preliminary results from
dendrochronological (hereafter dendro) dating provided the link between the historically
known fires and the actual archaeological firelayers(Hansen1998).
Countingfire-layersandrelatingthesequence
to historically known fires is as already mentionedanunreliablemethodfordatingpurposes
if other dating methods cannot supplement it.
First,onecannotbesurethatallthefire-layers
found on a site actually represent a fire known
fromwrittensources.Thewrittensourcesmust
alsobethoroughlyanalysed.Thiswasnotdone
until1979(Helle1998).Laterexcavations,such
as Domkirkegaten 6 BRM 245 (Dunlop et al.
1994, 112) and also the Bryggen excavation itselfhaveshownthattherehavebeenseveral‘unknown’/localfiresinBergenandthatthefire-layersdonotalwayscoverthewholesite(Christensson1988).Consequently,itisnotmethodologically advisable to use the number of fire-layers
aloneasameansofabsolutedating.
Herteig was aware of the problems when establishing the ‘fire-layer chronology’ and much
effort has been devoted to dating the series of
fire-layers and thus the periods in the Bryggen
material through the archaeological evidence.
Studiesoftheceramicmaterialandthedendro
sampleswereundertakeninthelate1980s.Some
of these studies were taken into account when
publishing the stratigraphical analysis and the
datingoftheBryggenmaterialin1990and1991
(Herteig1990;Herteig1991).
The other excavations carried out between
1955 and 1979 followed the main principles of
the methods developed during the Bryggen excavation.Thefire-layersequencesatthesmaller
excavation sites were not, however, so clear (eg
Larsen1967a).Andas‘theBryggenmethod’depended upon the stratigraphy of fire-layers, the
lackofdiscernibleandextensivefire-layersseems
tohavemadeananalysisoflayersanddatingof
thematerialdifficult.Theartefactmaterialwas
notstudiedinconnectionwiththereportwork.
Fouroftheexcavationscarriedoutintheperiod between 1955 and 1979 are used here as
sourcesforstudyingearlyBergen.Fourteenotherinvestigationscanbeusedassourcesfordetermining the natural topography. Two investigationsthrowlightuponthechurchtopography.
Investigationscarriedoutfrom1980until1998
Excavationanddatingmethods
The main methodological principle applied between 1980 and 1998 was stratigraphical excavation,asattheBryggenexcavations.However,
allculture-layersandstructureswerenowdocumented and given individual numbers. Layers
wereexcavatedonebyonestratigraphically,unlessspecialcircumstancesdidnotallowit,artefacts were recorded according to the numbered
layer they were located in. The archaeological
stratawereinterpretedasremainsofactivitiesin
a literal sense of the word according to principles developed in collaboration with the Polish
archaeologist Andrzej Golembnik (Golembnik
1995).This method is today referred to as the
‘singlecontextmetod’.Theculture-layersplayed
a decisive role when performing an analysis of
the material and dividing the archaeological
remains into phases. A relative chronology was
established through stratigraphical analysis of
the material. The relative chronology consisted
ofphases,dividedintothreestages:thefoundationstage,theactivitystage,andthedestruction
stage(Myrvoll1991,72).Suchadivisionofthe
layers made it possible to distinguish between
primarily and secondarily deposited layers. In
the terminology of The Excavation Unit/NIKU
primarylayerswere,layersthathavebeenaccumulatedoverashortorlongperiod,butwhich
havenotbeendisturbedormovedonalateroccasion.Secondarylayerswerelayerswhichconsistofmassesmovedfromtheiroriginalplaceof
depositionorwhichhavebeendisturbed(Christensson1988).
Thedatingprinciplewasthatartefactsshould
beanalysedinordertoobtainanabsolutedatefor
thedifferentphases.Potterywasthemostimportantfindsgroup,butotherartefactgroupssuch
as combs and shoes have also been used. Dates
basedondendrochronology,the14Cmethodand
4Generalpresentationofthearchaeological,botanicalandtopographicalsources
39
thethermoluminiscence(TL)methodwereused
when available. Finally, when an approximate
absolute dating frame was established, attempts
weremadetogivethephasesamoreaccuratedate
bycomparingthesedateswiththerecordeddates
oftownfires.Thiswasbecausesomeofthephysicallydocumentedfire-layersmightrepresentthe
firesknownfromthewrittensources(cfChristensson,Dunlop,andGöthberg1982).
IntheinvestigationscarriedoutbyTheExcavation Unit/NIKU the historically documented
fire-layershaveoftenbeenusedwhengivingan
absolutedate.Butthemethodologicalprinciple
hasbeenthatanabsolutedateshouldbeobtained
throughthearchaeologicalmaterialorthrough
naturalscientificmethods.Secondly,amoreaccurate date might be established by comparing
thearchaeologicaldatesforthefire-layers,with
thetownfiresknownfromthewrittensources.
In this way the recorded town fires would becomeasupplementtotheotherdatingmethods
andcountingfire-layerscanbeavoided.
In principle the dating method seems ideal.
Still,thedatesobtainedarenotunproblematic.
Thesizeoftheinvestigationsvaries.Someexcavationswerefairlylarge,withdiversifiedanddatablefinds.However,somewereverysmallopen
areaortrenchexcavations,withtoofewfindsto
establish a firmer dating frame. In some of the
smallerinvestigations‘theidealdatingmethod’,
could not be applied, and the old method of
counting fire-layers has often been used. This
useoffire-layerspresupposesthatweknowthe
numberoftownfiresthathavestruckthegeneral
areaofthesite.
ThenumberofgreattownfiresthatdevastatedBergenbefore1250hasbeenasomewhatcontroversialtheme(cfHansen1994b;Dunlopand
Sigurdsson1995).Atthepresentstateofresearch
thereisnogeneralconsensusonthenumberand
the date of major town fires before 1250. The
written sources give information about several
conflagrations in the years before 1250. There
isgeneralagreementthatoneshouldexpectthe
firesin1170/71,1198and1248tohavestruck
thenorthernandmiddletownareas.Atseveral
sitesafire-layerdatedtoc1225/1230hasbeen
recorded in addition to the fire-layers that may
correspond to the three fires known from the
writtensources(Dunlop1998).Thesites,where
40
the 1225/1230 fire is recognised, more or less
surroundtheBryggensite.Acloseinvestigation
oftheearliestfiresattheBryggensitehasdemonstrated that the first fire documented here is
olderthanformerlyrecognisedandtheformerly
assumed1225/1230fire(Hansen1994b)ismost
likely to be identical with the recorded fire in
1198(Hansen1998).Thisresultalsohasconsequencesforthedatingofthesurroundingsites,
and in some cases a discussion of the absolute
chronology at sites investigated between 1980
and1998isnecessary.
Seventy investigations undertaken between
1980 and 1998 can be used as sources for the
naturaltopography.Twentyofthesealsoprovide
culturehistoricalsourcesforstudyingearlyBergen.
Recentstudiesofthemedievalchurches
The medieval churches have been investigated
archaeologicallyandthroughmasonryandstyle
studies.ThestudiesarepublishedinTheChurchesofNorway,Bergen(LidénandMagerøy1980;
Lidén and Magerøy 1983; Lidén and Magerøy
1990).Iwillusethispublicationofthematerial
asthepointofdeparturehere.
Botanicalinvestigations
Since the 1940s botanical investigations have
beenperformedasindependentresearchprojects
or in connection with archaeological investigations in Bergen. Through the identification of
indicator species (pollen and macrofossils) and
mechanicalprocessesinvolvedinthedeposition
ofthebotanicalmaterial,botanistshaveidentifiedthreegeneraltypesofenvironmentspredating c 1070 in the Bergen material: natural environments, agricultural environments (cultivation/grazing)andpopulatedenvironments.
In 1979, botanical investigations were carriedoutonmarinedepositsattheBryggensite.
Organic layers were recorded and dated to periodsextendingintoprehistory.Thelayerscontained macrofossils, identified as latrine and
other household waste, and pollen that clearly
reflectedanthropogeneousactivity.Theorganic
materialinthemarinelayerswaswellpreserved
andthepresenceofthemarinedinophycea-cyste
Operculodiniumcentrocarpumwaslow.Thiswas
interpretedasaresultofafastaccumulationof
the layers, indicating that the waste had been
dumped in the sea (Krzywinski and Kaland
1984,26).Itwasthenarguedthatintraditional
‘Plaggenboden’agriculture,whichwasconsidered
the most common tradition for manuring the
landonthewestcoastofNorwaybackintoprehistory,allhouseholdandanimalwastewasused
as fertiliser on the fields. Therefore the householdwasteinthemarinesedimentsattheBryggensiteindicatesasurplusofwasteinrelation
tofields,whichagainindicatesahigherpopulationthannormalonanordinaryfarm.Furthermore,theyoungestlayerscontainedpollenfrom
specieswhichdidnotgrowinwesternNorway,
indicating the import of grain. On this basis,
KrzywinskyandKalandsuggesteda‘Merovingian/Early Viking Age pre-urban ‘coastal settlement’anda‘LateViking/EarlyMedieval‘early
town’ on the northern shore of the Vågen Bay
(Krzywinski and Kaland 1984, 31-33). These
arguments formed the basis for a more elaborateanalysisperformedbyHjelleinhermaster’s
thesisin1986.Sheanalysedsimilarwaste-layers
containing macrofossils and pollen indicating
importedgoods.Hersamplesweretakeninthe
earlymedievalVeisaninletbetweenHolmenand
thenortherntownarea,andsomeweredatedto
theVikingage.Sheinterpretedthelayersasindirectevidenceofadensersettlement,acentral
gatheringpointfortheexchangeofgoods,inthe
close vicinity of the sampling location (Hjelle
1986,55-57,61,71).Inlaterinvestigationsinthe
southerntownarea,fast-accumulatedlayerswith
onlypollenpresenthavealsobeeninterpretedas
theremainsofwaste-layersdumpedintothesea
(Hjelle1998).Thepresenceofimport-indicating
pollen in naturally deposited layers has formed
themainargumentforadensersettlementinthe
vicinityofthesamplinglocations(Hjelle1994,
164;Hjelle1998).
The arguments presented in 1984 and 1986
foradensersettlementpredatingtheendofthe
eleventhcenturywerebasedonthepresenceof
naturalscientificdataandonindirectreasoning.
Themostcentralpremiseseemstobethatwaste
wouldnotbethrownintotheseainaruralsettlement,butusedasfertiliser.Thereforewastelayers deliberately dumped in the sea are interpretedtoreflectadenserpopulationthannormal
onanordinaryfarm.
Twofactorscharacterisethelayersinthe1984
and1986studies:(1)thelayersaccumulatedfast
and (2) macrofossils, which could be identified
asanthropogeneouswaste,wereidentifiedinaddition to pollen. In my view both these factors
mustbepresentwhenalayerisinterpretedasa
waste-layerdumpedintotheseaintheclosevicinity of the sampling location because: ad (1)
other mechanical factors than actual dumping
mayhavecausedthefastaccumulationofalayer.
Such factors may be natural as well as human;
achangeofcultivationsystemsinanareamay,
forinstance,causeasuddenfastaccumulationof
layers. The fast accumulation of a layer cannot
inisolationbeusedasaconclusiveevidencethat
wastewasdumpedinthesea.Ad(2)inaddition
topollen,macrofossils,whichcanbeidentified
asanthropogeneouswaste,mustalsobepresent
inthelayer.Because,whereaspollenmaytravel
over longer or shorter distances by for instance
air or water (Hjelle 1986, Section 5.2), macrofossilsarenotsomobileandtendtobemorelocalinorigin(Robinson,Kristensen,andBoldsen
1992, 68). They would probably not travel far
withsurfacewater.Thepresenceofmacrofossil
materialaccordingly‘ties’theplacewheremasses
weredumpedtotheclosevicinityofthesampling
location.Thepresenceofthemacrofossilsinthe
materialinthe1984and1986studiesindicates
that these waste-layers were in fact dumped in
theseaintheclosevicinityofthesamplinglocation,andwerenotforinstancefirstusedasfertiliseronfieldsandlatercarriedintotheseaby
naturalmechanicalforces.
Polleninafastaccumulateddepositmaystem
either from (1) a pollen containing waste-layer
depositedintheseaor(2)frompollen,capsulatedinawaste-layerusedasfertiliseronfieldsand
laterwashedintotheseathewithsurfacewater.
Thereisnowayofdeterminingwhichthrougha
botanicalsamplealone.Inthepresentstudy,fastaccumulated deposits that contain only pollen
andnotmacrofossilidentifiedastheremainsof
anthropogeneouswastearethereforenotconsideredassufficientevidencethatwastemasseswere
dumpedintheseainthevicinityofthesampling
location. Fast-accumulated deposits containing
thistypeofmacrofossilsmay,however,convincinglyrepresentwastemassesdumpedinthesea
inthevicinityofthesamplinglocation.
4Generalpresentationofthearchaeological,botanicalandtopographicalsources
41
Resultsfromfivebotanicalinvestigationsare 5GENERALMETHODOLOGICAL
usedinthepresentstudyassourcesfortheearly APPROACHES,DEFINITIONS
history of Bergen and the natural topography.
Ecofacts from dated contexts are used in the ANDDEMARCATIONS
Asseenabove,thearchaeologicalandbotanical
samewayasotherarchaeologicaldata.
data stem from excavations carried out within
different scholarly traditions, this has a great
Randomobservations
In Bergen, archaeological random observations impactontheinformationpotentialandonthe
havebeenrecordedsincethenineteenthcentury. methodsIchoosetoapply.Itisimportanttotry
Theobservationsweremadebyinterestedlaypeo- toactivatethesourceswhetherexcavatedduring
ple,ofteninconnectionwithconstructionwork. thenineteenthcenturyormorerecently,soasto
Mostoftheinformationfromrandomobserva- makethebestoftheinformationpotentialinhertionsmadethroughtheyearshasbeencollected entintheavailablesources.
in connection with the regular excavations, or
localhistorianshaverecordedthemonotheroc- TheBergenarea
casions.Fiveoftheseobservationsprovideuseful TheareaaroundtheVågenBay,herecalled‘the
informationforthereconstructionofthenatural Bergenarea’isincludedinthestudy(cfFigure
topography.Otherinformationhasbeengained 1)andIstudythisareaasonesite.Thenatural
whenartefacts,foundduringgroundworkwere topographicalfeaturesoftheBergenareaaround
giventotheUniversityMuseumofBergen.The 1000differedconsiderablyfromthoseoftoday.
accessioncatalogue(tilvekst)forTheUniversity Centuries of building activities and land reclaMuseumcontainsadescriptionofthe‘strayfinds’ mationhavechangedthelandscapethoroughly.
-artefactshandedovertothemuseumthrough Reconstructing the natural topography around
theyears,butgenerallythefindscircumstances 1000isnecessary,asabackgroundforthespaarevague.Inordertolocaliseactivityfromthe tial analyses. The reconstruction of the natural
period under investigation, outside areas with topography is based on data derived from arregularexcavations,Ihavestudiedallstrayfinds chaeological,botanical,geo-technicalinvestigafrom the Bergen area. However there were no tionsandinvestigationsinconnectionwithother
artefacts that can be dated typologically to be- groundwork. The methodological approach to
thereconstructionisfoundinChapter6,Appentweentheninthcenturyandc1170.
dix 1 presents the data behind the reconstruction.
Geo-technicalinvestigations
Geo-technicalinvestigationsinconnectionwith
constructionworkhavealsocontributedimpor- Adiachronicapproach
tant information and several investigations are Theperiodfromtheninthcenturyuntilabout
usedassourcesforthenaturaltopography.Ihave 1170isstudiedinordertocreateadifferentiated
gonethroughrelevantreportsonsubsoilcondi- understandingoftheemergenceofBergenasa
tions from Norsk teknisk byggekontroll (NOTE- town.Ihavechosenadiachronicapproachand
BY)thisismaterialfromgeo-technicalinvestiga- the material is divided into five archaeological
tionsperformedinconnectionwithconstruction timelevels/horizonsthatserveasthechronological framework. In order to establish the beginworkandinsomecasesresearchprojects.
ningandendofthehorizonstheabsolutechronologiesoffivesiteshavebeenstudied.Theseare:
Maps
GrunnkartBergen1992servesasasourceforthe Koengen(site1),Bryggen(site6),Finnegården
reconstructionofthenaturaltopography,several 6a (site 26), Vetrlidsalmenningen (site 30) and
contour lines on bedrock are taken from this Domkirkegaten 5 (site 38). The time spans of
thefivehorizonsaredefinedonthebasisofthe
map.
beginningandendofphasesinthearchaeologicalmaterialfromthesefivesites.
42
• Horizon1coverstheperiodfromtheninth
century to the first decades of the eleventh
century:c800-c1020/30
• Horizon 2 covers the period between c
1020/30andthelastquarteroftheeleventh
century:c1020/30-c1070
• Horizon3coverstheperiodbetweenc1070
andc1100
• Horizon4coverstheperiodbetweenc1100
andthelate1120s
• Horizon5coverstheperiodbetweenthelate
1120sandc1170
The younger horizons also happen to coincide
witheventsknownfromthewrittensources.As
alreadymentionedKingOlavKyrreissupposed
tohavefoundedBergenabout1070.Theperiod
1103-1122wasthereignofKingØysteinMagnusson(Helle1982,113)whowasaveryactive
initiatorinBergen.In1170,therelicsofStSunniva were transferred from Selje to Bergen and
Bergenwasofficiallyestablishedasanepiscopal
residence at about this time (Helle 1982, 92).
Furthermore, a fire struck the town of Bergen
in1170/71(Helle1998,23),correspondingwell
withaconvenientfire-layerthatmarkstheendof-phaseatmanyarchaeologicalsites.Theshort
timespansprovidedbythehorizonsgivetherare
opportunitytodiscussthesourcesforstudying
earlyBergencloselytothehistoricalcontext.
Structuresassignedtoahorizonarethosethat
wereinusebytheendoftheperiodrepresented
bythehorizon.If,hypothetically,twophasesof
structures are dated to the period between the
1120sand1170,onlytheyoungeststructuresare
assigned to horizon 5. Culture-layers and artefacts/ecofacts are assigned to a horizon accordingtothephaseofstructurestheyareassociated
withoraccordingtotheirhistoryofdeposition
(cfp68).
Spatialanalysis
The material is analysed spatially. As a general
methodIwillvisualiseandanalysethesources
throughmaps,anapproachinspiredbytheVisualImpactAnalysis,usedinlandscapegeography
(Emmelin1984;Hansen1994b).Thismethod
implicatestheproductionofmapsasameansto
visualisepatternsandtendenciesinthematerial
and analyse relationships between the natural
topography and buildings/physical structures,
peopleandactivities/artefactassemblagesinthis
setting,patternsandrelationships,whichmight
otherwisebedifficulttodiscern.
Themapsconsistofseverallayersofinformation, where the reconstructed natural topography for the Bergen area around 1000 serves as
thebackgroundlayer.2Geographicalnorthconstitutesnorthonthemaps.3
When applying the Visual Impact Method
it is essential that as many structures as possiblecanbereconstructedandthattheextentof
culture-layersisreconstructedonsiteswhereno
other physical remains have been documented,
thus enhancing visibility. Data comprise buildings,pits,thoroughfares,fences,andsettlement
tracesthatcannotbegivenapreciseinterpretation. Usually, the reconstruction suggested by
theexcavatorisfollowed.Whentheexcavatorhas
describedthematerialwithoutfurtherinterpretation,Ihavereconstructedthesizeorextentof
structuresbyusingcontemporaryparallelmaterialfromBergen.Forinstancepassagesbetween
buildings were often founded on 2 m x 2m
stonefilled timbered caissons in the tidal area.
Abovethetidalarea,however,actualremainsof
thesurfacesofthepassagesarenotalwayspreserved.Wherethe2mx2mcaissonsarefound,
apassageisthereforereconstructed.
Classificationofthematerialintobasic,
supplementaryorgeneralbackground
sources
Inordertoactivatethesources,butnotlosetrack
ofthevaryingdegreesofuncertaintiesinherent
inthematerial,thearchaeological,botanicaland
writtensourcesaredividedintothreecategories
based on the validity of the dates and spatial
location of the material. The sources are thus
classifiedasrespectivelybasic,supplementaryor
generalbackgroundsourcesforthefivehorizons
usingthevarietyofdatingmethodsdescribedin
thissection.Thewrittensourcesaredividedinto
categoriesinthepresentchapteranddrawninto
thediscussionswhenrelevant.Thearchaeological
andbotanicalsourcesaredividedintocategories
andhorizonssitebysiteinChapter7wherethey
arealsovisualisedonmapshorizonbyhorizon.
Basicarchaeologicalandbotanicalsourcesare:
• Well-locatedphysicalremainsdirectlydated
5Generalmethodologicalapproaches,definitionsanddemarcations
43
totheperiodbetweentheninthcenturyand
c 1170, by artefacts, masonry, stratigraphy
and/ordendrochronology
• Well-located remains of abbeys, castles,
churches and churchyards, documented
through archaeology or written sources,
erectedorunderconstructionintheperiod
beforec1170
Archaeological and botanical general backgroundsourcesare:
• Material which cannot be dated more preciselythanrelativelyas‘olderthan’theoldest
datablematerialinaninvestigation
When dating sources indirectly by horizontal
linkstodatedsourcesatothersitesinthevicinitytheindirectlydatedsourcesbecome‘interreSupplementary archaeological and botanical lated‘withthedirectlydatedsources.
sourcesare:
• Archaeological material, which cannot be Informationderivedfromthewrittensourcesis
dated through the methods mentioned alsodividedintothecategoriesofbasicandsupabove,but,whichonthebasisofbroadcar- plementarysources.About1135,OrdericusVibon14(14C),thermoluminiscence(TL),and talismentionsBergenasatownlocatedonthe
pollendatesorindirectmeansofdating,may coast of Norway (Ordericus Vitalis V, 220-21;
represent early Bergen. The indirect means Helle 1982, 3), this is the only contemporary
of dating are the establishment of vertical writtensourcethatmentionsBergeninthepe(time depth) or horizontal (contemporane- riod investigated. The four thirteenth century
ity)linksbetweendirectlydatedsourcesand Kingssagas:Morkinskinna,Fagerskinna,Heimundated sources. These links may be made skringla, and Sverre’s saga and the Orkneyinga
considering:
saga have, however, been used as main sources
• Changes in the orientation of struc- for aspects of the early history of Bergen. The
turesthroughtime
firstthreeKingssagasdescribeeventsthattook
• The number of phases below better placeinBergenintheyearsbetweenthe1130s
datedmaterial
and1160.Morkinskinnawasmostlikelywritten
• The estimated age of timber struc- inNorwayinthelasthalforquarterofthethirturesthatwerenotdestroyedinfire
teenthcenturybyanIcelander.Anolderversion
• Reusedwoodinyoungerphases
apparently existed and may have been written
• Horizontalpatternsinthematerialon downasearlyasbetween1217and1222based
closelylocatedsitesdiscernedthrough on older manuscripts (KLNM XI 704-705).
theVisualImpactAnalysis
Fagrskinna was recorded between 1220 and
• Parallels in building technique and 1240inNorwayperhapsbyanIcelandicauthor,
thechoiceofmaterials
using older manuscripts, amongst others prob• The location of ‘vacant’ sites, that is sites ablyanolderversionofMorkinskinna,assources
where the natural topography has been (KLNM IV 139-140). Heimskringla was most
reached, but where structures or culture- likelywrittenbytheIcelanderSnorreSturlason,
layers other than cultivation layers are not probablyaround1230.Snorreusedoldermanpresentintheperiodsrepresentedbythevar- uscripts, most likely both Morkinskinna and
ioushorizons,exceptwheresitesarelocated Fagrskinna, and perhaps also oral tradition as
ontopofprotrudingbedrock.4
sources (KLNM VI 299-302). The three sagas
• Thelocationofmonuments(abbeys,castles, are remote in time to the events described but
churcheschurchyards)describedinthewrit- theydo,amongother,citescaldicpoemsthatare
tensourcesaserectedorunderconstruction consideredtobereliableassourcesfortheevents
priortoc1170,butnotarchaeologicallylo- described(KLNMXV,386ff).ThefourthKings
cated(seebelow)
saga,Sverre’ssaga,wasprobablywrittenbetween
• Dendrodatedreusedtimbers
themiddleofthe1180sand1210.Itisbasedon
eyewitnessesandiscommonlyconsideredtobe
areliablesourcefortheeventsdescribed,itmay
howeverbepoliticallybiased(KLNMXVII55144
558). The Orkneyinga saga was probably writtenattheendofthetwelfthcentury,theparts
thatareofrelevancehere-thosethatconcernthe
deedsandwhereaboutsofRagnvaldKale-may
havebeenwrittenasearlyasc1165(KLNMXII
699-702).
The sagas, to a large extent, describe events
thattookplaceinconnectionwithtwelfthcentury successional disputes with the Norwegian
kingsandclaimantstothecrownascentralactors. The town of Bergen appears as the scene
of the events, localities are mentioned but seldom given a detailed description. A detailed
topographical description of Bergen is thus not
anaiminthesagas.Inasocietywhereoraland
writtentraditionswerestrong,thesagas,evenif
theyareremoteintimetotheeventsdescribed,
maystillbequitereliableconcerningtheevents
andthepersonsinvolved.Theymay,however,be
lessreliableassourcesforsecularbuildingtopography.Thetownareawasdevastatedbyfiresin
1170/71and1198inthesefiresthesecular(timber-)buildingtopographywas,atleastpartially,
destroyed.Itcannotbeexcludedthatthirteenth
centuryBergen,knowntothesagawriters,was
quitedifferentfromtwelfthcenturyBergen.This
consequently weakens the value of the sagas as
sources for the early twelfth century secular
buildingtopography.Nevertheless,thesagasmay
stillserveasbasicsourcesfortheidentificationof
churchesandmonumentsthathavebeenarchaeologicallyinvestigatedoridentified,astheseprobablyhadthesamelocationthroughouttheMiddle
Ages. The sagas serve as supplementary sources
forthelocationofchurchesandmonumentsthat
arenotlocalisedphysicallythrougharchaeology
andtothesecularbuildingtopography.Forother
aspectsofearlyBergen,informationinthesagas
willbeconsideredalongtheway.
Inadditiontothesagas,severalwrittensources
dateanddescribetownfiresthroughtheMiddle
Ages. The extents and dates of these fires have
beenanalysedbyHelle(1998),andhisworkwill
beusedasareference.
risingthearchaeologicalandbotanicalmaterial
inChapter7,Ishallusethefollowingprocedure
asfaraspossible:arelativechronologyisestablished through a stratigraphical analysis of the
structuresand(fire-)layersonthesite.Secondly,
therelativechronologyismadeabsolutebydatingthroughavarietyofmethods.
Generally, the stratigraphical analyses presented in the site reports serve as the relative
chronologyofthesite.Wherenorelativechronologyhasbeenworkedout,Ihaveanalysedthe
materialintheattempttoestablishone.Theabsolutedatefortherelevantmaterialisobtained
throughdendrochronology,potteryandtosome
extent masonry, shoe and comb typology. Also
14
Candthermoluminiscense(TL)samples,and
thepresenceofpollenofCentaureacyanus(cornflower) are considered. When more traditional
meansofdatingarenotavailableanattemptis
madetoestablishverticaland/orhorizontallinks
between the undated material and better-dated
sources.Verticallinksrefertothetimedepthin
the material on a site, horizontal links refer to
the contemporaneity between material on one
sitewithmaterialonothersites.
The different dating methods imply varying
degreesofaccuracy.Dendrochronologyprovides
themostaccuratedate,thatiswhenasampleis
takenfromwoodthathasnotbeenreusedand
which has intact outer tree rings. Dendro materialmaythusprovidenarrowpostquemdates
forwhenabuildingwaserectedoraphasebegan.Dendrosamplescanalsoindicatehowlong
building activities continued within the phase
theyweretakenfrom.Samplesfromreusedtimbermaybeofhelpdatingbuildingactivitiesin
earlier phases, if it is possible to estimate how
manytimestheactualtimberwasreusedbefore
it ended up in its final context. Unfortunately,
this type of information is seldom available in
archaeologicalcontextsandinterpretationsmust
bemadeconsideringthegeneralpatternsinthe
materialfromtheactualsite(cfHansen1998).
Inthistextadendrosampledatedtoforexample
1103isreferredtoas‘after1103’.Whentheouter
Datingthearchaeologicalandbotanical
treeringwasalsothelastyearofgrowthforthe
material
tree,thedateisreferredtoasforinstance‘after
On several archaeological sites an absolute 1103/04’, where the winter of 1103/04 was the
chronologyfortheoldestmaterialhasnotbeen fellingyearofthetree.AllsamplesfromBergen
workedoutpreviously.Whendatingandcatego- have been taken from pine. Sapwood statistics
5Generalmethodologicalapproaches,definitionsanddemarcations
45
concerningthenumberoftreeringshasnotbeen
applied.5
Before I started on this project, the main
part of the archaeological material from before
c1190hadgenerallynotbeendatedindetail(cf
Hansen1994b,7)anddidnotseemtoinclude
stratathatcouldbedatedfurtherbackthanthe
1120s.Inordertoobtainmorereliabledatesfor
thelowermostlevelsofthearchaeologicalmaterial, I started out by taking 61 dendro samples
from stored timbers from several Bergen sites.6
Dendrosamplestakenearlierduringexcavations
were also re-examined thus a body of 85 dendro samples dating to the years before 1170 is
nowavailable,thesamplesarelistedinAppendix
2.Thereuseofwoodwasverycommonatsite
6,Bryggen,especiallyinfoundationsandother
substructures(Hansen1994b,Eskurs1).Thus,
when taking new samples the timbers were examinedfortracesofreuse,thesampledspotwas
alsoexaminedfortracesofsurfaceworkorwear
inordertodetermineiftreeringscouldbemissing.Forboththenewsamplesandsamplestaken
duringexcavationsthefindscontextwasstudied
throughtheoriginalsitedocumentation.These
effortsweremadesothatthedatedsamplescould
beevaluatedinrelationtoreuseandtothereliabilityofthedatesprovided.Samplesthatturned
outtobe‘crucial’formydatingofthesources
have been re-examined after the dating results
werefirstready.7
The dendrochronological analyses have thus
produced ‘fresh’ dating material for the oldest
archaeological phases at the central sites. The
new dating material along with pottery from
theBryggensiteformthebasisforare-examinationofthechronologyofthissiteandtheoldest
phasesherehavebeendatedfurtherbackintime
than formerly assumed (Hansen 1998). This
newchronologyservesasapointofdeparturefor
mystudyofsourcesfromtheBryggensite.The
remainingnewdatingmaterialisdiscussedwhen
reviewingtherelevantsitesinChapter7.
Pottery generally provides a wider dating
frame than dendro samples. The existence and
coexistence of different pottery types provide a
post quem date of how long activity must have
lastedinaphase.Insomecases,potterymayalso
indicateanestimatedantequemdateforaphase
ifverycommonwaresaremissing.Inthisstudy
46
onlythepresenceofwaresisusedwhendating.
Potteryusuallyhasanestimatedstartandendof
productiondate.Thesedatesarerarelydirectly
documentedthroughwell-datedkilnfindsorliterarysourcesforexample,butaremoregenerally
established through finds of (or the lack of)
sherds in other well-documented contexts, implyingamarginofuncertaintyoneithersideof
thedates.
Otherdatingmethodsusedinthisstudyare
studiesofmasonry,shoesandcombs-allwith
rather wide dating frames. Masonry typology
gives a wide dating for when a building was
erected,whereasshoesandcombsindicateadate
forhowlongactivitycanhavelastedinaphase.
Typologyonmasonry,shoesandcombsis,however,onlyappliedasameansofdatingwhenno
othermethodsareavailable.
Dates provided through the carbon 14 (14C)
andthethermoluminiscence(TL)methodsalso
give wide dating frames. Quartz or feldspar is
thesamplingmaterialwhendatingthroughTL,
botharepresentinceramics.Datesprovidedby
theTLmethodreflectthedateofthelasttime
the sampling material was heated up to more
than 500 o C, this method is therefore useful
when dating fire-layers. The dates provided are
givenwitha+/-5-7%uncertaintyfordatesfrom
theMiddleAges(Mejdahl1988).
The14Cmethodgivesadateforwhenorganic
material,fromwhichthesampleistaken,ceased
tolive.Thusnutsortwigswithashortgrowth
periodprovidemoreprecisedatesthancharcoal
fromatreetrunk,forexample,becausethelatter
maybeinfestedwiththe‘oldwoodproblem’if
thecharcoaldoesnotstemfromtheoutermost
treerings.Whennothingelseisstated14Cdates
have been calibrated according to INTCAL98
(Stuiver and van der Plicht 1998) and they are
interpretedusingtheOxCalRadiocarbonCalibration Program8. The OxCal program gives a
graphicpresentationoftherelationshipbetween
the measured 14C date and the historical date
according to the calibration. The presentation
is based upon a statistical analysis and shows a
probabilitydistributionofthematchesbetween
the 14C date and the calibration curve. Even
though the graphic presentation of the OxCal
calibrationprogrammakesitpossibletonarrow
the time intervals with the highest probability,
thesetimeintervalsarestillinflictedwithahigh
degreeofuncertainty.Becauseoftheuncertainties involved in dating through TL and 14C,
datesprovidedbythesemethodsareconsidered
to be better than nothing, but they cannot be
usedalonewhendatingabasicsource.
IntheBergenarea,thepresenceofpollenof
Centaureacyanus-cornflower-maygiveawide
post quem date for the deposition of the layer.
Dating deposits through the presence of this
pollenisbasedonmaterialfromsitesinBergen:
At Nedre Korskirkealmenning/Vågsalmenningen(1998)(site37)pollenofCentaureacyanus
wasnotpresentinadeposit14Cdatedtobetween
730-860.Intheoverlyingdeposit, 14Cdatedto
between810-970pollenofthespecieswas,however,present(Hjelle1998,Section5.2,5.3).At
theBryggensite(site6)thepollenwasnotpresent
inadeposit14Cdatedtotheseventhoreightcenturies(KrzywinskiandKaland1984,29,31).At
theKoengensite(site1)layer11,whichwas14C
datedwithtwopeaksofprobabilityto780-790
or810-1000,containedpollenofCentaureacyanus.Accordingtothematerialfromthesesites,
it appears that pollen of this plant was not introduced until (roughly estimated) the ninth
centuryintheBergenarea(seealsoHjelle1986,
58). Consequently, when they are documented
in otherwise undated contexts, it is reasonable
toassumethatthecontextsstemfromtheninth
centuryorlater.Datesbasedonthepresenceof
Centaureacyanuspollenareconsideredastentative, and cannot be used alone when dating a
basicsource.
Whenthematerialcannotbedateddirectly,
attemptsaremadetomakeverticalorhorizontal
linksbetweentheundatedmaterialanddirectlydated sources. When making vertical links the
number of phases ‘below’ directly-dated phases
andchangesinthegeneralorientationofstructuresonasitemaybeconsidered.Anevaluation
of how long timber buildings and structures
couldlastinBergen,iftheywerenotdestroyed
byfire,mayalsobedrawnuponwhendiscussing
the time depth of the material. The maximum
ageforatimberbuildingdependsonthestructuretypeandfactorssuchasthequalityofthe
buildingmaterial,foundations,climateetc.
AtDomkirkegaten6(site38)thestructuresin
phase10werebuilt‘after1128’andreplacedby
structuresinphase9,built‘after1156’(Komber
etal.1994,111):aperiodofabout25years.At
Finnegården6a(site26)structuresinphase12
were dated to ‘after 1103’ structures in the followingphaseweredatedto‘after1118’.Accordingly,thestructuresinphase12were15-20years
old when replaced. At the Bryggen site (site 6)
buildings that were constructed in the 1120s,
were still in use when the fire in 1170/71 occurred(Hansen1998),thusstandingforabout
50years.Theseexamples,thoughfewinnumber,
showthatbuildingscouldlastforatleast25to
50 years in twelfth century Bergen unless they
weredestroyedbyfire.
Whenhorizontallinksaremade,thecontemporaneityofundatedmaterialwithdirectlydated
sourcesiselucidated;parallelsinbuildingtechnique,thechoiceofmaterialsandotherpatterns
inthematerialdiscernedthroughtheVisualImpactAnalysismaybedrawnupon.Usingverticalandhorizontallinks,whennofirmerdating
evidenceisavailable,cannotbeusedalonewhen
datingbasicsources.
As mentioned, Bergen has been ravaged by
a number of extensive fires through the centuries, many of which have been recorded in the
writtensources(Helle1998).Whendatingthe
archaeologicalmaterialIwilltrytorelatethearchaeologicallydatedfire-layerstothehistorically
knownfires,assumingthatsomeofthearchaeologically documented fires may be identified as
ahistoricallyrecordedfire.Inthiswaythewritten sources serve as an additional and accurate
meansofdating.Therecordedfiresofrelevance
herearethetownfiresof1170/71and1198.
Ifthepublishedmaterialorexcavationreports
fromthesitesdonotanswerthequestionsofdatingandlocalisation,thesequestionsarediscussed
throughtheavailabledocumentationmaterialin
sofarasitispossible.
Assigning basic archaeological and botanical
sources for a horizon is mostly straightforward
sincethequestionofdatingiswellelucidatedby
direct and narrow dates. As for supplementary
orgeneralbackgroundsources,however,thereis
oftennostraightanswertothequestionofdating.InthesecasesIwilldiscussalternativedatingpossibilitiesforthematerial,andeventually
choosethedatingalternativeinvolvingtheleast
‘coincidences’ as the most plausible, or I may
5Generalmethodologicalapproaches,definitionsanddemarcations
47
choosetoomitthematerialfromthestudy.
Thebasicsourcesprovidethemainplatform
inmystudy.Butasweshallseetheyarescarce
in the earliest horizons. When interpretations
are made without basic sources I shall evaluate
thetendenciesinthesupplementarysources.In
thesecasestheinterrelatedsourcesthataredated
indirectly through horizontal patterns do not
count as independent sources. In chapter 7 the
archaeological and botanical material is dated
and assigned to horizons as basic, supplementaryorgeneralbackgroundsourcesonthebasis
ofdatafromthesinglesitesorthroughlinksto
materialfromsitesintheirclosevicinity.Ifthe
verticalandhorizontallinks,drawnuponwhen
assigningthesupplementarysourcesforhorizons
inChapter7,arestrengthenedbypatternsemerginginChapter9,whenImovefromthemicro
scaleofthesinglesitesandtheircloservicinities
andstudythematerialthroughabroaderspatial
analysis,thismaystrengthenthereliabilityofthe
interpretations(thedates)madeinChapter7.To
avoidcircularargumentsthepatternsdiscerned
whenzoomingoutmustbebasedonsourcesthat
arenotinterrelatedwithoneanother.
Implicitly my approach is that patterns discernedonamicrolevel-thesinglesiteandits
vicinities-areevaluatedinthelightofpatterns
discerned when zooming out and studying the
materialinabroaderspatialanalysis.Whenbasicsourcesarescarce,tendenciesinthematerial
basedonsourcesthatarenotinterrelatedwillbe
emphasised,whereasdetailsdependingonsingle
orinterrelatedsupplementarysourcesareconsideredlesstrustworthy.
Theplotasananalyticunit
Due to the variety in the methods applied on
thedifferentsitesIhavechosenascaleofinvestigationthatliesclosetothelowestcommondenominatorforthematerial.Inordertoactivate
thematerialthesmallestanalyticunithastobe
operational on as much material from as many
sitesaspossible.Atseveralsitescontextinformationontheartefactsistoocrudetosupportan
analysisatthelevelofbuildingsorpassages,for
example.Theaccuracyofinformationonstratigraphyandfindslocationonthesitesvariesfrom
descriptionsonagenerallevel,todescriptionsof
finds-locationwithingridsystemsvaryingfrom
48
8 m x 8 m to 2.5 m x 2.5 m, and to descriptionsinrelationtobuildingsandculture-layers.
Quantitativecomparisonsbetweensuchcontext
unitsaredifficulttoestablishandtheyarea-historic.Ihavechosentheplotasthelowestanalytic
unit.Theplotisaunitthatmakessensehistorically(themedievaltownspeopledidliveonaplot
whereastheydidnotliveinagrid).Furthermore
given the large size of the material, the plot is
manageableasananalyticunit.
I have studied the distribution of joining
sherdsandotherartefactsfromsite6,Bryggen;
site26,Finnegården6a,andsite27,Finnegården
3a.Thecross-fitartefactsstemfromallperiods
oftheMiddleAges.AsseeninFigure6cross-fit
artefactsderivedfromdifferentbagsoffinds(site
6) or different layers (sites 26 and 27) seldom
or never crossed the historical plot boundaries
onthesesites;outofthe64examplesonlytwo
crossedaboundary.Myobservationscorrespond
with studies of medieval Lund and Sigtuna in
Sweden (Roslund 1997, 41, 43) and indicate
thatpeopleintheMiddleAgesgenerallydidnot
throw garbage and waste onto the neighbour’s
plot and in this way respected the property
boundaries.Onecannotexclude,ofcourse,that
the cross-fits within plots may stem from loads
of garbage taken from somewhere else and unloaded on one plot. This explanation, however,
seemsquitehypotheticalintheperiodunderinvestigationhere.
Thefindspotsofjoiningartefactsthusimply
that artefacts found on a plot were most likely
usedthereaswelland,asapremiseIwillassume
thatmaterialfoundonaplotreflectsmajorinitiativesanddailyactivitiescarriedoutbyresidents
orvisitorstothisplot.
When interpreting the artefacts found on a
plot, as indicators of how the plot was used it
is possible to activate the artefacts as potential
sourcesaslongaswecanascribethemtoaspecificplot.PlotboundariesareidentifiedanddiscussedinChapter9andtheplotswillbeusedas
analyticunits.
The plots are labelled by the site number
and a letter, plot 6/C is thus plot C on site 6.
Whenplotboundariescannotbeidentifiedata
site,dataisanalysedwiththesiteastheanalytic
unit.Aunitthencomprisestheapproximateexcavatedareainagivenhorizonataspecificsite,
Figure6.Cross-fitartefactsatsite6,Bryggen;site26,Finnegården6aandsite27,Finnegården3a.(Theinformation
oncross-fitartefactsstemsfromLüdtke1989,15;BlackmoreandVince1994,73,8,andfromtheoriginal
documentationfromthethreesites)
5Generalmethodologicalapproaches,definitionsanddemarcations
49
these analytic units are not labelled by a letter,
but merely by the site number. On a few sites,
trenchinvestigations,stratafromtheearlyyears
ofBergen’shistoryaremerelyfoundin‘pockets’
in the sections, not as a continuous blanket of
layersandstructures.Datafromeach‘pocket’of
stratawillalsobetreatedasananalyticunit.The
analyticunitsare,liketheplots,labelledbythe
sitenumberandaletter.Unit30/EisaccordinglyunitEonsite30.Whenstudyingthefunction
of buildings the single buildings will represent
theanalyticunit.
Someartefactswerefoundintheboundaryarea
betweentwoplotsandtheycouldnotbeassigned
to either of the plots with any certainty. These
artefactshavenotbeenincludedinmyanalysis,
astheydonotmakeupasignificantnumber.Artefactsfoundbetweenplot26-27/Band27/Care
anexceptiontothisrule,asthisgroupmakesup
themajorityoffindsfromthetwoplots,andthey
havebeenincludedintheanalysis.
Levelofinquiry
Thearchaeologicalmaterialreflectsamyriadof
single activities spanning from the accident of
breakingapottoorganisingtheplotsystemofa
town.Actorsfromdifferentlevelsofsocietyhave
intentionallyorunintentionallycarriedoutdifferent activities. In principle the archaeological
materialmaythusreflectactivitiescarriedoutby
individualsfromdifferentsocialcategories,representingthemselvesormoreresourcefulactors.
Myscaleofinquiryisnotactivitiescarriedout
bysinglepeopleassuch,butratheractivitiescarriedoutbypeoplerepresentingdifferentgroups
ofactors:suchasrepresentativesforthekingor
representatives for the townspeople and visitors
orguestsofthetown.
Landuse,terminology
Fourmaintypesoflandusearereflectedinthe
sources:(1)secularoccupation,(2)monumental
manifestations,(3)cultivation,and(4)anatural
topography.Churches,churchyards,monasteries
oraroyalresidenceareexamplesofmonumental
manifestations.Culture-layersthatindicatecultivationinabroadsense(pastures,meadows,arablefields)aredesignatedascultivationlayers.A
siteisgenerallycharacterisedbysecularoccupationwhenstructuresorculture-layersotherthan
50
cultivationlayers,monumentalconstructions,or
tracesoftheregulationofanareaintoplotsare
identified(asitemaythushavebeendividedinto
plots without being occupied/taken into use).
Determining whether the land was taken into
useornotisoftenquitestraightforward,because
structures and culture-layers can be observed
visually during excavation, unless preservation-
or weather conditions have disturbed the possibilities of documentation. However, it cannot
beexcludedthatstructuresorculture-layerswere
overlookedduringexcavationorthatthenatural
subsoilwasnotreachedonasite.Thelanduseon
thesitesmaybecharacterisedby:
• ‘Secular occupation’, when the site is occupied/built on/settled, and structures or
culture-layers other than cultivation layers,
boundary indicating structures or monumentalconstructionsarefound.
• Monumentalmanifestations,whenchurches,
churchyards,monasteriesorroyalresidences
arefoundarchaeologicallyorotherwisedocumented.
• ‘Cultivation’,identifiedthroughbotanicalor
archaeological investigations of the natural
subsoil/cultivationlayers.
• ‘Vacantareas’,whennostructuresorculturelayershavebeenidentifiedabovethenatural
subsoil.
Approachestotheartefactmaterial
As a point of departure all artefacts from contexts dated to before c 1170 are considered as
potential sources for my study of early Bergen.
Textiles and rope have, however, been omitted
fromthestudybecausetheydemandspecialinsight and analysis and have not yet been fully
publishedbyspecialists.Theremainingartefacts
arestudiedandclassifiedaccordingtothespecificaimsofthestudy,itisthusnotanaimtogive
anexhaustivepresentationoranalysisoftheartefactmaterialassuch.Theartefactsarestudied
inseveralsteps.First,theartefactsareidentified
through their context and assigned to horizons
and categories. Second, the artefacts are classifiedinordertogetanoverviewoverwhatisthere
tobestudied.Finally,theartefactsofrelevance
areanalysedanddiscussedthematicallythrough
aqualitative,contextualandspatialapproach.
Divisionoftheartefactsintocategories
Within the horizons, artefacts and ecofacts
areassignedtoartefactcategoryIorIIaccordingtothehistoryofdepositionofthelayersin
whichtheywerefound.Basically,culture-layers
with four different histories of deposition may
be present in the material: (1) In situ culturelayers, which are layers found in their original
and functional context. (2) Culture-layers that
areredepositedbutprobablynottransportedso
far.Theselayers,andtheartefactsbelongingto
them, may represent ordinary everyday products and rubbish accumulated within a plot or
aproperty.Althoughtheyarenotfoundintheir
functionalcontext,theyprobablystillrepresent
activitiesthattookplaceintheclosevicinity.(3)
Culture-layers,whichhavebeenredepositedand
transportedwhenusedasfill-masses,forexample
in connection with construction work, and (4)
culture-layers/artefactstransportedbyfluvialaction,suchasbyastream.Thetwolattertypesof
culture-layerswiththeirartefactsmaybefound
farawayfromtheirfunctionalcontext.
Itsionlypossibletoobtainthedetailedinformationnecessaryforaclassificationofthelayers
accordingtoallthesecategoriesfortherecently
excavated material. It is therefore, possible to
distinguish between two categories of artefacts
only:categoryIthatrepresentsartefactsfromin
situlayersandlayerswhichbelongtotheredeposited material that has not been transported
far. And artefact category II that represents artefactsfromfill-massesusedduringconstruction
workandartefactstransportedbyfluvialaction.
Byusingthesetwobroadercategoriesmostofthe
archaeological material can be activated in my
analysis.
Ithasnotbeenpossibletoconnectanyartefactsdirectlytomonumentalorotherstructures
investigatedpriorto1955.Thefollowingcriteria
areusedtodefinethetwocategoriesofartefacts
foundonsitesexcavatedbetween1955and1979.
ArtefactsincategoryIcomprise:
• Artefactsfromfire-layers,includingartefacts
describedasinoraboutafire-layerinthefield
documentation
• Artefactsfromcontextsdescribedas0-10cm
underafire-layerinthefielddocumentation
• Artefacts which are described as under a
fire-layer but above the structure that was
destroyed/wentoutofuseinthefire,likea
floororthesurfaceofapassage
Category II consists of: the remaining artefact
materialfromagivenphase.
Thematerialfrominvestigationscarriedoutbetween1979-1998issortedintothetwocategoriesbythefollowingcriteria.CategoryIconsists
ofartefactsfrom:
• Insitufire-layersandotherlayersdescribed
asinsitu
• Layers that have accumulated in the activity or destruction stage of the horizon, but
whicharenotintheirfunctionalcontext
Category II consists of the remaining artefact
material, that is, material from the foundation
stageofthephasethatrepresentsahorizon,materialfromfluviallayersisalsoincluded.
In some cases the material does not fit into
the time frames given by the defined horizons.
Hypotheticallyaphaseatasitemayhavelasted
from c 1100 to the 1150s: the structures were
builtc1100,butartefactsinthephaseallstem
frominsitulayersinthebuildingandthusrepresenttheyearsaround1150.Thestructuresare
then a source for horizon 4 (1100-1120s). But
theartefacts,representingthetimearound1150,
aretooyoungtosignifyactivityinhorizon2.In
thiscasetheartefactsareassigned,ascategoryII
findstothefollowinghorizon5(1120s-1170),as
thisiswheretheywouldhaveendedup,hadthey
beenusedasfill-massesinconnectionwiththe
constructionofthefollowingphase.
Whenstudyingthefunctionofbuildingsonly
artefacts of category I are drawn upon. Otherwise,whenstudyingthedistributionofartefacts,
both category I and II artefacts are included
withoutconsideringtheclassificationintosource
types.Thepicturethatcouldhavebeendrawn
based on the category I finds alone would ideallyrepresentapictureofa‘moment’bytheend
oftherespectivehorizons.However,thispicture
would have many ‘white spots’ since several of
the find-yielding sites, as we shall see, did not
producefindsthatcouldbeclassifiedascategory
Ifinds.WhenincludingthecategoryIIfindsan
accumulated picture of many ‘moments’ of activity that have passed within the duration of
a horizon can be obtained. To some extent the
5Generalmethodologicalapproaches,definitionsanddemarcations
51
activity traces may also stem from the preceding horizon if artefacts from this horizon were
redepositedinfoundationlayersandclassifiedas
category II finds to the next horizon. This is a
problem that must be accepted, as most of the
materialhasnotbeendocumentedinsuchdetail
sothatacleardistinctioncanbemadebetween
redepositedlayersfroma‘present’anda‘preceding’horizon.
Cutcheon (1997) and Egan (1998). The classification of combs follows Wiberg (1977) and
Flodin(1989)withafewsupplementsfrommy
side. Inger Kellmer’s unpublished manuscript
and notes on the combs from the Bryggen site
havebeenveryusefulduringthestudyofcombs.
Wastefromcombproductionhasbeenidentified
withreferencestoRytter(1991).Ihaveclassified
theremainingartefactsbycomparingformand
material with published or otherwise accessible
Classificationoftheartefacts
illustrations and descriptions of artefacts from
Asmentionedearlier,someartefactgroupsfrom medieval Norway, Denmark, Sweden, GreentheBergenmaterialhavebeenstudiedandclas- landandEngland.
sifiedbyothers.Ihavedrawnuponthesestudiesasfaraspossible.Theremainingmaterialhas Aqualitative,contextualandspatialapproach
been classified as part of post-excavation work. InthelastpartofChapter7Ihavearguedthat
Thequalityandvalidityofthelatterworkvaries therepresentativityofthematerialinrelationto
andare-classificationhastoawideextentbeen thevarietyofwhatwasonceinuseandofwhat
necessary.Insofarasitispossible,classification waspreservedinthegroundshouldberegarded
systemsestablishedthroughformerstudieshave asfairlycomparablefromsitetositeintheearly
beenusedasitisbeyondtheaimheretodevelop periodofthetown’shistory.Asfarasthefindsnewsystems.InafewcasesIhavedevelopedes- frequency from site to site within the horizons
tablished systems further to meet the needs of isconcerned,therearesomanymethodological
thepresentstudy.
circumstancesinvolvedthataquantitativeanalyThe vast majority of pottery from the Bry- sisofthematerialacrossthefind-bearingplots/
ggen site was classified during post-excavation sites will not be possible. A qualitative, contexworkbyARoryDunlopandIanReed.Dunlop tualandspatialapproachisthereforeapplied.
hasalsoclassifiedmostofthepotteryexcavated
Asalreadyargueditislikelythatwhatwasacafter 1980. In addition, Pingsdorf and Paffrath tually found in an analytic unit, was also used
ceramicsfromtheBryggensiteandpotteryfrom there.Inalllikelihoodproductionwasteortools
southeastEnglandhavebeenstudiedbyrespec- found on a plot stem from activities there. A
tivelyHartwigLüdtke(1989),LynnBlackmore qualitativeapproachimpliesafocusonartefacts
and Alan Vince (1994). All these pottery clas- andecofactsthathaveactuallybeenfoundand
sificationshavebeenuseddirectly.Ihaveclassi- assignedtothehorizons,andananalysisofwhat
fiedpotterythathadnotbeenstudied,withthe theymayrepresent.Presenceratherthanabsence
kindhelpofDunlop.IngvildØye’sclassification isthusevaluatedassignifiersofdailyactivities.
oftextileequipmentfromtheBryggensite(Øye
When studying the artefacts or ecofacts, the
1988)isusedandappliedtothematerialfrom single finds are regarded as significant sources
other sites. The classification of fishing equip- forvariousactivitiesstudied.Incaseswherethe
ment from the Bergen area (Olsen 1998) and practicalfunctionofanartefactisambiguousthe
weapons from Bergen (Nøttveit 2000) are also itemcannotbeusedinisolationwhenidentifygoing to be used. Tallysticks are classified ac- inganactivity.Thelocalcontextthatisthefind
cordingtoGrandell(1988),footwearandleath- spotoftheartefactandotherfindsfromtheplot/
erwasteaccordingtotheprinciplesoutlinedby unitisthendrawnintothediscussion.
Larsen(1992).OtherleatherarticlesareidentiInthesurveysofdailyactivitiesartefactsare
fied through Marstein (1989). Wooden objects considered as sources regardless of the number
are classified according to Mårtensson (1976), ofobjectsandregardlessofthenumberorclass
Christensen(1985),Weber(1990)orFuglesang (basicversussupplementary)oftheanalyticunits
(1991aand1991b).Metalobjectsareidentified involved.However,toregardsinglefindsorfinds
throughFærden(1990),Hurley,Scully,andMc- fromasinglesupplementarysourceasconclud52
ing evidence to an activity is considered as too
hazardous,asthisapproachpresupposesthatall
artefacts have been classified, dated, localised
andsoforthcorrectly.Thisisofcoursetheideal
situation,buthardlyareality,giventhehuman
element involved in all these procedures. Accordingly,thetendencyinthedistributionpatternoffindsacrosstheplots/unitsisemphasised
whenevaluatingthereliabilityofthematerialas
sources for studying daily activities. And when
asourceforaspecificsubjectinahorizonstems
fromasinglesupplementarysourceorfrominterrelated sources alone it cannot be used in
isolationasconclusiveevidence.Pollenisnotaccountedforinnumbersbuttheirpresenceisused
asasource.
The significance of the absence of certain
groupsoffindsisonlydiscussedinafewcases
onselectedplots/unitswhereatleasthalfofthe
availablebuildinglandontheplotwasexcavated,orwherethenumberofartefactsassignedto
theplots/unitsperm2islargeenoughtocarrya
meaningfulquantitativeanalysis.Theonlyplots/
unitsthatmeetthesecriteriaarefoundinhorizon5.Theseareplots6/C,6/D,6/Eand6/G.
Forhorizon5Ihaveconsideredmorethanfour
findsperm2(cfTable28)asasufficientnumber
ofartefatsforaquantitativeevaluation.
6RECONSTRUCTIONOFTHE
NATURALTOPOGRAPHY
ABOUT1000
I will now reconstruct the natural topography
thatisgoingtoserveasthebackgroundlayerin
maps produced throughout the study, and as a
physicalsettingforinitiativesandactivitiesthat
tookplaceintheBergenareauntilc1170.
Stateofresearch
Theextentoftheavailablebuildinglandabout
1000-1100 has been discussed throughout the
historyofresearchonBergen.Researchershave
mainlybeeninterestedintheoriginalshoreline
oftheVågenBayandthelimitofbuildingland
towards the hill Fløyfjellet. Koren-Wiberg was
the first to make a reconstruction of parts of
6Reconstructionofthenaturaltopographyabout1000
the natural topography based on underground
observations, he made a reconstruction of the
Vågen shoreline in the middle town area (Koren-Wiberg1921,15ff).ThetownengineerØW
Grimnes later supplemented this picture by reconstructing the shoreline by Holmen and in
thesoutherntownareabasedonboreholesfrom
testdrillingandwrittensources(Grimnes1937).
BasedonarchaeologicalresultsfromtheBryggen
site,HerteigfoundthattheearlymedievalshorelinealongthenorthernshoreofVågenactually
ran25-30mfurthertothenorththansuggested
byKoren-Wiberg,thusleavingamuchnarrower
strip of building land between the Vågen Bay
andFløyfjellet(Herteig1969,126ff).EgillReimerspresentedthefirstdetailedreconstruction
ofthe0-5mabovesealevel(masl)contourlines
tothesouthandwestoftheChurchofStMary’s
in the northern town area, his reconstruction
wasbasedonarchaeologicalexcavationsinthis
area(Reimers1974).
In 1976 H K Fritzvold, an engineer, wokred in collaboration with Helle and produced
a map of the shoreline in Bergen about 10001100 published in Helle’s town history (Fritzvold1976,Tegning1;Helle1982).ThereconstructioncoveredtheVågenBayandtheBayof
Alrekstadvågen, today’s Lille Lungegårdsvann.
Ithaditsmainemphasisontheshoreline,but
contourlinesbetween-8and20maslwerealso
reconstructed in some places. The reconstructionwasbasedondatafrombuildingprojects,
reports from archaeological excavations, older maps, boreholes and surveys by Fritzvold
(Fritzvold 1976). Along with a map Fritzvold
gave an account of the methods and data behind the reconstruction, thus it has been possible for other researchers to use and evaluate
hismapandsupplementitwithnewdata.All
laterreconstructionsofthenaturaltopography
have in principle been supplements or modificationsofthismap.Krzywinski’scomputerised
reconstructionofthenaturaltopographyinthe
Holmenarea(Krzywinski1991),themappresentedbyMyrvollin1993basedonFritzvold’s
map and data from archaeological excavations
fromthelate1970suntil1993(Myrvoll1993,
87), and also the map presented by myself in
1994 (Hansen 1994b) supplement his reconstruction.
53
The present reconstruction of the natural topography also presents an adjustment of
former reconstructions based on up-to-date information. Contour lines and archaeologically
documented streams/small rivers in the Bergen
area about 1000 are reconstructed. The reconstructionofthenorthern,middleandsouthern
town areas is largely based on my earlier work
(Hansen1994b).Sincesomemodificationsand
new sources have come forward, I have chosen
topresentthesourcesusedforthepresentreconstructionasawhole.
A map is a graphical presentation of an interpretation of data and it is a strong medium.
Accordingly it is important to account for the
methods behind the production of the map, to
presentthesourcesforthemap,anddiscussthe
problematicpartsofthereconstruction.
Methodologicalapproachesand
premisesforthereconstructionofthe
naturaltopographyabout1000
The sources for the pre-urban topography are
divided into basic and supplementary sources.
Basicsourcesare:
• Height and orientation of moraine or bedrockrecordedduringarchaeologicalinvestigationsorothergroundwork
• Bedrock contours from Grunnkart Bergen
1992(basedonaerialphotographs)
• BedrockcontoursfromGeneralkart1879/80
(basedontrigonometry)
Supplementarysourcesare:
• Measurementofbedrockandmorainefrom
test drilling. These data are considered less
secure than data from groundwork as they
oftenshowthelevelwherebedrockandnot
moraine was encountered. According to
Fritzvold(Fritzvold1976,5)theremayoften
beasmuchas1-3mofmorainicmasseson
topoftheassumedbedrocksurface.Insome
casesitisthusrealistictoadd1-3mtothe
measurements of bedrock from boreholes.
Another aspect to be considered is that a
probeborehaslimitedpenetrationabilityin
firmmassesandthedrillmaystopatblocks
ofstoneandcompactmorainemasses.This
54
•
•
•
•
•
may have consequences for the evaluation
ofthebedrocksurfaceandthethicknessof
massesaboveit(Fritzvold1976,7)
DatafromKoren-Wiberg’sinvestigations,as
thesemeasurementsareusuallydocumented
as metres under the pavement not in masl.
ThelevelofthepavementinKoren-Wiberg’s
daysisestimatedonthebasisofGeneralkart
1879/80.
The level of the lowest recorded culturelayer,asthislevelisassumedtorepresenta
maximumheightofthenaturalsurface
Generalinformationonobservationsofmoraineorbedrockfromolderinvestigationsor
groundwork
The main configuration of the mountains
aroundmedievalBergen
Theorientationofthebuildingsintheposttwelfth century town of Bergen. It is assumed, that the orientation of the harbour
constructions and culture-layers later than
thetwelfthcenturyBergenreflectsthemanmade or natural landscape they were built
in.Theassumptionisbuiltonobservations
intheBryggenmaterial:attheBryggensite
(site6)theoriginaleleventhcenturynatural
surfacewasreachedandthetwelfthcentury
harbourfrontwasexcavated.Laterharbour
fronts were also excavated. It is clear that
buildings closest to the waterfront always
related directly to the contemporary waterfrontandperpendiculartothewaterfront(cf
figuresinHerteig1990;Herteig1991).
InthereconstructionIassumethat:
• Theleveloftheseawasthesameabout1000
astoday(Herteig1969,100).Thereforethe
contourlineof+/-0isdefinedastheshorelineabout1000(Fritzvold1976,5).
• Thewaterlevelbynormalhighandlowtide
was the same in the Middle Ages as today
(Herteig1969,100).Todaytheaveragehigh
tidesealevelis0.46maslforBergen,theaverage spring tide level is 0.62 masl and the
highest sea level observed at storm surge is
1.53 masl (personal communication Norges
Sjøkartverk,Stavanger).
• The transition between culture-layers and
moraineorbedrockrepresentsthepre-urban
surface if nothing else is indicated. In fact
thisassumptionmaybesomewhatunrealisticsincetheremusthavebeenagrowthlayer
abovethemorainicdeposits.However,inthe
mapscalepresentedheresuchdetailsarenot
significant.
• Itisassumedthattheorientationofbuildings
andculture-layersinthetownthatemerged,
reflect the pre-urban landscapes they were
placedin.
The map has been produced with an equidistanceof1masfarasthiswaspossiblethrough
the available sources. The maps (Figure 62)
where the sources for the natural topography
are presented have an equidistance of 1 m in
the northern, middle and southern town areas;
outside these areas an equidistance of 5 m has
been considered sufficient in the presentation.
Onothermapsthroughoutthestudywherethe
natural topography serves as a background, an
equidistanceof5misshown.Thecontoursfor
theseabottomhaveonlybeenreconstructedin
afewareas,sincedataisscarce.Streamsorsmall
riversareonlydrawnwheretheyaredocumentedarchaeologically.Thecontourlineshavebeen
drawnmanuallybyinterpolatingbetweenpoints
withaknownlevel.Inareaswherebasicsources
are scarce, the contour lines have been drawn
based on the nearest basic sources and an estimatebasedonthesupplementarysources.Fora
numberofareasreconstructionsbasedonbasic
sources,andmethodssimilartothoseusedhere,
are already available. Where no new data have
cometolight,Ihaveusedthesereconstructions.
Ihavedrawnthemapmanually,althoughcomputerprogrammesareavailablefordoingtheinterpolationjob(seeegChristophersen,Cramer,
andJones1989).Amanuallyproducedmaphas
theadvantagethatdata,whichcannotbegiven
objectivenumbersmayalsobetakenintoconsideration,suchastheobservationthatbuildingsin
themedievaltownreflecttheterraintheywere
builtin.Bydrawingthemapmanuallyitisthus
possibletohaveamorequalifiedreconstruction,
evenofareaswherethebasicsourcesarescarce.
Themethodofinterpolatingknownpointsproduces a picture of a smooth and less detailed
landscape. It is therefore important to visualise
where the representation is built upon basic or
supplementarysources.
6Reconstructionofthenaturaltopographyabout1000
Figure62inAppendix1presentsthesources
behindthereconstruction.Numbersonthemap
refertothelistofsourcesandtothediscussions
behindthecourseofthecontourlines,foundin
Appendix 1. The reconstruction of the natural
topography serves as the background layer for
mapsproducedthroughoutthepresentstudy.
Majorfeaturesofthereconstructed
naturaltopographyabout1000
AsshowninFigure62,theVågenBaystretchesnorthwest-southeastintothelandfromthe
inner coast of western Norway. In about 1000,
Vågen was deeper and wider than today. The
originalnorthernshorelineranasmuchas130
mnorthofthemodernquayfrontandthebay
extended some 300 m further eastwards than
today.AstripoflandseparatedVågenfromthe
BayofAlrekstadvågenthatwasalsomuchdeeper than today’s Lille Lungegårdsvann. Then as
todaytheNordnespeninsulamadeupthesouthern shore of Vågen. Along the northern shore,
whengoingfromthewesttotheeast,onewould
firstencountertheHolmenpromontorythatwas
separatedfromthelatertownareabytheVeisan
inlet.FromthemouthofVeisan,theshoreran
inslightcurvesbeforereachingtheheadofthe
Vågen Bay. The curves formed a shallow bay
between the middle and the southern town areas, and a small promontory made up part of
the southern town area. Holmen was a rather
flatpieceoflandwell-suitedforsettlementand
withthehighestpointabout10masl.TheeasternshoreofVeisanwasmadeupofamorainic
tongue,andsuitablebuildinglandwasfoundas
astripoflandbetweenVeisan,theVågenshorelineandthe15mcontourofthehillFløyfjellet.
Betweenthenorthernandthemiddletownareas
a protruding rock by the Vågen shoreline rose
up 8-9 m and formed a natural topographical
landmark.
55
7EVALUATIONOFTHE
ARCHAEOLOGICALAND
BOTANICALSOURCES
clearlybeidentifiedasamongotherswoodchips,
latrine-,kitchen-andbreweryrefuse,anddung
(Hjelle1986,55).Thelayersmaythusindicate
thatwastewasdumpedintheseainthevicinity
The archaeological and botanical sources will ofthesamplinglocation(cfp51ff).
nowbeevaluatedinordertoidentifyandclassify
thematerialthatcanelucidatetheearlydevelop- Dates
mentofBergen.Theindividualsitesarepresent- A 14C sample from the lowermost layer, layer
edtopographicallyandaregivenasitenumberas 11, was dated with two peaks of probability to
reference. The ‘secular’ sites are discussed with 780-790or810-1000.Layer14was14Cdatedto
areferencetothepresentstreetaddressandthe 1160-1255andlayer23was 14Cdatedto1180museumnumberoftheexcavation.Themonu- 1300.9 This leads to the question whether polmental manifestations are referred to by their lenzone6representsacontinuousdepositionof
contemporarynamesortheirbuilder.Areference materialfromlayer11andonwardsorwhether
forsitenumbers,streetaddress/monumentsand there were breaks in the deposition of material
museum number is given in Table 21 (p 10X). betweenlayers11and14.Thethicknessofthe
Figure22(p10X)presentstheinvestigatedareas layers may perhaps throw some light on this
and monuments. The presentation and evalua- question. The layer series from the bottom of
tionofthearchaeologicalandbotanicalmaterial layer11tothetopoflayer14is50cmthickand,
compriseasfaraspossible,andwhennecessary: accordingtothe14Cdatesavailable,representsa
• an evaluation of the dates applied in the roughly estimated time span of a maximum of
materialandadivisionofthematerialinto 475andaminimumof160years.Accordingto
basic,supplementaryorgeneralbackground Hjellethelowpresenceofphyto-planktoninthe
sourcesandintohorizons1,2,3,4or5
samples implies that the layers accumulated so
• adiscussionofthelocationofasiteoramon- fast that water plants did not have a chance to
ument
flourish (Hjelle 1986, 40). An accumulation of
• anassignmentoflayerswithartefactsoreco- only50cmthroughaperiodofminimum160
factstoartefactcategoriesIorIIaccording yearscannotbecharacterisedasafast,continutotheirhistoryofdeposition
ousaccumulationbutratherimpliesthatthelay• aroughreconstructionofbuildings,passages ersweredepositedinmanysequences.Unfortuandothermajorfeaturesatthelocalities
nately,itisnotpossibletogiveacloserestimate
ofwhenthesinglelayersbetweenlayers11and
Structures and layers assigned to the five hori- 14inpollenzone6weredeposited.
zons are drawn onto the natural topography
A deposition in sequences does not mean,
onmapspresentedinFigures23-27,thereader however,thatthedepositionofwastewasnota
should have these maps at hand when reading relatively recurrent event. What is documented
thechapter.
and dated in pollen zone 6 may be the single
‘bucket’or‘load’ofwaste,whichwasthrownout
at one random location. It is not unreasonable
TheHolmenarea
toassumethatotherlocationswerealsousedas
awastedumpthroughtheyears,thedeposition
Site1,Koengen(Botanicalinvestigationin
ofwastemaythereforehaveoccurredmorefreVeisanbyKariLoeHjelle)(1986)
quentlythanthethicknessanddatesofthepolHjelle performed an analytic pollen investiga- lenzone6layersimply.Theearliestpossibledate
tionofmarinesedimentsandorganicdepositsin ofwastedeposition,representedinpollenzone6,
themedievalVeisaninlet,locatedtothewestof goesbacktobetween780-790orbetween810Holmen.Onlythematerialfrompollenzone6 1000 but the activity may have lasted through
isdatedtotheperiodstudiedhere.Layers11-30, manyyears.
thatispollen-zone6,accumulatedfastandconTheoldestlayerinpollenzone6,represented
tainedbothpollenandmacrofossilswhichcould bylayer11isusedasasourceforhorizon1.Since
56
thematerialisdatedthroughawide 14Cdateit
is used as a supplementary source. The layers
inpollen-zone6containedpollenofweedsthat
mayindicatetheimportofgrain.10Theimports
indicatingpollenarefoundinlayer11aswellas
in later layers (Hjelle 1986, 59). Ecofacts from
layer11areassignedtohorizon1,CategoryII.
I cannot determine if any of the later layers in
pollen-zone 6 represent horizons 3-5 (c 1070-c
1170),sincetheselayershavenotbeendatedin
detail.Thelaterlayersarethereforeomittedfrom
mystudy.
93) also built Christchurch minor (Hkr 18931901, III 226). The church was built in wood
andlaterreplacedbyastonechurch.Thetimber
churchwas,accordingtothesaga,completedin
the reign of Olav Kyrre, no details are known
aboutitslayoutorsize.Thechurchwaslocated
at Holmen in the churchyard of the ChristchurchCathedral(LidénandMagerøy1980,144;
LidénandMagerøy1990).Itmayserveasabasic source for horizons 3, 4 and 5. The material cannot elucidate activity on the site before
horizon3.
Table1.Site1,Koengen(1986)
‘Phase’
Layerslater
than11
Layer11
pollenzone6
Archaeological
evidence
Pottery
Other
Naturalscientificdates
Dendro
TL
Layersolder
thanpollen
zone6
14
Dating
Horizon
Sourcetype
(B/S/G)
C
Peaksof
probability
between
780-790or
810-1000
1
Omittedfrom
thestudy
S
Beyondthe
periodofthe
study
Databasedon(Hjelle1986)
Site2,TheChristchurchCathedral
(StoreKristkirke)
AccordingtoHeimskringla,OlavKyrrestarted
the erection of the Christchurch Cathedral on
Holmen(Hkr1893-1901,III226).In1170the
relicsofStSunnivaweretransferredfromSeljeto
theChristchurchCathedral(Storm1880,151),
this may indicate the formal completion of the
twelfth century cathedral (Lidén and Magerøy
1980,145).GerhardFischerinvestigatedtheremains of the church in 1929. The excavations
showedthatthenavewas21-22mwideandthat
thechurchwasabasilica.Fischersuggestedthat
theoriginallengthofthechurchdidnotextend
57m(LidénandMagerøy1980,147;Lidénand
Magerøy 1990). The church serves as a basic
sourceforhorizons3,4and5.Thechurchyard
ismentionedinthewrittensources(Lidénand
Magerøy1990),buttheboundaryandexactlocationisnotknown.Thematerialcannotelucidateactivityonthesitebeforehorizon3.
Site3,Christchurchminor(LilleKristkirke)
AccordingtoHeimskringla,OlavKyrre(10667Evaluationofthearchaeologicalandbotanicalsources
Site4,TheChurchoftheApostles
(Apostelkirken)
AccordingtoMorkinskinna,thetwelfthcentury
ChurchoftheApostleswasbuiltbyKingØystein Magnusson (1003-1123) (Msk 352, 384).
Presumably the church was a timber church
(LidénandMagerøy1980,137),nodetailsare
known about its layout or size. During excavations at Holmen, in the 1940s, Fischer found
whatheassumedwerere-usedwallboards,roof
tilesandahalf-colonettefromthefirstApostles’
church. Later investigations, however, indicate
thattheremainsdonotstemfromatwelfthcenturychurch.Itisthereforeunlikelythattheystem
fromthefirstChurchoftheApostles(Lidénand
Magerøy1990,36-38).Sincethechurchprobablywasbuiltbetween1103and1123itmayserve
asabasicsourceforbothhorizons4and5.The
churchwaslocatedatHolmen.Fischeralsobelievedtohavefoundthewallofthechurchyard
ofthefirstApostles’church,however,thiswall
doesnotdatebacktothetwelfthcentury,accordingtoanewinvestigationofthedatingmaterial
performedbyDunlop(Dunlop1996a,3.2).The
57
second Apostles’ church was built at the same
siteasthefirstchurch(LidénandMagerøy1980,
137).WallsofthesecondApostles’churchwere
identifiedbyexcavationsinthe1950s(Lidénand
Magerøy1980,137)andshouldthusindicatethe
approximatelocationofthefirstchurchaswell.
Againstthisbackground,thelocationofthefirst
Apostleschurchisusedasabasicsourceforhorizons 4 and 5. The material cannot elucidate
activityonthesitepriortohorizon4.
Site5,ØysteinMagnusson’shallatHolmen
AccordingtothewrittensourcesØysteinMagnusson (1103-1122) built a large timber hall
at Holmen. About 1180 it was still standing,
though it was in a somewhat poor condition,
and it was considered the largest and most renowned timber hall in Norway (MHN 64;
Hkr 1893-1901, 285, 294; Msk 352, 384; Ågr
94;Helle1982,115).Thehallmayaccordingly
beusedasabasicsourceforhorizons4and5.
DuringhisinvestigationsatHolmenFischerbelievedthathehadfoundthehall.Butaccording
to the investigation of the dating material it is
unlikelythattheremainsfoundbyFischerdate
asearlyasthetwelfthcentury(Dunlop1996a,
10), the exact location of Øystein’s hall is still
unknown.Accordingtothewrittensources,the
hallwaslocatedclosetothefirstChurchofthe
Apostles, which in turn was succeeded by the
secondChurchoftheApostles.Sinceweknow
thelocationofthesecondChurchoftheApostleswemaythenalsoknowtheapproximatesite
for Øystein’s Hall. On this basis Øystein’s hall
is located close to the Church of the Apostles.
Thematerialcannotelucidateactivityonthesite
beforehorizon4.
Thenortherntownarea
Site6,Bryggen(1955-1979)BRM0
Theexcavationsatsite6,theBryggensite,was
anopenareainvestigation,coveringabout5700
m2(Herteig1990,9),onlyanareaofabout2000
m2 is, however, relevant for my study. Asbjørn
E Herteig published his stratigraphical analysis
anddatingoftheBryggenmaterialin1990and
1991(Herteig1990;Herteig1991).Herteigdivided the site into four areas connected to his58
toricallyknowntenements:TheGullskogården
area,Søstergården-Engelgården-andBugården
(Figure 7). Chronologically, the site was dividedinto‘periods’,eachterminatedbyafire,and
further subdivided into ‘phases’ (Herteig 1990;
Herteig 1991). Period 1 comprised the oldest
documentedstructuresonthesite.
The stratigraphical analysis presented by
Herteig in 1990 and 1991, serves as my point
of departure when discussing the stratigraphy
of the Bryggen material. New interpretations
basedondendrochronologyand/orstratigraphy
are, however, introduced. As an initial stage in
thepresentstudyIhavere-evaluatedthedating
of the Bryggen material before fire V (Hansen
1998), this serves as the chronological point of
outsethere.InordertoanalysetheoldestperiodsoftheBryggenmaterialmoreclosely,Ihave
madeanupdatedversionoftheso-calledH-post
database(fortheartefactmaterialfromperiods
1-4,period4endsabout1250),containinginformation about the context of artefacts and
otherfindsfromtheexcavation.11Thisupdated
versionservesasabasisinmystudyoftheartefactsfromsite6.
Thebotanicalmaterial
InSøstergårdenbotanicalsampleswereinvestigatedfrommarinedepositspredatingstructures
from period 2 at the Bryggen site (Krzywinski
andKaland1984).
Dates
Thebotanicalmaterialindicatedhumanactivitiesinthevicinityofthesitepriortoperiod2.
Thedepositsweredatedby14C(Krzywinskiand
Kaland 1984), but only the youngest deposit,
‘unit7’isrelevantasasourcefortheperioddiscussedhere.A14Csample12takenfromhazelnut
shells dated the youngest deposit to BP 970+/-
40 BP (calibrated through Stuiver (1982)) and
represents a historical date within the period
1000-1070or1090-1150(onesigma)(KrzywinskiandKaland1984).
KrzywinskiandKalandhaveearlierdiscussed
thedepositfrom‘unit7’.Thecompositionofthe
depositandtheoldestdatesprovidedbythe 14C
samplehavebeenusedasargumentsindatingthe
urbansettlementtothebeginningoftheeleventh
century (Krzywinski and Kaland 1984). I will
Figure7.Site6,Bryggen.
Thedifferentstagesinthe
excavationofthesiteand
namesofthetenements.
(AfterHerteig1990,10
andHerteig1991,12)
discussthedatingframefortheaccumulationof
the‘unit7-deposit’inmoredetail.Thedatefor
theendoftheaccumulationofthedepositshould
be looked into more closely. The profile drawn
up in connection with the botanical investigationistheonlysourcehere,asthearchaeological
documentation does not contain any details on
thestratigraphyinthispartofsite6.13Judgingby
theprofileinFigure8,the‘unit7-deposit’seems
to make up the surface of the building-land
(beacharea)whenperiod2started(inthe1120s
(Hansen1998)),indicatedbycaisson41,dendro
7Evaluationofthearchaeologicalandbotanicalsources
datedto‘after1126’,14andbuiltdirectlyontop
of the deposit. An analogous situation, where
similarcaissonswereconstructed,showsthatthe
caissonswereplaceddirectlyonthebuildingsite
withoutanypriorpreparationofthesurface(Golembnik 1993, Figure 8). In all probability, no
layerswereremovedfromthebuildingsitebefore
caisson 41 was placed on the beach thus makingthe‘unit7-deposit’theyoungestdepositon
thesite.Theterminationofthedepositionofthe
layershouldthereforebedatedtoshortlybefore
1126/thebeginningofperiod2.
59
Thecompositionandthicknessofthedeposit
throws light on the duration of the accumulation.‘Unit7’hadahighcontentofdisintegrated
woodchips,hazelnutshellsandmossesand‘the
compositionofthedepositwasidenticalwiththe
laterwastedepositsofthemedievaltown’(KrzywinskiandKaland1984,33).Thecomposition
points towards a rather quick and/or intensive
accumulation, otherwise the organic componentswouldhavebeenwashedawaybythesea.
Thelowcontentofmarinedinophycea-cystealso
implies a fast accumulation of the layer (KrzywinskiandKaland1984,26).Thecontentofthe
‘unit7-deposit’thusimpliesarelativelyfastaccumulationandthethicknessofthedepositpoints
inthesamedirection.
Sincethecompositionwasidenticalwiththe
laterwastedepositsofthemedievaltown(KrzywinskiandKaland1984)onemaycomparethe
speed of accumulation of such medieval waste
depositswith‘unit7’.The‘unit7-deposit’hada
maximumthicknessofabout50cm(Figure8).
depositwasnotaresultof120-130yearsofrelativelyfastaccumulationbutrathertheresultof
a maximum 20-35 years. A period of 20 to 35
yearsbefore1126/thebeginningofperiod2coincides with a date within the 1090-1150 peak
ofthe 14Cdate,ratherthanwiththe1000-1070
peak,assuggestedbyKrzywinskiandKaland.
If‘unit7’accumulatedduringaperiodof20
to35yearsbeforethe(late)1120s,theaccumulationmayhavestartedbetweenthe1090sandc
1100. In this case the deposit can be used as a
sourceforbothhorizons1and2inSøstergården.
Thestratigraphicalrelationshipbetween‘unit7’
andcaisson41indicatesthatthedeposithadaccumulatedjustbeforethecaissonwasbuilt‘after
1126’,consequently‘unit7’maybeusedasabasicsourceforhorizon4intheSøstergårdenarea.
Whetherthedepositalsodatesbacktothetime
periodcoveredbyhorizon1isbasedonweaker
arguments.Ontheonehand,thecompositionof
thedepositcountsinfavourofafastaccumulation,maybeamaximumofabout20years.On
Figure8.Thestratigraphyofprofile220atsite6,Bryggen.(AfterKrzywinskiandKaland1984Figure3)
In comparison, waste deposits at the Bryggen
site accumulated to a roughly estimated thicknessof70cmintherespectively30and50year
time span between 1170-1198 and 1198-1248
(Herteig1990,Plates1and2).Thisgivesanaverage annual accumulation of 2.3 cm between
1170 and 1198 and 1.4 cm between 1198 and
1248.Ofcourse,suchacomparisonisproblematicdependingonthedegreeofbuildingactivity
and production of waste. Still, the comparison
givesusthenotionthataroughly50cmthick
60
theotherhand,thecomparisonofthicknessof
deposits implies that the deposition could have
lastedupto35yearsandwecannotexcludethat
thedepositactuallydatesbacktothelateeleventh
century.Onthebasisofthis,thedepositwillbe
usedasasupplementarysourceforhorizon3
inSøstergården.Noculture-layersorstructures
weredatedtohorizons2or1,thisinformation
isusedasasupplementarysourceforhorizons
1and2.
Figure9.Thedendrodated
postsinthejettyatsite6and
similarpostsatBorgund,
Sunnmøre.(Borgund1961
Æ7,48,50X/166,70Yand
45,4X/166,4Y;BRM0,Bryggen
PlanO6XIV,Bilag1)
Artefactcategories
Ecofacts from the deposit may be used as categoryIIfindsinhorizons3and4.
Thearchaeologicalmaterial
Period1,localisation,dates
Herteig divided the constructions in period
1 into phases 1.1 and 1.2. In addition a ‘widespreadlayerofsmallstones’laidoutonthebeach
andajettywereassignedgenerallytoperiod1.
Theconstructionsfromphase1.1and1.2were
localisedintheGullskogårdenarea,thejettywas
recordedinSøstergården,andthestonelayerwas
recordedbothintheGullskogårdenareaandin
thewesternmostpartofSøstergården.Thenaturalsubsoilwasprobablyreachedbytheexcavators in these areas. Engelgården and Bugården
werenotthoroughlyexcavatedbelowthelevelof
7Evaluationofthearchaeologicalandbotanicalsources
period2andthenaturalsubsoilwasprobablynot
reachedhereexceptintwotrenches,onebythe
jettyandonebytheabovementioned‘unit7’.
Herteigassumedthatthestructuresassigned
tophase1.1representedmorethanonebuilding
phase(Herteig1990,125;Herteig1991,97)and
newdendrodatescombinedwithstratigraphical
observationssuggestthatanumberofstructures,
assignedtoperiod2,shouldbeconsideredaspart
ofphase1.2.Thiscallsforareconsiderationof
someoftheearlieststructuresfoundatthesite.
ThejettyandassociatedlayersintheSøstergårdenarea
Thetwopostsinthejetty,assignedtoperiod1,
andlocatedintheSøstergårdenarea,weredendrodatedtorespectively‘after1026’and‘after
1029’.15ThesamplesweretakenbyReimersand
myself. We observed a rectangular cut in the
61
lowerpartofeachofthesejetty-posts,cutsthat
couldnotbeexplainedaspartofthejettyconstruction, and thus would imply that the posts
werereusedinthejetty(Hansen1998,93).Afterhavinglookedcloserattheoriginalsitedocumentation,itis,however,clearthatthecutsdid
have a function in the jetty. A horizontal thin
beam(stillinsituwhenexcavated)ranthrough
thetwoholestopreventthepairofpostsfrom
sinkingintothedepositsofthesandybeach.A
similar arrangement has also been documented
injettiesattheBorgundsiteinSunnmøre,Norway(Herteig1975,28,Figure4).Thetwoholes
inthesite6postswereratherlargecomparedto
thehorizontalbeam,asimilardifferenceinthe
proportions was, however, also observed in the
Borgundmaterial(Figure9).Therewerenootherindicationsofreuseontheposts,barkwasnot
preservedbutthesurfaceofthetimberhadnot
beenworked.Afewtreeringsmaystillhavebeen
worn off, the early 1030s thus seems a reliable
dateforthetimber.Thejettyisconsequentlyassignedtohorizon2asabasicsourceforstudyingtheearlyhistoryofBergen.Thepostsfrom
thejettymayhavebeenvisibleabovetheground
(site documentation, profile 184 and plan O6
XIV bilag 1) (when referring to profiles, plans
andgridsinthefollowingreferencesaremadeto
theoriginalsitedocumentationunlessotherwise
stated)untiltheendofHerteig’speriod1,dated
tothe1120s(butwedonotknowifthejettywas
stillinuse),andserveasabasicsourceforhorizons3andhorizon4.
The stratigraphical relationship between a
post in the jetty and the surrounding layers is
recordedintheoriginalsitedocumentation(profile222andplanO6XIV,enclosure1).Profile
222wasa15mlongcross-sectionofthebeach.It
showsthatthreelayersaccumulatedaroundthe
jetty before the post was superposed by period
2constructions.Theselayerscompriselayer‘1’:
a roughly 10 cm thick ‘yellow grey sand layer’,
deposited on top if this was layer ‘2’: about 30
cm thick and consisting of ‘yellow grey sand
and pebbles’. On the top of here was layer ‘3’:
a roughly 15 cm thick ‘dark brown layer with
pebbles,shellsandwoodwaste’.Iinterpretlayer
3aspartofthe‘widespreadlayerofsmallstones
spreadonthebeach’,generallyassignedtoperiod
1(plansP6XII,O6XI,N6XIV)andassigned
62
bymetohorizon4asabasicsource(cfbelow).
Nostructureswereindicatedinlayers1and2,
implying that the jetty was the only constructioninthisareabeforelayer3waslaidout.The
description of layer 1 as well sorted, indicates
thatnaturalforcesmayhavedepositedit.Layer
2,wasnotwellsortedandprobablyreflectshumanactivitiesonthebeachoftheSøstergården
area.Asthestonelayer(3)maybeassignedasa
basicsourceforhorizon4thelayersprecedingit
must have been deposited after the early 1030s
butbeforetheperiodrepresentedbyhorizon4,
thatisintheperiodsrepresentedbyhorizons2
and3.Thereisnofirmevidencetonarrowdown
thedateofthelayerdepositions,butitmaybe
argued that some time went by from when the
constructionofthejettytookplace‘after1029’
-intheearly1030s-andlayer2wasdeposited
onthebeach,enoughtimeforlayer1toaccumulate.Thusitismorelikelythatlayer2represents
horizon3ratherthanhorizon2.Thuslayer2is
accordinglyassignedtohorizon3.Asthedating
ofthematerialisnotbasedonsolidevidenceit
willbeusedasasupplementarysourceonly.No
structures or culture-layers predating horizon 2
havebeendocumentedintheSøstergårdenarea.
This information is used as a supplementary
sourceforhorizon1.
Thestratigraphicalrelationshipbetweengroupsof
structuresassociatedwithperiod1intheGullskogården
area
In the Gullskogården area constructions that
havebeenassignedtophase1.1inperiod1consist of two palisade-built fences.16 Anticipating
events a bit I shall label these respectively 6/C
and6/B.17Twenty-onescatteredposts,someof
themwithinfence6/Cwere‘witheveryreservation’interpretedasabuildingbyHerteig(building 497). Some of the posts outside fence 6/C
were interpreted as traces of a cellar building
(Herteig1991,97).18Thestructuresassignedto
phase1.1mayrepresentmorethanonebuilding
phase.Ihavetentativelyseparatedthestructures
intoanolderandayoungerlevelofstructuresby
astratigraphicalanalysis.
According to the stratigraphy (profile 69/3)
two layers accumulated over fence 6/C when
it went out of use: first a 15 cm thick layer of
‘brown fill-masses’, then an about 20 cm thick
layerof‘lightbrownfinesandandgravel’.When
comparingtheprofilewithsiteplans(Q3X,XI,
andXII),thesandandgravellayerseemstohave
beendepositedoveratleasttwopostsinsidethe
fence as well,19 implying that these were older
thanthesandandgravellayer.Atleastninepost
holesfromphase1.1(somestillwithpostsintact)
hadbeendugthroughthelightbrownfinesand
andgravellayer.20Theyprobablybelongedtoa
building and must be later than the sand and
gravellayer.Thefence6/Candatleasttwoposts
mustbelongtoaphaseolderthanthenineposts.
Asthefencewascoveredwith‘brownfill-masses’
beforethesandandgravellayerwasdeposited,it
islikelythatthefencewasoutofusewhenthe
ninepostswereerected.Thusthephase1.1structuresmaybesubdividedintophase1.1.1:fence
6/Candassociatedstructures,andphase1.1.2:
the nine posts. The nine posts are tentatively
reconstructedasapost-builtbuilding.Theconstructionofthis‘9-postbuilding’inphase1.1.2
wassucceededbythedepositionofacoarsegravellayerintheareasouthofthebuilding(towards
thebeach).Building45datingtophase1.2,was
mostlikelybuiltontopofthegravelsinceone
ofthepostsfromthebuilding,cutthroughthe
coarselayer(profiles69and31).Consequently,
the ‘9-post building’ must have been built beforebuilding45.Thelimiteddocumentationof
the stratigraphy makes it difficult to determine
whether ‘the 9-post building’ went out of use
before phase 1.2 or if the structure still existed
in phase 1.2 contemporaneously with building
45.Oneofthepostsin‘the9-postbuilding’was
scorchedbyfire,apossibleindicationthatitwas
burntinfireVIII(the1120s(Hansen1998)),togetherwithbuilding45.This,ratherweak,evidenceistakenasanindicationthat‘the9-post
building’ from phase 1.1.2 lasted until the end
ofperiod1andthusexistedcontemporaneously
withbuilding45forsomeyears.
Thepossiblecellarbuildingandthescattered
posts in the area south of the ‘9-post building’
mustclearlyhavebeendemolishedbeforebuilding45waserectedinphase1.2,astheywerecoveredbythe‘wide-spreadlayerofstoneslaidon
thebeach’,thatwasprobablycontemporarywith
building 45 (see below). A reused timber log21
frombuilding45inphase1.2canbeinterpreted
asastaveoriginallybelongingtoacellarbuild7Evaluationofthearchaeologicalandbotanicalsources
ing (Reimers in prep). Could this stave originally have come from the possible cellar buildingthatmustpredatebuilding45accordingto
thestratigraphicalevidence?Thestratigraphical
evidence does not contradict such a suggestion
anditseemsthatthestavesinthepossiblecellar
buildingwerepulledupwhenthisstructurewas
demolished as only post holes could be traced
at the site. Anyhow the reused cellar building
stave from building 45 implies that the Gullskoenareacontainedmorethanjust‘the9-post
building’ before phase 1.2. The cellar building
may originally have belonged either to phase
1.1.1 or 1.1.2, probably the latter, if the reused
timber stems from the possible cellar building.
A couple of factors speak in favour of such an
interpretation: The reused cellar building stave
wasinsuchgoodshapethatitcouldbereused
whereastheremainsofthefencefrom1.1.1were
mostlylefttorot.Therewasalsomorethanone
levelofstructures,predatingbuilding45,inthe
areawherethepossiblecellarbuildingwaslocated.Andif-again-thereusedtimberinbuilding45stemsfromthepossiblecellarbuildingI
wouldpresumethatthebuildingbelongedtothe
youngestlevelofstructureshereandnottothe
oldest.Accordingly,Ifinditmostplausiblethat
thepossiblecellarbuildingandassociatedposts
belongtophase1.1.2andthattheoldestlevelof
scattered posts in the area south of ‘the 9-post
building’belongtophase1.1.1.
Howistheperiod1materialdated?
The material from phase 1.2 is well-dated and
will serve as a point of departure when dating
theolderphasesatsite6.Themainconstruction
inphase1.2wasbuilding45,well-datedthrough
dendrochronology, built c 1110 and later destroyedinafiredatedtothe1120s(Hansen1998,
123), thus dating phase 1.2 to between c 1110
and the 1120s. Another feature was ‘the widespread layer of small stones laid on the beach’.
This layer was recorded in the areas of Gullskogården, Søstergården and the northern part
ofEngelgården(Herteig1991,111).Thestones
weredepositedonthebeachinseveralsequences
(grids Q3 and P3). In the Gullskogården area
the layer was later than the scattered posts assignedtophase1.1.2(theassumedcellarbuilding) but most likely contemporary with build63
ing45.Structuresandlayersfromphase1.2may
serveasbasicsourcesforhorizon4.
Itismoreproblematictodatethebeginning
andendofphases1.1.1and1.1.2.Timberfrom
‘the 9-post building’ (1.1.2) was dated through
dendrochronology to ‘after 1069’. The sample
was characterised as being of good quality by
TerjeThun.Butwecannotdetermineiftreerings
weremissingorifthepostwasreused(Hansen
1998,93).22Thelatterisalwaysapossibilitythat
shouldbeconsideredwhendatingthroughasole
dendrosample.Ifsometreeringsweremissing
thesamplewouldstilldatethefellingyearofthe
timberto‘relativelyshortlyafter1069’.Butsince
wedonotknowifthesamplecamefromreused
woodornot,therearetwopossiblescenarios.(1)
Ifthepostwasnotreusedandhadalltreerings
intact-ormissedjustafew,thesamplewould
date the beginning of phase 1.1.2 (the possible cellar building and ‘the 9-post building) to
‘relativelyshortlyafter1069’:(horizon3).This
wouldprobablydatetheendofphase1.1.1with
fence6/Candassociatedpoststobeforethelast
quarteroftheeleventhcenturyandthussuggests
thatthisphasebegansome25-50years(cfp60)
earlierinthesecondquarteroftheeleventhcentury(horizon2).(2)Ifthepostwasreusedonce,
weshouldadd25to50yearstothe1069date
andthiswoulddatethebeginningofphase1.1.2
to the beginning of the twelfth century (horizon4).Thisinturnwoulddatetheendofphase
1.1.1totheendoftheeleventh/thebeginning
ofthetwelfthcenturyanddatethebeginningof
thisphasetentativelyto25-50yearsbefore:the
lastquarteroftheeleventhcentury(horizon3).
Isitpossibletofindsupportforeitherofthese
scenarios in other material from the Bryggen
site?Followingthesecondscenario,thesuggestedcellarbuildinginphase1.1.2wouldhaveto
be‘crammed’intohorizon4:itwouldhavetobe
builtabout1100andabandonedbeforethewidespreadlayerofsmallstoneswaslaidoutcontemporaneouslywiththeconstructionofbuilding45
‘after1110’(Thelatterperhapsreusingmaterial
fromthecellarbuilding).Thetimedepthinthe
sourcesdiscernedthroughthestratigraphicalrelationship between the possible cellar building
andstructuresassignedtohorizon4(the‘widespread layer of small stones’ and building 45)
thusfavoursscenarioIasthemostrealistic.
64
The jetty, dated to the early 1030s, was located about 17 m south-west of fence 6/C and
associated posts. If we date the beginning of
phase1.1.1accordingtothefirstscenariotothe
secondquarteroftheeleventhcenturythisdate
correspondswellwiththedateforthejetty,the
jettyandthefencewithassociatedpostswould
be contemporary. If we zoom out and have a
lookatthespatialrelationbetweenfences6/B,
6/CandthejettyvisualisedinFigure24,apatternemerges:fences6/Band6/Cmakeupthe
boundariesoftwoplots(cffootnote17).Ifhypotheticallyathirdplot‘6/D’,ofthesamewidth
asplot6/C,waslocatedtotheeastofplot6/C,
the jetty would run straight up to the easternmostboundaryofthethirdplot,connectingthe
jettytotheplot.Ifthehypotheticalplotwasreal
this would be an indication that the jetty was
contemporary with the palisade fence bounded
plotsandviceversa.Ishallreturntothispoint
againinpages155ffand183ff.Ifwefollowthe
secondscenariothejettywouldnotbeassociated
withanyknownstructuresinthenortherntown
area. When considering these circumstances I
find that the dating suggested in the first scenario seems more plausible at the present state
ofresearch.
Myconclusionisthusthatphase1.1.1(fences
6/Band6/Candassociatedstructures)mayhave
startedduringthesecondquarteroftheeleventh
century and it probably ended before the last
quarteroftheeleventhcentury,thusservingas
asourceforhorizon2.Phase1.1.2(‘the9-postbuilding’andthepossiblecellarbuilding)most
likely began ‘after 1069’, the 9-post building
possiblylasteduntilthe1120sandthusservesas
asourceforhorizons3andhorizon4.Thepossiblecellarbuildingwasdemolishedbeforephase
1.2beganandthusservesasasourceforhorizon
3only.Sincethematerialfromphases1.1.1and
1.1.2istentativelydateditcanonlybeusedasa
supplementarysource.Thequestionofthehypotheticalplotisasalreadymentionedresumed
onabroaderbasisinpages155ffand183ff,and
an attempt is made to strengthen the proposed
dates.Therewerenotracesofstructuresorculture-layersbelowphase1.1.1,thisinformationis
usedasasupplementarysourceforhorizon1.
Intheareawestoffence6/C,no insitustructures,exceptfence6/B,weredocumentedprior
to period 2, however reused timbers (dated to
‘after1024and‘after1040’),foundonplot6/B
inhorizon5(cfbelowandChapter9)maystem
from activities on plot 6/B (cf p 65ff) this informationisusedasasupplementarysourcefor
horizon2.
Structuresassignedtoperiod2reconsidered
Acomplexofstructuresconsistingofsix2mx2
mstone-filledcaissonsthatformedthefoundationforapassageandaquayfronthasbeenassignedtoperiod2andassociatedwithamongst
others,building502inthisperiod(Herteig1991,
Plate 14). Three of the caissons were, however,
dendro dated to respectively ‘after 1104’, ‘after
1106’ and ‘after 1108/09’ (caissons 29, 27 and
28)(Hansen1998,Table2),implyingthatthey
shouldratherbelongtophase1.2associatedwith
among others building 45 dated to ‘after 1110’
(cfabove).Thesampledpiecesofwoodshowed
no signs of reuse, which strengthens the reliabilityofthedates.Thestratigraphicalrelation
betweencaisson29andbuildings502and45respectivelysupportstheassociationofthecaissons
withbuilding45fromperiod1.2.Accordingto
plan O3 XI, caisson 29 was built on the same
levelintheterrainasbuilding45,whereasbuilding502,succeedingbuilding45,musthavebeen
builtontopoffill-massesthathadaccumulated
ontocaisson29(planO3,X)(Figure10).Isuggestthatcaissons27,28,29,30,31,and32that
formedapassageandaquayfrontshouldallbe
assignedtoperiod1.2.Theymayserveasabasic
source for horizon 4. As they were still in use
throughperiod2theyalsoserveasabasicsource
forhorizon5.Apostfrombuilding45wasreusedwhenthenorthernpartofthepassagewas
repairedinperiod2(cfHerteig1991,94ff).
Building 66 in Gullskogården has been assigned to period 2 (Herteig 1991, 87ff). Three
dendro samples from this building were dated
to respectively ‘after 1024, ‘after 1040’ and ‘after1127’23.Thefirsttwosamplesweretakenin
1997/98 and produced dates that were much
earlierthanexpectedforbuilding66,beingassignedtoperiod2.Thethirdsamplewasthereforetakenin1999.Thissamplegaveayounger
datethatplacesthebuildingsafelywithinperiod
2.Thetwooldersamplesweretakenfromposts
whereonlythebottompartwaspreserved,and
7Evaluationofthearchaeologicalandbotanicalsources
theissueofreusecouldnotbedecidedupon.The
samplingspotsdidnothavesignsofhavingbeen
workedup.However,accordingtoThun’sevaluationofthesamplesbothmaymisssometreerings,sosomeyearsshouldbeaddedtothedate
oftheoutermosttreering.Thefactthattherewas
alargedifferenceintheageofthedatedtimbers
suggests that the two older samples stem from
reusedwood.Evenifsometreeringsweremissingthetimbersstilloughttostemfromthemiddle quarters of the eleventh century. This may
indicatethattherewasactivityintheareawhere
building66waslaterbuilt.Thisinformationhas
alreadybeenassignedasasupplementarysource
forhorizon2above.
TheBryggensitefromhorizon1to4:majorfeatures,
artefactcategories
Nostructuresorculture-layerscouldbeassigned
to horizon 1. A jetty and several constructions
were assigned to horizon 2, these are: fence 6/
C built in palisade technique (assigned by me
to phase 1.1.1) this fence makes up the north,
westandsouthboundariesofplot6/C.Another
fenceindicatesaplot6/Btothewestof6/C,no
insitustructureshavebeenassociatedwiththis
plot, however reused timbers from building 66
assigned to period 2 may stem from activities
here.Iftherewasalsoaplottotheeastof6/C,
thejettywouldrunstraighttowardstheeastern
cornerofthisplot,providingtheplotwasofthe
samewidthasplot6/C.Thishypotheticalplot
is labelled 6/D and it is on a preliminary basis
assignedtohorizon2alongwith6/Band6/C.I
willreturntothefactualexistenceofthisploton
abroaderbasisinpages183ff.Twopostswithin
plot6/Cmaybecontemporarywiththeboundaryindicatingfence,asmaysomeofthescattered
posts south of fence 6/C. It is not possible to
determinewhatkindofconstructionstheposts
were part of. No artefacts have been associated
withhorizon2.
Duringhorizon3,thejettyintheSøstergården
area was still visible, but we cannot determine
whetheritwasstillinuse.A30cmthicksand
andgravellayerwasdepositedonthebeachby
thejetty.IntheGullskogårdenarea,nineposts
mayrepresentabuildingherecalledthe‘9-post
building’.Apossiblecellarbuildingandassociatedpostshavebeenassignedtohorizon3.Before
65
Figure10.
PlanO03XandO03XI,
site6,Bryggen
66
the ‘9-post building’ was erected, the site was
prepared by depositing sand and gravel, fences
6/Band6/Cfromphase1.1.1werethuscovered
leavingnovisibletracesofplotboundariesonthe
siteinhorizson3.Noartefactshavebeenassociatedwithhorizon3.
In horizon 4, the ‘9-post building’ was perhapsstillinuse,whilethepossiblecellarbuilding had been demolished. A gravel layer was
filledintopreparethebuildinglandfortheconstructionofbuilding45fromphase1.2,andthe
buildingwasnowerected.Layersofsmallstones
were then spread over the beach to consolidate
theground.Arowofcaissonsthatservedasthe
foundation of a passage and a quay front were
also constructed. Fire-layer VIII (with its contentsofartefacts)istheonlylayerthathasbeen
ascribedtophase1.2inthesitedocumentation
ofartefactcontexts,however,othercontexts,and
thusartefacts,havealsobeenconnectedtohorizon4ascategoryIandIIfindsaccordingtothe
criteriaoutlinedearlier.
areaintheyearsbeforefireVII.Thismaterialis
thereforeplacedinhorizon5,andcanbeused
asabasicsource.
Majorfeatures,artefactcategories,period2
Constructionsfromphase2.2,assignedtohorizon5,include25buildings,33caissons,anumber
ofpitsandmooringposts.Quayfrontsandfive
passageswithoutnumbersintheoriginaldocumentationarereconstructedonthebasisofthe
caissonsthatmeasured2mx2m.Thestructures
formatleast8rowsofbuildingsextendingdown
the morainic slope to the waterfront. Between
the buildings, the passages provide access from
the quay front to the buildings. The built-up
areaisreconstructedaccordingtoHerteig(1990,
1991)andMoldung(2000).Onlyfire-layerVII
hasbeenascribedtophase2.2inthedocumentationofartefactcontexts,butartefactsfromother
contexts can also be connected to phase 2.2 as
categoryIandIIfindsaccordingtothecriteria
outlinedearlier.
Period2,dates,location
Period2attheBryggensiteconsistsoftwophases:2.1and2.2.Phase2.2isdefinedasthe‘level’
of structures that burnt in fire VII (1170/71).
Onmostofthesiteonlyonephaseofstructures
wasfound.Themajorpartofthestructuresfrom
phase 2.2 therefore make up the first and only
‘level’ofstructuresatthesiteinperiod2.Ina
few places, however, the structures from phase
2.2wereprecededbyan‘unburnt’levelofbuildings, which make up phase 2.1 (Herteig 1990;
Herteig 1991). Structures from period 2 were
recorded in the Gullskogården, Søstergården,
Engelgården and Bugården areas. The northernmostpartofEngelgården,andthesouthernmostpartofBugårdenwere,however,notdocumentedindetail.Period2iswell-datedthrough
both dendrochronology and pottery. The main
partofthestructuresfromphase2.2wasunder
constructionfromthe1120suntilthefirstpart
of the 1130s. The buildings representing the
‘secondgeneration’ofstructuresperiod2,were
underconstructionfromthelate1130sandinto
the1150s(theyoungestdendrosamplefromperiod2wasfrom1149).Period2endedinafire
datedto1170/71(Hansen1998).Thematerial
definedwithinphase2.2representsthebuilt-up
7Evaluationofthearchaeologicalandbotanicalsources
Site7,ØvreDreggsalmenningen(1989)BRM
298
The excavation at site 7, Øvre Dreggsalmenningen, about 35 m northwest of site 6, was a
trenchsurveywhere14profileswereinvestigated
andanalysed(Dunlop1989e).Botanicalmaterialhasalsobeenstudiedinconnectionwiththe
investigation(Hjelle1989).
Dates
The oldest material above the natural deposits
wasnotdated.Thepresenceofpollenanda 14C
datecan,however,beofhelpindatingthematerial.Theoldeststratainprofiles1,2,3,5and8
canbedividedintothreephases:layer19/20representsphase1,beingtheoldestlayerabovethe
naturalsubsoil(layer21).Phase2isrepresented
by layer 18, and building K19, a pit-house, destroyedinafirerecordedaslayer17/31.Levelling
layer30representsphase3.
Thepresenceofpollenfromcentaureacyanus
indicates that phase 1 should be dated to the
ninthcenturyorlater.A14Csample24fromcharcoalinthephase2fire-layer17/31isdatedwith
thehighestprobabilitytobetween860and1020
(Figure 11). The sample does not provide the
67
Table2.Site6,Bryggen(1955-1979)BRM0
Archaeologicalevidence
Buildingphase/
Sitearea
2.2
Gullskogården,
Søstergården,
Engelgården,
Bugården
2.1
Youngestpottery
Other
typespresent
LondonCoarse,
LondonFineEarly
Style,DevStamford,
FrenchType,Near
Stoneware
1.2
Gullskogården
Søstergården,
Engelgården
1.1.2
Gullskogården,
Søstergården
1.1.1
Gullskogården
Paffrath
Thejetty,
Søstergården
Reusedpostsin
building66
‘Unit7’
Søstergården
Lackofculturelayersorstructures
Naturalscientificdates
Dendro
14
Dating
Horizon Source
type
(B/S/G)
C
1127
(1135-1149)
Ends1170/71
1100
(1104-1129)
1100
(1107-1110)
Begins1120s
Endsc1135-1140
Begins‘after1110’
Ends1120s
1069
Activity‘after1069’ 3-4
Ends1120s
Stratigraphical
relationshipto1.1.2
5
B
4
B
Begins2ndquarter 2
ofeleventh
century
Endsbeforec1069
(1026)1029
Activityfromthe 2-4
early1030s
‘after1024’
Activityinthe
2
‘after1040’
middleofthe11th
century
Thicknessofdeposit,
1000-1070
3or4
stratigraphical
or
relationtocaisson41
1090-1150
Belowunit7
1and2
S
S
B
S
S
S
DatabasedonHansen1998.Datesinboldaretheyoungestintheconstruction/phase
bestdatingevidencesincewedonotknowmore
preciselyfromwhereitderived,thesamplemay
alsohavehadan‘oldwoodproblem’,thusprovidingatooearlydate.Stillitmayprovideawide
postquemdatefortheestablishmentofphase2or
ofactivitiesduringthephase.Accordingtothis,
thesampleindicatesthatbuildingK19oractivities associated with the building dates back to
theeleventhcenturyorearlier.Thethirdphase:
levellinglayer30containednodatingevidence.
Thecultivationactivitiesrepresentedinphase
1mayhavetakenplaceintheninthcenturyor
laterandceasedwhenthesettlementrepresented
by phase 2, building K19 and associated layers
wasestablished.Itisnotpossibletodetermine,
whenphase2wasestablishedandphase1ceased
onthebasisofthematerialfromsite7alone.
Ifwelookatsitesintheclosevicinityofsite7,
patternsemergethatmaybeofhelpplacingphase
2atsite7intomychronologicalframework.At
68
site 11, Dreggsalmenning 20 (cf p 110ff) two
palisade-builtfencesmakeuptheboundariesfor
threeplotsthatranperpendiculartotheVeisan
shore. The plots and fences are tentatively assigned to horizon 2 as supplementary sources
becausetheywerequitelikelyconstructedcontemporaneously with identical fences and plots
onsite6Bryggen.Thefences/plotsonsite6are
tentativelyassignedtohorizon2assupplementarysourcesthroughverticalandhorizontallinks
tobetterdatedsources,thefencesandplotsare
thusnotdateddirectly(cfp89ff).Theorientation of the pit-house corresponds well with the
orientationoftheplotsatsite11(cfFigure24).
Itisthereforereasonabletosuggestthatbuilding
K19andassociatedlayersmaybepartofasettlementgenerallyassociatedwiththeplots.When
choosingtheyoungestpeakofprobabilityforthe
14
Cdatefromlayer17/31thedateofabout1020
supports an assignment of the pit-house or as-
Figure11.14Csamplefrom
layer17/31site7,Øvre
Dreggsalmenningen
Figure12.14Csample
fromphase9/10site8,
Dreggsalmenningen14-16
sociatedactivitiestohorizon2.Andthe14Cdate
infactindirectlysupportsthattheregulationof
thepalisade-boundedplotsshouldbeassociated
with horizon 2 rather than with younger horizons.Basedonthecorrespondingorientationof
thepit-houseandplotsonsite11andtheyoung7Evaluationofthearchaeologicalandbotanicalsources
estpeakofprobabilityforthe 14Csample,Iassignthepit-houseinphase2tohorizon2.
It is not possible to date fire-layer 17/31 itself, the fire may in theory have occurred immediatelyaftertheconstructionofthebuilding
oritmayhaveoccurredmuchlater,sothedate
69
oftheendofthesettlementphaseisdifficultto
determinemoreprecisely.Stillitishardlylikely
thatapit-housewasinuseformorethan25-50
years,whichistheestimated‘lifetime’fortimberstructure(cfp60ff).Ithereforeassumethat
thebuildingwasnotinuseintheperiodcovered
byhorizon3.
Tosummarise,itisproposedthatactivityin
the area around site 7, represented by layer 20,
began in the ninth century or later and ceased
beforethefirstdecadesoftheeleventhcentury,
whenthesettlementrepresentedbybuildingK19
and associated layers was established. The settlementtracesprobablydonotdatetothetime
spancoveredbyhorizon3.Onthisbackground,
layer20isassignedtohorizon1andthesettle-
broaderbasisinChapter8.Layer19/20K19in
profile 2 was interpreted as a pit-house. Layers
17/31,and18,and30inprofiles1,2,3,5and8
wereassociatedwiththeconstruction.Thelayers
areassignedtohorizon2categoryI(17/31)and
categoryII(18,30).
Site8,Dreggsalmenningen14-16(1986and
1990)BRM237
Theexcavationatsite8,Dreggsalmenningen1416locatedabout15m,tothewestofsite7was
carriedoutintwosteps.In1986about550m2
wereexcavated,in1990about80m2,however,
fortheoldestphasesundisturbedculture-layers
onlymadeupabout40m2.Thenaturalsubsoil
wasreachedattheexcavation.25
Table3.Site7,ØvreDreggsalmenningen(1989)BRM298
Archaeologicalevidence
Profile/
Pottery
Strata
1,2,3,5,8/
K19,Layers
17/31,18
1,8/Layer
20
Other
Naturalscientificdates
Dendro
TL
Coherencewith
horizon2material
fromtheBryggen
site?
Presenceofpollen
fromcentaurea
cyanus
14
Dating
Horizon
Sourcetype
(B/S/G)
C
Highest
Begins2ndquarterof
probability eleventhcentury
between Ends?
820and
1020.
Betweentheninth
centuryandthe2nd
quarterofeleventh
century
2
S
1
S
Databasedon(Dunlop1989e;Hjelle1989)
mentassociatedwithbuildingK19andlayers18,
17/31 and 30 is assigned to horizon 2. As the
materialisdatedby14Candcentaureacyanusand
thedateforthetransitionfromphase1tophase
2cannotbesecurelyestablishedthematerialwill
beusedasasupplementarysource.Thematerial cannot elucidate activities on the site after
horizon2.
Generallanduse,artefactcategories
Dunlopinterpretslayer19/20,inprofiles1and
8 as a cultivation layer (Dunlop 1989e) (cf the
broad definition of cultivation p 67). Pollen in
the layer indicates meadow vegetation on the
sampling location, pollen of barley (Hordeum)
andwheat(Triticum)indicatecultivationorsettlementinthevicinityandthepresenceofcentaurea cyanus (Hjelle 1989) may signify household waste from a settlement in the vicinity. I
willdiscussthelocationofthissettlementona
70
Dates
Datesfromsite8arebaseduponpottery,14Cand
TL.Golembniksuggeststhatphase10/1986and
phase9/1990(hereafter10/9)endedinthemiddle of the twelfth century. The younger phase
9/8endsinthesecondhalfofthetwelfthcentury and phase 8/7 lasts until the end of the
twelfthcentury.Adateforthebeginningofthe
oldestsettlementphasehasnotbeenproposed.I
willattempttodatethebeginningofactivityat
thesite.
Phase10/9wasdestroyedbyafireanddated
by 14CandTL.The 14Cdate(1030-1190)was
takenfromcharcoalthatprobablyderivedfrom
buildingmaterial(Figure12),andtheTLdate
(1190+/-40)derivedfrommaterialburntduring
thefire.The14Cdateshouldthusprovideawide
dateforthebeginningofthephaseandconstructionofthebuilding.TheTLdateprovidesawide
datingframeforthetimeofthefire.26
Table4.Site8,Dreggsalmenningen14-16(1986/90)BRM237
Archaeologicalevidence
Phase
1986/9
1990/8
Pottery
Other
Andenne,Paffrath,Soft
FiredBlackware,North
French
NaturalScientificdates
Dendro
1986/10
1990/9
Thenaturalsubsoil
TL
1190
+/-40
14
C
11601270
10301190
Dating
Horizon
Source
type
(B/S/G)
Beginsc1150 5
endsc1170
B
Beginsc1100 4
endsc1150
Priorto
1-3
phase1986/10
1990/9
S
S
DatabasedonGolembnik1986and1990
Whenusingtheoldestpossibledateprovided
theTLdateindicatesthatthefireoccurredafter
c1150.Adateoftheendofphase10/9tosometime around ‘after 1150’ is not in conflict with
thedatesuggestedbyGolembnikfortheendof
the succeeding phase 9/8 to the second half of
thetwelfthcentury.The14Cdategivestoowide
aframetoelucidatethequestionogthedateof
thebeginningofphase10/9.
Before the building in phase 10/9 was constructed,thebuildingsitewasdevelopedthrough
‘largerpreparationofanewbuildingspacealong
thesandyembankment’(whichoriginallymade
upthesite)(Golembnikinprep-a,8).Whenthe
building from phase 10/9 burnt down, ‘serious
earthwork’wascarriedoutalloverthesiteand
newbuildingswerenotbuiltonthesameplace
asintheprecedingphase(Golembnikinprepa, 9). The changed layout of the site from the
oldestphasetothesecondphaseonthesitemay
perhaps indicate that phase 10/9 lasted quite
sometimebeforeitwasstruckbyfire.Ifweestimatethatatimberbuildinglasted25-50years
ifnotstruckbyfire,amaximumperiodofabout
50yearsforphase10/9couldbeexpected(cfp
60ff).Theoretically,thephasemayhavestarted
asearlyasabout1100.Withasuggestedc1100c1150date,phase10/9doesnotfitdirectlyinto
the scheme of horizons proposed here. Accordingtotheprinciplesoutlinedabove(p68ff)The
structuresinthephaseshouldbeusedasasource
for horizon 4 and since the phase is not welldateditisusedasasupplementarysourceonly.
Theartefactsfoundinphase10/9areallfound
infire-layer20andaretooyoungtobeusedas
sources for activity in horizon 4. The artefact
material should, therefore, represent horizon 5
7Evaluationofthearchaeologicalandbotanicalsources
materialasCategoryIIfinds.Sincethefiremust
haveoccurredbetween1150andthesecondhalf
ofthetwelfthcentury,theartefactmaterialcan
be used as a basic source. The lack of culturelayersorstructuresassociatedwithhorizons31 is used as a supplementary source for these
horizons.
Thedatefortheendofphase9/8to‘thesecond
halfofthetwelfthcentury’alsoseemsplausible
accordingtotheceramicevidence.Thisgivesan
approximatedatingframeforphase9/8to‘after
c 1150-c 1170’. This dating frame places phase
9/8 in horizon 5, as the phase is well-dated it
canbeusedasabasicsource.
The botanical investigations in connection
with the excavation (Hjelle undated) confirm
Golembnik’sinterpretationthatthespacealong
thesandyembankmentwaspreparedthoroughly
beforeconstructionworkbeganinphase10/9.
Majorfeatures,artefactcategories
Noculture-layersorstructurescouldbeassociatedwithhorizons1-3.Inhorizon4/phase10/9,
site8waslocatedbetween2.5and4.5maslon
thewesternandnorthernsidesofthemorainic
tongueeastoftheVeisaninlet.Preparationand
levellingofnewbuildinglandprecededtheconstructionofbuildingsatthesite.Building158,
located on the western and lowest part of the
site,wastheonlypreservedstructureinhorizon
4(Golembnikinprep-a).Onthenorthernside
ofthesite,wheretheterrainslopedtowardsthe
north as well as towards Veisan, layer 684 was
recorded and interpreted as the remains of occasionalratherthanofcontinuousactivity(Golembnik in prep-b). On the remaining part of
thesite,thelevellingconnectedwithlaterphases
71
disturbed the phase 10/9 surface. Three layers
havebeenassignedtothephase,findsfromthese
layersareattributedtohorizon5ascategoryII
finds.
Inthefollowinghorizon5/phase9/8,theremains of two buildings, K166, K145/152/157,
and a pavement K144/154 were found in the
western part of the site. In the middle of the
site,fourposts,K136,wereuncovered.Aninterpretationofthefunctionofthesepostshasnot
beensuggested(Golembnikinprep-a),buttheir
presenceshowsthattheareawasbuilt-upinthis
phase.Intheeasternpartofthesitetheremains
of a building K102/104 were uncovered. The
northernpartofthesitewasopen(Golembnik
inprep-b).Twenty-threelayershavebeenassociatedwithhorizon5/phase9/8,elevenlayersare
assignedtoartefactcategoryIandtwelveareassignedtocategoryII.
Dates
I will now evaluate the proposed dates for the
lowest/oldest structures at site 9, through the
artefactmaterial.Thefencethatdividesthesite
intotwoparts,building8andwell1northofthe
fence,buildings10,11,12,13,aswellascaissons
1 and 2 south of the fence are discussed. Furthermore,building14andanumberofscattered
structures north of the fence are discussed.27
Figure14showsthestratigraphicalrelationship
betweenstructuresatthesite.Asprimarilyverticalrelationsbetweenthestructureshavebeen
documentedandonlyafewhorizontalconnectionscanbemadebetweentheverticalgroups,
itisimpossibletodeterminewhichstructuresare
contemporarybasedonthestructuresalone.To
getabetterunderstandingoftherelativeandabsolute dating, I have studied pottery and shoes
from contexts connected to the structures. An
account of the collection and dating of the arSite9,Sandbrugaten5(1967)BRM3
tefact assemblages is given in Appendix 3. The
Theexcavationatsite9,Sandbrugaten5,located datescanonlybetakenaspostquemdatessince
about 11 m to the southwest of site 8, was an wedonotknowifthefindsderivefrominsituor
open area investigation, which covered about redepositedlayers.
480m2.Structuresfromthesitehavebeenpresentedinanarchivereportbyexcavationsuper- Structuresnorthofthefence
visorArneJLarsen(Larsen1967a).Inthereport, Building 8 and well 1 are the oldest ’regular’
abriefstratigraphicalanalysisofthestructuresis buildings/structures,northofthefence.Howevpresented.Noabsolutechronologyhasbeensug- er,underneaththebuildingandintheareanorth
gested for the structures and the artefacts have of the fence other structures, were also found,
notbeenanalysed.
whichhavenotbeenaccountedforinthereport.
Larsenassumedthatstructuresbuiltdirectly Building8andwell1arecontemporary.Pottery
onordugintothesterilemoraine,weretheoldest inartefactassemblage2,depositedunderbuildonthesite.Thesestructurescomprisebuilding8 ing8,gaveapostquemdateofbuilding8andthe
andwell1,fourbuildings:10,11,12and13-,a welltoafterc1225.Potteryfromassemblage1
16mlongpalisade-builtfence(Figure13),and depositedinthewellsupportsthisdate.Incontwo2mx2mstone-filledcaissons(caissons1 clusion,building8andwell1areprobablylater
and2)(Larsen1967a,42).InapreviousstudyI thanthetwelfthcentury.Underneathbuilding8
have argued that building 8 and well 1, build- anassemblageofstoneswaslocated(N-11/plan
ings10and11andthefencecouldbecontem- 1andplan2),asecondassemblageofstoneswas
porarywiththeoldeststructuresinphase10/9 documented closer to the fence (N -11/plan1),
(1986/90)attheneighbouringsite8,Dreggsal- interpreted by Larsen as a possible pavement
menningen 14-16. I have further argued that (diaryp15).Asthetwostoneassemblageswere
buildings12and13andcaissons1and2,could bothembeddedin/ontopofthenaturalsubsoil,
be contemporary with the structures in period theymayhavebeencontemporaryandperhaps
2atsite6,Bryggen(Hansen1994b,44ff).The alsopartofthesamepavement.Thepavements
orientation of the structures and geographical canonlybedatedrelatively:thefirststoneassemclosenessofsite8andsite9,aswellastypologi- blageisolderthanbuilding8andthesecondascalsimilaritiesbetweenthestructuresonsite6 semblagemaybecontemporarywiththefirst.In
andsite9weremymainarguments.
conclusion,theareainthenorthernpartofthe
72
Figure13.Thepalisade-builtfenceatsite9,Sandbrugaten5.(Negative67and70,photoArneJLarsen)
Figure14.ThestratigraphicalrelationshipbetweenartefactassemblagesdescribedinAppendix3andmajorstructuresandlayers
atsite9,Sandbrugaten5
siteappearstohavebeenopenwithoutbuildings
Theobservationthatthenorthernpartofthe
untilthethirteenthcentury,butsectionsofthe sitecontainednobuildingsuntilafterc1225is
areamayhavebeenpavedatanearlierstage.
usedasasupplementarysourceforhorizons1
7Evaluationofthearchaeologicalandbotanicalsources
73
to 5. Prior to the erection of the building and
well,sectionsoftheareanorthofthefenceseem
tohavebeenpaved.Asthisdatacannotbedated
except relatively to sometime before the thirteenth century, I choose to omit the material
fromthestudy.
Thefenceandstructuressouthofthefence
Theremainsofbuildings10and11andthefence
liestratigraphicallybelowbuildings6and7.The
orientation of the fence differed from that of
buildings10and11.Asthefenceandthebuildingsdidnotoverlap,thestratigraphicalrelationship between the structures is not immediately
clear.Accordingtothereport,thefencewasdug
intothesterilemoraineandbuildings10and11
werepartlybuiltuponfill-masses.Someofthese
massesaccumulatedontothefence(N-11/plan
4,mrk46-49),showingthatthefencewasconstructedbeforethebuildings.Stillthefencemay
havebeeninusewhenthebuildingswereconstructed and used, judging by the level of the
structures.Someofthemassesunderbuildings
10 and 11 contained artefacts (assemblage 5)
datedbroadlytosomewherebetween1050-1200
onthebasisofshoematerial.Thisgivesabroad
post quem date for the erection of buildings 10
and11.Whenthefenceandbuildings10and11
wentoutofusetheconstructionsweresucceeded
bybuildings7and8andapassage.Artefactsin
assemblage4maystemfromthelayersbetween
buildings10and11andthepassage.Theyoungesttypeofpotteryinthisassemblagedatesthe
passageto‘after1215/25’,showingthatthefence
andbuildings10and11mayhavebeeninuse
intothethirteenthcentury.
Tosumup,thefencewasbuiltbeforebuildings10and11,whichwereconstructed‘afterc
1050-1200’. Both the fence and the two buildings may have been in use into the thirteenth
century.Thereisnoclearevidenceforwhenthe
fencewasconstructedmoreprecisely,exceptthat
thishappenedbeforebuildings10and11were
built. The difference in the orientation of the
buildings in relation to the fence may suggest
that the fence and the buildings were built accordingtodifferentplans,indicatingthatsome
timepassedbybetweentheconstructionofthe
fence and the construction of the buildings. I
willreturntothisbelow.
74
Noartefactassemblagescouldbefoundwith
clear information about their stratigraphical
relation to buildings 12, 13, 14 and caissons 1
and2.Assemblage10must,however,havebeen
depositedclosetobuildings13,14orcaisson2,
as these were the only constructions in grid L
-11/plan 8. We do not know if the assemblage
wascontemporarywiththestructures,sincethe
fill-massesthatcontainedtheartefactsmayhave
beendepositedaftertheconstructionswentout
ofuse.Thedatingofthematerialsuggeststhat
theassemblagecanbedatedbroadlytobetweenc
1050-1200,indicatingthatcaisson2andbuildings13and14werestillinuseorwentoutofuse
duringthisperiod.
Assemblage11istentativelydatedtoafterthe
end of the twelfth century, after c 1200, and
was found within the walls of building 12. As
thecontextissomewhatunclear,itisdifficultto
determinewhetherthefindsbelongtolayersrepresentingtheuseofthebuildingoriftheystem
from fill-masses spread over the area at a later
stage.Noartefactscouldbeassociatedwithcaisson1.Theartefactsshowthatbuildings13and
14andcaisson2probablywereinuseorwentout
ofusebetweenc1050andc1200.Assemblage
11indicatesthatbuilding12wasinuseorwent
outofuseafterc1200.Thestratigraphicalrelation between building 14 and caisson 2 shows
at least two phases of structures; where building14belongedtotheolderphaseandcaisson
2totheyounger.Whenthecaissonwasbuilt,a
slightlyneworientationofstructuresonthesite
wasintroducedcomparedwithbuildings10and
11aswellasbuilding14.Ithereforesuggestthat
buildings10and11belongedtothe‘olderphase’
along with building 14. As mentioned earlier,
buildings10and11werebuiltuponfill-masses,
someofwhichaccumulatedontothefence,and
theorientationofthefencedifferedfromthatof
thebuildings.Onthebasisofalltheseobservations,Isuggestthatseveral‘phases’ofstructures
werepresentinthispartofthesite:phase(1)the
fence,(2)thefill-massesthataccumulatedonto
thefence,(3)buildings14,10and11,andfinally
(4)caisson2.
Thetwo2mx2mstone-filledcaissonsatsite
9areidenticaltothoseassignedasbasicsources
for both horizon 4 and horizon 5 at site 6 and
for horizon 5 at sites 27 and 28 (cf pages 93ff,
132ffand135ff).Consideringthisbackground,
thecaissonsfromsite9mayalsohavebeenconstructedduringhorizon4orhorizon5andare
assignedtentativelytothesehorizons.Buildings
12-14weremostlikelybuiltascellarbuildings,a
constructiontypeknownperhapsfromasearly
asphase1.1.2(horizon3)anddefinitelyfromperiod2(horizon5)atsite6(cfp89ff).Asbuilding14mustbeolderthancaisson2,assignedto
horizons4and5,building14maybetentatively
assignedtohorizon3.Thisisnotinconflictwith
atypologicaldateforthebuildingbasedonmaterialfromsite6.Sincebuildings10and11have
thesameorientationasbuilding14,theyarealso
assignedassourcesforhorizon3.Buildings10
and11mayhavebeeninuseintothethirteenth
century (cf above), so they are also assigned to
horizons4and5.Buildings12and13fitintoa
buildingpatternassociatedwiththe2mx2m
stone-filledcaissonsonsite6inhorizons4and
5consequentlyIassignthesebuildingstohorizons4and5alongwithcaissons1and2.The
suggesteddatesarenotinconflictwiththewide
datesprovidedbytheartefactmaterial.Sincethe
datesfortheSandbrugaten5structuresareindirect, the structures are used as supplementary
sources.
Ifweacceptasapremisethatbuildings10and
11werebuiltcontemporaneouslywithbuilding
14representinghorizon3,andthatmassesaccumulatedontothefencebeforebuildings10and
11 were built, then we may assume that some
timepassedbybetweentheconstructionofthe
fenceandtheconstructionofbuildings10and
11. This assumption is also supported by the
differingorientationofthebuildingsversusthe
fence.Assumingthatbuildings10and11representhorizons3-5onthesite,itisreasonableto
let the fence represent horizon 2. The culturelayerswhichaccumulatedontothefencewould
thenbelongtohorizons2or3.Iwillelaborateon
thesepresuppositions.
Aswehaveseen,the16mlongpalisade-built
fence was constructed before buildings 10 and
11. The fence had the same orientation as the
structureintheoldestphase(10/9)atsite8-the
phasewasdatedtentativelyto1100-1150andassignedtohorizon4asasupplementarysource(cf
p98ff).Inthesucceedingphaseatsite8(phase
9/8,horizon5),theorientationofthestructures
7Evaluationofthearchaeologicalandbotanicalsources
hadchangedsomewhatandnolongercorresponds
withthefenceatSandbrugaten5.Thisindicates
thatthefencewasconstructedincorrespondence
withtheoldestbuildingatDreggsalmenningen
14-16ormorelikelyviceversa-thebuildingwas
builtincorrespondencewiththefence.Thisimpliesthatthefencewasolderthanoratleastcontemporaneouswiththeoldeststructuresatsite8.
Typologicalaswellasotherfactorsmaypointin
thesamedirection:atsite6palisade-builtfences
identical to the fence at site 9 were assigned to
horizon2asasupplementarysource(cfp92ff)
thetypologicalcoherencebetweenthestructures
mayinitselfbeanindicationthatthestructures
werecontemporary.Thefencesatsite6werecovered by deposits when they went out of use in
thefollowingphaseatthesite,theywerethusno
more visible. The fact that identical techniques
andmaterialwereusedwhenbuildingthefences
atthetwosites,stronglysuggestthatthefenceat
site9wereconstructedwhilethefencesatsite6
werestillvisibleinthelandscape.Alsothefact
that the orientation of the fence at site 9 corresponds with the orientation of the transverse
fencesatsite6,inthesensethatitrunsparallel
to the Vågen shoreline, points towards a connectionbetweenthefences.This,seentogether
with(1)theorientationofstructuresassignedto
horizon 5 at site 8, and (2) the change in the
orientationofthestructuresatsite9,fromwhen
thefencewasbuilttowhenbuildings10and11
were constructed, implies that the fence at site
9 may belong to a plan that predates the earliestbuildingsontheVågenshoreline.-Thisplan
wasnotquitesatisfactorywhentheareasouthof
thefencewasbuilton.Althoughnoneofthese
arguments are conclusive, when seen together,
theypointtowardsanassociationofthefenceat
Sandbrugaten5withthefencesatsite6inhorizon2.Thefenceatsite9isthereforeassigned
tohorizon2.Asthedatingmaterialisindirect
andtoalargeextentdependsontheindirectdate
ofthefenceatsite6,thefenceatsite9isused
asasupplementarysource.Asmentionedabove
this fence may still have been in use until the
beginningofthethirteenthcenturyandisalso
assignedasasourceforhorizons3-5,againas
a supplementary source.28 The culture-layers
thataccumulatedontothefenceareassignedto
horizons2or3asIcannotdeterminewithany
75
certaintywhichhorizonitmaybelongto,thusit
servesasageneralbackgroundsourceonly.The
lackofstructuresorculture-layersthatcouldbe
associatedwithhorizon1,southofthefenceis
usedasasupplementarysourceforhorizon1.
Buildings10,11and14wereassociatedwithhorizon3.Building14isassumedtohavegoneout
of use in horizon 4 while buildings 10 and 11
maystillhavebeeninuse.Inhorizons4and5,
two2mx2mcaissonsdemarcatethepresence
oftwopassages.Buildings12and13arealsoasMajorfeatures
signedtohorizons4and5.Northofthefencea
Tosumup,thematerialfromtheSandbrugaten pavementdatedtobeforethefirstquarterofthe
5sitehasbeendatedthroughverticalandhori- thirteenthcenturywaslocated,however,itisnot
zontallinkstobetterdatedsourcesfromsitesin usedfurtherasasourceinmystudy.
thevicinity,andcanonlyserveasasupplementarysource.Nostructuresorculture-layerscould Artefactcategories
beassociatedwithhorizon1.Thepalisadefence The layers at site 9, Sandbrugaten 5 generally
thatranacrossthesitehasbeenassociatedwith cannot be included as a source since they were
horizon2.Thisfencemayalsohavebeenpresent not documented in any detail. Only artefacts
inhorizons3-5.Waste-layersdepositedontothe fromassemblage5maycomefromaclosedconfencehavebeenassociatedwithhorizons2or3. textthatcanbebroadlydatedtotheperiodunTable5.Site9,Sandbrugaten5(1967)BRM3
Archaeologicalevidence
Dating
Structures
Nobuildings
beforebuilding
8andwell1
Caissons1
and2
Pottery
Other
GrimstonDecorated
Building8andwell1constructed
ware,Ardenburg,
afterc1225
ScarboroughII
Shoematerial.Sametypeof
Begins‘after1110’
structuresasstructuresatsites Endsafter1170
6,28and27
Buildings12-13
Shoematerial.Sametypeof
Begins‘after1110’
buildingpatternasatsite6,
Endsafter1170
horizons4and5
Buildings10
Shoematerial.Stratigraphical Contemporarywithbuilding14(?)
and11
relationshipto‘thefence’.Same Endsafter1170(afterc1225?)
orientationasbuilding14
Culture-layers
Stratigraphicalrelationshipto
depositedonto
thefence.
thefence
Thefence
Olderthanbuildings10and
Builtin2ndquarterofeleventh
11.Differsinorientationfrom century.
buildings10and11.Same
Endsinthirteenthcentury
orientationasphase10/9
structuresatsite8.Typological
andfunctionalcoherencewith
horizon2materialatsite6
Culture-layers
Stratigraphicalrelationshipto
underbuildings
buildings10and11
10and11
shoematerial
Building14
Stratigraphicalrelationship
tocaisson2.Sametypeof
structuresasstructuresatsite6.
‘Thepavement’
Stratigraphicallybelow
Olderthanbuilding8
building8,ontopofthe
naturalsubsoil
Thenatural
Priortohorizon2
subsoilsouthof
thefence
DatabasedonLarsen1967,originaldocumentationandHansen1994b
76
Horizon
Source
type
(B/S/G)
1-5
S
4-5
S
4-5
S
3-5
S
2or3
S
2-5
S
3
S
3
S
?
G
1
S
derinvestigation.Theartefactsfromassemblage
5 are assigned to horizon 3 as supplementary
sourcesofcategoryIIastheyareassociatedwith
culture-layers immediately below buildings 10
and11andwedonotknowhowtheyweredeposited.
Site10,Sandbrugaten3(1953)
Locationofthesite
A note dated 22/5 1953, probably written by
Cato Enger gives us information about observations made at ‘Hoteltomten v/Sandbrugaten’,
(‘thehotelsitebySandbrugaten’)(Enger1953).
TheobservationsweremostlikelymadeatHotel Slottsgården, today’s Dreggen Hotel, which
openedin1956atSandbrugaten3(cfHartvedt
1994,239).Thissiteissituatedtothesouthof
site9.
Dates
Threeorfour2mx2mlog-builtcaissonshad
been picked up from the middle of the site.
When Enger arrived one was still in situ. The
caissonshadbeenpreservedataheightofabout
1 m and they had been placed directly on the
naturalsand.Fromthedescriptionitislikelythat
thecaissonsareofthesametypeasthosedocumented at several other sites by the waterfront
inthetwelfthcentury.Inwell-datedcontextsall
thesecaissonsareassignedtohorizon4or5(site
6,site28,andsite27).Ontypologicalgrounds
thecaissonsfromsite10maybeassignedtohorizons4and5asasupplementarysourceand
thematerialcannotelucidateactivityonthesite
priortohorizons4or5.
The‘easternstreet’mustbeBrynjulfgaten,which
no longer exists. When placing the caisson accordingtothedescriptioninthenoteandbythe
helpofanoldmapshowingtheareabeforestreet
regulations in the late 1940s and early 1950s
(Bergen1913-30),itformsalinewiththe2m
x2mstone-filledcaisson1atsite9,makingthe
localisationofthecaissonatsite10veryconvincing.Theremainingthreeorfourcaissonsatsite
10werelocatedin‘themiddleofthesite’.Ifthe
caissons follow the caisson-pattern seen on site
6,theyshouldbeplacedatadistanceofabout6
mfromoneanotheralongalinerunningat90
degreestothewaterfrontoralongthe+/-0masl
contourlineofthenaturaltopography.The2m
x2mcaissonatsite9,Sandbrugaten5andthein
situcaissonfromsite10makeupthebeginning
ofarowofcaissonsthatranat90degreestothe
waterfront.
The +/- 0 contour, which has been reconstructed through data from surrounding sites,
runsthroughthemiddleofsite10.Thelocation
of the remaining caissons at Sandbrugaten 3 is
reconstructedusingthisinformation.
Site11,Dreggsalmenningen20(1967)BRM4
The excavation at site 11, Dreggsalmenningen 20, was an open area investigation coveringabout740m2.ExcavationsupervisorArneJ
Larsen’s archive report gives a brief account of
the relative chronology for some of the structures(Larsen1967b).EarlierIhaveattemptedto
sortoutanddatetheoldeststructuresatthesite
(Hansen1994b).Togodeeperintothequestion
of dating, I have now studied selected artefact
assemblages(Appendix4).
Locationofthestructures
The in situ caisson was placed in the southeasterncornerofthesite,about8.3mfromthe
neighbouring house and 8 m from the edge of
thepavementintheeasternstreet(Enger1953).
Dates
Structures uncovered down to the moraine are
presumed to be the oldest on the site. These
structurescomprisetheremainsoftwopalisade-
Table6.Site10,Sandbrugaten3(1953)
Archaeologicalevidence
‘Phase’
Pottery
4or5logbuilt
caissons
Plotboundary
Naturalscientificdates
Other
Dendro
Typologicalcoherencewith
caissonsatsite6,site28and
site27,horizon5
Reliesonplotidentification
onsite9
7Evaluationofthearchaeologicalandbotanicalsources
TL
14
Dating
Horizon
Source
type(B/
S/G)
C
5
S
4-5
S
77
Fire-layer2
Potteryinassemblages1,3and4datefire-layer
2toafterc1215/25.Assemblage7givesasimilar
datetoafterc1215/25forthedepositionoffillmassesafterbuilding12burntdown.Thefire,
whichscorchedbuilding12,mayalsobedated
to the years after c 1215/25 and may therefore
inalllikelyhoodbeidenticalwithfire2.Fence
1,fence2andbuildings4,8,9,10and12were
probably all destroyed in or replaced by other
structures after fire 2, dated to after 1215/25.
Thisgivesanantequemdateforthestructures
belowthefire-layer.
Structuresbelowfire-layer2
Figure15.Apalisade-builtfenceatsite11,
Dreggsalmenningen20.(Negative154,photoArneJLarsen)
built fences (fence 1 and fence 2) (Figure 15),
fourbuildings(4,8,9,and10),andthreedrains
(3-5).Allthesestructureswerestratigraphically
superposed by fire-layers. The fire-layer above
buildings8,9and10andfence1wasdescribed
as ‘fire-layer 2’ (Larsen 1967b). The fire-layers
abovebuilding4andfence2werenotindirect
contact with fire-layer 2. Still, Larsen assumed
that these fire-layers also represented fire 2
(Larsen1967b,9,14).Building12alsobelongs
to the lowermost documented structures and
maylikewisehavebeendestroyedinafire.Figure 16 shows how the structures are related to
fire-layer2andtothefirewhichscorchedbuilding12.Thetwofencesdividethesiteintothree
plots,2911/A,11/Band11/C:buildings8and12
belongtoplot11/C,buildings4,9and10and
drain3and4toplot11/B.Nostructureshave
been identified on plot 11/A. Drain 5 replaced
fence2beforefire2occurred.
Sevenartefactassemblagesfromthesitehave
beendiscussedinAppendix4.Theseassemblagesindicatethedateoffire-layer2,thefire-layer
abovebuilding12andthestructuresunderthe
fire-layersbyprovidingwidepostquemdates.The
contextoftheassemblagesisgenerallynotwelldocumented,thedatesarethusratheruncertain.
First,Iwilldiscusswhetherthefire-layerabove
building12maybeidenticalwithfire2.Then
the construction dates for the structures below
fire2arediscussed.
78
Asthestratigraphicalrelationshipandtherelative chronology between the structures under
fire-layer2isnotclear,theconstructionofevery
singlestructureshouldideallybedateddirectly.
This is not possible on the basis of the documentedevidenceandotherapproachesmustbe
sought.
As seen in Figure 16 there were two levels
of structures below fire-layer 2: on plot 11/C,
buildings8and12weretheonlystructuresbelowfire2,theymakeuponelevelofstructures.
On plot 11/B, building 9 was represented by
twofloorlevels,building10wastorndownand
filled-inbeforefire2,anddrain3replaceddrain
4.Building4existedthroughthewholeoflevel
1and2.Fence2betweenplots11/Band11/A
wasreplacedbydrain5beforefire2.Thecrucial
point here is whether any of the remains from
levels1and2areearlierthanc1170.
Only one artefact assemblage can be related
tothestructuresbelowfire2:assemblage2that
was uncovered under one of the floor levels of
building9orbuilding10onplotB.Whetherit
belongstolevel1or2,isimpossibletoascertain.
A comb of type D430 was found in the assemblage,datingbuilding9orbuilding10broadly
toafter1170/71.Thisdatecorrespondswiththe
afterc1215/25dateoffire2.Thecombcannot
dateotherstructuresonthesitebecauseallthree
plotsonthesitewerenotnecessarilybuiltupon
simultaneously. We therefore have to consider
the construction date for the other structures
throughothermeans.
It is not clear whether building 4 actually
burnt in fire 2, but it predates the fire. Both
Figure16.ThestratigraphicalrelationshipbetweenartefactassemblagesdescribedinAppendix4andtheoldeststructuresand
layersatsite11,Dreggsalmenningen20
buildings 8 and 12 were scorched by fire and
were thus probably still standing when fire 2
occurred ‘after c 1215/25’. If the three buildingstheoreticallyhadbeeninusefor25to50
yearswhentheyweredestroyedorwentoutof
use ‘after 1215/1225’, they may perhaps have
beenconstructedasearlyaswithintheperiod
covered by horizon 5 (cf p 60ff). Building 12
wasmostlikelyacellarbuilding.Parallelstothe
building are found at site 6 probably through
thewholeofthetwelfthcentury(cfp92ffand
Herteig1992).Theconstructiondateforbuilding 12 therefore cannot be narrowed through
analogousmaterial.Neitherbuildings4and8
can be dated more closely through the available material, nor can the construction date
for drains 3 and 4 be determined. Drain 3,
however, replaced drain 4, indicating that the
drainage function dates back some time. It is
notunlikelythatdrain3or
4werecontemporarywithbuilding9orbuilding
10thatlayadjacenttothedrain.Basedonthe
available information, I tentatively suggest that
buildings4,8,12andperhapsalsobuilding9or
building10wereconstructedasearlyasduring
the period covered by horizon 5. The drainage
functionrepresentedbydrain3or4mayalsogo
back into horizon 5. The material is, however,
poorlydatedandisthereforeusedasasupple7Evaluationofthearchaeologicalandbotanicalsources
mentarysourceonlyforhorizon 5.Nobuildingscouldbeassociatedwiththeperiodbefore
horizon5.Thisinformationisusedasasupplementarysourceforhorizons1-4.
Thestratigraphicalrelationshipbetweenfence
1and2andthebuildingsonsite11isnotdocumented. At site 9, the palisade-built fence was
tentativelyassignedtohorizon2.Therewereno
buildingsnorthofthefenceuntil‘afterc1225’
andsouthofthefenceculture-layersandbuildings were tentatively assigned to horizons 3-5
(cfp101ff).Inotherwordsthebuildingsatsite
9wereprobablynotconstructedasearlyasthe
fence on this site. The situation at site 11 may
be similar, and we cannot immediately assume
thatthefencesatsite11werebuiltcontemporaneouslywiththedocumentedbuildingsonthis
site.Itisthusimpossibletodeterminethedate
fortheconstructionofthefencesthroughmaterialatsite11alone,andIshallattempttolink
thefencestomaterialfromsitesinthevicinity.
Takingintoconsiderationthatthefencesatsite
11,site9andsite6wereidenticalasforbuilding
techniques and the choice of materials, it seems
reasonabletosuggestthatthefencesatthethree
sites were built contemporaneously. As already
pointedout,inthediscussionofthepalisade-built
fenceatsite9,thecircumstancethatthefencesat
site 6 were covered by deposits in the following
79
phase furthermore provides a strong argument
thatthepalisade-builtfencesbeyondsite6were
builtwhilethefencesatsite6werestillvisiblein
thelandscape.Onthisbasisitseemslikelythat
thefencesatsite11werebuiltatthesametimeas
thefencesatsite6.Asseenabovethefencesatsite
6arenotdirectlydated,buthavebeenassociated
withhorizon2asasupplementarysourceonly.In
thediscussionofthedateofthepithouseatsite
7,itwasarguedthatthepit-housemightwellbe
partofasettlementgenerallyassociatedwiththe
palisade-bounded plots. This was based on the
locatedbytheVeisanshoreline.Twofencesthat
demarcateplots(11/A,11/B,and11/C)werethe
onlystructuresthatcouldtentativelybeassigned
tothishorizon.Inhorizons3and4thesituations
seemtobethesameasinhorizon2.Horizon5
isrepresentedbystructuresfoundontwoofthe
plots:building4,adrainandpossiblyalsobuildings9or10onplot11/Bandbuildings8and12
onplot11/C.Noartefactshavebeenassociated
withthestructuresinhorizons2-5.Activitiesat
thesitepriortohorizon2cannotbeelucidated
throughthematerial.
Table7.Site11,Dreggsalmenningen20(1969)BRM4
Archaeologicalevidence
Phase
Pottery
Buildings4,8,12,9or10
Nobuildingsorculturelayerspriortohorizon5
Fences1and2
NaturalScientificdates
Other
Dendro TL
Stratigraphicalrelationto
fire-layer2.Maximumlife
expectancyofbuildings.
14
Dating
Sourcetype
(B/S/G)
5
S
1-4
S
2-5
S
1
S
C
Typologicalresemblance
tostructuresatsite6and
site9.Sameorientationas
buildingatsite7
Thenaturalsubsoil
Horizon
Priorto
horizon2
Databasedondocumentationmaterialfromthesite
correspondence between the orientation of the
buildingandtheplotsatsite11.Basedona 14C
datetheconstruction-oractivity-phaseofthepithousecouldnotbelaterthanabout1020.Thus,
iftheassociationbetweenthepit-houseandthe
fencesatsite11isreal,whichdoesnotseemfar
fetched,thedateofthepit-houseorassociatedactivitiessupportstheideathatthefencesatsite11
(andthefencesatsites6and9aswell)shouldbe
assignedtohorizon2.Accordinglythefencesat
site11areassignedtohorizon2.
Sincethefenceswerestillinusewhenthefirst
buildingswereconstructedonsite11theymay
alsorepresenthorizons3-5here.Thedatingof
the material is founded on horizontal links to
supplementary sources and is accordingly used
asasupplementarysourceonly.Thelackofmaterialdatingbacktohorizon1isusedasasupplementarysourceforthishorizon.
Site12,Dreggsalmenningen10-12(1972)
BRM42
Site 12 at Dreggsalmenningen 10-12 covered
about735m2,butwasonlypartiallyinvestigated
archaeologically.Areportwithabriefdescription
ofthearchaeologicalobservationsbutwithouta
stratigraphicalanalysisordatingisavailable(Reimers1972b).
A2mx2mstone-filledcaissonwasidentifiedatthissite.Baseduponthedatesofsimilar
caissonsatsites6,27and28,thecaissonatsite
12servesasasourceforhorizons4and5.The
datingevidenceis,however,notsatisfactoryand
the material must be used as a supplementary
source. No artefacts have been assigned to the
caisson.Thedocumentedmaterialcannotelucidateactivitiesonthesitepriortohorizon4.
Site13,Dreggsalmenningen10-16(1986)
BRM242
The excavation at site 13, Dreggsalmenningen
Nostructuresorculture-layerscouldbeassoci- 10-16, was a trench survey with 8 profiles, the
ated with horizon 1. In horizon 2, site 11 was naturalsubsoilwasreachedduringtheinvestiga
Majorfeatures,artefactcategories
80
Table8.Site12,Dreggsalmenningen10-12(1972)BRM42
ArchaeologicalEvidence
Profile/Plan Pottery
KK/I-4,
I-5
Naturalscientificdates
Other
Dendro
Typological
coherencewithsite6
TL
14
Dating
Horizon
Sourcetype
(B/S/G)
C
‘After1110’
and1170/71
4-5
S
DatabasedonReimers1972
tion.Areportwithastratigraphicalanalysisand materialfromthesitealone.Excavationsupervidatesisavailable(Dunlop1986b).
sorsCliffordLongandLyderMarstranderinterpretedthetwocaissonsaspartofapierbuiltin
Dates
connectionwiththeconstructionoftheChurch
Inallthephasessettlementwasdestroyedthrough ofStMary(LongandMarstrander1980,23).I
totalorpartialfires(Dunlop1986b).Theexca- findthisinterpretationplausiblesincethepieris
vationsupervisorDunlopsuggestedthattheold- orienteddirectlytowardsthechurch.Thestandestphase8wasdestroyedinthe1198townfire ingChurchofStMarywasprobablyunderconandthelaterphasesinfiresdatedtoc1230,1248 structionfromc1140(LidénandMagerøy1990,
and 1332. The absolute dates were based upon 99). It may, however, have had a predecessor,
pottery(Dunlop1986b).However,theceramic erected as early as c 1100 (Lidén 1993, 74 and
evidenceisverysparse:nosherdswerefoundin cfsite23).Theoretically,thepiermaythushave
phase 8. Three sherds of Paffrath derived from been built as early as c 1100. Elsewhere I have
phase 7, indicating that tha phase belonged to discussedthedateoftheoldestphasesatsite14
thetwelfthcentury(Lüdtke1989,32).Inphases (Hansen1998)andmadethecasethattheoldest
6and5therewerenodatablesherdsandinphase phase,phase1,endedatthesametimeastheend
4asherdofYorkshirewaresuggestedadateto ofperiod2(1170/71)atsite6.Thisgivesatentaafterc1190(cfReed1990,30).
tivedatingframeforphase1tobetweenc1100
Accordingtomydatingcriteriaforbasicsourc- and1170/71,placingphase1inhorizons4and
es,theevidenceisnotpreciseenoughtogivean 5.Asthebeginningoftheoldestphaseatsite14
absolutedateofthematerial.Isitpossibletogive cannotbedatedthroughartefactsornaturalsciatentativedateinotherways?Thecaissonatsite entificmethods,Iwillseeitinabroadercontext
12, assigned to horizons 4 and 5, was situated byevaluatingthegeneralpatternsinthematerial
lessthan10mfromthetrench.Becauseofthe (cfp185ff).Thedateofthebeginningofphase
spatialclosenessitmightbelikelythattherewas 1toc1100canonlybeconsideredastentative
activityattheneighbouringsite13aswell.How- and as a supplementary source for horizon 4.
everassites12and13aremostlikelylocatedon Thedatefortheendofthephaseismorereliable
two different plots (as we shall see later on in asitisbasedonthestratigraphicalrelationship
Chapter9),thespatialclosenesscannotbeused to site 6. Phase 1 may thus be used as a basic
asameansofdating.Consequently,thematerial sourceforhorizon5.
isomittedfromthestudy.
Majorfeatures,artefactcategories
Site14,Dreggsalmenningen(1979)BRM83
The excavation at site 14, Dreggsalmenningen
(1979),locatedbetweensite13andsite6,wasan
openareainvestigationcoveringabout288m2.
The natural subsoil was probably not reached
duringexcavation.Areportwithastratigraphicalanalysisanddatingsuggestionsbasedonpotteryisavailable(LongandMarstrander1980).
Two4mx4mstone-filledcaissonsrepresent
the oldest documented phase. It is not possible
to date the beginning of phase 1 through the
7Evaluationofthearchaeologicalandbotanicalsources
Thematerialcannotelucidateactivitiesonthe
sitepriortohorizon4.Inhorizons4and5,site
14waslocatedinopenwateralongthenorthside
of the Vågen Bay, between -2.5 and -3.0 masl
(LongandMarstrander1980,22).Theconstructionsinphase1consistoftwo4mx4mstonefilledcaissons,interpretedasapierbuiltforcarryingheavyloads(LongandMarstrander1980,
23).Onlyonelayerhasbeenassignedtophase1:
LayerA336.Thislayer,however,hasalsobeen
assignedtophase2.Becauseoftheuncertainties
81
Table9.Site14,Dreggsalmenningen(1979)BRM83
Phase
1
Archaeologicalevidence
Naturalscientificdates
Pottery
Dendro TL
Other
Stratigraphical
relationtosite6
andrelationtoSt
Mary’s
14
Dating
Horizon
Sourcetype
(B/S/G)
between
c1100and
1170/71
4-5
S,B
C
DatabasedonLong1980andHansen1994b
itwillnotbeincludedhere.Consequently,noartefactshavebeenassignedtohorizon5.
Site15,Stallen,Svensgården(1980/82)
BRM90
Theexcavationatsite15,Stallen,about50meast
of site 6 was an open area investigation, which
coveredabout75m2(TopArk).Thenaturalsubsoilwasreachedduringexcavation.Areportwith
a stratigraphical analysis and dates based upon
potteryandTLisavailable(Dunlop1984a).
Dates
Thedatingoftheoldestphase,phase10,isindirectsincethedatingmaterialfromthisphase
isnotsatisfactory.Theoverlyingphases9and8
aredateddirectly,however.AccordingtoDunlop phase 8 was terminated by a fire, well-dated by pottery, dendrochronology and TL. The
firewasidentifiedasthetownfireof1248.The
beginning of phase 8 is dated to after c 1220
throughGrimstonwaresdepositedinthebeginning of the phase. Thus Dunlop assumed that
thefirethatterminatedphase9shouldbedated
toc1220/30.Adatefortheprecedingfire-layer,
whichmarkedtheendofphase10,wasobtained
by‘countingfirelayers’andthusassumingthat
thisfire-layerwasidenticalwiththeknowntown
fireof1198(Dunlop1984a,34-37;Dunlopand
Sigurdsson1995;Dunlop1998).
Grimston wares have been produced in differenttypes,withdifferentdates.The‘Grimston
Decoratedware’,31whichischaracterisedbythe
applicationofplasticdecoration,oftenofananthropomorphic character, is traditionally dated
toafterc1220(JenningsandRogerson1994).
‘Grimston ware’, which is plain without elaborateplasticornaments,isdatedfromtheendof
thetwelfthcentury(Reed1990,31).
TheGrimstonwarepresentinphase8atthe
Stallensitehasnoplasticornaments,eventhough
some of the sherds are so large that we would
82
expectornamentstobevisibleiftheyhadbeen
presentonthevessel.Thesherdsmaytherefore
beclassifiedastheplain‘Grimstonware’dating
fromtheendofthetwelfthcentury,ratherthan
tothe‘GrimstonDecoratedware’.Consequently,
thebeginningofphase8andthefirethatended
phase 9 may be dated to as early as the beginningofthethirteenthcenturyortheendofthe
twelfthcentury.Thisdatesuggeststhatthefire
isidenticalwiththetownfireof1198,whichaccordingtothewrittenrecords,reducedallofthis
partofthetownintoashes(Helle1998,25-28).
Suchaninterpretationisnotinconflictwiththe
ceramicassemblageinphase9.However,theceramicmaterialfromphase10cancannotgivea
closedateforthebeginningorendofthisphase.
Consequentlyitistemptingyetagainto‘count
fire-layers’andassumethatthefire,whichended
thephase,isidenticaltothemajortownfireof
1170/71,whichprecededthetownfireof1198.
At site 6, situated at a distance of about 60-70
mboththesefiresarealsopresentinthematerial(cfp82ff).Thephasemay,accordingtothis
interpretation, be used as a source for horizon
5.However,asthematerialcannotbedateddirectly it must be considered a supplementary
source.
Anintensedepositionoflayersinitiatedphase
10. The contents of these layers indicate that
therewassettlementintheareabysite15prior
tophase10aswell(Dunlop1984a,46-47).Itis
notpossibletodatethisactivityexceptasolder
thanhorizon5,thematerialisthereforeusedasa
generalbackgroundsourceforhorizons1-4.
Majorfeatures,artefactcategories
Forty-fivelayerswereassociatedwiththeinitiatingstageofhorizon5/phase10.Thefindsfrom
theselayersareassignedtohorizon5ascategory
II finds. After the deposition of layers, two or
three foundation substructures were built. The
phaseendedinafire,whichleftathickfire-layer
over most of the site (Dunlop 1984a, 29, 46).
Findsfromthefire-layerareassignedtocategory
I.Thestructureswereinterpretedasfoundations
forbuildingsorfor‘awalkwaybehindthequay
front’(Dunlop1984a,46).Ifinditplausibleto
assume that part of the site was occupied by a
passage. In later phases a passage was situated
hereandtherewascontinuityfromtheoldestto
thelaterphasesatthesiteintermsoftheorientationandlocationofstructures.Italsoseemed
thatthesitewaslocatedwithinoneplotduring
allphases(Dunlop1984a,54).Ifthestructures
in phase 10 were built according to the same
building pattern as the following phases 9-1, a
passagewouldgenerallyoccupythewesternmost
half of the site, the easternmost part would be
occupiedbyabuildingoratransversegangway
betweenbuildings.Thestructuresatsite15are
therefore reconstructed as a passage and as an
undefinedbuilt-upareainphase10.
graphicalanalysisanddatesisavailable(Dunlop
1989a).Thephasesaredatedbypottery.None
ofthestratacanbedatedtoearlierthanc1170.
Thenaturalsubsoilwasreachedin‘plan1’and
profile12only.Theoldestculture-layersinprofile12weredatedtothefifteenthcentury(Dunlop1989a,20),however,sincetheywerelocated
ontopofabedrockoutcrop,theyarenotusedas
asourcehere(cfp57ff).Theoldestlayersabove
themoraineinplan1were‘notolderthanthe
late twelfth century’ dated on the basis of the
presence of Low Countries Highly Decorated
ware(Dunlop1989a).Sincethelowestculturelayersfromplan1canbedated,thisinformation
isusedasasupplementarysourceforhorizons
1-5.
Site17,Nikolaikirkealmenningen(1985)
BRM202
The excavation at site 17, Nikolaikirkealmenningen,wasanopenareainvestigationthatcovSite16,Bryggeparken(1989)BRM287
eredabout10m2.Areportwithastratigraphical
Atsite16,Bryggeparken,threesmallareaswere analysisanddatesbaseduponpotteryisavailable
investigated:(‘plan’1,2,3)coveringrespective- (Dunlop1985a).Thenaturalsubsoilwasreached
ly 7m2,9m2,and18m2(Hansen1994b,58) andtheoldestculture-layersabovebedrockcould
and13profilesintrenches.Areportwithstrati- be dated to the middle of the thirteenth cenTable10.Site15,Stallen,Svensgården(1981)BRM90
Phase
8
9
10
Priorto
phase10
Archaeologicalevidence
Naturalscientificdates
Pottery
Other
Andenne,Cookingpot,Dev
Stamford,‘Grimstonware’,
Londonarea,Miniatures,North
French,Paffrath,Pingsdorf,Soft
firedBlackware,
Andenne,Cookingpot,London
area,Paffrath,Pingsdorf,Soft
firedBlackware,
Dendro
Andenne,Cookingpot,
Relative
Paffrath,Pingsdorf,Shellyware, chronology
SoftfiredBlackware,
-‘counting’
fire-layersand
thecloseness
tosite6
Activity
indicated
throughthe
presenceof
redeposited
culture-layers
inphase10
TL
14
Dating
Horizon
Source
type
(B/S/G)
5
S
1-4?
G
C
Begins
after1198
Ends1248
Begins
after
1170/71
Ends1198
Beginsc
1120s
Ends
1170/71
DatabasedonDunlop1984
7Evaluationofthearchaeologicalandbotanicalsources
83
tury.Thismaterialwaslocatedinacleftinthe
bedrock. It is unlikely that older culture-layers
hadsystematicallybeencleanedoutofthecleft
(Dunlop1985a,8).Therefore,itseemsreliable
thatthisareawasnotoccupieduntilthemiddle
ofthethirteenthcentury.Thisinformationcan
beusedasasupplementarysourceforhorizons
1-5.
Site18,Koren-WibergsPlass(1980)BRM143
The excavation at site 18, Koren-Wibergs Plass
was a trench registration, comprising two profiles.Abriefreportisavailable(Myrvoll1980).
Thenaturalsubsoilwasreachedinprofile1.The
oldest culture-layer was a fire-layer (5) without
anyfinds.Inlayer4abovelayer5asherdofRedslipped Proto-stoneware was found. According
totheprevailingdateforthisware(Lüdtke1989,
32),layer4mustbelaterthanc1240,indicating
thatlayer5maybefromthefirsthalfofthethirteenth century. The material indicates that the
areawasnotoccupiedinhorizons1-5.Sincethe
lowermostmaterialcanbedated,itcanbeused
asasupplementarysourceforhorizons1-5.
Site19,Wesenbergsmauet(1989)BRM297
The investigation at site 19, Wesenbergsmauet,
was a trench excavation where 13 profiles were
studied.Areportwithastratigraphicalanalysis
anddatesforthelaterdepositsisavailable(Dunlop1989d).Theoldestlayersabovethenatural
subsoilmaybedatedtotheendofthetwelfthor
the beginning of the thirteenth century on the
basisofceramicmaterial(Dunlop1989d).The
materialcanthereforebeusedasasupplementarysourceforhorizons1-5.
Site20,Øvregaten39(1981)BRM94
Theexcavationatsite20,Øvregaten39wasan
open area investigation, which covered about
70m2.Thenaturalsubsoilwasreachedduring
theexcavation.Areportwithdatingsuggestions
basedonpotteryisavailable(Dunlop1982).
Dates
InhisexcavationreportDunlopgivesnodating
suggestionsfortheoldestphasesinthematerial,
phases9and10.Phases7,8and9weredestroyed
in fires. Through the presence of Scarborough
IIpotteryinanoccupation/destructionlayerin
84
phase7,Dunlopsuggestsadateof1225-1230for
thefire,whichdestroyedphase7.Onthisbasis
healsosuggeststhattheprecedingfirewhichdestroyedphase8,maybeidenticaltotherecorded
townfireof1198.Thefirethatdestroyedphase
9andmarkedthebeginningofphase8,haslaterbeeninterpretedastherecordedtownfirein
1170/71(Dunlop1998,135).Aftertheoriginal
reportwascarriedout,two14Cdatesfromphase
10 and layer 147 have been obtained (Dunlop
1982, Dating appendix). Together with the ceramicmaterialtheyprovidenewevidenceforthe
absolutechronologyofphases9and10,andalso
helptodateofphase8.
The pottery in phase 8 consists of Paffrath,
London Shelly and Developed Stamford wares.
These types are found from the middle of the
twelfth century (Lüdtke 1989, 32; Reed 1990,
28;BlackmoreandVince1994,33).Judgedby
thepotteryalonephase8maythusbedatedto
anywhereinthelasthalfofthetwelfthcentury
or later. A 14C date from layer 147, which may
beassociatedwithphase8,impliesasomewhat
earlierdateforthebeginningofthephase8than
theproposed1170/71date.
FirstIwilltakeacloserlookattherelationship between layer 147 and phase 8. Layer 147
wasoneofseverallayersunderfire-layer149that
werenotassignedtoaphaseintheoriginalreport(NVProfileB)(Dunlop1982,27).Froma
stratigraphicalpointofview,however,thelayers
under layer 149 should belong to phases 8 and
9,asfire-layer149wasbelievedtorepresentthe
firethatdestroyedphase8.Asmentionedabove,
phase9wasalsodestroyedinafire,leavingafirelayeronmostofthesite.Theexactsamestratigraphicalsituationasonthesiteingeneralcould
befoundinNVprofileB:fire-layer149,which
endedphase8,wasprecededbyafire-layer,firelayer40.Consideringthesimilaritybetweenthe
stratigraphicalsituationsinNVprofileBandthe
mainpartofthesite,itislikelythatfire-layer40
represents the fire that destroyed phase 9. This
layershouldthusbeassignedtophase9.Layer
147 must belong to the beginning of or to the
occupationalstageofphase8,asthelayerliesdirectlyontopoffire-layer40(NVprofileB).The
remains of an unnumbered structure between
layer147andfire-layer149indicatethattheactivitieswhichcausedthedepositionoflayer147,
werenotthelasttotakeplaceinphase8.A 14C
date from layer 147 should thus provide a date
fortheearlierpartofphase8ratherthanforthe
destructionofthephase.
The14Csamplefromlayer147wastakenfrom
bogmyrtle(Myricagale),ashrubwhichshould
notbeseriouslyinflictedby‘theoldwood’problem. With the highest probability the sample
may stem from sometime before 1160 (Figure
17),implyingthatactivityinphase8mayhave
startedbeforec1160.ThepresenceofDeveloped
Stamford ware in layer 70 dates activity in the
Figure17.14Cdatefrom
layer147inphase8site20,
Øvregaten39
Figure18.14Cdatefrom
layer24inphase10
Øvregaten39BRM94
7Evaluationofthearchaeologicalandbotanicalsources
85
beginningofthephasetoafterc1150(cfReed
1990, 28). Based on the ceramic evidence and
thepeakofprobabilityforthe14CsampleIsuggest1150-1160asatentativedateofthebeginningofphase8.Accordingtothistentativedate,
Dunlop’sdatingofthebeginningofphase8may
besomewhatlate.Itisinfactmorelikelythatthe
1170/71firemarkstheendofphase8ratherthan
thebeginning.Thereisnowayofdocumenting
thispossibilityonsolidevidence.However,with
thepresentevidenceavailableIfindittobethe
bestsuggestion.32
Ifphase8beganbetweenc1150and1160this
givesatentativedatefortheendofphase9.The
scarce ceramic material from phase 9 does not
contradictadateofbetween1150and1160for
the end of phase 9; one sherd of Paffrath ware
datesactivityinthephasewidelytothetwelfth
centuryorlater(Lüdtke1989,32).Thisleadsto
thequestionofwhenphase9began.Datafrom
phase9itselfdoesnotprovideevidenceforadate
forthebeginningofthephase.Iwillthereforego
ontophase10andexaminetheevidencefrom
thisphase.
A14Csamplefromlayer24intheconstruction
stageofphase10wasdated.Thesamplewastakenfromtwigssoitisprobablynotinflictedwith
the ‘old wood problem’. Several peaks of probabilitystandoutwithintwomainareas(Figure
18).Sincephase8mostlikelystartedsometime
between 1150 and 1160 and phase 9 came inbetween,thedateofthebeginningofphase10
mustbepushedbackwards.Infact,thepeakbetween1080and1160fitswellinthesequenceof
phasesandthe 14Cdategivesanindicationthat
thebeginningofphase10maybeplacedwithin
thistimeframe.
Thedatesprovided,howevertentative,givea
frameofdatesforphases9and10.Asphase 9
endedinafire,itisimpossibletosayhowlong
the phase lasted. The structures in phase 10,
however,leda‘naturaldeath’,implyingthatthe
structures lasted for perhaps 25-50 years before
phase9wasinitiated(cfp60ff).Ifweadd25-50
yearstothe14Cdatefromphase10,thisgivesan
estimateddatefortheendofphase10/beginning
ofphase9tobetween1105/1130-1185/1210;the
oldestalternativeobviouslybeingthemostrealistic.Thisis,ofcourse,toowideadatebutthe
availablematerialdoesnotsupportafirmerdate.
86
In conclusion, as phase 10 with some probability may have started between 1080-1160,
preferably in the first part of this time span,
the material may perhaps represent horizon 3.
Phase 8 is dated tentatively to c 1150/1160-c
1170.Thismakesphase8acandidateforhorizon5,andphase9maythereforebeacandidate
forhorizon4.Asthematerialcannotbedated
morepreciselythesourcemustbeconsideredas
supplementary.Nostructuresorculture-layers
couldbeassociatedwithhorizons1-2andthis
informationisusedasasupplementarysource
forthesehorizons.
Majorfeatures,artefactcategories
As we have seen no structures or culture-layers
could be associated with horizons 1-2. In horizon 3/phase 10, site 20 was situated between
6.5 and 8.5 masl. Only one construction was
assigned to the phase: K21, interpreted as part
ofafloor.WhenK21wasconstructed,layers83
and94wereprobablydepositedtosupportand
drainK21(Dunlop1982,29-30).Theselayers
are assigned to horizon 3, category II. Layer
24acompactexcrementlayerwasfoundontop
of K21 and was associated with this structure
(Dunlop1982,29-30).Layer24musttherefore
beassignedtohorizon3,categoryI.
Inhorizon4/phase9K20wasconstructedand
wasK20interpretedaspartofabuilding.Layer
63waslaidoverthebuildingsitebeforeK20was
constructed, while layer 72 was deposited duringphase9.Phase9endedinafire,represented
byfire-layers105and40(Dunlop1982).Asthe
firestruckbetweenc1150and1160,thematerial
fromtheoccupationandfire-layersisnotrepresentativeforhorizon4-activity.Accordinglythe
artefacts from these layers are assigned to horizon5,categoryII(cfp68ff).
The structure in horizon 5/phase 8, K23/
K25/K26,isinterpretedasabuildingwithaninternalfireplace,K24,andadrain,K27.Outside
thebuilding,K26isinterpretedasacourtyard
(Dunlop1982,42-43).Elevenlayersareassociated with the construction and development of
thephase8structures.Theselayersareassigned
tohorizon5,categoryII.Sixlayersareassociatedwithactivitiesduringthephase,andareassignedtocategoryI.
Table11.Site20,Øvregaten39(1981)BRM94
Archaeologicalevidence
Phase
8
Pottery
DevStamford
Other
Naturalscientificdates
Dendro
TL
9
10
‘Natural
death’
Thenatural
subsoil
Dating
14
C
Between Beginsbetween1150and1160,
950and ends1170.
1160
Begins?Endsbetween1150
and1160
1080Beginsbetween1080and
1160?
1160?Ends25to50yearslater
Priortophase10
Horizon
Source
type
(B/S/G)
5
S
4
S
3
S
1-2
S
Databasedon(Dunlop1982)
Site21,Klingesmauet(1989)BRM299
Dates
During the investigation at site 21, Klingesmauet, 16 profiles were documented. A report
with stratigraphical analysis and dates is available (Dunlop 1989f). Botanical material has
alsobeenstudiedinconnectionwiththeinvestigation (Hjelle 1989). The natural subsoil was
reached during the excavation. The oldest archaeologicaldeposits,K39andassociatedlayers,
havebeendatedbypottery,and14C.Thesecond
phaseofstructures33isdatedto‘possiblythelatertwelfthcentury’.Thefirstphaseofstructures
wasdestroyedinafire,datedto‘sometimeinthe
twelfthcentury(Dunlop1989f,23,28).
The dating material is very sparse and it
is hardly possible to get closer to a firmer date
based on the available material. If the second
‘phase’wasbuiltinthelatetwelfthcenturyitis,
however,temptingtosuggestthatthefirstphase
mayrepresenttheperioduptoabout1170and
thusmaybeusedasasourceinthepresentstudy.
Theceramicevidence(inphase1),theyoungest
type of pottery found being Normandy Gritty
waredatingtoabout1160/70inLondon(Vince
1991, Figure 7), is not in conflict with such a
suggestion.Figure19showsthatthe14C-sample
from the first phase may date to sometime between890and1160,withthreepeaksofprobability.Wedonotknowtheoriginalcontextof
thewood,butitprobablyoriginatesfromthesettlement represented by the first phase and thus
datesactivitiesduringthephaseratherthanactivitiesattheendofthephase.Thesamplewas
takenfromcharcoal,andmaybeinflictedbythe
‘old wood problem’; we may therefore have to
add some years to the maximum age provided.
7Evaluationofthearchaeologicalandbotanicalsources
Accordinglythe14Cdatefromtheoldestphaseis
notinconflictwiththeproposeddate.
Based on the available material, and bearing
the uncertainties in mind, I suggest that K39
andassociatedlayerscanbeusedasasourcefor
horizon 5. As the material is not well-dated it
canbeusedasasupplementarysourceonly.The
horizon5-phaseofstructuresonlycoveredthe
NEpartofthetrench(profiles12-16).Thelack
of culture-layers or structures associated with
horizons 1-4inthispartofthetrenchisused
asasupplementarysourceforthesehorizons.In
theSWpartofthetrench(profiles1-11)theoldestphaseofstrataabovethenaturalsubsoilwas
datedto‘possiblythelatetwelfthcentury’onthe
basisofpottery(Dunlop1989f,28).Thelackof
culture-layersorstructuresassociatedwithhorizons1-5intheSWpartofthetrenchisusedasa
supplementarysourceforthesehorizons.
Majorfeatures,artefactcategories
Site21,KlingesmauetwaslocatedonthemorainicterraceatthefootofFløjfjellet.Thesurfaceof
thenaturalsubsoilslopedfromamaximumelevation of 14.5 masl down to 7.5 masl towards
theSW.Nostructuresorculture-layerswereassigned to horizons 1-4. In horizon 5 structures
andculture-layerswereonlyrecordedinprofiles
12to16betweenthe10and15maslcontours.
K39,astonefoundationforabuildingwasdocumented,DunlopsuggeststhatK39wasastable
or byre. The botanical investigations support
thisinterpretation,butcannotestablishwhether
dungfromthebuildingwasfromcoworhorse
(Hjelle 1989, 7). According to the location of
layersassociatedwithK39,thefoundationmust
have been part of the NE wall of the building
87
Figure19.14Cdate
fromlayer65inthe
oldest‘phase’atsite
21,Klingesmauet
(Dunlop1989f,21,28).Eightlayerswereassoci- Dates
atedwiththeuseofK39,allhavebeenassigned Theoldestphases,phases17and16arenotdattocategoryI,horizon5.
ed, and the following phases 15 to 13 are only
given a wide date to the later twelfth century.
Site22,Kroken3(1984)BRM223
Dunlopsuggeststhatphase12endedin1170/71
Theexcavationatsite22,Kroken3,wasanopen andphase11in1198(Dunlop1987,52).
areainvestigation,whichcovered10-20m2.The
The date suggested for the end of phase 12
naturalsubsoilwasreached,areportwithstrati- seemstoooldwhenlookingattheceramicmategraphicalanalysisanddatesbaseduponceramic rialfromtheolderphases.AsherdofYorktype
material is available and botanical material has wareinphase14impliesactivitiesbytheendof
also been investigated (Dunlop 1987; Hjelle thetwelfthcentury,thatisafterc1170,accord1987).
ing to recent dates for York type wares (Reed
1990, 30; Armstrong, Tomlinson, and Evans
1991).Inphase15asherdofNorthFrenchtype
Table12.Site21,Klingesmauet(1990)BRM299
Archaeological
evidence
Profiles
Pottery
Other
Oldeststructuresand
Normandy
culture-layersinprofiles Grittyware
12-16
Thenaturalsubsoilin
profiles12-16
Thenaturalsubsoilin
profiles1-11
Databasedon(Dunlop1989f )
88
Naturalscientificdates
Dendro
TL
Date
Horizon
14
C
Between890 Broaddatebefore 5
and1160+
c1170
‘ownage’of
thesample
Priortohorizon5 1-4
Priorto‘possibly
thelatetwelfth
century’
1-5
Sourcetype
(B/S/G)
S
S
S
ware was found. This ware could have reached
Bergenshortlybeforec1170assherdsarefound
insmallamountsattheBryggensiteasearlyas
in period 2, which ended about 1170 (Hansen
1998,114).Consequently,itislikelythatphase
15,ratherthatphase12endedabout1170.
Inphase15onlyoneconstructionwasrecorded:apitK38,perhapsdugtoextractsandfrom
theunderlyingmorainicdeposits(Dunlop1987,
54).Thephasemayhavelastedforashortwhile
only, only a few days or weeks perhaps. Also
phase16,representedbyaditchoradrain,signifiesalimiteduseofthearea(Dunlop1987,54).
Thereisreallynowayofdatingtheendand
beginningofphase16,asthereisnodatingmaterial.Thesameappliestophase17,whichhadno
constructions.Pollensamples(showinganopen
landscape where grazing took place and grain
perhapswasgrown)weretakenfromthephase,
buthavenotbeendated(Hjelle1987,66).The
materialfromphases16and17isdifficulttofit
intothesystemofhorizonsandtocharacterisein
termsoflanduse,whetheritwasruralorurban
cannotbedetermined.Thematerialistherefore
omittedasasourceinthepresentstudy.Asphase
15 probably can be dated to the years around
1170,itbelongstohorizon5.Iconsiderthedatingsatisfactorysothematerialcanbeusedasa
basic source. The material prior to phase 15 is
omittedfromthestudy.
Majorfeatures,artefactcategories
As we have seen, only one construction can be
contributed to horizon 5/phase 15: this was
K38,presumablyasandpit.Twofill-layerswere
connectedtoK38.Theartefactsfromtheselayers must have been transported to the pit, and
shouldbelongtohorizon5,categoryIIfinds.
Site23,TheChurchofStMary(Mariakirken)
Site23,TheChurchofStMary,hasbeeninvestigated through building archaeology and style
studies. The church is mentioned for the first
timeinwrittensourcesinconnectionwithevents
thattookplacein1183(Ss1920,83;Holtsmark
1961, 117). Masonry studies of the standing
buildingsuggestthatthetwelfthcenturychurch
was under construction in the period between
1140and1180(LidénandMagerøy1990,99).
Thismakesthechurchabasicsourceforhorizon5.Thetwelfthcenturychurchwasabasilica
withasquarechancelandtwotowerstothewest
(LidénandMagerøy1980,11-20).Mostofthe
twelfth century building is incorporated in the
standingchurch.
Underneaththetowersofthestandingchurch
a foundation wall was recorded in 1974. The
wallwasprimarytothetowersandhadanorientationdifferentfromthatofthestandingchurch.
AccordingtoLidénandMagerøy,thewallmay
indicatethatanotherbuildingwaserected-orat
leastinitiatedherebeforetheconstructionofthe
standingChurchofStMary.Soapstonechipsin
thecoreofthewallindicatethatthebuildingwas
achurch.Furthermore,reusedsoapstoneashlars
inthestandingStMary’smayoriginatefroman
olderbuildingphase(LidénandMagerøy1980,
12). Based on these facts and on the fact that
the standing St Mary’s was initiated later than
a number of other churches in the town area,
LidénhassuggestedthatthestandingChurchof
StMarymayhavehadapredecessordatingback
toabout1100(Lidén1993,74).Pilastersinthe
northaisleareofanolderstylethanthosefound
elsewhereinthechurch.Lidénsuggeststhatthe
pilastersmayhavebeenformedbythelodgethat
worked at the Christchurch Cathedral at Holmen(LidénandMagerøy1990,173).
Table13.Site22,Kroken3(1985)BRM223
Archaeologicalevidence
‘Phase’
14
15
Pottery
YorkTypeware
NorthFrenchtypeware
Naturalscientificdates
Other
Dendro
16
TL
14
Dating
Horizon
Sourcetype
(B/S/G)
Endsafterthe1180s
Beginsandends
about1170
5
B
C
Nodata
Omitted
Databasedon(Dunlop1987)
7Evaluationofthearchaeologicalandbotanicalsources
89
At site 6, which covers the southern part of
St Mary’s churchyard at different times in history, remains were found that were interpreted
tentativelyasapassageandassignedtoperiod2.
Threelevelsofburialswererecordedunderneath
thewoodwork(Herteig1991,74).Ifthewoodworkwaspartofthepassagebetweenbuildingrows 3 and 4 at the Bryggen site in period 2,
as indicated by Herteig, it may have been constructedasearlyasthe1120s,whichisthedate
for the beginning of period 2 (Hansen 1998).
Stratigraphical evidence shows that the northernmostpartofthepassage,betweenrows3and
4,wasamongthefirststructurestobebuilton
this plot in period 2; a post from building 45
in the preceding period was thus reused in the
passage foundations (referred to as ‘post 72’ in
Herteig1991,94-97)whenthenorthernpartof
the passage was repaired. Thus it is likely that
thispartofthepassagemaybedatedtotheearly
partofperiod2,inthe1120s.
If the woodwork is part of the passage, the
threelevelsofburialsunderneaththewoodwork
must be older than the early part of period 2.
ThiswouldsuggestthatwearedealingwithburialspredatingthestandingStMary’s.34Oneway
ofgatheringfurtherinsightintothismatteristo
studytheorientationofthegravesfoundrespectivelyunderandoverthewoodwork,asachurch
and burials associated with the church usually
havethesameorientation(egEide1974).
The skeletons documented in squares R02
and R03, at site 6, are all of the same orientation as the standing St Mary’s, and there is no
differenceintheorientationoftheskeletonson
eithersideofthewoodwork.Thisisastrongindicationthattheburialsarecontemporarywith
thestandingChurchofStMaryandnotwithan
olderchurch.Thismaysuggestthatthewoodworkwasnotreallypartoftheperiod2passage.
Nodirectstratigraphicalrelationcouldbedocumented between the woodwork and the period
2-passage (cf Herteig 1991, 74). Furthermore,
the material from this part of the Bryggen site
isdifficulttointerpret,asthestratigraphy/structuresintheareahavebeenpenetratedbyburialsagainandagain.Thereforeitispossiblethat
the woodwork was not part of the passage betweenrows3and4inperiod2,butratherpart
of a construction in the churchyard itself. The
90
evidencefromthispartoftheBryggensiteistoo
inflicted with uncertainties to be able to carry
theevidencenecessarytodatetheChurchofSt
Maryintotheeleventhcentury.
To conclude it is difficult to ‘build’ a predecessortothestandingStMary’sonthesparse
evidenceavailable.Still,thewallwastherealong
withtheashlarsandiftheyrepresentanearlier
StMary’s,itistoointerestingtobeignored.On
thisbasisLidén’ssuggestionthatStMaryhada
predecessorisfollowed.AccordingtoLidénthis
‘earlyStMary’s’mightdatebacktoabout1100.
Thepresumedchurchwillbeusedasasupplementarysourceforhorizon4.
TheextentofthechurchyardtowardsVågen
isseenatsite6,Bryggen.Theburialsinsquares
R02 and R03 must be contemporary with the
standingStMary’sandaredatedtothetwelfth
centurysincetheyareallstratigraphicallybelow
the Church of St Lawrence, mentioned in the
writtensourcesforthefirsttimein1208,(Lidén
andMagerøy1980,147withreferences).Theextentoftheburialsistakenasthedelimitationof
thechurchyardwhichbelongedtothestanding
StMary’sanditisusedasabasicsourceforthe
extentofthechurchyardinhorizon5.Inanearlierstudy,theextentofthechurchyardtowards
the west, north and east has been estimated to
be10-20mfromthechurchbuilding(Hansen
1994b, 72, Figure 16), this extent is used as a
supplementarysourceforhorizon5.Thematerialcannotelucidateactivityonthesitepriorto
horizon4.
Site24,TheChurchofStPeter(Peterskirken)
The Church of St Peter, or rather the churchyard, is first mentioned in the written sources
inconnectionwitheventsin1183(Ss1920,83;
Holtsmark1961,117).Thechurchruinandthe
churchyard were partly uncovered by KorenWiberg in 1920. The building and churchyard
wallcanbelocalisedthroughhisplans,mapsand
more recent excavations (Hansen 1994b, 77).
Thechurch,measuring11.9mx24-25m,hada
rectangularnaveandthechancelwasofthesame
widthasthenave.Inthe‘westfront’alateRomanesque/earlyGothicportalwasdocumented
andloosebuildingstonesinearlyEnglishGothic
stylewerefoundinthevicinity.Lidénsuggests
thatthelattermayberelatedtoalaterrebuild-
ingofthechurch.Theportalisnotcommented
asprimaryorsecondarytotheoriginalbuilding
(LidénandMagerøy1980,171),butissomewhat
tooyounginstyletobelongtoachurchdating
tothelatetwelfthcentury.Still,Lidénsuggests
thatStPeter’swasbuiltbetweenc1120and1180
onthebasisofatotalviewofthesourcesforthe
twelfthcenturychurchesinBergen(Lidénand
Magerøy1990,11;Lidén1993,74).Ishalltoa
certainextentfollowhisproposal,butsincethere
arenoconcretesourcesthattiethechurchsafely
totheperiodbeforec1170,thechurchandthe
churchyardareusedasasupplementarysource
onlyfor horizon 5.Thematerialcannotelucidateactivitiesonthesitepriortohorizon5.
Site25,TheChurchofStOlavontheHill
(OlavskirkenpåBakkene)
TheChurchofStOlavontheHillismentioned
forthefirsttimeinthewrittensourcesinconnectionwitheventsin1181(Ss1920,57;Holtsmark1961,82).AccordingtoMorkinskinnaand
HeimskringlathechurchwasbuiltbyKingHarald Gille(Gilchrist)afterhisvictoryoverKing
MagnustheBlindin1134-1135(Hkr1893-1901,
III376;Msk400).Alltracesofthechurchwere
gone in the seventeenth century when Edvard
EdvardsenwrotehishistoryofBergen.Theexact
location of the church is thus not known. The
churchmayhavebeenatimberbuilding(Lidén
andMagerøy1980,170).
Thechurchmayserveasabasicsourceforhorizon5sinceitwasapparentlyunderconstructionorinuseintheyearsupuntil1170.Ihave
discussedthelocationofthechurchindetailin
apreviousstudy(Hansen1994b,81-84)andaccordingtotheseresults,thechurchislocatedto
theareaofNedreStølen6.Thelocationofthe
churchservesasasupplementarysourceforhorizon 5.Thematerialcannotelucidateactivity
onthesitepriortohorizon5.
Themiddletownarea
Site26,Finnegården6a(1981)BRM104
Dates
Theexcavationatsite26,Finnegården6a,was
an open area investigation, covering about 40
m2.Thenaturalsubsoilwasreachedduringthe
7Evaluationofthearchaeologicalandbotanicalsources
excavation. A report from 1982, with supplementsfrom1983and1998,withstratigraphical
analysisanddatesbaseduponpottery,TLsamplesanddendrochronologyisavailable(Dunlop
1982(1998)).
In1991and1998newdendrodateswereobtained,allfromtimberwithnosignsofreuse.35
Dating results from the dendro material have
been integrated in Dunlop’s dating framework
forthesite.Dunlopthusconcludesthattheoldestphaseonthesite,phase12,startedc1110/20,
whilephase11startedc1130andendedinthe
1170/71 fire (Dunlop 1982 (1998)). The four
dendrosamplesfromphase12weredatedtobetween1099and1103.Sincetherewerenosigns
of reuse on the dated timbers, the phase may
actually have started earlier than proposed by
Dunlop, perhaps shortly after 1103. The dated
timbers from phase 11 also indicate a slightly
earlier start for this phase than proposed by
Dunlop;thedatesto1112and1118thusindicate
thatphase11startedinthe1120sratherthana
decadelaterassuggestedbyDunlop.Thesedatingsuggestionsarenotinconflictwiththepottery dates from the site. The ceramic evidence
fromthesitesupportsDunlop’sdatefortheend
ofphase11to1170/71.Inconclusion,adateof
phase12to‘after1103’-the1120sandphase11
tothe1120s-1170/71seemslikely,placingphase
12inhorizon4andphase11inhorizon5.The
materialiswell-datedandcanbeusedasabasic
source.
Therearesomeindicationsthatphase12did
not represent the first settlement at or near the
site. Some of the timbers in a triangular logbuilt and stonefilled caisson (K37) from phase
12hadvariousnotches,apparentlywithoutany
functioninthefinds-context.Thenotchessuggestthatsomeofthetimberswerereusedinthe
phase12context(Dunlop1982(1998),43).A
dendro sample from one of the reused timbers
wasdatedin2001andshowedthatthetimber
had been cut shortly after 1090.36 The timber
may originally have been used in a constructionbuiltabout10yearsbeforecaissonK37in
phase 12. K37 was built ‘after 1102’ according
tothepreviouslydatedsamples.Thestructures
fromphase12werebuiltonabeachdepositwith
tracesofhumanactivities.Oneoftheseactivities
musthavebeentheconstructionofthephase12
91
structures(Dunlop1982(1998),43),butsome
oftheactivitiesmayalsopredatethephase.With
thepresenceofthereusedtimbersandthepossibletracesofactivitiespriortophase12,thepossibility of a settlement phase prior to phase 12
seemsplausible.
Thelackofinsitustructurespredatingphase
12canbeexplainedbythesite’slocationinthe
tidal zone close to the original shoreline; on
stretches along the Vågen shoreline, with welldated traces of settlement prior to horizon 4,
structures did not extend all the way down to
thesea,butwerefoundsome25-30mfromthe
shoreline(site6).HenceIsuggestthattheareain
thevicinityofsite26wasoccupiedpriortophase
12.Thedendrodatefromthereusedtimberin
K37suggeststhatactivitydatesatleasttoshortly
after 1090, thus the pre-phase 12 material can
be assigned to horizon 3. As the location and
thedateofthematerialarenotwell-founded,the
materialisusedasasupplementarysource.No
structures or culture-layers could be associated
uncoveredatthesite:inthenorthernpartK37,a
triangularstone-filledcaisson,andinthesouthernpartK42,interpretedasthefoundationofa
quayorabuilding.Onlythreelayerswererecognisedinphase12,theyweredepositedduringthe
phaseandcanbeassignedtohorizon4,category
I.Phase11,representinghorizon5,wasinitiated
byapartialdemolitionofstructuresfromphase
12,followedbyintensivelayerdeposition.This
was intended to elevate the ground surface for
building. In the northern part of the site K36
wasconstructed,inthesouthernpartK41,and
inbetweenthesetwoconstructionsK38waslocated. The northern and southern parts of the
siteseemtohavebeenbuiltseparatelybutcontemporaneously.K38mayhaveformedpartofa
wickerhurdle.K36andK41arebothinterpreted
assub-constructionssupportingopenareas.Sixteenlayerswereassignedtophase11,twolayers
areassignedtohorizon5,categoryIand14to
categoryII.
Table14.Site26,Finnegården6a(1981)BRM104
Phase
11
12
Priorto
phase12
Archaeologicalevidence
Naturalscientificdates
Pottery
Other
Andenne,Cookingpots,Developed
Stamford,Humber,LondonBrown,
Paffrath,Pingsdorf,SoftFiredBlack
ware,York,
Dendro
1112(1118)
1099,1100,1102
(1103)
1090(reused
timberinphase
12)
Thenatural
subsoil
TL
Dating
14
Horizon Source
type
(B/S/G)
C
1120s-1170/71 5
B
1103/1110-
1120s
‘After10901103/1110’
4
B
3
S
1-2
S
Priorto
horizon3
Databasedon(Dunlop1982(1998))andmyowninvestigations
Dates/potteryinboldaretheyoungestintheconstruction/phase
withhorizons1-2,thisinformationisusedasa Site27,Finnegården3a(1982)BRM110
supplementarysourceforthesehorizons.
The excavation at site 27, Finnegården 3a, was
anopenareainvestigationthatcoveredabout80
Majorfeatures,artefactcategories
m2.Thesitewaslocated18msouthofsite26.
Beforehorizon4,site26waslocatedinthetidal Thenaturalsubsoilwasreachedduringtheexcazonebetweenabout0maslandabout+1masl. vation.Areportwithstratigraphicalanalysisand
Nostructuresorculture-layerscouldbeassoci- preliminarydatesbaseduponpotteryisavailable
atedwithhorizons1and2.Inhorizon3thearea (Golembnik 1993). Dendro samples have been
inthevicinityofsite26mayhavebeensettled. datedafterthereportwasfinished(Golembnik
Inhorizon4,twoconstructions(phase12)were 1993,AppendixIIIandsamplestakenbyReim92
ers and myself in 1997/98). My dates here are
The beginning of phase 2 is dated to ‘after
based upon the dendro samples, pottery from 1144’;thebeginningofthesucceedingphase3
site27,anddatesfromthenearbysite26.
is,however,datedby8dendrosamplestoaslate
as‘after1213’.Thesampleswerealltakenfrom
Dates
structuralelements(Golembnik1993,Appendix
Theoldestphaseatsite27isphase1whereonly III),andthedatesoughttobereliable.Thedatculture-layersandnostructureswerefound(Go- ingframeforphase2isaccordingly‘after1144’
lembnik1993,8-10).Theterminationofphase toc1213.Phase2issub-dividedintosixstages.38
1canbedatedindirectlybyevidencefromphase At the beginning of the phase, three caissons
2:adendrosamplefromcaisson53,builtinthe supported by levelling layers were built simulbeginningofphase2,anddatedto‘after1144’. taneously. These constructions and layers conThedatedlog37showednosignsofreuseandthe stitute stages 1-3 in phase 2 (Golembnik 1993,
outer tree rings were intact, thus giving a reli- 11-18).Fromstage4thefirstsherdsof‘Grimston
abledateforthebeginningofphase2.Theend ware’appear(Golembnik1993,TableVII),acofphase1shouldaccordinglybedatedpriorto cording to the traditional dating this ware was
c1144.FindsofPaffrathsherdsprovideawide notproducedbeforetheendofthetwelfthcendateforactivitiesinphase1tothetwelfthcen- tury(Reed1990,31).Stages4-6maytherefore
turyorlater,withoutgivingaclosedateforthe representactivityatthesitefromtheendofthe
beginningofthephase.Atthenearbysite26,ac- twelfthorthebeginningofthethirteenthcentivitiesmayhavestarted‘after1090’.Becauseof tury,thusitislikelythatstages1-3representthe
theclosenessofthesites,itislikelythattheold- thirdquarterofthetwelfthcentury.Stages1-3
estculture-layersinphase1atsite27maycor- areusedasabasicsourceforhorizon5.
respondtoorresultfromactivitiesinthephase
priortophase12orinphases12or11atsite26. Majorfeatures,artefactcategories
Phase 1 at site 27 should therefore be dated to Site 27 was situated below sea level between
betweenc1090andc1144,andmayrepresent about-0.5and-1.7maslwhenthefirsttracesof
horizon3andhorizon4.Thematerialinphase human activities were accumulated in phase 1.
1isonlydatedindirectlyandcanonlybecon- Nostructuresorculture-layerscouldbeassigned
sidered as a supplementary source. In the lay- tohorizons1and2.Inhorizons3and4,phase
ersunderphase1,nowaste-layerspredatingthis 1 layers, that layers representing detritus from
phasewereobserved(Golembnik1993,8).This constructionworkontheshore,weredeposited
informationisusedasasupplementarysource perhapsasadeliberatelevellingofthearea.The
forhorizons1and2.
initial levelling in phase 1 was followed by six
Table15.Site27,Finnegården3a(1982)BRM110
Archaeologicalevidence
Phase
2,stages
4-6
2,stages
1-3
1
Naturalscientificdates
Pottery
Other
Dendro
Andenne,Grimston,
Paffrath,Pingsdorf,
LondonShelly,Dev
Stamford,York
Andenne,Paffrath,
1144
Pingsdorf,LondonShelly,
York
Cookingpot,Paffrath
Closeness
tothe
Finnegården
6aBRM104
site
TL
Thenatural
subsoil
14
Dating
Horizon
Sourcetype
(B/S/G)
C
Beginsattheend
oftwelfthc
Endsc1213
Beginsafter1144 5
Endsinlate
twelfthc
Beginsafterc
3-4
1090
Endsc1144
Priortophase1
1-2
B
S
S
DatabasedonGolembnik1993andmyowninvestigations
Dendrodatesinboldaretheyoungestintheconstruction/phase
7Evaluationofthearchaeologicalandbotanicalsources
93
layersaccumulatedduringthephase(Golembnik
1993,9-10).Whetherthelayersweredeposited
duringhorizon3orduringhorizon4isunclear.
Thefindsfromtheselayersarethereforeassigned
ascategoryIIfindstohorizon4.Inhorizon5/
phase2,stages1-3,three2mx2mcaissonsand
two mooring posts were constructed supported
by 25 levelling layers. Finds from the levelling
layersareassignedtohorizon5categoryII.
Site28,Rosenkrantzgaten4(1978/79and
1981)BRM76
The excavations at site 28, Rosenkrantsgaten
4,werecarriedoutintwocampaigns,covering
altogether about 450 m 2 (Lindh 1979; Ekroll
1981). Only the material from 1978/79 is relevant to my study. A report with an account
of the stratigraphical relationship between the
buildings is available (Lindh 1979). The level
of the natural subsoil was documented during
theexcavation.
Dates
Ihaveearlieranalysedmaterialfromsite28inordertodatetheoldestphases.Typologicalresemblancebetweenthestructuresatsite28/phase1,
site6,phases2.1and2.2,andthestructuresat
site27,-andasimilarnumberofphasesatthe
sites,indicatedthatthestructuresfromphase1
atsite28werecontemporarywithphase2.1at
site 6 and phase 2 at site 27. However, dendro
samplesofthestructuresinphase1atsite28produceddatesthatweresomewhatolderthanthe
typologicaldateprovidedbysites6and27and
prevailing at that time (Hansen 1994b, 51 and
Ekskurs3).Asmynewevaluationoftheoldest
materialatsites6and27(cfp85ffandp132ff)
hasprovidedanearlierdateforthesephases,the
typologicaldateofthestructuresatsite28isalso
olderandnowcorrespondswiththedendrodate
of the phase. Accordingly, phase 1 at site 28 is
datedasfollows:phase1wasbuiltintwostages,
astone-layercalledAmarkstheendofthephase
(Lindh1979,5-9).39Thebeginningofthefirst
stageisdatedbydendrochronologyto‘after1128’
andthebeginningofthesecondstageisdatedto
‘after1141’.Thedendrosamplesweretakenfrom
structural elements and seem reliable. I assume
thattheconstructionsinphase1arerepresentativefortheyearsupuntilthelastquarterofthe
94
twelfthcentury,andprobablytotheendofthe
century.Thematerialfromphase1/stages1and
2isthereforeusedasabasicsourceforhorizon
5.Thustheymayserveasasourceforhorizon5.
The documented data cannot elucidate activity
onthesitepriortohorizon5.
Artefactcategories
Inordertobeabletouseartefactsfromphase
1 as a source for horizon 5, I have identified
deposits that ought to belong to the period
before1170.Thestone-layerdenotedAinthe
reportwasdepositedatthesitewhenthecaissonswentoutofuse(Lindh1979,8)andprovides a fairly reliable upper limit for artefact
assemblages,whichmaybeassociatedwiththe
period when the caissons were in use. Having studied artefact assemblages associated
with the caissons and located under layer A,
Ihavedismissedallassemblageswithmaterial
tooyoungfora‘pre-1170context,andIhave
sortedouttheassemblages,whichwerefound
on the same level as assemblages with ‘post1170 material’. The remaining artefacts may
be identified as probable horizon 5 material
and they are assigned to category II as they
havebeendepositedinfill-masses.
Majorfeatures,artefactcategories
Thematerialfromthesitecannotelucidateactivitiesonthesitepriortohorizon5.Inthebeginningofhorizon5site27waslocatedintheVågenBayatabout-2.0maslandabout-1.5masl.
Phase1atthesitewasbuiltupintwostages.The
firststagecomprisescaissons2and6,interpreted
asthefoundationofapier(Lindh1979,7).The
pierwasprobablybuiltshortly‘after1128’.After
the construction of the pier, layers were deposited in the area. Later, ‘after 1141’ yet another
set of caissons was built. One caisson is interpretedasarepairofthestage1pierthatwasin
useuntiltheendofthephase.Threeothercaissonsareinterpretedasfoundationsforbuildings.
Eight mooring posts also belong to the second
stageofthephase.Artefacts,assignedtohorizon
5havebeenidentifiedthroughtheirrelationship
tothestructures.Alltheidentifiedartefactsare
assignedtohorizon5ascategoryIIfindsasthey
weredepositedinfill-masses.
Table16.Site28,Rosenkrantsgaten4(1978/79)BRM76
Phase
1,Stages1
and2
Archaeological
evidence
Pottery
Other
Naturalscientificdates
Dating
Dendro
TL
1127-11281137-1141
14
Horizon
Sourcetype
(B/S/G)
C
Beginsafter1128
5
Endsafter1141/Late
twelfthc?
B
DatabasedonHansen1994b
Dendrodatesinboldaretheyoungestintheconstruction/phase
Site29,Vetrlidsalmenningen2,Kjøttbasaren
(1996and1997)(BRM490)
Site 29, Vetrlidsalmenningen 2, comprised a
numberofexcavatedareas,onlytheareacalled
“hul2”(about4.5m2)isrelevantinmyconnection.Theoldestdocumentedmaterialin‘hul2’
wasphase6,astheinvestigationdidnotproceed
beyond this phase and the natural subsoil was
notreached.Inphase6twotimberuprights,K9
andK17,werefound.Theyareinterpretedasin
situpolesina2mx2mcaissonofatypewelldocumented at many other sites (site 6, 9, 27,
28)Inthefollowingphase7aredepositedbeam
wasfound.Thisislikelytobefromthephase6
caisson. No dating material was available from
phase 6, however the redeposited beam from
phase 7 was dendro dated to ‘after 1128/29’.40
On the basis of the parallel material from the
neighbouring site 27 Dunlop dates the start of
phase6toc1175(Dunlop1999).Accordingto
the newest dendro dates from site 27, caissons
at this site were, however, dated to ‘after 1144’
(see above), thus making Dunlop’s c 1175 date
forthebeginningofphase6somewhatlate.The
caissonsatsite27haveparallelsatsite6andat
site28whereseveralcaissonsaredendrodatedto
the1120s.Adateto‘the1120s’correspondsvery
wellwiththe‘after1128’dateprovidedbytheredepositedcaissonbeamatsite29.Ifindithighly
likelythatthecaissoninphase6atsite29should
be dated to ‘after 1128’. The caisson thus representshorizon5atthissite.Sincethecaisson
isdatedindirectly,througharedepositedtimber
in the following phase, the caisson is used as a
supplementary source. There were no artefact
findsinphase6.Layerswerenotdocumentedin
anydetailandtheyareomittedfromthisstudy.
Thematerialcannotelucidateactivityonthesite
priortohorizon5.
Site30,Vetrlidsalmenningen(1991/92)
BRM342
The excavation at site 30, Vetrlidsalmenningen
BRM 342, comprised 69 profiles in trenches
andtwoopenareasofapproximately4m2(‘V3’)
and9m2(‘V5’).Areportfor‘V3’withastratigraphicalanalysisanddatesbaseduponpottery
isavailable(Hansen1992).AreportfortheremainingpartoftheinvestigationisinpreparationandDunlop,theexcavationsupervisorhas
kindlyprovidedthepreliminarymanuscriptfor
mydisposal.Itcontainsastratigraphicalanalysis
ofstrataanddatesforstratigraphicalsequences
in the material are suggested, based upon pottery,14C,dendrochronologyandthestratigraphicalrelationshiptoinvestigationsinthevicinity.
Botanical material was collected in connection
withtheinvestigationsbutnoreportisavailable.
Thenaturalsubsoilwasreachedinseveralprofilesandopenareas.
Profiles9-13
Inprofiles9-13(calledanalyticunit30/A,cfp
65ff)constructionsK26,K27,K28,K34,K41,
Table17.Site29,Vetrlidsalmenningen2,Kjøttbasaren(1996and1997)BRM490(NIKUprojekt22321)
‘Hul2’
Phase6
Archaeologicalevidence
Naturalscientificdates
Pottery
Dendro
TL
Reusedbeam
infollowing
phase:1128
Other
Typologicalcoherencewiththe
site6period2.0,site27phase2
andsite28phase1material
14
Dating
Horizon
Sourcetype
(B/S/G)
C
Beginsafter 5
‘1128’,ends?
S
Dendrodatesinboldaretheyoungestintheconstruction/phase
7Evaluationofthearchaeologicalandbotanicalsources
95
K42,K47,K50,K231areinterpretedaspartof
apierandrepresenttheoldestphaseofactivity
here. The post K41 was dated through a combinationof14Canddendrochronologytoc900.
Therewasstillbarkonthepost,indicatingthat
it was in situ. The pier was later incorporated
inayoungerstructure(K49/48),interpretedas
‘somekindofwall,possiblyintendedtostopthe
spreadingofdumpedlayersoreventokeepthe
streaminitschannel’,thisstructurepresumably
burntattheendofthetwelfthcentury(Dunlop
inprep).Judgedbythelocationofcaissonsassignedtohorizon5atsites27,28,and29,the
shallowbaybysite30oughttohavebeenalmost
filledoutduringhorizon5.Itisthuslikelythat
thepierdidnotfunctionasapieranymoreduring horizon 5. Until horizon 4 it may however
havebeenstandingfreelyinthebay.Asitdidnot
fallintodisrepair,itwasprobablyusedasapier
untilitwasincorporatedinthe‘wall’structure.
Consequentlythepierisassignedtohorizons14asabasicsourceandthewallstructureisassignedtohorizon5asabasicsource.Sincenone
ofthelayersassociatedwiththestructurescontained artefacts and the analysis of these layers
isnotyetcompletedinthereport,Ihavechosen
nottoincludelayersfromthispartoftheexcavationinmyinvestigation.
Profiles14-19and23
Inprofiles14-19and23(analyticunit30/B)a
number of fluvial layers were found above the
natural subsoil. They show that a small river
haditspathhereorintheclosevicinity.Dunlop
divides the earliest material into horizon VIII,
phasesA,B,andCandsuggestsc900-1198asa
datingframeworkforthelayers.PhaseAended
inafiredatedto1198.Thefluviallayerscontain
anumberofartefactsthatcanbeusedasasource
inthepresentstudyifweareabletonarrowdown
thedatesforthethreephases.Accordinglyitis
important to discuss which layers/phases may
represent the period before c 1170 and in turn
mayserveasasourcehere.Thereishardlyany
dating evidence, but if as a point of departure
phaseAendedinthe1198fire,thisprovidesan
upper limit for the date of phases A, B and C.
We can assume that the deposition of phase A
took‘sometime’.A14Csample41takenfromnutshellsfromphaseBisdatedtobetween1030and
96
1190. This implies that the upper limit for the
date of the phase B material is found before c
1190andcorrespondswiththeassumptionthat
thedepositionofphaseAtook‘sometime’.The
14
CdatealsoimpliesthatthedateforthedepositionofphaseBdidnotgobacktothe900s,but
mayratherbefoundlater‘afterc1030’.Thismay
suggest that phases B and C are representative
forourhorizons2and/or3and/or4and/or5.
Since it is not possible to specify which of the
horizonsthematerialmayactuallyrepresent,the
artefactsfromthelayersinphaseCandBwill
allbeassignedascategoryIIfindstohorizon5
(cfp68ff).Asthedatingofthematerialisweak,
thematerialisusedasasupplementarysource
only. The presence of the stream is assigned to
horizons1-5asasupplementarysource.
Profiles26,27,28and29
Inprofiles26,27,28and29(analyticunit30/
C)constructionsK96,K105,K109,K114,K115
and K124 are interpreted as a bridge over the
small river and built about 1150. The material
is therefore assigned to horizon 5. The bridge
is not well-dated and can be used as a supplementarysourceonly.Itisneitherpossibletodeterminehowlongthebridgewasinuse,norto
determinewhichoftheassociatedlayersshould
representhorizon5.Ihavechosennottoinclude
layersfromthispartoftheexcavationinmyinvestigation. The bridge represented the oldest
phase of structures above the natural subsoil.
This information is used as a supplementary
sourceforhorizons1-4.
Profiles36-40
Profiles36-40(analyticunit30/D)showstratigraphical sequences of pits, layers and a ditch/
channel dug into the natural subsoil. The sequencesarereferredtoasA-E.A:Thelowermost
andoldeststructuresarethreepostholesK147149,allstratigraphicallyolderthan11layers:B:
519,522-524,562,564-567,610and609.These
layersarestratigraphicallyolderthansevenpits
insequenceC:K139,K154-K158andK192and
oneditch/channel,K142,whicharestratigraphicallybelowsequenceD:thepitK167.Sequence
E:fire-layer588/619coverspit167.Thepostsin
sequenceAareinterpretedaspartofonestructure,thelayersinBareinterpretedaslandclear-
anceandcultivationlayersandthepitsinCand
Dareinterpretedassandextractionholes.
Dunlop suggests that E, the fire-layer, may
beidenticalwiththe1248townfire,thusgiving
an upper date for the sequence. Two 14C dates
fromsequenceCaredatedto980-1160ADand
1020-1180AD.A14Cdatefromoneoftheeleven
layersinsequenceBisdatedto1000-1160AD.
The 14Csamplesaretakenfromcharcoalinfillmassesfromtheconstructions,thecontextisnot
clearandthesamplesmaybeinflictedbythe‘old
wood problem’. The wide dates provided may
thereforebeevenwider.SequenceC,pitK158,
contained two sherds of Andenne ware and pit
K192onesherdofunknownprovenance.42Andennewareisproducedfromtheeleventhcentury(Reed1990,38)andlikethe14Cdates,the
sherds provide a wide date. The fabric of the
sherd of unknown provenance is rather hard
fired,oxidisedandwhite/grey,andwithoutglaze
orpaint.Becauseoftheratherhardfiredfabric
the sherd should be attributed to the twelfth
ratherthantheeleventhcentury.The 14Cdates
fromsequenceCindicatethatthestructuresare
mostlikelyfromc1160orolder.Thestructures
inCmay,therefore,belaterthanc1100andolderthanc1160.ThesequenceoflayersinBmay
alsobelongtothetwelfthcenturyortheymay
beolder.ThethreepostholesinAcanonlybe
datedrelativelyandareolderthanthestructures
inB.Theymaybecontemporarywiththepier
inprofiles9-12andactivitiesassociatedwiththe
pier,butthisisonlyaneducatedguess.Nolayers
have been associated with the use of the posts
andthefunctionofthestructuresisunclear.
Inconclusion:ithasnotbeenpossibletodivide
thesinglestructuresfromsequencesAtoCinto
the system of horizons. However, the activities
representedinBandCmaygenerallybecharacterisedasextensivelanduse.Inhorizons4-5the
area around profiles 36-40 was thus characterisedbynon-intensivelanduserepresentedbythe
structures in sequences B and C. The material
canbeusedassupplementarysourcesonly.BeforethisthepostsinsequenceAsuggestthatthe
areawasoccupied.Thetracesofoccupationmay
go back into the eleventh century, representing
horizons1-3.Sincethematerialfromsequence
A is merely dated relatively as older than the
materialfromB-Citisusedasageneralback7Evaluationofthearchaeologicalandbotanicalsources
groundsource.Thelayersincludedhereareonly
theartefact-yieldinglayerssincethereportdoes
nottreatallthelayersassociatedwiththestructures. Only three layers contained finds, all of
whicharefoundinsequenceC.Thelayersmay
havebeentransportedfarbytheriverbeforethey
endedwheretheywerefound.Thereforetheyare
assignedtocategoryIIandtreatedwiththehorizon5material(cfp68).
V3
InV3(analyticunit30/E)astructureinterpreted as a separation basin43 was the oldest structureabovethenaturalsubsoil.Thebasinseems
tohavebeenusedandcleanedupinseveralstagescoveredbyphases11to4.Thesinglephases
cannotbedatedwithanyaccuracy.However,a
widedateof‘afterc1100’forphases7-4,awide
dateofactivitiesinphase3to‘aftertheendof
thetwelfth/beginningofthethirteenthcentury’,
andatentativedateoftwelfthcenturyforphases
11-8(Hansen1992),showthatthebasinwasin
useformostofthetwelfthcentury.Thematerial can count as a source for horizons 4 and
5.Phase10,basin1/K208/K214mayrepresent
horizon4andthephase4basin,K146mayrepresenthorizon5.Asthephasesarenotsecurely
dated,thematerialwillbeusedasasupplementarysource.Thelackofculture-layersandstructuresthatcouldbeassignedtohorizons1-3is
usedasasupplementarysourceforthesehorizons.Thelayersweretransportedbywatertothe
site,andtheydonotreflectthefunctionofthe
basin,ratheractivityinthevicinityandtheyare
assigned to category II. Since the phases/layers
cannotbedatedmorepreciselyIfinditreasonabletoassignthelayersfromphases11-7tohorizon4/categoryII.Layersfromphases6-4are
assignedtohorizon5/categoryII.
Site 31, Øvregaten/Finnegårdsgaten (1979
and1980)BRM86
Theexcavationsatsite31,Øvregaten/Finnegårdsgaten, were carried out in 1979 and 1980, and
comprised three profiles at the Kristi Krybbe
School(1979)andanopenareainvestigationof
about150m2(1980).Areportcoversbothinvestigations.Stratigraphicalanalysisisgivenforthe
material from the open area investigation and
datesarebasedonceramicmaterial.Themate97
Table18.Site30,Vetrlidsalmenningen(1991/92)BRM342
Profile:Strata
Profiles9-13:K27/K28/K41/
K42/K47/K50/K231
Profiles14-19and23:
layers:176/188/
189/190/194,177,179,181,
220-223,238-241,274-277,
224-232.
Profiles26-29:
K96/K105/K109/K114/K115/
K124
Archaeologicalevidence
Naturalscientificdates
Pottery
Dendro TL
C890
Other
Stratigraphical
relationshipwith
fire-layerdated
totheendof
twelfthcentury.
Stratigraphical
relationshipwith
fire-layerdated
totheendof
twelfthcentury
Thenaturalsubsoilinprofiles
26-29
Profiles36-40:
Andenne,
K139,K142,K154-158,K192, anda
andcultivationlayers:519, possible
522-524,562,564-567,609, twelfth
610
century
sherd
Profiles36-40:
Stratigraphical
K147/K148/K149
relation-shipto
twelfthcentury
structures
Phase4basinK146and
Stratigraphical
artefact-yieldinglayersfrom
relationshipto
phases6-4:292,294,314,
latetwelfth/
316,324,332,334,335,338,
earlythirteenth
348,350,356,378,428,429,
century
456,457,572
structures
V3andprofiles56-57:
Stratigraphical
Phase10Basin1:K208/K214
relationshipto
andartefact-yieldinglayers
latetwelfth/
fromphases11-7:297,300,
earlythirteenth
331,462,490,507,612,613,
century
616,786,791
structures
ThenaturalsubsoilinV3and
profiles56-57:
14
C
Ca8901020
Dating
Horizon Source
type
(B/S/G)
Constructedc900
1-5
Inuseuntilendof
twelfthc
Representativefor
5
theyearsuptoc1170
Priortohorizon5
980-1160 Representativefor
and
activityfromc1100
1020toc1170
1180
Representativefor
activitybeforec
1100?
B
S
5
S
1-4
S
4-5
S
1-3
G
Representativefor
5
activityinthemiddle
andendofthe
twelfthcenturyuntil
c1170
S
Representative
foractivityinthe
beginningofthe
twelfthcentury
4
S
Priortoc1100
1-3
S
Databasedon(Hansen1992;Dunlopinprep)
rial from the profiles is only described and not
dated(Christensson1980c).Thenaturalsubsoil
wasreachedatbothexcavations.
Theoldeststructuresandculture-layersfrom
the open area investigation were dated broadly
to1250(Christensson1980c,25).Burialsfrom
theChurchofStMartinwerefoundintheprofilesrepresentingthefirstactivityonthispartof
thesite,anddatingfromthemiddleofthethirteenth century at the earliest, according to the
date of the church (Lidén and Magerøy 1980;
Hansen1994b).Theoldestactivitytraceswere
found above morainic masses and information
98
onthelackofoccupationcanbeusedasasupplementarysourceforhorizons1-5.
Site32,TheChurchofStNicholas
(Nikolaikirken)
The Church of St Nicholas is mentioned for
the first time in written sources in connection
with events in 1160 (Hkr 1893-1901, III 417),
thechurchwasalsomentionedinwrittensourcesinconnectionwitheventsthattookplacein
1181and1183(Ss1920,54,83).In1895Bendixeninvestigatedpartsofthechurch,thusthe
location of the church is known. According to
Lidén’sanalysisofthematerial,basedonstylistic
evidence,thechurchbelongstothesamegroup
ofbuildingsasStMary’sandStCross,andLidén
suggests that St Nicholas’s may have been constructedshortlyafter1130(LidénandMagerøy
1990, 99). This makes the church a basic
source for horizon 5. According to a sixteenth
century chronicle, Bergens Fundas, the church
wasfoundedbyKingØysteinMagnussonwho
reignedfrom1103to1123(LidénandMagerøy
1990,99).StNicholas’smayhavebeenabasilica,atowertothewestwasofthesamewidthas
the nave (Lidén and Magerøy 1983, 160). The
largewesttowerindicatesthatthechurchhada
galleryforthechurchpatron(Lidén1993,79).
Thissupportsthesuggestedconnectionbetween
thekingandthefoundationofthechurch,and
thechurchmaythereforealsoserveasasource
for horizon 4, as the written source is remote
intime,thechurchisusedasasupplementary
sourceforhorizon4.
Ihaveearlierdiscussedtheorientationofthe
church and the location of the churchyard on
the background of Bendixen’s publication and
recentexcavations(Hansen1994b,72-77),and
basemylocationofthechurchaccordingtothis
study.Theextentofthechurchyardistreatedas
asupplementarysourceforhorizon4andasa
basicsourceforhorizon5.Thematerialcannot
elucidateactivitiesonthesitepriortohorizon4.
Site33,TheChurchofStColumba(Steinkirken)
It is commonly assumed that Steinkirken, (the
StoneChurch),isidenticalwiththeChurchof
St Columba (Lidén and Magerøy 1983, 139).
Steinkirkenismentionedforthefirsttimeinthe
written sources in connection with events that
took place in 1181 (Ss 1920, 54; Holtsmark
1961,82).TheChurchofStColumbaismentioned for the last time in medieval sources in
1427(LidénandMagerøy1983,139).Theexact
location of the church is not known. Based on
anevaluationofthesourcesofthetwelfthcentury churches in Bergen Lidén suggests that St
Columba’swaserectedbetweenc1120and1180
(LidénandMagerøy1990,11;Lidén1993,74),
however,sincetherearenoconcretesourcesthat
tiethechurchtotheperiodbeforec1170Ishall
usethechurchasasupplementarysourceonly
forhorizon5.ElsewhereIhavediscussedthelo7Evaluationofthearchaeologicalandbotanicalsources
cationofStColumba’sonthebasisoftheavailablewrittenandarchaeologicalsources(Hansen
1994b,84-87),inthepresentstudythelocation
isbaseduponthisdiscussion.Thechurchisthus
tentativelylocatedtotheareaaroundØvregaten
15-17.Thelocationofthechurchmaybeusedas
asupplementarysourceforhorizon5.Thematerial cannot elucidate activity on the site prior
tohorizon5.
Thesoutherntownarea
Site34,LilleØvregatenfriområde(1994)BRM
465
The excavation at site 34, Lille Øvregaten, was
an open area excavation that covered 33 m2. A
report with a stratigraphical analysis and dates
baseduponpotteryand14Cisavailable.Theoldestmaterial, 14Cdatedtobetween795and410
BC, is from the late Bronze Age or early Iron
Age.Thereisabreakinactivitiesuntiltheoldest
medievalremainsabovethenaturalsubsoilappear,theyweredatedtothethirteenthcentury
(Hansen1995b).Thelackofstructuresandculture-layersduringhorizons1-5isusedasasupplementarysourceforthesehorizons.
Site35,Korskirken(1984)BRM200
Theexcavationatsite35,KorskirkenBRM200,
was an open area investigation, covering about
10m2andlocatedtothechurchyardofStCross
(site39).Ihavediscussedthedatingoftheoldest
materialfromthissiteinapreviousstudywhere
I concluded that the oldest deposits (phase 1)
above the natural moraine could not represent
thetwelfthcentury.Thephasewasdatedtoafter
c1250throughthepresenceofSaintongeware
and Proto Stoneware (Hansen 1994, 62-64).
Thematerialdoesnotindicateactivitiesinhorizons 1-5. This information is used as a supplementarysourceforthesehorizons.
Site36,Skostredet10(1992)BRM346
Theexcavationatsite36,Skostredet10,wasan
openareainvestigation,whichcoveredabout160
m2. No report is available from the excavation.
Excavation supervisor Andrzej Golembnik has
kindly provided oral information of relevance
here(cfHansen1994,65).Thesitewasoriginal99
lylocatedabout30mfromthenorthernshoreof
Vågenatabout-1.5maslclosetoariverbed.The
oldestphaseofstructuresisdatedtotheendof
thethirteenthcenturythroughthepreliminary
examinationofpottery.Priortophase1,organic
layersweredeposited.Thelayershavenotbeen
datedandmaythusbefromanytimebeforethe
endofthethirteenthcentury.AccordingtoGolembnik,thesedepositsmusthaveaccumulated
overalongtimeanddonotrepresentanintentional dumping of masses. The layers differ in
characterfromtheharbour-relatedlayers,which
Golembnik was familiar with from his excavationsatsite27andtheydidnotcontaintypical
harbour related finds, such as rope and ballast
stone. His preliminary conclusion is that there
was no harbour here before the first constructions were built in phase 1, and that the shore
wasnotusedasawaste-dump.Thisinformation
isusedasasupplementarysourceforhorizons
1-5.
Site37,NedreKorskirkealmenning/
Vågsalmenning(1998)BRM544
The excavation at site 37, Nedre Korskirkealmenning/Vågsalmenning, was an open area
andtrenchinvestigation.Someofthebotanical
materialisrelevanttothepresentstudy(Hjelle
1998).
Dates
The relevant botanical sample was taken in a
profile about 60 m from the nearest (eleventh
century-)shoreofVågen,atadepthof-3.1masl,
correspondingtophase6atthesite.Thelower
part of layer 40 contained pollen zone 3a, the
upperpartofthelayercontainedpollenzone3b.
Thebeginningofpollenzone3bwas 14Cdated
tobetweenAD810and970.Asthe 14Csample
wastakenfromthebeginningofpollenzone3b,
itislikelythattheactivitiesrepresentedbypollen
zone3beganearlierandlastedlongerthanthe
dateimplies(Hjelle1998).Thebotanicalmaterialfrompollenzone3,phase6,maybeusedasa
sourceforhorizon1.Asthematerialisdatedby
14
Citmaybeusedasasupplementarysource.
Afterphase6therewasabreakintheaccumulationoflayers.Thesucceedinglayer/phase5was
datedby 14Ctobetween1295and1420.Other
sources, however, document that there was ac100
tivityintheVågsbunnenareabeforetheendof
thethirteenthcentury.Therefore,thelackofdepositsonthesitemustreflectthedeposition-or
preservationconditionsonthelocationbetween
phases6and5.Thelayersabovephase6arenot
usedasasourceinthepresentstudy.
Generallandusereflectedinthematerial
Layer 40, a fast-accumulated marine sediment,
consistedofdarkbrowngyttjawithonlyasmall
amount of macroscopic material (Hjelle 1998,
Section5).44Pollenzones3aand3bcontained
pollen from a variety of grain, herbs and other
plants. From the beginning of zone 3b, pollen
indicatingtheimportofgrainwaspresentinthe
sample.45AccordingtoHjellethefastaccumulation of layer 40 may be explained either as an
intensificationofagriculturallanduseorasthe
depositionofwasteinthesea.Duetothepresenceofimport-indicatingpolleninzone3b,she
concludesthatthelayermostlikelyaccumulated
asaresultofwaste-dumpingintheseaatsome
distance from the area of investigation (Hjelle
1998).OnthisbasisHjellesuggests‘somekind
ofdensersettlement’intheVågsbunnenareain
theVikingperiod(Hjelle1998).
Thepresenceofimport-indicatingpollenina
sample does not in itself illuminate on the history of deposition of a layer. Household waste
with or without import-indicating pollen may
have been used to manure fields. In Chapter 4
I have argued, on a methodological basis, that
deposits,whichcontainonlypollenandnomacrofossils, positively identified as remains of human waste, cannot count as sufficient evidence
that waste masses were dumped in the close
vicinity of a sampling location (cf p 51ff). Accordingly, the deposit at site 37 cannot stand
alone as evidence that waste was dumped into
theseainthevicinityofthepointofregistration.
At the nearby site 36, no indications of intentional dumping of masses were found, prior to
theoldestarchaeologicalphasedatedtotheend
ofthethirteenthcentury.Site36waslocatedat
adistanceofabout70mfromsite37.Thetwo
sitesarerelativelyclosetoeachother,bothwere
locatedatsomedistancefromthenorthernshore
ofVågen.Altogetherthen,therearenoindicationsthathouseholdwastewasthrownintothe
sea.Ifindthatthequestionofgenerallanduse
cannotbesettledthroughthematerialfromsite
37alone.Iwillresumethisquestiononabroader
basisinChapter8.
The presence of pollen-indicating household
wasteisassignedtohorizon1asasupplementarysource.Datafromthesitecannotelucidate
activityinhorizons2-5.
the stratigraphical level of 463 until phase 10
arefromtheninthcenturyoryounger,butolder
thanphase10.Layer463wasamarinesediment
of naturally deposited gravel with a few wood
chips, the material also comprised import-indicatingpollen.AccordingtoHjelle,thepresence
ofimport-indicatingpollenmayreflectachange
Table19.Site37,NedreKorskirkealmenning/Vågsalmenningt(1998)BRM544
Archaeologicalevidence Naturalscientificdates
‘Phase’
Pottery
Layer40pollen
zone3
Other
Dendro
TL
Dating
C
Between810-970
andsomewhat
earlier
Horizon
Sourcetype
(B/S/G)
1
S
14
Databasedon(Hjelle1998)
Site38,Domkirkegaten6(1987)BRM245
Theexcavationatsite38,Domkirkegaten6,was
an open area excavation, which covered about
300m2.Areportwithastratigraphicalanalysis
anddatesisavailable(Komber,Dunlop,Sigurdsson, and Hjelle 1994). Botanical material was
analysed in connection with the investigations
(Hjelle 1994). The natural subsoil was reached
atthesite.
Dates
The archaeological phases are dated through a
combinationofpottery,14Canddendrochronology (Komber, Dunlop, Sigurdsson, and Hjelle
1994). The oldest activity phase at site 38 was
phase10,dendrosamplesfromphases9and10
provideadateof‘after1128’to‘after1160’for
phase10(cfdatainKomber,Dunlop,Sigurdsson, and Hjelle 1994, 112). Phase 10 thus correspondstohorizon5.Sincethearchaeological
materialiswell-dated,itcanbeusedasabasic
source.
Thebotanicalmaterialfromlayersbelowthe
oldestregularculture-layersinphase10wasnot
dated,exceptrelativelytotheoldestphaseinthe
archaeological material (Komber, Dunlop, Sigurdsson,andHjelle1994).Polleninthesamples
may,however,provideawidedateforsomeofthe
deposits predating phase 10. The relevant pollensamplesweretakenfromseverallayers:508,
507and463,508beingtheoldest,and463the
youngest stratigraphically. In layer 463, pollen
oftheimport-indicatingweedCentaureacyanus
waspresent.Thismayindicatethatlayersfrom
7Evaluationofthearchaeologicalandbotanicalsources
inactivitiesandtheestablishmentofadensersettlementinthearea(Hjelle1994,160,161,164,
167).Frommypointofviewthematerialfrom
site38isinsufficientasevidencethathousehold
wastewasdumpedintotheseainthevicinityof
the site, and as evidence of a settlement in the
vicinity of the site (p 51ff). The material may
just as well represent cultivation, where human
waste was used to fertilise fields, near the site.
In which case the waste indicated through the
import-indicatingpollenmayhavecomefroma
settlement nearly anywhere in the Bergen area.
Since the character of the general land use reflectedinlayer463isambiguous,andthedate
providedbythepollenissowide(horizon1-4),
Ichoosetoomitthematerialasasourceforthe
characterofactivitiesatthesiteduringhorizons
1-4. The lack of settlement at the site prior to
horizon5is,however,usedasasupplementary
sourceforhorizons1-4.
Majorfeatures,artefactcategories
Inhorizons1-4,priortophase10,andinhorizon5,phase10,site36waslocatedinthebeach
zone between +/-0 and 1 masl. In horizons 14,theareawasnotsettled.Inhorizon5asmall
square log built caisson and several posts were
locatedabout5-10mfromthenormalhightide
(Komber,Dunlop,Sigurdsson,andHjelle1994,
72). The excavation supervisor Jochen Komber
suggeststhatthecaissonwaspartofabuilding
andthatfiveofthepostsmadeupaquayfront.
Theremainingstructuresareinterpretedasthe
foundation of a counterbalanced hoist (Komb101
er, Dunlop, Sigurdsson, and Hjelle 1994, 214).
Threelayerswereassignedtohorizon5;(phase
10)layer421wasdepositedfirstbeforeanyconstructionswerebuiltandwasprobablydeposited
whenlevellingoftheareastartedinordertoraise
itabovespringtideorstorms(Komber,Dunlop,
Sigurdsson, and Hjelle 1994, 71). Since we do
notknowwheretheartefactsinlayer421derive
from, they are assigned as category II finds to
horizon5.Duringphase10,layers420and470
were deposited. Layer 420 is thought to derive
fromtheuseoftheconstructionrepresentedby
thesmalllog-builtcaisson(K311).SinceLayers
420and470aredepositedinsitu,artefactsfrom
these layers are assigned as category I finds to
horizon5.
Theextentofthechurchyardisnotknownin
spiteoflaterinvestigationsinthevicinityofthe
church(Hansen1994b,81).Thematerialcannot
elucidateactivityonthesitepriortohorizon5.
Site40,TheChurchofStCross(Korskirken)
TheChurchofStCrosswasmentionedforthe
firsttimeinwrittensourcesinconnectionwith
events, which took place in 1181 (Ss 1920, 54;
Holtsmark 1961, 82). Based on masonry studiesLidénhassuggestedthatthetwelfthcentury
churchwascompletedbefore1160.Hesuggests,
withsomereservations,thattheconstructionof
St Cross may have begun before the constructionofthestandingChurchofStMary,initiated
around 1140 (Lidén and Magerøy 1990, 99).
Table20.Site38,Domkirkegaten6(1987)BRM245
Archaeologicalevidence
Naturalscientificdates
Other
Dendro
9
Youngestpottery
typespresent
DevStamford
10
Andenne
Singlecomb
Phase
Before10,
nosettle-
ment
TL
1115(11541157)
Dating
14
Horizon
C
‘after
1158’-early
thirteenthc
‘after1128’5
‘c1158’
Afterninth
1-4
centurybefore
‘c1128’
1128
Pollenof
centaureacyanus
Sourcetype
(B/S/G)
B
S
DatabasedonKomberetal1994
Dendrodatesinboldaretheyoungestintheconstruction/phase
Site39,TheChurchofStOlavinVågsbotn
(OlavskirkeniVågsbotn)
The Church of St Olav in Vågsbotn was first
mentionedinthewrittensourcesinconnection
with events in 1181 (Ss 1920, 54; Holtsmark
1961,81).Thechurchhasbeeninvestigatedby
LidénwhohasalsoanalysedevidencefromearlierinvestigationsperformedbyBlixinthe1880s
(Lidén and Magerøy 1983). Based upon style,
among other cyma reversa mouldings that have
parallelsinStMary’sandStCross,thetwelfth
century church may have been built between c
1120and1180(Lidén1993,74).ThismakesSt
Olav’sabasicsourceforhorizon5.Thetwelfth
centurychurchhadarectangularnaveandthe
chancel was narrower than the nave. The west
tower was also narrower than the nave. The
twelfthcenturyStOlav’sisincorporatedintoday’sCathedralchurch.
102
TheChurchofStCrosscanaccordinglybeused
asabasicsourceforhorizon5.
Thetwelfthcenturychurchwasarectangular
building,thechancelwasofthesamewidthas
thenave,whichhadnoaisles.Thetwelfthcentury building is incorporated in the choir and
naveofthestandingchurch(LidénandMagerøy
1983, 112). The extent of the twelfth century
churchyardisnotknown,buttheburialground
didnotcontinuesofarnorthastosite35,Korskirken(1984)BRM200untilafterc1230-40
(Hansen1994b,71).Thematerialcannotelucidateactivityonthesitepriortohorizon5.
TheNordnesandNonneseterareas
Site41,Rådstuplass2-3,‘Vestlandsbanken’
(1963)BRM20
Theexcavationsatsite41,Rådstuplass2-3was
anopenareaexcavationthatcoveredabout550
m2.Judgedbytheceramicevidence,theoldest
structures above the natural subsoil date from
the fifteenth or sixteenth centuries (Site documentation,Rådstuplass2-3BRM20).Thisinformationisusedasasupplementarysourcefor
horizons1-5.
Site42,Nygaten2(1991)BRM333
The excavations at site 42, Nygaten 2, was an
openareainvestigation,coveringabout770m2.
Large parts of the site were disturbed by modernactivities.Theoldestculture-layersabovethe
natural subsoil can be dated to the end of the
sixteenth century on the basis of pottery (Site
documentation,Nygaten2BRM333).Thisinformationisusedasasupplementarysourcefor
horizons1-5.
Site43,TheMunkelivBenedictineAbbey
withtheChurchofStMichael
(Munkelivkloster)
AccordingtothewrittensourcestheBenedictine
abbey of Munkeliv was founded by King Øystein Magnusson (1103-1122) (MHN 64; Hkr
1893-1901,III284p;Msk352;Ågr94),historians seem to agree on this point (Helle 1982,
137-139).OnstylisticgroundsLidénalsofinds
itlikelythatthechurchbuildingmaydateback
tothe1120s(LidénandMagerøy1990,73-87).
TheAbbeyofMunkelivmaybeusedasabasic
sourceforhorizons4and5.
Thelocationofthechurchandchurchyardis
wellknownthroughNicolaysen’sexcavationsin
the 1860s and observations in connection with
constructionwork.Thetwelfthcenturychurch
wasalongchurchwithanapsidalchancelatthe
east,itwasabout35mlongandabout14mwide
(LidénandMagerøy1980,150-151).Thelocation of the church and the churchyard is used
asabasicsourceforhorizons4and5.Munkebryggen,aquayonthesouthernshoreofVågen,
belongedtotheabbey.Theexactlocationofthe
quayisnotknown,butitisnotunreasonableto
assumethatitwasplacedontheVågenshoreline
7Evaluationofthearchaeologicalandbotanicalsources
closetotheabbey.Ihavenottriedtolocalisethe
quayinmoredetail.Thematerialcannotelucidateactivityintheareabeforehorizon4.
Site44,StJohn’sAugustinianAbbey
(Jonskloster)
The date of the foundation of the Augustinian
AbbeyofStJohnisnotknown.Lidénhasargued
thattheabbeywasfoundedinthe1150sasthis
coincideswiththeestablishmentofchaptersconnected to the cathedrals in Norway (Lidén and
Magerøy1980,142).Informationinthewritten
recordsalsoimpliesthattheabbeywasfounded
in the 1150s (Helle 1982, 6, 142). Stylistic detailsonbuildingstonesfromthechurchdemonstratebuildingactivityonthemonumentinthe
1180-90s(LidénandMagerøy1980,143).This,
however,isnotincompatiblewithanearlierfoundationoftheabbey.Itisnotunusualthatafter
thefoundationofanabbey,theconstructionof
buildings was carried out over a long period of
time(cfEide1986;Hommedal1987).Basedon
implicationsinthewrittensourcestheabbeyof
StJohnmaybeusedasabasicsourceforhorizon
5.StJohn’swaslocatedontheNordnespeninsula. Church-foundations were located in connection with groundwork in 1895: The church
tower was found in the area around Fortunen
2,otherfoundationswerefoundatStrandgaten
10-12. The churchyard was located north and
eastofthechurch,andburialshavebeenfound
on several occasions (Lidén and Magerøy 1980,
142-144).Thelocationoftheabbeythusseems
certainandmaybeusedasabasicsourceforhorizon5.Sincethebuildings,whichrepresentthe
periodcoveredbyhorizon5,arenotknown,only
theapproximateareacoveredbythelatermonumentisusedasanillustrationoftheabbey.Jonsbryggen,apier/quaybelongedtotheabbey.The
locationofthequayisnotknown.Mostlikelyit
was placed on the Vågen shoreline close to the
abbey(cfHelle1982,290).Aspecificlocalisation
ofthequayisnotattempted.Thematerialcannot
elucidateactivityintheareapriortohorizon5.
Site45,TheChurchofAllSaints
(AlleHelgensKirke)
The Church of All Saints is first mentioned in
the written sources in connection with events
in 1181 (Ss 53). Based on an evaluation of the
103
Table21.Sitenumber,streetaddress/monument,museumnumber
Site
Streetaddress/monument
number
Site1
Koengen
Site2
Site3
Site4
Site5
Site6
Site7
Site8
Site9
Site10
Site11
Site12
Site13
Site14
Site15
Site16
Site17
Site18
Site19
Site20
Site21
Site22
Site23
Site24
Site25
Site26
Site27
Site28
Site29
Site30
Site31
Site32
Site33
Site34
Site35
Site36
Site37
Site38
Site39
Site40
Site41
Site42
Site43
Site44
Site45
Site46
1986
TheChristchurchCathedral
Christchurchminor
TheChurchoftheApostles
ØysteinMagnusson’shallatHolmen
Bryggen
ØvreDreggsalmenningen
Dreggsalmenningen14-16
Sandbrugaten5
Sandbrugaten3
Dreggsalmenningen20
Dreggsalmenningen10-12
Dreggsalmenningen10-16
Dreggsalmenningen
StallenSvensgården
Bryggeparken
Nikoliakirkealmenningen
Koren-WibergsPlass
Wesenbergsmauet
Øvregaten39
Klingesmauet
Kroken3
TheChurchofStMary
TheChurchofStPeter
TheChurchofStOlavontheHill
Finnegården6a
Finnegården3a
Rosenkrantzgaten4
Vetrlidsalmenningen2Kjøttbasaren
Vetrlidsalmenningen
Øvregaten/Finnegårdsgaten
TheChurchofStNicholas
TheChurchofStColumba
LilleØvregatenfriområde
Korskirken
Skostredet10
NedreKorskirkealmenning/Vågsalmenning
Domkirkegaten6
TheChurchofStOlavinVågsbunnen
TheChurchofStCross
Rådstuplass2-3‘Vestlandsbanken’
Nygaten2
TheMunkelivBenedictineAbbeywiththeChurchofStMichael
StJohn’sAugustinianAbbey
TheChurchofAllSaints
TheNonneseterconvent
twelfthcenturychurchesinBergen,Lidénsuggests a c 1120-1180 date for the church (Lidén
1993). However since there is no concrete evidencethatplasesthechurchbefore1170,Iwill
usethechurchasasupplemantarysourceonly
for horizon 5. The exact layout, size or location of the church is not known. The church
may initially have been built in wood (Lidén
andMagerøy1990,35).PartofastonewallbehindMagistratbygningenmayhavebeenpartof
the hospital that was connected to the church
104
Yearofinvestigation
Museumnumber/project
number
Botanicalinvestigationin
VeisanbyKariLoeHjelle
1929-
1955-1979
1989
1986and1990
1967
1953
1967
1972
1986
1979
1980and1982
1989
1985
1980
1989
1981
1989
1984
BRM0
BRM298
BRM237
BRM3
BRM4
BRM42
BRM242
BRM83
BRM90
BRM287
BRM202
BRM143
BRM297
BRM94
BRM299
BRM223
1981
1982
1978/79and1981
1996/97
1991/92
1979/80
BRM104
BRM110
BRM76
BRM490
BRM342
BRM86
1994
1984
1992
1998
1987
BRM465
BRM200
BRM346
BRM544
BRM245
1963
1991
BRM20
BRM333
in1276.Awoodendrain 14Cdatedtobetween
1300-1390,foundduringatrenchexcavationat
Allehelgensgate3-5,mayalsobelongtothehospital(Hansen1994a).The 14Cdateshowsthat
thedrainmostlikelystemsfromthehospital,as
noothertracesofoccupationhavebeentracedin
thevicinityofthesiteatforinstancesite41and
site42.Accordingly,theChurchofAllSaintsis
placed in the area around Magistratbygningen
and Allehelgensgate 3-5. The location of the
churchmaybeusedasasupplementarysource
for horizon 5 since the exact location of the thefurtherwegobackintime,probablyalsoremonument is not known. The material cannot flectingasmallerscaleofactivities.
elucidateactivityintheareapriortohorizon5.
Thesourcesforhorizon1areallbutonesupplementaryorgeneralbackgroundsources.The
Site46,TheNonneseterconvent
supplementary sources are based on 14C dates
(NonneseterKloster)
orthepresenceofpollen,thesedatingmethods
‘Nonneseter’ismentionedinthewrittensourc- providewidedatingrangeswithinwhichactives in connection with events that occurred in itymostlikelytook place intheBergenarea. I
1134/35 (Hkr 1893-1901, III 326; Msk 400), willhavetocarryalongthechronologicaluncerhistoriansseemtoagreethatthenameofthecon- taintiesinherentinthematerialwhenanalysing
ventwasusedanachronistically(egHelle1982, anddiscussingitasabackdropfortheyounger
6).Onthebasisofstylisticstudiesofthevault horizons.
inthesurvivingwesttower,Lidénarguesthatit
Mostofthesourcesthathavebeenassigned
cannotbedismissedthatthechurchatNonne- tohorizons2and3areclassifiedassupplemenseterwasunderconstructionin1135(Lidénand taryandinmanycasesstructuresandcultureMagerøy 1990). The convent was most likely layers within each site have been dated tentafounded at the latest c 1150 (Helle 1982, 141) tively to the horizon, taking into use patterns
andmayserveasabasicsourceforhorizon5.
inthematerialfromtheindividualsiteorfrom
Theconventwaslocatedbythenorthernshore sites in the close vicinity. Attempts have thus
of Alrekstadvågen. The twelfth century church beenmadetomakeverticalaswellashorizontal
hadarectangularlayout,thewesttowerisstill linksbetweentheundatedsourcesanddirectly
standing. The church was 11-11.5 m wide and datedsources.Iwillreiteratehowcentralstrucabout33-34mlong,includingthetower(Lidén turesinhorizon2wereassignedtothishorizon
and Magerøy 1990, 46). The churchyard was (cfFigure20).
located on the north side of the church, to the
Only two structures were dated directly to
south of the church the remaining part of the horizon 2, these are the jetty at site 6, dendro
convent was found (Lidén and Magerøy 1980, dated to shortly ‘after 1029’ (B) and the site 7
167-68). The location of the convent may be pit-housethat,accordingtothehighestpeaksof
usedasabasicsourceforhorizon5.Themate- probability for a 14C date, may have been conrialcannotelucidateactivityintheareapriorto structedintheyearsabout1020orearlier.Inadhorizon5.
ditiontwopostsinbuilding66,probablyderived
fromreusedtimbers,weredatedtorespectively
sometime‘after1024’and‘after1040’andmay
Thespatialandtemporaldistribution thusindicateactivityintheGullskogårdenarea,
atsite6,inhorizon2.
ofthesources
HavingshownthroughverticallinkstoyoungThetemporaldistribution
erphasesthattheremustbeacertaintimedepth
Table 22 presents sources assigned to the hori- inthematerialsfromsites6andsite9andthat
zonsaccordingtosourcecategories.Thenumber palisade-builtfencesmakeuptheoldestphasesat
or ‘extent’ of the sources cannot be quantified sites6,9andperhapsalsoatsite11,Iarguedthat
ormeasuredexactly,becauseonsomesitesparts thefencesatthethreesiteswerebuiltcontempoof the material have been considered as a basic raneously.Thefenceswerebuiltusingthesame
source,whereasotherpartshavebeenconsidered techniqueandmaterials,hencethefactthatthe
assupplementaryorgeneralbackgroundsources. fences at site 6 were covered by fill-masses and
Thesitesandthematerialalsovaryinsize.
were no longer visible in the phase succeeding
The proportion of basic sources becomes that of the fences at this site, strongly suggests
smallerthefurtherwegobackintimeandthe thatthefencesatsites9and11werebuiltwhile
reliabilityofthematerialassourcesfortheho- thefencesatsite6werestillvisible.Thisimplies
rizons is proportionally reduced. Likewise the that the fences at site 9 and site 11 were built
numberofsourcesthatreflectactivitydecreases contemporaneouslywiththesite6fences.
7Evaluationofthearchaeologicalandbotanicalsources
105
Table22.Thetemporaldistributionofsourcesforhorizons1-5
Basicsources
Horizon5
(1120s-c1170)
Horizon4
(c1100-1120s)
Horizon3
(c1070-c1100)
Horizon2
(c1020/30-1070)
Horizon1
(c800-c1020/30)
21
Supplementarysources
(activity/cultivation)
15
Generalbackgroundsources
1
9
12
14
2
4
6
16
3
2
5
16
3
1
4
20
1
Attempts were then made to link the fences
tothedirectlydatedjettyandpit-houseassigned
to horizon 2. At site 6 and site 11 the fences
clearly demarcated plots. The pit-house was located closely to the plots at site 11 and was of
thesameorientationastheplots.Thiswasseen
asanindicationthattheplotsandthebuilding
mightbecontemporary.Andsincethebuilding
wasprobablybuiltbeforeorabout1020(S),this
wouldindicatethattheplotswerelaidoutbefore
oraboutthistime.Atsite6ahorizontallinkwas
madebetweenthejetty(B)andthefences/plots
whenarguingthatifathirdplotexistedeastof
plot6/C,andthisplotwasofthesamewidthas
plot6/C,thejettywouldrunstraightuptothe
easterncornerofthisplot,thuslinkingthejetty
withtheplots,andsuggestingthatthestructures
were contemporary, that is from shortly ‘after
1029’.Thereusedtimbersfoundinhorizon5indicatedthattherewasactivityinthegeneralarea
duringhorizon2aswell.
Thesupplementarysourcesassignedtohorizon
3atsites6and9inthenortherntownareahave
been dated indirectly to this horizon through
verticallinkstowell-datedmaterialfromhorizon
4andthroughverticallinkstothesourcesthat
wereassignedtohorizon2.Theindirectdatefor
thestructuresatsite6wassupportedbyapost
dendrodatedtoshortly‘after1069’,butassigned
to the horizon as a supplementary source only,
becausethequestionofreusecouldnotbesettled
forthepost.Itisquitecertainthattherewasactivityinthenortherntownareaduringhorizon
3,butitcannotbeascertainedthatabsolutelyall
structuresassignedtohorizon3haveinfactbeen
rightfullyassignedtothehorizon.Thisisespeciallyrelevantforthestructuresatsite9,asthey
weredatedtypologicallythroughstructuretypes
thathavewidedatesatsite6.
106
Supplementarysources
(nodocumentedactivity)
11
Inthemiddletownarea,activity(beyondsite
30)wasonlyassociatedwithsites26and27,the
dateofactivityonsite27isinterrelatedwithsite
26(cf60ff),soIshallnotgofurtherintosite27
here. The notion of activity at or about site 26
priortohorizon4wasbasedonthepresenceof
dendrodated,reusedwoodinhorizon4.Ifind
ithardtoexplainhowthiswoodendedupatthe
siteinhorizon4ifitdidnotoriginatefromactivityinthevicinityatanearlierstage(horizon3).
InChapter9,theplotsandplotsystemsinthe
town area are going to be identified, visualised
throughtheVisualImpactAnalysis(cfp56ff)
anddiscussed.Iftheverticalandhorizontalpatternsdrawnuponwhenassigningthesourcesfor
horizon 2 are strengthened by patterns emergingwhenwemoveoutfromthemicroscaleof
thesinglesitesandtheirclosevicinitiesthismay
strengthen the reliability of my assignment of
structures to horizon 2. This in turn may also
strengthen the assignment of structures to horizon3,asthelowertimelimitforactivitiesin
the Bergen area and the general time depth of
the Bergen material will be further elucidated.
Thequestionofchronologywillberesumedin
Chapter9(p183ff).
Asfarashorizons4and5areconcernedthe
proportion of basic sources is more satisfactory
andthesourcesprovideareliablebasisforinterpretations.
Thespatialdistribution
TherelevantsitescovermostoftheBergenarea
buttoavaryingdegree.Figure22showsamap
of all investigated areas and monuments that
serve as sources for Bergen before c 1170. The
early archaeological investigations carried out
at Holmen only focused on the monumental
architecture of the area. With the exception of
7Evaluationofthearchaeologicalandbotanicalsources
Figure20.Theverticalandhorizontalrelationshipsbetweencentralstructuresassignedtohorizons2-5
107
remainsofmonumentalarchitecturenoculturelayersorstructurespredatingc1200havebeen
documented here in the early years of research
(Dunlop 1996a). Investigations in recent years
havenotproducedculture-layersolderthanthe
fifteenthcentury(Dunlop1996a)andmysurvey
ofartefactsfromtheareahasnotproducedfinds
thatmaypredatec1170(cfp46).Thismustin
partbeexplainedastheresultofthedestruction
oftheolderstratacausedbymorerecentactivities;therecertainlymusthavebeenactivityalso
beyondtheactualmonumentsatHolmenbefore
the fifteenth century. Altogether investigations
beyondthemonumentscannotprovidesources
for activity at Holmen in the period under researchhere.SincetheexcavationsatHolmenare
notrelevantassourcesinthestudy,onlymonumentsareshownonthemaps.
The northern town area is best covered by
investigations. There is rarely more than 50 m
betweenthesites,andasaroughestimateabout
13000m2and77profilesintrencheshavebeen
investigatedovertheyears.Thiscomprisessome
16%oftheareabetweenthe15and0maslcontours,excludingtheprofilesintrenches.Incomparison,theroughestimatesforthemiddleand
southerntownareasare580m2and72profiles
amounting to about 3 % of the building land
betweenthe25masland0maslcontoursinthe
middle town area, and 1344 m2 amounting to
about5%ofthebuildinglandbetweenthe15
masland0maslcontoursinthesoutherntown
area. The numbers in square metres include
churchesandchurchyards,wherethesearelocated, but not trench investigations or sites below
the0maslcontour.Thefiguresareonlyinterestingtoacertainextent,asthenumberofinvestigatedsquaremetresorcubicmetresorprofilesin
trenchesisnotdirectlyproportionaltothelevel
ofinsightachieved.Theinvestigationsdifferin
the detail of information provided, due to the
differentmethodsapplied.Recentsmalleropen
area-andtrenchinvestigationswithlongprofiles
that‘crosssection’thetownarea,supplyrelativelybroadinformationaboutanareainspiteofthe
relativelyfewsquareorcubicmetresexcavated.
Themapshowsthatmostoftheexcavatedsites
are located close to the waterfront, and several
arebelow0maslintheperiodunderinvestigation.Investigationslocatedabovethetidalzone
108
andinthehigherpartsofthetownareaarethus
relativelyfewandactivitiesheremustbestudied
throughothersitesandsources.
The northern town area must be considered
well covered by excavations, though the area
just above the tidal zone could be covered better.Mostofthelargersiteswereexcavatedbefore
1979andsomeevenbefore1955,withalossof
importantdetailsasaconsequence.Themiddle
townareaisrepresentedbyfewerinvestigations.
Thesitesarespreadoutandcompriselongprofiles that cross-section the sloping terrain from
thefootofFløyfjellettothewaterfront.Furthermore, most of the secular sites were excavated
after1979andrefinedmethodsofanalysiswere
applied, these factors may to some extent outweighthelownumberofsites.Inthesouthern
townareainvestigationsarescarce.Wehaveno
sitesfromtheareawestoftheChurchofStCross
(site40)andthetwolargestsitesareeitherbelow
sealevelorinthetidalzone.However,thesites
have been investigated in recent years through
detailedmethodsofanalysisincludingbotanical
studiesofthelayers,whichtosomeextentcompensatesforthescarcityofsites.
TheNordnesandNonneseterareashavebeen
investigated through a number of larger and
smaller excavations I have only included a few
‘secular’siteshere(sites41and42)toshowthat
thesecularpartsoftheearlyandhighmedieval
towndidnotextendintotheseareas.Onlythe
monastic and ecclesiastic institutions here are
relevantassourcesforearlyBergen.
Therepresentativityoftheartefactand
ecofactmaterial
Altogethernearlytenthousand(9798)artefacts
havebeenassignedtotheperiodbeforec1170.In
addition,botanicalandosteologicalsourceshave
been associated with the period under investigation.Table23showsthenumberofartefacts
inhorizons2-5dividedintosourceandartefact
categories. Table 24 shows the distribution of
relevant data from botanical investigations. All
theartefactsandecofactsstemfrombasicorsupplementary sources and none can be associated
directlywiththemonumentalsites.Everyfragmentofanartefactiscountedasoneartefactunlessthefragmentsareobviouslypartofthesame
object.Acombmaythusconsistofmanyfrag-
mentsbutonlycountsasoneartefact,whereasa
sherdofpotterycountsasoneartefact.Ihavenot
madespecialattemptstorefitfragments.Pollen
and macrofossils are not quantified and merely
thepresenceofecofactsrelevantasasourcefor
thesubjectsunderdiscussionwillbedealtwith.
Table23.Thenumberofartefactsassignedtohorizons2-5
(N=9798)
Horizon
2
3
3
4
4
4
5
5
5
5
Source
category
S
S
S
B
B
S
B
B
S
S
Artefact
category
II
I
II
I
II
II
I
II
I
II
Numberof
artefacts
1
3
18
51
90
115
2677
6707
6
130
B=basicsource,S=supplementarysource
Table24.Relevantbotanicalsources
Site/plot/unit
Site1,Koengen
Site6Bryggen
Site7,Øvre
Dreggsalmenningen
Site37,Nedre
Korskirkealmenning
Site38,Domkirkegaten6
Site21,Klingesmauet
Horizon
1
3-4
2
Ecofacts
Pollen,macrofossils
Pollen
Pollen
1
Pollen
1-4
5
Pollen
Pollen
Thetablesshowthatthereisalargedifference
inthenumberofartefactsandecofactsassigned
to the four find-yielding horizons and hardly
anyfindscanbeassignedtotheoldesthorizons.
Several factors influence the representativity of
thematerial-inrelationtowhatwasonceinuse
-andinrelationtothenumberofartefactspreserved in the ground. I will therefore elaborate
uponthis.
Artefactsendedupintheculture-layerswhen
lost or thrown out as garbage or when lost or
leftbehindduringforinstanceafire.Somewere
probablyreusedagainandagainandlargequantitiesoforganicmaterialmusthaveendedupas
firewoodorweredestroyedduringfires.Theselectionofitemspreservedinthegroundisthus
7Evaluationofthearchaeologicalandbotanicalsources
notrepresentativeinaonetoonerelationshipof
whatwasactuallyinuse.Organicmaterialthat
burns is probably underrepresented in relation
toitemsofstoneandpottery,forexample.Also
metalobjectsmaybeunderrepresented;theymay
havebeenreusedormelteddown,andarepreserved relatively badly in the ground. Such circumstancesshouldbecommonforallthesites.
Ideally,thevarietyoforganicversusmineraland
metalfindsthatendedupinthegroundshould
thusbecomparablefromsitetosite.
The local conditions for the preservation of
organicmaterialandmetalsoneachsitedepend
onthetopographicallocationofthesiteandthe
characteroftheculture-layers.Intheperiodunderinvestigationmostoftheartefactproducing
sitesarelocatedclosetothewaterfront,afeware
located at the foot of Fløyfjellet. The artefactyieldingsitesarealmostalllocatedondryland
-thelayerswerenotreallywaterlogged-andthe
topographicallocationofthesitesisthusrather
homogeneous.Alsothecompositionoftheculture-layers from site to site is rather uniform
throughthehorizonsunderstudy.Thedeposits
are characterised by a variety of mineral layers
withlargedrainagecapacityandratherthinorganiclayers.Theconditionsforthepreservation
ofmetalsandorganicmaterialsarenotoptimal,
butthisrelatestoallthesitesandthenumberof
itemspreservedinthegroundshouldthusideally
becomparablefromsitetosite.
Ifconditionswereideal,thenumberandvarietyofartefactsandecofactsfromthesitesshould
becomparablebetweenthesites.However,specialproblemsinherentinthematerialinfluence
the number of artefacts and ecofacts that may
serveassourceshere.Theseproblemsarerelated
to the selection of artefacts that were collected
during excavation, to the number of artefacts
thatcouldbeidentifiedbymeassourcesforthe
horizonsoneachsite,andtothenumberofartefacts I retrieved in the museum storerooms.
Thesespecialproblemsmustbeconsideredwhen
decidinghowtostudythesources.
Tables25-28showtheapproximatefinds-frequency per excavated m2 at find-yielding analyticunitsinhorizons1-5.46Asseeninthetables
thereisalargevarietyinthenumberofartefacts
per square metre on the artefact-yielding plots
or units. Some of the differences may reflect a
109
varyingintensityorcharacterofactivitiesonthe
plots/units,however,thevaryingdocumentation
methodsandalsotheprinciplesforcollectingthe
osteological and botanical material (other than
pollen) must account for the real large differencesinthefinds-frequencywithineachhorizon
onthesites.
In some parts of site 6, culture-layers other
than fire-layers were removed by machine and
basically only structures were recorded (plot 6/
F). In the areas of plots 6/E, 6/F and 6/G the
levelbelowhorizon5wasnotthoroughlyexcavated(cfp85ff).Atsite15andatsite20culturelayers were partly removed by machine so that
onlyprofilesand‘pockets’ofculture-layerscould
be investigated thoroughly. Trench excavations
generallyproducefewfinds,whichisaproblem
inherentinthisexcavationmethod.
Table25.Horizon2(c1020/30-c1070)approximate
numberofartefactsperexcavatedm2attheartefact-yielding
unit(N=1)
Analyticunithorizon2
approximateexcavated
area
Site7,Øvre
Dreggsalmenningen
(trench)
ArtefactscategoryI
andIItotal
1
Artefacts
perm2
excavated
**
**Anestimatehasnotbeenmadefortrench
excavations
Table26.Horizon3(c1070-c1100),approximatenumber
ofartefactsperexcavatedm2attheartefact-yieldingplot/unit
(N=21)
Plot/unithorizon3
approximateexcavated
area
Site9,Sandbrugaten5and
site10,Sandbrugaten3
plot9-10/B(c320m2)
Site20,Øvregaten39unit
20/A(c70m2)
ArtefactscategoryI Artefacts
andIItotal
perm2
excavated
18
0.05
3
0.04
‘All’ artefacts have been collected systematicallyatallthesitesexcavatedfrom1955andafter
(exceptatsite10),however,therehasbeenadevelopmentindefininganartefactthatisworthwhile
tobedocumented;slagisthusonesourcegroup
thathasnotbeencollectedsystematicallyduring
excavationsbefore1980.Wastefrombone,antler,hornworkingisalsoagroupofmaterialsthat
probablywasnotcollectedsosystematicallyuntil
110
after1980.Neitherecofactshavebeencollected
orstudiedsystematically.Theosteologicalmaterialsometimesfiguresinthefinds-catalogues(eg
sites26and27),butaccordingtovaryingprinciples,whilebotanicalmaterialisrarelyrecorded
in the catalogues. All these methodological circumstancesinfluencethefindfrequencyperexcavatedsquaremetreconsiderably.
As mentioned earlier the documentation
method applied at several large sites excavated
before 1980 has not allowed me to relate artefacts to the horizons, although the presence of
various structures indicates activities. At site 6
theareathatwaslaterusedasaburialgroundfor
theChurchofStMary,thatisthenorth-easternmostpartsofplots6/Band6/C,thedocumentationoffinds-contexthasbeendifficulttohandle.
Thismayprobablyaccountforthelackofartefactsassignedtohorizons2and3fromthisarea.
Atsites9,10and11,Iwasnotabletorelateany
artefactstohorizon5.Atsite28ithasonlybeen
possibletoassignafewfindstohorizon5.Itis
likely that the artefacts that could be assigned
tothehorizonsrepresentarandomselectionof
what was retrieved during excavations. Consequently,eventhoughIwasonlyabletoidentify
afractionofthefindsthatwereretrievedduring
excavation,theidentifiedfindsshouldrepresent
anunbiasedselection.
InthearchivesIhavenotbeenabletoretrieve
all the artefacts that were recorded in the databases. However, many were described, drawn or
otherwisedocumentedsothattheycouldstillbe
included in the study. Only in the cases where
slagandpossiblewastefromantler,boneorhorn
workingwaslistedandnotretrieveddidthe‘loss’
offindspresentarealproblem.Thisisdiscussedin
thestudiesoftheseartefactgroupsinChapter11.
Tosumup,therepresentativityofthematerial
concerningthevarietyofwhatwasonceinuse
andofwhatwaspreservedinthegroundshould
beregardedasfairlycomparablefromsitetosite
intheearlyperiodofthetown’shistory.Interms
ofthefinds-frequencyfromsitetositewithinthe
horizonsthereare,however,somanymethodologicalcircumstancesinvolvedthataquantitative
analysis of the material across the find-bearing
plots/siteswillnotbepossible.Insteadaqualitative,contextualandspatialapproachwillbeapplied(cfp71ff).
Table27.Horizon4(c1100-1120s),approximatenumberofartefactsperexcavatedm2atthe7artefact-yieldingplots/units
(N=252)
Plot/unithorizon4approximateexcavatedarea
Site6,Bryggenplot06/B(c480m2)
Site6,Bryggenplot06/C(c440m2)
Site6,Bryggenplot06/D(c250m2)
Site26,Finnegården6aplot26/A(c17m2)
Site26,Finnegården6aandsite27,Finnegården3aplot26-27/B(c43m2)
Site26,Finnegården6aandsite27,Finnegården3aplot26-27/B-C(c26m2)*
Site27,Finnegården3aplot27/C(c34m2)
Site30,Vetrlidsalmenningenunit30/E(c4m2)
Artefactscategory
IandIItotal
45
28
3
28
28
50
52
18
Artefactsperm2
excavated
0.09
0.05
0.01
1.64
0.64
1.92
1.52
4.5
*Artefactsfromanareacoveringabout26m2areassignedtoplotBorC
Table28.Horizon5(1120s-c1170),approximatenumberofartefactsperexcavatedm2atthe24artefact-yieldingplots/units
(N=9100)
Plot/unithorizon5approximateexcavatedarea
Artefactscategory
IandIItotal
Site6,Bryggenplot06/A(c72m2)
Site6,Bryggenplot06/B(c480m2)
Site6,Bryggenplot06/C(c440m2)
Site6,Bryggenplot06/D(c250m2)
Site6,Bryggenplot06/E(c218m2)
Site6,Bryggenplot06/F(c200m2)
Site6,Bryggenplot06/G(c150m2)
Site8,Dreggsalmenningen14-16plot08/A(c17m2)
Site8,Dreggsalmenningen14-16unit08/B(c17m2)
Site8,Dreggsalmenningen14-16unit08/D(c3m2)
Site15,StallenSvensgårdenandsite16,Bryggeparkenplot15-16/A(c82m2)
Site20,Øvregaten39unit20/A(c70m2)
Site21,Klingesmauetunit21/A(trench)
Site22,Kroken3,unit22/A(c10m2)
Site26,Finnegården6aplot26/A(c17m2)
Site26,Finnegården6aand27,Finnegården3aplot26-27/B(c43m2)
Site26,Finnegården6aandsite27,Finnegården3aplot26-27/B-C(c26m2)*
Site27,Finnegården3aplot27/C(c34m2)
Site28,Rosenkrantzgaten4plot28/B(c137m2)
Site28,Rosenkrantzgaten4plot28/C(c60m2)
Site30,Vetrlidsalmenningenunit30/A(trench)
Site30,Vetrlidsalmenningenunit30/B(trench)
Site30,Vetrlidsalmenningenunit30/D(trench)
Site30,Vetrlidsalmenningenunit30/E(c4m2)
Site38,Domkirkegaten6unit38/A(c286m2)
2
717
2084
3144
973
187
905
13
51
19
24
43
2
13
245
149
229
132
56
30
1
23
6
35
17
Artefactsperm2
excavated
0.02
1.48
4.75
12.55
4.45
0.93
5.93
0.75
2.99
6.33
0.28
0.61
**
1.30
14.40
3.45
8.80
3.88
0.40
0.50
**
**
**
8.75
0.05
*Artefactsfromanareacoveringabout26m2areassignedtoplotBorC
**Anestimatehasnotbeenmadefortrenchexcavations
7Evaluationofthearchaeologicalandbotanicalsources
111
Figure21.Generallegendformaps
112
Figure22.Investigatedsitesandmonuments
7Evaluationofthearchaeologicalandbotanicalsources
113
Figure23.Structuresandlayersassignedtohorizon1(c800-c1020/30)
114
Figure24a.Structuresandlayersassignedtohorizon2(c1020/30-c1070),thenortherntownarea
7Evaluationofthearchaeologicalandbotanicalsources
115
Figure24b.Structuresandlayersassignedtohorizon2(c1020/30-c1070),themiddletownarea
116
Figure25a.Structuresandlayersassignedtohorizon3(c1070-c1100),Holmen
7Evaluationofthearchaeologicalandbotanicalsources
117
Figure25b.Structuresandlayersassignedtohorizon3(c1070-c1100),thenortherntownarea
118
Figure25c.Structuresandlayersassignedtohorizon3(c1070-c1100),themiddletownarea
7Evaluationofthearchaeologicalandbotanicalsources
119
Figure26a.Structuresandlayersassignedtohorizon4(c1100-c1120s)
120
Figure26b.Structuresandlayersassignedtohorizon4(c1100-c1120s),thenortherntownarea
7Evaluationofthearchaeologicalandbotanicalsources
121
Figure26c.Structuresandlayersassignedtohorizon4(c1100-c1120s),themiddletownarea
122
Figure27a.Structuresandlayersassignedtohorizon5(1120s-c1170)
7Evaluationofthearchaeologicalandbotanicalsources
123
Figure27b.Structuresandlayersassignedtohorizon5(c1120s-c1170),thenortherntownarea
124
Figure27c.Structuresandlayersassignedtohorizon5(c1120s-c1170),themiddletownarea
7Evaluationofthearchaeologicalandbotanicalsources
125
Figure27d.Structuresandlayersassignedtohorizon5(c1120s-c1170),thesoutherntownarea
126
PARTII
MAJORINITIATIVESANDDAILYACTIVITIESINEARLYBERGEN
Havingsingledouttheavailablesourcesofrelevancetotheperiodfromtheninthcenturyto
c1170,Iwillnowanalyseanddiscussthesources
acrossthesitesinsixpart-studiesandundersix
themes.Onlysiteswheresourcesfortherespectivethemesareavailableareincludedinthestudies,writtensourcesarediscussedwhenrelevant.
Inordertokeepintouchwiththeuncertainties
inherent in the material, structures and layers
are referred to as basic (B), supplementary (S)
orgeneralbackgroundsources(G).Thespecific
methodological approaches are accounted for
themebytheme.
8HORIZON1(C800-C1020/30),
ABACKDROP
In this study I will investigate where activity
was located on the northern shore of Vågen
fromabouttheninthcenturyuptoc1020/30
anddiscusshowthisactivitymaybecharacterisedintermsofgenerallanduseandintermsof
urbanversusnon-urbansettlement.Thematerialfromhorizon1issparseandtheanalysisis
mainlygoingtoserveasabackdroptohorizons
2-5. I will discuss data geographically going
fromnorthwesttosoutheastaswellaschronologically.
Locationandgenerallanduse
Materialthatcouldbeassignedtohorizon1has
been documented only at a few sites along the
northern shore of Vågen (cf Figure 23). In the
northwesternmost area in the Veisan inlet, organiclayersfromsite1(S)containedhousehold
waste dumped in the close vicinity. The layers
8Horizon1(c800-c1020/30),abackdrop
were 14Cdatedwithinthetimeframes780-790
or 810-1000 and contained both pollen and
macrofossils among others wood-chips, latrine,
kitchenandbreweryrefuseanddung.Thepollenspectreshowedpollenfromplantswhichdid
notgrowinwesternNorwayatthetimeofdeposition(Hjelle1986,55,58)(cfp78ff).
Can the settlement that produced the waste
be located more precisely? Being situated close
totheregistrationpoint,theHolmenareanorthwest of Veisan may be a possible location for a
settlement. However the archaeological sites at
Holmencannotelucidatetheperiodcoveredby
horizon1(cfp157).Largepartsofthesouthand
easternshoreofVeisanhavebeeninvestigatedarchaeologically (cf p 157 and Figure 22) but no
insitutracesofoccupationolderthanhorizon2
have been identified here. The terrain north of
Veisanseemstohavebeentoosteeptobeattractive as building land if other possibilities were
open. By the process of elimination, it seems
unlikely that the eastern and northern shores
hosted a settlement during horizon 1. To summarise, this leaves Holmen as a likely location
forasettlement.Thepossibilityofsomekindof
settlementatHolmenbroadlydatedtotheninth
ortenthcenturiescannotbeexcludedbutisnot
substantiated through the existing archaeologicalmaterial.
Inthenortherntownareaonlyonesiteproducedrelevantmaterialforhorizon1.Atsite7(S)
acultivationlayerdatedtentativelytosometime
betweentheninthcenturyandthefirstdecades
oftheeleventhcenturywasrecorded(cfp95).
Polleninthelayerindicatedmeadowvegetation
atthesiteandwheat-orbarleygrowingorsettlementinthevicinity.Importindicatingpollen
127
signifyinghouseholdwastewasalsofound.The
pollenmustapparentlyhavebeentransportedto
site 7 from a settlement in the vicinity. As accountedforinChapter7(p157andFigure22)
thenortherntownareaisfairlywellcoveredby
archaeological sites. Since no settlement traces
could be assigned to horizon 1 here, I find it
probablethatthesettlementreflectedindirectly
in the material from site 7 was not located in
the northern town area. A possible explanation
for the presence of pollen indicating household
waste at site 7 can be that surface water which
had been in contact with fields fertilised with
household waste, transported the pollen to site
7. In that case, the material from site 7 would
notonlyindicatemeadowsatthesiteitself,but
alsofieldsintheclosevicinity.Thelatterwould
nothavebeenrecordedarchaeologicallywithout
botanicalinvestigations,whichhavebeencarried
outinonlyafewplaces.Basedontheavailable
data-howevervague-Isuggestthatthegeneral
landuseatsite7mayreflectmeadowsatsite7
andfieldsfertilisedbyhouseholdwasteinthevicinity.
Atsite37(S)alayer, 14Cdatedtoalmostthe
sameperiodasthesite1-depositsinVeisan:810970,accumulatedinthesea.Thelayercontained
import-indicating pollen probably signifying
household waste. As shown in Chapter 7 there
werenomacrofossilstotiethepointofdepositionofthehouseholdwastecloselytosite37and
there were no indications that household waste
wasdumpedintheseaattheneighbouringsite
36.Therearethusnoindicationsinthematerial
fromsite37oritsclosevicinitiesthatthewaste
producingsettlementwaslocatedclosetosite37.
AsaccountedforinChapter7(p157andFigure22)thesourcesforactivityinthesouthern
townareaarefewandfarbetween.Stillthefact
thatnotracesofoccupationdatedtotheperiod
beforehorizon5werefoundatanyofthewelldocumentedsitesmaybeaslightindicationthat
thepollenfromsite37didnotoriginateatasettlementinthisarea.Thepollen,alongwithother
household waste may rather have been spread
on arable fields somewhere in the Bergen area,
andmayhavebeenwashedintotheVågenwith
thesurfacewater.Thusthepollenwouldreflect
agricultural activities in the catchment area of
128
streamsleadingtotheVågenBay,ratherthana
settlementinthevicinityofsite37.
The traces of land use discussed so far all
stem from supplementary sources, as they were
all dated by broad 14C dates or the presence of
Centaureacyanus.Itisnotpossibletodetermine
whethertheactivitiesindicatedatthesiteswere
contemporary in a narrow sense. The sources
are, however, not interrelated. As they all indicate activity in the Bergen area during horizon
1,thegeneraltendencythatactivitiesofvarious
kinds were carried out during the wide period
representedbyhorizon1,oughttobereliable.
At site 30 an approximately 2 m wide pier
(B) (analytic unit 30/A) was found, extending
intothewateroftheoriginalsmallbayandbuilt
about 900 (B). Three posts (G) in unit 30/D
interpreted as part of one structure were constructed further up on the shore contemporary
with or perhaps somewhat later than the pier.
No layers have been associated with the use of
thepostssoitisdifficulttodecidetheirfunction
moreprecisely,butthepostsdopossiblyindicate
somekindofsettlementhere(cfp138ff).The
jettyisconsideredareliablesourceforhorizon1,
whereastheassignmentofthepoststohorizon1
isnotwell-founded.
Urbanornon-urban?
On the basis of the available botanical and archaeologicalmaterial,canwedeterminewhether
thesettlementtracesfoundinVeisanandinthe
middletownareawereurbanornon-urban?
The‘Holmensettlement’
KariLoeHjellehasdiscussedthematerialfrom
site1inrelationtothestructuralandfunctional
urbancriteriapresentedbyHelleandNedkvitne
(1977)(cfp20).Sheconcludesthattheanticipated settlement by Veisan was probably more
concentrated than rural settlements around
Bergen,anditwaspermanentincharacter.Furthermore she argues that pollen indicating import of grain implies international trade and a
centralplaceforthetradingofgoods,andindirectlyindicatesthatspecialisedeconomicactivitieswerecarriedout.Accordingly,thesettlement
mayfulfilthefunctionalandstructuralcriteria
foratown(Hjelle1986,61-62).Hjelledoesnot
explicitly define the settlement at Holmen as a
town,butclaimsthat:‘...assofarasonecantie
themedievaldepositstothedefinitionofatown
onecantietheVikingAgedepositstothesame
definition’(Hjelle1986,62)(mytranslation).
Herconclusionsarebasedontwomainarguments. As the botanical material, dated to the
Viking Age, reflects the same activity as layers
datedtothetwelfthorthirteenthcenturieswhen
Bergen was definitely urban, the waste-layers
from the Viking Age may also represent nonagrarianactivitiesofamorepermanentcharacter
(Hjelle1986,55-57,61-62).Theactualmaterial,
however,comprisesonlyafewlayersoflittlevolumeandthematerialcanhardlycountasbeing
representativeforneithertwelfthnorthirteenth
century ‘urban activity’ nor ‘Viking Age activity’assuch.Hjelle’ssecondargumentisbasedon
thepremisethatwastewasnotthrownintothe
seaonaruralsiteasitcouldbeusedasfertiliser
on the arable land, thus waste thrown into the
seaindirectlyreflectsmentalitiesorstrategiesofa
non-ruralpopulation.Asthispremisealsorelates
tootherinvestigations,andfrequentlyhasbeen
usedinthedebateofearlyurbanisationinBergen(cfp51ff)Iwilldiscussthematerialbehind
thepremise.
A central question is the nature of the ordinaryfarm,whichthepremiserefersto.AccordingtoHjelle,botanicalinvestigationsinseveral
agriculturalareashavenotproducedwaste-layers
depositedinbasins,theinvestigationoftheVikingageandmedievalfarmatLurekalvenserves
ashermainexample(Hjelle1986,56-57).InvestigationsofsedimentsintheKaasaBaycloseto
theVikingAgeandmedievalfarmatHøybøen,
ontheislandofSotra,showednodepositscomparablewiththoseinBergen(Krzywinski1991).
Basedonmaterialfromthesesitesitisconcludedthatinruralareaswastewasnotdumpedin
theseabutusedonthefieldsasfertiliser(Hjelle
1986,56;Krzywinski1991,148).Acrucialquestionisthenwhetherthesefarmsarecomparable
toapossiblefarmnearVågenBay.
LurekalvenandHøybøenwerelocatedinthe
outercoastaldistrictwestofBergen,Lurekalven
onasmallisland.Soilwasascarseressourcehere
andthe‘Plaggenboden’landuse,whereallwaste
was used as fertiliser, was typical for farms in
8Horizon1(c800-c1020/30),abackdrop
coastal areas west of Bergen in the Viking and
MiddleAges(Kaland1979;KrzywinskiandKaland1984).Incontrast,‘Bergen’islocatedfurtherinlandbytheinnerfjordsandconnectedto
awidevalley.Grassfrommeadows,notheather
wasthemainfodderhere(cfKaland1979;Hjelle
1994,164).Theoutercoastaldistrictheathland
farmsmaythereforenotserveassatisfactoryparallelstoapossiblefarmintheBergenareaand
do not provide sufficient basis for the premise
thatwastethrownintotheseaindirectlyreflects
adenserpopulationthanwasnormalforanordinaryfarm.Onthebasisofthematerialfromsite
1alonewecannotdeterminewhetherthepossiblesettlementatHolmenwasdenserinstructure
thansettlementsincomparableareas.
Hjelleconcludesthatspecialisedtradingactivitieswerecarriedoutinthesettlement.Thisconclusionisbasedonpollenindicatingthepresence
ofgraingrownoutsideNorwayandthenotion
thatthesettlementwasdenserinstructurethan
settlementsinthesurroundingarea.Aswehave
seen, the latter notion is not strong, leaving us
withtheimport-indicatingpollen.Thepresence
of pollen of foreign origin may not necessarily
countasevidenceofaneconomydifferentfrom
that of a Viking Age rural settlement. Trading
connectionsofamorelimitedscaleandwithin
anon-urbanspheremaywellhavebeenfoundin
theperiodrepresentedbyhorizon1.Theimport
of grain in itself is not dependent on an urban
structureandargumentsbasedontheimport-indicatingpollenarethusnotconclusive.
In conclusion, I find that Helle and Nedkvitnesstructuralandfunctionalcriteriaforthe
settlementtoqualifyasatowncannotbetested
satisfactorily,andwecannotdeterminewhether
thebotanicaltracesfromVeisansignifyanurban
ornon-urbansettlementonthebasisofthematerialfromsite1alone.ExcavationsofKaupang
inVestfold,BirkaandHaitabuhaveshownthat
craftswereimportantactivitiesintheseViking
Agetownsandwasteandblanksfromthefabricationofcombs,glassbeadsandmetaljewelleryarecommonfindgroupshere.Inconnection
withthesetownslargeburialgroundshavealso
beenidentified(egAmbrosianiandClarke1995
(1991)).Thesurveyofstrayfindsandfindsfrom
regular excavations from the Bergen area have
129
notproducedanycraftsindicatingartefactsthat
canbedatedtotheVikingAge.Theuseofdata
exsilentioisproblematicfromamethodological
pointofview,thelackoffindsmaybeexplained
withreferencestoalackofinvestigationsinrelevantareasandmethodologyandsuchreferences
arecertainlyrelevantfortheHolmenarea.Still,
thislackoffindsmayalsobeaslightindication
thatnoVikingAgeurbansettlementresembling
thoseatKaupang,BirkaandHaitabuwaslocatedtotheHolmenareaduringhorizon1.
Looking at the material from site 1 in isolation,onemightsuggestthatthewaste-layersdid
notstemfromasettlementonlandbutfroma
‘householdonaboat’instead.Thewaste-layers
might have been deposited in the Veisan from
boatsanchoringforthenightintheshelteredinlet.Theactivitytracescouldthenbeexplainedas
originatingfromamuchusedanchorageplace,
forinstanceusedbytravellersontheirwayupor
downthecoast.Suchplaceshavebeenfoundin
numbers along the Danish and Swedish coasts
(egCallmer1991;Ulriksen1998),however,the
activity traces from site 1 cannot, however, be
seeninisolation,andtheanchorageplacetheory
doesnotexplaintohowhouseholdwastesignifiedbyimport-indicatingpollenapparentlyendedupinotherplacesintheBergenarea(cfsites
7and37).
I have made the case that the deposits from
site7andsite37mayindirectlyreflectagricultural activities in the Bergen area. If we accept
this,itfollowslogicallythattheagriculturalactivities were conducted from a settlement that
hadaccesstoimportedgrain,andthesettlement
associatedwithVeisanandtentativelylocatedto
Holmenspringstomind.Althoughoursources
arelimited,theymay,whenseentogether,suggest that the culture-layers found in Veisan reflectasettlementwhereagrarianactivitieswere
carriedout,perhapslocatedatHolmenandwith
fieldsintheBergenarea.Thesettlementapparently had international contacts and imported
grainperhapsasearlyasintheninthcentury.
Aswehaveseenearlier,researchershave,with
theplacenameBjorgvinasapointofdeparture,
discussedthepresenceofafarmwiththisname
intheBergenarea(Lorentzen1952,43-44,with
references;Herteig1969,129-134,withreferenc130
es).Basedupontopographicaldata,placenames
and the study of boundaries between farms in
thevicinityofBergen,Hellehassuggestedthat
theBjorgvinfarmwaslocatedintheareaaround
Vågen, the farm may have been as large as the
royalestateatAlrekstadandmayhavebeenroyal
property(Helle1982,71-85).TheBjorgvinname
may,accordingtothephilologistDASeip,belongtoagroupofvin-names,whichhadalready
beenintroducedatthebeginningoftheViking
Age(c800)(Helle1982,85).Suchadateisnot
inconsistentwiththewidedatesprovidedinthe
botanicalmaterial.Itistemptingtosuggestthat
the proposed settlement at Holmen was identicalwiththesupposedBjorgvinfarm.Untilmore
firmarchaeologicalevidenceisavailablefromthe
area,however,thisproposalmustbeconsidered
merelyasahypothesis.
ThePieratsite30
Howdoesthepierandpossiblyalsothepostsat
site30addtothispicture?Asthepier(30/A)is
probablynotolderthanc900,itappearstobe
youngerthanthematerialdiscussedsofar.Due
to the broad dates and the general character of
theactivitiesrepresentedbythebotanicalmaterial,itcannotbeexcludedthatthestructuresat
site30andthesuggestedsettlementatHolmen
werealsoinuseatthesametime.
Thepieritselfmusthavefunctionedasalanding-placeforgoodsandpeoplecarriedbyboat,
andthepossiblepost-constructionfurtherupthe
beachcouldhavebeenalmostanything,perhaps
ashedoraboathouse.Thegeneralcharacterof
the activity is difficult to grasp the structures
beingfewinnumberandnolayershavingbeen
documented in connection with the structures.
Although the profiles of site 30 stretched as a
cross-sectionoftheslopingterrainfromthefoot
ofFløyfjelletinthewesttotheshoreofVågen
in the east (cf p 138ff), there are no traces of
contemporarystructuresorculture-layersinthe
remaining units at the site. Likewise there was
noevidenceofactivityintheperiodcoveredby
horizon 1 neither at site 34 about 30 m to the
south of the site 30 trenches, at site 31 to the
westnoratsites26and27,locatedabout79m
from the trenches of site 30. The pier and the
post-constructionatsite30wereapparentlynot
partofadenselybuilt-upsettlementinhorizon
1.Thepierwaslocatedabout350meastofHolmen with ample possibilities for a much closer
landing-place for the suggested Holmen settlement.
Aswehaveseen,theroyalestateatAlrekstad
waslocatedabout2kmsouthofthemouthof
VågenaboveAlrekstadvågenBay,thelaterStore
Lungegårdsvann.Theroyalestatemusthavehad
a landing-place for goods and people and with
enough space for boathouses. The location of
this landing-place has been discussed over the
years. Alrekstadvågen, the closest alternative to
AlrekstadhasbeenconsideredalesslikelycandidatethantheVågenBay(Koren-Wiberg1921,
21;Lorentzen1952,47;Herteig1969,134-136;
Helle1982,74-75),becausetheAlrekstadvågen
Bay is more likely to freeze in the winter than
Vågen. Furthermore, Alrekstadvågen is less accessible by larger boats than Vågen (Herteig
1969, 134-136; Helle 1982, 74-75) and harder
todefendandescapefromthanVågen(Herteig
1969,136).BoththeHolmenarea(Koren-Wiberg 1921) and Vågsbunnen close to the later
ChurchofStCrosshavebeensuggestedaspossible locations for the landing-place (Lorentzen
1952;Herteig1969;Helle1982).
TheNorwegiankingsoftenfrequentedAlrekstadandotherwestNorwegianfarmsfromKing
HaraldHårfagreandonwards(Hkr1893-1901,
I 155, 161; Helle 1982, 72) and Alrekstad may
thushavebeenaroyalestatealreadybytheend
oftheninthcentury.Thisdateisnotinconflict
with the archaeological date of the pier. Since
the area around Vågen was most likely owned
bythekingbeforeatownemergedhere(Helle
1982, 71-85 with references) it is possible that
Alrekstad was free to establish a landing-place
withintheVågenarea.Thesecircumstancesdo
notprovethatthepierandassociatedstructures
atsite30representalanding-placeconnectedto
Alrekstad, but they certainly do not contradict
suchaninterpretation.
area. The finds from Veisan are best explained
as representing a settlement where agricultural
activitieswerecarriedout,itmayhavebeenlocatedatHolmenandprobablyhadfieldsinthe
Bergenaera.
9PLOTSANDPLOTSYSTEMSIN
THETOWNAREA
In this chapter I will first identify boundaries
throughthesources thathavebeenassignedto
horizons 2 to 5 with varying certainty. If the
boundariesformsystemsthismaystrengthenmy
assignmentofthematerialassourcesfortherespectivehorizons.Iwillthereforediscusswhether plot systems were present in the town area
fromhorizon2tohorizon5.Havingdiscerned
twodifferentplotsystems,Iwillevaluatecentral
datesthatapplytothesourcesfromhorizons2
and3.Finallyareasincludedintheplotsystems
discernedaretentativelyreconstructed.
Plotboundaries
PlotsinearlyBergenareidentifiedthroughthe
presenceofoneormoreofthefollowingboundary indicators (cf Schia 1987a; Christophersen
andNordeide1994,122-123):
• Palisadefences
• Systematicdifferencebetweenculture-layers
depositedoneachsideofan‘invisible’line
• Systematic coherence in the orientation of
structuresoneachsideofan‘invisible’line
• Wallalignments
• Churchyards
• Shorelines
• Eavesdrops
And‘diagnostic’structures:
• Thedoubletenementbuildingpattern
Theboundaryindicatorsaremostapplicableon
thelargersiteswhereabroadviewofthebuilt-up
area can be achieved and where the settlement
Conclusions
wasdenselybuilt.AtsmallersitesIwillusethe
To conclude, there are no traces of occupation presenceof‘diagnostic’structuresasanadditionthatcancountasconcludingevidenceofanur- almeanstoidentifyboundariesandplots.The
bansettlementinhorizon1.ThepieratVetrlid- characteristic2mx2mcaissonsfoundatmany
salmenningen was not part of a wider built-up sites are such structures. At site 6 the building
9Plotsandplotsystemsinthetownarea
131
pattern, usually referred to as the double tenement system, was identified in horizons 4 and
5.Thetypicalbuildingpatternonaplotinthis
systemischaracterisedbytworowsofbuildings,
apassage,andeavesdropsthatrunat90degrees
totheVågenwaterfront.Thepassageismostoftenlocatedbetweenthebuildingsalongthemiddleaxisofthetenement(Herteig1985,11)and
the eavesdrops demarcate the lengthwise plot
boundaries towards neighbouring tenements.
Some ‘double tenements’ consist of only one
row of buildings (Herteig 1985, 11), but when
referring to the typical double tenement layout
in this study, it consists of two rows of buildings flanked by eavesdrops that demarcate plot
boundaries.Atsite6,passagesassignedtohorizons4and5wereclearlyfoundedon2mx2m
stone-filledcaissons.Suchcaissonsalsoindicate
thewaterfrontextensionofthebuilt-uparea.At
largersiteswheresuchcaissonsarepresentthey
arealwaysassociatedwithpassagesthatrunbetweenbuildingrowsortheydemarcatethewaterfrontextensionoftheplot.Thiscaissontype
thus appears to be ‘diagnostic’ for the ‘double
tenement’buildingpatternandassociatedplots.
When similar caissons are found on other and
smallerwaterfrontsites,theymostlikelysignify
atenementpatternsimilartothatatsite6inhorizons4and5.Atsmallersiteswhereonly2m
x2mstone-filledcaissonshavebeenidentified
the location of plot boundaries may be reconstructedbyprojectingthelocationofeavesdrops
inphasesfollowingthe‘caissonphase’,because
thelocationofeavesdropsappeartobeverystableinareaswherethispatternhasbeenstudied
indetail(cfMoldung2000).
Except for palisade fences, the boundary indicators all depend on observable patterns of
constructions or culture-layers, this presents a
problematsmallsitesandwhenonlyafewstructuresorculture-layerscanbeobserved.Insuch
cases, conditions that are specific for the single
sitehavebeenconsideredwhenidentifyingplots.
Theidentifiedplotsand-whereplotscouldnot
beidentified-theanalyticunitsarelabelledaccordingtoprinciplesoutlinedabove(cfp65ff).
Plot boundaries are plotted onto maps on
Figure28toFigure32accordingtotheirdates.
Boundariesidentifiedaccordingtotheboundary
132
indicators outlined above are drawn in a solid
lineonthemaps,andboundariesthatareidentifiedthroughconditionsspecifictothesinglesite
are drawn in a dotted line. Boundaries outside
thesitesarereconstructedinabrokenline.
Horizon2(c1020/30-c1070)
Inhorizon2,plotboundarieswereidentifiedat
threesites.Theseareinthenortherntownarea
only (Figure 28). The boundaries were all demarcatedbypalisadefences.Atsite6afence(S)
formedtheboundariesofaplot(Herteig1991,
97)labelled6/C.Anotherpalisadefence(S)may
indicateasecondplot(6/B)westofplot6/C.If
therewasathirdplotaswell-eastofthewell-definedplot-ajetty(B)wouldrunstraighttowards
theeasterncornerofthisplot,providingtheplot
wasofthesamewidthasplot6/C.Onthisbasis
ahypotheticalplot6/Dhasbeenreconstructed
onapreliminarybasis(cfp89ff).
The plots cannot have extended all the way
downtotheVågenshorelineasthefencesofplots
6/Band6/CtowardsVågenranparalleltothe
shoreline, delimiting the plots from the shore.
Thejettythatmayhaverunfromthehypotheticalplot6/Dtotheshorelineandafewscattered
postsinfrontofplot6/Cwerealsoassignedto
horizon2.Allinallthefencesthatranparallelto
theVågenshorelineappeartohavemarkedthe
extentofthebuilt-upareatowardstheshore.The
endofplot6/CtowardsVågenwas11.6mwide,
thelengthupthemorainicslopeisunknown.
Atsite9,apalisadefencethatranparallelto
theVågenshoreline,dividedthesiteintonorthern and southern parts. The area north of the
fence is labelled plot 9/A. There are no indicationsthatthefenceformedthetransversesouthernendofplotsinthesamewayasthesouthernmostpalisadefenceatsite6.
Twopalisadefences(S)clearlydividedsite11
intothreeplots,11/A,11/B,and11/C,thatran
at90degreestotheVeisanshoreline.Onlythe
size of plot 11/B could be measured. The plot
most likely extended to the Veisan shoreline,
where it was about 11.6 m wide by the reconstructed shoreline. The plot was approximately
12.1mwideabout15mfromtheshoreline.How
fartheplotextendedupthemorainicslopeisunknown,thepit-houseatsite7(S)hadthesame
Figure28.Boundariesidentifiedinsourcesassignedtohorizon2(c1020/30-c1070)
Figure29.Boundariesidentifiedinsourcesassignedtohorizon3(c1070-c1100)
9Plotsandplotsystemsinthetownarea
133
orientationastheplotsatsite11.Ifthepit-house according to the double tenement system. By
waslocatedonplot11/B,theplotcouldbe55-60 projectingtheboundariesfromhorizons3to5
mlong.
atsite9ontosite10thewaterboundextensions
oftheplotsidentifiedatsite9emerge.Theplots
Horizon3(c1070-c1100)
atsites9and10arelabelled9-10/Band9-10/C.
Inhorizon3,boundarieswerealsoidentifiedin Based on the reconstruction of the natural tothe northern town area only (Figure 29). The pographytheplotswereabout38mlong,they
palisade fences from horizon 2 at site 11 were ranat90degreestotheVågenshoreline.
also assigned to horizon 3 (S) and plots 11/A.
11/B,and11/Carereconstructed.Atsite6,there Horizon4(c1100-1120s)
werenolongerclearboundaryindicators.Only At site 11, the boundaries (S) from horizon 3
twobuildings,the‘9-postbuilding’(S)andthe continued in horizon 4 (Figure 30). At site 6,
‘possiblecellarbuilding’(S)wereassignedtenta- building45(B),apassageandaquayfront(B),
tivelytohorizon3atsite6.Culture-layershave and perhaps the 9-post building (S), formed a
not been documented in such detail that any doubletenementthatconformedtothewidthof
boundaries can be identified. Still, for several plot6/Cfromhorizon2,plot6/Cisreconstructreasons,itislikelythatthetwobuildingswere edinhorizon4aswell.Atsite8,theoldesttraces
relatedtoadefinedplot.Firstofallthelocation ofoccupationwereassignedtohorizon4(S)but
ofthetwobuildingswasrelatedtotheplotsize boundaries cannot be identified in the phases
fromtheprevioushorizon2plot(6/C)further- predatingc1170.Inthesucceedingphases,howmore the width of this plot was still respected ever,whenstructuresarebetterpreserved,asysin the succeeding horizons 4 and 5. Such con- tematicdifferenceintheorientationofstructures
tinuityinthelocationofboundariescanhardly inthenorthern,westernandeasternpartsofthe
be a coincidence. Therefore plot 6/C is recon- site can be observed clearly. The topographical
structedinhorizon3aswell.Theplotranat90 conditionsaswellasthebuildingpatternindidegreestotheVågenshorelineinhorizon3.At cate four plots. As the oldest material at site 8
site9,boundariesintheareasouthofthefence servesasasource(S)tohorizon4,Ifinditlikely
maybereconstructedinhorizon3bylookingat that the area was divided into plots already in
patterns in the younger material. South of the thishorizon.Iwillreturntothelocationofthe
palisadefencethebuildingpatternonsite9,in boundariesbelow.
horizons4and5,wasparalleltothatatsite6in
At site 14, a pier (S), interpreted as the seahorizons 4 and 5; the double tenement system wardextensionofastreet,alsoformedboundahadbeenintroduced.Thisisshownbypassages riesinthelandscape.Thepier/streetwasabout4
foundedon2mx2mstone-filledcaissons(S) mwideandranat90degreestothewaterfront,
andflankedbybuildings(S).Twopassagesare itprobablyextendedupthemorainicslope.The
reconstructed on the basis of the caissons. Ac- areaislabelledplot14/A.
cordingtothenumberofpassagestwoplotsare
Inhorizon4,theboundariesofaplotforthe
reconstructedinthesouthernpartofthesitein presumedearlyStMary’s(site23)inthenorthhorizons4and5,theyarelabelled9-10/B and erntownareaaretoouncertaintobeusedasa
9-10/C(seebelow).Returningtohorizon3,the boundaryindicator.Inthemiddletownarea,the
eastern wall extension of buildings 10 and 11 wallaroundthechurchyardoftheChurchofSt
(S) is in line with the reconstructed boundary Nicholas (S) (site 32) denotes the plot for this
betweenplots9-10/Band9-10/Cinhorizons4 church. Other plot boundaries could not be
and5,implyingthattheplotsinhorizons4and identifiedinhorizon4.
5werestructuredbyaplot-systemthatalready
existedinhorizon3.Henceplots9-10/Band9- Horizon5(1120s-c1170)
10/Carereconstructedinhorizon3aswell.
In horizon 5, there was continuity in the locaAtsite10,2mx2mstone-filledcaissons(S) tionoftheboundariesfromhorizon4.Inaddiindicatethatsite10,inhorizon4,wasbuiltup tion,site6wasnowclearlydividedintoseveral
134
Figure30.Boundariesidentifiedinsourcesassignedtohorizon4(c1100-1120s)
Figure31.
Boundariesand
buildingrowsat
site6,Bryggen.
(Modifiedfrom
Herteig1991,
Plate14and
1990,Figure85,
Figure56)
9Plotsandplotsystemsinthetownarea
135
plotsthatranat90degreestotheVågenshoreline(Figure31).‘Buildingrows4and3’47inthe
Gullskogårdenexcavationareamakeupplot6/C
still with the same width as the 6/C plot from
horizon 2. To the east of 6/C, building rows
2 and 1, also in the Gullskogården area, now
clearly make up a plot. The eastern boundary
of this plot coincides excactly with the eastern
boundaryofthehypotheticalplot6/Dassigned
tohorizon2,andwiththelocationofthejetty
from horizon 2. The horizon 5 plot is also labelled 6/D. Building row X and row Y in the
Søstergårdenexcavationareamakeupplot6/E.
Plots6/C,6/D,and6/Ewereoccupiedbytwo
rowsofbuildingsandapassagethatranbetween
the building-pairs, they thus make up typical double tenements with eavesdrops and wall
alignments indicating the boundaries between
plots.Intheremainingpartsofthesitethelayoutofthestructuresisnotimmediatelyclearand
boundarieswillhavetobediscussedonabroader
basis.The‘Engelgårdennorthrow’madeupthe
westernmost building row and caisson 64 was
partofthefoundationforapassageinthearea
thatIsuggestmakesupplot6/G.Ifweprolong
the alignment of the westernmost wall around
St Peter’s churchyard (site 24) towards the waterfront,itwouldcoincidewiththewesternmost
extension of building 203 in the ‘Engelgården
north row’, and thus form the westernmost
boundaryofplot6/G.Theeasternboundaryof
plot6/Gislessclear.However,ifweprolongthe
easternmostalignmentofcaisson88astheeasternboundary,enoughspaceisleftbetweenthis
boundaryandthepassageforasecondbuilding
row.Theplotwouldthenalsorepresentatypical double tenement. On this basis the eastern
boundary of plot 6/G is reconstructed. East of
6/G,labelled6/H,thereareindicationsthatthe
areawasoccupied.Theeasternlimitofthisplot
cannotbedeterminedthroughthematerialfrom
site 6 alone and will be discussed on a broader
basisbelowintheanalysisofplotsystems.The
areabetweenplot6/Eand6/Gbelongedpartly
totheSøstergårdenexcavationareaandpartlyto
the Engelgården excavation area. This plot, labelled6/F,isdefinedbytheboundariesofplots
6/Eand6/G.
IntheGullskogårdenarea,westofplot6/C,
136
buildingrows5and6andtheircommonpassage
havebeeninterpretedasawidedoubletenement
by Herteig. Furthermore, row 7 has been suggestedastheeasternmostrowofanotherdouble
tenement west of rows 6 and 5 (Herteig 1991,
108ff).Thisinterpretationdoesnotleaveroom
forthestreet(14/A)thatwaslocatedwestofsite
6atsite14,theareamaythereforehavebeenorganisedinadifferentwayfromthatsuggested.I
willdiscussthisonabroaderbasiswhenanalysingplotsandplotsystemsbelow.
At site 11 yet a boundary may be localised
through the wall alignment of building 12 (S)
towardstheareasouthwestofthesite,thisplotis
labelled11/D(Figure32).TheChurchofStPeter(S)(site24)maynowhavebeenlocatedaway
fromthewaterfrontandsurroundedbythewalls
of the churchyard (S). The plot of the Church
ofStPeterislabelled24/A.Thepresenceofa2
mx2mcaisson(S)atsite12indicatesthatthis
area was characterised by the double tenement
building layout and thus divided into plots. At
site15apassage(S),hasbeenassignedtohorizon 5. There was continuity in the orientation
andlocationofstructures,fromtheoldesttothe
youngestphasesdocumentedatthissite,soIassumethatthesitewaslocatedwithinoneplot.If
theplotwasofaboutthesamelengthastheplots
atsite6,site16wouldbepartoftheplotatsite
15,theplotisthereforelabelled15-16/A.Atsite
21,theoldeststructuresandculture-layerswere
assignedtohorizon5(S).Thestructuresandassociatedlayerswerelimitedtothenorth-eastern
halfofthetrench.Thenorth-easternpartofthe
trench is labelled unit 21/A, the south-western
partislabelledunit21/B.
TheextentoftheburialsatStMary’schurchyard,documentedatsite6,indicatesthesouthern boundary of the churchyard (plot 23/A).
Towardstheotherthreesidesofthechurch,the
churchyard has been estimated to be 10-20 m
wide(S).Attheremainingsitesinthenorthern
townarea,plotboundariescannotbeidentified
in horizon 5. The material from these sites is
treatedwithinanalyticunits.
Inthemiddletownareaseveralplotboundariescanbeidentified,inadditiontotheplotof
St Nicholas (B) (plot 32/A) (Figure 32). The
western and eastern parts of site 26 were built
Figure32a.Boundariesidentifiedinsourcesassignedtohorizon5(1120s-c1170)
Figure32b.
Boundaries
identifiedin
sourcesassigned
tohorizon5
(1120s-c1170)
9Plotsandplotsystemsinthetownarea
137
separately(B)inhorizon5andaneavesdropran
betweenthetwopartsintheimmediatelyfollowingphases,wherethepreservationofstructures
wasbetter.Itisthereforelikelythatthetwoparts
of the site belonged to two separate plots, they
arelabelled26/Aand26-27/B(seealsobelow).
Atsite27,three2mx2mstone-filledcaissons(B)indicatethatthesitewasbuiltupina
‘doubletenement’layout.Inphases3and4,that
followedthephasethatrepresentshorizon5,the
excavated site was divided into two plots, the
boundarybeingindicatedbyaneavesdrop(Golembnik1993,Figures21,31).Theplotboundaryfromthesephasesisprojectedontohorizon
5.Thewesternmostplotatthesiteisidenticalto
plotBatsite26.Thecommonplotonthetwo
Finnegården sites is labelled 26-27/B, and the
westernmostplotatsite27islabelled27/C.
Atsite28,seven2mx2mcaissons(B)indicate that the double tenement building pattern
was also established here and that the site was
dividedintoplots.Thepresenceofthisbuilding
pattern in the succeeding phases supports this
interpretation(cfLindh1979,figures).Thepositionofeavesdropsinphase2atthesiteimplies
thepresenceofthreeplotsatthesite.Inhorizon
5 a pier built on three of the caissons was the
forerunner for a passage on the middle plot. It
cannotbedeterminediftherewereanyplotshere
earlierthanhorizon5,asactivityonthesiteprior
to horizon 5 cannot be elucidated through the
available sources. The plots are labelled 28/A,
28/B,and28/C.
Site29wasprobablycharacterisedbythedoubletenementsysteminhorizon5,indicatedby
a2mx2mstone-filledcaisson(S).Usingmaterialfromearlierphasesandboundaryindicators similar to those used in the present study,
Dunlophasreconstructedaboundarybetween
twotenementsacrossthiscaisson(Dunlop1999,
Figure22).Twoplotshavethusbeenidentified,
labelled29/Aand29/B.
I have identified boundaries using the sources
thatwereassignedtothehorizonswithvarying
certainty.Iwillnowattempttodiscernpatterns
inthematerialthroughabroadspatialanalysis
ofthesourcesanddiscusswhethertheplotswere
laidoutaccordingtooverallsystems.
138
Oneorseveralplotsystems?
By the term plot system I refer to the overall
principlesofhowtheplotswerelaidout.These
principlesmayhavebeenbasedon(1)thestandardusedwhendividingareasintoplotsandon
(2) the relation to the shorelines of Veisan and
Vågen.Inordertodeterminewhethermoresystems are present in the material from horizon
2tohorizon5,Iwillstartoutbystudyingthe
system(s)oftheplotsinthenortherntownarea,
as plots have been assigned to horizons 2, 3, 4
and5here.
In horizon 2 the identified plots were related
totheshorelinesofVeisanandVågenintwodifferentways.Whileplots11/A,11/B,and11/Cat
site11mostlikelyextendeddowntotheVeisan
shoreline, plots 6/B and 6/C along the Vågen
waterfront did not extend to the shoreline, but
wereboundedbyfencesabout30mfurtherup
the beach instead. The fence at site 9 was also
withdrawnfromtheVågenwaterfront,running
parallel to the shoreline about 38 m further up
thebeach.Thetwohorizon2plotsthatcanbe
measuredwereofalmostexactlythesamewidth
(seeTable29formeasuresoftheplotsinthetown
area).Thesecircumstancessuggestthattheplots
were laid out according to the same standard -
andinasystemthatwasdirectedtowardsVeisan
ratherthanVågen.
Inhorizon3,plots11/A,11/B,and11/Cwere
stillorientedtowardstheVeisanshoreline.Along
Vågen,however,thebeachwasnowdividedinto
plotsthatextendeddowntotheVågenshoreline.
Thechangeiswell-documentedespeciallyatsite
6 as the lengthwise boundaries were respected,
whereas the crosswise were not. This may indicatethattheVeisanfocusedsystemofplotsfrom
horizon2wasreplacedbyanewsysteminhorizon 3, a system that was directed towards both
theVågenandVeisanshores.
In horizons 4 and 5, the plot boundaries
showedcontinuityinthelocationfromhorizon3
andextendeddowntotheshoresofeitherVeisan
orVågen.Inconclusion,thenortherntownarea
appears to include two plot systems: a Veisanbound system probably introduced during horizon 2, rearranged into the Veisan and Vågenboundsystemprobablyintroducedduringhorizon3andmaintainedduringhorizons4and5.
In the middle town area the first clear plot
boundaries appear along Vågen in horizon 5.
Thesettlementmay,however,gobacktohorizon
3accordingtoasupplementarysourcefromsite
26.Isitpossiblethatplotswerelaidoutinthe
middletownareabeforehorizon5inspiteofthe
lackofclearboundaryindicatorsinthematerial?
Thelackofidentifiedboundariesatthemiddle
town area sites before horizon 5 may partly be
explained by the topographical location of the
sitesintheVågenBayoronthewaterfront.Only
site26islocatedabove+/-0maslbeforehorizon
5andthesizeofthissite,coveringonly40m2,
mayexplainthelackofclearboundarieshere.
Ifinditlikelythatpeoplewhosettledinthe
Bergen area in horizon 3 would settle on land
thatwasdividedintoplots.Firstofall,because
thenortherntownareawasclearlydividedintoa
plotsysteminhorizon3,theconceptionofsuch
a division existed in Bergen when the middle
townareawasoccupied-probablyinhorizon3
(S).Furthermore,thekingseemstohaveowned
thelandintheBergenareabeforethetownwas
established (cf Helle 1982, 77-79 with references).Basedonthis,Ifinditunlikelythatnew
townspeoplecouldsettleanywheretheypleased
uponarrival.Soifpeopleoccupiedtheareaby
site26asearlyasduringhorizon3itisreasonabletosuggestthatboundarieswerelaidoutin
the middletownareaasfarbackasinhorizon
3,eventhoughithasnotbeenrecordeddirectly
thisearly.
The plots that were identified along Vågen
inthemiddletownarea,fromhorizon5,allextended down to the Vågen waterfront. In this
sensetheplotsystemseemstocorrespondtothe
systemfromthenortherntownareainhorizon
3.Thestandardormeasuresoftheplotsinthe
two town areas, however, diverge. Although
most of the measures are approximate, as they
are partly based on the reconstruction of the
naturaltopographyandthevaryingaccuracyof
the documentation, they clearly indicate different plot sizes as far as width is concerned (Table29).Inthenortherntownareathewidthsof
theplotsthatcouldbemeasureddonotchange
fromhorizon2throughhorizon5,infactthey
areclosetoidentical.Excludingplot24/A(StPeter’schurchyard),theplotwidthinthenorthern
9Plotsandplotsystemsinthetownarea
town area ranges from 10.0 m to 13.0 m (dependingonwherethemeasurementsaretaken),
withanaverageof11.7m.
If,hypothetically,thenortherntownareaplots
werelaidoutaccordingtoasystemwhereabout
11.5mwasthestandardwidthfora‘modelplot’-
thewidthof23.3mfortheseawardsboundaryof
StPeter’schurchyard(site24)wouldfitintosuch
a system by spanning the width of two ‘model
plots’.Thedistanceofabout24mbetweenthe
eastern side of the pier/ street (14/A) at site 14
andthewesternboundaryofplot6/Catsite6
alsocorrespondswellwithasystemcharacterised
byabout11.5mwideplots.OnthisbasisIsuggestthatthenortherntownareawasdividedinto
plotsaccordingtoasystembasedonastandard
ofa‘modelplot’about11.5minwidth.
Inthemiddletownarea,onlythewidthsoftwo
plots, 28/B and 26-27/B, could be measured
withsomeaccuracy(cfTable29).Withanaveragewidthofapproximately17.25m,bothwere
considerably wider than those that could be
measuredinthenortherntownarea.Doesthis
showthatthemiddletownareawasdividedinto
plotsaccordingtoadifferentsystemthanthose
inthenortherntownarea?48Oraretheapparentlywiderplotsinthemiddletownareajustan
irregularityinthesameoverallhorizon3system,
implyingthatthewidthofplotsissecondaryto
therelationofplotstotheVågenshoreline?
Ifinditlikelythatalltheplotsthatextended
totheVågenwaterfrontmayhavebeenpartof
the same overall system. Because if the middle
townareaplotsdatebacktohorizon3,asargued
above, and if we accept as a premise that there
wasachangeofplotsystemsfromhorizon2to
horizon3inthenortherntownarea,itseemsunreasonablethattwodifferentsystemswereintroducedinthetownareaatthesametime.
Howcanthedifferenceofwidthmeasuresin
themiddletownareaversusthoseofthenorthern
townareathenbeexplained?Again,ifweaccept
that the northern town area had already been
dividedintoapproximately11.5mwideplotsat
an earlier stage, a probable explanation for the
differenceinwidthcanbethattheabout11.5m
standardandthephysicallocationofthelengthwiseboundariesstructuredthesystemthatwas
introducedinthenorthernareainhorizon3.In
139
Table29.Plotswherethelengthorwidthcanbemeasured,horizons2-5
Plot/
11/B
11/C
9/B
9/C
6/C
6/D
6/E
6/F
6/G
6/H
24/A
Northerntownarea
Horizon
2-5
5
Widthbytheshore
11.6
Widthabout15m
fromtheshore
12.1
4-5
4-5
Length
>55.060.0
5
5
5
5
11.0
11.7
11.7
11.3
10.0
12.0
12.1
11.8
12.5
13.0
12.0
11.9
23.3
11.9
32.0
38.0
38.0
themiddletownarea,however,nooldersystem
existedwhenplotsweretobelaidoutduringhorizon 3 and the width of plots could be determinedwithoutconsiderationofanoldersystem.
Tosumup,Ihavearguedthattwoplotsystems existed from horizon 2 to horizon 5. Accordingtothistheoldestsystemwasestablished
inthenortherntownareaduringhorizon2and
consisted of approximately 11.5 m wide ‘model
plots’.TheseplotsextendeddowntotheVeisan
shorelinebutnottotheVågenshoreline.Iregard
thisasanindicationthatthissystemwasdirected
towardstheVeisaninletratherthantowardsVågen.Anewplotsystemwasprobablyintroduced
inthenorthernandmiddletownareas,during
horizon 3 and maintained through horizon 5.
WithinthissystemtheplotstowardsVågenextended all the way to the waterfront. This plot
system was thus more oriented towards Vågen
Baythantheoldersystem.Inthenortherntown
area,theplotwidthandthelocationoflengthwiseboundarieswereprobablystructuredbythe
older system where the ‘model plot’ was about
11.5mwide.Inthemiddletownarea,plotswere
laidoutaccordingtothesamesysteminterms
ofthefocusonVågenBay,butthewidthofthe
plotsdivergedfromthoseofthenortherntown
area.
140
2627/B
Middletown
area
2-5
Widthatthelandsideendofplot
28/B
19.0
5
14.5
5
5
5
18.8
15.7
32.0
Evaluationoftheplotsystems
discernedandcentraldatesofsources
assignedtohorizons2and3
Theexistenceofplotsinthenorthernandmiddletownareashasbeenshownthroughrespectively fourteen and five well-identified boundaries.Thereshouldthusbenodoubtthatboundaries existed in these town areas in the period
understudy.Patternsformedbythewell-identifiedplotsanddiscernedthroughabroadspatial
analysis of the sources give the suggested plot
systemsareliableempiricalbasis.Uncertainties
concerningtheplotsandplotsystemsdiscerned
shouldthusbemoreattachedtothechronology
ofthestructuresthantothefactualexistenceof
boundariesandsystems.
Thesourcesfromhorizons2and3weretentativelyassignedtothesehorizonsthroughpatterns
inthematerialonthesinglesitesandtheirclose
vicinities.Howdothesuggestedplotsystemscoincidewiththesepatterns?Andcanthepatterns
discernedthroughthebroadviewofthesources
strengthen the initial assignment of the rather
poorlydatedmaterialtohorizons2and3?
The palisade fences at sites 9 and 11 were
thought to be contemporary with the palisade
fencesatsite6onthebasisofanumberofcircumstances (cf the discussions on pages 89ff,
103ff,110ff,and155ff).Aboveithasnowbeen
shownthatthewidthsofplots11/Band6/Cwere
almostidentical,thisisyetanothercircumstance
supporting the hypothesis that the palisade
fencesandassociatedplotswerelaidoutcontemporaneously. Furthermore it is shown through
boundary indicators and diagnostic structures,
toalargeextentbasedonbasicsources,thatthe
widths of the palisade-bounded plots at site 6
were respected in the succeeding horizons, and
that the plots identified in the northern town
areaaltogetherseemtobepartofageneralsystemwhereapproximately11.5mmayhavebeen
thewidthofthe‘modelplot’.Thisalsosupports
thehypothesisthatthepalisadefencesatsites6,
9 and 11 were constructed contemporaneously.
In this respect the patterns discerned through
the broad spatial analysis lends support to the
suggestedcontemporaneityofthefences.
Thepalisadefencesandassociatedplots6/B
and6/Catsite6weretentativelyassignedtohorizon2amongstotherthingsthroughthehorizontal link made between the well-dated jetty,
assigned to horizon 2 and a hypothetical plot
6/Deastofplot6/C.Whentheareacoveredby
thehypotheticalplot6/Dwasbuiltoninhorizon5,afactualplot6/Disclearlydiscernedin
thematerial.Thelocationofthehorizon2jetty
corresponds exactly to the eastern boundary of
this plot. This co-location of the jetty and the
boundaryishardlyacoincidenceandIconsider
thisastrongindicationthatthejettyfromhorizon2actuallydidleaduptoaplotinhorizon2.
Thehypotheticalplot6/Dassignedtohorizon2
isthusclosetobeingestablished.Thisstrengthensthehorizontallinkmadebetweenthewelldated jetty and the palisade-bounded plots at
site6,andthussupportstheassignmentofthe
palisade-built fences to horizon 2. This also
strengthens my general assignment of sources
fromsites6and9tohorizon3,asthesuggested
timedepthofthematerialfromthesesitesisbetterestablished.
This in turn strengthens my suggestion that
oneandthesameplotsystemwasintroducedin
thenorthernandmiddletownareasduringhorizon3;thissuggestion,ismainlybasedonthe
premise that the change of systems seen in the
northerntownareatookplaceduringhorizon3.
Allinallitseemsthatthepatternsdiscerned
inthematerialwhencarryingoutabroadspatial
analysissupportmyinitialassignmentofcentral
sourcesinthenortherntownareatohorizons2
9Plotsandplotsystemsinthetownarea
and3.Thisdoesnotmeanthatthesourcesfor
horizons2and3arenowwell-datedorthatthe
lastwordissaidonthematter.However,atthe
presentstageofresearch,thedatespresentedhere
representanalternativethatdoesnotinvolvetoo
many unlikely ‘coincidences’. Thus I maintain
the dates suggested in Chapter 7, bearing in
mindtheuncertaintiesinherentinthematerial
inthefollowinganalyses.
Theextentofthetwoplotsystems
Thehorizon3system
Iwillfirstreconstructtheareacoveredbythehorizon3system,becausethereconstructionofthe
extentofthehorizon2systemmustbebasedon
whatcanbeinferredfromthehorizon3system.
Judgingbytheidentifiedboundaries,thelength
oftheplotsseemstohavevaried,soIhavenot
triedtoreconstructtheplotsintheirfulllength.
Figure33showsthesuggestedreconstruction
of the horizon 3 system. In the northern town
area,thebuildinglandbetweenStPeter’schurchyard(plot24/A)fromhorizon5andplot15-16/
Aconformtotheapproximately11.5m‘model
plot’systemwhendividingthedistancebetween
St Peter’s churchyard’s easternmost boundary
and the passage at plot 15-16/A by 11.5. The
same applies to the area between plot 6/C and
theeasternboundaryofthepier/street14/A.
Thestretchoflandbetweenthewesternlimit
of the pier/street and the western boundary of
plot9-10/Cismorecomplicated.Areconstructionoftheplotsinthisareadependsonwhether
ornotoneincludesthepier/street(14/A),which
isassignedtohorizon4(S),asanoriginalpartof
thehorizon3system.Theareasouthofthefence
atsite9mayhavebeendividedintoplotsduring
horizon3(S)andwasthus,probably,included
inthehorizon3systemfromthebeginning.If
thestreetwasnotplannedorbuiltuntilhorizon
4, one would expect that there was insufficient
space for the street. The distance between the
westernboundaryofplot9-10/Candtheeastern
boundaryofthestreetisabout73.5mandthe
distance between the western boundary of plot
9-10/C and the western boundary of the street
is about 69.5 m. Dividing these measurements
by11.5therewouldberoomforrespectively6.1
141
Figure33.Theareacoveredbythehorizon3plotsystem
Figure34.Site8,Dreggsalmenningen14-16.Theorientationofstructuresyoungerthanhorizon3
142
or5.8‘modelplots’betweenthestreetandplot
9-10/B. Based on this calculation it is slightly
more likely that the street was not included in
the horizon 3 system from the beginning. The
calculation,however,isproblematic:the11.5m
standardforamodelplotisonlyanestimateand
asmallchangeofthestandardwouldchangethe
calculationandindicateanotherconclusion,the
calculationisthereforedisregarded.
If, hypothetically, the street was originally a
part of the plot system introduced in horizon
3, one could also argue that the Church of St.
Mary’s was part of the original system.49 Some
circumstances may point in that direction; the
orientationofthelaterStMary’scertainlysuggeststhattherewasroomenoughforthechurch
when it was constructed and that the church
thereforewasanearlyelementinacomprehensivetownplan.Also,onemaysuggestthatwhen
theplotsonsite6wereprolongedtowardsVågen
bytheintroductionofanewplotsysteminhorizon3,theywereshortenedattheotherend,thus
makingspaceforachurch.OnthisbasisIsuggestthatthestreet(14/A)andaplotearmarked
for a church (23/A) may have been part of the
plotsystemintroducedinhorizon3.Whetheror
notachurchandastreetwerealsoconstructed
during horizon 3 cannot be established at the
presentstateofresearch.
Betweenplots9-10/Cand11/Cthecurvature
ofthenaturaltopographymakesitdifficultto
reconstructplotsjustbymeasuringoutapproximately11.5mwide‘modelplots’alongtheshore
by the mouth of the Veisan inlet. As already
mentioned no boundaries can be identified at
site8inthephasespredatingc1170.Inthefollowingphases,however,asystematicdifference
intheorientationofstructuresinthedifferent
partsofthesitebecomesclear(Figure34).Observingtheorientationofstructuresfromhorizons4and5andyoungerstructuresonsite8,
it seems clear that the northern part of site 8
belongedtoaplotthatranmoreorlessparallel
toplots11/A-C.Thisareaislabelledplot8/A.
Thesourcesfromthisplotincludelayer684in
horizon4andanumberofculture-layersinhorizon5.Thewesternpartofsite8seemstohave
belongedtoaplotthatranmoreorlessparallel
tothefence(plot9/A)atsite9.Thisareaisla9Plotsandplotsystemsinthetownarea
belledunit8/Bandincludesbuilding158with
associatedstructuresinhorizon4andbuildings
K166,K145/152/157andassociatedstructures
inhorizon5.Regardingthemiddleandeastern
partsofsite8,theycannotbeassignedtoany
definite plots at the present stage of research.
However,thefourpostsK136inthemiddleof
thesiteinhorizon5,maybelongtooneanalytic
unit:8/C,andbuildingK102/104intheeasternpartofthesitemayalsobetreatedasone
unit:8/D.
Inthemiddletownarea,noattempthasbeen
madetoreconstructplotsthathavenotbeenrecordedarchaeologicallybecausethesizeofthe
plotsseemstovary.However,Iassumethatthe
wholeofthemiddletownareaalongtheVågen
shorelinewasalsoregulatedintoplots.
In summary, the plot system assigned to
horizon 3 probably covered the area along the
Veisan and Vågen shorelines in the northern
and middle town areas. Whether the system
alsoincludedtheareasclosertothefootofFløyfjelletandthesoutherntownareaisimpossible
todecideonthebasisoftheavailablematerial.
Plotsmayperhapsalsohavebeenparcelledout
forapier/streetandachurch,wheretheChurch
ofStMarywaslaterbuilt.Ifso,theplotsystem
maybecharacterisedasarathercomprehensive
townplan.
Thehorizon2system
Basedonthelocationofthepalisade-builtfences,
thehorizon2plotsystemoughttohavecovered
theareabysite6,thenorthernpartofsite9and
theshoreofVeisanatleastuntilsite11.Wedo
notknow,however,howfareastandnorththe
systemextendedbeyondthisareatowardsFløyfjelletortowardstheeastalongtheVågenBay.
Aswehaveseenabove,thedistancebetweenthe
easternmost boundary of St Peter’s churchyard
(24/A) and the reconstructed passage at site 15
correspondswellwithaplotsystembasedonthe
approximately11.5mstandardwidth.Iftheplot
width between 24/A and 15-16/A indeed was
structured by the suggested approximately 11.5
mmodelplot,asarguedabove,thisareamight
havebeendividedintoplotsofthissizealreadyin
horizon2.Otherwise,theorganisersofthenew
plotsystemintroducedinhorizon3wouldhave
143
Figure35.Theareacoveredbythehorizon2plotsystem
been free to apply other measurements to the
plotshereastheymayhavedoneinthemiddle
townarea.Thisisnotwell-foundedempirically
butneverthelessseemsquiteprobable.Ithussuggestthattheareaeastofsite6wasdividedinto
approximately11.5mwideplotsalreadyduring
horizon2.Atsite6,thefence(s)thatmarkedthe
transverse plot boundaries ran along the beach
ridge between 2 and 3 masl, the extent of the
regulated area is reconstructed along this line.
Theprotrudingrockreconstructedbetweenthe
northern and the middle town areas may have
functioned as a natural topographical ‘barrier’
againstanextensionofthehorizon2systemfurthereastintothemiddletownarea.Itcannotbe
determinedhowfartowardsFløyfjelletthesystemwasapplied.Figure35showstheareathat
144
may have been included in the horizon 2 plot
system.
Conclusions
BoundarieswerediscernedalongtheVeisanand
Vågenshorelinesinthenorthernandthemiddle
townareasonly.Twoplotsystemsmayhaveexistedintheperiodunderstudy.Theexistenceof
plotsandplotsystemsiswell-foundedempirically.Thedateoftheplotsystemsisnotsowellsubstantiated,butdatingtheearliestplotsystemto
horizon2andthelatersystemtohorizon3seems
tobethebetteralternativeatthepresentstageof
research.Theearlysystemprobablycoveredthe
northerntownarea,inthissystemtheplotsseem
to have been directed more towards the Veisan
shorelinethantowardstheVågenshoreline.The
latersystemcoveredboththenorthernandthe
middle town areas, this system seems to have
beenmoreorientedtowardstheVågenwaterfront
thanthefirst.Theearlylengthwiseplotboundariesapparentlystructuredthelocationofthenew
plotsinthenortherntownarea,astherewascontinuityinalllengthwiseboundariesdocumented
in the northern town area, the approximately
11.5m‘model’widthoftheplotswasthusmaintained.Plotsofvaryingwidthswerelaidoutin
themiddletownarea.Thenewplotsystemmay
possiblyhaveincludedspaceforachurchanda
street,inwhichcaseitmaybecharacterisedasa
rathercomprehensivetownplan,however,thisis
notsowell-foundedempirically.
10TOWHATEXTENTWASTHE
BERGENAREA‘OCCUPIED’?
In this chapter I will take a closer look at the
questionsconcerningtheextentandcharacterof
landusewithintheplotsandunits(cfp67).The
relationship between structures/buildings and
thetidalzoneandshorelineswillalsobestudied,
andIwilldiscusswhetherthestructures/buildingsalongtheVeisanandVågenshorelineswere
confinedto
• theareaabovethetidalzone,indicatinglow
pressureonbuildingland
• trespassed the physical boundary into the
tidal zone, indicating pressure on building
land
• orexpandedbeyondthewaterfront,indicating high pressure on building land and/or
investmentindeeperharbours
Horizon2(c1020/30-c1070)
Theexcavatedsiteswithindicationsofthegenerallanduseinhorizon2onlycoverafewofthe
plotsinthesystemthatwasprobablyintroduced
in the northern town area. Along the Veisan
shoreline the plots at site 11 were probably not
occupied (S) (Figure 36). At site 7, a pit-house
K19(S)andafire-layer(S)thatcoveredanarea
ofatleast100m2indicatethatthissitewasoccupied.Whetherthestructureswerelocatedon
plot11/Bsite11(cfabove)cannotbedetermined
10TowhatextentwastheBergenarea‘occupied’?
onthebasisoftheavailablesources.Site9was
divided into plot 9/A oriented towards Veisan
and a southern area oriented towards Vågen.
Southofplot9/A,aculture-layer(G),deposited
ontothefencemayreflectactivityinthisareain
horizons2or3;however,sincethequestionof
chronologycannotbesettledIshallhavetoomit
thelayerasasourceforthepresentquestion.On
site 6, the jetty (B) stretching across the beach
towardsplot6/Dindicatesthatthisplotwasoccupied.Butthelackofculture-layersandstructures (S) in the close vicinity implies that the
jetty was the only structure on this part of the
beach(cfp85).Totheeast,thelackofstructures
and culture-layers in ‘unit 7’ (S) indicates that
thispartofthebeachwasnotsettled.Onplot6/
C,atleasttwoposts(S)wererecorded,indicatingthatthisplotwasoccupied.Scatteredposts
(S)outsidetheplot,towardsthewaterfront,may
alsobelongtohorizon2,thusindicatingthatthe
beach immediately beneath plot 6/C was built
on.Thestructuresareconfinedtotheareaabove
thetidalzone.Onplot6/Bnoinsitustructures
orculture-layers(S)havebeendocumented,but
reusedtimbersfrombuilding66,locatedonthe
plotinhorizon5implythattheplotwastaken
intousealreadyduringhorizon2(S)(cfp65).
On site 15 redeposited culture-layers (G) have
been documented prior to horizon 5. If they
originatedinhorizon2theyimplythatplot1516/Awasoccupiedalreadyduringthishorizon.
On sites 8, and 16-21 the lack of structures or
culture-layers(S)assignedtohorizon2,implies
thattheanalyticunitswerevacant.
Inthemiddletownareathepier(B)fromhorizon1wasstillinuse(unit30/A).Furtherupthe
shore(unit30/D),thethreeposts,perhapsdated
tohorizon1(G)mayalsostillhavebeeninuse.
Thepierandthepostsbothimplythatthearea
wasoccupied.Thebuiltareawasprobablylimited
totheclosevicinityofsite30,astherewasalack
ofstructuresandculture-layersontheneighbouringsites26,27,31,and34andinprofiles26-29
andV3(units30/Band30/E)atsite30.
The lack of structures and culture-layers at
sites35-36and38inthesoutherntownarea,also
implythattheareawasvacant.AtNordnesand
intheNonneseterareatherewerenotracesofsettlementinhorizon2.
145
Figure36.Horizon2(c1020/30-c1070)occupiedandvacantanalyticunits
146
ChristchurchMinor,wascompletedinthereign
ofOlavKyrrewhereastheconstructionworkon
the Christchurch Cathedral had not come far.
MaritNybøarguesinherstudyoftheChurch
ofStAlbanyatSelje,thattheconstructionofthe
Christchurch Cathedral must have been under
way in the reign of Olav Kyrre, otherwise the
churchcouldnothaveservedasamodelforthe
ChurchofStAlbanyatSelje(Nybø2000,192193).Thiscouldbeanargumentforintenseactivities at Holmen in horizon 3. On the other
hand, a planned church in Bergen could also
haveservedasamodel,sotheargumentshould
not be given too much weight, when estimatingtheextentofactivityatHolmeninthedays
of Olav Kyrre. All in all, we may still presume
thatafairamountofbuildingactivitytookplace
at Holmen in horizon 3. We do not know the
numberofindividualsthatwereinvolvedinthe
building processes, but the construction workers must have been accommodated somewhere.
Asweshallseeitseemsthatthetownareawas
scarcely occupied, so one must probably look
elsewhere for the settlement that housed construction workers. There was ample space for
housingconstructionworkersatHolmen,butas
we have seen, the sources cannot elucidate the
extentofasettlementhere.
NocleartracesofoccupationonplotsdocumentedalongtheshoreoftheVeisaninletcould
be assigned to horizon 3. At site 9, as many as
three buildings (S) may have been constructed
abovethetidalzoneonplot9-10/B,thusindicating occupation along the Vågen shoreline. On
plot9-10/Cnotracesofoccupation(S)couldbe
assignedtohorizon3,thismayperhapsindicatea
vacantplot.Plot6/Cwaslevelledwithsandand
gravelbefore‘the9-postbuilding’(S)andapossible cellar building (S) were constructed, both
abovethetidalzone.Thejettyfromhorizon2,
probablyassociatedwithplot6/D,maystillhave
beeninuse.Sandandgravellayers(S)weredepositedonthebeacharoundthejettyperhapsas
Horizon3(c1070-c1100)
earlyasinhorizon3(cfp85ff),indicatingthat
According to the written sources King Olav the plot was occupied. ‘Unit 7’, a marine layer
KyrrebegantheerectionoftheChristchurchCa- depositedinVågenjustoutsideplot6/Eor6/F,
thedralatHolmen(B)(site2),andhealsobuilt has also been tentatively assigned to horizon 3
Christchurchminor(B)(site3)onthechurch- (S),thusindicatingthatoneoftheseplotswas
yardoftheChristchurchCathedral(Figure37). occupiedthisearly.Atplot15-16/A,redeposited
Summary
To sum up, data from 30 analytic units in the
three town areas have been analysed as sources
for horizon 2.50 Occupation was indicated in
sevenofthese.Dataindicatingoccupationcome
fromtwobasicsources,threesupplementaryand
twogeneralbackgroundsources,andareindeed
ratherscanty.Evenso,ifthesourcesarecorrectly
assignedtohorizon2,theydoindicatethatsome
plots and units were occupied in the northern
townareaandnearsite30.
In23ofthe30unitstherearenostructuresor
culture-layersthatcouldbeassignedtohorizon
2, this information is used as a supplementary
sourceforhorizon2.Thevacantunitsalongthe
footofFløyfjelletwerenotinterrelatedandsince
theypointinthesamedirection,thisimpliesthat
theareawasnotoccupied.Thusitseemsreliable
thattheseareasinthenorthernandmiddletown
areaswerenotoccupied.Thevacantunitsinthe
southerntownareaarenotinterrelatedandalso
point in one and the same direction, implying
thatthisareawasnotoccupied.However,given
thelackofdataanddistancebetweenthesources,nostrongconclusionscanbemadeaboutthe
generallanduseinthispartoftheBergenarea.
Accordingtothetrendsdiscernedinthematerial,itseemsthattheareasclosetotheVeisan
andVågenshorelinesinthenortherntownarea
and site 30 (unit 30/A and 30/D) in the middle town area were occupied. Where structures
otherthanthejettiesorpiersweredocumented,
they were confined to the area above the tidal
zone.Thisimplieslowpressureonbuildingland.
Altogether the evidence is scanty, but seems to
drawinthesamedirection,implyingthatsome
plotsorunitswereoccupied,butthat,generally,
pressurewaslowonbuildinglandduringhorizon2.Ifthegeneralbackgroundsourcesareerroneouslyassignedtohorizon2,theimpression
oflowpressureonbuildinglandbecomeseven
stronger.
10TowhatextentwastheBergenarea‘occupied’?
147
culture-layers(G)mayperhapsindicateactivity
in the vicinity. No structures or culture-layers
(S)weredocumentedonsite17,indicatingthat
thisareawasvacant.
On site 26 in the middle town area, reused
timber (S) at plot 26/A indicates that the area
wasoccupiedabovethetidalzone.Redeposited
culture-layers(S)atsite27likewiseindirectlyindicatesettlementinthearea.Onsite30thepier
(site 30/A) (B) and the three posts (site 30/D)
(G)fromhorizon1,representedtracesofoccupationhere,whereastherewerenotracesofoccupationinunits30/Band30/E.
Atsite20,aconstructioninterpretedaspart
ofabuilding(S)wasassignedtohorizon3and
represents the first traces of occupation at the
footofFløyfjellet.Betweensite20andsite30in
themiddletownareanootherremainsofoccupationhavebeenfoundattheinvestigatedsites
(sites16,18,19,21/A21/B,and31),indicating
thattheseareaswerevacant.
Inthesoutherntownarea,fromsite30tothe
headofVågenBay,nostructuresorculture-layers
havebeenassignedtohorizon3.IntheNordnes
and Nonneseter areas no traces of occupation
havebeendocumented.
Summary
Insummary,twochurcheswereinitiatedatHolmen,botharedocumentedthroughbasicsourcesandtheirpresenceisconsideredreliable.Data
from 30 analytic units in the three town areas
havebeenusedassourcesforthesecularsettlement in horizon 3. Traces of occupation were
documented in ten units in the northern and
middle town areas only. The settlement traces
weredocumentedthroughonebasic,sevensupplementaryandtwogeneralbackgroundsources.
Sincethebasicsourcesforthepresenceofsettlement traces are so scarce, the picture drawn of
occupiedunitsmaynotberelieduponinevery
detail.Thegeneraltrendsinthematerial,however,oughttobequitereliable,asnotallthesupplementarysourcesareinterrelated.
The lack of settlement was implied in 20
units. The vacant units along the foot of Fløyfjelletandinthesoutherntownareamaystillbe
evaluatedalongthesamelinesofthinkingasfor
148
horizon2.Theextentofvacantareasatthefoot
of Fløyfjellet in the northern and middle town
areas,indicatedbythesources,oughttobequite
reliable,datafromthesoutherntownareamay
beevaluatedasforhorizon2.Thevacantplots
byVeisanaredocumentedthroughsourcesfrom
twositesthatwerenotinterrelated,thisenhances
the likelihood of vacant plots in this area, and
oughttobetrustedasageneraltrend.Thelack
ofsettlementonplotsalongVågenwasonlyimplied by one supplementary source; that is the
lackofstructuresatplot9-10/C,andcannotbe
usedasconclusiveevidencethatsettlementwas
missinghere.Ifthetracesofactivity,assignedto
plot15-16/Acannotbedatedasearlyashorizon
3, they do however support the idea that some
plotsalongVågenwerenotyetinuse.Itisdifficult to be conclusive in this matter, so I will
letthequestionofvacantplotsalongVågen,in
horizon3,remainunanswered.
Altogether then, according to the general
trends in the material, most of the occupied
plots/sitesseemtohavebeenlocatedbytheVågenshoreline.Thismayimplythattheareawas
consideredtheprimeareaforsettlement.Where
buildingsweredocumented,theywereconstructedabovethetidalzone,indicatingthatpressure
on building land was low. Along Veisan, some
plotswereprobablystillvacant.Alongthefoot
ofFløyfjellettheavailablebuildingspaceseems
tohavebeenvacantwhenlookingapartfromthe
settlementatsite20inthenortherntownarea.
Theevidenceisgenerallyscarce,butseemsto
beinthesamedirection,leavingtheimpression
ofarathersparselybuilttownarea.Itseemsthat
pressurewaslowonsecularbuildingspaceand
emphasis was on the monumental side. If the
general background sources are erroneously assignedtohorizon3,thisimpressionisstrengthened.
ThesagassaythatafterOlavKyrre’sfoundation of Bergen, the town soon became a place
formanyrichmen(Hkr1893-1901,III226;Fsk
305; Msk 289; Helle 1982, 86). If by this descriptionthesagasmeanttosaythatBergensoon
becameadenselyinhabitedorpopulatedplace,
the description does not correspond well with
thegeneraltrendsinthearchaeologicalsources.
Thesagasarequiteremoteintimetotheevents
Figure37.Horizon3(c1070-c1100)occupiedandvacantanalyticunits
described (cf p 57ff) this may be one explanationforthediscrepancybetweenthesourcecategories. Another aspect to consider is that it is
hard to determine what the chroniclers meant
by‘many’and‘soon’.Thearchaeologicalsources
areveryspecificandcanbequantifieddirectlyas
10TowhatextentwastheBergenarea‘occupied’?
opposedtothedescriptionhandedovertousby
the medieval chroniclers. And ‘soon’ may from
thepositionofthethirteenthcenturychroniclers
bemuchlaterthanhorizon3,orhorizon4for
thematter.ConsequentlyIhavechoosentodisregardthedescriptionprovidedbythesagasasa
149
sourceabouttheextenttowhichtheBergenarea (S),accumulatedaswell,probablyreflectingacwasoccupied.
tivityonplots6/Eor6/F.Atplot15-16/A,redepositedculture-layers(G)mayindicateoccupaHorizon4(c1100-1120s)
tioninthevicinity.Notracesofoccupationwere
AtHolmen,archaeologicalexcavationshavenot documentedonsite17.
revealedanynon-monumentalstructuresthatcan
In the middle town area, a triangular stonebedatedtohorizon4(Figure38).Accordingto filledcaisson(B)wasregistredonplot26/A.On
thewrittensourcestheChristchurchCathedral plot26-27/Bthefoundationofaquayorabuild(B)(site2)andChristchurchMinor(B)(site3) ing (B) was registred. Both constructions were
werestillstandingandtwonewmonuments,the locatedinthetidalzone,indicatingpressureon
Church of the Apostles (B) and Øystein Mag- buildingland.Detritusfromconstructionwork
nusson’slargetimberhall(B)wereconstructed. (S)andsixlayers(S)weredepositedattheneighAlongtheVeisanshorelinenotracesofoccu- bouringsite27.Theselayersindirectlyindicate
pationhavebeendocumentedontheplotsatsite that the shore behind the site (plot 27/C) was
11.Onsite8,layer684(S)wasrecordedonplot occupiedinhorizon4.Atsite30,thepier(30/
8/A, and interpreted as remains of occasional A),erectedinhorizon1(B),wasprobablystill
activity (Golembnik in prep-b). In the Bergen standing.
sourcesIhavegenerallynotbeenabletodistinAtthefootofFløyfjelletinthenorthernand
guish between culture-layers representing occa- middle town areas traces of secular settlement
sional occupation and layers representing con- werefoundatsite20only.K20(S),interpreted
tinuousoccupationbasedonthepremisesused as part of a building, was thus assigned to hobyGolembnik(seealsoChapter13).Therefore, rizon 4. At sites 16, 18, 19, 21 and 31 no cullayer684canonlybeusedtorepresentsecular ture-layersorstructures(S)couldbeassignedto
occupation in general. Building K158 (S) was horizon 4, indicating vacant areas. On site 30,
constructedinunit8/B.Onplot9-10/Btowards cultivationlayersandanumberofpits(S)interVågen,buildings13,10and11andcaisson2(S) pretedassandextractionholesrepresentlanduse
have been associated with horizon 4, and indi- atsomedistancefromtheVågenshoreline(30/
catethatthisplotwasoccupied.Onplot9-10/C, D). According to the categories used here the
building12andcaisson1likewiseindicatethat pits represent ‘occupation’ whereas the cultivathe plot was occupied (S) and on plot 12/A a tionlayersrepresentcultivationinabroadsense,
caisson(S)impliesthatthisplotwastakeninto sothesourceissomewhatambiguousastothe
use.Two4mx4mstone-filledcaissons(S),in- characteroflanduseintheunit.Whenlooking
terpretedasthefoundationforapier,implythat atthesourceinthecontextofsurroundingunits,
plot 14/A was taken into use. A variety of cul- abasin(S)usedinconnectionwithfresh-water
ture-layersandstructuresatsite6showthatat supplyhasbeendocumentedintheclosevicinleastplots6/B-Gwereoccupied:onplot6/Cthe ity,thisbasinindicatesthattheareawasnowoc‘9-postbuilding’(S),erectedduringhorizon3, cupied(30/E).AccordinglyIinterpretthecultimaystillhavebeeninuse.Itismorecertainthat vationlayersasbelongingtoasettledareainthe
agravellayerwasdepositedtopreparethebuild- closevicinity.
inglandinfrontofthe‘9-postbuilding’towards
TheChurchofStNicholas(site32)(S)may
Vågenandthatbuilding45,apassageandaquay have been located at the foot of Fløyfjellet in
frontwerealsoconstructedinthetidalzoneon the middle town area. Burials and walls of the
thisplot(B).Postsfromthehorizon2jetty,as- churchyardmaydenotetheextentofthechurchsociatedwith6/D,werestillvisibleonthebeach, yard. The church was oriented SSE – NNW.
indicatingthatthejettywasperhapsalsostillin Withapointofoutsetinthesomewhatirregular
use. Sand and gravel layers were deposited on orientationoftheChurchofStNicholasithas
thebeacharoundthejetty(S).Finally,layersof beenarguedthatthechurchhadtoadapttoan
smallstones(B)werespreadoverplots6/B-Gto areathatwasalreadysodenselybuiltoutthatthe
consolidatetheground.‘Unit7’,amarinelayer ideal orientation could not be achieved (Helle
150
Figure38.Horizon4(c1100-1120s)occupiedandvacantanalyticunits
1982,149).Thelackofsecularoccupationtraces
onsitesintheareaalongthefootofFløyfjellet
does not support such an interpretation. Furthermoretheorientationofthechurchcoincides
well with the general topographical orientation
ofthemountainsidebehindthechurch(Hansen
1994b, Figure 32). The building was located
10TowhatextentwastheBergenarea‘occupied’?
onaterraceatapproximately20to25maslon
the mountainside just below a point where the
mountainrisesrathersteeply.Ifthechurchhad
beenorientedstrictlyE-Wtherewouldnothave
been enough space for a large building. The
chancelwouldhavefacedthemountainsideand
thebuildingwouldhaveturned‘itsback’tothe
151
town,losingitsimpressiveanddominatingpositionasalandmark.Inthenortherntownareaa
predecessor to the standing St Mary’s (S) (site
23)mayhavebeenerectedinhorizon4.
In the southern town area the sites are few
and far between. No traces of occupation have
beendocumentedonanyofthesites.Notraces
ofoccupationhavebeenlocatedintheNonneseterareainhorizon4.Accordingtothewritten
sourcestheabbeyofMunkeliv(site43)(B)was
established at Nordnes. Munkebryggen, a pier/
quayonthewesternsideofVågenalsobelonged
totheabbey,butthelocationofthequayisunknown.
foundedandthequestionofvacantlandherein
horizon4remainsopen.
Altogether,thesourcesaremorefirmlyfoundedthanintheearlierhorizons,andtrendsinthe
material should be reliable. Accordingly, it appearsthatsomeVeisanshorelineplotswerenow
occupied.AlongVågen,occupationwasindicated
onmostdocumentedplots,bothinthenorthern
andthemiddletownarea.Wherebuildingshave
beendocumentedinthenortherntownarea,they
areaboveorontheedgeofthetidalzone,indicatinglowpressureonsecularbuildingspace.Inthe
middletownarea,buildingsmayhavetrespassed
intothetidalareaindicatingpressureonbuildingland.Passagesthatledacrossthetidalzone
Summary
to quays at the waterfront were built on some
Tosumup,asmanyasfivemonumentalbuild- plots along the Vågen shoreline. Along the foot
ingsmayhavebeenerectedatHolmen,onNor- ofFløyfjellet,thechurchofStNicholasprobably
dnesandinthenorthernandmiddletownareas occupiedlargepartofthebuildingland,butsecduring horizon 4. This interpretation is based ularsettlementwassparseandonlydocumented
on three basic sources and two supplementary attwosites,oneinthenorthernandoneinthe
sources;ifthesupplementarysourcesareerrone- middletownarea.Thisalsosupportstheimpresouslyassignedtohorizon4thesourcesstillwit- sionoflowpressureonbuildingspaceforsecular
nesslargeinvestmentsinmonumentalmanifes- settlementinhorizon4.Largeresourceswereput
tationsintheBergenareainthishorizon.
intoconstructingthemonuments.
Excluding the monumental sites, data from
35analyticunitsinthetownareaswereusedas
sources.Secularoccupationwasindicatedat20 Horizon5(1120s-c1170)
plots/unitsandinthenorthernandmiddletown The Christchurch Cathedral (B) (site 2),
areasonly.Thesettlementtracesderivedfrom9 ChristchurchMinor(B)(site3),theChurchof
basicsources,10supplementarysources,andone the Apostles (B) (site 4) and Øystein Magnusgeneral background source. The basic sources son’stimberHall(B)(site5)stillconstitutethe
nowmakeupamuchmorecomfortableshareof monumental buildings at Holmen (Figure 39).
thesources,andthetrendsinthematerialought Thenumberofsourcesthatproducedtracesof
to be quite reliable in terms of the presence of occupationinthetownareaincreasesconsideroccupationinthenorthernandmiddletownar- ablyfromhorizon4totheperiodcoveredbyhoeas.
rizon5.StartingbytheeasternshoreofVeisan,
Notracesofsettlementwerefoundinfifteen plots 11/B and 11/C were now probably occuunits(S).ThevacantunitsalongthefootofFløy- pied by buildings (S). Plot 11/A may still have
fjelletandinthesoutherntownareamaystillbe beenvacantasnotracesofoccupation(S)were
evaluatedasintheprevioushorizons:theextent documented. Culture-layers, three buildings, a
of vacant areas at the foot of Fløyfjellet in the pavementandfourposts(B)werefoundatsite8,
northernandmiddletownareas,indicatedbythe thestructuresandculture-layersshowthatanasources, is quite reliable, whereas data from the lytic units 8/A-D were occupied. Plots 9-10/B
southern town area are scarce, and the impres- and9-10/C,orientedtowardsVågen,wereprobsionofvacantbuildinglandhereisnotsowell- ablyoccupied(S)plot12/A(S)wasprobablyocfounded. The vacant plots by Veisan are docu- cupiedtoo.
mentedthroughsupplementarysourcesfromone
The area just surveyed most likely corresiteonly.Thelackofsettlementisthusnotwell- spondstotheareadescribedas‘insideSandbru’
152
(Sandy-Bridge) in the sagas. According to the
sagas,severaltenementswerelocatedinthisarea
whenafiercefighttookplaceherein1155(Hkr
1893-1901, 386ff; Fsk 353ff; Msk 455ff; Helle
1982,6withreferences).Thiscorrespondswell
withthepicturegiventhroughthearchaeologicalsources.
Further to the east, along Vågen, the two
large stone-filled caissons that were tentatively
assignedtohorizon4werestillinfunctionasa
pier(B)(14/A).Thepierwascloselylocatedto
site6wheretracesofoccupation(B)havebeen
documentedonplots6/A-Eand6/G.Thearea
associatedwithplot6/Fhasnotbeenexcavated
indetailnorwell-documented,exceptalongthe
waterfrontwhere2mx2mstone-filledcaissons
wererecorded.Postsabovethewaterfront,however,showthattheplotwasoccupiedinhorizon
5.Asfarasplot6/Hgoes,thispartofthesiteis
notwell-documented,butmooringpostsbeyond
thewaterfrontimplythattheplotwasoccupied.
Theconstructionsatsite6comprise27buildings
and332mx2mstone-filledcaissons.Mooring
postswerelocatedalongthewaterfront,several
pits also belong to the horizon. The plots were
nowclearlyoccupiedbybuildingsorganisedin
tenements that ran perpendicular to the Vågen
shorelineandoccupiedthetidalzone.Thisimplies pressure on building land here. Plot 6/G
hadclaimedsomelandfromtheseainthesense
thattheoutermostcaissonswerelocated6.5m
fromtheoriginalshoreline,anexpansionofthe
built-up area into the sea had thus taken place
andindicateshighpressureonbuildingland.On
site15activitybeganwiththedepositionoflayers
(S)thatmadeupthefoundationforastructure
interpreted as a passage (S) and an undefined
structure(S).Plot15-16/Awasthusoccupiedin
horizon5.Onsite17thelackoftracesofoccupationindicatesthatthisareawasstillvacant.
South of the protruding rock by the Vågen
shoreline,site28waslocatedinthebaysome1520mfromtheoriginalshoreline.Apierfounded
onthree2mx2mstone-filledcaissons(B),a
rowofcaissons(B)thatranalongthewaterfront,
and8mooringposts(B)showthatplots28/A-C
werenowoccupied.Thelocationofthecaissons
clearly shows that an expansion of the built-up
areaintotheseahadstartedhere.Thismayin10TowhatextentwastheBergenarea‘occupied’?
dicatehighpressureonbuildingland.Atsite26
partialdemolitionofthestructuresfromhorizon
4 was followed by massive layer deposition (B)
that elevated the ground surface for building.
Constructions (B) interpreted as the foundation for a walkway behind the quay front were
then built. On site 27 to the southwest of site
26, three 2 m x 2 m stone-filled caissons (B),
ranparalleltotheshoresome5-15mfromthe
originalshoreline.Theoriginalshallowbayhere
thusseemstohavebeenfilledinandthebuilt-up
areaextendedintothebay.Southwestofsite27a
similarcaisson(S)waslocatedatsite29atsome
distance from the original shoreline. This caissonindicatesthatplot29/Aor29/Bbehindthe
structurewasprobablyoccupied.Thelocationof
thecaissonsupportsthepictureofanexpansion
ofthebuilt-upareaintoVågenattheexpenseof
thesmallbay.Thisexpansionisrecordedatsite
30 as well; the pier that dated back into horizon1wasnowonreclaimedland,incorporated
inastructure(B)perhapsintendedtostopthe
spreadingofdumpedlayersortokeepthestream
that ran down the sloping terrain, in its channel.Altogethertheexpansionofthebuilt-uparea
intothebayindicateshighpressureonbuilding
land.
AlongthefootofFløyfjelletasandextraction
pit(B)atsite22showsthattheareawasinuse
forsomekindofoccupation.Totheeastofhere,
at site 21, the stone foundations for a building
(S)andanumberofassociatedlayers(S)show
thattheNEpartofthesite(21/A)wasprobably
occupied.IntheSWpartofthesite(21/B)no
culture-layers or structures could be associated
with horizon 5, indicating that this area was
vacant (S). At site 20 a building with a courtyard,aninternalfireplaceandadrain(S)indicate that this site was occupied. The structures
andculture-layersfromsites20-22weretheonly
traces of secular occupation encountered along
Fløyfjellet in the northern town area. Between
site20andsite30notracesofsecularoccupation (S) could be assigned to horizon 5 at sites
16,18,19and31,indicatingthatthisareawas
vacant.Atsite30abridgeoverthestreamthat
ran down the sloping terrain was documented
15-20mfromtheoriginalshoreline(S)(30/B).
The bridge may have secured passage between
153
themiddleandthesoutherntownareas.Further
towards Fløyfjellet (30/D), the sand extraction
pitsfromhorizon4probablystillrepresentedthe
landuse(S)andthebasin(S)usedinconnection
with freshwater supply was still located here as
well(30/E).Theculture-layersandstructuresat
site30aretheonlytracesofsecularoccupation
encounteredatthefootofFløyfjelletinthemiddletownarea.
Churches and churchyards occupied large
partsofthenorthernandmiddletownareasin
horizon5.InthenortherntownareatheChurch
ofStOlav’sontheHill(B)(site25)mayhave
been located (S) north of site 21. The central
northerntownareawasoccupiedbytheChurch
of St Mary (B) (site 23) and its surrounding
churchyard(B,S).StMary’swasquitewelloriented according to an east-west axis. Southeast
ofStMary’s,thechurch(S)andchurchyard(S)
ofStPeter(site24)mayhavebeenlocated.The
‘westend’ofStPeter’sfacedVågen.Inthemiddle
townareatheChurchofStNicholas(B)(site32)
surrounded by the walls of the churchyard (B)
was still present in the landscape. The Church
of St Columba (S) (site 33) and the associated
churchyard(S)perhapsoccupiedashareofthe
buildingareabelowStNicholas’.
Inthesoutherntownareanotracesofoccupation(S)havebeendocumentedatsite34located
at the foot Fløyfjellet. At site 35, which is the
onlyrelevantsitebetweensite30andtheChurch
ofStCross(site40),notracesofoccupationhave
beenlocated,thismayindicatethatthesmallbay
bysite30wasonlybeingfilledinfromthemiddle town area and not from the southern town
area.TheChurchofStCross(B)waslocatedon
the promontory that extended into the Vågen
Bay.AlsotheChurchofStOlav’sinVågsbunnen(B)(site39)wasnowerected.Onlytworelevantexcavationshavebeencarriedoutbetween
the two churches. At site 38, on the beach at
the head of the Vågen bay, a small square logbuiltcaisson(B)andseveralposts(B)represent
thefirsttracesofoccupationhere.Thematerial
fromsite38cannotelucidatewhetherthegeneral
area was occupied, but as the two churches in
the southern town area were well oriented east
west according to the Christian standard this
may perhaps be taken as an indication of am154
plebuildingspacewhenthechurcheswerebuilt.
The material from site 36 (S) may tentatively
supportthenotionthatthesoutherntownarea
wasnotbuiltoutinhorizon5,exceptintheinnermostareaattheheadoftheVågenBaybysite
38.Theevidencefromthesoutherntownareais,
however,ratherscantyduetothescarcenumber
ofexcavatedsites.
IntheNonneseterarea,theNonneseterconvent(site46)(B)wasnowerectedonthenorthern shore of Alrekstadvågen. The churchyard
waslocatedonthenorthsideofthechurch,the
remainingpartoftheconventwaslocatedtothe
south.
At Nordnes the abbey of Munkeliv (B) was
still standing and further two institutions were
foundedonthepeninsula.About300mwestof
Munkeliv,StJohn’sAbbey(site44)(B)waslocated. Jonsbryggen, a pier/quay on the western
side of Vågen, also belonged to the abbey, but
theexactlocationofthequayisunknown.The
Church of All Saints (S) was probably located
furtherwestonthelandtonguebetweenVågen
BayandtheAlrekstadvågen.
Summary
Summingup,asmanyasfivenewchurches,in
addition to the (second?) Church of St Mary,
mayhavebeenfoundedinthetownareainhorizon5,threeofthesearedocumentedthrough
basicsourcestwothroughsupplementary.Inthe
NonneseterandNordnesareastwomonasteries
and possibly also one church were established.
Themonestarieswerewell-documentedthrough
basicsourceswhereasthechurchisasupplementary source. All in all extensive investments in
monumentsisconsideredwell-founded.
Datafrom45plotsorunitshavebeenusedas
sourcesforsecularoccupationinthetownareain
horizon5andsecularoccupationwasnowdocumentedinallthreetownareas.Tracesofsettlementweredocumentedin34analyticunits,in
22casesbybasicsourcesandin12unitsbysupplementarysources.Withthelargeshareofbasic
sources,thewidepresenceofoccupiedplotsand
unitsshouldbeconsideredwell-documented.
Vacant plots and units were indicated in 10
units.TheunitslocatedatthefootoftheFløyfjellet are not interrelated and as they point in
Figure39.Horizon5(1120s-c1170)occupiedandvacantanalyticunits
10TowhatextentwastheBergenarea‘occupied’?
155
thesamedirectionthelackofsecularsettlement
inthisareaoughttobequitereliable.Thelackof
occupationtracesatsite17closetoVågenmay
be real, as the earliest culture-layers at the site
werelocatedinacleftinthebedrockanditisunlikelythatolderculture-layershadsystematically
beencleanedoutofthecleft(cfp119).TheexistenceofvacantplotsbyVeisanisdocumented
throughalonesupplementarysource,sothelack
of settlement here should be taken with some
reservationsandthequestionremainsopen.The
vacantunitsinthesoutherntown,areaaspointedoutseveraltimesalready,arenotinterrelated
andseemtopointinthesamedirection,implyingthatthesoutherntownareawasnotdensely
builtupon.Thisimpressionisnowsupportedby
theorientationofthetwochurchesinthistown
area.Nevertheless,thesourcesforthesouthern
townareaarefewandfarbetween,sofirmconclusionscannotbemade.
Accordingtothetrendsinthematerial,most
of the plots along the Vågen shoreline seem to
havebeenoccupiedinhorizon5.AlongVågen
therewasonlyalackofsettlementtracesatsite
17. Also along the Vågen shoreline, buildings
were constructed throughout the whole length
oftheplots,makinguseofthetidalzone.This
indicates that there was pressure on building
land. On some plots the built-up area had also
expanded into the Vågen basin claiming new
building space, and seeking deeper water. This
indicates high pressure on building land. The
secularsettlementhadincreasedalongFløyfjelletinthenortherntownareabuttherewasstill
vacantbuildingland.
In conclusion the sources indicate pressure
andinsomeplacesevenhighpressureonsecular
buildingspacealongthewaterfrontsinthenorthernandmiddletownareas.AtthefootofFløyfjellettherewasstillamplebuildingspace,though
churchesnowoccupyagoodshareoftheland.In
thesoutherntownareasettlementwasdocumentedatonesite.Theorientationofthechurchesin
thispartoftownmayindicatethattheareawas
not densely built upon when the churches were
establishedinhorizon5.Thismayindicatethat
there was low pressure on building land here,
thoughnofirmconclusionscanbereached.
156
Conclusions
Accordingtothetrendsdiscernedinthescarce
sources assigned to horizon 2 settlement was
mostly found in the areas closest to the Veisan
andVågenshorelinesinthenortherntownarea,
the sources altogether imply low pressure on
buildinglandinthishorizon.
In horizon 3, there was considerable activity
on the monumental side. The sources concerningsecularsettlementarescarce,butaccording
tothegeneraltrendsinthematerial,mostofthe
occupiedplotsorunitsseemtohavebeenlocated
bytheVågenshoreline.AlongVeisansomeplots
mayhavebeenvacant,andatthefootofFløyfjelletsettlementwaslimitedandonlydocumented
atonesiteinthenortherntownareaandatsite
30inthemiddletownarea.Thereseemstohave
been low pressure on building land on plots
whereoccupationwasdocumented.
Altogether,thesourcesforhorizon4aremore
firmlyfoundedthanintheearlierhorizons,and
trends in the material should be reliable. Large
resourcesweredirectedatthemonumentalconstructions.Inthetownareasitseemsthatsome
Veisan plots were now occupied. Along Vågen,
occupation was indicated on most documented
plots,bothinthenorthernandthemiddletown
areas.Thelocationofdocumentedbuildingsindicates low pressure on secular building space
inthenortherntownarea,whereaspressureon
buildinglandwasregistredinthemiddletown
area.AlongthefootofFløyfjellet,thechurchof
StNicholasprobablytookupalargepartofthe
buildingland,butsecularsettlementwassparse
and only documented at two sites, one in the
northern and one in the middle town area. In
the southern town area no secular occupation
wasfound,butthesourcesforthetownareaare
scarcesotheimpressionofvacantbuildingland
isnotwell-founded.
Inhorizon5basicsourcesmakeupthemajorityofsources,andtrendsinthematerialshould
be reliable. In the whole Bergen area large resourcesweredirectedatmonumentalconstructions. According to the trends in the material,
mostoftheplotsalongtheVågenshorelineseem
tohavebeenoccupiedinhorizon5.AlongVågen
thelackofsettlementtraceswasdocumentedat
onesite.AlongtheVågenshore,therewaspres-
sure on building land and on some plots even
highpressureonbuildingland.ThesecularsettlementincreasedsomewhatatthefootofFløyfjellet in the northern town area but there was
stillvacantbuildinglandherealthoughchurches
nowprobablyoccupiedagoodshareoftheland.
In the southern town area the few available
sources may indicate that pressure on building
land was low here, though no firm conclusions
canbemade.
Altogether the sources imply that it took a
longtimebeforethemajorityofthedocumented
plotsandunitsinthetownareaswereoccupied
and taken widely into use. The actual occupationofthetownareaswasthusalongandslow
process.
Placesofproduction
11CRAFTSANDPRODUCTION
INEARLYBERGEN
Thenatureandorganisationofcraftsandproduction
InthischapterIwillstudythenatureandorganisationoftheproductiveactivitiesidentifiedinearlyBergen.Iwilldiscusswhetherproductiveactivitiesdiscernedinthematerialmayhaveprovideda
fundamentaleconomicbasisfortheemergenceof
atownintheBergenarea.First,placeswhereproductive activities were carried out are identified.
Secondthenatureandorganisationoftheidentifiedactivitiesarediscussed,andfinallytheimportanceoftheseactivitiesasafundamentaleconomicbasisfortheemergenceofthetownisdiscussed.
Iwilldiscussthepresenceofthefollowingcrafts
andproductiveactivitiesinearlyBergen:
• Combmaking
• Miscellaneousantler,bone,hornandwhale/
walrusboneworking
• Shoemaking
• Otherleatherworking
• Metalworking
• Stoneworking
• Woodworking
• Skinning
• Textileproduction
• Fishing
• Hunting,warandgame
• Agriculture
• Basiccooking
• Foodandbeverageprocessing
11CraftsandproductioninearlyBergen
Crafts and production have left their traces
throughtools,blanks,rawmaterials,production
wasteandthroughthefinishedproducts.When
identifyingwhereproductiveactivitieswerecarriedoutthefinishedproductsarenotconsidered.
A number of multifunctional tools, including
honesandpossibletoolsmadeofmetalarealso
excludedfromthestudybecausetheycannotbe
usedassignifiersofaspecificactivity.Placesof
productionareidentifiedhorizonbyhorizon.
The sources are to a large extent presented
in tabular form, since a comprehensive textual
presentation of all artefact types for each unit,
each town area etc would become too detailed
anddifficultreading.Thetablesshouldthusbeconsultedcloselyalongwiththetext.
WhenstudyingthenatureoftheproductiveactivitiesIwilldiscusswhethertheactivitieswere
aimedathouseholdconsumption,orforsaleand
thus carried out by professionals. Sale in this
context is used in a broad sense, meaning distributionbeyondthehouseholdoftheproducer
(cfp40).IuseKarinGjølHagen’sdefinitionof
professionalism, where professional production
isproductionforsale,asopposedtoproduction
for consumption within the household (Hagen
(1988) 1994, 29-31). When studying how the
productiveactivitieswereorganisedIwilldiscuss
whether the artisans/producers were sedentary
residentsoftheplotorsettlementwhereproductiontookplace,oriftheytravelledbetweenseveralplacesofproduction.
In earlier research the nature and organisationofproductiveactivitiesintheearlyMiddle
Ageshavebeenstudiedthroughthequantityand
distribution of production waste, the character
oftoolsneededfortheproductionandthequality and level of standardisation of the finished
products.Writtensourcesforvarioustradeshave
been drawn into the discussion (eg Christophersen1980,127ff;Roesdal1980,105ff;Ambrosiani 1981, 32ff; Christophersen 1982, 109;
Øye1988,131;Bergquist1989,22;Flodin1989,
128; Færden 1990, 277; Christophersen and
Nordeide1994,216;Ulriksen1996,119;Rytter
1997;Hagen(1988)1994).
157
I will discuss the nature and organisation of
theproductiveactivitiesthrough(1)thelinkbetween production waste and the finished products,(2)thespatialdistributionofthefinished
productsinandbeyondBergen,(3)thecharacteroftools,skillsandknowledgeneededforthe
production,(4)thelevelofstandardisationofthe
finished products and (5) the distribution patternforproductionwaste.
Trying to link production waste to finished
products, – that is rendering probable that finishedproductsfoundinBergenarereflectedin
theproductionwasteandtoolsandviceversa-it
may be shown that the finished products were
orcouldhavebeenproducedandpurchasedin
Bergen.
Thespatialdistributionofthefinishedproducts inBergen may shed light uponthe nature
oftheproduction.Whentwoormoreitemsare
so similar that it seems probable that the same
artisan or workshop made them, the items are
denominated ‘twin’ products. If twin products
are found in different analytic units I find it
likelythattheyhavebeendistributedbeyondthe
producershouseholdandthus,accordingtothe
definitionusedhere,wereproducedforsale.The
finishedproductsareincludedinthediscussion
sofaraspossiblewithinanacceptableeffort.
Whenspecialisedtools,ahighlevelofskills,
and knowledge of how an up-to-date product
wasformedarerequiredfortheproductionthis
alsoimpliesthattheproducerswereprofessionals. As opposed to this, production involving
household tools and ‘common knowledge’ only
may reflect household production. If the productswerestandardised,itmaybeanadditional
indication that they were produced professionally(cfHagen(1988)1994,99).
The quantity and distribution of production
waste reflect different ways of organising production.Wasteassemblagesaredividedintotwo
widecategories:‘small’or‘large’.Thecategories
aredefinedforeachactivitytypewithreference
tostudiesofsimilaractivitiesinothermedieval
towns in Scandinavia. Thus an assemblage is
characterised as small or large respectively if it
fallswithinthequantity-categoriesthatareconsidered as small or large for equivalent artefact
categories in studies of other medieval Scandi158
navian towns. The different activities produce
waste in varying quantity, volume and weight.
Whencharacterisingwasteassemblagesassmall
orlargeIhavechoosentoletthenumberoffragmentsformthebasis.
In Lund and Trondheim small amounts of
productionwastelocatedinmanydifferentfind
spots were interpreted as traces of professional
ambulating combmakers and metalworkers,
whereaslargeamountsofwastewereinterpreted
as traces of stationary artisans (Christophersen
1980, 127; Bergquist 1989, 128; Flodin 1989,
128).Basedonsmallamountsofwastedistributed on many places of production Jens Rytter
hassuggestedthattheartisansofKonghellemay
either have ambulated, or they may have been
resident‘part-time’artisans,withtheproduction
of combs as one of several trades (Rytter 1991,
92).
Iftheproductionwascarriedoutprofessionally,Iassumethatmanysmallwasteassemblages
imply that the producer was either a part-time
professional resident artisan or an ambulating
artisan. Household production may also be reflected in small amounts of waste scattered on
many plots. Large waste concentrations are interpreted as the result of long-term production
byfull-timeresidentprofessionals.Apermanent
workshopmayalsoreflectresidentprofessionals.
Since the quantity of production waste reflects
the size of the production, I assume that small
amounts of waste reflect a limited production
aimed at an interurban market, whereas large
amountsofwastemayimplythattheproduction
wasaimedatawidermarket.Stoneandtimber
buildings signify large-scale stone and woodworking,whenquantifyingtheseactivitytraces
Imerelyconsidertheirpresence.
Based on similarities between Viking Age
combsfromRussiatotheeastandIrelandtothe
west, Kristina Ambrosiani has suggested that
professional combmakers travelled and worked
withinlimitedregionsthatoverlapped(Ambrosiani 1981, 32ff). If Bergen items have ‘twins’
outsideBergenanditisprobablethattheitems
were produced outside Bergen this may imply
that the producers were professional ambulatingartisans.Ishallthereforestudy‘twin’productsfromlocalitiesoutsideBergen,thisisdone
throughpublishedorotherwiseaccessibleillustrationsofartefacts.
As a point of departure I presume that four
types of producers theoretically may have been
presentinearlyBergen:(1)householdproducers,
(2) professional sedentary full-time producers,
(3) professional sedentary part-time producers,
and(4)professionalambulatingproducers.The
productiveactivitiesarelinkedtotheseproducer
types(Table30).
ing and miscellaneous antler, bone, horn and
whale/walrus bone working. These find groups
fromBergenhavenotpreviouslybeenpublished
orstudiedindetail.Asmentionedearlier,Inger
Kellmerstudiedthecombsfromsite6whereshe
alsotoucheduponthewastematerialfromcomb
production.51Ihavereclassifieditemsthatwere
originallyclassifiedastheremainsofcombmakingandthatwereretrievedinthearchives,that
is 85 % of this find group, supplementing my
observationswithKellmer’snoteswhenpossible
X
X
X
X
Professionalambulatingproducers
X
X
X
X
X
X
Householdproducers
Professionalsedentaryfull-timeproducers
Twinproductsoutside
Bergen
X
X
Permanentworkshop
X
Professionalsedentarypart-timeproducers
Afewlargewaste
assemblages
TwinproductsinBergen
X
Specialisedskillsand
knowledge
Householdtools/common
knowledge
X
Specialisedtools
Manysmallwaste
assemblages
Table30.Producertypesandhowtheymaybereflectedinthesources
X
fortheremaining15%.Combblanksconsistof
discardedtoothsegmentsandconnectionplates.
Theseobjectsarequitecharacteristicandeasyto
identify(Figure40).Thegroupofcomboffcuts
containsoffcutsthatcouldbeassociateddirectly
withcomb-productionasopposedtooffcutsthat
couldnotbeidentifiedintermsofendproduct.
Twotoolscouldbeassociatedwithcombmaking
(Figure41)bothweremadefromtinesofantler.
Thefirsttoolwasabout7cmlong,witha2cm
x2cmquadraticcross-section,fourholesofdifferentdiameterspenetratedtheitem.Iinterpret
itasawiredrawerusedwhenformingrivetsout
ofthinrolledsheetsofmetal.52Thesecondtool
derives from the same find spot, and based on
sizeandshapeitwasapparentlypartofthesame
toolkit, the tool is interpreted as a punch (cf
Placesofproduction
Ambrosiani1981,Figure62).Someofthecomb
productionwasteandblankshavebeenclassified
Combmakingandmiscellaneousantler,
according to species by osteologist Anne Karin
bone,hornandwhale/walrusboneworking
Hufthammer, as probable reindeer antler (cf
Anumberofartefactgroupshavebeenstudiedin footnote63).
ordertoidentifyproductionareasforcombmakIn addition to combs, other items of miscelProductionasafundamentaleconomicbasis
WhenaddressingtheimportanceoftheproductiveactivitiesasaneconomicbasisfortheemergenceofthetownIwilldiscusswhetheranyof
the productive activities identified were fundamentalfortheriseofthetown.Productiveactivitiesthatwerecarriedoutonahouseholdbasisand
thoseservingan‘interurbanmarket’didnotadd
‘value’tothetowncommunityandmaytherefore not have been of fundamental importance
totheinitialriseofthetown(cfChristophersen
1982,108).Asopposedtothisproductiveactivitiesthatservedawidermarketmayhaveplayed
animportantpartasanindependenteconomic
factorintheearlytown.
11CraftsandproductioninearlyBergen
159
laneous antler, bone, horn and whale/walrus
bonewereproduced.Thesourcesforthisactivity
comprisevariousblanksandoffcuts.Iwasonly
abletolocateabouthalfofthematerialoriginally
cataloguedasoffcutsinthemuseumstorerooms,
fortheremaininghalfIhavehadtorelyonthe
original finds catalogues for a classification of
the finds. It appears that a distinction between
hornandantlerhasnotbeenmadeinthefindscatalogue for site 6. In the material that could
bechecked,artefactscataloguedashornactually
turnedouttobeantler.Accordinglythesite6materialthatwascataloguedashorn,andcouldnot
berecheckedhasbeenreclassifiedashorn/antler.
Whentheartefacthasbeendescribedasgoat-or
ox-horn,Ipresumethattheartefactwasactually
horn.Therawmaterialoftheblanks(otherthan
comb blanks) has been specified as far as possible,thisclassificationhasbeencarriedoutonly
visually by me. The category ‘bone’comprises
miscellanous bone, the category ‘whale/walrus’
comprisesbonefromtheanimals,notivory.
Therewerenotracesofcombmakingpriorto
horizon5andnotracesofmiscellaneousantler,
bone,hornandwhale/walrusboneworkingprior
tohorizon4.
(cfTable31).TheplacesofproductionwerelocatedonplotsalongtheVågenshorelineinthe
northernandthemiddletownareas,theywere
identifiedthroughbasicaswellassupplementary
sources.Thetendencyinthematerialthatseveralplacesofproductionexistedduringhorizon
4,isthusconsideredtrustworthy.
Horizon5(1120s-c1170)
AlongtheVågenshorelineinthenortherntown
area,antler,whale/walrusboneandperhapsalso
hornwasworkedonplot6/B,gamingpiecesof
whale/walrusbonewereoneproductaccording
to a blank found there (cf Table 32). On plot
6/C,antler,bone,aswellaswhale/walrusbone
was worked, combs, needles/pins and line runners were products. On plot 6/D, antler, bone,
hornaswellaswhale/walrusbonewasworked.
Combs, gaming pieces and needles/pins were
someoftheproducts.Onplot6/Eantler,bone
andperhapsalsohornwereworked,needles/pins
ofantlerwereoneoftheproducts.Onplot6/G,
antler and bone were worked, combs were one
of the products. By the Vågen shoreline in the
middletownarea,antlerandbonewereworked
onplot26/A.Onplot26-27/B,antlerandperhapsalsohorn(26-27/BC)wereworkedandon
Horizon4(c1100-1120s)
plot27/Chornandperhapsalsoantler(26-27/
Tracesofmiscellaneousantlerworkinghavebeen BC)wasworked.Onplot28/C,antlerwasalso
identifiedatfourofthesevenfind-yieldingana- workedandcombsproduced.Antlerwasworked
lyticunits.Onplot27/C,needle/pinsofantler bythesmallriverthatrandowntheslopingterwereproduced,accordingtoablankfoundthere rain (unit 30/B). In horizon 5 the activity of
6/B
26/A
26-27/B•
26-27/BC•
27/C
(1)#
(2)
3
4
(6)
(9)
ArtefactcategoryIIinbrackets
#Antler
Numbersinboldrefertobasicsources,inplainsupplementarysources
•Artefactsfrombothbasicandsupplementarysources
160
Whale/walrusbone
offcut
Hornoffcut
Horn/antleroffcut
Boneoffcut
Antleroffcut
Trialpiece
Miscellaneousblanks
Linerunnerblank
Needle/pinblank
Wiredrawer
Punch
Comboffcut
Combblank
Plot/unit
Gamingpieceblank
Table31.Horizon4(c1100-1120s),miscellaneousantler,bone,hornandwhale/walrusboneworking(N=25)
Figure40.Toothsegmentsandconnectionplates
Figure41.Punchandwire
drawer:aBRM0/86590/02;b
BRM0/86590/01
11CraftsandproductioninearlyBergen
161
combmakingwasthusidentifiedonfourofthe
24 find-yielding analytic units and miscellaneous antler, bone, horn and whale/walrus bone
working was identified on 12 of the 24 findyieldingunits.Theproductionplaceswerelocatedinboththenorthernandmiddletownareas.
Theproductionplacesforcombswereidentified
throughbasicsourcesonly,theproductionplaces
formiscellaneousantler,bone,hornandwhale/
walrusboneworkingwereidentifiedthrough10
basicandtwosupplementarysources.The tendencyinthematerialthatproductionplacesfor
combmakingandformiscellaneousantler,bone,
horn and whale/walrus bone working existed
duringhorizon3,isthustrustworthy.
Shoemakingandotherleatherworking
The sources reflect that both shoemaking and
‘otherleatherwork’werecarriedoutinearlyBergen.Shoemakersareheredefinedasindividuals
orgroupsofspecialistswhoproduceditemsthat
weremadefrom‘new’leather.These‘shoemakers’mayhaveproducedotherleatheritemsthan
shoes,butthesourceshavenotbeenstudiedin
suchdetailthatconcreteproductscanbelinked
tothewaste,therefore,forthesakeofsimplicity
theyaretermedshoemakers.Theactivityof‘otherleatherwork’coverstheproductionofarticles
thatwerefabricatedfromreusedleather.
The Bergen town regulations of 1282 imply
thatshoemakerswerealsotannersandthattan-
6/B
(4)
1
6/D
(1)
#
(1)
6/E
12
(32)
1
(30)
(1)
(1)
1
(1)
#
1
+
(1)
#
(1)
#
(2)
+
(2)
+
(1)
+
1
(10)
16
(24)
2
(5)
1
1
(4)
(4)
(3)
6
(5)
(2)
(2)
(6)
(5)
15-16/A
26/A
26-27/B
1
#
(2)
1
(5)
1
(9)
(2)
26-27/BC
(9)
27/C
28/C
(7)
(1)
30/B
ArtefactcategoryIIinbrackets
#Antler
*Whale/walrusbone
+Bone
Numbersinboldrefertobasicsources,inplainsupplementarysources
162
(2)
(6)
28/B
(9)
(1)
Whale/walrusbone
offcut
Hornoffcut
Horn/antleroffcut
Boneoffcut
Antleroffcut
Trialpiece
Miscellaneousblanks
Linerunnerblank
(1)
*
6/C
6/G
Needle/pinblank
Wiredrawer
Punch
Comboffcut
Combblank
Plot/unit
Gamingpieceblank
Table32.Horizon5(1120s-c1170),combmakingandmiscellaneousantler,bone,hornandwhale/walrusboneworking
(N=254)
(4)
11CraftsandproductioninearlyBergen
Horizon3(c1070-c1100)
Traces of shoemaking and other leatherwork,
tentatively dated to horizon 3 were found in
thenortherntownareainoneofthetwofindyieldinganalyticunits(plot9-10/B)(Table33).
There is no indication in the documented culture-layersthattanningwascarriedoutonany
ofthesites.Sincethetracesofshoemakingand
otherleatherworkingstemfromonesupplementarysourceonly,thematerialisconsideredastoo
inflicted with uncertainty for the identification
ofshoemakingandotherleatherworkingduring
horizon 3 and it is not included in the further
discussions.
9-10/B
(2)
Wastetype2
(otherleatherworking)
Plot/unit
Last
Table33.Horizon3(c1070-c1100),shoemakingand‘other
leatherworking’(N=3)
Wastetype3
(shoemaking)
ningandshoemakingwereco-located(NgLIII
14). The physical remains of tanning may be
chalk,hair,bark(Schia1975,24),birdoranimal
manureandlargevatsusedforsoakingthehide
(Larsen1992,86ff).The1282regulationsrelate
toshoemakinginthehighmedievaltownofBergen more than 100 years later than the period
studiedhere.Theorganisationoftheshoemaker’scraftmayhavedifferedfromtheearlyurban
craft.ShoemakersintwelfthcenturyBergenmay
havetannedtheirownleather,butshoemaking
could have taken place on other locations than
thatoftanning.Whentryingtoidentifyproductionareasforshoemaking,leather-wasteand/or
shoemaker’stoolshavebeenconsideredsufficient
evidencetoidentifyshoemaking.Thecompositionofculture-layersandtracesofchalkproductionare,however,alsoregarded.
I have divided the leather waste into three
types using the methods developed by Larsen
(Larsen1970,34;Larsen1992,35)(Figure42).
Type 1 comprises pieces of leather where the
edges are torn. The leather in this waste group
has no traces of having been reused and representsrubbish.Type2comprisespiecesofleather
withstitchingalongsomeedgesandstraightcuts
alongothers.Thiswastegrouprepresentstheremains of leather items that were reused as raw
materialfornewitems.Type3comprisespieces
ofleatherthathadnotracesofstitchingandthat
werecutfromalargerpieceofleather.Thistype
compriseswastefromtheregularfabricationof
itemsfromleatherthathadnotbeenusedpreviously.Larsenhasstudiedshoesandleatherwaste
fromtheGullskoenareaatsite6(Larsen1992),
butdidnotdistinguishbetweentypes1and2in
thisstudy.Ihavethereforereclassifiedtheleather
materialfromtheGullskoenareatogetherwith
the remaining leather material from early Bergenbeforec1170.Areaswhereshoemakingtook
place are identified through the distribution of
wasteoftype3andtoolsofthecraft.Alast53is
theonlytoolthathasbeensafelyidentifiedasa
shoemaker’stool(Figure42).Areaswhere‘other leatherworking’ was carried out are studied
throughthedistributionofwasteofgroup2.
No traces of shoemaking or ‘other leatherworking’werefoundbeforehorizon3.
(1)
ArtefactcategoryIIinbrackets.Numbersinbold
refertobasicsources,inplainsupplementarysources
Horizon4(c1100-1120s)
In the middle town area traces of shoemaking
were found in three of the seven find-bearing
analyticunits(Table34).Thereisnoindication
in the documented culture-layers that tanning
wascarriedoutonanyoftheplots/sites.Theexistenceofplaceswhereshoesweremadeduring
horizon 4 is documented through basic as well
assupplementarysourcesandisthusconsidered
trustworthy.
163
26/A
26-27/B•
26-27/BC•
27/C
Wastetype2
(otherleatherworking)
Last
Plot/unit
Wastetype3
(shoemaking)
Table34.Horizon4(c1100-1120s),shoemakingand‘other
leatherworking’(N=47)
2
(3)
(19)
(20)
ArtefactcategoryIIinbrackets.Numbersinbold
refertobasicsources,inplainsupplementarysources
•Artefactsfrombothbasicandsupplementarysources
Horizon5(1120s-c1170)
Waste of type 3 signifies shoemaking on 13 of
the24find-yieldingunits(Table35).Onplot6/
Calast54wasfoundinadditiontothewaste.No
layers with tanning related contents have been
recorded on any of the plots/sites. Three limeslaking pits were, however, located in an open
areaofplot6/C,and3lumpsofchalkwererecorded on the same plot, a lump of chalk was
recordedonplot6/G.Itisnotlikelythatthese
remains are related to tanning, because both
plots 6/C and 6/G were excavated in their full
lengths and no vats, necessary for soaking the
hides,weredocumented.Furthermore,therewas
Figure42.Leatherwasteandashoemaker’slast(BRM0/54784/01).(FromLarsen1991,34-35).(DrawingsbySveinSkauge)
164
notenoughrunningwaterontheplotsfortanneries to be operated, and no layers with bark
or other tanning-related ingredients have been
recorded.Therearethusnoindicationsthattanningtookplaceonplot6/Cor6/Gduringhorizon5.Wasteoftype2wasfoundinnineofthe
find-yieldinganalyticunitsandshowthat‘other
leatherwork’wascarriedouthere.
Theproductionplacesforshoesaswellasfor
‘other leatherworking’ are found on the plots
alongtheVeisanandVågenshorelinesandatthe
footofFløyfjellet.Shoemakingwasdocumented
through11basicsourcesandtwosupplementary
sources,‘otherleatherwork’througheightbasic
andonesupplementarysource,thetendencyin
thematerialthatproductionofshoesand‘other
leatherwork’tookplaceduringhorizon5,isthus
consideredwell-founded.
6/B
(4)
6/C
6
(51)
15
(50)
(21)
6/D
6/E
Wastetype2
(otherleatherworking)
Last
Plot/unit
Wastetype3
(shoemaking)
Table35.Horizon5(1120s-c1170),shoemakingand‘other
leatherworking’(N=643)
(2)
(1)
6/F
(20)
29
(52)
2
(13)
(1)
1
(2)
8/B
268
(8)
(1)
20/A
(1)
(4)
26/A
1
(1)
(1)
6/G
26-27/B
(2)
26-27/BC
(33)
27/C
28/B
(41)
(2)
28/C
(2)
30/B
(2)
ArtefactcategoryIIinbrackets
11CraftsandproductioninearlyBergen
(5)
(1)
Metalworking
Metalworkinghasnotbeenstudiedpreviouslyin
theBergenmaterial.IntheUrbanCodeof1276
metalworkersaredividedintoseveraltrades(Bl
1923,byskipingenChapter8).Itisnotpossible,
however, to make such fine distinctions in the
presentarchaeologicalmaterial.Acommondistinctioninarchaeologicalstudiesofmetalworking is made between smithing and casting, the
formeridentifiedthroughslagfrommetalworking,thelatterthroughwasteandequipmentassociated with casting (Bergquist 1989; Færden
1990).Ihaveclassifiedthematerialaccordingto
theseprinciples.Theworkshopofametalworker
maybecharacterisedbysootandcharcoalinaddition to production waste (Bergman and Billberg1976;Færden1990,193).Suchlayersare,
however, a common ingredient in the makeup
oftheculture-layersinBergenasfiresdevastated
the town repeatedly through the Middle Ages.
Basedontheavailabledocumentationsootand
charcoal layers cannot be used as a source for
metalworkinginthepresentstudy.
The finds collection of slag is probably not
representativeforwhatwasactuallyfound(cfp
159ff).Onlysevenartefactswereoriginallyclassifiedasslagandtheslagwasfoundonalmost
asmanydifferentplots/sites.Asmentionedearlier, slag was not collected systematically during excavations carried out before 1980. In addition, slag may derive from non-metallurgical
processes(Bergquist1989,46).Onlyfourofthe
sevenpiecesof‘slag’havebeenretrievedinthe
archives,andanevaluationoftheoriginalclassificationoftheremainingfindshasnotbeenpossible. Hence, slag is not considered as evidence
ofsmithingunlessitwasidentifiedbyme.One
fragmentofaclayfurnaceliningisalsofoundin
thematerial,andmayrepresentsmithing.
Cruciblesandmouldswereusedwhencasting
(Bergquist1989,26)(Figure43).Crucibleshave
been catalogued together with pottery. Since
moulds is a finds category that has been given
attentionevenintheoldestdocumentationmaterial,Iassumethatthethreeartefactsclassifiedas
mouldswereactuallymoulds-eventhetwothat
couldnotberetrievedinthemuseumstorerooms.
Iffoundalongwithcruciblesormouldsoffcuts
fromcopperalloyandfinemetalsmayrepresent
165
e
c
a
b
d
Figure43.Crucibles,apossibletouchstone,andamould:a,c,ecruciblesBRM104/2280,BRM104/2311,BRM104/2326;
bBRM0/64456touchstone?;dBRM110/4949mould
therawmaterialsusedwhencasting.Thesefinds
can,however,notstandinisolationasevidenceof
metalworking.Likewise,concentrationsofnails,
rivets,bitsandpiecesofironfoundtogetherwith
slag,mayrepresentrawmaterialforsmithing(cf
Ulriksen1996,42).Ontheotherhand,thiscategoryoffindsmay,justaslikelyrepresentrubbish
oritmayevenbepartofstructuresorartefacts
ontheplot.Ifindittoodifficulttodistinguish
betweenironrawmaterialandrubbishandhave
choosentoomitthefindgroupfromtheanalysis,
exceptinthecaseofsite8buildingK158.Weights
and balances may also have been used in connectionwithmetalworking(Pedersen2001,24)
166
Thesetoolscan,however,alsobeassociatedwith
trade,andcannotinisolationbeanindicationof
eitheractivity.Oneitemhasbeenclassifiedasa
possibletouchstoneonthebasisofitsshapeand
thehardnessofthestone.55Atouchstone(Figure
43)wasusedwhentestingthevalourofgold(cf
Biddle1990,76,278p).Therewerenotracesof
goldonthestonefromBergen,andastheclassificationissomewhatuncertain,theitemcannot
inisolationbeasourceformetalworking.
‘Thesmiths’booths’arementionedinthesagasinconnectionwithafightthattookplacein
Bergenin1155.Thefightisdescribedinseveral
reliablesources(Helle1982,6),thatgiveade-
tailed topographical description of the location
ofthesmiths’booths.Accordingtothedescription, the booths should be close to a tenement
probablylocatednearorperhapsslightlytothe
eastofsite9/10.Goingdowntothistenement
fromthebooths,onewouldprobablycomefrom
theareaofsite8locatedonthetopofthemorainic tongue that characterised the landscape
just here. The saga passage is used as a supplementarysourceforhorizon5.Therearenoindicationsofmetalworkingpriortohorizon4.
Horizon4(c1100-1120s)
Inhorizon4(Table36),cruciblesshowthatcasting was carried out on two plots by the Vågen
shorelineinthemiddletownarea,offcutsoffine
metalonplot6/Binthenortherntownareamay
also be an indication of the activity of casting,
but cannot in isolation be used as evidence of
metalworking.Inunit30/E,locatedinthemiddle town area at the foot of Fløyfjellet, a fragment of a clay furnace lining reflects smithing.
In addition, copper alloy offcuts may perhaps
signify casting of copper alloy here. Altogether
metalworkinghasbeenidentifiedinthreeofthe
7find-yieldinganalyticunits.Twounitswerebasicsources,onewassupplementary.Thesources
arefew,but,duetotherepresentativityproblems
inherentinthematerial,theyprobablyrepresent
a minimum of places where metalworking was
carried out. The tendency in the material that
metalworking took place during horizon 4 is
thusconsideredreliable.
6/B
26/A
26-27/B•
30/E
(1)
1
(2)
(1)
(3)
ArtefactcategoryIIinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
•Artefactsfrombothbasicandsupplementary
sources
11CraftsandproductioninearlyBergen
Touchstone?
Weight/balancearm
Offcutfinemetal
Offcutcopperalloy
Slag
Mould
Crucible
Plot/
unit
Furnacelining
Table36.Horizon4(c1100-1120s),metalworking(N=8)
Horizon5(1120s-c1170)
In the northern town area, along the Vågen
shoreline,castingofcopperalloytookplaceon
plot6/Caccordingtocruciblesandotherfinds
fromthisplot(Table37).Apossibletouchstone
and a weight from the same plot may perhaps
also signify metalworking. Tools of trade were,
however,alsofoundonplot6/C(Table67),so
the weight cannot be conclusively associated
withmetalworking.Castingofcopperalloyand
otherfinemetalstookplaceonplot6/D,inadditionsmithingisindicatedthroughslag.Acrucible and a piece of slag show that casting and
smithingwerecarriedoutonplot6/G.
BytheVeisan,slaginunit8/Bindicatesthat
smithingtookplaceduringhorizon5.Thelocationofthesmith’sboothsdescribedinthewrittensources(cfabove)correspondswellwiththe
contextoftheslagfoundhere.Theslag,found
inbuilding158,wasfoundtogetherwithalarger
concentrationofironbitsandpieces.Thebuildingburnedsometimeafter1150(cfp98ff),slag
and iron bits were found in the fire-layer and
may reflect activities in the particular building
sometimeafter1150.Aswehaveseen,theevents
describedaroundthesmiths’boothstookplace
in 1155. Site 8 was excavated and documented
thoroughly,butthesitewasalsoverydisturbed
bylateractivities(cfp98ff)andonlyafraction
of building K158 and hardly any culture-layers
outside the building were intact. Any waste located outside the building may thus have been
removed by later activities. It is not impossible
that building 158 could have been one of the
smiths’booths.
Offcutofcopperalloywasfoundonplot20/
A,locatedatthefootofFløyfjelletinthenorthern town area, but cannot in isolation provide
evidenceofmetalworking.BytheVågenshorelineinthemiddletownarea,cruciblesindicate
castingonplot26/A.Abalancearmfitsintothe
pictureoffinemetalworking.Nofindsindicatingtradewerefoundinthiscontextinhorizon5
thisstrengthensthenotionthatthisbalancearm
should be associated with metalworking rather
thantrade.Onplot26-27/Bcruciblesshowthat
casting went on, the same applies to plot 27/C
where a mould was found together with crucibles.Offcutofcopperalloyfoundbetween26167
27/Band27/Cfitswellintothepictureofmetalworkingoneitherplot.
Alongthesmallriverbysite30,offcutoffine
metal was found in unit 30/B. Further up the
slopingterrain,inunit30/E,cruciblesandslag
indicatethatbothcastingandsmithingwerecarriedout.Thedepositsinunit30/Bwerefluvial
layers and the fine metal offcut may well have
been transported from 30/E to 30/B by fluvial
action. The fine metal offcut could therefore
representactivityfurtherupthemorainicslope
around unit 30/E rather than around unit 30/
B.Inthatcase,onlyoneactivityareaformetalworkingisrepresentedonsite30.
was also evidence of smithing. The three units
comprised two basic and one supplementary
source. On one plot smithing alone was documented, also through a basic source. This plot
mayperhapscorrespondwiththesmith’sbooths
mentioned in written sources. Altogether the
tendencyinthematerialthatcastingandsmithing took place in Bergen during horizon 5, is
consideredwell-founded.
168
Touchstone?
Weight/balancearm
Offcutfinemetal
Offcutcopperalloy
Furnacelining
Slag
Mould
Crucible
Stoneworking
Stoneworking in large-scale in early Bergen is
representedthroughthemonumentalbuildings
thatwereerectedduringtheperiodunderstudy.
Lime-slaking pits and lumps of chalk that perhaps served as raw material for mortar may be
Table37.Horizon5(1120s-c1170),metalworking(N=42)
otherindicators.Inaddition,stoneworkingofa
smallerscaleisdocumentedthroughoffcutsand
Plot/
blanksfromthetownarea.
unit
The artefact category ‘discus’ comprises flat,
circular slabs of slate with a diameter of 8.513.5cmandlessthanonecmthick.Theymay
be interpreted as blanks but have also be inter6/C
(2)
(1)
(1)
(1)
(1)
pretedasbelongingtoanoutdoorgameactivity
6/D
1
1+
(Herteig1969,198).Alonetheycannotsignify
(1)
(1)
1
(1)
(1)
small-scalestoneworking.Offcutsofstonehave
(1)
6/G
1
1+
probablynotbeendocumentedorcollectedsys8/B
(1)
tematicallyatthesites.Mostlikelyonlyoffcuts
+
that stand out as somehow special have been
20/A
(1)
26/A
3
documentedorcollectedasthissourcegroupis
(2)
(1)
oftenconsideredaspartofthemakeupofalayer
26ratherthanasanartefactinthelayer.Noneof
27/B (4)
26therecordedlayersfromearlyBergenhavebeen
27/
(5)
(1)
describedascontainingsteatiteoffcut.56Noneof
BC
thefindsthatwereoriginallyclassifiedasoffcut
27/C (1)
(2)
28/B (1)
ofsteatiteorslatewereretrievedinthemuseum
30/B
(1)
storerooms, I feel confident though, that the
30/E (2)
(1)
+
findswereoriginallyclassifiedcorrectlybecause
thestonetypesarenotdifficulttorecognise.The
ArtefactcategoryIIinbrackets
+cataloguedbytheauthor
stonetypeswerequarriedoutsidetheBergenarea
Numbersinboldrefertobasicsources,inplain
andmuststemfromstonesthathavebeencarsupplementarysources
ried into town as finished items, raw materials
orboth.Altogether,thedocumentedmaterialis
probablynotrepresentativeforwhatwasfound
To sum up, remains of metalworking have during excavations and the representation of
been identified in 9 of the 24 artefact-yielding small-scalestoneworkingplacesmustbeconsidanalytic units in horizon 5. Casting took place eredasaminimum.
in8unitscomprisingsevenbasicandonesupplementarysource.Inthreeoftheseunitsthere
Horizon2(c1020/30-c1070)
Apieceofsteatiteoffcutwasassignedtohorizon
2. It may reflect small-scale stoneworking carriedoutinoraroundunit7/A.Asthefindstems
from one supplementary source only, it is consideredastoouncertainfortheidentificationof
small-scalestoneworkingduringhorizon2andis
notincludedinfurtherdiscussions.
‘Discus’
(1)
Large-scale
stoneworking
‘Discus’
Miscellaneousblank
Offcutslate
Offcutsteatite
Chalk
Lime-slakingpit
Large-scale
stoneworking
Site
Spindlewhorlblank
Table39.Horizon3(c1070-c1100),stoneworking(N=1)
Large-scale
stoneworking
6/C
30/E
StMary’s
StNicholas’
Munkeliv
Abbey(site43)
‘Discus’
Miscellaneousblank
Offcutslate
Plot/unit/site
Offcutsteatite
Horizon3(c1070-c1100)
Accordingtowrittensourcesthebuildingofthe
ChristchurchCathedralwasinitiatedatHolmen
during horizon 3. This information is a basic
sourceandisconsideredreliable.
No traces of small-scale stoneworking have
beenassignedtothehorizon.
Spindlewhorlblank
Table40.Horizon4(c1100-1120s),stoneworking(N=5)
Chalk
7/A
ArtefactcategoryIIinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
Lime-slakingpit
Miscellaneousblank
Spindlewhorlblank
Offcutslate
Small-scalestoneworking
Large-scale
stoneworking
Large-scale
stoneworking
Offcutsteatite
Chalk
Lime-slakingpit
Unit
Large-scale
stoneworking
Table38.Horizon2(c1020/30-c1070),stoneworking(N=1)
initiatedduringhorizon4.Inthemiddletown
areatheChurchofStNicholas(S)mayhavebeen
erectedandatNordnestheMunkelivAbbeywas
initiated (B). The monuments are assigned to
horizon 4 as two supplementary and one basic
source respectively, should the supplementary
sourcesbeerroneouslyassignedtohorizon4,the
tendencythatlarge-scalestoneworkwascarried
outinthishorizonisstillconsideredreliable.
Steatitewasworkedonplot6/CbytheVågen
shorelineinthenortherntownareaandinunit
30/EatthefootofFløyfjelletinthemiddletown
area(Table40).Small-scalestoneworkisindicatedthroughbothabasicsourceandasupplementarysource.Thesourcesarefewinnumber,
butduetotherepresentativityproblemsinherent
inthematerialprobablyrepresentaminimumof
places where small-scale stoneworking was carriedout.Thetendencydiscernedinthematerial
thatthisactivitywascarriedoutinhorizon4,is
thusfoundreliable.
Small-scalestoneworking
1
(1)
x
x
X
ArtefactcategoryIIinbrackets
Numbersandxinboldrefertobasicsources,
numbersandxinplainsupplementarysources
Horizon5(1120s-c1170)
In horizon 5 large-scale stoneworking probably
Christchurch
X
Cathedral
beganinnineplaces.Inthenortherntownarea
Numbersandxinboldrefertobasicsources,
the (second?) church of St Mary (B) and the
numbersandxinplainsupplementarysources
churchofStPeter(S)wereerected,inthemiddletownareaStColumba(S),andinthesouthern town area the churches of St Olav (B) and
Horizon4(c1100-1120s)
Inthenortherntownareathepossiblepredeces- St Cross (B). In the Nordnes and Nonneseter
sortothestandingStMary’s(S)mayhavebeen areastheestablishmentofthetwomonasteriesSt
Small-scalestoneworking
11CraftsandproductioninearlyBergen
169
John’sAbbeyandtheNonneseterConventtook
place.
Large-scalestoneworkisalsoimpliedontwo
plots by the Vågen shoreline in the northern
town area by the presence of lime-slaking pits
and lumps of chalk on plot 6/C and perhaps
also through the lump of chalk found on plot
6/G.TheplotswerelocatednearStMary’sand
StPeter’srespectivelywheremortarwasprobably
neededinconnectionwithconstructionandrepairwork.Large-scalestoneworkisdocumented
through several basic sources and the tendency
inthematerialthatthisactivitywascarriedout
duringhorizon5isconsideredreliable.
footofFløyfjellet.Inadditiontothis,‘discuses’
onplot6/Cmayperhapsindicatestoneworking
onthisplot,however,thisartefactcategorycannot in isolation provide evidence of small-scale
stoneworking. Small-scale stoneworking was
documented through four basic and one supplementarysource.Thetendencyinthematerial
thatthisactivitytookplaceduringhorizon5,is
thusconsideredreliable.
ArtefactcategoryIIinbrackets
Numbersandxinboldrefertobasicsources,
numbersandxinplainsupplementarysources
Woodworking
Large-scalewoodworkingisobviouslyrepresentedbythetimberbuildingsandstructuresuncoveredatalmostalloftheplots/siteswhereactivity has been documented archaeologically. The
distribution of these sources is not listed here.
Monumental buildings, not recorded archaeologicallybutknownthroughthewrittenrecords,
alsoshowthatlarge-scalewoodworkwascarried
out in the early town. The timber monuments
that were initiated during the period under investigationarelistedinthetablesfortherelevant
horizonsbelow.
Blanks, tools and waste - lathe-turned cores
only-havebeenstudiedwhenidentifyingareas
with small-scale woodworking. Lathe-turned
coresmayhavebeenusedaswhippingtops(toys)
inwhichcasetheywereprobablyremovedfrom
their original place of production, though not
necessarily from the plot where they were produced.Still,lathe-turnedcorescannotinisolation be indicators of small-scale woodworking.
There may be tools for woodworking among
themetalfinds,knivesmayobviouslyhavebeen
usedforcarving,butasmentionedearlierIhave
notbeentoabletosortoutspecialisedtoolswith
anydegreeofcertaintyandthismetalfindscategoryisomittedinmysurvey.Asalreadypointed
out,woodblanksandwastewereprobablyoften
burntasfirewood,thisofcourseaffectstherepresentativityofthematerial.Thenumberofareaswheresmall-scalewoodworkingwascarried
outmustthereforebeconsideredasaminimum.
Small-scalestoneworkingwasdocumentedon
threeplotsbytheVågenshorelineinthenorthern
townarea,onplot28/BbytheVågenshoreline
inthemiddletownareaandonplot30/Eonthe
Horizon3(c1070-c1100)
AccordingtobasicsourcesChristchurchMinor
wasbuiltinwoodatHolmen(Table42)inhorizon3.Thepresenceoflarge-scalewoodworking is thus considered reliable. No small-scale
6/C
Large-scale
stoneworking
X
3
6/D
6/E
6/G
28/B
30/E
StMary’s(site23)
StPeter’s(site24)
StColumba(site
33)
StOlav’sinVågsbunnen(site39)
TheChurchofSt
Cross(site40)
StJohnsAbbey
(site44)
Nonneseter
Convent(site46)
170
1
‘Discus’
Miscellaneousblank
Spindlewhorlblank
Offcutslate
Offcutsteatite
Chalk
Large-scale
stoneworking
Plot/unit/site
Lime-slakingpit
Table41.Horizon5(1120s-c1170),stoneworking(N=33)
Small-scalestoneworking
1
(1)
(1)
9
2
(1)
1
1
(4)
(1)
(1)
X
x
x
X
X
X
X
Needle/pinblank
Rabbet
X
Needle/pinblank
X
Rabbet
TheChurchoftheApostles
(site4)
Lathe-turnedcore
Site
Monumentalbuilding
Table43.Horizon4(c1100-1120s),woodworking(N=1)
6/B
6/C
6/D
6/E
6/F
StOlav’sontheHill(site25)
TheChurchofAllSaints
(site45)
(1)
(3)
(1)
Needle/pinblank
Horizon4(c1100-1120s)
According to basic sources the Church of the
ApostleswasbuiltatHolmeninhorizon4.The
presenceoflarge-scalewoodworkingisthusconsidered reliable. No small-scale woodworking
has been recorded in the sources for horizon 4
(Table43).
Rabbet
Plot/site
Lathe-turnedcore
Table44.Horizon5(1120s-c1170),woodworking(N=10)
Numbersandxinboldrefertobasicsources,
numbersandxinplainsupplementarysources
Monumentalbuilding
ChristchurchMinor(site3)
Lathe-turnedcore
Monumentalbuilding
woodworking has been recorded in the sources inisolationprovideevidence.Small-scalewoodforhorizon3.
working was thus documented on at least two
plots.Thesourcesforsmall-scalewoodworking
Table42.Horizon3(c1070-c1100),woodworking(N=1)
are assigned to horizon 5 as basic sources. The
numberofartefactsissmall,butduetotheinherentproblemsofrepresentativitytherecorded
Site
number of places where small-scale woodworkingtookplaceshouldprobablybeconsidereda
minimum. The tendency in the material that
thisactivitytookplaceduringhorizon5isthus
consideredreliable.
(1)
(1)
(1)
X
X
ArtefactcategoryIIinbrackets
Numbersandxinboldrefertobasicsources,
numbersandxinplainsupplementarysources
Numbersandxinboldrefertobasicsources,
numbersandxinplainsupplementarysources
Horizon5(1120s-c1170)
TheChurchofStOlav’sontheHill(B)andperhapsalsotheChurchofAllSaints(S)werebuilt
inhorizon5,presumablyinwood.Thepresence
of large-scale woodworking is thus considered
reliable.
In the northern town area along the Vågen
shorelinearabbetandaneedle/pinblankshow
thatsmall-scalewoodworkingtookplaceattwo
plots,thefindsoflathe-turnedcoresonadditionallytwoplotsmayperhapsindicatethatwoodworking was also carried out here, but cannot
11CraftsandproductioninearlyBergen
171
Skinning
Cutandchopmarksonosteologicalmaterialindicatethatanimalswereskinned(Hufthammer
1987, 64ff). Four craniums of cats and five of
dogscomprisethesourcesforskinninginearly
Bergen.
6/D
2
(1)
6/E
(1)+
Craniumdog
Plot
Craniumcat
Table45.Horizon5(1120s-c1170),tracesofskinning:
skullsofcatsanddogs(N=12)
(2)
(1)
(2)+
ArtefactcategoryIIinbrackets
+studiedbyHufthammer
Numbersinboldrefertobasicsources,inplain
supplementarysources
Table 45 shows that skulls assigned to horizon
5,asbasicsources,werefoundonplots6/Dand
6/Einthenortherntownarea.57Oneofthecat
skulls and two of the dog skulls from plot 6/E
have been studied earlier, all three skulls had
choporcutmarksandshowthatskinningwas
carriedouthere(perscomHufthammer2002).
Cut and chop marks are very common in the
materialfromthewholemedievalperiodinthe
Engelgårdenarea(Hufthammer1987,64ff).On
thisbasisthereisagoodchancethatsomeofthe
threecatandtwodogskullsfromplot6/Dalso
indicate skinning. However, skinning has only
beendocumentedsafelyonplot6/E.Thepresenceoftheactivityisdocumentedthroughabasicsourceandisconsideredreliable.
Textileproduction
Artefacts that could be associated with textile
productionhavebeendividedintotwogroups:
‘textiletools’and‘possibletextiletools’.Thefirst
group contains artefacts that are characteristic
andnoteasilymistakenforitemswithotherareasofuse.Thesecondgroupcomprisesartefacts
that are less distinct and may have had other
areas of use. Weights and needles/pins belong
to the group of ‘possible textile tools’. In other
172
studies of textile production weights have been
analysedalongwithtextiletools(egØye1988,
Hagen(1988)1994,Nordeide1989).Studiesof
textiletoolsandfishingtacklefromBergenhave
however, demonstrated the difficulties in distinguishing positively between weights used as
warp-weightsandthoseusedasnet-weights.In
thesestudiestheweightandtosomeextentthe
shapeoftheobjectshasbeenusedasanindicator
of function, but the context of the weights has
been considered the most important indicator.
Accordingly,iftheweightswerefoundwithtextiletoolstheywereinterpretedas‘possiblewarpweights’,iffoundalongwithfishingtacklethey
wereinterpretedaspossiblefishingrelatedequipment. (Øye 1988, 70; Olsen 1998, 57, 87ff).
When, in the present study, weights are found
astheonlyindicatoroftextileproductionIwill
discussthefunctionoftheconcreteweightson
the basis of weight as described in Øye (1988,
69)andonthebasisofcontext.Needles/pinsare
alsotraditionallyusedassourcesfortextileproduction,butmayalsohavehadseveralotherareasofuse(Øye1988,97ff).Whenneedles/pins
arefoundastheonlysourcefortextileproduction,Iwilldiscussthefunctionoftheparticular
needles/pins.Notextiletoolscouldbeassociated
withhorizon2.
Horizon3(c1070-c1100)
Onepossibletextiletoolwasassignedtohorizon
3 (S); the weight on plot 9-10/B by the Vågen
shoreline in the northern town area was, however, found along with fishing tackle and thus
probably functioned as a net-weight. Textile
productioncouldnotbedocumentedinhorizon
3(Table46).
9-10/B
(1)
*
ArtefactcategoryIIinbrackets
*Mostlikelyanet-weight
Numbersinboldrefertobasicsources,inplain
supplementarysources
Textile
tools
Plot
Possibletextiletools
(1)
*
6/B
26-27/
BC•
Textile
Production
witnessed
positively
Flax-beater
Flax-comb
Knifebeater
Linen-smoother
Long-toothedcomb
Needle/pin
Reed-hook
Shears
Warpingpaddle
WeightT/F
Windingpin
Textile
Production
witnessed
positively
Spindlewhorl
Drop-spindle
Spindlewhorl
Plot
Possibletextiletools
Flax-beater
Flax-comb
Knifebeater
Linen-smoother
Long-toothedcomb
Needle/pin
Reed-hook
Shears
Warpingpaddle
WeightT/F
Windingpin
Textile
tools
Table47.PossibletextiletoolsinHorizon4(c1100-1120s)
(N=4)
Drop-spindle
Table46.PossibletextiletoolsinHorizon3(c1070-c1100)
(N=1)
(1)
+
ArtefactcategoryIIinbrackets
*Mostlikelyanet-weight
+Mostlikelynotatextiletool
Numbersinboldrefertobasicsources,inplain
supplementarysources
Horizon4(c1100-1120s)
•Artefactsfrombothbasicandsupplementary
Notextiletoolswereassignedtohorizon4inthe sources
sevenfind-yieldingunits.However,five‘possible
textile tools’, were assigned to the horizon (Table 47). The weight on plot 6/B by the Vågen
shoreline in the northern town area was found
along with fishing tackle and should rather be
associated with fishing. The needle/pin found
betweenplots26-27/Band27/Cinthemiddle
townareawasofØye’stypeA(Øye1988,Figure
IV.2).Thesizeandshapeoftheheadindicates
thatthisparticularneedle/pincannothavebeen
usedforsewing,ratheritmayhavebeenusedas
a pin to fasten garments (Cf Øye 1988, 99ff).
Theneedle/pincanthereforenotbetakenasan
indicationoftextileproductiononeitherofthe
plots.Inunit30/Elocatedin
11CraftsandproductioninearlyBergen
themiddletownareaatthefootofFløyfjelleta
weightofØye’stypeA(Øye1988,TableIII.4.2)
was found, weighing about 162 g. Judging the
weight and shape, the artefact was most likely
used as a net-weight (cf Øye 1988, 69), and it
cannotbetakenasanindicationoftextileproduction.Inconclusiontextileproductioncannot
bedocumentedinhorizon4.
Horizon5(1120s-c1170)
Inhorizon5textiletoolswerefoundoneightof
the24find-yieldingunitsandoneitherplot2627/Borplot27/C(Table48).Thefindssignify
textileproductionintheseunits.Weightsclassifiedas‘possibletextiletools’werepresentinfour
additionalunits.Weightsatplot6/F,6/Gandin
unit8/Bwerefoundtogetherwithfishingtackle,
andarethusmostlikelynet-weights.Theweight
onplot15-16/AwasofØye’stypeB,thefunction of the weight cannot be determined with
any degree of certainty. A needle/pin of Øye’s
typeAorBwasfoundonplot26-27/B,andmay
havebeenusedfortextileproduction.Aspindle
whorl was found between 26-27/B and 27/C
and indicates textile production on either plot.
Tosumup,textileproductionwasdocumented
on9plots.Inadditiontothis,aweightthatcan173
Table49.Horizon3(c1070-c1100),fishingtackleand
Fishing
Fishing tackle has been classified according
to Ole Mikal Olsen’s classifications (1998). I
havedividedartefactsintotwocategories:fishingtackleand‘possiblefishingtackle’.Thefirst
groupcomprisesartefactsthatareidentifiedaccordingtofunction,thesecondgroupcomprises
weightsthatmayreflecteithertextileproduction
orfishing.Thefunctionoftheweightsthatare
consideredasindicatorsoffishingrelatedactivities, is discussed and evaluated case by case on
criteriasimilartothoseaccountedforundertextiletools.
Nofishingtacklecouldbeassociatedwithhorizonspriortohorizon3.
Horizon3(c1070-c1100)
A weight tentatively assigned (S) to horizon 3
wasfoundonplot9-10/Binthenortherntown
area(Table49).Theweightandshapeoftheobjectindicatethatitwasusedasanet-weight(cf
p227).Sincetheevidencestemsfromonesupplementary source only, it is considered as too
uncertainfortheidentificationoffishingduring
horizon3andisnotincludedinthefurtherdiscussions.
174
Fishing
positively
identified
Weight
Possible
Fishing
tackle
Nettingneedle
Linerunner
Linesinker
Float
notbeclassifiedintermsoffunctionmayhave possiblefishingtackle(N=1)
served as either a warp weight or a net-weight.
Fishingtackle
The production places were located along the
VeisanandVågenshorelinesinboththenorthern and the middle town areas and at the foot
ofFløyfjelletonplot20/Ainthenortherntown Plot
area.Theplacesofproductionweredocumented through eight basic and one supplementary
source,thetendencyinthematerialthattextile
wasproducedduringhorizonisthusconsidered
reliable.
9-10/B
(1)
X
ArtefactsofcategoryIIareinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
Horizon4(c1100-1120s)
Fishingtackleassignedtohorizon4wasfound
in the northern town area on plot 6/B and in
themiddletownareaonplot26/A.Fishingrelated activities are thus documented on two of
sevenfind-bearingplots,inhorizon4(Table50).
Fishing was documented through basic sources
only, the tendency in the material that fishing
tookplaceduringhorizon4,isthusconsidered
reliable.
Table48.TextiletoolsandpossibletextiletoolsinHorizon5(1120s-c1170)(N=177)
6/C
5(3)
(1)
6/D
3(7)
1
6/E
6/F
(1)
1
(1)
(1)
1
(2)
3
(1)
1
(2)
1
4
(12)
(6)
1
1
(1)
6/G
8/B
8/D
15-16/A
1
20/A
26/A
26-27/B
(1)
(1)
1
Windingpin
WeightT/F
4
(5)
10
(13)
27
(8)
(3)
(3)
*
6
(1)
*
(1)
*
X
(1)
X
X
X
X
(1)
?
(1)
(1)
(1)
+
(1)
26-27/BC
28/B
Warpingpaddle
Textile
Production
witnessed
positively
Shears
(1)
3
(2)
(6)
Reed-hook
(1)
Knifebeater
(1)
Needle/pin
2(3)
Flax-comb
6/B
Flax-beater
Drop-spindle
Spindlewhorl
Plot
Long-toothed
comb
Possibletextiletools
Linen-smoother
Textiletools
(1)
(1)
(1)
(1)
(3)
X
X
X
X
X
ArtefactsofcategoryIIareinbrackets
*Mostlikelynet-weights
?Functionuncertain
+Mostlikelyatextiletool
Numbersinboldrefertobasicsources,inplainsupplementarysources
Table50.Horizon4(c1100-1120s),fishingtackleand
possiblefishingtackle(N=3)
6/B
26/A
(1)
Fishing
positively
identified
Weight
Linerunner
Linesinker
Float
Plot
Possible
Fishing
tackle
Nettingneedle
Fishingtackle
(1)
1
X
X
ArtefactsofcategoryIIareinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
11CraftsandproductioninearlyBergen
Horizon5(1120s-c1170)
Inhorizon5,fishingtacklewasfoundintenanalyticunitsalongtheshoresofVeisanandVågen
inthenortherntownareaandinthreeunitsalong
theVågenshorelineinthemiddletownarea.In
addition, a weight was found in unit 20/A locatedatthefootofFløyfjellet.Thisweightwas
foundtogetherwithtextileequipmentandmost
likely functioned as a warp-weight (cf p 228).
Onplot15-16/Ayetanotherweightwasfound,
thefunctionofthisweightcannotbedetermined
withanycertainty(cfp228).Allinall,fishing
was well-documented in 10 of the 24 artefactyieldingunits,andaweightmayindicateeither
textileproductionorfishingrelatedactivitieson
one plot (Table 51). Fishing was documented
175
through ten basic sources, the tendency in the likelyshowsaminimumofplaceswhereweapmaterialthatfishingwasanactivityduringhori- onswereowned,ratherthanarealpictureofthe
distribution(Nøttveit2000).
zon5,isthusconsideredwell-founded.
No weapons could be assigned to horizons
Table51.Horizon5(1120s-c1170),fishingtackleand
priortohorizon5.
possiblefishingtackle(N=128)
(1)
(1)
(2)
2
(1)
8/B
15-16/A
20/A
26-27/B
26-27/
BC
27/C
28/B
(1)
(1)
(1)
X
X
X
X
X
X
(1)
X
X
(3)
X
X
(1)
(1)
X
ArtefactsofcategoryIIareinbrackets
*Mostlikelyawarp-weight
?Functioncannotbedetermined
Numbersinboldrefertobasicsources,inplain
supplementarysources
Table52.Horizon5(1120s-c1170),weaponsofwar,
huntingandgame(N=9)
Plot
6/B
6/C
Slingwar/
hunt/game
(3)
WeightsT/F
Nettingneedle
(1)
4
(5)
10
(13)
27
(8)
(3)
(3)
6
(1)
(1)
(1)?
(1)*
Arrowhead,hunt
6/E
6/F
6/G
(1)
Horizon5(1120s-c1170)
Weapons have been found on four plots in the
northerntownareaandononeplotinthemiddle
townareainhorizon5(Table52).Allplotswere
locatedbytheVågenshoreline.Weaponsofwar
were found on two plots and hunting weapons
ontwoplots.Abowthatmayhavebeenusedeitherforhuntingorwarwasfoundonafifthplot.
Theweaponshaveallbeendocumentedthrough
basic sources. It is thus well-documented that
weaponsforbothhuntingandwarwereowned
by townspeople during horizon 5 and hunting
wasprobablypartofthetownspeople’sstrategy
forgatheringfoodoracquiringfur.
Spearhead,war
6/D
2
Fishing
positively
identified
Arrowhead,war
6/C
3
(2)
4
(3)
6
(4)
Possible
fishing
tackle
Bowhunt/war
6/B
Linerunner
Float
Plot
Linesinker
Fishingtackle
1
1
1
(1)
2
(1)
Huntingandwar
Weaponsassignedtotheperiodunderstudyhave
been classified according to Ole Magne Nøttveit’sclassification(2000).Inaddition,leather
items identified as slings (cf Dahlbäck 1983,
264; Marstein 1989, 115) are included. Slings
wereusedasaweaponandashuntingequipment
aswellasingamesaswell(KLNM,VII322ff,
XVI229).Someoftheidentifiedweapons,such
assomeofthearrowheads,wereforhuntingand
indicatethathuntingwaspartofthetownspeople’sstrategyforgatheringfoodoracquiringfur.
Otherweapons,suchasspearheads,weremeant
forwarorclassdistinction.Weaponswereprobablysomethingtheownercaredwellfor,thusthe
distribution of weapons on the plots/sites most
176
6/D
6/G
26-27/B
1
(1)
Numbersinboldrefertobasicsources,inplain
supplementarysources
ArtefactsofcategoryIIareinbrackets
AgricultureinearlyBergen
Agricultureinthiscontextisdefinedastheproductionofplantsandanimalhusbandry(cfØye
1998.7).Theosteologicalmaterialisnotavailableasasourceforagricultureinhorizons2to
5andthebotanicalsourcesare,withoneexception,notrelevant.Øyehasgonethroughthemuseumstoreroomsinordertoidentifytoolsused
inagriculture,thetoolsidentifiedbyherwereall
Horizon5(1120s-c1170)
In unit 21/A, at the foot of Fløyfjellet in the
northern town area, construction K39 was interpreted as a stable or byre through remains
of dung within the building (Table 53). It was
notpossibletodeterminewhetherthedungwas
fromhorsesorcattle(Dunlop1989f,28;Hjelle
1989,7).Iftheconstructionwasahorse-stable
andnotabyreitdoesnotreflectagricultureas
such,sincehorsesweremostlyusedfortransport
andhorsefleshwasnoteatenintheMiddleAges
(Øye 1998, 44, 53). Since the function of the
buildingisambiguous,itcannotstandaloneasa
sourceforagriculture.
On plot 6/G by the Vågen shoreline, in the
northerntownarea,apossiblesicklewasfound
andmayreflectharvesting.Theitemstemsfrom
abasicsource,butwassobadlypreservedthatit
cannotbeusedasconclusiveevidenceofagricultureinearlyBergen.
Altogether the sources of agriculture during
horizon5cannotthrowlightuponthepresence
ofagriculturalactivities.
11CraftsandproductioninearlyBergen
Plot
6/G
21/A
Possiblesickle
Table53.Horizon5(1120s-c1170),agriculturesources
(N=2)
Stableorbyre
youngerthanhorizon5.Duringmyownreclassification of the finds from early Bergen I have
only retrieved one possible tool. Culture-layersthatcontainanimaldungandartefactsmay
give information as to whether the inhabitants
of early Bergen were also agriculturists. Before
1980dungwasnotrecordedasacomponentin
culture-layers, and at sites excavated after 1980
noneoftherecordedlayerscontaineddungaccording to the original documentation. Excrement-layershavebeenrecordedatallexcavations
since 195558. At site 21 layers 67 and 68 were
describedbythearchaeologistsaslayers‘possibly
withexcrement’(Dunlop1989f,20).Thebotanicalanalysisofthelayers,however,alsoidentified
dung(Hjelle1989,7).Thisimpliesthataclear
distinctionbetweendungandexcrementisnot
alwaysfeasiblewithoutabotanicalanalysisofthe
deposits. Most of the sites excavated after 1955
hadlayerswhereexcrementwaspartofthecomposition,anditcannotbeexcludedthatsomeof
this‘excrement’wasactuallydung.Theabsence
ofdungcanthereforenotbeusedasasourcefor
theabsenceofanimalhusbandry.
1
X
ArtefactsofcategoryIIareinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
Basiccooking,foodandbeverageprocessing
Foodmusthavebeencookedandconsumedon
all the occupied plots/sites in the early town.
Werefoodandbeveragesalsoprocessedfromraw
materialtorefinedproductsinallanalyticunits,
orcandifferentformsofspecialisationbeidentified?Refinementoffoodstuffsandbeveragespresupposesthatthesettlementonaplot/sitewhere
suchactivitiestakeplacewaswellestablishedas
opposed to being sporadic; food and beverage
processing depend on the presence of extensive
facilities, such as large containers and access to
a fireplace where large amounts of water could
beheated.Sausagemakingforexamplepresupposesthatanimalswerekilledanddressed,this
isamoretime-consumingactivitythanthebasic
cookingtaskofboilingporridgeforexample.
Inordertoelucidatewhetherdifferentforms
ofspecialisationtookplacewithinpreparationof
foodandbeverages,placeswherebasiccooking
was carried out and where more time-demanding food and beverage processing took place
are identified. Tools used when processing raw
materials into refined foodstuffs and botanical
evidenceofbeerbrewingareusedassourcesfor
thelatteractivity.Thetoolsthathavebeenidentified are: sausage pins - used when processing
meatintosausagesforstorageorimmediateconsumption(cfWeber1990,76ff),skewers-used
when drying fish, grinding slabs, and a cross -
perhapsusedwhenprocessingmilk(Ågotnesin
prep).Asapremise,whenidentifyingwheresausagesweremade,Iassumethatthesausagepins
were removed before the sausage left the place
of production or storage and were served. The
toolsrepresentingbasiccookingaresteatiteves177
Table54.Horizon2(c1020/30-c1070),foodandbeverage
processing(N=1)
7/A
Baking-stone
Steatitevessel
Basiccooking
Myricagale
Milk-processing
cross
Grindingslab
Skewer
Plot
Sausagepin
Foodandbeverageprocessing
(X)
ArtefactcategoryIIinbrackets
Horizon3(c1070-c1100)
PollenofMyricagalewasrecordedin‘unit7’on
eitherplot6/Eor6/F(KrzywinskiandKaland
1984,24),anditmayindicatebeerbrewingon
eitherplot.However,sincedatastems
Table55.Horizon3(c1070-c1100),basiccooking,food
andbeverageprocessing(N=4)
Steatitevessel
Myricagale
Grindingslab
Basiccooking
Baking-stone
6/Eor
6/F
9-10/B
Skewer
Plot
Milk-processingcross
Foodandbeverageprocessing
Sausagepin
selsandbakingslabsofslate-usedwhenbaking
flat-breadforstorageorimmediateconsumption
(cf Weber 1990, 62), most likely also for heating other foodstuffs over the hearth (Ågotnes
inprep).59Manywoodenartefactsmayalsoreflectbasiccooking,butaclearidentificationof
theirfunctionisnotstraightforward,sotheyare
notincludedhere.Neitheristhedistributionof
cookingvesselsofpotterydrawnintothediscussion,asthisartefacttypehasnotbeenclassified
asanindividualcategoryatallsites.Steatitevesselsandbakingslapswillthereforesufficetorepresentbasiccooking.Ihavenotre-classifiedthe
sausagepins,thesteatitevesselsnorthebaking
slapsfromsite6,asthismaterialwasdifficultto
accessinthemuseumstorerooms.Theitemsare,
however, easy recognisable and there is a good
chance that they were properly identified duringtheoriginalfinddocumentation.Evenifall
theitemswerenotclassifiedcorrectly,thelarge
numberoftherespectiveartefactsareregardedas
quiterepresentativeforwhatwasactuallyfound
duringexcavation.60Pollenofmyricagalerecordedthroughbotanicalinvestigationsmayindicate
beerbrewing,astheplantwasacommoningredient in beer in the twelfth century (KLNM,
XX689ff;Kjersgaard1978,84ff).Asmyricagale
alsogrowswildandhadotherareasofuse(Høeg
1976,457ff),itspresenceisonlytakenasanindicationofbeerbrewing.
(1)
(2)
X
ArtefactcategoryIIinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
•Artefactsfrombothbasicandsupplementary
Horizon2(c1020/30-c1070)
Inhorizon2pollenofmyricagalemayimplythat sources
beerbrewingtookplaceinthevicinityofsite7/A
inthenortherntownarea(Table54).Sincedata
stems from one supplementary source only, the
evidenceofbeerbrewinginhorizon2isconsideredastoouncertainandisnotincludedinthe
furtherdiscussions.
from one supplementary source only, the evidenceofbeerbrewinginhorizon3isconsidered
astoouncertain,andisnotincludedinthefurtherdiscussions.Basiccookingwasdocumented
by sources assigned tentatively to horizon 3 on
plot 9-10/B, again since data stems from one
supplementarysourceonly,theevidenceofbasic
cookinginhorizon3isconsideredastoouncertaintobeincludedinfurtherdiscussions(Table
55).
Horizon4(c1100-1120s)
Basic cooking was documented in five of the
seven find-bearing analytic units in horizon 4
(Table56).Foodwasalsoprocessedintwounits
178
in the northern town area and on two plots in
themiddletownarea.Basiccookingwasdocumentedthroughthreebasicandtwosupplementary sources, food and beverage processing was
documentedthroughthreebasicandonesupplementarysource.Themainpatterndiscernedin
thematerialthatbasiccookingandfoodprocessing were carried out during horizon 4, is thus
consideredwell-founded.
possiblyhavebeenbrewedinthreeplaces,cereals ground in one place and milk processed in
oneplace.Foodandbeverageprocessingisdocumentedthroughtenbasicsourcesandtwosupplementarysources.Eachofthevarietiesoffood
andbeverageprocessinghavebeendocumented
throughatleastonebasicsource,sothetendency
thatfoodandpossiblyalsobeverageswereprocessed during horizon 5, and that all the varietiesoffoodandpossiblyalsobeverageprocessing
Table56.Horizon4(c1100-1120s),basiccooking,foodand wererepresented,isconsideredreliable.
beverageprocessing(N=36)
Table57.Horizon5(1120s-c1170),basiccooking,foodand
beverageprocessing(N=1265)
(1)
(1)
6/E
(1)
(2)
(1)
ArtefactcategoryIIinbrackets
Numbersinboldrefertobasicsources,inplain
supplementarysources
•Artefactsfrombothbasicandsupplementary
sources
6/F
6/G
8/A
8/B
15-16/A
(138)
(3)
3
(2)
(2)
3
(2)
(1)
(1)
1
Steatitevessel
6/D
Myricagale
(5)
16
(38)
49
(122)
107
(244)
Milk-processing
cross
6/C
6/B
Grindingslab
(7)
1
Skewer
(11)
Basiccooking
Baking-stone
(2)
Steatitevessel
(2)
1
1
Plot/unit
Sausagepin
Foodandbeverageprocessing
Myricagale
Grindingslab
Basiccooking
Baking-stone
6/B
6/C
26/A
26-27/B•
26-27/BC•
27/C
30/E
Skewer
Sausagepin
Plot/unit
Milk-processingcross
Foodandbeverageprocessing
20
(25)
60
(58)
33
(71)
4
(14)
13
(30)
21
(16)
1
(9)
(3)
19
(5)
(1)
(11)
(2)
(37)
(3)
7
(1)
2
3
(2)
20/A
(X)
Horizon5(1120s-c1170)
21/A
X
Inhorizon5,basiccooking(Table57)wasdoc- 26/A
(1)
1
(1)
umented in 17 of the 24 find-yielding analytic
units,locatedalongtheVeisanandVågenshore- 26-27/B (2)
(1)
(2)
linesinthenorthern,middleandsoutherntown 26-27/BC (2)
(19)
(4)
(5)
(2)
areas and at the foot of Fløyfjellet in the mid- 27/C
(3)
(5)
dle town area. Basic cooking was documented 28/B
(1)
(1)
through14basicsourcesandthreesupplemen- 28/C
(1)
tarysourcesandthepresenceofthisactivitydur- 30/B
30/E
(10)
(1)
ing horizon5isthuswell-foundedandconsid38/A
(X)
1
eredreliable.
ArtefactcategoryIIinbrackets
Foodandbeverageprocessingwasdocumentedin12ofthe24artefact-yieldingunitslocated Numbersinboldrefertobasicsources,inplain
supplementarysources
along the Veisan and Vågen shorelines in the
northern, middle and southern town areas and
atthefootofFløyfjelletinthemiddletownarea.
Sausagesmayhavebeenmadeinnineplaces,fish
may have been dried in three places, beer may
11CraftsandproductioninearlyBergen
179
Summary
A number of productive activities have been
identifiedinearlyBergenthroughtheavailable
sources (Table 58) Large- or small-scale stoneworking and possibly beer brewing were documentedthroughartefactassemblagesassignedto
horizon2assupplementarysources.Astheevidencefortheseactivitiesstemsfromsinglesupplementary sources, it has been considered too
uncertain to be included in further discussions
ofproductiveactivities.
In horizon 3 large-scale stoneworking and
large-scale woodworking at Holmen have been
identified through basic sources, and the presenceoftheseactivitiesisconsideredreliable.In
thenortherntownareashoemaking,otherleatherwork,fishing,basiccooking,andfoodandpossiblyalsobeverageprocessingwereindicatedby
supplementarysources.Sincetheindividualactivitiesweredocumentedthroughsinglesupplementarysourcesonly,theirpresenceinhorizon3
isconsideredtoouncertaintobeincludedinthe
furtherdiscussionsofproductiveactivities.
In horizon 4, miscellaneous antler, bone,
hornandwhale/walrusboneworking,shoemaking and other leatherworking, metalworking,
large and small-scale stoneworking, large-scale
stoneworking, fishing, basic cooking, and food
processing were activities all indicated through
bothbasicandsupplementarysources,thepresenceoftheseactivitiesinhorizon4isthusconsideredreliable.
In horizon 5, combmaking, miscellaneous
antler,bone,hornandwhale/walrusboneworking, shoemaking, ‘other leatherworking’, metalworking, large and small-scale stoneworking,
large and small-scale woodworking, skinning,
textileproduction,fishing,hunting,basiccooking,andfoodandpossiblyalsobeverageprocessinghaveallbeendocumentedthroughbothbasicandsupplementarysources.Thepresenceof
theseactivitiesisconsideredreliable.
Whatwasthenatureoftheproductive
activitiesandhowwerethey
organised?
Combmaking
Altogether81combshavebeenassignedtohorizons4and5,comb-blanksarerelatedtohorizon
5only.Thecombscompriseavarietyofcompositesingleordouble-sidedcombs,heldtogether
byrivetsmadeofrolledsheetsofcopperalloy.61
Allbuttwocombsstemfrombasicsourcesand
thepresenceofcombsisconsideredreliable.As
mentioned earlier the combs are classified accordingtothesystemdevelopedbyWibergand
Flodin with a few supplements from my side
(Wiberg 1977, 202-209; Flodin 1989, 29-33)
(Table59).62
Within each general type of combs there
areseveral‘variationsoverthesametheme’.All
blanksfromtheproductionplacesidentifiedin
?
X
X
X
X
X
X
X
X
X
?
X
X
?
?
X
X
Xinboldarebasedonbasicsourcesinadditiontosupplementary,xinplainisbasedonsupplementary
sourcesonly
?Datastemfromsinglesupplementarysourcesandareconsideredtoouncertaintobeincludedinfurther
discussions
180
Beverageprocessing
Foodprocessing
Basiccooking
Agriculture
Fishing
X
?
X
X
Hunting/war/game
Textileproduction
Skindressing
Small-scalewoodworking
Large-scalewoodworking
X
X
X
X
X
Small-scalestoneworking
?
X
X
Large-scalestoneworking
X
X
?
X
X
Metalworking
Leatherworking
2
3
4
5
Shoemaking
Combmaking
Horizon
Antler,bone,hornand
whale/walrusboneworking
Table58.Productiveactivitiesdocumentedfromhorizon2throughhorizon5
X
horizon 5 can be linked to comb types found
amongthefinishedproducts:ablankfromplot
6/Eshowsthatcompositesinglecombswereproduced here. On plot 6/G, at least three differentcombtypeswereproduced(E5-2,E5-3and
yet a single or double type) and on plot 28/C
compositedoublecombsweremade.Thecomb
typesfoundinhorizons4and5areall,except
typesE1,E5-1andD2,foundamongtheblanks
as well. This link between blanks and the finished products shows that most of the finished
combswere,orcouldhavebeen,producedand
purchasedinBergenduringhorizon5.
brosiani1981;Christensen1986),skillsandnot
least knowledge of what an up-to-date comb
looked like. Bergen comb types have parallels
in material from Oslo (Wiberg 1977; Wiberg
1987),Trondheim(Flodin1989)inNorway,in
Lund(Blomquist1942),Viborg(Nielsen1969)
andSchleswig(Ulbricht1984,Tafel71no1)in
medievalDenmark,inSigtuna(Floderus1941,
89) and Lödöse in medieval Sweden. This also
showsthattheproductswerestandardised.On
this basis there should be no doubt that the
combs associated with horizons 4 and 5 were
producedbyprofessionalcombmakers.
6/E
6/G
26/A
26-27/B
26-27/BC
27/C
28/B
28/C
30/E
38/A
1
(2)
1
(2)
DU
D2
1
(5)
(1)
D1
(1)
(1)
EU
(1)
(1)
E6-1
(1)
(1)
E5-5
E5-2
E5-1
E4
E3-b
E5-3tworowsofrivets
6/D
E5-3onerowofrivets,
oneprofile
Horizon4(c1100-1120s)
26/A
Horizon5(1120s-c1170)
6/B
(2)
6/C
(2)
E3
E1
Plot/Combtype
E5-3onerowofrivets,
noprofile
Table59.Combsassignedtohorizons4andhorizon5andaccordingtocombtype(N=81)
1
(1)
(3)
(1)
(1)
(2)
1
(2)
2
(2)
(1)
(1)
1
1
(1)
(1)
(1)
(1)
(1)
(1)
(1)
(1)
1
(2)
(1)
(1)
(1)
(1)
1
2
(2)
(3)
1
(1)
(1)
(1)
(1)
(1)
(2)
(1)
(1)
(1)
(1)
(1)
(1)
1
ArtefactcategoryIIinbrackets
Numbersinboldrefertobasicsources,inplainsupplementary
Were the combs produced professionally? In
somecases,twoorthreecombsaresosimilarthat
it is reasonable to assume that the same combmaker or workshop may have produced them.
The twin combs belong to horizon 5 and they
arefoundondifferentplots(Figure44andsee
alsoFigure45)indicatingthatthecombmakers
sold combs to people from various households.
Inadditiononemayarguethattheproduction
ofthecombsrequiredspecialisedtools(cfAm11CraftsandproductioninearlyBergen
Waste from comb production was found on
four of the artefact-yielding analytic units in
horizon 5 (Table 32). The presence of combmakingduringhorizon5isconsideredreliable.
Theamountofwasteoneachproductionplace
wasveryscarce,varyingfromoneto75blanks
and offcuts and must be characterised as small
becausetheyfallwithinthequantitycategories
characterised as small in studies from contemporary Scandinavian towns (eg Christophersen
181
Figure44.TwincombsfromBergen.TypeE5-3onerowofrivets,oneprofile:aBRM0/77536plot6/B,bBRM104/2383plot
26/A;typeE5-3tworowsofrivets:cBRM0/43711plot6/D,dBRM0/64328plot6/C,eBRM110/5483plot26-27/BC;
typeE-1:fBRM76/11106plot28/C,gBRM76/9807plot28/B,hBRM110/4605plot27/C;typeE5-3onerowofrivets,no
profile:iBRM0/45464plot6/D,jBRM0/72946plot6/C
182
Figure45.TwincombsoftypeE5-3,withtworowsofrivets,fromBergen,Lund,Viborg,SchleswigandLödöse.Bergen:a
BRM76/12652plot28/B,bBRM104/1987plot26-27/B,cBRM104/2276plot26/A,dBRM104/2369/01plot26/A,e
BRM110/5483plot26-27/BC;Lund:fL.U.H.M.15310282:A,gK.M.22802aVIII(Blomquist1943,144-145);Viborg:
h11B165(Nielsen1969,Figure26)FotoTuriThomsen;Lödöse:i2700-68-CA35(PerscomSoniaJeffery2002Lödöse
Museum);Schleswig:jnonumber(Ulbricht1984,Figure71:1)
11CraftsandproductioninearlyBergen
183
Figure46.Twincombs
oftypeE5-2fromBergen
andLund.Bergen:aBRM
104/2275plot26/A;Lund:
bK.M.8480Annegatan
(Blomquist1943,144)
1980, 126ff; Flodin 1989; Rytter 1997). The
sparseamountofwasteimpliesthattheartisans
primarilyproducedforaninterurbanmarketand
thesmallamountofwastedistributedonseveral
plotsimpliesthattheartisanseitherambulated
or were part-time occupied residents of Bergen
duringhorizon5.
Severalofthecombsfromhorizon5havecontemporarytwinsinothertownsinScandinavia.
I have found twins in Lund, Viborg and Schleswig in medieval Denmark and in Lödöse in
medievalSweden.63Thesimilaritiesbetweenthe
combs found in Bergen, Lund, Viborg, Schleswig and Lödöse (Figure 45 and Figure 46) are
sostrikingthatthesameartisansorworkshops
shouldprobablybeseenbehindthetwinsordirectcontactsbetweentheartisansmusthaveexisted.Eitherwaythestrongresemblancebetween
thecombsfromavarietyofplacessuggeststhat
the combmakers were ambulating artisans. In
contemporary Trondheim (Flodin 1989, Figure11and12)andKonghelle(Rytter1997,91)
wastefromcombmakingshowsthesamedistributionpatternasthatinBergen.Thismaysupportthenotionthatsomecombmakerstravelled.
184
In contrast, the material from contemporary
Lund seems to suggest that combmakers here
were more permanently settled artisans (Christophersen1980,126ff).
Altogether the distribution pattern for production waste in Bergen and other contemporary towns in Norway and ‘twins’ among the
finished products in Bergen as well as in other
townsstronglysuggestthatthecombmakersrepresentedinhorizon5inBergenwereprofessional
ambulatingartisanswhotravelledfromplaceto
place,workedhereforalimitedperiodoftime
andthenwenton.WhenworkinginBergenthey
mainlyservedaninterurbanmarket.Theremust
havebeenmoretravellingartisansorworkshops
working in the Scandinavian area at any given
time,theymusthavebeeninspiredbyeachother’sworkandasaconsequencemanyvariations
overthesamebasiccombtypesweredeveloped.
11CraftsandproductioninearlyBergen
Apairofcompasseswasnecessarywhendecorating the ornamented gaming pieces and may
have required specialist skills and tools. The
decorationsonthetwoornamentedneedles/pins
musthavebeenincisedwithaknifeoranother
sharppointedtoolbycarversthatwerefamiliar
withthelooksoffashionabledressorhairaccessories(Figure47).Theideasbehindtheneedles/
pinsfromBergenarealsorecognisedinneedles/
pins from Trondheim, although the latter are
found, in older contexts (Christophersen 1987,
Figurep73).Boththedecoratedgamingpieces
fromhorizons4and5andtheneedles/pinsfrom
horizon5may,therefore,havebeenproducedby
professionalspecialists.
Horizon4(c1100-1120s)
6/B
6/C
26/A
1*
1
Horizon5(1120s-c1170)
6/B
(2)
(1)
6/C
1
(1)*
6/D
3
(1)* (2)
(1)
(2)
6/E
(2)
6/G
26/A
2627/B
27/C
28/B
30/B
(1)
Unspecifiedtool
Spindlewhorl
Runicinscription
Rowlock
Needle/pin
Linerunner
Plot
Ice-skate
Table60.Horizons4and5productsofantler,bone,horn
andwhale/walrusbone(N=45)
Gamingpiece
Miscellaneousantler,bone,hornandwhale/
walrusboneworking
Finishedproductsofmiscellaneousantler,bone,
hornandwhale/walrusboneworking,foundin
theBergenmaterialareallrepresentedamongthe
blanks except the spindle whorls. This link betweenthefinishedproductsandproductionwaste
makes it likely that finished products found in
Bergenwereorcouldhavebeenmadehere.The
activity of miscellaneous antler, bone, horn and
whale/walrusboneworkingwasdocumentedin
fourofthesevenartefact-yieldingunitsinhorizon
4 and in 12 of the 24 artefact-yielding units in
horizon5;thepresenceoftheactivityinhorizons
4and5isconsideredreliable.
Theamountofwasteretrievedateachplaceof
productionspannedfromtwotoninefragments
inhorizon4andoneto68fragmentsinhorizon
5(Table31andTable32).Thewasteassemblages
arecharacterisedassmall,astheyfallwithinthe
quantitycategoriescharacterisedassmallinstudies from contemporary Scandinavian towns (eg
Christophersen1980,126ff;Flodin1989;Rytter
1997). The distribution pattern and amount of
waste may therefore be interpreted as either the
result of household producers and/or part-time
residentproducersand/orambulatingproducers.
The sparse amount of waste implies production
foraninterurbanmarket.
The finished products can be divided into
those that could be produced with the use of
household tools like a knife and no specialised
skillsorknowledgeandthosethatrequiredsomewhatmorespecialisedskills,toolsandknowledge.
Providing that the raw materials were available
and based on the level of skills, knowledge and
specialisationoftoolsrequired,thefirstgroupof
findsmayhavebeenproducedashouseholdproduction.Thesecondgroupismorelikelytohave
beenproducedbyskilledprofessionals.
Table60showsantler,bone,hornandwhale/
walrusboneitemsfoundinhorizons4and5,they
allstemfrombasicsources,andtheirpresenceis
considered reliable. All the items, except ornamentedgamingpiecesandtwoornamentedneedles/pins,mayrepresenthouseholdproduction.It
isthuslikelythatmiscellaneousantler,bone,horn
andwhale/walrusboneworkingwascarriedout
onahouseholdbasisduringhorizons4and5.
(1)
(1)
1
(1)
(1)
1
(1)*
(1)
(1)
(1)
(1)*
1
(1)
(1)
(1)
(1)
(2)
(1)
(1)*
(1)
(1)
(1)
(1)
(1)
(1)
ArtefactcategoryIIinbrackets
Numbersinboldthelatterbasicsources,inplain
supplementarysources
*Ornamentedgamingpiecesandneedles/pins
ItisnotpossibletoascertainwhethertheprofessionallyproduceditemsweremadeinBergen
by part-time resident or by ambulating specialists,soIwillleavethequestionopen.
185
Inconclusion,basedonthedistributionpatternforwasteandthecharacteroffinishedproductsitislikelythathouseholdproducerscarried
outmiscellaneousantler,bone,hornandwhale/
walrus bone working during horizons 4 and 5.
Professional artisans also probably worked in
Bergen during both horizons 4 and 5. In horizon4itcannotbeestablishedwhethertheprofessionalswerepart-timeresidentsorambulated.
Productionmustmainlyhaveservedaninterurbanmarketbothduringhorizons4and5.
Shoemaking
Leather waste (type 3) cannot be linked so directly to the finished products, as was the case
forthecombblanks.Still,toolsfromhorizon5
andleatherwasteoftype3frombothhorizons4
and5reliablyreflecttheproductionofshoesin
thetownarea.Itisthereforelikelythatatleast
someoftheshoesfoundwereorcouldhavebeen
producedandpurchasedinBergen.
Weretheshoemakersprofessional?Atotalof
1082shoes,thatissolesoruppers,werefoundin
horizons4-5,theymostlystemfrombasicsources
andthemainpatternoftheirspatialandchronologicaldistributionisconsideredreliable.Ofthe
656uppers,217weredecoratedwithembroider-
d
a
c
b
e
Figure47.Ornamentedgamingpiecesandneedles/pins:a,b,cgamingpieces;dBRM0/53003;eBRM0/81009
186
Figure48.TwinshoesfromBergen
11CraftsandproductioninearlyBergen
187
Figure49.EmbroiderypatternsC5andG2fromTrondheimandOslo.(PublishedinSchia1977,Figure44;Schia1987,
Figure22;Marstein1989;Smedstad1991,Figure32)
iesvaryingfromasinglelinetomoreelaborate
patterns,theembroideryontheupperswasclassified according to Larsen 1992 (Larsen 1992,
Plate 1). Before the upper was decorated, the
embroidery pattern was scored onto the leather
surfacewithaknife.Theindividualshoemakers
scoredtheleatherwithapersonal‘touch’.Aswith
thecombs,theembroideriescanbeclassifiedin
188
severaltypesandwithineachtypetherearevariations.Ihavestudiedthepatternandthe‘touch’
andthespatialdistributionofupperswithelaborateembroiderypatternsoftypeA2,C5andG2
inordertoidentifythe‘touch’ofconcreteartisansorworkshopsbehindtheseshoes.64
Threeexamplesof‘twin’shoes(fromdifferentpairsofcourse)wereidentifiedinthemateri-
alassignedtohorizon5,theembroiderypatterns
ontheseshoesandthe‘touch’oftheshoemaker
aresosimilarthatthesameshoemakerorworkshop must be seen behind the products. The
twin shoes were found on different plots. This
indicates that shoemakers sold shoes to people
outside their own household and were professional(Figure48).Furthermore,theproduction
ofshoesfoundinbothhorizons4and5required
specialisedtools,theyhadahighquality65andthe
shoetypesaswellastheembroiderytypesrepresentedintheBergenmaterialhaveclearparallels
incontemporaryOsloandTrondheim(compare
Schia1977;Schia1987b;Marstein1989;Larsen
1992).ThisshowsthattheBergenshoemakersin
bothhorizons4and5knewhowanup-to-date
shoeshouldbecutandinmanycasesdecorated,
andthattheproductswerestandardised.These
factorssuggestthattheshoesweremadeprofessionallyinbothhorizons4and5.
The assemblages of production waste from
shoemaking varied from two to 20 shreds of
leatherwasteinhorizon4andoneto276shreds
inhorizon5(Table34andTable35).Theyfall
within the quantity category characterised as
small in contemporary Oslo (Tørhaug 1998,
51).Wastefromshoemakinghasadistribution
pattern similar to that of combmaking: small
amountsofwasteinseveralanalyticunits.This
indicatesthattheshoemakersdidnotworkona
permanent basis or as full-time shoemakers on
theproductionplacesinearlyBergen.Italsoindicatesthattheyproducedmainlyforaninterurbanmarket.
Were the shoemakers itinerant artisans like
thecombmakers?IftheBergenembroiderypatternshavetwinsinothercollections,thiswould
present a convincing argument for such a suggestion. Unfortunately, embroidery on shoes is
asubjectthathasnotbeenwidelystudied.Embroideriesarecommoninthepublishedmaterial
fromcontemporaryOslo(Schia1975,189Figure123),theyseemtobecommonincontemporary Trondheim as well (cf Marstein 1989).
Publishedillustrationsoftheembroideredshoes
arescarce,however.Figure49showsexamplesof
embroideries of Larsen’s type C5 and G2 from
contemporaryOsloandTrondheim.Whencomparingwiththeembroideriesfromhorizon5on
11CraftsandproductioninearlyBergen
Figure48itappearsthatthesamevariationsover
embroiderytypesC5andG2wereappliedinthe
three towns. The illustrations from Trondheim
and Oslo are, however, somewhat standardised
andthevariationsoftheparticularC5andG2
themesarenotsospecialisedorcomplicatedthat
theycanbetakenasthe‘fingerprint’ofoneartisan or workshop. In order to identify ‘twins’,
theembroiderieswillhavetobestudiedinmore
detail,whichisbeyondthescopeofthisstudy.
Thusithasnotbeenpossibletoidentifypositive
twinshoesinothercollectionsinScandinavia.A
concludingargumentforambulatingshoemakers
cannotbeproduced.Still,onecannotignorethe
factthatthesimilaritiesbetweentheembroideriesfromhorizon5inBergenandcontemporary
Trondheim and Oslo are striking. Waste from
shoemakersincontemporaryOsloalsoshowsthe
samepatternofdistributionasthatoftheearly
Bergen material (Tørhaug 1998, 94-95), supportingthenotionthatsomeshoemakersambulated.Altogether,inthelightofthedistribution
ofproductionwasteinBergenandOsloandthe
close paralells between embroideries in Bergen,
Trondheim and Oslo, I find it very likely that
shoemakersofBergeninhorizon5wereorganisedinawaysimilartothatofthecombmakers:
professionalambulatingartisansthatworkedin
alargeregion,producingitemsmainlyforaninterurbanmarket.Astrongcasecannotbemade
fortheorganisationofshoemakersinhorizon4;
according to the distribution pattern for waste
foundinBergentheymayhavebeeneitherresident part-time artisans or ambulating artisans,
andthequestionremainsopen.
‘Otherleatherwork’
Wastefrom‘otherleatherwork’cannotbelinked
directly to the finished products but it is likely
thatatleastsomeoftheproductsof‘otherleatherwork’weremadeinBergen,sincetheactivity
is well-documented during horizon 5. ‘Other
leatherwork’wasdocumentedon10plots/sitesin
horizon5,thewasteassemblagesspannedfrom
oneto81shredsofleatherwaste(Table35)and
must be characterised as small (Tørhaug 1998,
51).Thequantityanddistributionofproduction
waste indicate that the production was carried
out by household producers, professional part189
Figure50.Examplesofcrudeandfine‘otherleatherwork’:aBRM0/85396/01,child’sshoemadefromagrownup’sshoewith
embroidery;bBRM0/45983/01,knife-sheathmadefromashoewithembroidery
190
timeorambulatingartisans,andtheproduction
mustmainlyhaveservedaninterurbanmarket.
Theidentifiedproductsofotherleatherwork
are all from horizon 5, they stem from basic
sources and the main pattern of their distributionintimeandspaceisconsideredreliable.The
productscompriseshoesthathavebeenrepaired
eitherwiththread(3finds)orwithleatherstraps
andsimilartechniques(24finds),achild’sshoe
madeoutofagrownup’sembroideredshoeand
aknife-sheathmadefromanembroideredshoeupper(Figure50).Alltheseproductsmayhave
been made in Bergen. The question is whether
theyweremadeprofessionallyornot.
The products can be divided into fine and
crudework,thefirstcategorybeingcharacterised
bytheuseofthread,fineneedlesandprobably
alast,thelatterbytheuseofleatherstrapsand
a pointed instrument for piercing holes in the
leather.Judgedbythetoolsandskillsinvolved,
professional leatherworkers should probably be
seen behind the finer work, whereas household
producers-thetownspeople,maybeseenbehind
theotherproducts.Theshoesthatwererepaired
withthreadincludingtheaforementionedchild’s
shoeareexamplesofthecategoryofprofessional
work.Shoesrepairedwithleatherstrapsandthe
sheaththatwassewntogetherwithstrapsareexamplesofthegroupofitemsmadeonahouseholdbasis.Itseemsthatbothprofessionalsand
household producers carried out ‘other leatherworking’duringhorizon5.
Wastefrom‘otherleatherwork’isoftenfound
inthesamebagofcategoryIfinds,orinthesame
categoryIlayerasthatofwastefromshoemaking
(egbags45534,45544,45593allfromplot6/D
and85400,85457fromplot6/G).Thismayindicate that the distinction between shoemakers
andcobblersmaynotyethavebeenintroduced
(cf Larsen 1992, 88), and that the professional
andprobablyambulatingshoemakerscarriedout
thefinerrepairworkduringhorizon5.
Inconclusion,inthelightofthedistribution
pattern for waste, the character of the finished
products from horizon 5 and the frequent colocation of waste from shoemaking and ‘other
leatherwork’ during this horizon, it seems that
‘other leatherwork’ may have been carried out
bothbyhouseholdproducersandbyprofessional
11CraftsandproductioninearlyBergen
artisans. Some of the professional artisans may
havebeenidenticaltotheambulatingshoemakers.
Metalworking
Metalworking,studiedinitiallyassmithingand
casting,hasleftreliabletracesinhorizons4and
5. First the nature and organisation of casting
willbediscussed.Therearefewitemsmadein
copper alloy and no items of fine metals have
beenfound.Furthermore,theproductsofcasting are not so well preserved and have been
treatedbyvariousmethodsduringconservation.
A visual comparison between items has generallynotbeenfruitfulandithasnotbeenpossible
tomakeadirectlinkbetweentheartefactsthat
indicate production and the finished products.
Sincewehavetracesofproduction(Table36and
Table37)itislikelythatatleastsomeofthecasts
foundinBergenwerealsomadehere.
Tracesoftheactivityofcastingwerefoundin
horizons 4 and 5. The find assemblages varied
from one to three fragments in horizon 4 and
onetofivefragmentsinhorizon5(Table36and
Table37),andmustbecharacterisedassmall(cf
Bergquist1989).Castinghasleftawastedistributionpatternsimilartothatofcombandshoe
production:smallamountsofwastescatteredon
several plots. The distribution pattern implies
thatcastingwascarriedouteitherbyhousehold
producers,byresidentpart-timeprofessionalsor
by ambulating artisans. The activity of casting
requiresspecialistskillsandknowledge,thismay
initselfimplythatthecastingsmithsofhorizons
4 and 5 were professional. The producers may
thushavebeeneitherresidentpart-timeprofessionals or ambulating artisans. The production
may chiefly have served an interurban market
duringhorizons4and5.
Slagandclayfromafurnacelining,indicating smithing has a distribution pattern similar
tothatofcastingandtheothertradesdiscussed
above;inhorizon4afragmentofafurnaceliningwasretrievedandinhorizon5thenumberof
slagpiecesoneachplot/sitevariesbetweenone
and three specimens per analytic unit. In spite
of the problems of representativity attached to
the artefact groups, the number of fragments
associatedwithsmithingmustbeconsideredas
191
Figure51.TwinkeysfromBergenandTrondheim.Bergen:aBRM0/72983;Trondheim:bN10579/S139(Christophersen
1987,Photop87)
192
small(cfBergquist1989).Thedistributionpatternimpliesthatsmithingwascarriedouteither
byhouseholdproducersorbyresidentpart-time
professionals or ambulating artisans. Like casting, the activity of smithing requires specialist
skillsandknowledge,thismayinitselfimplythat
the smithing artisans of horizons 4 and 5 were
professional. Furthermore three keys for barrel
lockswerefoundinhorizon5.Thekeysaremade
from iron with a string of copper alloy twisted
around the key as ornament.66 The three keys
presentvariationsofthesamekeytype.Theartisansbehindthesekeyswereup-to-dateonthe
looksofcontemporarykeys,asstrongparallelsto
theBergenkeysfoundinforinstanceLund,Copenhagen, Novgorod, York, Trondheim and in
Greenlandindicate(cfMårtensson1976,Figure
358; Christophersen 1987, Photo p 87; Roesdal
1993;Berglund2001,269).Oneofthekeysfrom
Bergenhasa‘twin’inTrondheim(Figure51).As
the Bergen and Trondheim keys were found in
two different towns the artisan who made the
keysmusthavesoldarticlesoutsidehishousehold.
Altogether this suggests that the metal workers
behindforinstancethesekeyswereprofessional.
Someoftheproducersmaythus,whenalsoconsidering the distribution pattern for waste, have
been either resident part-time professionals or
ambulatingartisans.Theproductionmaychiefly
haveservedaninterurbanmarket.
IfthetwinkeysweremadebythesameworkshopinBergenandTrondheimrespectively,they
weremostlikelyproducedbyanambulatingartisanorworkshop.Thissuggeststhatsomeofthe
smithsambulated.IncontemporaryTrondheim
thedistributionpatternfortheactivityofcasting
andsmithingissimilartothatofhorizons4and
5 in Bergen (cf Bergquist 1989, 121). The evidence of smithing in contemporary Trondheim
supportsthattheTrondheimkeymayhavebeen
made locally, and the distribution pattern for
wastefromTrondheimsupportsthatsomesmiths
ambulated. I suggest that some of the smithing
artisansworkinginBergenduringhorizon5were
professionalambulatingsmiths.Thissuggestion
findssupportinpatternsdiscernedinthewaste
materialfromBergenandcontemporaryTrondheimandOsloandinthepresenceofthetwin
keysinBergenandTrondheimduringhorizon5.
11CraftsandproductioninearlyBergen
Intermsoftheorganisationofthetradeduring
horizon4,thatiswhethertheartisanswereparttime resident or ambulating artisans, cannot be
determined,sothequestionisleftopen.
Aninterestingdetailinthematerialisthatin
horizon4,thefurnaceliningfragmentwasfound
alongwithoffcutsofcopperalloy.Inhorizon5
all but one occurrence of slag was found along
withcruciblesandoffcutsofcopperalloy.Perhaps
thesefindsshouldbeassociatedwiththeworkof
thecastingsmithsandviceversa?Perhapssomeof
thesmithswerenotsospecialisedbutknewhow
tohandlebothmethods?Itishardtodecideon
thescarcematerialavailable,andnoconclusions
canbereachedonthisquestionhere.
Inonecaseslagappearsalonewithoutanyindicationofcasting,thisisinunit8/Binbuilding
158.Asdiscussedearlier,thisbuildingmaywell
havebeenoneof‘thesmith’sbooths’mentioned
inHeimskringla(cfp214ff).Ifalocalitycalled
‘thesmith’sbooths’existedin1155onewouldexpectthistobepermanentworkshopsforprofessional smiths. Of course, it cannot be excluded
thatambulatingsmithsreturnedtospecialworkshops and worked there for some time. On the
other hand, many products of smithing such
as nails, rivets and various tools seen in the archaeologicalmaterialfromhorizons4and5,were
goodsneededintheeverydayhouseholdandone
wouldexpectthatthedemandfortheseproducts
waslargeenoughforastationarysmithtoreside
hereandsupplytheurbancommunity.
To conclude, there are many uncertainties
associated with the archaeological evidence of
metalworking in early Bergen. It cannot be ascertainedwhetherthemetalworkersofhorizon4
werepart-timeresidentorambulatingartisans,so
thisquestionisleftopen.
Thecontoursofdifferentcategoriesofsmiths
have emerged. In horizon 5, some smiths may
have been stationary and perhaps supplied the
townwitheverydaycommodities.Someprobably
ambulated and supplied a large area with more
rare products, such as the twin keys of Bergen
andTrondheim.Thekeyscanhardlybecharacterised as luxury items, as close parallels to the
Bergenkeysarerathercommonandthekeysare
notmadeofespeciallyrareorpreciousmetals.
193
Stoneworking
Large-scalestoneworkingwasindicatedthrough
thechurchesandotherinstitutionsinitiatedfrom
horizon3throughhorizon5.Thecraftsmenbehindthestone-builtmonumentsmusthavesold
their expertise and labour and were professionalsinabroadsenseoftheword.Thecraftsmen
may perhaps have been organised in Bauhütten
or‘lodges’-workshopsofskilledcraftsmen,this
ishowtheywereorganisedinEurope(Lidénand
Magerøy1990,73;Ekroll1997,112).Therewas
nolocalorNorwegiantraditionforbuildingin
stone when the first stone monument was initiatedinBergenbyKingOlavKyrreabout1070.
LidénfindsitlikelythatthefirstlodgeofcraftsmenthatworkedinBergenintheperiodrepresentedbyhorizon3wasmadeupbyforeignersor
Norwegiansthathadlearnedtheircraftabroad.
JudgedbyAnglo-Normanmouldingonpilasters
inthestandingChurchofStMary,theearlylodge
wasinspiredbyAnglo-Normanarchitecture.The
earlylodgeworkedattheChristchurchCathedral
atHolmenandmayhaveformedthepilastersof
StMary’s(LidénandMagerøy1990,73).
Inthebeginningofthetwelfthcentury,represented by horizons 4 and 5, a new lodge arrivedinBergen.Theworkofthislodgeisseenin
allthetwelfthcenturychurchesinBergen.This
lodgehadaClassical/Lombardbackground.Art
historianshavediscussedwhetherthecraftsmen
cametoBergenviathecathedralinLundorvia
thecathedralinSpeyerintheRhineland.Lidén
findsitmostlikelythatthecraftsmencamefrom
Lund around 1120 (for a detailed discussion
see Lidén 1990, 73-87). About the middle of
thetwelfthcenturythechurchbuildersofBergenwereagaininspiredbytheEnglish,perhaps
through English lay brothers who, at that time
workedattheCistercianmonasteryofLyseclose
toBergen.Lidénsuggeststhatthecraftsmenof
the‘Classical/Lombardlodge’nowtriedoutnew
formsandcombinedthemwiththeearlierones
(LidénandMagerøy1990,88).
Theprojectofbuildingachurchwasalengthy
one,theChristchurchCathedralwasunderconstructionforperhapsaslongacentury(cfp80)
and for instance the standing St Mary’s (initiated during horizon 5) must have been under
constructionforseveraldecadesastheworkwas
194
begunbyClassical/Lombardinspiredcraftsmen
but finished under English/Classical/Lombard
inspiration. Craftsmen on many levels were required,perhapsthemainarchitectsambulated,
but most of the lower level workers must have
lived in Bergen. Anyhow, we may assume that
the many monumental building sites of early
Bergen provided full-time work for many residentstoneworkers.
The mortar or plaster from the lime-slaking
pitsonplot6/Cmusthavebeenproducedtobe
usedoutsidetheboundariesoftheplot,because
no structures using mortar or plaster have been
recordedinhorizon5norinthefollowingphase
ontheplot.Perhapsthelimewasmeantforthe
maintenanceofthenearbychurchofStMaryor
forotherstonebuildings?Distributinglime,the
limeproducersonplot6/Cinhorizon5wouldbe
professionalsandthenprobablystationaryones.
In conclusion, professional large-scale stoneworkers were probably present in Bergen from
horizon3throughhorizon5inconnectionwith
themonumentalbuildingsites.Inadditionprofessionallarge-scalestoneworkers,whoproduced
lime may have been present during horizon 5.
Given the limited size of the production, they
probably produced for an interurban market
only.
Small-scale stoneworking was reliably documentedthroughsmallamountsofoffcutofsteatiteandafewblanksscatteredonseveralplotsin
horizons4and5.Aspointedoutthedocumented areas for small-scale stoneworking are probably not representative for the real number of
productionareas,anditcannotbeexcludedthat
other types of stone than steatite was worked.
The waste and blanks may indicate household
production,productionbypart-timeresidentor
ambulating artisans. The production may have
beenaimedataninterurbanmarketonly.
When leaving out the stone items that were
surelybroughtintoBergenasfinishedproducts
or blanks (hones, baking slabs, steatite vessels,
and grinding stones), only a small number of
other stone products are left, comprising steatiteproducts:mouldsforcasting,fishingtackle,
warp-weightsandspindlewhorls.Spindlewhorls
and fishing tackle in other types of stone and
slate discuses are also found in horizons 4 and
5.Theproductscanbedividedintotwogroups.
Inthefirstgroup,itemsmayhavebeenproduced
usingordinaryhouseholdtoolslikeaknifeonly,
and the finish of the objects was rather crude.
Inthesecondgroup,itemswereproducedusing
more specialised tools such as a lathe, and the
finishoftheobjectswasfine.
Steatitewarp weights, fishingtackle and not
least the steatite spindle whorls may belong to
the first group. The spindle whorl blank from
plot6/Dinhorizon5wasofsteatite(Table41).
Steatitepiecesmusthavebeeneasytocomeby
at the monumental building sites and in some
casessteatitevesselshavebeenreusedasrawmaterial for weights, all these factors suggest that
the products were made locally by household
producersduringhorizons4and5.
Thirteenspindlewhorlsturnedinserpentine
may belong to the second group of finds (Figure 52). Several factors suggest that they were
made professionally. Geological analysis of five
of the whorls show that the whorls have a uniformchemicalcomposition.67Itisthusnotunlikelythattheyoriginatefromthesamequarry.
Thewhorlsvarysomewhatinshapeandsize,but
all are neatly finished compared to the steatite
whorls.Furthermore,therawmaterialissohard
thatspecialtoolsmusthavebeenrequiredwhen
working it. The presence of one whorl made
from pottery68 and one cast in metal69 shows
thatspindlewhorlswereanarticlethatwouldbe
boughtifmadeforsale.The13serpentine/diabasewhorlswerefoundonsixdifferentplotsin
horizon5(6/B,6/C,6/D,20/A,26/Aand28/
B).Iftheyweremadebythesameartisan/workshop,thedistributionofthewhorlsondifferent
plots suggest that professional craftsmen made
them. Another indication that they were made
professionallyisthefactthattenmorewhorlsof
thesamestonewerefoundinyoungercontexts,
implyingproductionovertime.
It is not possible, to determine whether the
Figure52.Turnedspindlewhorls.BRM0/45060,BRM0/45222,BRM0/45847,BRM0/54529,BRM0/63860,BRM
0/64396,BRM0/64557,BRM0/64558,BRM0/65017,BRM0/73103,BRM76/10967,andBRM94/1066,BRM
104/2261
11CraftsandproductioninearlyBergen
195
whorls were made in Bergen or imported into
Bergen,aswecannotlinktheproducttoaplace
ofproduction.Giventhatmanyotherstoneitems
were imported it is likely that the whorls were
alsoimportedfromsomewhereoutsideBergen.
Altogether,thesourcesimplythatsmall-scale
stoneworkingwascarriedoutonahouseholdbasisduringhorizons4and5.Therewasnoindicationthatprofessionalsmall-scalestoneworking
wascarriedoutinearlyBergen.
Woodworking
Large-scale woodworking is represented in the
ordinaryhouseconstructionsandinthemonumental building projects initiated during horizons3,4and5.Buildinginwoodwastradition
inNorwayandthebuildingsonthetownplots
wereprobablybuiltbythosewhoweregoingto
use them in a broad sense. As with the largescale stonework, the craftsmen who built the
monumental timber buildings must have sold
their expertise and labour and were in this respectprofessionals.Thecraftsmenmayormay
nothavebeenorganisedinlodges.Itispossible
thatthecraftsmenwerenativecraftsmenrather
thanforeigners,giventhatwoodwasthetraditional material for buildings in Norway. Since
wedonotknowforhowlongthemonumental
timber buildings were under construction, the
material is too scarce to give a qualified opiniononwhetherthecraftsmenwerestationaryor
ambulatingprofessionalsandthequestionisleft
open.
Inconclusion,professionallarge-scalewoodworkersweremostlikelypresentinBergenfrom
horizon 3 through horizon 5 in connection
withthemonumentalbuildingsites.Howthese
workerswereorganisedisuncertain.
Small-scalewoodworkingisrepresentedinthe
finished products and was well-documented in
horizon5throughtoolsandproductionwaste.
Asmentionedearlier,thewastefromsmall-scale
woodworking is most likely underrepresented.
Wehavenoindicationoftherealdistributionof
productionareasandtherealextentofproduction;thedistributionpatterngivenbythesourcescannotbeusedasanindicatorofthenature
oftheproduction.
Lathe-turnedcoresindicatetheproductionof
196
lathe-turnedvesselsfoundonseveralplots.The
vessels are, however, rather uniform with no
specialexternalcharacteristicstodiagnose‘twin
products’.Turningrequiresalathe,whetheror
notthiswasstandardequipmentinahousehold
in twelfth century Norway is hard to say. According to the Urban Code of 1276, turners
were considered a separate group of craftsmen
(Bl 1923 8,1) (KLNM, XVII 470ff). Whether
thisappliestothetwelfthcenturyaswellis,impossibletodetermine.Aneedleblank,theonly
identified blank among the finds, cannot form
the basis of a discussion. This blank may have
been cut for use within the producer’s householdorforsale.Thefinishedproductsinwood
comprise a multitude of items spanning from
thesimplestsausagepintoornamenteditemsof
varying sizes and functions. Most of the items
(excluding the turned vessels) may have been
producedusinghouseholdtoolslikeaknife,and
requirednospecialskills.Theymaythusrepresenthouseholdproduction,anditislikelythat
small-scale woodworking was carried out on a
householdbasisduringhorizon5.
A group of items is distinguished by being
ornamented and demands further discussion.
On the basis of carvings on wood Signe Horn
Fuglesanghasarguedforaworkshopofprofessional woodcarvers in eleventh century Trondheim(Fuglesang1981;Fuglesang1984).Ihave
studiedcarvingsonwoodfromearlyBergen,excludinggamingpiecesinanattempttoidentify
asimilarworkshopinBergen.Theornamented
pieces of wood (Figure 53) are all assigned to
horizon 5 and derive from basic sources. The
17findsmakeupaheterogeneousgroup,comprisingthreespoons–noneofwhichhavebeen
retrieved in the museum storerooms but two
wereidentifiedthroughdrawings70,onelidfor
adrinkingvessel71,twopossibleplugswithanimal heads,72 four lids for containers73 and one
undefinedutensil.74Thefunctionoftheremainingsix75objectsismoreuncertain.Alltheornaments are unique and none stand out with especiallyhighartisticqualities.Thecarvingsare
cutorincisedwithaknife-whichwassurelya
standard tool in every household. Hence there
isnothingintheavailablematerialthatpoints
towardsaworkshopofprofessionalwoodcarvers
Figure53.Ornamenteditemsinwood
11CraftsandproductioninearlyBergen
197
Figure53b.Ornamenteditemsinwood
198
Figure53c.Ornamenteditemsinwood
in Bergen. Rather the carvings may have been
madeonahouseholdbasis.
To sum up, most of the small-scale woodworking represented in the finds from horizon
5mayhavebeenmadeonahouseholdbasis,the
turnedvesselsmaybeanexceptiontothis,but
thereisnoconcreteevidenceoflocalprofessional
productionintheperiodstudiedhere.
Skinning
Inthetownregulationfrom1282(NgLIII,14)
skinners are mentioned as a separate group of
craftsmen.Theavailablesourcesforskinningin
earlyBergenare,however,tooscarcetobeused
asasourceforthenatureandorganisationofthe
tradeforthisperiod.
Textileproduction
Whendiscussingthenatureandorganisationof
11CraftsandproductioninearlyBergen
textile production in early Bergen tools are the
onlysourcestudied.Textiletoolswerefoundon
nineplotsinhorizon5andthepresenceoftextile
productioninthishorizonisconsideredwellestablished.Theidentificationofprofessionalversusamateurtextileproducershasbeenattempted
by Øye (1988) and Gjøl Hagen ((1988) 1994)
onmaterialfromBergenandTrondheimrespectively.GjølHagenhasasapremiseforherstudies that the upright loom, represented by warp
weightsinthearchaeologicalmaterial,combined
withtextileswovenintwo-shafttechniqueand
alowlevelofstandardisationreflectsproduction
forhouseholdconsumption.Asopposedtothis
the horizontal loom combined with three-shaft
textiles with a high degree of standardisation
would reflect the production of textiles for sale
(Hagen(1988)1994,73-99).Øyealsoassociates
theuprightloomwithproductionoftextilesfor
199
householduse,thehorizontalloomwithprofessionalproduction(Øye1988,131).
AsIhavenotstudiedthefragmentsoftextile
inthematerialfromearlyBergen,theycannotbe
drawnuponasasourcehere.Ifthereisadirect
connection between the upright loom and productionoftextilesforhouseholduse,thepresence
ofwarp-weightsonlyandnohorizontalloomsin
horizon5reflectthattextileswereproducedfor
householduseonly.Adirectonetooneconnection between the upright loom and production
oftextilesforhouseholdusecannotbetakenfor
granted as Icelandic frieze, woven on upright
looms,wasproducedforsaleandexportatleast
from the thirteenth century onwards (KLNM,
XIX 409; Hagen (1988) 1994, 98). Based on
thepresenceofwarp-weightsaloneitcannotbe
determinedwhethertheproductionoftextilesin
Bergenwasforhouseholduseonlyorforsaleas
well.
InconclusiontheinvestigatedsourcesfortextileproductioninearlyBergencannotelucidate
thenatureandorganisationofthisproduction.
Fishing,huntingand,farming
Theactivitiesoffishing,hunting,76andfarming
are‘primaryactivities’,whoseproductscannotbe
tracedbyarchaeologicalmethodsaloneandthe
natureandorganisationoftheproductioncannot
begraspedonthesparseosteologicalandbotanicalevidenceavailable.
The tools for fishing assigned to horizons
4 and 5 and those for hunting, assigned to horizon 5 were most likely part of the ordinary
household equipment in contemporary coastal
Norway.Olsenpointsoutthatfishingdemands
insightinhowtouseaboat,howtohandlethe
fishing tackle, and how to ‘read’ the weather,
knowledgeoflocalfishinggroundsisalsoanasset(Olsen1998,121).Onemightarguealonga
similar line of thinking in terms of agriculture
andhunting.Thiskindofknowledgeisperhaps
considered somewhat specialised today, it must
however, have been common in early medieval
Norway.Onecannotargueonthisbasiswhether
ornotfull-timespecialistscarriedoutfishing,agriculture or hunting. On a common-sense basis
onemayarguethatifonewasafull-timefisher,
hunteroragriculturist,Bergen,beingatownwas
200
probablynottheoptimalplacetosettlein.Hence
whenfishing,inhorizons4and5andhuntingin
horizon 5 are recorded, the activities were most
likelyjustoneofseveralstrategiesofthehouseholdeconomyinanearlyurbancontext.
Basiccookingandfoodandbeverage
processing
Theproductsofbasiccookingandfoodandbeverage processing have not been investigated archaeologically.Basiccookingwasrepresentedby
findsonrespectively6of7and17of24ofthe
artefact-yielding plots/sites in horizons 4 and 5
andtheactivitywasconsideredwell-documented
in both horizons. Basic cooking definitely involvedtheuseofordinaryhouseholdequipment
and common knowledge. It is therefore likely
thatthisactivitywascarriedoutonahousehold
basis.Alsothevariousformsoffoodprocessing;
theirpresencebeingwellestablishedthroughthe
sources for horizons 4 and 5, were most likely
carried out with the use of ordinary household
equipmentandcommonknowledge,sotheywere
probablyalsocarriedoutonahouseholdbasis.
Beer brewing was possibly reflected in the
sourcesforhorizon5.Thetoolsandknowledge
involvedinbeerbrewingwereprobablyalsointegratedinanordinaryhousehold,andtheactivity
mayhavebeencarriedoutashouseholdproduction.Still,thirteenthcenturysourcestellofsaleof
beer,thusimplyingthatbeerwasthenbrewedby
professionalsinadditiontobeingbrewedforthe
household (KLNM, VI 224). May such professionalbrewingbefoundatanearlierstagealso?
WhenRagnvaldKalefromAgdervisitedBergenbetween1115and1120(horizon4)hedrank
andsleptinUnn’stenement,whichfromthedescription,musthavebeenaninn(Orkn1913-16;
Holtsmark 1970, 92-94; Helle 1982, 114). And
when King Sigurd Munn (Sigurth Haraldsson)
was killed in 1155, corresponding to horizon 5,
hewasalsoinatenementdrinking.SigridSæta
wasthehostessofthistenement(Hkr1911,591;
HoltsmarkandSeip1975,679).Thesepassages
implythatinnswherevisitorscouldbuybeverages,presumablybeer,andlodgingwerearealityin
Bergenperhapsalreadyduringhorizons4and5.
Hellefindsitlikelythatthedescriptionofthe
town in 1115-1120 is anachronistic, describing
Bergenattheendofthetwelfthcenturyorsomewhatlater(Helle1982,114).Ifinditreasonable
to assumethatinnswerearealityinBergenat
leastasearlyashorizon5(the1120stoc1170).
Because if we accept that various ambulating
artisansvisitedthetownduringhorizon5,itit
likelythatthesevisitors,andprobablyalsoother
visitorsinstillincreasingnumbers,couldnotdependmerelyonthehospitalityofthetownspeople
(cfKLNM,V701ff).Visitorstothetownwould
needaccommodation,andasfortheartisans,a
placeforatemporaryworkshopaswell.Itisthus
likelythatatleasttheinnvisitedbyKingSigurd
Munninhorizon5wasarealitythisearly.
In addition, some people must have let out
suitablepremisesforatemporaryworkshop,duringhorizon5whenambulatingartisansvisited
Bergen.Thismaybeillustratedbybuildingsat
site6,wheremorethanonetypeofactivitycarriedoutbyprofessionalambulatingartisanswere
indicatedbyartefactsofcategoryI(Table61).
Combmaking
Antler,bone,hornand
whale/walrusboneworking
Shoemaking
Leatherworking
Metalworking
Table61.Buildingswiththepresenceofmorethanone
productiveactivitycarriedoutbyprofessionalambulating
artisans,indicatedbyartefactsofcategoryI.
X
X
X
X
X
X
X
BuildingsinHorizon5
(1120s-c1170)
6/40(plot6/C)
6/196(plot6/G)
The likely presence of inns and premises for
leasesuggestthatnewurbantradeshadbeenintroducedinBergenduringhorizon5,tradesconductedbypeoplewhoaspartoftheirstrategyfor
makingalivinginBergen,gaveservicetovisitors
ofthetown.
Thedistributionofsausagepinsinrelationto
toolsforbasiccookingshowsaninterestingpatternandmayalsoshedsomelightonthepresence
of service-related trades. As already mentioned,
the production of sausages was probably an activitycarriedoutonahouseholdbasis.Still,sausagepinswerenotfoundinnearlyasmanyana11CraftsandproductioninearlyBergen
lyticunitsinhorizon5,asweretoolsforordinary
cooking (Table 57). This may be an indication
thatsausagemakinginsomerespectswasaspecialisedactivity.
Eightpercentofallfindsassignedtoanalytic
unitsinhorizon5wereclassifiedassausagepins.
Thisshowsthattheartefactisverycommonand
indicatesthat,whenitisnotfoundinsomeunits
in horizon 5, this may have not only methodologicalbutculturehistoricalexplanationsaswell.
Asmentionedearlier,Ihaveasapremiseforthe
discussionthatthefindspotsforsausagepinsreflect the place where the sausages were made or
stored,notwheretheywereeaten.Ialsoholdasa
premisethat‘everybody’atesausages.
Arethereanyspecialfactorsthatcharacterise
theanalyticunitswheresausageswerenotmade
orstored?Figure54showsthatonseveralanalytic
unitsinhorizon5,wherebasiccookingtookplace
butno,orrelativelyfew,sausagepinswerefound,
the professional and probably ambulating artisans (combmakers, shoemakers and metalworkers)hadmadeavisit.Thisrelatestoanalyticunits
6/G,8/A,8/B,26/A,30/B,and30/E.Sausage
pinswerealsoabsentonplot27/C,butsomepins
werefoundbetweenplot27/Cand26-27/Band
itcannotbeexcludedthatthepinsstemfrom27/
C.Thisgeneralpatterninthematerialmayimply
thatambulatingartisansdidnotmaketheirown
sausages, but instead had to buy their sausages.
Thethoughtinitselfisnotunreasonable,consideringthemanyprocessesinvolvedwhenmaking
sausages(cfp235).Fromthisitfollowslogically
that some people must have made sausages for
sale,andthuswereprofessionalsausagemakers.
Sincesuchaconclusionisbasednotonlyon
thepresenceofartefacts,butalsoontheabsence,
aquantitativeevaluationofthematerialiscalled
for.Unfortunately,whenconsideringartefactassemblages from the plots with few sausage pins
but many artisans, only the artefact assemblage
fromplot6/G,horizon5,qualifiesforareliable
quantitative analysis (cf p 71ff). On plot 6/G
(horizon5)whereambulatingartisansofvarious
kindshadstayed,only0.55%ofthefindswere
classified as sausage pins, implying that sausage
makingwasnotarecurringactivityhere.Bearing
inmindthemethodologicalproblemsinherentin
thematerial,thedistributionpatternforsausage
201
Figure54.Horizon5(1120sc1170),sausagepins,basic
cookingtoolsandproduction
wastefromambulatingartisans:
combmakers,shoemakersand
metalworkersasapercentageof
thetotalnumberoffindsfrom
artefact-yieldinganalyticunits
202
pinsstillappearstoimplythatsausageswerenot
made on/in all plots/analytic units during horizon5.
If ‘everybody’ ate sausages, patterns in the
sources imply that sausage makers, besides producing for their own consumption, sold their
products to hungry visitors such as ambulating
artisans. Accordingly these sausage-making specialists must have been professional. And they
weremostlikelyresidentsofBergenrepresenting
anewurbanservice-relatedtrade.
Whethertheinnkeepers,peoplewithpremises for lease and the professional sausage makers
carriedouttheirservice-relatedtradesonafulltimebasiscannotbedeterminedontheavailable
sources. But their likely presence suggests that
newurbanservice-relatedtradeshadbecomepart
ofthetownspeople’seconomyduringhorizon5.
Asvisitorsmadeuseoftheservicesprovidedone
mayarguethattheservice-relatedtradesserveda
‘widermarket’asopposedtoaninterurbanmarket.
Summary
Table62sumsupthesuggestednatureandorganisationoftheproductiveactivitiesidentified
from horizon 3 through horizon 5. The nature
and organisation of skinning and textile production,carriedoutduringhorizon5couldnot
be established. And it could not be established
whetherlarge-scalewoodworkerswerefull-time
professionalresidentsorambulatingprofessionalsduringhorizons3to5.Neitherwasitpossible
tomakeastrongcaseforwhethertheactivitiesof
antler,bone,hornandwhale/walrusboneworking,andshoemakingduringhorizon4andantler,bone,hornandwhale/walrusboneworking
duringhorizon5werecarriedoutbypart-time
residentprofessionalsorambulatingartisans.
It seems likely that fishing, hunting, basic
cooking, small-scale stone and woodworking,
some sausage making, and some miscellaneous
antler,bone,hornandwhale/walrusboneworkingwereallactivitiescarriedoutonahousehold
basisduringhorizon4.Duringhorizon5,additionalactivitieswerecarriedoutonahousehold
basis, these were hunting, some beer brewing
andfoodprocessingotherthansausagemaking.
Furthermore,large-scalestoneworkingwasmost
11CraftsandproductioninearlyBergen
likelycarriedoutonafull-timeprofessionalbasis
inhorizons3to5.Andsomemetalworkingmay
have been carried out by full-time professional
residentsduringhorizon5.Combmaking,shoemaking,someleatherworking,andsomemetalworkingweremostlikelycarriedoutbyprofessionalambulatingartisansduringhorizon5.
Commonforalltheabove-mentionedactivities is that they must primarily have served an
interurbanmarket.
Inadditionsomebeerbrewing,somesausage
making, innkeeping and the activity of letting
outpremiseswereprobablycarriedoutbyfulltime or part-time professional residents during
horizon5.Theseservice-relatedtradesmayhave
servedawidermarketthanthetown,asvisitors
of the town probably made use of the facilities
andservicesprovided.
Wereanyoftheproductiveactivities
fundamentalfortheemergenceof
Bergen?
Fishing, hunting, miscellaneous antler, bone,
horn and whale/walrus bone working, some
‘other leatherworking’, small-scale wood and
stoneworking,basiccookingandsomefoodand
beverage processing were probably all activities
carried out on a household basis. Accordingly,
noneoftheseaddedvaluetothetowncommunity,andnoneofthemcouldinthemselveshave
providedafundamentaleconomicbasisforthe
riseofthetown.
Ambulating professional shoemakers (who
also repaired shoes), combmakers and metalworkers who came to Bergen for short visits,
were most likely artisans that served large areaswithstandardisednon-luxuriousitems,and
theymayprimarilyhaveservedtheinterurban
marketwhileworkinginBergen.Theirpresence
inBergenreflectedinthematerialfromhorizon
5 must be seen as secondary, and as a consequence of an established community. Consequentlytheirpresencemaynothaveservedasa
fundamentaleconomicbasisfortheriseofthe
town.
Itcouldnotbeestablishedwhethertheactivitiesofantler,bone,hornandwhale/walrusbone
203
5
4
5
Hunting/war/game
Basiccooking
Beerbrewing
Sausagemaking
Foodprocessingotherthatsausage
making
Textileproduction
Fishing
4
5
5
4
5
5
4
5
5
4
5
5
3+
4+
5+
Lettingoutroomforcontemporary
workshops
3
4
5
Small-scalewoodworking
Large-scalewoodworking
Small-scalestoneworking
4
5
5
4
5
Large-scalestoneworking
Metalworking
‘Otherleatherwork’
5
5
Innkeeping
Professional
sedentaryfulltime
Professional
sedentaryparttime
Whether
Professional
sedentaryparttimeor
ambulating
cannotbe
established
Professional
ambulating
4
5
Skindressing
Cannotbe
established
Household
Shoemaking
Combmaking
Productive
activity/
Producers
Antler,bone,hornandwhale/walrus
boneworking
Table62.ThenatureandorganisationofproductiveactivitiesindicatedinBergenbeforec1170
5+
5+
5+
5+
5+
5+
5+
5+
4
5
5
3+
4+
5+
Numbersdenotehorizonswheretheactivitieswererecorded
+Denotesthatthenatureandorganisationoftheactivitymaybemanifold
working,andshoemakingduringhorizon4and
antler,bone,hornandwhale/walrusboneworkingduringhorizon5werecarriedoutbyresident
part-time professionals or ambulating artisans.
Regardlessoftheorganisationoftheseactivities
duringhorizon4,andforantler,bone,hornand
whale/walrusboneworkingalsoduringhorizon
5, the small amounts of waste left behind suggestthattheartisansproducedforaninterurban
marketonly.Accordingly,theactivitiesdidnot
add value to the town community and cannot
haveservedasafundamentaleconomicbasisfor
theriseofthetown.
The only ‘full-time professional’ productive
activity documented in early Bergen is largescale stoneworking, carried out by craftsmen
engagedintheconstructionofthemanymonumentalbuildingserectedthroughhorizons3-5.
Presumably the artisans were integrated in the
204
householdofthemonumentfoundersduringthe
periodofconstruction.Inspiteofbeingprofessional, their production thus took place within
a ‘household’, and their presence should not in
itself be seen as a fundamental economic basis
fortheriseofthetown.Alongthesamelineof
thinking the presence of professional sedentary
or ambulating large-scale woodworkers should
notbeseenasafundamentaleconomicbasisfor
theriseofthetown.
Innkeeperswithbeveragesforsaleandlodgingsforrent,sausagemakersandpeoplewholet
outpremisesfortemporaryworkshopsmayrepresentagroupofurbanprofessionalswhowere
active in Bergen from horizon 5. The activities
ofthesenewurbanservice-relatedtrades,carried
outbypart-timeorfull-timeprofessionals,may
intimehaveaddedvaluetothetowncommunity
astheactivities,inpart,werepaidforbyvisiting
travellers.Thenewtradesshouldalsobeseenas
aconsequenceoftheexistenceofacommunity
ratherthantriggeringofftheriseofthetown.
Inconclusionnoneoftheproductiveactivities
documentedinearlyBergencanhaveservedas
afundamentaleconomicbasisfortheemergence
ofthetown.Theproductiveactivitieswereratherpresentasaconsequenceoftheexistingsettlement,thoughtheirpresencemustalsohavehada
synergeticeffectonthegrowthofthetown.
12TRADE
RawmaterialsanditemsnotproducedinBergen,
show that goods from both far and near were
broughtintotownandusedhereandthedistributionoftwincombsandshoeshasdemonstratedaninternalredistributionofgoodswithinthe
town. However, when elucidating the existence
oftradeasadailyactivityandtheimportanceof
tradefortheusersofBergen,onlytracesoflongdistancetradewillbediscussedandinvestigated.
Thus the entering and departure of goods in
andoutofBergenareconsidered,nottheinternalredistributionofgoodswithinthetown.As
mentionedearlier,thebroadterm‘long-distance
trade’coversbothtradethatwaspartofaninternational network and trade limited to Norway
(cfp40p).
Did the physical organisation of the town
meetthedemandsoflong-distancetrade?Were
majorinitiativestakentochoosethebestnatural
harbours and to improve harbour and working
conditionsalongthewaterfront?InTrondheim
eleven and twelfth century investments in harbourfacilitieshavebeeninterpretedasareflectionofthedevelopmentincontemporaryfreight
carriers (Christophersen and Nordeide 1994,
91ff; Christophersen 1997). I presuppose that,
if access to the sea from the town plots was a
priorityfortheplannersofBergenthismayindicatethatseatransportingeneralwasconsidered
important to the planners. And if accessibility
forseagoingandcoastalfreightcarriersseemsto
havebeenaprioritywhenplotswerelaidoutin
the town areas, this may be an indication that
long-distancetradewasconsideredimportantfor
12Trade
theplannersofBergen.Ialsopresupposethatif
workingconditionsinthetidalzoneandaccess
tothewaterfrontwereimprovedthisisanindicationthatseatransportingeneralwasconsidered important by the users of Bergen. And if
harbourfacilitieswereimprovedsothattheymet
thedemandsfordeepharboursofcontemporary
carriers,thismaybeanindicationthatlong-distancetradewasconsideredimportantfortheusersofBergen.
When evaluating harbour conditions I will
apply a methodological approach similar to
that of the Trondheim study. The ship wreck
Skuldelev3isusedasanexampleofaneleventh
century coastal carrier, Skuldelev 1 as an eleventhcenturydeepseacargocarrier,andLynæs,
dendrodatedtoc1140isusedasanexampleof
twelfthcenturydeepseacargocarrier.Thethree
shipshaddraughtsofrespectively0.84m,1.28
mand1.50mwhenfullyloaded(Crumlin-Pedersen1985,85-88).Theprofilesofthebottoms
of Skuldelev 1 and 3 (Crumlin-Pedersen 1985,
Figures 4 and 7), indicate that these boats had
tobeloadedandunloadedwhilestillfloating,77
-iftheyweredrawnontothebeachtheywould
heel over severely, probably causing damage to
thehull.Unloadingandloadingtheseboattypes
wouldthereforerequiremooringpossibilities;a
quayplacedinwaterwithampledepthandwith
directaccesstoland,ormooringpostslocatedon
ample depth. The latter solution would involve
theuseofbargesoragangplank.
Whenfacilitiessuchasquaysormooringposts
placed at ample depth are constructed, I interpretthisastheattempttomeetthedemandsof
contemporarycarriers.Whenaccesstotheshore
andworkingconditionsinthetidalzoneareimprovedIinterpretthisasanindicationthatsea
transport in general was considered important
fortheinitiatorsbehindtheactivities.
Thedistributionoftoolsoftrade:tallysticks,
owner’s marks, weights and balances, and the
presence of possible storage buildings serve as
sources as to where goods entered town and
where commodities may have been stored. Tallystickswereprobablyusedduringloadingand
unloading of cargo from ships and comprise
sticks of varying forms marked with knife-cuts
alongtheedges(Grandell1988,66).Atallystick
205
from Trondheim (dated to between 1050 and
1150) had the inscription: ‘Sigmund owns this
sack’(ChristophersenandNordeide1994,256),
inBergensimilarinscriptionsarefoundintwo
cases,inslightlylatermaterial.78Theinscriptions
support that the sticks were closely associated
withtradeinbulkcommodities.Whenfoundin
situinabuildingatallystickprobablysignifiesa
placewherecommoditieswereaccountedforora
storageroom.Owner’smarksarelabelsofwood
oftenwithanamewrittenonitinrunes.(Grandell1988,69).Thelabelswereprobablyattached
tothecommoditiesandmaydenotetheowner,
eitherthebuyerorseller.Whenanowner’smark
isfoundinsituinabuildingthebuildingmaybe
interpretedasastoragebuilding.Ifinditlikely
thattheuseoftallysticksandowner’smarksimpliesthatbulkcommodities,asopposedtosmall
volumeluxuryitems,reachedBergen.Thepresence of tallysticks and owner’s marks thus suggestthepresenceoflong-distancetradewithbulk
commodities. One weight and a balance arm
werefoundinhorizon5,theseartefactsareoftenusedasindicatorsoftrade(egCallmer1991,
29),theymay,however,alsobeassociatedwith
metalworking(Pedersen2001)andcantherefore
notbeseeninisolationassourcesfortrade.No
coins were assigned to the period under study,
neitheronsitesexcavatedbeforenorafter1980.
Metaldetectorshavegenerallynotbeenusedin
Bergenbutifcoinshadbeencommonintheperiodunderstudy,onewouldexpectthatatleast
some had turned up, in spite of relatively poor
preservationconditionsformetal.Thescarcityof
coins is therefore probably real. Christophersen
hasdiscussedtheuseofcoinsinearlymedieval
Scandinavia.Hearguesconvincinglythatcoins
werenotusedas‘generalpurposemoney’(Christophersen1989,134-137).Theabsenceofcoins
isthereforenotusedasasourceherewhendiscussingtheactivityoftradeinearlyBergen.79
Inearlierstudiessomeofthebuildingsfrom
early Bergen have been interpreted as storage
housesonthebasisoftheirlocation,thelackof
fireplacesorboxbenches,andspecialconstructional details. The function of these buildings,
andbuildingswheretoolsoftradearefoundas
categoryIartefacts,isdiscussedthroughcategory
Ifindsfromthebuildings(cfp68ff).Thearte206
factgroupsthatservedassourcesfortheproductiveactivitiesdiscussedinChapter11aredrawn
upon as sources for activity in the buildings.
Thenumberofbuildingsusedforthestorageof
goods and the degree of specialisation of these
buildingsmaybeanindicationoftheintensity
oftrade-relatedactivitiesthattookplace(Carelli
1999,480;Sarfatij1999,275).Thismayalsobe
significantforunderstandingtheimportanceof
long-distancetradeforearlyBergen.
Direct or indirect contacts between Bergen
and eastern and western Norway respectively
arestudiedthroughhonesfromEidsborginthe
eastandDarkGreySchisthoneswithawestern
Norwegian origin (Mitchell, Askvik, and Resi
1984).80Thescopeofdirectorindirectcontacts
betweenBergenandforeignplacesofproduction
isstudiedthroughthepresenceofpotteryfrom
places of production in present day England81,
Germany,82 France83 and the Low Countries,84
andstavebeakerstraditionallybelievedtohave
anorthGermanorigin(Weber1990,94).Being
awarethattheseartefactsmayreflectconsumptionratherthantrade(Carelli1999,77;Hodges
(1982)1989,57ff),Istilltakeitthattheyreflect
an indirect or direct contact with distant harbours.Owner’smarkswithauseof‘Greenlander’runes,thatmayimplyanownerfromIceland
orGreenland(Hagland1986,24;Johnsen1990,
N682, N737), are used as a source for contact
betweenBergenandNorthAtlanticdestinations.
Pollen from plants85 that did not grow in contemporary Norway may indicate the import of
graintoBergen(KrzywinskiandKaland1984,
33). Passages in the written sources also imply
traderelationswithbothlong-distancelocalities
inNorwayandabroad,theyaredrawnintothe
discussionwhenrelevant.
Whenreferringtostructuresandthenatural
topography,thesourcesarelabelled(B)forbasic,
(S)forsupplementaryand(G)forgeneralbackgroundsources.Thisisanexpressionofthegeneralreliabilityofthematerialassourcesforthe
horizonsandtothereconstructionofthenatural
topography.
Table63.Trade-indicatingsources
Infrastructure
Improvementofharbourfacilities
Storagebuildings
Toolsoftrade
Tallystick
Owner’smark
Possibletoolsoftrade Weight
Balance
Internationalcontacts Pottery,stavebeakers,runic
inscriptions,writtensources,pollen
Nationallong-distance HonesfromEidsborg,andfrom
contacts
westernNorway
Horizon2(c1020/30-c1070)
Aplotsystemwaslaidoutinthenortherntown
area,atthepresentstateofresearch,thisisbelievedtohavetakenplaceduringhorizon2(cf
p 183ff). The plots along Veisan probably included the Veisan shoreline, whereas the plots
alongVågendidnotextendtothewaterfront.In
thissensethesystemwasmoredirectedtowards
VeisanthantowardsVågen(cfp180ff).
Veisanwasstillamarinebasinintheeleventh
century(B)(Figure55a).Therecordedbedrock
thresholdbetweenVågenandVeisanwasquite
high,probablyat-0.3mbelowsealevel(S)and
atthenormalhightidetherewouldonlybeabout
0.76mwaterabovethethreshold.Ajokerinthe
reconstruction is, however, that perhaps only
partofthemouthofVeisanhasbeenrecorded(cf
Appendix1,point1andfootnote89).ThepossibilityofanaturalfairwaybetweenVågenand
Veisan in an area not covered by investigations
cannotbeexcluded,andIshallhavetoleavethe
questionofthethresholdtoVeisanopen.
Thereconstructionofthenaturaltopography
onthewesternbankofVeisan(B)showsagently
slopingbeachsuitableforloadingandunloading
if boats were drawn onto the beach. The boats
must, however, have been fit for beach landing.
The natural harbour conditions in Veisan were
Figure55a.Horizons2and3,harbourconditions,thenortherntownarea
12Trade
207
Figure55b.Horizons2and3,harbourconditions,themiddletownarea
thusnotespeciallysuitableforcontemporarycargocarriers.
Along Vågen, in the northern town area, the
seabedhasbeendocumentedbysite6only.The
beachslopedgentlywithagradientofabout1:121:17 between the 1 masl contour and the +/-0
contour resulting in a wide tidal area (B). The
gradientbetween+/-0mand-3mwasabout1:3
(B), contemporary carriers thus could not moor
alongside here directly, but would have to moor
beyondtheunderwatershelf.Thesourcesforthe
reconstructionofthebeachandtheseabedbysite
6,arebasic,andconsideredreliable.
Altogethercontemporarycargocarrierswould
have to moor beyond the underwater shelf both
inVeisanandbytheVågenshoreline,andwould
havehadtounloadandloadusingbarges.This
maybeanindicationthatseatransportingeneral
andlandingconditionsforcontemporarycarriers
were not the only issue when dividing the land
intoplots.
The jetty (B) that ran across the beach from
plot6/Dmayhaveprovidedaccessfromtheplot
to boats anchored beyond the underwater shelf
(Herteig1990,132).Itishoweveruncertainifthe
jettyactuallyextendedintothebay.Thejettywas
theonlydocumentedattempttoimproveharbour
208
facilitiesorworkingconditionsbythewaterfront
alongVågen,inthenortherntownarea(cfTable
64).Thenaturalharbourconditionsaswellasthe
improvementofharbourandworkingconditions
by plot 6/D are considered well-founded. The
improvedaccesstothewaterfrontonplot6/Din
thenortherntownareamayindicatethatworking
conditionsatthewaterfront,andthusseatransportingeneral,wereconsideredimportantforthe
usersofthisplot.
Inthemiddletownarea(Figure55b)thepier
builtinhorizon1(B)(unit30/A)wasstillstanding
andmusthavebeenkeptup.Hencethepierwas
apparentlystillconsideredimportantfortheusers.
Theseabedhasnotbeendocumentedaroundthe
pier,butjudgingbythereliableseabedcontours(S,
B)fromsites26and27thepiermusthavereached
atleastthe-1to-1.5mcontourandcontemporarycarrierscouldprobablygoalongsidehereat
hightide.Thepresenceofthepierandthegeneral
harbour conditions are considered well-founded.
Thepierisinflictedwithproblemsofinertia;being from the early tenth century, it was built to
meet needs that apply to a period much earlier
thanhorizon2.Andasweonlyhavetheverypierstructureasasourceforthefunctionofthepier,
thepiercannotbeusedasasourcetoilluminate
whetherornottheusersofsite30consciouslymet
the demands of eleventh century carriers. If the
pier met the demands of eleventh century carriers,thiswasmerelyacoincidencesinceitwasbuilt
yearsbeforethecarrierswerethoughtof.Thepier
therefore cannot be used as a source relating to
thepresentquestion.Notoolsoftrade,indicators
ofinternationalcontactnorstoragebuildingswere
assignedtohorizon2.Butthenhardlyanyfinds
havebeenassignedtohorizon2atall,andthelack
offindscannotbeusedasasource.
Tosumup,harbourconditionswereperhaps
notthemainpriorityfortheinitiatorbehindthe
regulationofthenortherntownareaintoplots.
Theimprovementofworkingconditionsatthe
Vågenwaterfrontonplot6/Dindicatesthatsea
transport in general was considered important
for the users of this plot. There are no indicationsthattradewascarriedoutinthetownarea
duringhorizon2
Iceland/Greenland
France
LowCountries
Germany
England
MultifunctionalStoragebuilding
Balance
Weight
X
Owner’smark
6/D
Tallystick
Plot/
site
Improvementofharbourconditions
Table64.Horizon2(c1020/30-c1070),trade-indicating
sources
NumbersandXinboldrefertobasicsources,
numbersandxinplainrefertosupplementary
sources
Horizon3(c1070-c1100)
During horizon 3 the natural harbour conditionsinthenortherntownareawereidenticalto
theconditionsinhorizon2(Figures55aandb);
Both by and along the Vågen shoreline (at site
6)carrierswouldhavehadtomoorbeyondthe
underwater shelf. In the middle town area, the
coast and seabed has been documented at sites
12Trade
26 and 27 only. The beach sloped gently with
a gradient of about 1:20 between the +1 and
the+/-0contours,theseabedhadagradientof
about1:10between+/-0and–2mbelowthesea
level(B,S).Thereconstructionofthenaturalis
basedmostlyonbasicsourcesandisconsidered
reliable.
Atthepresentstateofresearch,theintroduction of a new plot system is believed to have
taken place during horizon 3 (cf p 183). Plots
along Vågen were extended towards the Vågen
waterfrontandtherebygainedimmediateaccess
totheVågenBayinboththenorthernandthe
middletownareas.Thenewplotsystem’sfocus
ontheVågenshorelinemayindicatethatthose
whoregulatedthelandconsideredaccesstothe
seaandseatransportingeneralasimportant.
Agravellayer(S)wasprobablyspreadonthe
beach by the jetty (plot 6/D), it may have stabilised the beach, and made work here easier.
This may indicate that sea transport in general
wasconsideredimportantfortheuseroftheplot,
butasthematerialstemsfromonesupplementary
sourceonlyitcannotactasasourceinisolation.
Inthemiddletownarea,activitywasdocumented
atsites26and27only(whendisregardingsite30,
thatisinflictedwithproblemsofinertia).There
was probably activity in the vicinity of the two
sites(S),butthecharacteroftheactivityisuncertain,andnoattemptstoimproveharbourfacilities
havebeendocumented(S).Theimprovementof
workingconditionsonthebeachwasdocumentedthroughonesupplementarysourceonlyinthe
northernandmiddletownareas,whenexcluding
thematerialfromsite30.AccordinglyIfindthat
thetendencyinthematerialthatseatransportin
generalwasconsideredimportantfortheusersof
theplots,isnotwell-documented.Stillasseenin
Chapter 9 a reliable tendency in the material is
thatoccupiedplotsweremostlylocatedattheVågenshoreline.Thismayinitselfbeanindication
thataccesstotheseaandthusthatseatransport
ingeneralwasconsideredimportantfortheusers
ofthetownplots.
Thepierfromhorizon1atsite30/A,Vetrlidsalmenningen (B) probably still functioned as a
pier,butisnotusedasasourceforthepresent
questionduetoproblemsofinertia.
Notoolsoftrade,internationalcontact-indi209
catingfindsnorstoragebuildingswereassigned
tohorizon3.AhonefromeasternNorway,tentativelyassignedtohorizon3(S)onplot9-10/B,
indicatesdirectorindirectcontactbetweenBergen and eastern Norway. However, as the materialstemsfromasinglesupplementarysource
only it is not considered reliable as a source to
indicate such contacts, and is not included in
furtherdiscussions.Pollenofplantsthatdidnot
grow in contemporary Norway may have been
foundasearlyashorizon3onplot6/Eor6/F
(S),indicatingthatforeigngrainwaslandedin
Bergen.Thisinformationalsostemsfromasinglesupplementarysourceonly,andcannotstand
alone, it is therefore not considered in further
discussions.
6/D
x
6/Eor
6/F
910/B
Pollen
Iceland/Greenland
France
LowCountries
Germany
England
WesternNorway
EasternNorway
Multifunctionalstoragebuilding
Balance
Weight
Owner’smark
Tallystick
Plot/
site
Improvementofharbourconditions
Table65.Horizon3(c1070-c1100),trade-indicating
sources
x
(1)
ArtefactcategoryIIinbrackets
NumbersandXinboldarebasicsources,xand
numbersinplainaresupplementarysources
Tosumup,seatransportingeneralmayhave
beenaconsiderationwhenthetownareawasredesigned.Accesstotheseaandseatransportin
general may also have been a priority for those
whohadaplotinthetownarea.DirectorindirectcontactbetweenBergenandeasternNorway
andtheimportofgrainwasindicatedinsupplementary sources but the information is consideredtoouncertaintobeincludedinthefurther
discussions(Table65)
210
Horizon4(c1100-1120s)
Inthenortherntownarea(Figure56)apier(S)
may have been constructed (14/A) and extended into the Vågen Bay reaching a water depth
of-2.5to-3.0m.Thepierwasdimensionedfor
heavyloadsandcontemporarycoastalaswellas
seagoingcarrierscouldberthheredirectly.The
pierwasassignedtohorizon4asasupplementarysourceonly,itspresenceinthelandscapethis
early is thus not well-documented, and cannot
be used as concluding evidence that the needs
of contemporary carriers were met during this
horizon.
Onplots6/B-6/Fseverallayersofsmallstones
(B) were laid out, providing a firm surface for
theunloadingandloadingofshipsandforother
workonthebeachandinthetidalzone(Herteig
1991,111).Largershipswould,however,stillhave
tobeunloadedandloadedbymeansofsmaller
boats.Onplot6/Cfurtherstepsweretakento
improveworkingconditionsacrossthetidalzone
andbythewaterfront;arowof2mx2mstonefilledcaissons(B)wereconstructedformingthe
foundationsforaquayalongthewaterfrontand
apassagethatranfromthewaterfronttotherear
partoftheplot.Thequayfrontwaslocatedbythe
+/- 0 contour, thus contemporary carriers were
probablynotabletoberthhere,butwouldhave
tomoorindeeperwaterfurtheroutintheharbourbasinandunloadbymeansofbarges.The
passage,however,providedbetteraccessfromthe
waterfronttothebuildingsattherearpartofthe
plot.Similarpassagesandquayfrontsmayhave
beenconstructedatplots9-10/B,9-0/Cand12/
A(S),wheretheseabedwasquiteshallow,ithas
beenreconstructedwithagradientofabout1:10
betweenthe+/-0andthe-1mcontours(B,S),
and contemporary carriers could probably not
goalongsidethequaysdirectly.Fromsite27in
the middle town area, waste-layers (B) had accumulatedontheseabottombeyondplots2627/B and 27/C, thereby reducing the depth of
theVågenBay.Atsite26atriangularstone-filled
caisson(B)mayhavebeenpartofapassagethat
ledacrossthetidalzonefromthebuilt-outpart
oftheplottothewaterfront,probablyasituation
identicaltothatdocumentedonplot6/C.Itis
interpretedasanattempttogiveaccessacrossthe
tidalzoneonplot26/Aor26-27/Bandthereby
Figure56a.Horizon4(c1100-1120s),harbourconditions,thenortherntownaera
Figure56b.Horizon4(c1100-1120s),harbourconditions,themiddletownaera
12Trade
211
improvingworkingconditionsbythewaterfront.
Theimprovementofworkingconditionsandaccesstothewaterfrontwasdocumentedthrough
basicsourcesinboththenorthernandthemiddletownareas,andthegeneraltendencyinthe
materialthattheseactivitiestookplaceinboth
thenorthernandmiddletownareasisconsidered
reliable. The improved access to the waterfront
andworkingconditionsaretakenasanindicationthatseatransportingeneralwasimportant
fortheusersofthetownplots.Thepier(B)in
unit 30/A may still have been in use as a pier,
butisnotusedasasourcehereduetoproblems
ofinertia.
A tallystick, found in building 45 on plot
6/C in the northern town area, reflects trade
withbulkcommoditiesonthisplot.Thebuilding(B)hasbeeninterpretedasawarehouseora
boathousethroughconstructiondetails(Herteig
1969, 98ff, 113; Moldung 2000, 24). Artefacts
ofcategoryIfromthebuildingshowthatgeneral
cookingandsmall-scalestoneworkingalsotook
place here. It thus may seem that the building
was‘multifunctional’,thatisithadotherfunctionsinadditiontobeingawarehouse.
DirectorindirectcontactswithGermanyand
theLowCountriesaredocumentedthroughpotteryonthreeorfourplots.Ononeandfourplots
respectivelyhonesfromeasternandwesternNorwayindicatecontactbetweenBergenandthese
areas. Pollen from weeds not native to Norway
were found on either plot 6/E or 6/F (B) and
indicate that grain was a commodity imported
to Bergen. The tallystick, the contact-indicatingsourcesandthepollenstemfromfivebasic
sources and two supplementary archaeological
sources,inadditionamultifunctionalwarehouse
was also recorded through a basic source. The
tendency in the archaeological and botanical
materialthattradewasnowpartofthetownspeople’s economy is considered reliable on this
basis.
Passages in the Orkneyinga saga tells that
RagnvaldKalefromAgderwenttoGrimsbyin
Englandabout1115-20,hewasinthecompany
of salesmen and carried along goods for sale.
Upon his return to Norway he went to Bergen
and later he sailed from Trondheim to Bergen
(Orkn 1913-16, 141; Holtsmark 1970, 93-94).
212
These passages may imply that Bergen was a
nodeinanetworkforbothcoastaltrafficonthe
Norwegian coast and traffic across the North
SeatoEnglandabout1115-20,correspondingto
horizon 4. This information corresponds with
thetendencyinthearchaeologicalandbotanical
sources,andisconsideredreliable.
Inconclusion,seatransportingeneralseems
tohavebeenconsideredimportantfortheusers
ofseveralplotsinVågeninthenorthernandthe
middletownareasandthesourcesaltogetherimply that trade with bulk commodities was now
partoftheeconomyofthetownspeopleofBergen.
Horizon5(1120s-c1170)
Sometimeduringhorizon5,Veisanwasprobably
notfitasaharbouranymore(cfHjelle1986,67).
Inthenortherntownarea(Figure57)waste-layers
(B)depositedduringhorizon4andamounting
to a thickness of 0.5 m at site 6 had accumulated along the Vågen shoreline (cf also Figure
8).Thishadanimpactonthewaterdepthclose
totheshore.Thepierbysite14(B)wasinuse
and the water depth was most likely sufficient
forevenlargecarrierstoberthhere,thesourceis
consideredreliableandisanindicationthatthe
needsofcontemporarycarriersweremet.
Passagesandquayfrontstructures,likethose
seenatplot6/Cinhorizon4,werenowbuilton
mostdocumentedplotsalongtheVågenshorelineinboththenorthernandthemiddletown
areas.Inthenortherntownareathequayswere
mostlylocatedbetweenthe+/-0and-1mcontours,onplot6/G(B)theoutermostcaissonsin
thequay,however,extendedtothe-2mcontour.
Mooringposts(B)werelocatedbeyondthequay
frontsbetweenthe-1mand-2mcontours.In
spite of the decreased water depth, due to the
depositionofwaste-layersinthebay,theseabed
alongVågeninthenortherntownareawasprobablydeepenough,evenatlowtide,forcontemporaryseagoingcarrierslikeLynæstobemoored
bythemooringposts.Theboatscouldthenbe
loadedandunloadedbytheuseofagangplank
betweenthevesselandthequay.
In the middle town area, at site 26, wastelayers deposited during horizon 4 and layers
depositedintentionally(B)hadraisedtheland.
26-27/BC•
27/C
30/E
1
Pollen
Iceland/Greenland
France
LowCountries
(1)
Germany
(1)
England
Balance
Weight
Owner’smark
WesternNorway
X
X
X
X
EasternNorway
6/B
6/C
6/D
6/E
6/Eor6/F
6/F
6/G
9-10/B
9-10/C
14/A
26/A
26-27/B•
Tallystick
Improvementofharbour
conditions
Plot/site
Multifunctionalstoragebuilding
Table66.Horizon4(c1100-1120s),trade-indicatingsources
1
X
X
X
x
x
x
X
(1)
(2)
(1)
1
2
(1)
(1)
3
(3)
NumbersandXinboldrefertobasicsources,numbersandxinplainreferto
supplementarysources
•Artefactsfrombothbasicandsupplementarysources
Onplots26-27/B,27/Cthequayfrontcaissons
(B)werethuslocated7-11mfromtheoriginal
shorelinebetweentheoriginal+/-0mand-1m
contours,onplot26-27/Band27/Cthisgavea
waterdepthoflessthan1minfrontofthequay
(B)(Figure58).Mooringpostsweredocumentedbeyondthequayatplot26-27/Baboutthe-1
mcontour(B).Onplots29/Aand29/Basimilar
locationofthequayfront(S)maybeseen.On
plots 28/A, 28/B and 28/C the quay front (B)
waslocatedabout20mfromtheoriginalshoreline (S), between the -1 m and -2 m contours
(S).Apierextendedalmosttothe-2mcontour
onplot28/Bandmooringposts(B)werelocated immediately beyond the quays. At normal
tideorlowtidethequaysandmooringpostsof
plots26-27/Band27/Cwerenotlocatedinsufficientlydeepwaterforavesselwiththedraught
ofLynæs.Boatswithadraughtofmorethan1
mwouldthusprobablyhavetobemooredfurtherfromtheshoreandunloadedandloadedby
12Trade
the help of barges (Figure 58). On plots 28/A,
28/Band28/C,thedepthbythequayfrontwas
sufficientforvesselswithadraughtof1.5mat
normalandhightidetogoalongsidethemooring posts and use a gangway during unloading
andloading.Onplot28/Bthepierprovidedsufficientdepthforlargevesselstolandevenatlow
tide.Inunit30/A,thepierfromhorizon1was
nowincorporatedindrylandstructures(B).
Inthesoutherntownarea,structures(B)interpretedasaquayfrontandthefoundationofa
counterbalancehoist,werefoundonsite38.The
seabottomattheheadoftheVågenBayprobablyslopedgentlywithagradientof1:30between
the+/-0mandthe-1mcontours.Dataforthis
reconstruction are scarce, so it may not be entirelyreliable,neverthelessthestructuresonsite
38werelocatedabovethe+/-0mcontourinthe
tidalzone(B)andevenathightidelargerboats
cannothavegonealongsidethequayhere.Larger
boatsmusthavemooredbeyondthequaysand
213
Figure57a.Horizon5(1120s-c1170),harbourconditions,thenortherntownaera
Figure57b.Horizon5(1120s-c1170),harbourconditions,themiddletownaera
214
Table67.Horizon5(1120s-c1170),toolsoftrade
12Trade
6/B
6/C
6/D
6/E
26/A
26-27/B-C
3
3
(5)
5
(3)
(4)
Balance
Weight
Plot
Tallystick
unloading and loading must have been carried
outbythehelpofsmallerboats.
The sources that elucidate improvements of
working and harbour conditions during horizon5inthethreetownareasarealmostentirely
basedonbasicsources.Thetendencyinthematerialthatsubstantialimprovementsinworking
conditions in the tidal zone were undertaken,
and that harbour conditions were improved to
meetfreightcarriers’demandsfordeeperwater,
isconsideredreliable.
Tools of trade were found in four of the 24
artefact-yielding units in the northern town
areaandonplot26-27/Bor27/Cinthemiddle
town area (Table 67). On plot 26/A a balance
armwasfound,thefunctionofthebalancearm
is ambiguous and as waste from metalworking
wasalsofoundontheplot,thebalancearmmay
stemfrommetalworkingratherthantrade(cfp
214ff).Thetoolsoftradewereassignedtohorizon5throughbasicsources,sotheirassociation
withthehorizonisconsideredreliable.
Owner’smark
Figure57c.Horizon5(1120s-c1170),harbourconditions,thesoutherntownaera
(1)
(1)
3
(3)
(1)*
(1)
Numbersinboldarebasicsources,inplain
supplementarysources
*possiblyfrommetalworking
Someofthetoolsoftradewerefoundinbuildings(B),implyingthatthebuildingswereeither
warehouses for commodities or houses where
commodities were accounted for. In addition,
buildings41and66onplot6/Bhavebeeninterpretedasstorageroomsonthebasisofconstruction details; these buildings had open corners
that may have provided ventilation for stored
215
Figure58.Site27,Finnegården3a.Waterdepthbythequayfront.(ModifiedfromGolembnik1993,Figure5)
216
goodssuchasstockfish,andwereunfitaspermanentdwellings(Herteig1969,94ff;Helle1982,
126). Table 68 shows the five buildings that
couldbeassociatedwithtradeandtheproductiveactivitiesrepresentedbyartefactsofcategory
Iinthebuildings.Bothbuildingsandartefacts
wereassignedtohorizon5asbasicsourcesand
theirassociationwiththehorizonisconsidered
reliable.
Threebuildingsbelongtoplot6/Dandtwo
to plot 6/B making up half the documented
buildings on these plots. This large proportion
oftrade-relatedbuildingsimpliesthattradewas
veryimportanttotheusersoftheseplots.Inaddition,tobeingassociatedwithtradeorstorage
ofcommodities,thetrade-relatedbuildingswere
allusedforpurposesthatwedonottraditionally
relatetotrade.Infourofthebuildingssausage
pins were found; but since sausages were often
dried and stored, the finds are not incompatible with the interpretation of the buildings as
storage houses. Fishing/textile tools comprisingweightsonlywerealsofoundinfourofthe
buildings.Iftheweightsrepresentfishing,their
presence corresponds well with the function of
thebuildingsasstoragerooms,thoughperhaps
not what we understand as storage rooms for
commodities of trade. As in building 45 (from
horizon4),thereweretracesofgeneralcooking
inallthehousesandfurthermoretheactivitiesof
antler,bone,hornandwhale/walrusboneworking, shoemaking and/or small-scale stoneworking were represented in two houses. According
tothefinds,thefivetrade-relatedbuildingsseem
tohavebeenusedforseveralpurposesandwere
notmeantforstorageofcommoditiesalone,they
werethus‘multifunctional’.Thismodifiessomewhattheimpressionoftradeasveryimportant
forthetownspeople.
In22ofthe24artefact-yieldinganalyticunits
inhorizon5(Table69)artefactsindicatingdirectorindirectinternationalcontactshavebeen
found.Andin16and7unitsrespectively,hones
fromEasternandwesternNorwayhavealsobeen
found.Theartefactgroupsareallrepresentedby
basicsources,andthetendencythatinternational, as well as national contacts were a reality is
consideredwell-founded.Thewidedistribution
ofitemstransportedoverlongdistancesindicate
X
X
X
X
X
Owner’smark
X
X
X
X
X
Tallystick
X
X
X
X
Small-scaleStoneworking
Textile/fishing?
Generalcooking
X
X
X
X
X
Foodandbeverage
processing
6/B/66
6/B/41
6/D/484
6/D/485
6/D/486
Shoemaking
Horizon5
Plot/
building
Antler,bone,hornand
whale/walrusboneworking
X
X
X
Xinboldarebasicsources,xinplainsupplementary
sources
6/C
X
2
6/D
X
3
6/E
X
6/F
6/G
X
X
6/H
8/A
8/B
8/D
1516/A
20/A
21/A
22/A
26/A
X
2627/B
2627/
BC
27/C
28/B
28/C
29/A
29/B
30/A
30/E
30/E
38/A
X
3
1
(2)
3
5
(6) (6)
8
1
(6) (16)
(4) (8)
(1)
5
(1)
1
(3)
1
(2)
LowCountries
Germany
England
WesternNorway
2
15 40 20
(1) (7) (17) (26)
1
44
1
15
(2) (134) (138) (46)
74
1
59
32
(8) (13) (193) (87)
(1) (12) 8 (44)
(77)
(8) (24) (22)
(2) 42 15
(54) (21)
5
(1)
(3)
1
1
(2)
2
(4)
(4)
(4)
Iceland/Greenland?
X
X
France
6/A
6/B
Toolsoftrade
Improvementof
harbourconditions
Multifunctional
storagebuilding
Table68.Horizon5(1120s-c1170),thefunctionof
buildingscontainingtoolsoftradeorinterpretedasstorage
rooms
EasternNorway
Table69.Horizon5(1120s-c1170),trade-indicating
thatlong-distancetradewasimportanttopeople sources
onmostoftheplotsinthetown.Thisimpression
is supported by the mention in written records
ofanumberofcoastalcarriersthatvisitedBer- Plot
genontheirwaynorthwardstoTrondheimwith
commoditiesin1162(Hkr1893-1901,III438439;Helle1982,162;Hkr1982(1979),656).
1
(1)
1
1
1
1
(5)
1
(1)
X
(1)
(5)
(1)
6
37 14
(22) (46) (32)
(4)
3
1
(21) (8)
(3) (10) (10)
(3)
In conclusion it is well-founded that harbour
andworkingconditionswereimprovedinallthe
townareas.ThusitseemslikelythatgoodlandX
(1)
(2) (9) (11)
X
(1) (2) (2) (1)
ing possibilities for contemporary carriers were
X
(2) (3)
considered important by the users of the town
X
plots. Tools of trade and the wide distribution
X
(1)
of contact-indicating artefacts throughout the
(1)
(2)
whole town area are also well associated with
(1)
(1) (4)
X
1
horizon 5 and their presence may indicate that
(1)
trade had become an important activity to the
NumbersandXinboldrefertobasicsources,in
inhabitantsofBergen.
plainsupplementarysources
Conclusions
Thenaturallandingconditionsforcontemporary
carriers were not especially favourable during
horizon 2 so harbour conditions were probably
nottheonlypriorityfortheinitiatorbehindthe
regulationofthenortherntownareaintoplots.
Improvement of the working conditions at the
Vågenwaterfrontonplot6/Dindicatesthatsea
12Trade
217
transportingeneralwasconsideredimportantto
theusersofthisplot.Itcouldnotbedetermined
whethertradewasanactivitycarriedoutduring
thehorizon.
During horizon 3 access to the sea, and sea
transportingeneralmayhavebeenapriorityfor
theusersofatleastoneplotinthetownarea.DirectorindirectcontactbetweenBergenandeasternNorwayandtheimportofgrainwasindicatedinsupplementarysourcesbuttheinformation
isconsideredtoouncertaintobeincludedinthe
furtherdiscussions.
Workingconditionsinthetidalzoneaswell
asaccesstothewaterfrontwereimprovedinboth
thenorthernandthemiddletownareasduring
horizon4andseatransportingeneralseemsto
havebeenimportanttotheplotusersalongVågen. Trade-indicating sources show that longdistancenationalandinternationaltradeinbulk
commoditieswasnowpartoftheeconomyofthe
townspeopleofBergen.
Substantial improvements of harbour- and
workingconditionsacrossthetidalzoneinthe
threetownareaswerecarriedoutduringhorizon
5.Itthusseemslikelythatgoodlandingpossibilities for contemporary carriers were considered
important to the users of the town area. Tools
oftrade,trade-indicatingbuildings,andcontactindicatingartefactswidelydistributedthroughoutthewholetownareaindicatethattradehad
becomeanimportantactivityfortheinhabitants
ofBergen.
13THECHARACTEROFTHE
SETTLEMENTINTHETOWN
AREA
In the oldest phase of occupation at the Folkebibliotekstomten(TheLibrarysite)siteinTrondheimthesettlementwasinterpretedastemporary
becauseoftheprovisionalcharacterofstructures
identifiedandthelimitedaccumulationofculturelayers(ChristophersenandNordeide1994,267).
InstudiesofeighthcenturyRibe(Denmark)and
Viking Age Kaupang in Tjølling (Norway) the
character of the constructions has been used as
218
anarchaeologicalcriterionwhenelucidatingthe
character of the settlement (Frandsen, Madsen,
and Mikkelsen 1988, 8; Jensen 1992; Ferveile
1994; Skre, Pilø, and Pedersen 2001, 10). Due
tothefragmentarystateofthesources,especially
in the earliest horizons, such criteria cannot be
adaptedhere.Insteadthecharacterofthesettlement on the town plots is studied through the
presenceofcertaindailyactivitiesandgroupsof
artefacts.Dailyactivitiesmayreflectthecharacteroftheurbancommunity,theymayelucidate
whethersettlementsonthetownplotswerewell
established and had a permanent character, or
whetherthetownplotwasusedseasonallyoroccasionallybypeopleonassignmenttothetown
foralimitedperiodoftimeorpeoplethatwere
justpassingthrough.
Ifinditlikelythatiffoodandbeverageswere
processedonaplotthismaybeanindicationthat
thesettlementherewaswellestablishedandhada
permanentcharacter.Thedistributionofsources
forfoodandbeverageprocessing(cfp235ff)is
drawnuponhere.
Ialsofinditlikelythatthepresenceofwomen and young children indicates a settlement
that was well established and had a permanent
character.Intheearliestand‘seasonal’phasein
Trondheimnotracesofwomenorchildrenwere
found (Nordeide 1989, 34; Christophersen and
Nordeide 1994, 237, 269; Nordeide 1999, 46),
lendingsupporttothisnotion.Textiletoolshave
typicallybeenassociatedwithfemaleactivitiesin
medievalstudies(Øye1988;Rui1993;Nordeide
1999;Hagen(1988)1994)andareusedhereas
indicatorsofwomen.Iwillonlyuseartefactsin
thecategory‘textiletools’(cfp226ff)asasource
forthepresenceofwomenasonlythesefindsare
positivelyidentifiedastextileequipment.
Toysandshoesizesareusedassourcestoindicatethepresenceofyoungchildren.Medieval
authorsmostcommonlydividedchildhoodinto
infantia,0-7years,puertia,7to12yearsforgirls
and7-14yearsforboysandadolescentia12or14
toadulthood.Theviewofinfantiaasaperiodin
which the child was helpless and dependent on
adults is expressed by several medieval writers
(Shahar(1990)1992,22ff).Inmystudyayoung
childisapersonthatwasstilldependentonits
parents/adultsandwhowasstillnotfitforwork.
Horizon3(c1070-c1100)
A child’s shoe (S) from plot 9-10/B and pollen
ofmyricagale(S)fromplot6/Eor6/Fmayindicatethatthesettlementsherewerewellestablishedandhadapermanentcharacter(Table70).
The evidence of beer brewing and the presence
ofyoungchildrenrespectivelyis,however,documentedthroughonlyonesupplementarysource
eachandisthusconsideredtoouncertaintobe
includedinthefurtherdiscussion.
9-10/B
6/Eor6/F
Whippingtop
(lathe-turnedcore)
Toy-sword
Toy-boat
Toy-horse-figure
Child’sshoes
Textileproduction
Plot
Foodandbeverage
processing
Table70.Horizon3(c1070-c1100),sourcesforthe
characterofthesettlements
(1)
(x)
ArtefactcategoryIIinbrackets
Xandnumbersinboldarebasedonbasicsources,
xandnumbersinplainarebasedonsupplementary
sources
13Thecharacterofthesettlementinthetownarea
Horizon4(c1100-1120s)
Food processing was documented on four of
thesevenartefact-yieldingplots/units,twoplots
werelocatedinthenortherntownareaandtwo
plots in the middle town area (Table 71). The
presence of this activity is considered reliable,
beingdocumentedthroughbothbasicandsupplementary sources. The sources thus indicate
thatwell-establishedsettlementsofapermanent
characterwerepresentinthenorthernandmiddletownareas.
Whippingtop(latheturnedcore)
Toy-sword
Toy-boat
(X)
X
X
(x)
Toy-horse-figure
6/B
6/C
26/A
27/C
Child’sshoes
Plot
Textileproduction
Table71.Horizon4(c1100-1120s),sourcesforthe
characterofthesettlements
Foodandbeverage
processing
Thiswouldseemtocorrespondwiththeperiod
ofinfantia.Youngchildreninthepresentcontext
are7yearsofageoryounger.Todaysize28,185
mm,istheaveragesizeshoesforchildren7years
ofage(Ruth2000).86Onaverage,medievalmen
andwomenwereabout4.3%shorterthantheir
modern counterparts (Bennike 1993).87 As the
proportionsofthehumanbodyareunchanged,
the average size for the shoe of a seven-year-old
childoughttobe177mmoraboutsize26-27.88I
choosetoapply175mmasthedivisionbetween
shoesforsmallchildrenandshoesforolderchildren and grownups. This figure does not take
intoconsiderationindividualvariationsandmust
be considered a working hypothesis rather than
asanaccuratemeasurement.Ihavemeasuredthe
length from toe to heel on 345 soles or uppers.
Onshoeswithapointedoranelongatedtoethe
toe was not included. Thirteen soles or uppers
were 175 mm, size 26-27, or smaller. Toys are
identified through parallel material from medievalcontexts.
Nosourcesforthepresentquestioncouldbe
safelyassociatedwithhorizon2.
ArtefactcategoryIIinbrackets
Xandnumbersinboldarebasedonbasicsources,
xandnumbersinplainarebasedonsupplementary
sources
Horizon5(1120s-c1170)
Textile tools indicating the presence of women
werefoundinnineofthefind-bearinganalytic
unitsinhorizon5,coveringthenorthern,middle
and the southern town areas. Finds indicating
thepresenceofyoungchildrenwerefoundonsix
ofthefind-yieldingplotscoveringthenorthern
andthemiddletownareas,andfoodprocessing
wasdocumentedin12or13units,coveringthe
threetownareas.Asshownearlier,thepresence
of the activities of textile production and food
processingisconsideredwellestablished,asthe
activities are mainly indicated in basic sources.
Thepresenceofchildrenisdocumentedthrough
eight basic sources and is thus also considered
well-founded. All in all, well-established settlementsofapermanentcharacterwerereflectedin
thesourcesfrom15ofthe24find-yieldinganalyticunits,coveringallthethreetownareasand
theirpresenceisconsideredwell-documented.
219
6/B
6/C
6/D
6/E
6/F
6/G
8/D
20/A
21/A
26/A
26-27/B
26-27/
BC
27/C
28/B
28/C
38/A
X
(X)
X
(X)
X
(X)
(X)
(X)
X
(X)*
(x)
x
X
(X)
X
(X)
X
(X)
(X)
Whippingtop(latheturnedcore)
Toy-sword
Toy-boat
Toy-horse-figure
Child’sshoes
Textileproduction
Plot
Foodandbeverage
processing
Table72.Horizon5(1120s-c1170),sourcesforthe
characterofthesettlements
(1)
(1)?
1
(3)?
(2)
1
(5)
(1)
(1)?
(2)
(1)
(1)?
1
(1)
(X)
(x)
(X)
(X)
(X)
(X)
(1)
(X)
(X)
(X)
(X)
(1)
(1)
ArtefactcategoryIIinbrackets
Xandnumbersinboldarebasedonbasicsources,
xandnumbersinplainarebasedonsupplementary
sources
*Only0.55%ofthetotalnumberofartefacts(cfp
260ff )
?Mayhavebeenusedasatoy
Conclusions
Fromtheavailablesourcesitisdifficulttohavea
qualifiedopiniononthecharacterofthesettlementsthatmostlikelyoccupiedatleastsomeof
theplotsorunitsinthetownareaduringhorizons2and3.Fromhorizon4andonwardsthe
presenceofwell-establishedandpermanentsettlements is well-documented. Well-established
andpermanentsettlementswerefoundinboth
thenorthernandthemiddletownareasduring
horizon4.Inhorizon5,well-establishedpermanent settlements were documented in all three
townareas.
220
PARTIII
THESYNTHESIS
14HOW,WHEN,BYTHE
How,when–andbywhom?
INITIATIVESOFWHOM,AND
SettlementtraceshavebeenlocatedintwoplacWHYDIDBERGENEMERGEASA esintheBergenareaduringhorizon1(c800-
c1020/30).ActivitytraceswerefoundinVeisan
TOWN?
andhavebeententativelyassociatedwithasetSixpartstudieshavenowelucidatedmajorinitiatives and daily activities that took place in the
Bergen area from the ninth century to c 1170.
Theinsightandconclusionsfromthepartstudies are drawn upon in the synthesising discussions of how, when, by the initiative of whom
andwhyBergenemergedasatown.Thisisthe
themeforthelastpartofthethesis.
IhaveinferredthatactioninahierarchicsocietylikethatoftheearlyNorwegiancentralkingdom, was initiated top-down and bottom-up.
Resourcefulactors,thekingorrepresentativesof
theking,tookthetop-downinitiative,whereas
the bottom-up initiative was taken by people
from lower levels of society, here represented
by ‘the townspeople’ and visitors of the town.
Throughthenarrowtimescopesrepresentedby
thehorizonssomeoftheinitiativesanddailyactivitiescanbelinkedtohistoricallyknownpersonsandtheactivitiescanbeseeninthecontext
of the wider society. Groups of actors are thus
goingtobelinkedtomajorinitiativesanddaily
activities,adding‘bytheinitiativeofwhom’and
also‘why’tothelistofquestionsaddressed.
The first questions to be addressed are how
andwhenthetownofBergenfirstsawthelight
ofday.WasBergenfoundedordidthetowngrow
organicallyoutofanolderurbansettlement?If
founded,howandwhendidthishappen,andby
theinitiativeofwhom?Andhowandbywhom
wasthetownsettledthroughtheyears?
tlement where agrarian activities were carried
out.ThesettlementwasperhapslocatedatHolmen with its fields in the general area around
VågenBay.Inthemiddletownarea,apierwas
documentedandinterpretedasalanding-place.
The pier was not part of a wider built-up area.
Whenthenortherntownareawasdividedinto
plots,itseemsthatthishappenedonapieceof
landnotpreviouslyoccupiedbyanon-ruralsettlement.Thelocationoftheplotswasrespected
intheyearstocome,thisshowsthatthelayoutof
thenortherntownareawasconsideredasanact
ofafundamentalcharacterbytheeleventhand
twelfth century users of the Bergen area. The
oldNorseverbsetjadenotestheactoffounding
somethingfromthegroundandthusapplieswell
totheactoflayingoutthenortherntownarea
(cfp25ff).
Whatcanthematerialassignedtohorizon2
tellusaboutthetypeofplacethatwasfounded?
Landparcelledintoplotsiscommonlyseenasan
indicator of the founded town or marketplace,
the latter being characterised by occasional as
opposedtopermanentsettlement(cfpages20ff
and38).Itcannotbedeterminedonanempirical basis whether the initiator of the layout of
thenortherntownareahadatownoraseasonal
marketplaceinmindwhenregulatingtheland.
However,becauseBergenintimedevelopedinto
apermanentlysettledurbancentreIfinditmost
likelythattheinitiatorhadplansforatownand
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
221
notaseasonalmarketplacewhenlayingoutthe
land.ThusIpresumethatwhenplotswerelaid
outinthenortherntownareatheideaofatown
wasmaterialisedandatownwasfoundedfrom
thegroundintheBergenarea.
Who did this planning? Former research on
thehistoryofBergenhasshownthattheBergen
areawasmostlikelyownedbythekingbeforea
townemergedhere(Helle1982,77-79withreferences).Ifthiswasthecase,itislikelythatonly
thekingcouldpossesstheauthoritytodividethe
land into plots. Furthermore the plots seem to
havebeenlaidoutaccordingtoanoverallplanreflectedintheregularityoftheplot-widths.Both
thesefactorssuggestthatacentralauthority-the
king-shouldbeseenasresponsibleforthelayoutofthenortherntownarea.Aswehaveseen,
thedatingofthefirstplotsysteminthenortherntownareaisnotbasedonveryfirmevidence,
but an assignment of the system to horizon 2
(c1020/30-c1070),isthemostreliablealternativeatthepresentstateofresearch(cfp183).
Thepit-houseinunit7/Awasprobablyconstructedabout1020accordingtoawide14Cdate,
the jetty on plot 6/D was constructed shortly
‘after1029’,accordingtodendrodatesfromthe
construction. Both the pit house and the jetty
wereofthesameorientationastheplotsintheir
surroundingsanditisreasonabletoassumethat
bothconstructionswerebuiltincorrespondence
withtheplotlayoutandnotviceversa.Itisthereforealsoreasonabletoassumethatthepithouse
andthejettywereconstructedshortlyafterthe
regulationoftheland.Ifso,thenortherntown
areamayhavebeenpartofaplanforatownthat
materialisedsometimeabout1020/30.
Between1015and1026/27OlavHaraldsson
(laterStOlaf)aimedtowinrecognitionasking
bytheNorwegiansandcreateacentralkingdom.
TheNorwegianaristocracyseemstohavebeen
dividedbetweensupportersofOlavandacentral
Norwegianmonarchyandthosethatwishedto
upholdtheoldersystemwheretheoverlordship
ofadistantforeign(Danish)kingwasexercised
through native lords. From the middle of the
1020s resistance to Olav seems to have grown.
Norwegian aristocrats collaborated with the
Danish/EnglishKingKnutdenMektige(Knut
the Powerful), who came to Norway in 1028,
222
gainedrecognitionaskinganddroveOlavHaraldssonintoexile.Norwaythenbecamepartof
Knut’s Scandinavian Empire (1028-34) (Rumble1994,6).In1030,OlavreturnedtoNorway,
butfellatthebattleofStiklestad,wherepeople
fromTrøndelag,westernandnorthernNorway
areknownashisopponents.Breakingthetradition of ‘distant overlordship’, Knut after a few
yearsimposedhisyoungsonSvein(Swein)under the tutelage of his mother Alfiva (Ælfgifu)
askinginNorway(1030).Thisdirectruleand
probable attempt at enlargening royal rights in
Norway was probably less acceptable than distant overlordship for the native aristocracy and
intheyearsbetween1030and1034/35achange
of policy within the Norwegian aristocracy appearstohaveemerged.Ifwecantrustthewritten records, the leading aristocracy now saw
the advantages of an independent Norwegian
centralmonarchy,amonarchythatwasatservicetothearistocrats.By1034,SveinandAlfiva
wereforcedtoleavethecountryandsoonafter
OlavHaraldsson’syoungsonMagnusOlavsson
was set up as king and sworn allegiance to by
Norwegianaristocrats(Andersen1977,128-146;
Sawyer1994,20-22).
If the date of the laying out of the land was
about1020/30,bothOlavHaraldssonandKnut
denMektigeortheirrepresentativesarepossible
founder-candidates. Can arguments that make
onemoreprobablethantheotherbepresented?
DuringthereignofOlavHaraldssonthecentral kingdom was strengthened and more land
thaneverwasundertheking.Olavmayhaveintroducedroyaladministrationandcontrolona
locallevel,andheisknownforhisintroduction
ofChristianityastheofficialreligioninNorway
(Andersen1977,134ff).ThesagasFagerskinna
andHeimskringlarelatethatOlavfoundedthe
townofBorg(HelleandNedkvitne1977,212),
today’s Sarpsborg in eastern Norway. Apparently,Olavwasapersonwithwide-rangingideas
and he may have had the capacity to plan and
foundatowninwesternNorway.Thehistorian
Per Sveaas Andersen characterises the written
sources on Olav as the richest of any medieval
Norwegianking(Andersen1977,109).Thefact
that Olav is not mentioned in connection with
Bergen in the otherwise rich written sources,
maycountindisfavourofOlavasafounderhere,
butcannotbegivendecisiveweight.
Knutisnotknownasatownfounderinthe
writtensources.ThehistorianAlexanderRumblehaslisteddatablepoliticaleventsinthereign
of Knut, based on the written records (Figure
59). According to his studies there is no mentionofeventsbetween1028and1035(Rumble
1994),thismayindicatealacunaintherecords,
andthelackofmentionofKnutasafounderof
Bergenshouldnotbegivendecisiveweight.England,towhichKnutsucceeded,hadacomplex
urban network comprising towns with markets
andadministrativefunctions,asearlyasinthe
eleventhcentury(Hill1994,101).InhisEnglish
backgroundKnutcouldverywellhavefoundthe
inspiration for founding a town in other parts
of his empire. During the reign of Knut plots
were laid out in Lund in medieval Denmark
and this town was thus redesigned (cf Andrén
1980,49;Carelli2001,107ff).InDenmark,influencefromEnglandisseenintheorganisation
of the Danish church, in the introduction of a
royalcoinagebasedontheEnglishpatternand
probably in the administrative division of the
land (Skovgaard-Petersen 1977, 191-204; Lund
1994,27-46).ThereisnodoubtthatKnutwasa
kingwithwide-rangingideasandinitiativeand
assuchmayhavehadthecapacitytoplanand
foundatowninwesternNorway.
It seems that no decisive arguments can be
presented pro or contra Olav or Knut through
thekings’historicalcontexts,andtherelatively
widedateofabout1020/30fortheearliestdocumented activities on the plots cannot settle the
question. What is the central conclusion here
thenisthattheplan-probablyforatown-was
materialised and Bergen was probably founded
when the northern town area was divided into
plots by a king. According to the most plausible interpretation of the available sources, this
mostlikelyhappenedduringthefirstdecadesof
the eleventh century, that is in the years about
1020/30.
Whooccupiedandinvestedintheplots,who
were the ‘townspeople’? In order to elucidate
thesequestionsIshalldiscussthematerialacross
horizons 2 to 5. Several circumstances indicate
thatactivity andsettlementon the plots inthe
town area were established through bottom-up
initiative,andnotunderthedirectcontrolofthe
respectivekings,duringhorizons2to5.Firstof
allaccordingtothetrendsinthematerial,ittook
alongtimebeforeamajorityofthedocumented
plotsinthenorthernandmiddletownareaswere
actuallysettledandused.Thismayinitselfbe
anindicationthattheindividualplotusersfrom
horizon2throughhorizon5builtonplotswhen
theywantedand,atleastonsomelevel,ontheir
owninitiative.Furthermore,thereareexamples
thattheindividualplotswerebuilton/settledindependentlyofeachotherandnotaccordingto
anoverallplan,whenapassageandaquaystructure were built during horizon 4 on plot 6/C
(B)togetbetteraccessfromtherearpartofthe
plottothewaterfront,asimilarimprovementof
workingconditionswasobservedonplot26/Ain
themiddletownarea(B),usingadifferenttype
ofsubstructureasfoundation.Passagesandquay
frontsidenticaltothoseconstructedonplot6/C
duringhorizon4(c1100-1120s)werelaterbuilt
onseveralplotsalongtheVågenwaterfront,but
notuntilhorizon5(1120s-c1170)(sites6,27,28
and29).Theseexamplesdemonstrateindividual
bottom-up initiative from the plot users implyingthatprivateplotowners/usersasopposedto
thekingshouldbeseenbehindtheactualoccupationoftheplotsinthetownareafromhorizon
2throughhorizon5.
Butwhoweretheplotusers-thetownspeople? Or who were they representatives for? In
early medieval Norway, giving away land or
moneywasacommonwayofcreatingandsecuringalliances.KnutdenMektigeis,asarelevant
example,knownforhislargemonetarygiftsto
Norwegianlordspriorto1028(KLNM,IX2628). When the king during horizon 2 laid out
thenortherntownarea,hemayhavegivenplots
toinfluentialpeopleinreturnfortheirpastand
futureloyalty.Whenthekinginearlymedieval
Norway gave away land he also gave away the
allodialrightstothelandandthepieceofland
wouldthusstayinthefamilyofthereceiversfor
generations to come (KLNM, IX 26-28). This
may probably also apply to the period under
study here. According to written sources from
thehighandlaterMiddleAges,landinBergen
was then owned by private magnates, the king
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
223
Figure59.Selectlistofpoliticalevents,1024-42.Basedonwrittenrecords.(ModifiedfromRumble1994,Table1.1)
orbythemonasticinstitutions(Lorentzen1952,
76;Helle1982,78,281-284;Ersland1994,7577). Ersland has shown that along the Vågen
shoreline, in what corresponds more or less to
the northern and middle town areas, land was
mostlyinprivatehandsinthelateMiddleAges
(Ersland1989,241,249,271,279;Ersland1994,
Figure12,75ff).Duetoallodialrightsattached
toland,theownershipinthehighandlaterMiddleAgesmaygofarbackintime,perhapstothe
periodstudiedhere.
Returningtohorizon2onemayarguethatif
thekinggaveawayatleastsomeoftheplotsin
the northern town area during horizon 2, this
wouldexplainwhytheplotsinthehorizon2plot
systemseemtohavestructuredthewidthofthe
plotsinthesystemintroducedduringhorizon3
(cfp180ff).Infact,ifsomeoftheplotsinthe
northerntownareadidnotalreadyhavedifferentownersbeforehorizon3,(c1070-c1100)the
founderofthehorizon3systemcouldhavebeen
free to layout the northern town area without
having to consider the earlier system. AccordinglyIsuggestthatatleastsomeoftheplotsin
thenortherntownareaweregivenawayduring
horizon2.Theplotsmay,withreferencetolater
224
medievallandownership,havebeengiventoinfluentialpeople-inreturnforpastandpresent
loyalty.Thesepeopleortheirfamiliesmaystill
have owned the plots when a new plot system
wasintroducedinhorizon3inthemiddletown
areaandthenortherntownareawasre-regulated. The townspeople of early Bergen may thus
havebeenmagnatesortheirrepresentatives.
How did the townspeople receive the king’s
ideaofatown?Tracesofoccupationwerefound
onafewplotsduringhorizon2,butpressureon
buildinglandwaslow.Thescarcesourcesindicatethatafterthefoundationhardlyanymajor
initiativeswerecarriedintolifebythetownspeople. The king’s plans were apparently not well
received.
Anewmajorinitiative,horizon3
(c1070-c1100)
The story of how, when and by whom Bergen
wasfoundeddoesnotendhere.Anothermajor
initiative was taken some years later, when a
newplotsystemwasintroducedinthenorthern
andmiddletownareas(cfFigure33).Thedatingevidenceisnotquitesatisfactory,butdating
thenewplotsystemtohorizon3representsthe
best-sustainedalternativeatthepresentstateof
research(cfp183).
Themiddletownareawasnowprobablyparcelledintoplots.Inthenortherntownareathe
horizon2plotsystemwasrespectedintermsof
the width of the plots. The boundaries of the
shortsideoftheplotswere,however,movedand
focusseemstohaveshiftedtowardstheshoreof
VågenBay.Thenewtownplanmayperhapshave
includedspaceforachurchwhereStMary’s(site
23) was later built and for a thoroughfare, but
thisisnotsowellsubstantiatedandshouldmerely be considered as a hypothesis. Again, a king
shouldbeseenbehindtheinitiative;theBergen
areawas,asreferredtoearlier,mostlikelyroyal
property(Helle1982,77-79withreferences)so
itislikelythatonlythekingcouldparceloutthe
middletownarea.Thetimespanrepresentedby
horizon3correspondsmoreorlesswiththereign
ofOlavKyrre(1066-1093).AccordingtoHeimskringla, Olav Kyrre founded Bergen (sejta) (cf
p25ff).Thearchaeologicalsourcessuggestthat
Olavdidnotfoundthetownfromtheground,
ratherheinvestedfurtherinthetownscapewhen
parcelling out and including a virgin piece of
land in the townscape. Furthermore Olav built
Christchurch minor (site 3) and founded the
ChristchurchCathedral(site2)atHolmen,thus
includingthisareainthetownscape.
Alongthesamelineofthinkingaspresented
above Olav may, like his predecessor, have donatedplotsinBergentopresentandfutureallies
amongnativemagnatesorotherallies.According to Snorre Sturlason’s Heimskringla, Olav
gavelandtohisentrustedmanSkuleKongsfostreinBergen.ThislandstayedinSkule’sfamily
foryearstocome(Hkr1911,511;Holtsmarkand
Seip1975,584).SnorrewasfamiliarwithBergen
and Skule’s descendants and was probably well
informed on this matter, we should therefore
beabletotrustthesagahere(Helle1982,105).
This gives additional support to the suggestion
that Olav gave away land in Bergen to present
andfutureallies.
How did the townspeople receive the king’s
plansforatown?Accordingtothegeneraltrends
in the material assigned to horizon 3, most of
theoccupiedplotswerelocatedalongtheVågen
shoreline. Along Veisan some plots were most
likely still vacant, and at the foot of Fløyfjellet
settlement was only seen at one site. Along the
Vågen waterfront documented buildings were
constructedabovethetidalzone,indicatinglow
pressureonbuildingland.Inconclusion,oneis
left with the impression of little initiative from
thehandsofthetownspeople.Itseemsthatthe
king’s plans were not so well received by those
that were given a plot. With the king’s investmentsatHolmenandhisdivisionofthemiddle
townareaintoplots,themaininvestmentsinthe
new town were apparently in the hands of the
king.
Horizon4(c1100-1120s)
During horizon 4 several monumental buildings were initiated at Holmen as well as in the
northernandmiddletownareasandatNordnes.
Horizon4coincidesmoreorlesswiththereign
ofØysteinMagnusson(1103-1123)andSigurd
Jorsalfar(1103-1130).Øysteinhasbeenascribed
asthefounderoftheChurchoftheApostles(site
4)andalargetimberhall(site5),bothatHolmen. He is also ascribed as the founder of the
Munkeliv Benedictine abbey at Nordnes (site
43) and as the possible founder of the Church
ofStNicholasinthemiddletownarea(site23).
WiththefoundationofMunkelivonemayargue
that the Nordnes area was added to the townscape.Thepossiblepredecessorofthestanding
St Mary’s (site 23) and the pier at site 14 may
possiblyalsobelongtohorizon4.Accordingto
archaeological investigations the pier represents
thefirstofmanygenerationsofathoroughfare,a
street.Everytimethestreetwasrebuiltormaintained, construction work was carried out accordingtooneoverallplanandinoneturn.The
thoroughfareandconstructionstoeithersideof
thestreetwerebuiltindependentlyofeachother.
Thissuggeststhatthestreetwasacommonthoroughfare, an allmenning, administrated by the
‘public’asopposedtoprivateindividualowners
(Marstrander 1983). In later sources the commonthoroughfaresinBergenarereferredtoas
the king’s patrimony (NgL III 25; Helle 1982,
79, 282). Seen together with the archaeologicalobservations,thismaysuggestthattheking
should be seen behind the construction of the
common thoroughfare and its predecessor, the
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
225
pier.Asweshallseebelow,thestandingChurch
of St Mary may be tentatively associated with
a royal initiative. If a king was involved in the
foundationofthestandingStMary’sduringhorizon5itislikelythatakingwasalsoinvolvedin
thepossiblepredecessor.
St Nicholas’, the pier and the possible predecessor to the standing St Mary’s are assigned
to horizon 4 as supplementary sources. Even if
oneorallofthesesourcesareerroneouslyassociatedwithhorizon4,itisstillwell-documented
through the remaining sources that further investmentsinmonuments,institutionsandonthe
infrastructureofthetownweremadeinBergen
ontheking’sinitiative.
Regardingmajorinitiativestakenbythetownspeople,someplotsalongVeisanwerenowoccupied,andalongtheVågenshorelineoccupation
wasindicatedonmostplots.Withintheplotsin
thenortherntownareatherewasapparentlylow
pressure on building space, as the documented
buildingswereallconfinedtotheareaabovethe
tidalzoneorhadjustbarelycrossedintothetidal
zone. In the middle town area there may have
beenpressureonbuildingland.Onsomeplots
along the Vågen shoreline substantial improvementsofworkingconditionsonthebeachwere
carriedout,andinsomecasespassagesandquay
structureswerebuilt.Inthenortherntownarea,
at the foot of Fløyfjellet, there was settlement
but still ample building space. Well-established
householdsofapermanentcharacterweredocumentedonseveralplots.Allinall,itseemsthat
pressureonbuildinglandinthetownareaand
withintheplotswasnotintense.However,investmentsweremadetoimproveworkingconditions
inthetidalzoneontheVågen-boundplotsand
permanenthouseholdswereestablishedthroughoutthenorthernandmiddletownareas.
Tosumup,thekingnowinvestedfurtherin
Bergen,foundingecclesiasticinstitutionsandincluding the Nordnes area in to the townscape.
Townspeopleontheirsideinvestedinthetown
plotsbyestablishingpermanenthouseholdsand
improving working conditions by the Vågen
shoreline.
226
Horizon5(1120s-c1170)
Theperiodrepresentedbyhorizon5fallsmore
or less within the time of the civil wars where
joint kings and claimants to the crown fought
eachother.Whendiscussingtheinitiativesofthe
kinginthefollowingitisonlypossibletopina
nameonthekinginafewcases.
Twomonasteriesandachurchwerefounded
in the Nordnes and the Nonneseter areas and
settlement was documented in the southern
townareaforthefirsttime.Thesoutherntown
areaandtheNonneseterareaswereinthisway
included in the townscape. Furthermore, seven
churcheswerebuiltorrebuiltduringhorizon5.
As in the northern and middle town areas it is
likelythatthekingownedtheNonneseterarea
and the land in the southern town area (Helle
1982, 77ff with references), and therefore only
the king could possess the authority to include
these areas into the townscape. The Church of
StOlavontheHill(site25)wasbuiltbyKing
Harald Gille (Gilchrist) after his victory over
KingMagnustheBlindin1134-1135(cfp130).
Regarding the initiators behind the remaining
monumentalconstructionsthesourcesarevague,
butsomesuggestionscanbemade.
Basedonrecordsoflatermedievallandownership, the Nonneseter Convent (site 46) was
theninposessionoflandformerlyownedbythe
king. Helle suggests that the convent may also
have been founded by royal initiative (Helle
1982, 140) I will follow this interpretation.
StJohnsabbeyonNordnes(site44)wasofthe
Augustianorder,andithasbeensuggestedthat
theabbeywasfoundedasasupportforthecathedralchurchatHolmen(Helle1982,142with
references).AstheChristchurchCathedral(site
2)wasunderconstructionduetoroyalinitiative,
itwouldseemlikely,ifweholdasapremisethat
theaugustinianabbeywasfoundedasasupport
forthecathedral,thatthekingwasalsoinvolved
inthefoundationofStJohn’s.
Basedonthesizeofthechurchyard,thesize
ofthechurchbuildings,theelaboratewestfront
of the standing St Mary’s (site 23) and a possiblelargewestfrontonStCross(site40),ithas
been suggested that the king, in collaboration
with the townspeople founded (the standing)
StMary’sandStCross(Lidén1993,78).Ihave
arguedthatspaceforachurchwhereStMary’s
was later built may have been part of Olav
Kyrre’stownplan/theplotsystemintroducedin
horizon3.Thissuggestionisnotsowell-founded
empirically,butwouldcertainlysupportthenotionthatthekingwassomehowinvolvedasan
interested party when the standing StMary’s
waserected.Thecorrecteast-westorientationof
StCrossimpliesthatthechurchwasconstructed
whiletherewasstillamplespaceforthebuilding
and its churchyard, and this may perhaps suggest that St Cross was part of a superior town
planwhenthesoutherntownareawasincluded
inthetownscape.Ifso,thismaysupporttheidea
thatthekingwasinvolvedalsoasafounderalso
ofStCross.Alongthesamelineofthinkingthe
orientationofStOlav’sinVågsbunnen(site39)
might indicate that the church was part of an
initialplanforthistownarea.Therefore,ifthe
town area, as suggested here, was included in
thetownscapebyinitiativeoftheking,theking
mightwellbeassociatedwiththischurch.The
fact that the king was probably still the owner
oftheareaaroundStOlav’sinthehighandlate
MiddleAges(Helle1982,78;Ersland1994,Figure12,75ff)maysupportthisnotion.Inaddition,datafromsite38suggestthattheareaby
StOlav’swasusedunderstrictcontrol;nogarbage was dumped here during the first phases
on the site. Furthermore, a piece of jewellery
madeofcutquartzcrystalandfoundintheconstruction layers of the phase following horizon
5,mayhavebelongedtoapersonofhighsocial
status(Komber,Dunlop,Sigurdsson,andHjelle
1994,216).Allinall,thesourcesmayimplythat
StOlav’s in Vågsbunnen was founded involvingroyalinitiative.AllSaintsintheNonneseter
area(site45)(S)ismentionedasaroyalchapel
inlatersources(Helle1982,145withreferences)
and may therefore have been founded on royal
initiativeperhapsasearlyashorizon5.
LidénsuggeststhatStColumba(site33)(S)
andStPeter’s(site24)(S)werebuiltascorporate
churches.Thesuggestionisbasedonthename
of the patron saint for St Columba and on the
incorrect orientation of St Peter’s (Lidén 1993,
79).ThecircumstancethatStPeter’sapparently
was built on two ‘model plots’ in the northern
townarea(cfp180ff)showsthatthechurchwas
notpartoftheplotsystemthatwasprobablylaid
outinhorizon3bytheking.Thismaysupport
thatthechurchwasbuiltbyoneormore‘private’
founders. Since both churches are classified as
supplementarysourcesforhorizon5,nostrong
conclusions can be made as goes the erection
of ‘private churches’ during horizon 5, and the
questionofthefoundersofStPeter’sandStColumba’scannotbesettledonfirmevidence.
Evenifsomeoftheinitiativesascribedtothe
kingabovewerenotactuallyassociatedwiththe
king,butratherwerearesultofbottom-upinitiativesitoughttobeatrustworthytendencyin
thematerialthatconsiderableinvestmentswere
madeinthetownbytheking(s)duringhorizon
5.Atop-downinitiativewithaconnectiontothe
kingmaythusbearguedfortheinclusionofnew
landintothetownscapeandforthefoundation
ofseveralofthemonumentsknownfromthehorizon.Abottom-upinitiativeor‘private’founders may perhaps be seen behind at least two of
thechurches.
From the hands of the townspeople, secular
settlementwasseeninthewholetownarea(Figure27andFigure39).AlongtheVågenshorelinealmostalltheinvestigatedplots/sitesinthe
northernandmiddletownareasnowappearto
havebeenoccupied.Onlyatsite17inthemiddle town area, was there a lack of traces of occupation. Along the Vågen shoreline buildings
were constructed throughout the whole length
of the plots, and the tidal zone was taken into
use.Thisshowspressureonbuildinglandhere.
Settlement had expanded into the Vågen basin
seekingdeeperwateronseveralplots,achieving
newbuildingspaceandindicatinghighpressure
onthebuildingland.AtthefootofFløyfjellet,
secular settlement had expanded some in the
northern town area but there were still major
open areas with vacant building land. Between
site30andthechurchesofStNicholas(Site32)
and St Columba (Site 33) the few sites investigatedwerenotsettled,indicatingthatalsothis
areawasstillvacant.
In the southern town area settlement was
onlydocumentedatonesite.Thesourcesinthe
southern town area are mostly well-founded,
buttherepresentativityofthesourceshereis,as
discussedearlier(cfp157),notsatisfactory.The
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
227
orientationofthechurchesinthispartoftown
mayindicatethattheareawasnotdenselybuilt
onwhenthechurcheswereestablishedinhorizon5.Thismayinturnindicatethattherewas
notmuchpressureonbuildinglandinthispart
of town. Well-established households of a permanentcharacterweredocumentedinallthree
townareas.
Toconclude,theking(s)seemtohaveinvestedfurtherinthetownbyaddingtwonewareas
tothetownscapeandfoundingseveralecclesiastic institutions. The townspeople now invested
more extensively in their town plots. From the
handsofthetownspeopleanintensifiedpressure
onsecularbuildingspacewasseeninthenorthernandmiddletownareas,especiallyalongthe
Vågenwaterfront.Thetownspeopleperhapsalso
investedinchurches.
Conclusions
Thestoryofhow,whenandbytheinitiativeof
whomcontainsasuccessivechainofmajorevents.
Bergenwasprobablyfoundedthroughroyalinitiativeonlandwhereagriculturalactivitieswere
carried out. At the present state of research it
seemsmostlikelythatthishappenedwhenplots
werelaidoutinthenortherntownarea,inthe
yearsaround1020/30.Theideaofatownwas,
itseems,notwellreceivedbythe‘townspeople’
- probably magnates - who were given plots in
theplannedtown.Thetownareawasthusbarely
taken into use. During horizon 3, corresponding to the reign of Olav Kyrre, it appears that
the northern town area was redesigned, Holmen and the middle town area were added to
thetownscape,andmonumentswereinitiatedat
Holmen. Plots in the town area were probably
given to native magnates or other allies of the
king, but again the idea of a town was apparentlynotsowellreceivedbytheusersofthetown
plots;itseemsthatpressureonbuildinglandwas
sparseinthetownarea.Duringhorizon4,ØysteinMagnussonbuiltaroyalhallatHolmenand
foundedseveralecclesiasticinstitutionsincluding
oneatNordnes,therebyaddingthisareatothe
townscape. The townspeople were now getting
more active on the plots and well-documented
settlementsofapermanentcharacterwereestablished.Duringhorizon5thesoutherntownarea
228
andtheNonneseterareawereaddedtothetownscape,andtwomonestariesandperhapsasmany
aseightchurcheswerebuilt,mostoftheseinitiativeswereprobablybytheking.Thetownspeopleinvestedmoreintensivelyintheirtownplots
andperhapsalsobuiltsomeofthechurches.Itis
interestingtonoticehowthetowngrewinsteps
andexpandedphysically,asstillmoreareaswere
includedinthetownscape(Figure60).
WhywasBergenfounded?Andhow
didthetowndevelop?
Why was the town founded and why did the
kings,andintimealsootheractors,investinthe
town?Whenlistingimportantdeedscarriedout
bykingsorotherimportantactors,themedieval
chroniclers always emphasise towns that were
founded or strengthened, churches and monasteriesthatwerebuiltorreceivedlargegifts.There
shouldbenodoubtthat,inadditiontopractical
reasonsforsuchinvestments,prestigeaswellas
otherformsofsocialprofitwereimportantmotives when founding a town and investing in
monuments. For the townspeople investing in,
orperhapsjustlivingonaplot,atownplotmay
alsohavehadmorethanpracticalpurposes;the
townmayforinstancehaveattractedpeoplein
searchofadifferentlifestyle.
Fortheking,foundingandinvestinginatown
mustbeseenonapracticallevelasoneofseveralmeansinalargerplantocentraliseactivities
and/or functions that hitherto were decentralised,ortointroducenewactivitiesorfunctions.
Suchactivitiesandfunctionsmayhavebeeneconomic, jurisdictional, administrative, religious
orcultural(cfp20),militaryfunctionsmayalso
have been relevant. The motives for founding
and investing in a town vary according to the
historicalcontextoftheinitiators,butitislikely
that‘thelargerplan’atanytimewouldbeadvantageousfortheinitiator.Forthetownspeopleusingaplotinthenewtownmayhavebeenanasset.Orperhapsthekingmayhaveimplemented
strongincentivestoencouragetheplotownerto
takeatownplotintouse.Theking(s)musthave
had sufficient resources not only to materialise
thephysicalinfrastructureoftheplannedtown,
Figure60.Areasincludedinthetownscapefromhorizon2throughhorizon5
butalsotomakepeopleusethetown.Positiveas
wellasnegativemeansofenforcementmayhave
beenused.
InordertoelucidatewhyBergenwasfounded
and invested further into the intended and actualfunctionsofthetownarediscussed.Focusis
mainlyonthepracticalfunctions.Theintended
functionsarethosethattheking(s)hadplanned
for the town, whereas the actual functions are
thosethatwerecarriedintolifebytownspeople
orvisitorsofthetown.Itisreasonabletoassume
thatthetownspeopleintheearlyyearsusedthe
plotsinaccordancewiththefunctionsintended
bytheking(s).Intimethetownmayhavebegun
tolivealifeofitsownandnewfunctionsmay
havebeenintroduced.
Iholdasapremisethatsourcesthatcanbeassociatedwithroyalinitiativereflecttheintended
functionsofthetown,whereasinitiativescarried
outbytheusersoftheplotsanddailyactivities
carriedoutinthetownreflecttheactualfunctions of the town. Major initiatives and daily
activitiescarriedintolifebytheactorsareseen
inrelationtotheactors’contemporaryhistorical
context.
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
229
ThenewtownintheBergenarea,horizon2
(1020/30-c1070)
What can be understood about the intended
and actual function of the planned new town
in the Bergen area through the sources for horizon2(1020/30-c1070)?Occupationhasonly
beendocumentedonafewplotsorunitsandthe
sources could not provide a reliable picture of
dailyactivitiesactuallycarriedoutthere.Inother
words,tracesofsettlementandactivityassigned
tothenortherntownareainthishorizonareso
vaguethatwecanhardlyestablishhowthearea
actuallyfunctionedinapracticalsensethrough
thesesources.Iwillthusturnto‘circumstantial
evidence’ that may shed some light on the intendedfunctionoftheareainhorizon2.These
are: (1) The circumstance that at least some of
theplotswereprobablygivenawaytomagnates,
(2)thephysicallayoutoftheplotsinthehorizon
2systemandthebottom-upuseoftheplots,and
(3) the townspeople’s investments in improved
workingconditionsonthewaterfront.
Ad (2): I find that the facts, that palisade
fencesdelimitedtheplotsandthesuggestedbottom-upor‘private’useoftheplots,bothimply
thatthefunctionsoftheplotswereintendedto
include ‘private’ activities or activities carried
outbytheindividualplotowner-asopposedto
publicorofficialbusinessoractivitiessupervised
closelybytheking,orintendedtoservetheking
directly.Itisthereforeunlikelythatthenorthern
townareawasregulatedforforexamplemilitary
purposes.
Ad(3):atfirstsightthehorizon2plotsystem
appears to be directed towards the safe natural
harbouroftheVeisaninletandgivestheimpressionthataccesstotheseawasimportant.Howeveraswehaveseenearlier,thelandingconditions in Veisan were, probably unfavourable
for larger ships. May this imply that transport
of goods by boat was considered unimportant
whenfoundingthetown?Thisishardlyatrustworthyimplicationfromthematerial.Because,
whenthelocationofafuturetownwaschosen
onecould,mostlikely,notpickoutjustanypiece
ofland.Thecircumstancethattheking,aswe
haveseen,probablyownedtheBergenareaand
thefactthattheroyalestateAlrekstadlayclose
bymayhavebeenmoreimportantforthelocali230
sation of Bergen than a harbour with optimal
landingconditions.Furthermore,longtraditions
forusingtheVeisaninletasaharbourmayhave
existed;theinletmaystill,oruntilrecently,have
servedasalandingplaceforthesuggestedagrariansettlementhere.Therelativelypoornatural
landingconditionsinhorizon2shouldthusnot
be given too much emphasis when judging the
king’splansforthetown.
Accordingtothemaintrendinthematerial,
occupationduringhorizon2wasmainlylocated
bytheshoresofVeisanandVågen.Thejettyat
plot 6/D indicates that access across the tidal
zonetothewaterfrontwasconsideredimportant
bytheusersoftheplot.Assumingthatinthebeginningthetownspeopleusedtheirplotsinaccordancetotheking’splans,theirinvestmentsin
betterworkingconditionsbythewaterfrontmay
reflecttheking’sinitialplansforthetown.
Thus,thefollowingcanbeinferredaboutthe
intendedfunctionofthetownplotsinhorizon2:
1)Theplotswereprobablygoingtobeusedby
magnates.2)Activityofaprivatecharacterwas
goingtobecarriedoutthere.3)Goodworking
conditionsonthewaterfrontwereimportantfor
theactivities.InadditionIholdasapremisethat
thekingwouldbenefitfromsuchactivitiesifit
wascentralised.
Christophersen has suggested that the kings
oftheearlyNorwegiancentralmonarchysought
tocontroltheredistributionofgoodstraditionallycontrolledbythelocalelite.Bytransferring
theredistributionofgoodsfromthelocalelite’s
staples to new urban centres the king would
bothweakenthelocaleliteandhecouldgethis
share of the wealth (Christophersen 1989, 129,
144).Thishypothesispresupposesastrongcentral king and the use of negative means of enforcement.Othershavesuggestedthattheking
attractedmerchantstotheearlytownsbyofferingplotsandprotectingmarketpeace.Theking
wouldprofitfromthisbycollectingduesinreturnforprotection(egSkovgaard-Petersen1977,
140ff;Ros2001,19).Thishypothesisoffersan
example of a positive means of enforcement.
The two hypotheses may be regarded as exponents of different ways of understanding social
change(cfp33ff)andthereisadisagreementin
theirunderstandingoftheking’sroleandmore
subtlemotivesforthetownfoundationandhis
meansofenforcement.Butbothhypothesessee
‘the redistribution of goods’ or trade as importantfunctionsoftheearlytowns.Ahypothesis
thatBergenwasfoundedbythekingasacentral
placeforlong-distancetradetobecarriedoutby
thelocalelite,findsanechointhe‘circumstantialevidence’,thatisthesuggestedprivateownershipofthetownplots,theprivatecharacterof
activitiesintendedforthetownplots,andinthe
townspeople’sinvestmentsinimprovedworking
conditionsonthewaterfront.AccordinglyIsuggestthatoneoftheking’sintentionswastoestablishacentralstaplewheregoods,disposedof
byindividualplotowners-probablynativemagnates-werecollectedandenteredinanational
orinternationaltradingnetwork.Soapstonevessels,darkgreyschisthones,andHyllestadquernstonesfromwesternNorway(Mitchell,Askvik,
and Resi 1984; Myrvoll 1986; Christophersen
1989; Jensen 1990; Carelli 2001; Baug 2002)
mayhavebeenpotentialgoods.Thekingcould
benefitfromthisarrangementbycollectingdues
ontheprotectionofmarketpeaceortrade.
In addition, the king could benefit from a
centralisationofthecollectionoftheking’sown
dues and veitsler to Bergen (cf p 21ff). Surplus
from such incomes could be entered in an international trading network when shipped out
from Bergen. If the pier in analytic unit 30/A
represented a landing-place for the royal estate
atAlrekstaditmayalreadyhavefunctionedasa
landing-placeforroyalincomespaidinkindand
itmaywellhavebeenintendedtohaveanewrole
as theking’slanding-placebythetownofBergen.Unfortunately,theavailablesourcescannot
revealtheintendedoractualfunctionofthepier
andassociatedconstructions,sothefunctionof
thepierinrelationtotheplannednewtownis
merelyhypothetical.
If we consider the location of Bergen on a
macroscale,theareawaswelllocatedtoserveas
acentralstapleforwesternNorway.Bergenwas
closertoEnglandandthecontinentthanTrondheim.Thekingmayhaveseentheadvantageof
atownclosertoEuropeandperhapsalsoatsome
distance to Trondheim and Trøndelag, where
thepowerfulLadeearlsresided.Intheeleventh
centuryVikingraidsasawaytoraiseanincome
hadtobereplacedbyothermeansforkingsor
pretenderstothecrown.Theearlycentralkings
thus needed to find new ways of raising an incometosecuretheirpositioninsociety(cfBagge
2002, 204-207). Investing in a new town in
western Norway may have been an attempt to
raiseanincome.
Thecircumstancethattheplannedtown-apparently-wasbarelytakenintouseduringhorizon2callsforfurtherdiscussion.Apieceofland
dividedintoplotsisnot‘atown’untilpeopleuse
itasasuch.WhateverthemotivesbehindfoundingBergen,thefoundationmusthavebeenjust
one of several initiatives planned by the king,
becausethesuggestedcentralisationoffunctions
thatweretraditionallyinthehandsofmagnates,
wouldhavetoincludeanewsetofrulesforthese
activities. However, considering that the plans
according to the trends in the material were
notarealsuccessitisprobablethatthefoundingkingdidnotpossesssufficientresourcesfor
thenewrulestobeeffectedoraccepted.Theperiod around 1020/30 was rather turbulent and
neither Olav nor Knut were in power for long.
Theremaynothavebeenenoughtimefornew
traditionstobeestablishedwithinthetimespan
of Olav’s or Knut’s reigns respectively. Furthermore, the young King Magnus Olavsson, who
succeededKnut,wasunderthetutelageofEinar
TambarskjelveoftheLadeearls.Thisstrongassociation with Trøndelag and Trondheim may
have halted further investments by the king in
theBergentownproject.
To sum up, I suggest that the king about
1020/30plannedandfoundedatowninwestern
Norway. The hypothesis that one of the towns
plannedfunctionswastoserveasacentrewhere
magnates and the king could have goods in a
long-distance trading system finds some supportinthesources.Itissuggestedthattheking
plannedtocollectduesinreturnforprotection
ofmarketpeaceorasataxontrade.Theking,
however,didnothavethesufficientresourcesto
carryhisplansintoeffectwithintheshorttime
spanofhisreign.
OlavKyrre’sBergen,horizon3(c1070-c1100)
When Olav Kyrre invested further in Bergen
duringhorizon3.Whatwerehisintentionswith
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
231
thetown?Andhowdidthetownactuallyfunction?OlavinitiatedtheconstructionofChristchurchMinorandtheChristchurchCathedralat
Holmen,here-designedthenortherntownarea
andincludedthemiddletownareainthetownscape.
TheinvestmentsatHolmenhavebeenseenas
partofaplanforBergentobecomethebishop’s
residence and an ecclesiastic centre for western
Norway (Helle 1982, 90 with references). The
combinationofroyalseat/bishop’sseatiscommonintheNordiccountries,inOrkney,andin
Ireland.Againstthisbackground,Lidénsuggests
thatOlavhadintentionsofbuildingaroyalresidenceatHolmeninadditiontothechurcheshe
initiated there (Lidén and Magerøy 1990, 10).
If the pier at site 30/A was originally a landing-place for the royal estate at Alrekstad, and
ifthepierwasintendedtofunctionastheking’s
landing-place during horizon 2 as suggested
as a hypothesis above, then the fact that Olav
Kyrre’splanprobablyincludedthepierintothe
townscapeduringhorizon3lendssomesupport
to Lidén’s hypothesis. Because if the royal pier
wasincludedintothetownscape,thefunctions
ofthispiermostlikelywouldhavehadtoberelocatedtosomewhereelse-Holmenbeingalikely alternative. Again, the sources are too vague
foranystrongconclusionstobemade.However
theymaylendsometentativesupporttothehypothesisthatOlavalsohadaroyaladministrative
centreinmindwheninvestingfurtherinBergen.
Ifso,hemayalsohaveplannedforBergentobe
thekingsstapleforroyalduespaidinkind.
Olav also invested in the town area: I have
suggested that, in the horizon 3 plot system,
spacemayhavebeenreservedforastreetandfor
achurchwhereStMary’swaslaterlocated.The
plotforStMary’swasclearlylocatedinthemost
centralplaceinthenortherntownareaandmust
have been intended for a ‘town church’ as opposedtothechurchesatHolmen.Theexistence
ofspaceforthestreetandachurchalreadyduringhorizon3isnotsowell-foundedempirically,
sotoomuchemphasisshouldnotbeplacedon
thesesources.Nevertheless,itseemswell-foundedthatthenortherntownareawasre-designed,
andwiththelayoutofthemiddletownareathe
townalsogrewconsiderably,andnowextended
232
atleasttotheareathatwasoccupiedbythepreurban landing-place at site 30. It is thus clear
that Olav invested in the town area and must
havehadplansforthetownareainadditionto
the plans for a bishop’s seat and the suggested
royaladministrativecentre.
IhavearguedabovethatthetownplotsinOlav
Kyrre’sBergenwereprobablygivenawaytoinfluentialallies,mostlikelynativemagnates.Butnot
allplotswereoccupiedandtheywerenotusedin
asimilarway.Ihavesuggestedthattheplotswere
intendedfor‘private’activitiesasopposedtoactivitiesdirectedbytheking.Olav’splotsystem
wasseeminglydirectedmoretowardstheVågen
Baythantheoldersystemandtheneedformore
plots with better landing conditions may have
triggeredtheplanningofthemiddletownarea
wheninvestingfurtherinBergen.Again,investinginatownmustbeseenaspartoftheking’s
larger plan to centralise new or old functions.
Followingasimilarlineofthinkingasabove,the
focusonfavourablelandingconditionsandthe
private character of activities to be carried out
ontheplots,maysuggestthatOlavlikehispredecessorsplannedBergenasacentralstaplewhere
goods, disposed by magnates, could be entered
inalong-distancetradingnetwork.Accordingto
ecclesiastic rules bishops should have their seat
inatown(Helle1982,111).Olav’sinvestments
inthesecularpartsoftownthereforecorrespond
wellwithagreaterplanforBergentobecomea
bishop’sresidence.
Howdotheking’ssuggestedplanscorrespond
to the actual function of the town? Settlement
andactivitytracesinthetownareaarevaguein
horizon3,butaswehaveseensomeoftheplots
inthenorthernandmiddletownareasweresettled.Thedistributionofoccupiedplotsindicate
thattheVeisan-boundplotswere‘secondchoice’
asopposedtoVågen-boundplotsandthatbetter
landing conditions were considered important
bythetownspeople.Thepriorityoflandingconditionsmayindicatethatseatransportingeneral
waspartofthetownspeople’sstrategyforusing
theirplotsinBergen.
AtHolmenprofessionallarge-scalewoodworkersandtosomeextentalsolarge-scalestoneworkersmusthaveworkedonthemonumentalsites.
These people probably belonged to the king’s
household and their presence must have had a
certain synergetic effect on the development of
thetown.Wheretheworkerslived,andtheiractualimportanceforthegrowthofthetowncan,
however,notbeestablished.Thesourcesforthe
actualfunctionofthetownareaarevague,hintingthatseatransportwasconsideredimportant
foractivitiescarriedoutbythetownspeople.
Tosumup,itissuggestedthatOlavKyrreinvestedfurtherinBergen,planningthetownto
becomeanecclesiasticandperhapsalsoaroyal
staple/administrative centre. He may also have
planned the town as a central staple for goods
disposedbymagnatesandenteredinalong-distance trading network. The scarce sources implythatseatransportwasimportantforusinga
townplot.Nofactualactivitiesthatcouldserve
as a fundamental economic basis for the town
have been recorded. The limited extent of the
activitiesinthetownareaimpliesthattheroyal
plansforthetownareawerenotsuccessful.The
townspeopleapparentlydidnothavestrongincentivestouseaplotinthenewtown.Eitherthe
kingdidnotfollowuphisplansforactivitiesin
the town area with the necessary means of enforcement,orhedidnotpossesstheresourcesto
applythenewrulesanddevelopnewtraditions
forusingthetownplots.
Bergenduringhorizon4(c1100-1120s)
Aswehaveseen,thekinginvestedsubstantially
in Bergen during horizon 4 (c 1100-1120) as
well, and the townspeople started using plots
to a wider extent than before. What were the
king’splansforthetown?ØysteinMagnusson’s
constructionofthegreathallatHolmenshows
that a royal residence, and along with it probablyroyaladministrativefunctions,werenowlocatedinBergen.Hethusintroduced-orperhaps
strengthened already existing – royal functions
of the town. The foundation of the Church of
theApostlesatHolmen,theMunkelivAbbeyat
NordnesandprobablyStNicholasinthemiddle
town area suggests that Øystein also wished to
strengthenBergenasanecclesiasticcentre.The
locationoftheChurchofStNicholasinthegeographicalcentreoftheexistingtownareamaybe
Øystein’s contribution to the town plan or the
town’sinfrastructureasonemayarguethatthe
church ties the northern and the middle town
areas together across the natural topographical
barrier of the protruding rock on the coast of
Vågen.StNicholaswasthepatronofsailorsand
merchantsandthischoiceofpatronsaintforthe
church may imply that Øystein planned to encouragetradeinBergen(KLNM,XII288-291;
Helle1982,116).
Why so many churches? Not only Øystein
but also the other twelfth century kings probablyinvestedstronglyinecclesiasticinstitutions.
Howcanthekings’investmentsinchurchesand
monasteriesinBergenbeinterpreted?Ithasbeen
argued that the many new churches built duringthetwelfthcenturyinBergenbearwitnessto
thestrongphysicalexpansionofthetown(Helle
1982,149;Helle1992,26).Herethereisapresuppositionthatthechurcheswerebuiltas‘parish’churchestoserveanalreadyexistingpopulation. The limited pressure on building land in
the town areas up until horizon 4 and, in the
caseofthesoutherntownarea,probablyalsointo
horizon 5, however, suggests that the churches
werebuiltwhilethetown’sphysicalextent,and
thusprobablyalsoitspopulation,wasstillquite
limited. Accordingly, the churches cannot have
been built to serve an already existing population. Lidén has discussed the original status of
thechurchesbuiltorre-builtduringhorizon5.
Onthebasisofthesize,shapeandlocationofthe
churcheshesuggeststhatStMary’sandStCross
wereprincipalchurchesusedbythetownspeople
ingeneral.Theotherchurchesmighthavebeen
built as votive churches (St Olav’s on the Hill
wasbuilttofulfilavow)oraschurchesforindividualsorgroupsofpeoplewhowishedtohave
theirownplaceofworship(LidénandMagerøy
1990,17ff;Lidén1993).BothHelleandLidén
thus implicitly suggest that the churches were
builtmainlytoserveasplacesofworship.
As we have seen, Øystein invested in one or
perhapsasmanyasthreechurchesandanabbey,
andinhorizon5theking(s)investedinperhaps
asmanyasfivechurchesinadditiontotwoabbeys.Itislikelythattheremayhavebeenother
more subtle motives behind the initiatives than
mere dedication to Christianity and the aim to
providethetownspeoplewithaplaceofworship.
With references to Torstein Veblen’s theory
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
233
of ‘conspicuous consumption’ the archaeologist
A Jan Brendalsmo suggests that building large
stonechurchesinTrøndelagintheeleventhand
twelfthcenturieswasawayforthechurchfounderstoshowofftheextentoftheirsocialposition
or capacities (Brendalsmo 2001, 262ff). The
many twelfth century churches (mainly stone)
in Bergen were built at a period of time when
severaljoint-kings,rivallingkings,andpretenderstothecrownnaturallyhadastrongdesireto
showofftheirsocialcapacities.Itmaytherefore
notbeirrelevanttoexplainthemanychurchesin
Bergeninrelationtoconspicuousconsumption.
However there may also be additional explanations.
Untilatleastthemiddleofthetwelfthcentury
thechurchfounder/patron(privateortheking)
had great influence on matters concerning the
electionofpriestsandperhapsmoreimportantly
onmattersconcerningthedisposaloflandand
income under the church (Helle 1995, 22-23;
Krag 1995, 201-203). The churches may have
had incomes from land (Helle and Nedkvitne
1977, 221; Helle 1982, 151) and after the first
decadesofthetwelfthcenturyperhapsalsotithe
(Andersen1977,335ff).Onemayarguethat,if
thekingestablishedchurchesandotherecclesiasticinstitutionsinthetownandgavethemland
tocollectincomefrom,hecreatedanewgroup
oflandowners.Sincetheselandownershadtheir
basis in the town they would probably use the
town as a staple when canalising surplus from
dues into a long-distance trading system. This
wouldinturnstrengthenthetown’spositionas
astapleandacentreofcommercialactivitiesinvolvingamongothertheexportofstockfishfrom
northernNorwaytoEurope(seefurtherbelow).
Establishing churches in the town could thus
betheking’smeansofenforcementtomakethe
wheelsgoround.Explanationsofthegreatecclesiasticinvestmentsfromthekings’sidemayalso
befoundalongthislineofthinking.
This explanation finds parallels in Anders
Andrén’s ‘congested countryside’ theory, which
onamoregenerallevelappliestoScandinavian
towns that emerged between c 1000 and 1150.
Andrén sees the erection of so many churches
in Scandinavian towns during this period as a
productofthecentralkings’de-centralisationof
234
the right to execute sovereignty (Andrén 1985,
77-81;Andrén1989).Thekings’investmentsin
Bergen, including the foundation of the many
churchesandotherecclesiasticinstitutionsinthe
town,shouldnotbeseeninisolation.Onamore
generalleveltheyshouldbeseenasinvestments
madeinconnectionwithoverallstrategies,where
probablyastrengthenedpositionforthecentral
king(s)iscrucial.
Returning to Bergen and the period represented by horizon 4, Øystein is known as the
founderofchurchesandothermonumentsalong
thesailingroutebetweenBergenandthefishery
districtsinnorthernNorway.Thishasbeenseen
asaninvestmentinfacilitiesandinfrastructure
between the rich fisheries in the north of the
countryandBergen(Helle1982,116,note78).
ØysteinandhisbrotherSigurdJorsalsfartaxed
people who went fishing in Vågan in northern
Norway(F,XVI2;Helle1982,116)andobviouslyhadvestedinterestsindevelopingthefishing activities. In the light of this, Øystein’s investmentsinBergenmayalsobeseenaspartof
a plan to strengthen and control the export of
stockfish to Europe. In addition, following the
arguments presented for the earlier horizons, it
maystillhavebeenintheking’sinterestthatBergenfunctionedasastapleforgoodsinthehands
ofmagnates.
Doestheactualfunctionofthetownareacorrespondwiththeking’ssuggestedplans?Asseen
above, building activities on the town plots indicatethatpressurewasnotintenseonbuilding
space from the townspeople; substantial efforts
toimproveworkingconditionsontheshore,and
accesstothewaterfrontare,however,discerned.
The first tool of trade was found in horizon 4
andthetrade-indicatingsourcesimplythattrade
withbulkcommodities,bothinternationally-as
indicatedbypotteryfromtoday’sLowCountries
andGermany-andalongtheNorwegiancoast,
wasnowpartofthetownspeople’seconomy.The
saga’s description of Ragnvald Kale’s journeys
adds English harbours to the list of places that
hadcontactswithBergen.Tradefromthetown
plotsmayreflectthatgoodswerenowtransferred
throughthemagnates’townplotinBergenand
into a long-distance trading system. Potential
goods may, as earlier, have been soapstone ves-
sels,darkgreyschisthones,andHyllestadquernstones,nowstockfishwasprobablyaddedtothe
listofgoods.Tradefromtheplotsinthetown
area corresponds well with Øystein’s suggested
plans.
The sources also suggest that professional
shoemakers, metalworkers and antler, bone,
hornandwhale/walrusboneartisansworkedin
Bergen.Theymayeitherhavebeenresidentparttimeartisansorambulatingartisanswhovisited
Bergen for shorter periods. Either way, none of
theactivitiescanbeseenassignificantdeterminantsfortheriseofthetown,astheproduction
wassolimitedthatitmusthaveservedaninterurbanmarketonly.
To sum up, Øystein Magnusson may have
planned to strengthen Bergen as an ecclesiastic
and royal administrative centre. The king’s investmentsinBergencanalsobeseenasapositivemeansforhisassumedplanstoinfluenceor
controltheexportofstockfishfromthenorthof
Norway to Europe and his plans to strengthen
Bergenasastaplewheregoods,inthehandsof
magnates,weredirectedthroughBergenandinto
along-distancetradingnetwork.Theking’slarge
investmentsinmonumentsandecclesiasticinstitutionscanalsobeseenashisattempttoshow
offhissocialcapacity.Seentogether,thesources
showthatthetownspeopleofBergenwerenow
involvedinawiderinternationalnetworkwhere
commodities were exchanged. The king’s plans
forthetownareathusseemtohavebeenquite
successful.Øystein’sinvestmentsinBergenand
between Bergen and the fisheries to the north
may perhaps have triggered the growing interestamongthemagnatestouseatownplot.The
planned townscape was thus slowly developing
intoalivingurbancommunity.
Bergenduringhorizon5(1120s-c1170)
WhydidthekingsinvestfurtherinBergen?As
we have seen the sources are vague regarding
the initiator(s) behind most of the monuments
foundedduringhorizon5.Kingsmay,however,
betentativelyconnectedtomanyofthemonumentsandshouldmostlikelybeseenbehindthe
incorporationofthesoutherntownareaandthe
Nonneseterareaintothetownscape.IfStJohn’s
abbey was founded on royal initiative this may
beinterpretedasyetanotherroyalinvestmentin
Bergenasthebishop’sresidence.TherelicsofSt
SunnivaweretransferredtoBergenin1170and
placed in the Christchurch Cathedral (MHN
1880, 151-152). The construction of the cathedralmaythenhavereachedanendandthebishop’sseatwasformallyattachedtoBergen(Helle
1982,92,146;Lidén1993,10).Theking’spossibleinvestmentintheNonneseterconventcan
also have been intended to strengthen Bergen’s
roleasanecclesiasticcentre.
Seven churches were built or rebuilt in the
town area and in the Nonneseter area, five of
these may be tentatively associated with royal
initiativesandareseenastheking’sinvestment
inthetown.Asarguedabove,thevariouskings’
substantial investments in ecclesiastic institutionscanbeseenasbothawaytoshowoffsocialcapacitiesandasapositivemeanstotrigger
andencouragetradeactivitiesinthetown.The
presence of the gjaldker, a royal delegate who
collectedincomeforthekinginBergenin1159
(Hkr 1893-1901, 604; Holtsmark 1961, 692;
Helle1982,8)showsthatthetownspeoplenow
paidduestotheking,andimpliesthattheking
hadaninterestinbloomingcommercialactivities. The inclusion of the southern town area
intothetownscapeisalsointerpretedasaroyal
investment in the town. Analysis of ownership
tolandinmedievalBergenshowsthattheking
stillownedalargepartofthesoutherntownarea
in the thirteenth century (Ersland 1989, 257ff;
Ersland1994,Figure12).Thisimpliesthatthe
kingdidnotgiveawaylandinlargepartsofthis
townarea.Doesthissuggestthatthekingwas
nowplanningtoengagehimselfinactivitiesof
amore‘common’character,activitiesthatcould
notbeconductedfromHolmen?
In the town area, almost all the investigated
plots/sites in the northern and middle town
areas appear to have been settled in horizon 5.
However, there was probably not much pressureonbuildinglandinthesoutherntownarea.
ThoughVeisanwasnotsuitedasaharbourany
more,severalplotsalongVeisannowseemtobe
occupied. The same applies to the area at the
foot of Fløyfjellet in the northern town area.
Thenorthernandmiddletownareaswerethus
gettingmoredenselyoccupiedandevenlessat-
14How,when,bytheinitiativesofwhom,andwhydidBergenemergeasatown?
235
tractivespacewasfilledup.AplotinBergenhad
apparentlybecomeanassetworthusingforthe
townspeople. The large number of plots where
well established and permanent settlements are
documented supports this picture. The likely
presenceofprivatelyfoundedchurchesmaysuggest that the town was now so established that
townspeopleinvestedinactivitiesbeyondthose
conductedfromtheirindividualplots.
In the northern and middle town areas substantialinvestmentsinharbourfacilitiessuggest
thatgoodlandingconditionsforcontemporary
carrierswereconsideredimportantforthetownspeople.TradewasconductedbothinternationallyandalongthecoastofNorwayaswitnessedby
coastalcarriersontheirwaytoTrondheimfrom
Bergenduringhorizon5.Artefactsindicatingdirectorindirectcontactstoeasternandwestern
Norwayandfoundin16and7ofthe24artefactyielding analytic units respectively, also add to
thispicture;asdotheinternationalcontact-indicatingfindsretrievedfrom22ofthe24artefactyieldingunitsandwithaprovenancetoGreenlander/Icelandic,English,German,Frenchports
and ports in the Low Countries. The sources
thusshowthattradewasnowanimportantpart
of the townspeople’s strategies for using a plot.
Fivebuildingscouldbeconnectedtotrade,however,thebuildingsweremultifunctionalandnot
solely constructed with the purpose of trade in
mind.Allinall,though,tradeconductedfrom
thetownareanowseemstobeimportant.This
correspondswellwiththekings’suggestedplans
forthetownarea.
ProfessionalsausagemakerswereprobablyactiveinBergenduringhorizon5andrepresenta
new urban trade aimed at serving townspeople
or visitors of the town. Innkeepers and people
whohadpremisestoletforforexampleambulatingartisansmayalsobelongtoanewgroupof
specialiststhatmadealivinginthetown.These
service-relatedtradesaretheonlydailyactivities
documented - beside long-distance trade - that
can be argued to have played an independent
economicroleforthetown,astheyservedaninterurbanaswellasawidermarket(visitorstothe
town)withtheirservices.
Onplot6/Clime-slakingpitsindicatethatthe
residentsofthisplotdeliveredmortar,perhapsto
236
thenearbychurchofStMary’s.Andasmithmay
havehadapermanentworkshopinthenorthern
townarea.Thematerialalsosuggeststhatambulatingartisansorworkshops-combmakersand
miscellaneousantler,bone,hornandwhale/walrusboneworkers,shoemakersthatalsorepaired
shoes,andfinemetalsmithsnowpaidfrequent,
butshort,visitstoBergen.Thepresenceofthe
ambulatingartisansshouldprimarilybeunderstoodasaconsequenceoftheexistingurbancommunityratherthanviceversa,anditmayshow
thatBergenwasnowlargeenoughforavisit,but
notyetlargeenoughtoprovideamarketforfulltimeresidentspecialistsofthesetrades.
Theservice-relatedtradesandambulatingartisanscannotbeassociatedwiththeking’sinitial
plansforthetown.Theexistenceofsuchtrades,
however,showsthattheplannedtownscapewas
beginningtolivealifeofitsownandithaddeveloped into a living urban community. This
developmentwasprobablyduetothesynergetic
effectcausedbythetown’sroleasabishop’sresidence,anecclesiasticcentre,acentralstaplefor
royal and private goods, a centre for stockfish
trade,andasacentreforroyaladministration.
To sum up, the kings in horizon 5 probably
followedtheirpredecessors’plansforBergenasa
bishop’sseatandstrengthenedBergen’sroleasan
ecclesiasticcentre.Thekingsmayalsobetentativelyassociatedwiththefoundationorrenewal
ofasmanyasfivechurches.Itissuggestedthat
thekings’possibleecclesiasticinvestmentswere
aimedatbothshowingofftherespectivekings’
social capacities and at strengthening the town
asacentralstaple-tobeusedbybothmagnates
and the king(s) - by canalising tithes or other
incomesfromthechurchesthroughBergenand
into a trading system. Long-distance national
and international trade was now an important
part of the townspeople’s strategies for using a
plot in the town area. The actual function of
thetownthuscorrespondswellwiththekings’
assumed plans for the town. In addition to investingintheirprivateplotsgroupsoftownspeople may also have founded churches, and some
townspeople had developed new strategies for
maintainingalivinginthetown.Thetownhad
thusdevelopedfromaplannedtownscapeintoa
diversified,living,urbancommunity.
Conclusions
In horizon 2 (c 1020/30-c1070) a king, probablyOlavHaraldssonorKnutdenMektige,may
havecarriedintolifetheplansforatowninwesternNorway,thetownmayhavebeenplannedto
functionasacentralstapleforgoodsinthehands
of magnates and the king himself. The king’s
motivesforfoundingatownmayhavebeento
collectduesontradeoronprotectionofmarket
peace.Itseemsthattheplotsintheplannednew
townwerejustbarelytakenintouseduringhorizon2,anditappearsthattheking’splanswere
not a success. The king probably did not have
sufficientresourcestocarryintolifehisplans.
When King Olav Kyrre invested further in
the town during horizon 3 (c 1070-c 1100) he
planned to develop Bergen into an ecclesiastic
and perhaps also a royal administrative centre.
He may also have had plans for the town as a
central staple for goods disposed by magnates
andhimself.Activityinthetownareawasscarce
andthesourcescannotelucidatetheactualfunction of the town area. Altogether the limited
amount of activity in the town area, however,
suggeststhattheking’splansforthetownarea
were not successful. Again the king apparently
didnothavetheresourcestocarryouthisplans,
atleastforthetownarea.
Duringhorizon4(c1100-1120s)KingØysteinMagnussonplannedtostrengthenBergenas
aroyalandecclesiasticadministrativecentre.He
mayalsohaveinvestedinthetownaspartofa
plantoinfluenceandcontrolthetradeinstockfishfromnorthernNorwaytoEuropeandmay
havewishedtodirectthesurplusfromroyaland
othermanorialduesthroughthetownandinto
atradingsystem.Inthetownarea,tradewasthe
onlyrecordeddailyactivitythatmayhaveserved
asaneconomicbasisforthetownspeople.This
correspondswellwiththeking’sassumedplans
forthetownarea.
The various kings strengthened Bergen as a
royal and ecclesiastic centre during horizon 5
(1120s-c1170).Thekings’possibleinvestments
in the many ecclesiastic institutions may have
beenaimedatshowingofftherespectivekings’
socialcapacitiesandatstrengtheningthetownas
astaple.Inthetownarea,tradewasnowamore
visiblepartofthetownspeople’sstrategyforus15Conclusions
ingaplotinthetown.Newurbanservice-related
trades cover a wide spectre, and various ambulatingartisansfoundtheirwaytoBergen.These
factorsshowthatBergenhaddevelopedfroma
plannedtownscapeintoadiversified,living,urbancommunity.
15CONCLUSIONS
Thequestionsofhow,when,bytheinitiativeof
whom,andwhyatownemergedintheBergen
areahavenowbeenaddressedandacasestudyof
theprocessofurbandevelopmentinearlymedievalScandinaviahasbeengiven.Theperiodbetweentheninthcenturyandc1170wasstudied,
withamainemphasisontheperiodbetweenc
1020/30andc1170.
Through six part-studies of major initiatives
and daily activities the overall questions have
been elucidated. And initiatives and activities
were eventually linked to actors from different
levelsofthesocialhierarchyanddiscussedwith
thewiderhistoricalcontextasabackdrop.The
understanding of social change as a product of
theinterplaybetweenpeoplefromdifferentlevelsofthesocialhierarchyandtheirwiderhistoricalcontexthasstructuredthediscussions.
Byusingthearchaeologicalandbotanicalmaterial from various methodological approaches,
the qualitatively diversified material could be
activated whether excavated in the nineteenth
centuryormorerecentlyandthemainpartsof
theextensivebodyofsourcescouldbetakeninto
use.Thesourceshavebeendividedintocategoriesaccordingtotheirreliabilityandplotshave
served as the main analytic unit. The sources
were studied spatially using the production of
mapsandaqualitativeandcontextualapproach.
Furthermore, the material has been studied
throughadiachronicapproachwherethenarrow
timescopesoffivehorizonsgavetheopportunity
ofstudyingthesourceswithinverynarrowhistoricalcontextsandinsomecaseslinkingmajor
initiativestohistoricallyknownactors.
In the first part-study, activity and general
land use in the Bergen area between the ninth
centuryandc1020/30(horizon1)wasstudied.
237
AtthistimetheBergenareawasmostlikelyoccupiedbyasettlementwhereagrarianactivities
were carried out. This settlement is tentatively
locatedatHolmen,anditmayhavehaditsfields
intheBergenarea.Alanding-placeforboats,locatedinthemiddletownarea,andperhapsassociatedwiththeroyalestateAlrekstad,wasalso
foundintheBergenareabeforethefirstdecades
oftheeleventhcentury.
Inthesecondpart-studyIshowedthatplots
were laid out in the northern town area duringhorizon2(c1020/30-c1070).Inhorizon3
(c1070-c1100)theseplotswereredesignedand
the middle town area was also included in the
townscape.Thedatesforthesemajorinitiatives
arebasedonthebest-sustainedinterpretationof
the available sources at the present state of research.
Thethirdstudyelucidatedhowtheplotsand
otherpartsoftheBergenareaintimewereoccupiedandwereusedphysically.Thestudyshowed
thattheareasalongtheVågenshorelinewereconsideredmostattractive,andthattheactualsettlementofthetownwasalongandslowprocess.
Inthefourthstudythenatureandorganisation of crafts and production were studied in
order to elucidate whether productive activities
identifiedinearlyBergencouldhaveprovideda
fundamental economic basis for the rise of the
town. Fishing, hunting, miscellaneous antler,
bone, horn and whale/walrus bone working,
some ‘other leatherworking’, small-scale wood
andstoneworking,basiccookingandsomefood
andbeverageprocessingwereprobablyallactivitiescarriedoutonahouseholdbasis.Andnone
ofthesecouldinthemselveshavebeendecisive
fortheriseofthetown.
Ambulating professional shoemakers (who
also repaired shoes), combmakers and metalworkerswhoprobablycametoBergenforshort
visits only were most likely artisans supplying
largeareaswithstandardisednon-luxuryitems.
Theymayprimarilyhavesuppliedtheinterurban
market,whileworkinginBergen.Theirpresence
inBergenreflectedinthematerialfromhorizon
5wasseenassecondary,andasaconsequenceof
anestablishedcommunity.Theirpresencecould
nothaveservedasafundamentaleconomicbasis
fortheriseofthetown.
238
Itcouldnotbeestablishedwhethertheactivitiesofantler,bone,hornandwhale/walrusbone
working,andshoemakingduringhorizon4and
antler,bone,hornandwhale/walrusboneworkingduringhorizon5werecarriedoutbyresident
part-time professionals or ambulating artisans.
Regardless of the organisation of these activitiesduringhorizon4,andforantler,bone,horn
andwhale/walrusboneworkingalsoinhorizon
5, the small amounts of waste left behind suggestthattheartisansproducedforaninterurban
marketonlyandtheactivitiescannothaveserved
asafundamentaleconomicbasisfortheriseof
thetown.
The only ‘full-time professional’ productive activity documented in early Bergen was
large-scale stoneworking, carried out by craftsmen engaged in the construction of the many
monumental buildings erected in horizons 3-5.
Presumably, the artisans were integrated in the
household of the monument founders during
theperiodofconstruction.Inspiteofbeingprofessional, their production took place within a
‘household’,andtheirpresencecouldnotinitself
beseenasafundamentaleconomicbasisforthe
riseofthetown.Alongthesamelineofthinking, the presence of professional sedentary or
ambulatinglarge-scalewoodworkersshouldnot
beseenasafundamentaleconomicbasisforthe
riseofthetown.
Innkeeperswithbeveragesforsaleandlodgingforlease,probablesausagemakersandpeople
who let out premises for temporary workshops
wereseenasagroupofurbanprofessionalswho
were active in Bergen from horizon 5. The activitiesofthesenewurbanservice-relatedtrades,
carriedoutbypart-timeorfull-timeprofessionals, may in time have added value to the town
community as the activities, in part, were paid
forbyvisitingtravellers.However,thenewtrades
were also a consequence of the existence of a
communityratherthanatriggerfortheriseof
thetown.
InsumnoneoftheproductiveactivitiesdocumentedinearlyBergenwereseenasfundamental
fortheinitiationofthetown.Rathertheywere
presentasaresultoftheexistingcommunity,althoughtheirpresencemustalsohavehadasynergeticeffectforthegrowthoftheplace.
Inthefifthpart-studytheimportanceoflongdistance trade in bulk goods as a fundamental
factor for the rise of the town was studied. I
showedthatharbourconditionswereapriority
forthetownspeople.Thiswastakenasanindicationthatseatransportwasconsideredimportant
totheactorsfromthebeginning.Concretemeasurestomeetthedemandsofwaterdepthfrom
contemporarycargocarriersweredocumentedin
horizon5.Trade-indicatingsourcessuchastools
oftrade,findswitha‘long-distanceprovenance’
andstoragebuildingsshowedthatlong-distance
tradewithbulkcommoditiesfromhorizon4and
onwardswasanimportantactivityinthetown.
In the sixth part-study the character of the
settlementsonthetownplotswasaddressedin
order to elucidate whether well-established settlementsofapermanentcharacterwerefoundin
thetownarea.Fromhorizon4andonwardssuch
settlementswerewell-documentedandcertainly
present.
Theconclusionsfromthesestudieswereused
in the synthesising chapter where the overall
questions of the study have been addressed. In
this chapter, actors from the different levels of
the social hierarchy were connected to the major initiatives and the daily activities. The materialfromBergenwasinterpretedwiththeactors’ wider historical context as a backdrop. I
argued that Bergen was most likely founded
through royal initiative on land where agricultural activities were carried out. This probably
occurredaround1020/30,inthereignofKing
Olav Haraldsson (c 1015-1028) or King Knut
den Mektige (1028-1034/35), making either of
thesekingsortheirrepresentativeslikelyfounder
candidates.Thekingmayhavewishedtofound
BergenasacentralstapleinwesternNorwayfor
goodsdisposedbymagnatesandthekinghimself.Theplanmayhavebeenthatgoodswereto
enter a national and international trading network from the new planned town. The king’s
motives for the foundation may perhaps have
been to collect dues on trade or on the protectionofmarketpeace.Theideaofatownwasnot
wellreceivedbythemagnateswhowereprobably
givenplotsintheplannedtown.Althoughsome
plotswereoccupied,itappearsthattheplanned
town in the Bergen area was just barely settled
15Conclusions
duringhorizon2,andthattheking’splanswere
notasuccess.
WhenKingOlavKyrreinvestedfurtherinthe
townduringhorizon3(c1070-c1100)thenorthern town area seems to have been redesigned,
Holmenandprobablyalsothemiddletownarea
were added to the townscape, and monuments
wereinitiatedatHolmen.Plotsinthetownarea
were most likely given to magnates. The king
planned to develop Bergen into an ecclesiastic
and perhaps also a royal administrative centre.
He may also have had plans for the town as a
centralstapleforgoodsinthehandsofmagnates
and himself. Again, activity in the town area
seemstohavebeenscarce,andthesparsesources
couldnotshedlightonthetown’sactualfunction. All in all, though, the apparently limited
activityinthetownareasuggeststhattheking’s
plansforthispartofthetownwerenotsuccessful.Thekingperhapsdidnotfollowuphisplans
for the town area with the necessary means of
enforcement,orhedidnotpossesssufficientresourcestoenforcethenewrulesanddevelopnew
traditionsforusingtheplace.
Duringhorizon4(c1100-1120s),KingØysteinMagnussonbuiltaroyalhallatHolmenand
founded several ecclesiastic institutions includingoneatNordnes,therebyaddingthisareato
the townscape. The townspeople became more
activeontheplotsandsettlementsofapermanentcharacterwerenowclearlypresent.Øystein
plannedtostrengthenBergenasaroyalandecclesiasticadministrativecentre.Hemayalsohave
investedinthetownaspartofaplantoinfluence
andcontrolthetradeinstockfishfromnorthern
Norway to Europe and he may have wished to
directthesurplusfromroyalandothermanorial
dues through the town and into a trading system.Tradewastheonlyrecordeddailyactivity
that may have served as an economic basis for
the town, this corresponds well with the king’s
assumedplansforthetownarea.
During horizon 5 (1120s-c1170) the various
kings-itisnotpossibletonametheindividual
kings in this horizon - probably included the
southerntownareaandtheNonneseterareain
the townscape and built several churches. The
kings’investmentsinthemanyecclesiasticinstitutionsmayhavebeenaimedatshowingoffthe
239
respectivekings’socialcapacityandatstrengtheningthetownasastaple.Thetownspeoplenow
investedmoreintensivelyintheirtownplotsand
perhaps also built churches. Trade had become
amorevisiblepartofthetownspeople’sstrategy
forusingplotsinthetown.Thenewurbanservice-relatedtradescoverawidespectreofactivitiesandawidespectreofmostlikelyambulating
artisansfoundtheirwaytoBergen.Thesefactors
showthatBergenhadbeguntolivealifeofits
ownandhaddevelopedintoadiversified,living,
urbancommunity.
Mystudyhashaditsbasisinthecontemporaryarchaeological,botanicalandwrittensources.Previousstudieshaveonlyshedaratherdim
lightontheearliesthistoryofBergen,mainlybecausethearchaeologicaldatauntilnowhasnot
beenconsideredolderthanthefirstpartofthe
twelfthcenturyandtherearefewrelevantwrittenrecords.Withfreshdatingmaterial,acritical review of the archaeological and botanical
sourcesandnewmethodologicalandtheoretical
approachesIhavesuggestednewanswerstothe
classical questions of how, when, by the initiativesofwhomandwhyatownemergedinthe
Bergenarea.Inmanyrespectstheanswersthat
were suggested in the present study relate to a
localorhistoricalsituationspecificforNorway.
Stilltheyshouldprovevaluablewhendiscussing
theemergenceoftownsonamoregenerallevel
awell.
Bythemiddle1990stherewasgeneralconsensusthatBergenwasmostlikelyfoundedbyOlav
Kyrre, perhaps juridically (Helle 1982, 1992),
perhapsphysicallybyparcellingoutthenorthern
andmiddletownareasintoplots(Ersland1994),
butprobablyonasitepreviouslyoccupiedbya
densernon-ruralsettlement.Ihavearguedthat
Bergenwasindeedfoundedbyaking,butthis
mostlikelyhappenedhalfacenturybeforeOlav
Kyrre, and probably on a virgin piece of land.
Later,OlavKyrreinvestedfurtherinthetown,
byredesigningthealreadyexistingplotsystem,
byincludingmorelandinthetownscapeandby
initiating ecclesiastic building projects at Holmen.Intheyearstocomestillmoreareaswere
includedinthetownscape;thetownthusgrew
insteps.
Ersland has argued that Bergen, conforming
240
to the ‘typical process of town foundation’ in
northern Europe, may have consisted of many
‘plan-units’addedtothetownscapeatdifferent
timesinhistory(cfp25ff)(Ersland1994).The
picturethathasemergedthroughmystudycoincides well with his plan-unit hypothesis as a
principle,andinthisrespecttheprocessesdocumented in Bergen may resemble the ‘typical
process of town foundation’ in other northern
Europeantowns.
Thesuccessiveadditionofstillmoreareasto
thetownscapealsoprovidesapracticalexplanationtothe‘doublenucleussituation’intwelfth
century Bergen, suggested by several researchers (Dunlop 1985; Myrvoll 1987; Lidén 1993).
However, the wide extent of Myrvoll’s and
Lidén’ssouthernmostnucleuscannotbesubstantiatedthroughthesources.
My study has first and foremost provided a
morenuancedandvariedpictureoftheprocesses
involvedfromthefoundationofatowntowards
alivingurbancommunity.Ihaveshownthatthe
story of how, when, on the initiative of whom
andwhyBergencameaboutcontainschainsof
major initiatives and daily activities. The processofurbandevelopmentwasslowandinvolved
royal investments as well as investments from
thehandsofthetownspeople.Inthisinterplay
betweenactorsfromdifferentlevelsofthesocial
hierarchyandtheirwiderhistoricalcontext,Bergenintimedevelopedfromaplannedtownscape
-amaterialisedidea-intoalivingurbancommunitycharacterisedbyadiversityoffunctions.
APPENDIXES
APPENDIX1
Sandbrugaten5.Reimers’reconstructionofthe
-1.5 to +5.5 masl contour lines for these sites,
based on observations from archaeological excavations(Reimers1974)(B).
Sourcesforthenaturaltopography
3)
Dreggsalmenningen
14-16: archaeological inabouttheyear1000anddiscussionof
vestigation,morainicdepositswereencountered
thecourseofthecontourlinesinthe
between2.0and4.9masl.Contourlinesforthe
reconstruction
sitehavebeenreconstructedinthereports(Golembnikinprep-a;Golembnikinprep-b)(B).
Thenumbersinthelistbelowrefertopointsor 4) Dreggsalmenningen 10-12: archaeological inareasonthemapinFigure62a-f.Onthemap
vestigation,morainicdepositswereencountered
basicsources(B)arenumberedinboldnumbers
between 2.1 and 3.3 masl (Dunlop 1986b)
andcontoursbasedonthesearedrawninanun(B).
brokenline:
,supplementarysources(S) 5) TheareasouthofStMary’s:archaeologicalinare in italics and contours are drawn in a dotvestigation,morainicdepositsbetween4.0and
5.5masl(Reimers1965)(B).
ted line: …… ( cf Chapter 6). The sources on
thelistarenotorderedinanyconsistentway,the 6) Dreggsalmenningen: archaeological investigation,bedrockwasreachedbetween-2.5andreadershouldthusfirstlookatthemapthenfind
3.0masl(LongandMarstrander1980)(B).
thereferencesofinterest.Thereconstructionis
7)
ØvreDreggsalmenning:archaeologicalinvestigenerallynotdiscussedindetail;wherethemap
gation,morainicdepositsbetween5.3and4.8
isbasedonbasicsourcesthecontoursspeakfor
masl(Dunlop1989e)(B).
themselves,forsomeareas,however,adiscussion 8) Kroken:archaeologicalinvestigation,morainic
ofthecourseofthecontoursisgiven.
deposits were recorded at 5.0 masl (Dunlop
1987)(B).Culture-layerswererecordeddown
to a level of 2.9 masl through test drilling
1) The reconstruction of this area is based on
(KrzywinskiandHjelle1985)(S).
Fritzvold’smap.Thesourcesareboreholesand
Fritzvold’s survey of the area (Fritzvold 1976, 9) Nye Sandviksveien: archaeological investigation, bedrock and moraine deposits recorded
14ff)(S).Thefoundationplanforthebuilding
between 6.1 and 12.0 masl (Sognnes 1974)
atBradbenken1(Trumpy1954)showsthata
(B).
north-south oriented bedrock ridge ran across
the mouth of Veisan. The bedrock threshold 10) Klingesmauet: Dunlop has reconstructed the
contourlinesfor11-14maslonthebasisofarbetween Vågen and Veisan was at -0.3 masl
chaeological excavations in the area (Dunlop
(S).89KariLoeHjelle’sinvestigationfromKo1989f) (B). In the present reconstruction the
engenshowsthatVeisanwasamarinebasinat
level of the oldest recorded culture-layers at
least into the eleventh century (Hjelle 1986,
KlingesmauetBRM299areusedasasupple36,67,73;Hansen1994b,177).Thereforethe
menttoDunlop’sreconstruction(S).
threshold between Veisan and the Vågen Bay
musthavebeenbelowthesealevelabout1000. 11) Øvregaten 43: archaeological investigation,
morainewasencounteredat6.25masl(Chris(Seealsopoint115).
tensson1980b)(B).
2) Bryggen; Dreggsalmenningen 10-12, and
241
12) Øvregaten43:archaeologicalexcavation,bedrock was encountered at 7.0 masl (Dunlop
1989b)(B).
13) Øvregaten 41: archaeological excavation, bedrockwasrecordedbetween11.9and12.7masl
(B)(Larsen1975).
14) Øvregaten 39: archaeological excavation, morainic masses and bedrock were encountered
between6.5and8.4masl(Dunlop1982)(B).
15) Øvregaten 37/39: probe boreholes, moraine
wasrecordedbetween6.0and6.5masl(Larsen
andReimers1978)(S).
16) Wesenbergsmauet: Dunlop has reconstructed
thecontourlinesfor11-14maslonthebasisof
archaeological data (Dunlop 1989d) (B). The
level of the oldest recorded culture-layers in
profiles1-6andprofile13areusedasasupplementarysourceinadditiontoDunlop’sreconstruction(S).
17) Koren-WibergsPlass:archaeologicalinvestigation, bedrock was encountered between 14.1
and13.0masl(Reimers1971a)(B).
18) Koren-WibergsPlass:archaeologicalinvestigation, moraine was recorded at 8.7 masl (Myrvoll1980)(B).
19) Nikolaismauet: archaeological investigation,
bedrockwasencounteredat15.0masl(Dunlop
1984h)(B).
20) Øvregaten25:boreholes,bedrockwasrecorded
between7.3and8.2masl(Larsen1978)(S).
21) Øvregaten23:boreholes,bedrockencountered
at10.0and15.0masl(Reimers1977)(S).
22) The Church of St Peter: Reimers has documented part of the northern wall around the
churchyardofStPeter’s.Theleveloftheoldest
culture-layersareusedhereasasupplementary
source for the natural topography (Reimers
1979)(S).
23) TheChurchofStPeter:Reimersdocumented
theSWcornerofthechurchruin.Thetopof
theruinwasrecordedat5.68maslin(Bertelsen
andLarsen1971).AccordingtoKoren-Wiberg
thispartofthewallwaspreserveduptoabout
one m above the ground level of the church
(Koren-Wiberg1921).Onthisbasisthelevelof
4.7maslisusedasasupplementarysourcefor
thetopography(S).
24) Bugården N 4: archaeological investigation,
culture-layerswererecordeddowntoalevelof
3.6masl(BertelsenandLarsen1971)(S).
25) Koren-Wibergrecordedabuildingtothesouth
ofStPeter’sduringhisexcavationhere(KorenWiberg1921).AccordingtoKoren-Wibergthe
lowestfloorlevelrecordedinthebuildingwas
242
26)
27)
28)
29)
30)
31)
32)
33)
34)
35)
3.67 m lower than Øvregaten. Generalkart
1879-80showsthatØvregatenatthistimehad
anelevationofabout8.5masl,onthisbasisthe
lowestrecordedfloorlevelinthebuildingsouth
ofStPeter’smusthavebeenatalevelofabout
5.3masl.Thismeasurementisusedasasupplementarysourceforthenaturaltopography(S).
Bugården S and Bredsgården N: archaeological investigation, culture-layers were recorded
down to a level of 3.4 masl (Reimers 1973b)
(S).
Enhjørningegården: archaeological investigation, culture-layers were recorded down to a
levelof4.1masl(Dunlop1984f)(S).Remains
werealsofoundofabuildingK1whichisidenticalwithK19attheexcavationatBryggeparkenBRM287(Dunlop1989a).Theorientation
of this building probably reflects the curve of
anearlyseafront(S).
Stallen, Svendsgården: archaeological investigation,moraineandbedrockwererecordedbetween1.5and3.2masl(Christensson,Dunlop,
andGöthberg1982)(B).
Bryggeparken: Dunlop has reconstructed the
2-9 masl contours through information from
archaeological investigations (Dunlop 1989c)
(B). From profiles 1-6 at the Bryggenparken
BRM287site(Dunlop1989a)thelevelofthe
lowestrecordedculture-layersisusedasasupplementary source for the natural topography
(S).
Øvregaten by Nikolaikirkealmenningen: archaeologicalinvestigation,morainewasrecordedat9.0masl(Christensson1980a)(B).
Bellgården Steinkjeller: archaeological investigation,bedrockrecordedbetween4.9and6.5
(Reimers1973a)(B).
Rosenkrantzgate 7: bedrock was recorded in
connectionwithconstructionwork(Lindholm
1916)(S).
Rosenkrantzgate 4: archaeological investigation,bythehelpofearthaugerthelevelofthe
morainicdepositswasmeasuredatbetween-1.9
and-1.4masl(Lindh1979)(S).Theconstructions in the later phases at the Rosenkrantzgate4sitereflectthecurveofanearlyseafront
(Lindh1979)(S).(Seediscussionbelow).
Lodin Lepps Gate: archaeological investigation, culture-layers were recorded down to a
levelbetween3.3and6.6masl(Dunlop1990)
(S).
Nikolaikirkealmenningen: archaeological investigation,theleveloftheoldestculture-layers
inprofiles1-4(Dunlop1983a)servesasasup-
36)
37)
38)
39)
plementary source for the natural topography
(S).
Forstandersmauet 4: archaeological investigation,bedrockwasrecordedat27.3masl(Dunlop1991a)(B).
Below Forstandersmauet 4: Bendixen documentedtheChurchofStNicholas,builtona
terraceonthemountainslope.Thesouthwall
of the nave was founded on boulders, which
rested on the bedrock, and the north wall of
thenavewasfoundeddirectlyonorinthemorainicmasses.Thewallaroundthechurchyard
was uncovered on a terrace below the church
(Bendixen 1896). This information is used as
asupplementarysourceforthenaturaltopography (S). The area between the present day
ØvregatenandForstandersmauethasagradient
of0.8:1onanevenslope.Theslopeis,however,
notevenandtodaytheareaischaracterisedby
artificialterraces,soitisdifficulttorecognise
otherthanthegeneraloutlineofthepre-urban
landscape.OnthebasisofBendixen’sdescriptionofthenaturalsubsoilontheplotofStNicholas’,Ihavereconstructedtwoterracesonthe
mountainside.Oneupperterracelargeenough
for the Church of St Nicholas to be built as
a church with nave and a southern side aisle,
andalowerterracewherethechurchyardwas
placed.Thisreconstructionoftheareaaround
StNicholas’mustbetakenwithsomereservations,aswedonothavemanybasicsourcesin
thearea.
Øvregaten 11: Koren-Wiberg excavated the
building closest to Øvregaten, he found two
levels of building remains and two fire-layers
beforehereachedthemoraine.Thelatestfirelayer(1)wasrecordedalmostdirectlyunderthe
modern(c1900)level,fire-layer2wasrecorded
1.5-2.0 m under fire-layer 1 (Koren-Wiberg
1908b). If we assume that each fire-layer was
about 10- 20 cm thick the moraine masses
musthavebeenfoundabout2mbelowthec
1900buildinglevel.AccordingtoGeneralkart
1879-80 Øvregaten 11 was built at about 13
masl.Themorainemassesmustthenhavebeen
measuredatabout11masl(S).
Øvregaten 9: archaeological investigation,
moraine was recorded between 11.0 and 14.6
(Solberg1969;Reimers1972a)(B).Thisdata
issupplementedbymeasurementofbedrockup
to15maslthroughboreholes(Strømmen1969)
(S).InformationfromKoren-Wiberg’sinvestigationinthevicinity,showsthatbedrockwas
coveredbymoraineinthisarea(Koren-Wiberg
40)
41)
42)
43)
1908b)(S)sowemustaddsometothe15m
bedrockcontourline.
Finnegårdsgaten/Øvregaten:archaeologicalinvestigation,morainewasrecordedbetween9.0
and13.4masl(Christensson1980c)(B).
Øvregate 4: in the area, which was Dramshusen’s backyard until the c 1900 regulation
Koren-Wiberg found ‘timber remains’ about
1.5mbelowthesurface(Koren-Wiberg1900).
Theplaceisfoundbetweencontourlines9and
10 masl on Generalkart 1879-80. The timber
remains must therefore have been found at a
levelabout7.5-8.5maslandsterilemoraineor
bedrockmusthavebeenunderthislevel.This
informationisusedasasupplementarysource
forthenaturaltopography(S).
Finnegården: archaeological investigations,
moraine and bedrock was recorded between
0.5 and 0.8 masl at Finnegården 6a (Dunlop
1982(1998))(B)andbetween-0.3and-1.7at
Finnegården 3a (Golembnik 1993) (B). The
surfaceofthenaturalsubsoilhasaWNW-ESE
(geographical) orientation (Dunlop and Golembnikinprep).(Seediscussionbelow).
Bedrock was recorded by Koren-Wiberg in
1900 during the demolition of the seawards
buildingoftheDramshusentenement.Itisnot
straightforwardtopindowntheexactposition
orheightoftheobservation(seeHansen1994b
for an elaborate discussion of this). However,
it is quite certain that Koren-Wiberg did observebedrockabovethewaterlevel(about+0.5
masl)intheareaaroundpresentdayBryggen
3-4(Hansen1994b,p182).(Seediscussionbelow).
Discussionoftheareaaroundpoints33,42and43:
Koren-Wibergincorporatedthebedrockaspart
ofthepre-urbanshoreline(Koren-Wiberg1921),
sodoesFritzvoldbuthereliesonKoren-Wiberg
onthispoint(Fritzvold1976,12).Koren-Wibergdidnothaveinformationaboutthenatural
topographyintheareatothenorthofhisbedrockobservation,accordingtohisownoutlineof
thesitesheinvestigatedinthisarea(Koren-Wiberg 1908a). His reconstruction of the shoreline
shouldconsequentlybetakenwithreservations.
DunlopandGolembnikhavereconstructedthe
shorelineinthisareawithoutincorporatingKoren-Wiberg’s bedrock observation (Dunlop and
Golembnikinprep).Ithinkitismostrealistic
toreconstructthe-/+0contourlineonthebasis
ofdatafromtheRosenkrantz4BRM76site(see
243
point33).Thatis,thegradientoftheseabedand
theorientationofthebuildingsinlaterphases,
coupled with data on the natural subsoil from
thesitesatFinnegården(seepoint42).Ifthisis
doneKoren-Wiberg’sbedrockappearsasarock
inthesea.
44) Vetrlidsalmenningen:archaeologicalinvestigation, Dunlop has reconstructed the shoreline,
contour lines 5-14 masl and the course of a
stream which ran down the sloping terrain.
(Dunlop in prep) (B). Between 1 and 4 masl
thenaturalsubsoilwasnotreached.Themeasurementsofthelowestculture-layersinprofiles
25,28and29areusedassupplementarysources(S).(Seediscussionbelow).
45) ThetoweroftheStCross-churchwasfounded
onbedrock,whilethesideaisles‘hadnonatural
foundation’ (Lorentzen 1952, 27). According
to Generalkart 1879-80 the church is placed
betweenthe3and4maslcontourlineswhich,
when leaving some depth for the foundations
placesthenaturalsubsoilatabout2.5-3.0masl
aroundthechurch(S).(Seediscussionbelow).
46) ThechurchyardoftheStCross-church:Archaeologicalinvestigation,morainewasrecordedat
1.0masl.Medievalfill-massesinthisareahad
a north-southwards orientation (geographical)
(Dunlop1984j)(B).(Seediscussionbelow).
47) Hollendergaten 2: excavation by hand and
boreholes:Fritzvolddocumentedthe+/-0masl
contourline(B).(Seediscussionbelow).
48) Hollendergaten9:archaeologicalinvestigation,
bedrock was recorded at 1.14 masl (Reimers
1973c)(B).(Seediscussionbelow).
49) Outside Hollendergaten 8-10: groundwork,
remains of a boat was found about 3.5 m belowthemodernsurface(Lorentzen1952,27).
Thisinformationimpliesthatthenaturalsubsoilwasbeloworaboutthislevel.Accordingto
GrunnkartBergenthepresentdayHollendergatenliesatalevelofbetween2.0and3.0masl,
thenaturalsubsoilthen,mostlikely,wasfound
atoraboutalevelof-1.5to-0.5masl(S).(See
discussionbelow).
Discussion of the area around points 44 to 49:
The sources behind the reconstruction of the
natural topography between Vetrlidsalmenning
and the Church of St Cross are few. The slope
ofthefill-massesatStCrosschurchyard(point
46)indicatesthattherewasasmallbaybetween
Finnegården and the St Cross (Dunlop 1984j,
244
47;DunlopandGolembnikinprep).Thecourse
ofthe+/-0to2maslcontoursis,however,quite
uncertain. Fritzvold reconstructs a rather large
promontory by the St Cross, incorporating the
bedrockatHollendergaten9(point48)aspart
ofthemainland.Ithinkitismorerealistictointerpretthebedrockasarockintheseasincethe
observationsoutsideHollendergaten8-10(point
49), however vague they may be, indicate that
the+/-0contourliesbetweenthispointandthe
St Cross. This assumption is supported by the
structureofthebuildingtopography,asweknow
itfromoldermapssuchasPJWilster’smapfrom
theendoftheseventeenthcentury(Harris1991,
29).OnthismapthepredecessorofHollendergatenwascurved,asitistoday,anditmay,convincingly,reflectthepre-urbantopography.
50) The area around Sparebankgaten, Bankgaten,
Nedre Korskirkealmenningen and Skostredet:
boreholes, Fritzvold has reconstructed the
naturaltopographybetween-4and+/-0masl
(Fritzvold1976)(S).(Seediscussionbelow).
51) Skostredet 10: archaeological investigation,
morainewasrecordedat-1.5masl.Thedeposits
slope from north-east towards the south-west
(geographical). There were fluvial deposits on
the site. The archaeological data was supplementedbyinformationfromboreholes.Golembnikhasreconstructedthe+/-0to-2contour
lines for the area (Golembnik in prep-c) (B)
(S).Thefluvialdepositsatthesiteimplythata
streamhaditsoutletinthevicinity(S).
52) Skostredet17:oralinformationfromconstructionwork,“seasand”about+/-0andbedrock
about+/-0(Fritzvold1976,9)(S)
53) From Domkirkegaten 6 to Kong Oscarsgate:
Komber et al have reconstructed the +/-0-7
maslcontoursforthisareaonthebasisofdata
fromtheDomkirkegaten6BRM245andthe
Lille Øvregaten/Domkirkeplassen BRM 246
sites (B), supplemented with data from boreholes in the area between Kong Oscarsgate,
Øvre Korskirkealmenningen, Lille Øvregaten
and Domkirkeplassen (Komber, Dunlop, Sigurdsson,andHjelle1994)(S).Thereconstruction, of the +/- 0, 1 and 2 contours does not
present the natural topography before human
activitytookplaceinthisarea,butratherthe
topography after some filling in of the Vågen
Bayhadtakenplaceasaresultoferosioncaused
probablybyacombinationofhumanandnatu-
54)
55)
56)
57)
58)
59)
60)
ral processes (before phase 10, which began
about1130/40)(Komber,Dunlop,Sigurdsson,
andHjelle1994,75-81).The+/-0-2contours
are considered representative for c 1000-1100
(B).Fritzvold’s1979reconstructionofthe+/-0
contour probably reflects the natural topographylongbeforec1000-1100andcorresponds
withthelevelofthe‘sterile’blueclaylayer508
attheDomkirkegaten6site(Komber,Dunlop,
Sigurdsson, and Hjelle 1994, 77). Fritzvold’s
reconstruction of +/- 0 is therefore not used
here.
KongOscarsgate15-17:constructionwork,‘sea
sand’wasfoundabout3munderthemodern
surfaceinthegatewayatTanksSkole(Lorentzen
1952,172).Thisgatewayliesbetweenthe5and
6maslcontourlines(GrunnkartBergen1992),
thenaturalsubsoilshouldthusbeatabout2.0
or3.0maslhere(S).
Kong Oscarsgate 36: archaeological investigation,themorainewasencounteredbetween7.2
and7.5maslandat10.7maslinHeggebakken
(Reimers1971b)(B).
NedreHamburgersmauet5:archaeologicalinvestigation, moraine masses were recorded at
5.9masl(Dunlop1981b)(B).
The15-20mcontourlinesarevisibleasbedrock
several places. These contour lines have been
drawn according to Grunnkart Bergen 1992
(B).FromtheChurchofStOlavsinVågsbunnenandnorthtowardsØvreKorskirkealmenningenthereconstructionofthecontourlines
isaguideonly,onthebasisofthebuildingtopographyasweseeitonGeneralkart1879-80
(S).
Lille Øvregaten: archaeological investigation,
contourlines7and8maslarereconstructedin
thereport(Hansen1995b)(B).
Informationonthenaturalbedrocksurfaceof
Nordnes is taken from Generalkart 1879-80.
The measurements have been checked west
of Krudthuset by the aquarium (Dunlop and
Hansen1994c)andatNordnesgaten47(Dunlop1991b).Theoldmeasurementswerequite
accurateatthesepoints.(B).Wherebedrockis
not visible, I have modified the contour lines
from Generalkart 1879-80 according to my
survey of the landscape. While doing this I
have taken into consideration the elaborate
buildingactivitytheareahasbeenexposedto
duringhistory(S)
Strandgaten3,7,19,21-23,17/Strandkaien14
and Strandkaien 2, 4, 8/11, 16, 18/20: boreholes and observations in connection with
61)
62)
63)
64)
65)
66)
67)
68)
69)
70)
71)
72)
groundwork (S), Fritzvold has reconstructed
the bedrock contour lines for this area (Fritzvold1976).Howeverthecontoursfortheseabedaredrawnwiththedatafrompoint117in
mindonthepresentreconstruction,atpoint17
gyttjaat-3.10maslshowsthatloosesediments
hadfilleduptheheadofthebay.
Strandgaten 55-57: archaeological investigation, information on the natural topography
between +/- 0 and 2.6 masl (Dunlop 1986a)
(B).
Klostergaten 16: archaeological investigation,
observation of moraine and bedrock between
15.0 and 15.8 masl (Dunlop and Hansen
1994b)(B).
St Hansstredet: archaeological investigation,
bedrock observed between 7.1 and 8.3 masl
(Hansen1994c)(B).
Bekketomten: archaeological investigation,
bedrock was observed between 6.4 and 7.17
masl. The natural topography sloped from
southeast towards northwest (Dunlop 1988b)
(B).
Nagelgården 6: archaeological investigation,
bedrock was observed at 0.7 masl (Dunlop
1988b)(B).
Tollbualmenningen by the western corner of
Tollboden: archaeological investigation, the
lowest culture-layers observed were at a level
of 1.0-1.1 masl (Dunlop 1988c). This can be
usedasasupplementarysourceforthenatural
topography(S).
Strandgaten 80: archaeological investigation,
sterile sandy clay was observed at 5.8 masl
(Dunlop1988a)(B).
Nordnes 33: boreholes, the contour lines for
bedrockfrom+/-0to10maslarereconstructed
(Dunlop1983b)(S).
Østre Muralmenning: boreholes, moraine
wasobservedbetween6.4masland12.5masl
(Christensson1981)(S).
Nøstegaten 65a-91: archaeological investigation, bedrock was observed at about 1.2 masl
(Sletten1984)(S).
Knøsesmauet: archaeological investigation,
bedrockwasencounteredbetween0.4and1.0
m under the street level (Dunlop 1991d). AccordingtoGeneralkart1879-80thestreetruns
at an elevation between 2 masl and 22 masl.
I deduct about 0.5 m from the measurements
attheGeneralkartanddrawcontours2to22
masl for the natural topography on this basis
(S).
Det Gamle Rådhus: archaeological investiga245
73)
74)
75)
76)
77)
78)
79)
80)
81)
82)
83)
84)
85)
246
tion, moraine was measured between 3.8 and
3.97(Dunlop1980)(B).Bedrockwasobserved
at4.7maslto5.1masl(Dunlop1985b)(B).
Alle Helgensgate 3, Magistratsbygningen: archaeologicalinvestigation,bedrockobservedat
6.9masl(Dunlop1984i)(B).
Manufakturhuset:archaeologicalinvestigation,
moraine was measured between 1.55 and 3.7
masl(DunlopandKoch1985)(B).
Alle Helgensgate 3-5: archaeological investigation,morainewasdocumentedbetween3.8
and4.4masl(DunlopandHansen1993)(B).
Chr Michelsensgate: archaeological investigation, moraine was documented at 8.5 masl
(Hansen1991)(B).
Grønnevollen 2: archaeological investigation,
bedrock was recorded between 0.2 and 0.5 m
under the floor of the building, there was no
basement(Dunlop1984g).AccordingtoGeneralkart1879-80thebuildingisplacedbetween
4and5masl.Accordinglybedrockshouldbe
foundatabout4.5masl(S).
Rådstuplassen: archaeological investigation,
moraine deposits were recorded at 4.0 masl
(Dunlop1991e)(B).
Torggaten 1c-1d: archaeological investigation,
sterileblueclayabout1.5masl(KochUndated)
(S).
Lidohjørnet, Nedre Torgalmenning: archaeologicalinvestigation,sterilemasseswererecordedbetween-0.2and1.7masl(KochUndated)
(B).
Walkendorfsgate 5: archaeological investigation, a well, dug into the natural subsoil was
recorded. The bottom of the well was at 3.75
masl(Christensson1985),accordingtophotos
ofthewellmorainemassesappeartobeginat
about5.0masl(S).
StrømgatentowardsVestreStrømkaien:archaeologicalinvestigation,atrench0.50to0.60m
deep,wasdugandsterilemasseswereencounteredalongthebottomofthetrench(Dunlop
1984m).Accordingtothisthenaturalsubsoil
mustbequiteclosetothesurfaceasweseeiton
Generalkart1879-80(S).
Vincent Lunges gate: boreholes, bedrock and
morainewereencounteredbetween+/-and2.0
masl(Fritzvold1976)(S).
Bergen Rådhus 1956 and 1972: information
onthe1-3maslcontoursbasedonarchaeological investigations and boreholes (S) (Fritzvold
1976).
OleBullsPlass3:excavationbyhand,+/-0was
recorded(Fritzvold1976)(B).
86) Starvhusgaten between Torgalmenningen 14
and Olav Kyrresgate 11: archaeological investigation, information on moraine at 0.6 masl
(Fritzvold1976)(B).
87) OlavKyrresgate31:boreholes,thecontourlines
for1-2maslaregiven(Fritzvold1976)(S).
88) Permanenten, between Foreningsgaten and
NordahlBrunsgate:groundwork,information
onthe+/-0to3mcontours(Fritzvold1976)
(S).
89) Grieghallen: boreholes, the +/- 0 contour is
drawnbyFritzvold(Fritzvold1976)(S).
90) Strømgaten21:boreholes,the+/-0contouris
drawnbyFritzvold(Fritzvold1976)(S).
91) Torgalmenningen: archaeological investigation, bedrock was measured between 2.57-5.7
masl and 4.5-0.6 masl (Bjørndal and Dunlop
1992)(B).
92) Kaigaten: archaeological investigation, the
leveloftheoldestrecordedculture-layersisat
0.5masl(Koch1982b),sterilemassesmustbe
foundbelowthislevel(S).
93) Kaigaten 1c-5: archaeological investigation,
bedrockwasencounteredalmostdirectlyunder
the present day surface (Dunlop 1984k). The
contourlinesforthisareaarethereforedrawn
accordingtoGeneralkart1879-80.
94) Kaigaten4-6:archaeologicalinvestigation,moraine was recorded between 0.2 and 0.6 masl
(Göthberg1982)(B).
95) Badstuestredet2:archaeologicalinvestigation,
bedrock was encountered directly under the
floorlevelofthebuilding,thebuildinghadno
basement (Dunlop 1984d). The contour lines
forthisareaarethusdrawnaccordingtoGeneralkart1879-80.
96) Thecontoursfortheseabedareheredrawnon
thebasisofdatafrompoints51,53,106,110
and117andmustbetakenasasuggestion.AccordingtoFritzvold1976,theseabedwasmuch
deeper,butFritzvold’scontoursprobablyrepresent the bedrock surface, whereas the present
reconstruction suggests the surface of loose
sediments.
97) Olav Kyrresgate: groundwork, bedrock was
measured between 3.5 and 4.4 masl (Dunlop
andHansen1994a)(S).
98) Lungegårdsgaten, Marken: archaeological investigation, bedrock was encountered just belowtoday’ssurface(Dunlop1984l)(B).
99) Lungegårdsgaten 2: archaeological investigation, bedrock was encountered at about 9.2
masl(Christensson1980d)(B).
100) Marken 3: archaeological investigation, bed-
rock and moraine were encountered between
6.8and7.2masl(Dunlop1984e)(B).
101) Marken/Tverrgaten: archaeological investigation, bedrock was observed just below today’s
surface(Dunlop1984b)(B).
102) Marken/Tverrgaten4-6:archaeologicalinvestigation,bedrockwasrecordedbetween8and10
masl(Dunlop1984c)(B).
103) Nygaten 5: archaeological investigation, bedrock was recorded between 5.9 and 8.7 masl
(Johnson1988)(B).
104) Heggebakken/Sentrum: archaeological investigation, bedrock and moraine were recorded
between1.0and5.5masl(Koch1982a)(B).
105) Nygaten 2: archaeological investigation, bedrock was recorded between 3.0 and 6.5 masl
(Dunlop1991c)(B).
106) VågsalmenningenandOlavKyrresgate:boreholes,the+/-0and1.0maslcontourlinescan
bedrawn(Fritzvold1976)(S).
107) Halfdan Kjærulfs gate: archaeological investigation, bedrock and moraine were recorded
between3.3and4.0masl(Dunlop1993)(B).
108) KongOscarsgate67:archaeologicalinvestigation,morainemasseswererecorded0.8mbelow today’s surface (Dunlop Undated-b), that
isabout8.0masl(S).
109) Klosteret: archaeological investigation, morainemasseswererecordedbetween0.8and1.1
mbelowtoday’ssurface(DunlopUndated-a),
thatisabout25.0masl(S).
110) Rådstueplass 2-3: boreholes, the -4-2 contour
lines have been reconstructed for bedrock
(Fritzvold1976)(S).Archaeologicalinvestigation, the natural topography was documented
between+/-0and3.0masl(Næss1963)(B).
111) Sverresborgarea:thecontoursforbedrockare
taken from Generalkart 1879-80 (today parts
ofthebedrockformationsbetweenSverresborg
andHolmenareblastedaway).Ihavesurveyed
the area and compared today’s terrain with
Generalkart 1879-80. My impression is that
themapgivesafairlytrustworthypictureofthe
natural topography especially of course where
bedrockisvisible.Wherebedrockisnotvisible,
IhavemodifiedthecontourlinesfromGeneralkart1879-80accordingtomysurveyofthe
landscape.WhiledoingthisIhavetakeninto
consideration the elaborate building activity,
whichtheareahasbeenexposedtoduringhistory(S).
112) Holmen: the contours for bedrock are taken
fromGeneralkart1879-80.Ihavesurveyedthe
areaandcomparedtoday’sterrainwithGener-
alkart1879-80myimpressionisthatthemap
givesafairlytrustworthypictureofthenatural
topographywherebedrockisvisible(B).Where
bedrockisnotvisible,IhavemodifiedthecontourlinesfromGeneralkart1879-80according
to my survey of the landscape. While doing
thisIhavetriedtotakeintoconsiderationsome
of the elaborate building- and levelling activitytheareahasbeenexposedtoduringhistory
(FischerandFischer1980,11)(S).(Seediscussionbelow).
113) TheshorelinearoundHolmen:onoldermaps
severalrocksintheseaareseenalongtheHolmen shoreline. The shoreline on the present
reconstructioniscombinationof/compromise
betweenKartOverNordnesogFæstningenfrom
1872-73(Harris1991)andGeneralkart187980(S).
114) TheshorelinetowardsBradbenken:boreholes,
contourlines-8.0-1.0aretakenfromFritzvold
(1976)(S).
115) Veisan’sshorelineismostlytakenfromFritzvold
1976.Fritzvold’sreconstructionisbasedontest
drillingcarriedoutbyhim,andontestdrilling
carriedoutin1915and1916.(Fritzvold1976)
(S). Fritzvold’s reconstruction of the Veisan
shorelineisadjustedthroughinformationfrom
archaeologicalinvestigations(seepoints1,119,
123and124).
116) Skuteviken: all contour lines are copied from
Generalkart1879-80,themapdoes,however,
not give information on the height of all the
contourlines,whichmakesthecopyworkdifficultandsomewhatimprecise,thecontourlines
shouldbetakenasaguideonly.
117) Nedre Korskirkealmenning/ Vågsalmenningen: archaeological investigation, a deposit of
gyttja 14C dated to between 810-970 was recorded at -3.10 masl (Hjelle 1998) (B). This
levelismostlikelyrepresentativeforthelevelof
thenaturalsubsoilabout1000.
118) FromHeggentoKaigaten:boreholes,moraine
wasrecordedbetween-4and6.0masl(NOTEBY1978)(S).
119) Koengen: archaeological investigation, ‘beach
sand’ was encountered at 1.25 masl (Dunlop
1981a)(B).
120) Slottsgaten 3A: archaeological investigation,
the ‘sea bottom’ was recorded at -2 masl and
‘gravelwhichseemstobetheseabottom’at-1.0
masl(Enger1957,5)(B).
121) Bryggen BRM 0 area: the contours for -2 to
-3 are drawn on the basis of information on
theleveloftheoldestculture-layersonthesite
247
(Herteig 1990, 56, 90) (S) and on documentation of the natural topography published in
KrzywinskiandKaland1984,Figure3(B).
122) Holmen: archaeological investigation, the 4,
5and6maslcontoursaredrawnonthebasis
of the Håkonshallen BRM 474 site (Hansen
1995a)(s).The2and5masllinesaresupplementedwithdataformGerhardFischer’sinvestigationsinthearea(FischerandFischer1980)
(S).
123) Bergenhus-Bontelabo: archaeological investigation,greenish-bluemarinesandwasfoundat
2.2masl(Dunlop1989g)(B).(Seediscussion
below).
124) Bontelabo-Veisan: boreholes, in connection
withFritzvold’sreconstructionofthenaturaltopography,7boreholesweredrilledinthisarea.
Onthebasisoftheseboreholes,ofbedrockvisibleonthesurface,Generalkart1879-80and
other (unspecified, but probably test drilling
carriedoutin1915-16atKoengenRangerstasjonbyNSBseefurtherreferencesinFritzvold
1976,16)informationfromthearea,Fritzvold
reconstructsabedrockgullyfromBontelaboto
Veisan(Fritzvold1976,16)(S).(Seediscussion
below).
(Fritzvold1976,16).Thisimpliesthatthegully
betweenBontelabowaslikewisefilledwithloose
massestoalevelabovethesealevel.Havingthis
information at hand, Fritzvold still concluded
that it is not possible to exclude the possibility
thatseawatercouldenterVeisanfromtheBontelabosideabout1000(Fritzvold1976,16).
Since Fritzvold did his reconstruction of the
naturaltopography,moreobservationsofpre-urbandepositshavebeenmade.AttheBergenhusBontelabo site (1989) (point 123) marine sand
wasencounteredat2.2masl.Thepointofthis
excavationissoclosetothegully,asreconstructed by Fritzvold, that it is hard to imagine that
a stream could have flown here, without erodingawaythemarinesanddocumentedatthearchaeologicalsite(Figure61).ThusIsuggestthat
therehasnotbeenastreambetweenVeisanand
Bontelabo since prehistoric time when the sea
levelwashigher,andthatthethresholdbetween
BontelaboandVeisanwasmostlikelyabout2.2
maslorhigherabout1000.
125) HolmentowardsVeisan:archaeologicalinves-
tigation,morainewasrecordedbetween1and
Discussion of the area around points 123 and
1.3masl(Hommedal1999)(B).
124:inoldNorwegiantheplacenameHolmen
126)
Dreggsalmenningen20:archaeologicalinvestimeans ‘islet’ or ‘a small hill on a flat piece of
gation,morainewasrecordedbetween4.1and
90
land’ (Fritzner 1973 (1867)). Whether or not
0.9masl(Larsen1967b)(B).The+/-0contour
Holmen in ‘early historic time’ was completely
isreconstructedonthebasisofthegradientof
surroundedbytheseahasbeenasubjectofdisthenaturalsubsoilandontheorientationofthe
cussion.Stressingthefirstmeaningoftheword,
oldeststructuresonthesite(S).
Munch (1855) and Koren-Wiberg (1908) as- 127) Øvregaten25-29:archaeologicalinvestigation,
sumed that Holmen in early historic time was
bedrockwasencounteredat9.1masl(Dunlop
1996b)(B).
surrounded by water (Grimnes 1937). Against
this view Grimnes (1937) argued that the general outline of the topography between BontelaboandVeisanratherfavoursthesecondmeaningofthename.In1976bedrockmeasurements
from test drilling provided more information
on the bedrock formations between Bontelabo
and Veisan. The measurements indicated a below +/- 0 gully between Bontelabo and Veisan
(Fritzvold1976,16).Ifthisgullywasnotfilled
inwithloosedeposits,theseamayhaveaccessed
VeisanfromtheBontelabo-sideabout1000,thus
makingHolmenanislet.Othermeasurements,
however,showedthatmoraineorsandhasgenerallyfilledupgulliesinthebedrockformations
oftheHolmenareatoalevelof+/-0or1masl
248
249
Figure61.DetailoftheareabetweenBontelaboandVeisan
Figure62.Sourcesforthepre-urbantopography‘thenaturaltopography’.
250
Figure62a.Sourcesforthepre-urbantopography‘thenaturaltopography’.
251
Figure62b.Sourcesforthepre-urbantopography‘thenaturaltopography’.
252
Figure62c.Sourcesforthepre-urbantopography‘thenaturaltopography’.
253
Figure62d.Sourcesforthepre-urbantopography‘thenaturaltopography’.
254
Figure62e.Sourcesforthepre-urbantopography‘thenaturaltopography’.
255
Figure62f.Sourcesforthepre-urbantopography‘thenaturaltopography’.
256
APPENDIX2
DateddendrochronologicalsamplesfromearlyBergen
Constructiontype
Constructionnumber
Phaseinsitereport
Date
Signsofreuse?
Surfacework?
6/B
building
044
2.1
after1080
j
?
0/92689
6/B
building
044
2.1
after1100
j
?
0/92687
6/B
building
044
2.1
after1100
j
n
0/92685
6/B
building
044
2.1
after1008
j
j
0/92542
6/B
5
building
038
2.0
1104/05
j
?
0/91006
6/B
5
building
038
2.0
after1100
j
?
0/Dno01325
6/B
5
building
038
2.0
after1141
0/92832
6/B
5
building
041
2.2
after1127
n
n
0/92835
6/B
5
building
041
2.2
after1131
n
n
0/92836
6/B
5
building
041
2.2
after1133
n
n
0/92696
6/B
5
building
041
2.2
after1134
0/92694
6/B
5
building
041
2.2
after1135
n
n
0/92786
6/B
5
building
066
2.0
after1024
n
n
0/92785
6/B
5
building
066
2.0
after1040
n
n
0/92798
6/B
5
building
066
2.0
after1127
n
j
0/93053
6/B
5
caisson
025
2.2
1121/22
n
n
0/93054
6/B
5
caisson
025
2.2
after1083
n
n
0/Dno01413
6/B
5
caisson
026
2.0
after1120
0/92914
6/B
5
caisson
038
2.0
1138/39
n
n
0/92916
6/B
5
caisson
038
2.0
after1138
n
n
Sampletakenby
Plotnumber
0/92688
Horizon
Dendronumber
Legend:
’after 1144’ - some tree rings may be missing, the felling year for the tree cannot be established.
’1128/1129’-thepreservedoutertreeringrepresentsthelastyearofgrowth.’j’-yes,’n’-no.’surface
work?-surfaceworkonthespotwherethesamplewastaken
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
Hansenand
Reimers
Hansenand
Reimers
257
258
after1128
2.0
after1134
2.0
after1137
045
1.2
1108/09
n
j
building
045
1.2
after1100
n
j
4
building
045
1.2
after1107
n
?
6/C
4
building
045
1.2
after1109
n
n
0/92716
6/C
4
building
045
1.2
after1110
n
n
0/Dno01415
6/C
4
caisson
027
2.0
after1106
0/93040
6/C
4
caisson
028
2.0
1108/09
n
n
0/93039
6/C
4
caisson
028
2.0
after1098
n
n
0/92936
6/C
4
caisson
029
2.0
after1074
n
n
0/92935
6/C
4
caisson
029
2.0
after1104
n
n
0/Dno01075
6/C
5
building
040
2.0
after1103
0/Dno01170
6/C
5
building
040
2.0
after1149
0/Dno01110
6/C
5
building
498
2.0
after1122
j
2.0
1124/25
n
n
post
0/Dno01382
6/B
5
0/Dno01253
6/B
5
0/Dno01254
6/B
5
0/Dno01395
6/B
5
0/Dno01398
6/B
5
0/92714
6/C
4
building
0/92717
6/C
4
0/92705
6/C
0/92704
quay
structure
quay
structure
quay
structure
quay
structure
quay
structure
0/92907
6/C
5
quay
structure
0/93029
6/D
2
post
1.1
after1026
n
n
0/93028
6/D
2
post
1.1
after1029
n
n
0/Dno01155
6/D
5
building
2.0
after1068
0/92909
6/D
5
post
2.0
after1149
n
n
0/Dno01466
6/E
5
building
130
2.0
after1112
0/Dno01471
6/E
5
building
130
2.0
after1117
0/Dno01469
6/E
5
building
130
2.0
after1128
042
Sampletakenby
2.0
3
Surfacework?
Date
after1126
6/B
Signsofreuse?
Phaseinsitereport
Constructionnumber
2.0
Constructiontype
after1125
Horizon
2.0
Plotnumber
after1069
Dendronumber
1.1
0/Dno01537
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
During
excavation
During
excavation
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
Hansenand
Reimers
During
excavation
During
excavation
During
excavation
Constructiontype
Constructionnumber
Phaseinsitereport
Date
building
130
2.0
after1129
0/Dno01460
6/E
5
building
402
2.0
after1117
0/Dno01474
6/E
5
building
402
2.0
after1123
0/Dno01476
6/E
5
building
402
2.0
after1125
0/Dno01478
6/E
5
building
402
2.0
after1128
0/Dno01479
6/E
5
building
402
2.0
after1128
0/Dno01517
6/E
5
building
482
2.0
after1116
0/Dno01518
6/E
5
building
482
2.0
after1120
0/Dno01515
6/E
5
building
482
2.0
after1120
0/Dno01516
6/E
5
building
482
2.0
after1124
0/Dno01514
6/E
5
building
482
2.0
after1125
0/Dno01454
6/E
5
caisson
041
2.0
after1126
0/Dno01512
6/E
5
passage
2.0
after1124
0/Dno01485
6/E
5
post
2.0
after1113
0/Dno01449
6/E
5
post
2.0
after1123
0/Dno01484
6/E
5
post
2.0
after1123
0/Dno01459
6/E
5
post
2.0
after1124
0/Dno01480
6/E
5
post
2.0
after1124
0/Dno01513
6/E
5
post
2.0
after1124
0/Dno01468
6/E
5
post
2.0
after1128
0/Dno01427
6/F
5
passage
2.0
after1029
110/06096
26-27/B-C
5
caisson
053
2
after1144
n
n
104/02489/VIII
26/A
4
caisson
037
12
after1090
j
n
104/02489/VII
26/A
4
caisson
037
12
after1099
n
n
104/02489/X
26/A
4
caisson
037
12
after1100
n
n
104/02489/II
26/A
4
caisson
037
12
after1102
n
n
76/15190
28/B
5
caisson
001
1
after1137
n
Sampletakenby
Horizon
5
Surfacework?
Plotnumber
6/E
Signsofreuse?
Dendronumber
0/Dno01470
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
Hansenand
Reimers
During
excavation
259
260
Constructiontype
Constructionnumber
Phaseinsitereport
Date
caisson
001
1
after1139
76/15161
28/B
5
caisson
002
1
after1127
76/15175
28/B
5
caisson
002
1
after1128
76/15162
28/B
5
caisson
003
1
after1140
76/15209
28/B
5
caisson
003
1
after1141
76/15207
28/B
5
caisson
003
1
after1141
76/15195
28/B
5
caisson
004
1
after1141
76/15168
28/B
5
caisson
004
1
after1141
76/15197
28/C
5
post
1
after1141
490/00026
29/A
245/02985
38/A
5
n
post
16
4
1128/29
j
post
343
10
after1128
n
Sampletakenby
Horizon
5
Surfacework?
Plotnumber
28/B
Signsofreuse?
Dendronumber
76/15163
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
During
excavation
APPENDIX3
Elevenartefactassemblagesfromsite
9,Sandbrugaten5(1967)BRM3
Thedateofelevenartefactassemblagesarediscussedinordertothrowlightupontheabsolute
date for the material at site 9, Sandbrugaten 5.
Theassemblageshavebeenselectedbycomparinginformationfromtheartefactdatabasefrom
site9,91theoriginaldrawings,commentstothe
drawings and the report from the excavation
(Larsen 1967a). The assemblages were selected
withthespecificaimofdatingstructuresinthe
lowestlevelatthesite.However,inordertoobtainareliablepictureoftheabsolutechronology
ofthesiteanattemptwasmadetodateallclosed
contexts found through the documentation. A
closedcontextisdefinedasanartefactassemblage
witharelativelyclearconnectiontoastructure.
In the original documentation the structures
werenotgivennumbers.Instead,structureswere
described by square and levelling number and
agenerallabelforexample‘floor’,‘drain’.Contexts for artefacts were described according to
the original field information. The Norwegian
descriptionofthecontextiscitedinbracketsand
translated.ShoesareclassifiedaccordingtoSchia
1977,GrewanddeNeergaard1988,andLarsen
1992.InFigure14the11assemblagesareseenin
relationtothestratigraphyatsite9.
Assemblage1
AccessionnumbersBRM3/24-25,BRM3/7273,BRM3/232.
Context:‘O-11andO-10.Inthewell’.(‘Ibrønnen’). Well 1 is the only well in squares O-11
and O-10, the finds must therefore stem from
well 1. Well 1 is assumed to be contemporary
with building 8 (Larsen 1967a). When well 1
andbuilding8fallintodisuse,theyarebothsuperposedbybuilding9.Theartefactsinwell1
muststemfromthetimebeforebuilding9was
constructed, the finds can therefore give a post
quemdateforbuilding9.
Datingfinds:
BRM 3/72, one sherd of Grimston Decorated
ware.
According to the prevailing date for Grimston Decorated ware (Jennings and Rogerson
1994),building9musthavebeenconstructed
afterc1225.
Assemblage2
Accessionnumbers3/90-100.
Context:N-11.‘Underthefloor’.(‘Undergolvlaget’).
Accessionnumber3/170.
Context:N-11.‘ByandunderthefloorinN-11.
In the sand’. (‘Inntil og i underkant av gulv i
N-11. I sanden.’). The only floor which can be
localisedtosquareN-11isthefloorofbuilding
8,thusthefindsmustcomefrombyandunderneaththefloorofbuilding8,andtheyprovidea
postquemdateforbuilding8.
Datingfinds:
BRM3/92,1sherdofArdenburgpottery.
BRM3/93,1sherdofScarboroughIIpottery.
BRM3/94,1sherdofScarboroughIIpottery.
BRM3/95,1sherdofGrimstonware.
BRM3/96,1sherdofDecoratedGrimstonware.
BRM3/97,1sherdofDecoratedGrimstonware.
BRM3/98,1sherdofDecoratedGrimstonware.
BRM3/99,1sherdofGrimstonware.
BRM3/170,4sherdsofGrimstonware.
The sherds of Decorated Grimston ware give a
postquemdateforbuilding8toafterc1225,accordingtotheprevailingdateforthisware(JenningsandRogerson1994).
Assemblage3
Accession numbers BRM 3/530, BRM 3/535539.
Context:‘M-11,infill-massesoverthepavement’.
(‘I fyllmasser i overkant av gangbroen’). The
pavement may be the timber passage (kavlpassage),whichismentionedinthereportasthere
isnomentionofotherpavementsorpassagesin
squareM-11.Theartefactsmaythereforegivea
postquemdateforthedestructionofthetimber
passage.
Datingfinds:
BRM3/535,5sherdsofScarboroughIIware.
BRM 3/538, 1 sherd of Scottish White Gritty
wareorYorkWhiteware.
BRM3/539,1sherdofSiegburgStoneware.
261
Theyoungestfindinthisassemblageisasherd
ofSiegburgStoneware,whichisusuallyfound
fromc1300andon(Lüdtke1989,33).Thetimberpavementmaythereforehavegoneoutofuse
afterc1300.
BRM3/703/2,thesoleofBRM3/703/1.
BRM3/704/1,1upperofalowthongshoewith
slits in pairs, top band along the instep and a
band of decoration running up the middle of
the front piece from the toe to the instep (embroiderypatternA(Larsen1992)).Thetoewas
Assemblage4
pointedandtheheelwasrounded.
Accession numbers BRM 3/571-584, BRM BRM704/3,1fragmentofathongshoeupper
3/681-683
withslitsinpairs.
Context:‘M-11andM-12,infill-massesinthe
levelundertheeast-westgoingpassage’.(‘Ifyll- Fragmentsofthreeshoeswerefound.Allthree
masserinivåetunderØ-Vpassasjen’).Theeast- werethongshoesandatleasttwohadapointed
westgoingpassagemaybethewoodenpassage toe. High and low thong shoes are found in a
(kavlpassage)whichismentionedinthereport, widetimerange.InthematerialfromtheGullasthereisnomentionofotherpavementsorpas- skoenareaatsite6,Bryggentheshoetypesare
sagesinsquareM-11.Theartefactsmaytherefore foundinperiods2-7(Larsen1992)datedfrom
giveapostquemdateforthetimberpassage.
the1120sto1476(Herteig1990;Herteig1991;
Datingfinds:
Hansen1998).However,thetypesdominatein
BRM3/575,1sherdofScarboroughIIware.
the older periods, making up for almost 80 %
BRM3/576,1sherdofScottishGrittywareor oftheshoematerialinperiod2,andalmost50
YorkWhiteware.
% of the shoes in period 3 (Larsen 1992, FigBRM3/577,1secondarilyburntsherdofGrim- ure54).Periods2and3atsite6aredatedfrom
stonorHumberware.
the1120sto1198(Hansen1998).AttheFolkebibliotekstomtensiteinTrondheimthongshoes
The Scarborough II sherd is the youngest type arefoundinphases4-8(Marstein1989,10and
present and this gives a post quem date for the 87),dominatinginphases4-6,datedtobetween
timberpassagetoafterc1215/1225(Farmerand ‘1050andc1175’(ChristophersenandNordeide
Farmer1982).
1994,35).Shoeswithapointedtoeareknown
fromallperiodsintheGullskoenarea,butthey
Assemblage5
are most common in periods 2 and 3 (Larsen
AccessionnumbersBRM3/692-704.
1992,Figure55)datingfromthe1120sto1198
Context: M-12. ‘Under the stone layer in fill- (Hansen1998).TheFolkebibliotekstomtensite
masseswithwoodchips’.(‘Understeinlagifyll- material is not classified by toe shape alone, so
massem.Spon’).Onlybuildings10and11are comparisonisdifficult.
referredtoasa‘stonelayerinsquareM-12’,soit
is likely that that the stone layer is identical to Thechronologicaldistributionofshoematerial
buildings10and11.Ifthisisthecase,thenthe fromtheGullskoenareaatsite6andtheFolkefindsintheassemblagegiveapostquemdatefor bibliotekstomtensiteistoowideforanystrong
buildings10and11.Buildings10and11aresit- conclusions to be reached on the dating of asuated stratigraphically under fill-masses, which semblage5atsite9.However,ifweaddthefact
were under the east-west going timber passage that all three shoes in assemblage 5 are thong
(kavlpassage). These fill-masses are dated to shoesandthatatleasttwoofthemhadpointed
‘afterc1215/1225’onbasisofthepresenceofa toes, this high proportion may imply that the
ScarboroughIIsherdinassemblage4.
assemblage stems from the time period domiDatingfinds:
natedbythethongshoeandwheretheelement
BRM3/703/1,1upperofalowthongshoeIII ofpointedtoewasmostcommon-thatisinthe
with slits in groups, pointed toe and rounded periodbetweenc1050andc1198accordingto
heel.
the Gullskoen area and Folkebibliotekstomten
site materials. A broad date of assemblage 5 to
262
between 1050 and 1200 is in accordance with
thefactthatassemblage5wasfoundstratigraphicallyunderassemblage4,datedtoafterthefirst
quarterofthethirteenthcentury.Ithereforesuggestthatassemblage5datesbroadlytobetween
c1050and1200.Thisgivesasimilarpostquem
datingforbuildings10and11.
Assemblage6
AccessionnumbersBRM3/755-764.
Context: M-10/pl. 5. ‘In well 5’. (’I brønd 5’)
(Thiswellisthesameaswellno4inthereport).
Well4cutsthroughremainsofbuildingswhich
burnedinfire3andthewellwassuperposedby
fire-layer 2. Therefore well 4 must be younger
than fire 3 but older than fire 2. The finds in
well4cangiveapostquemdateforfire2.
Datingfinds:
BRM3/760,1sherdofacookingpot.
BRM3/761,1sherdofYorkWhiteware.
BRM3/762,1sherdofPaffrathware.
BRM3/763,1sherdofacookingpot.
BRM3/764,1sherdofacookingpot.
remainsoffoundationrafts.
AccessionnumbersBRM3/887-894.
Context: K-11/pl. 4. In fill-masses (‘I fyldlag’).
Onplan4weseesomethingwhichlookslikethe
remainsoffoundationrafts.
AccessionnumbersBRM3/904-911.
Context:M-10/pl10.Infill-masses.Inthelevel
undertimbers(‘Ifyldlag.Inivåetundertreverk’).
Ontheoriginaldrawingthewoodseemstolie
closetothenaturalsubsoil,thefill-massesmay
thusbelongtothefirststageoffilling-intheVågenbay.
AccessionnumbersBRM3/924-926.
Context:‘M-10/pl.9.’Onplan9weseethesame
woodasonplan10,thewoodseemstolieclose
to the sterile masses, therefore the fill-masses
maybelongtothefirststageoffilling-intheVågenbay.Assemblage8maystemfromfill-masses
from the first expansion into Vågen. The artefactsaremostlikelyredepositedandreflectthe
materialcultureinearlierphases.
Datingfinds:
BRM3/888,1sherdofAndennepottery.
BRM3/889,1sherdofAndennepottery.
Theyoungesttypeofpotteryinthisassemblage
isthesherdofYorkWhiteware.Thesherdgives Andenne ware is produced from the eleventh
apostquemdateforfire2toaftertheendofthe centuryupuntilthemiddleofthefifteenthcentwelfthcentury.(Reed1990).
tury(Reed1990,38).
Assemblage7
AccessionnumbersBRM3/803-812.
Context:L-11/pl.5.‘Inwell3’(‘Ibrønd3’).This
wellisnumber6inthereport.Well6liesstratigraphicallybelowfire2anditismostlikelyconnectedtobuildingswhichburnedinfire3.The
assemblage may give us a clue to the dating of
thisplaceinthestratigraphy.
Datingfinds:
BRM3/808,1sherdofSiegburgStoneware.
StonewarefromSiegburgisfoundfromc1300
(Lüdtke1989).Thesherdgivesapostquemdate
forfire2.
Assemblage8
Accession numbers BRM 3/881-886, BRM
3/895-900.
Context:L-11/pl.6.‘Infill-masses’(‘Ifyldlag’).
Onplan6weseesomethingthatlookslikethe
Assemblage9
Accession numbers BRM 3/901-903, BRM
3/1004-1010.
Context:M-11/pl.11.‘Foundinfill-massesunderthefloormadeofthinlogs’.(‘Funnetifyllmasserunderstrangedækket’).Thefloorbelongs
tobuilding5andtheassemblagewasfoundin
layers under the floor of building 5. The floor
consisted of thin logs laid side by side, trash
could easily fall between the floor-logs and be
depositedasculturallayersunderthefloor.The
layersunderthebuildingmaythereforestemeither from the time before building 5 was constructed or from the period when the building
wasinuse.
Datingfinds:
BRM3/903,1fragmentoflowthongshoeupperwithslitsinpairsandroundedtoe.
BRM3/1009/1,1shoesolewithpointedtoeand
roundedheel.
263
BRM3/1009/2,1highthongshoeupperwith
slitsinpairs,pointedtoeandroundedheel.
BRM3/1010/1,1frontpartofashoe-upperwith
abandofdecorationrunningupthemiddleof
the front piece from the toe to the instep (embroiderypatternA).Thetoewasskewed.
BRM3/1010/2,1upperoflowthongshoewith
slits in pairs, top band along the instep and a
band of decoration over the instep and up the
middleofthefrontpiecefromthetoetotheinstep(embroiderypatternC).Thetoewaspointedandtheheelwasrounded.
BRM 3/1010/3, sole with pointed toe and
roundedheel.
BRM3/1010/4,laceholepiecefromalowside
lacedshoe(variant1).
BRM3/949/1,1fragmentofalowthongshoe
upperwithdenselycutslitsandroundedheel.
BRM3/949/2,1fragmentoflowthongshoeupperwithslitsinpairs.
BRM3/949/3,1fragmentofalowshoeupper
withtracesoftopbandalongtheinstepandheelstiffener.Theheelwasrounded.
BRM3/949/4,1fragmentofathongshoeupper
withapointedtoe.
Fragments of seven shoes were found. At least
three of these were thong shoes and one was a
lowsidelacedshoe,therestcannotbeclassified.
Four out of six shoes had pointed toes. As we
sawunderthediscussionofassemblage5,ahigh
proportion of thong shoes and of the element
ofpointedtoemayindicatethattheartefactassemblagedatesfromthelateeleventhcenturyor
thetwelfthcentury.Thelowsidelacedshoeof
type1isnotfoundbeforephase5attheFolkebibliotekstomten site in Trondheim (Marstein
1989),thatisfromc1100(Christophersenand
Nordeide1994,35).IntheGullskoenareaatsite
6thetypeisfoundfromperiod2andonwards.
The shoes provide a tentative dating frame of
c1100-1200forassemblage9.
Andennepotterywasproducedfromtheeleventh
centuryupuntilthemiddleofthefifteenthcentury (Reed 1990, 38). Fragments of four shoes
werefound,threewereidentifiedasthongshoes.
Asseenunderthediscussionofthematerialfrom
assemblage 5 the chronological distribution of
thethongshoematerialfromtheGullskoenArea
atsite6inBergenandtheFolkebibliotekstomten
site in Trondheim provides a very wide dating
frameforthongshoes.Still,aswithassemblage
5,thefactthatasmuchasthreeoutoffourshoes
are thong shoes, would imply that assemblage
10shouldbedatedtotheperioddominatedby
thistypeofshoe,thatisintheperiodbetween
c1050andc1200.ThepresenceoftheAndenne
ware in the assemblage cannot help us narrow
this time span as Andenne ware was produced
for a long time. It is difficult to be conclusive
aboutthedatingofassemblage10onthebasisof
finds.Isuggestthatassemblage10maybedated
tentativelytobetweenc1050andc1200onthe
basisofthepredominationofthongshoesinthe
material.
Assemblage10
AccessionnumbersBRM3/946-949.
Context:L-11/pl.8.‘Infill-masses.0.94masl’(‘I
fyllmasser.0.94moh’).
The artefacts must have been found close to
buildings13,14orcaisson2,asthesewerethe
onlyconstructionsonplan8.Itis,however,uncertainwhethertheseartefactsarecontemporary
with the constructions, because the fill-masses
whichcontainedtheartefacts,mayhavebeendepositedwhentheconstructionswentoutofuse
andtheareawasfilledoutaspartoftheexpansionofthebuilt-upareaintotheVågenbay.
Datingfinds:
BRM3/947,1sherdofAndenneware.
Assemblage11
AccessionnumberBRM3/951.
Context: ‘L-10/pl. 6. -7300x/8300y/0.79 masl.
Infill-massescontainingwoodchips’,(‘Isponog
treflisholdige fyllmasser’). The assemblage was
foundwithinthewallsofbuilding12.
Accession numbers BRM 3/952-965, BRM
3/984-997.
Context: ‘L-10/pl.8. Within the walls marked
26-45, 50-58’. (‘Indenfor tilevæggene mrk. 2645, 50-58’). This corresponds to building 12.
Itisnotpossibletodecidewhetherthefindsare
contemporarywithbuilding12,orwhetherthey
belongtofill-massesspreadafteritwentoutof
use,sothecontextisunsafe.
264
Datingfinds:
Pottery
BRM3/956,1sherdofAndenneware.
BRM3/957,1sherdofScottishGrittywareor
YorkWhiteware.
BRM3/987,1sherdofYorkWhiteware.
BRM3/988,1sherdofAndenneware.
BRM3/986,1sherdofModernRedware.
Shoes:
BRM3/997/1,solewithroundedtoeandheel.
BRM3/997/2,pumpwithroundedtoeandheel,
decorationalongtheinstep(embroiderypattern
B(Larsen1992)),edgebandalongtheinstep.
Pumps are found in small quantities at site 6
from period 2 until period 6, dated from the
1120sto1413(Larsen1992).AttheFolkebibliotekstomtensitepumps(SUL1and2)arefound
insmallquantitiesfromc1100-1375(Marstein
1989),thusleavinguswithmuchtoowidedating frames to be of any help here. We have to
relyonthedatesprovidedbytheceramicmaterial.ThetypeofYorkWhitewareencountered
hereisfoundfromc1200(perscomAlanVince
1998)andthisistheyoungestwarefoundinthis
context.ThesherdofModernRedwaremustbe
considered an intrusion. I suggest that assemblage11maybedatedtentativelytoaftertheend
ofthetwelfthcentury(afterc1200).
scribedbyfindcontext,thatissquareandlevelling number and sometimes a general label for
example‘floor’,‘drain’.Thefindcontextsforartefactsweredescribedaccordingtotheoriginal
fieldinformation.Thedescriptionofthecontext
oftheassemblagesisgenerallypoorandonlyin
afewinstancescouldtheartefactsberelatedto
astructure.Onlyassemblageswithpotteryand
comb material have been analysed. Combs are
classifiedaccordingtoWiberg1977andFlodin
1989.Table73showspotteryandcombsinthe
assemblages.Figure16showstheassemblagesin
relationtothestratigraphyatDreggsalmenningen20.
Assemblage1
AccessionnumbersBRM4/1630-1633.
Context:W-9/pl1.‘Indrain1,underfire-layer
2’.Drain1correspondstodrain3inthereport.
AccessionnumbersBRM4/2215-2234.
Context:W-9/pl1‘Northofdrain1,underfirelayer 2’. Drain 1 corresponds to drain 3 in the
report.
AccessionnumbersBRM4/2270-2275.
Context:W-9/pl2‘Indrain1,underfire-layer2’.
Drain1correspondstodrain3inthereport.
Assemblage1shouldprovideanapproximatepost
quemdatefortheculture-layersbetweendrain3
andfire-layer2
AsherdofScarboroughIIwareintheassemblage
datestheculture-layersbetweendrain3andfirelayer2toafterc1215/25,accordingtotheprevailing date for this ware (Farmer and Farmer
APPENDIX4
1982).Thismaygiveasimilardateforfire-layer
Sevenartefactassemblagesfromsite11, 2inSquareW--9.
Dreggsalmenningen20BRM4(1967)
Table73.Potteryandcombsinassemblages1-7from
Thedatesofsevenartefactassemblagesaredis- Dreggsalmenningen20(1967)BRM4
cussedinordertothrowlightupontheabsolute
DatingArtefact
Assemblage 1
datefortheoldestmaterialatsite11,Dreggsal- Pottery:
4
menningen20site.Theassemblageshavebeen Andenne
selected by comparing information from the Blackwarelowfired
Cookingpot
artefact database from site 1192 with the origi- DevStamford
3
nal drawings, comments to the drawings and Grimston
thereportfromtheexcavation(Larsen1967b). GrimstonDecorated
Hedon
1
The assemblages were selected aiming to date Humber
structures in the lowest level at the site. In the London(unspecified)
1
original documentation structures were rarely LowCountriesHighlydecorated
NearStoneware
1
given numbers, instead the structures were de- Paffrath
1
2
3 4 5 6 7
1 1
1 1
1
2
3
3
4
1
1 1
1
3 1 1 2 1
2
1
1
2
265
Pingsdorf
ScarboroughII
ScottishWhitegritty
Shelly(unspecified)
SouthScandinavian
Yorkshire(unspecified)
Postmedievalredware
Uncertain
Combs:
2
1
2
1
assumesisfire-layer2(Larsen1967b).
AccessionnumbersBRM4/2649-2656
Context:X-9/pl1.‘Underfire-layer2’
2
AccessionnumberBRM4/2657
1
1
1
Context:X-9/pl3.‘Underfire-layer2’
1
6
8
2
4
AccessionnumbersBRM4/3137-3150.
E6/1 D4
Context: X-9/pl 1. ‘North of drain, under firelayer 2’. The drain corresponds to drain 5 in
Assemblage2
thereportsincenootherdrainsaredescribedin
AccessionnumberBRM4/4064.
thesquare.Drain5isunderafire-layer,which
Context: W-9/pl 1. ‘In a building, under floor, Larsenassumesisfire-layer2(Larsen1967b).
under fire-layer 2’. The building must correspondtobuilding9or10,asthesearetheonly Assemblage3providesanapproximatepostquem
buildingsinsquareW--9/pl1.
date for the culture-layers in and around the
draininsquareX-9andthefire-layerabove.
Accession number BRM 4/4064 is a comb of According to the Scarborough II sherds in the
typeD4.CombsoftypeD4werefoundinphas- assemblagetheculture-layersbetweenthedrain
es 8 and 9 at the Folkebibliotekstomten site in andfire-layer2maystemfromafterc1215/25
Trondheim.Thesephasesaredatedtobetween (FarmerandFarmer1982).Thismaygiveasim1225 and 1325 (Christophersen and Nordeide ilarpostquemdateforfire-layer2insquareX-9.
1994).ThetypeisnotverycommoninOslo;at
the‘Mindetstomt’siteonecombwasfoundun- Assemblage4
derfire5(Wiberg1977,209),datingtoc1350 AccessionnumbersBRM4/2674-2680.
(Molaug1977,111).Onewasfoundatthe‘Søn- Context:X-8/pl1.‘Underfire-layer2’.
dre Felt’ site in fire-level 5 (Wiberg 1987, 419) Assemblage4providesanapproximatepostquem
dating to between c 1275 and 1350 (Molaug date for the culture-layers under fire-layer 2 in
1987,313).Atsite6,Bryggen10combsofthis the square and this may give a similar date for
typehavebeenfound,8werefoundinperiods thefire-layerabove.
3and4datedto1170/71-1198,and1198-1248 According to the Scarborough II sherds in the
respectively,thetworemainingcombswerefrom assemblage the culture-layers under fire-layer
youngercontexts.93Ifthecombfromassemblage 2 may stem from after c 1215/25 (Farmer and
2hasasimilardatingframeastheexamplesfrom Farmer1982).Thismaygiveasimilarpostquem
Oslo, Trondheim and site 6 in Bergen it dates dateforfire-layer2insquareX-8.
buildings9or10toafterc1170/71.
Assemblage5
Assemblage3
AccessionnumbersBRM4/3114-3122
AccessionnumbersBRM4/1634-1644.
Context:U-9/pl6.‘Underbuilding,underfloor’.
Context:X-9‘Underfire-layer2’.
The building may be building 5 since this is
AccessionnumbersBRM4/1695-1700.
theonlybuildingwithafloorpreservedonthis
Context: X-9/pl 1. ‘In drain 2’. Drain 2 corre- plan.
spondstodrain5inthereportasnootherdrains
aredescribedinthesquare.Drain5isundera The presence of Grimston, Low Countries
fire-layer, which Larsen assumes is fire-layer 2 Highly Decorated, and York wares shows that
(Larsen1967b).
theassemblageshouldbedatedtoaftertheend
AccessionnumbersBRM4/1806-1807.
ofthetwelfthcentury(Reed1990,30-31;MadContext: X-9/pl 1 ‘In drain 3, under fire-layer sen1996,22).Thisgivesasimilardateforthe
2’. Drain 3 corresponds to drain 5 in the re- buildingandthefloor.
portsincenootherdrainswererecordedinthe
square.Drain5isunderafire-layerwhichLarsen
266
1
9 2
Assemblage6
AccessionnumbersBRM4/3190-3195.
Context: U-9/pl 6. ‘West of building’. The
buildingisprobablybuilding5.Theassemblage
shouldthenbefromthefill-massesabovebuilding12.Assemblage6shouldthusprovideanapproximatepostquemdateforwhenbuilding12
wentoutofuse.
ThepresenceofLowCountriesHighlyDecoratedwareindicatesthattheassemblagewasdepositedaftertheendofthetwelfthcentury(Madsen
1996,22).
Assemblage7
AccessionnumbersBRM4/4045-4061.
Context: U-9/pl 7. ‘Under passage, under firelayer 1’. The passage is probably contemporary
withbuilding5,whichislaterthanbuilding12
(according to plan 5-7/U-9). The assemblage
maythereforestemfromthelevelbetweenbuilding5and12.
AccessionnumbersBRM4/4197-4222.
Context:U-9/pl7.Underfire-layer1’.Theassemblagemaystemfromthelevelbetweenbuildings
5 and 12 as this corresponds to the level excavatedinplan7/U-9.
The assemblage provides an approximate post
quemdateforthedepositionoffill-massesontop
of building 12 when this building went out of
use. This may provide an approximate date for
thefirewhichscorchedbuilding12andcausedit
tofallintodisuse.
The presence of Scarborough II pottery indicatesthatassemblage7shouldbedatedtoafter
c1215/1225(FarmerandFarmer1982).Asherd
ofPostMedievalRedwaremustbeconsidered
anintrusioninthiscontext.
267
FOOTNOTES
1
‘Doubletenements’aredoublerowsofbuildingsthatrunat90degreestothewaterfront(cfp173ff ).
2
ThemapsareconstructedinthegeographicalinformationsystemMapInfousingthelocalcoordinate
system‘Bergenlokalekoordinatsystem’,where6000X/6000YislocatedbyDomkirkenthepresentday
cathedralchurch.Thiscoordinatesystemwasusedatarchaeologicalinvestigationsbetween1979and1995.
Between1955and1979alocalsystemdevelopedfortheBryggenexcavationswasappliedatallmajorsites,
CuratorEgillReimershaskindlyconvertedtheBryggencoordinatesintotheBergenlokalekoordinatsystem.
3
TheVågenBayisaccordingto‘townnorth’orientednorth-south.Asmymapsaredrawnonthebasisof
ageographicalcoordinatesystem,thelocaltraditionfordescribingfeaturesinBergenaccordingto‘town
north’cannotbefollowedonmymapstheVågenBayisthereforeorientednorthwest-southeast.
4
IntheBergenareathenaturaltopographywascharacterisedbyslopingmorainicsurfacesaswellassteep
hillsandrocks.Itislikelythattheinhabitantsaimedtoleveloutthemostextremedifferencesofheight
whenbuildingactivitiestookplaceinnewphasesanditmaybeproblematictodetermineiftheoldest
culture-layersatasiteactuallyreflectthefirstactivityatthelocation.Thisisonlyconsideredarealproblem
whentheoldestculture-layersarelocatedonhighpartsofbedrock.Asaruleofthumb,stratawhichare
locateddirectlyontopofprotrudingpartsofbedrockarenottrustedtobetheoldestremainsoflanduse
atalocation,unlessspecialcircumstancescallforit.
5
Themethodofinvestigatingthenumberofmissingtreeringsonadendrosample,throughsapwood
statisticshasnotyetbeendevelopedforpine(PerscomThomasSBartholinattheNationalMuseumof
DenmarkCopenhagen.June2004).
6
ThedendrosamplesweretakenfrommaterialinthestoreroomsoftheBergenUniversityMuseumincooperationwithcuratorandarchitectEgillReimers,whohadaprojectofhisown.TerjeThunofNTNU
datedthesamplesandre-examinedsamplesthatweretakenduringtheBryggenexcavationsintheyears
between1955and1979inadditiontothenewsamples.This‘dendroproject’wasfinancedthrougha
grantfromtheFacultyofHumanities,UniversityofBergen.
7
TerjeThunatNTNUhaskindlyre-examined‘crucial’samplesin1999,2001andin2004.
8
OxCalversion3.5CBronkRamsey(2000);(http://www.rlaha.ox.ac.uk/orau/index.htm).
9
T-5190,T5682,andT5473(Hjelle1986,36).
10
Centaureacyanus(cornflower),Papaverrhoeas(poppy),Papaverdubium(poppy),Papaverargemone
(poppy),Helianthemumnummularium(rockrose),Cytisustype(broom),andUlex(gorse).
11
AdetailedaccountforhowtheH-postdatabasewasupdatedisgiveninHansen1998,109-111.
12
T-3786,970+/-40BP(KrzywinskiandKaland1984,24).
13
Thisisoriginallyprofile220/1-3intheBryggendocumentationmaterial.Cfalsodiariesofexcavationunit
N7/1972atsite6,BryggensiteBRM0(TopArk).
14
Dendrono1454BRM0(Hansen1998,94).
15
DendronosBRM0/93029andBRM0/93028(Hansen1998,93).Sincetheyhaveturnedouttobe
centralforthedatingoftheearliesthorizonstheywerere-examinedbyTerjeThun.
268
16
Rowsofpairsofpostsplacedatintervalswithverticalplanksstandingsidebysideinbetween.Bothposts
andplankswereeitherrammedordugintotheground.(SeealsoFigure13).
17
AccordingtoboundaryindicatorsaccountedforinChapter9,thefencesdemarcateboundariesoftwo
plotslabelled6/Cand6/B.
18
‘...thebottomoftheenclosingwallbelowthefloor-levelofabuildingraisedonposts’,(Herteig1991,97).
Thisdescriptionisinaccordancewithabuildingtypethatislaterdenominated‘cellarbuildings’(Herteig
1992,287).
19
Q3planXIIthefenceandpostsmrk5and75.
20
Q3planXmrks1,4,5,6,18,19,20,21and34.
21
BRM0/92710.
22
Dendrono1537BRM0,Thun’scatrasno1344.Detailslikesignsofreuseormissingtreeringswerenot
documentedwhenthissamplewastakenintheearly1970s.
23
DendronosBRM0/92786,0/92785and0/92798.Thefirsttwosamplesweretakenin1997/98,from
postswhereonlythebottompartwaspreserved,andtheissueofreusecouldnotbejudged.Thetwoposts
produceddatesthatweremucholderthanexpectedforbuilding66,beingassignedtoperiod2,athird
samplewastakenin1999.Thissamplegaveayoungerdatethatcorrespondswellwithperiod2.
24
T-9162.
25
ExcavationsupervisorAndrzejGolembnikhaskindlyplacedhisunpublishedmanuscriptforhisreportat
mydisposal.
26
Golembniksuggeststhatthefirewhichendedphase10/9wasidenticalwith‘thefirebeforethehistorically
documentedonein1170/71’(Golembnikinprep-a,8).Bythishemostlikelymeansthatthefireis
identicalwithfireVIIIfoundatsite6,formerlydatedtoc1150(Herteig1991).Today,however,thefire
isdatedtothe1120s(Hansen1998).FireVIIIdestroyedbuilding45attheBryggensite,however,the
fireseemstohavebeenquitelocal,sinceithashardlybeenrecognisedoutsidethisbuilding(butseethe
discussionofthe9-postbuildingatsite6above).Consequentlywecannotassumethatthefire,which
destroyedphases10/9atDreggsalmenningen12-14BRM237,wasidenticalwiththeBryggenFireVIII.
27
Larsendidnotgive‘building14’anumber.Inthereportbuilding14isspokenofas‘apossiblebuilding
underbuilding5’(Larsen1967a).Thescatteredstructuresnorthofthefencearenotdiscussedinthe
report.
28
Thefenceitself,beingdugintothegroundmayhardlyhavesurvivedforsome150years,butthefunction
ofthefenceasademarcatorsurvivedandtheboundarysymbolisedbythefencewasnottrespassedby
structurestillafterhorizon5.
29
SeeChapter9foradefinitionofplotboundaries.
30
ClassificationaccordingtoFlodin1989.
31
UsingtheterminologyofClarkeandCarter1977todistinguishthedifferenttypesofwares(Clarkeand
Carter1977).
32
Asherdfromphase8wasclassifiedasof‘unknownorigin’.Afteracloserlookitturnedoutthatthesherd
isapieceofaplasticornamentfromahighlydecoratedvesselofScarboroughIIware.Thiskindofpottery
isnormallynotfounduntilthethirteenthcentury(FarmerandFarmer1982).Ifthesherdisfoundinits
rightcontextDunlop’sproposed1170-1198dateforphase8is25-30yearstooearlyandmyproposed
1150/60-c1170dateis55-60yearstooearly.Inotherwordsthesherdfitsbadlywiththegeneralpictureof
thematerialbybeingtooyoung.ConsequentlyIthinkthatthesherdwasintrusiveanditisnotincluded
inthefurtherdiscussion.
33
K21,K35,K9,K33.
34
InastudyofburialsinmedievalOsloatimespanof20yearsbetweeneachlevelofburialswassuggested
(Eide1974,227-230).Thiswoulddatethefirstlevelofburialsinthepresentcaseto60yearsbeforethe
1120s,thatisthe1060s.
269
35
Thesamplesfrom1998weretakenbyReimersandmyself.
36
BRM104/2486(VIIIa)Catrasno11002561.Theoutertree-ringinthesamplewasdatedto1090.
AccordingtoTerjeThun,whoanalysedthesample(2001),afewtree-ringsmayhavebeenmissing.Asthe
samplewastakenfromaplaceonthetimberwherethesurfaceseemedintact,andlackednotree-rings,this
canonlybeafewyears.
37
BRM110/6096,thesamplewastakenbyReimersandmyselfin1998.
38
‘Stages’istheterminologyusedbyGolembnik.
39
‘Stages’isLindh’sterminology.Thedocumentationofstonelayer‘A’inthesitereportisanexceptiontothe
rule:althoughitisnotafire-layer,ithasbeendescribedandgivenanumber(‘A’).
40
BRM490/27
41
T-10346,fromlayer190,BRM342/97.
42
SherdBRM342/951fromlayer620inPitK192.
43
Abasinusedinconnectionwithfreshwatersupply;sedimentsintherunningwaterfromthenearbystream
wereseparatedfromthewaterwhenlettingthewater-streamslowdowninabasin.
44
‘Mørkbrungytjemedlitemakroskopiskmateriale’.(Hjelle1998,section5).
45
Helianthemum(rockrose),Papaverrhoeas(poppy)andCentureacyanus(cornflower).
46
Thelocationoftheunits/plotscanbeseenonmapsinChapter9and10egFigure33andFigure39.
47
BuildingrownumbersandlettersrefertoHerteigsmainpublicationofthebuildingsatBryggen(Herteig
1990and1991).
48
Theadditionalboundariesinthemiddletownareacannotaddanythingtothediscussion:Thedistance
betweenthesouthernboundaryofplot26-27/Bandthenorthernboundaryofplot29/Bisabout34m,
measuredatarightanglebetweentheboundaries.Ifthisdistanceweredividedintothreeplotstheseplots
wouldbeabout11.3mwideandthusconformtothesystemofthenortherntownarea.If,however,the
distancewasdividedintotwoplotsonly,theseplotswouldbeabout17mconformingtotheaverageof
theknownplotsofthemiddletownarea,inconclusion,thismaterialdoesnotaddtothediscussion.The
distancebetweenthenorthernboundaryofplot29/Bandthepier/waterfrontconstructionsatsite30,
Vetrlidsalmenningenisabout25m.Thisdistanceisjustlargeenoughfortwoplotsofthenortherntown
areasizetobesqueezedin.Butsomeofthesouthernmosthypotheticalplotswouldthenbelocatedpartly
inthesmallriverthatrandownthemorainicslopehere...Aplotofabout17mwidthontheotherhand
wouldfitquiteneatlybetweenthenorthernboundaryofplot29/Bandthepier/waterfrontconstructions
atsite30,Vetrlidsalmenningenandleaveroomforthesmallriver.Itseemsthatthismaterialcanbe
interpretedeitherwaysoitdoesnotaddanythingtothediscussion.
49
Thereadermayrecollectthatthestreetwasassignedtohorizon4asasupplementarysourcethrougha
horizontallinkofthestreettothepresumedpredecessortoStMary’s(cfp115ff,126ff ).
50
Eachplotorsite/analyticunitcountsasoneunit,whereitcannotbeascertainedwhetherdataarederived
fromoneortwoplots(egdatafrom‘unit7’atsite6maystemfromplot6/Eor6/Forboth)suchdataare
treatedasrepresentingoneplot/analyticunit.Inhorizon2datafromsiteslocatedoutsidethehorizon2
plot-systemandthesettlementatsite30,Vetrlidsalmenningen,arecountedasoneunitpersite.
51
Kellmernevercompletedorpublishedherstudies.Shedidhoweverleavebehindnotes,withvaluable
observations.Ihavebeenfortunatetohavetheseatmydisposal.
52
RivetsinthecombsfromcontemporarycontextsinBergenoftenappear‘hollow’andmust,assuggestedby
PatriciaGalloway,havebeenformedfromrolledsheetsofbronzeorothermetalsratherthanfrompulled
wire(CfBiddle1990,266,footnote20).
53
BRM0/54784/01.
54
BRM0/54784/1.
55
BRM0/64456.Thepossibletouchstonewasanalysedbybothmicrosondeandscanningelectron
270
microscopebyHaraldFurnes,DepartmentofEarthScience,UniversityofBergen.Furnesconcludedthat
therawmaterialforthepossibletouchstoneis‘alavaorashallowintrusion’(Furnes2001).
56
Asystemforrecordinglayerscontainingsteatiteoffcut/chipshasbeeninuseasaroutineatall
investigationsinBergensince1955.
57
Hufthammerhaskindlyprovidedtheinformationaboutcutandchopmarksontheskullsofcatsanddogs
fromtheEngelgårdenareasite6,Bryggen(aboutthesameareaasplot6/E).
58
Atsitesexcavatedbetween1955and1979layerswereinprinciplecharacterisedaccordingtothemost
dominantfeatureofthelayer,manylayerswere,however,notcharacterisedatall.
59
AccordingtoAnneÅgotnes’observationsofsignsofusageonbakingslabsfromsite6,Bryggen,theslabs
musthavebeenusednotonlyforbakingflat-bread.Othersmallertypesofbreadand/orotherfoodstuffs
wereprobablyalsobakedorheatedonthestoneslabs.
60
InmaterialfrommedievalRibe(DK)pinsinitiallyclassifiedassausagepinswerethroughtheircontext
re-classifiedasskewere/strechersusedwhenstrechingskins.TheRibepinswerefoundsittingintheground
encirklingpatchesofdarksoil(Bencard1973).Inthepresentmaterialthe‘sausagepins’havenotbeen
foundinsuchcontexts.
61
Fourlong-toothedcombsandtwoflax-combsarepresentinthematerialbutnotincludedhere.
62
IhavedividedFlodin’stypeE5-3intothreetypes:(1)onerowofrivets,noprofile,(2)onerowofrivets,
oneprofile,(3)withtworowsofrivets.
63
ThecombsfromLund,ViborgandOslostemfromolderexcavationsandarealldatedbroadlytoc11501225(Blomquist1942,142-148;Nielsen1969,61;Wiberg1977,207).ThecombfromSchleswigis
datedtothetwelfthcentury(Ulbricht1984,46)thecombfromLödösebroadlyto1100-1200(Letter
fromSoniaJeffery,LödöseMuseum30/10/02).ThecombfromSchleswigisofbone(Ulbricht1984,49)
OsteologistAnneKarinHufthammerofBergenMuseumhaskindlyclassifiedthe‘twin’combsandsome
ofthecombblanksfromBergen.Shepointsoutthatthesurfaceofthecombsisgenerallysoworkedup
thatapositiveclassificationoftherawmaterialisnotpossiblewithoutaDNAtest.Avisualclassification
ofthematerial,however,indicatesthatthecombsweremadeofantlerofreindeer(perscomHufthammer
2002).Antlerofreindeerandafewantlersofelkhasbeenfoundinmedievalculture-layers,whereas
nospecimenofdeerhavebeenfound(Hufthammer1987,69).Thissupportsthenotionthatantlerof
reindeerwasthepreferredrawmaterialforcombmakinginNorway(Trondheim,Oslo),whereastheraw
materialfoundincombwasteinsouthernScandinavia(Lund,Konghelle,Schleswig)isantlerofdeercf
(Rytter1997,10).Thecombsfromtheotherlocationshavenotbeenclassifiedaccordingtorawmaterials.
64
Ionlystudiedshoeswithelaborateembroiderypatternsasopposedtothosewithmoresimplepatterns,
becauseIwantedtohaveasmanydetailsaspossibleatmydisposalwhencomparingthepatterns.Itisthus
possiblethatthetwinsidentifiedbymecompriseanminimumoftwinsintheavailablematerial.
65
Iattemptedtomeasuretheregularityofthestitchesintheseamsoftheshoesaccordingtothemethods
describedbyKethELind(Lind1991,192ff ).HavingfollowedLind’sproceduresforawhilenoshoeswith
irregularseamshadturnedupandIgaveupthetimeconsumingproject.Withafewexceptions,thatwere
allresultsofrepairorsecondaryuseoftheshoe,theseamsoftheshoesinhorizons4and5appeartobe
veryregular.
66
TheinlaidmetalofthethreekeyswasstudiedbyKirstiHaugeRiisøen,BergenMuseumthrough‘X-ray
diffractionspectroscopy(XRD).TheinlaidmetalonkeyBRM0/72983wascopperalloy,whereasno
inlaidmetalscouldbetracedonkeysBRM0/44749andBRM104/2771.BRM104/2771has,however,
clearlyhadastringtwistedaroundit.AsforBRM0/44747thiskeyhasbeentreatedinsuchawayduring
conservation,thatitisnotpossibletostudytheoriginalsurface(Riisøen2001).
67
BRM0/45060,0/45222,0/45847*,0/54529*,0/63860*,0/64396,0/64557*,0/64558,0/65017,
0/73103,BRM76/10967,andBRM94/1066*,BRM104/2261*.Inordertogetacloserdetermination
oftherawmaterialandpossibleoriginoftherockthewhorlswereexaminedgeologically.Whorlsmarked
with*werescanningelektronmicroscopeanalysedbyHaraldFurnes,DepartmentofEarthScience,
UniversityofBergen.AccordingtoFurnes,‘Theanalysedspindlewhorlswereallcharacterisedbya
271
relativelyhighcontentofMgOandFeO,andalowcontentofSiO2.Thiscompositionindicatesarich
olivinecomposition.InadditiontheA12O3contentishigh.Thisshowsthatthematerialcontainsone
ormoreA12O3-richcomponents;thesemaybeCa-richplagioclaseand/orA1-richspinell.Therock
isthereforemostlikelyanultramaficrockthattosomedegreehasbeentransformedintoserpentine.
Noneofthesamplesthusrepresentgreenstone’(Furnes2001).Thereareslightvariationsinthecolour
ofthethirteenspindlewhorls.IaskedgeologistØysteinJJansen,BergenMuseum,whoisaspecialistof
serpentine/steatitequarries,tojudgewhetherornotthe13whorlsinspiteofthecolourvariationsmay
stemfromthesamegeologicalsite/quarry.Jansenstudiedthewhorlsthroughmagnifyingglass.Based
onhisinvestigationandtheresultsfromFurnes’scanningmicroscopeanalysis,Jansenconcludedthat
thedarkestwhorlsarelesstransformed,whilethelightercolouredwhorlsaremadeofrockthatismore
transformedtowardsserpentineandtalk.Stillallthewhorlsmaystemfromonegeologicalsite/quarry,
sincevariationsinthedegreeoftransformationoftherockmaybefoundwithinafewm3onageological
site(perscomJansenJanuary2003).
68
0/65009.TherewasnoproductionofpotteryinmedievalNorway.
69
0/65009.
70
BRM0/46136,0/46161,bothdrawn,0/76420describedonlyintheoriginalfindlists.
71
BRM0/45525.
72
BRM0/55139,BRM110/5682.
73
BRM0/43752,0/54277,0/79851,and0/82145.
74
BRM0/63827.
75
BRM0/44989,0/45092,0/45542,0/46275,0/53081and0/73063.
76
Theactivitiesofwarandgamesarenotconsideredrelevantforthisdiscussion.
77
BeingnoexpertonboatsmyselfIowemyconfidenceinthisobservationtoahelpfuldiscussionwithJan
Bill,TheNationalMuseumCentreforMaritimeArchaeology,Denmark,(January2002)regardingthe
question.
78
BRM0/85675:‘Øyolvownsthissack’.Thefindscontextisdatedtobetweenc1170andc1198.84690:
‘Endreownsthissack’.Thefindscontextisdatedtobetweenc1198and1248.
79
Thereadermayalsorecollectthat,asamethodologicalapproach,presenceratherthantheabsenceof
activitiesisgenerallyemphasisedinthepresentstudy(cfp71ff ).
80
HelgeAskvikattheDepartmentofEarthScience,UniversityofBergenhasclassifiedthehonesaccording
toprinciplesoutlinedinMitchell,Askvik,andResi1984.
81
EastMidlands,Hedon,Humber,Grimston,London,Scarborough,Stamford,andTorkseywares.
82
Pingsdorf,andPaffrathwares.
83
NormandyGritty,unspecific‘Frenchtype’,and‘northFrench’wares.
84
Andenneware.
85
Heliantemunnumularium(rockrose),Centaureacyamus(Cornflower)andMalva(Mallow).
86
DataforRuth’sstudyisbasedonchildrenfromtheUnitedStatesandGreatBritain.
87
BasedonfiguresinBennicke1993,37,medievalmenandwomenwererespectively3.9%and4.8%
shorterthantheirmoderncounterparts,henceatanaverage,medievalpeoplewereabout4.3%shorter
thanthoseoftoday.
88
Themoderncontinentalshoesizeisobtainedbymultiplyingthelengthoftheshoeinmmby1.5
(Groenman-vanWaateringe1978,185).Larsen(1970,1992)andSchia(1975)used24cm,size36,
asthedividebetweenadultandchildshoesintheirstudies.Theymeasuredunpreservedleather.The
majorpartofshoesfromBergenhavetodaybeenpreservedthroughvariousmethods,thishascaused
shrinkagetotheleathersothatafterpreservationtheshoesareshorterthanwhennewlyexcavated.Soles
measuredanddrawnbeforepreservationwerehence15-25mmlongerthanwhenImeasuredthemafter
272
preservation.InOsloobservationsonleatherbeforeandafterpreservationshowthesametendency–after
preservationsoleshadshrunk25-30mmdependingontheoriginallengthofthesole(Schia1977,123).
Mymeasurementsaretakenfromamixtureofleatherpreservedthroughvariousmethodsandfrom
unpreservedleatherthathasdriedup,inadditiontothismostoftheshoeshadtobesoakedinlederweicher
beforemeasurementscouldbetaken.Whensoakedinlederweichertheleatherswellsandregainssome
ofitspre-preservationsize.Duetothevarioustreatmentsoftheleatherfromthepointofexcavation
untilmeasurementsweretakenIhavenotbeenabletoestablishthepreciserelationshipbetweenthe
postexcavationsizeoftheshoeandthepresentdaysize.Thispresentssomeproblemswheninterpreting
thematerial.Inordertomakeupforatleastoneoftheseproblems20mmisaddedtomeasurements
takenfrompreservedleatherthatwasnotsoakedinlederweicherbeforemeasurementsweretaken,this
shouldprovideacoarsecompensationfortheshrinkagecausedbypreservation.Themeasurements
presentedherearethusmoreorlessequivalenttomeasurementstakenofshoesthatwerenotpreserved.
Yetanotherproblematicquestionishowthesizeoftheunpreservedshoecorrespondstothemedieval
shoe?Thisquestionhasnotyetbeenansweredanditisbeyondthescopeofthepresentstudytodoso.
Schiasuggestedthatthemedievalsizeoftheshoemaybefoundbetweenthenewlyexcavated-shoeand
thepreserved-shoesize(Schia1977,123),hisstudyofthesolesfromOslowereneverthelessbasedon
unpreservedleather,alsoLarsen’sstudieswerebasedonmeasurementsofunpreservedleather.Ialsohaveto
interpretthemeasurementsavailable,bearinginmindtheuncertaintiesinvolved.
89
Fritzvold(1976)alsoreferstothedrawingsfromTrumpy’sproject260.Howeverheisnotdetailedinhis
reconstructionofthethresholdbetweenVeisanandVågenandIfounditnecessarytocheckTrumpy’s
observationsmyself.Itproveddifficult,however,tofindthedrawingssinceTrumpy’smaterialisinprivate
hands.Halfofdrawing9,containinginformationonbedrocklevelsforthemainbuildingatBradbenken
1wasaccessible(andthereforealsoonlyinformationonhalfofthebuildingsite!),withthekindhelpof
InstanesA/S,Bergen.
90
AccordingtoFritzner(1973)theplacenamehastwomeanings:1)Holmr:OmflydtLandafringe
Omfang,lidenØiAa,VandellerSø.2)Holmi=Holmr:Tue,Forhøiningsomhæversigopafogoverden
omgivendeFlade(Land,Ager,Myr).
91
Supplementedwithinformationonthe‘A5indexcards’,theA5indexcardsaretheoriginaldocumentation
forthelocationofartefactsonsite9.
92
Supplementedwithinformationonthe‘A5indexcards’,theA5indexcardsaretheoriginaldocumentation
forthelocationofartefactsonsite11.
93
Ihaveclassifiedthesecombsanddatedthemaccordingtomyupdated‘H-postdatabase’(cfp82)and
Hansen1998.
273
LISTOFFIGURES
Figure1.BergenonthewestcoastofNorway.TheBergenarea .................................18
Figure2.Fourteenmedievaltownsrelatedtotheperiodbefore1200inthedocumentaryrecords.
(ModifiedfromHelle1992,8) ..................................................22
Figure3.Koren-Wiberg’sreconstructionofthesettlementinBergenbeforeOlavKyrre.(Koren-Wiberg,
1921,48PlanIII).............................................................25
Figure4.Lorentzen’sreconstructionofBergenc1200.(Lorentzen1952,75).......................26
Figure5.Thedoublenucleussituation.(Myrvoll1993,87) ....................................28
Figure6.Cross-fitartefactsatsite6,Bryggen;site26,Finnegården6aandsite27,Finnegården3a.(The
informationoncross-fitartefactsstemsfromLüdtke1989,15;BlackmoreandVince1994,73,8,
andfromtheoriginaldocumentationfromthethreesites) .............................49
Figure7.Site6,Bryggen.Thedifferentstagesintheexcavationofthesiteandnamesofthetenements.
(AfterHerteig1990,10andHerteig1991,12) ......................................59
Figure8.Thestratigraphyofprofile220atsite6,Bryggen.(AfterKrzywinskiandKaland1984Figure3)60
Figure9.Thedendrodatedpostsinthejettyatsite6andsimilarpostsatBorgund,Sunnmøre.(Borgund
1961Æ7,48,50X/166,70Yand45,4X/166,4Y;BRM0,BryggenPlanO6XIV,Bilag1) ......61
Figure10.PlanO03XandO03XI,site6,Bryggen..........................................66
Figure11.14Csamplefromlayer17/31site7,ØvreDreggsalmenningen...........................69
Figure12.14Csamplefromphase9/10site8,Dreggsalmenningen14-16..........................69
Figure13.Thepalisade-builtfenceatsite9,Sandbrugaten5.(Negative67and70,photoArneJLarsen).73
Figure14.ThestratigraphicalrelationshipbetweenartefactassemblagesdescribedinAppendix3and
majorstructuresandlayersatsite9,Sandbrugaten5..................................73
Figure15.Apalisade-builtfenceatsite11,Dreggsalmenningen20.(Negative154,photoArneJLarsen).78
Figure16.ThestratigraphicalrelationshipbetweenartefactassemblagesdescribedinAppendix4andthe
oldeststructuresandlayersatsite11,Dreggsalmenningen20...........................79
Figure17.14Cdatefromlayer147inphase8site20,Øvregaten39 ..............................85
Figure18.14Cdatefromlayer24inphase10Øvregaten39BRM94.............................85
Figure19.14Cdatefromlayer65intheoldest‘phase’atsite21,Klingesmauet......................88
Figure20.Theverticalandhorizontalrelationshipsbetweencentralstructuresassignedtohorizons2-5.107
Figure21.Generallegendformaps...................................................... 112
Figure22.Investigatedsitesandmonuments............................................... 113
Figure23.Structuresandlayersassignedtohorizon1(c800-c1020/30)......................... 114
Figure24a.Structuresandlayersassignedtohorizon2(c1020/30-c1070),thenortherntownarea ... 115
Figure24b.Structuresandlayersassignedtohorizon2(c1020/30-c1070),themiddletownarea..... 116
Figure25a.Structuresandlayersassignedtohorizon3(c1070-c1100),Holmen.................. 117
Figure25b.Structuresandlayersassignedtohorizon3(c1070-c1100),thenortherntownarea...... 118
Figure25c.Structuresandlayersassignedtohorizon3(c1070-c1100),themiddletownarea........ 119
Figure26a.Structuresandlayersassignedtohorizon4(c1100-c1120s).........................120
274
Figure26b.Structuresandlayersassignedtohorizon4(c1100-c1120s),thenortherntownarea .....121
Figure26c.Structuresandlayersassignedtohorizon4(c1100-c1120s),themiddletownarea.......122
Figure27a.Structuresandlayersassignedtohorizon5(1120s-c1170) ..........................123
Figure27b.Structuresandlayersassignedtohorizon5(c1120s-c1170),thenortherntownarea......124
Figure27c.Structuresandlayersassignedtohorizon5(c1120s-c1170),themiddletownarea .......125
Figure27d.Structuresandlayersassignedtohorizon5(c1120s-c1170),thesoutherntownarea......126
Figure28.Boundariesidentifiedinsourcesassignedtohorizon2(c1020/30-c1070)...............135
Figure29.Boundariesidentifiedinsourcesassignedtohorizon3(c1070-c1100)..................135
Figure30.Boundariesidentifiedinsourcesassignedtohorizon4(c1100-1120s)...................137
Figure31.Boundariesandbuildingrowsatsite6,Bryggen.(ModifiedfromHerteig1991,Plate14and
1990,Figure85,Figure56)....................................................137
Figure32a.Boundariesidentifiedinsourcesassignedtohorizon5(1120s-c1170)..................139
Figure32b.Boundariesidentifiedinsourcesassignedtohorizon5(1120s-c1170)..................139
Figure33.Theareacoveredbythehorizon3plotsystem.....................................144
Figure34.Site8,Dreggsalmenningen14-16.Theorientationofstructuresyoungerthanhorizon3 ....144
Figure35.Theareacoveredbythehorizon2plotsystem.....................................146
Figure36.Horizon2(c1020/30-c1070)occupiedandvacantanalyticunits .....................148
Figure37.Horizon3(c1070-c1100)occupiedandvacantanalyticunits......................... 151
Figure38.Horizon4(c1100-1120s)occupiedandvacantanalyticunits......................... 153
Figure39.Horizon5(1120s-c1170)occupiedandvacantanalyticunits ......................... 157
Figure40.Toothsegmentsandconnectionplates...........................................163
Figure41.Punchandwiredrawer:aBRM0/86590/02;bBRM0/86590/01.....................163
Figure42.Leatherwasteandashoemaker’slast(BRM0/54784/01).(FromLarsen1991,34-35).
(DrawingsbySveinSkauge)....................................................166
Figure43.Crucibles,apossibletouchstone,andamould:a,c,ecruciblesBRM104/2280,
BRM104/2311,BRM104/2326;bBRM0/64456touchstone?;dBRM110/4949mould ..168
Figure44.TwincombsfromBergen.TypeE5-3onerowofrivets,oneprofile:aBRM0/77536plot6/B,
bBRM104/2383plot26/A;typeE5-3tworowsofrivets:cBRM0/43711plot6/D,dBRM
0/64328plot6/C,eBRM110/5483plot26-27/BC;typeE-1:fBRM76/11106plot28/C,g
BRM76/9807plot28/B,hBRM110/4605plot27/C;typeE5-3onerowofrivets,noprofile:
iBRM0/45464plot6/D,jBRM0/72946plot6/C.................................184
Figure45.TwincombsoftypeE5-3,withtworowsofrivets,fromBergen,Lund,Viborg,Schleswigand
Lödöse.Bergen:aBRM76/12652plot28/B,bBRM104/1987plot26-27/B,cBRM104/2276
plot26/A,dBRM104/2369/01plot26/A,eBRM110/5483plot26-27/BC;Lund:fL.U.H.M.
15310282:A,gK.M.22802aVIII(Blomquist1943,144-145);........................185
Figure46.TwincombsoftypeE5-2fromBergenandLund.Bergen:aBRM104/2275plot26/A;Lund:
bK.M.8480Annegatan(Blomquist1943,144)....................................186
Figure47.Ornamentedgamingpiecesandneedles/pins:a,b,cgamingpieces;dBRM0/53003;eBRM
0/81009...................................................................188
Figure48.TwinshoesfromBergen......................................................189
Figure49.EmbroiderypatternsC5andG2fromTrondheimandOslo.(PublishedinSchia1977,
Figure44;Schia1987,Figure22;Marstein1989;Smedstad1991,Figure32) .............190
Figure50.Examplesofcrudeandfine‘otherleatherwork’:aBRM0/85396/01,child’sshoemadefrom
agrownup’sshoewithembroidery;bBRM0/45983/01,knife-sheathmadefromashoewith
embroidery.................................................................192
Figure51.TwinkeysfromBergenandTrondheim.Bergen:aBRM0/72983;Trondheim:bN10579/S139
(Christophersen1987,Photop87) ..............................................194
Figure52.Turnedspindlewhorls.BRM0/45060,BRM0/45222,BRM0/45847,BRM0/54529,BRM
0/63860,BRM0/64396,BRM0/64557,BRM0/64558,BRM0/65017,BRM0/73103,BRM
275
76/10967,andBRM94/1066,BRM104/2261.....................................197
Figure53.Ornamenteditemsinwood ...................................................199
Figure53b.Ornamenteditemsinwood.................................................. 200
Figure53c.Ornamenteditemsinwood..................................................201
Figure54.Horizon5(1120s-c1170),sausagepins,basiccookingtoolsandproductionwastefrom
ambulatingartisans:combmakers,shoemakersandmetalworkersasapercentageofthetotal
numberoffindsfromartefact-yieldinganalyticunits ................................ 204
Figure55a.Horizons2and3,harbourconditions,thenortherntownarea .......................209
Figure55b.Horizons2and3,harbourconditions,themiddletownarea ........................ 210
Figure56a.Horizon4(c1100-1120s),harbourconditions,thenortherntownaera ................213
Figure56b.Horizon4(c1100-1120s),harbourconditions,themiddletownaera..................213
Table66.Horizon4(c1100-1120s),trade-indicatingsources.................................. 215
Figure57a.Horizon5(1120s-c1170),harbourconditions,thenortherntownaera................. 216
Figure57b.Horizon5(1120s-c1170),harbourconditions,themiddletownaera.................. 216
Figure57c.Horizon5(1120s-c1170),harbourconditions,thesoutherntownaera................. 217
Figure58.Site27,Finnegården3a.Waterdepthbythequayfront.
(ModifiedfromGolembnik1993,Figure5) ....................................... 218
Figure59.Selectlistofpoliticalevents,1024-42.Basedonwrittenrecords.
(ModifiedfromRumble1994,Table1.1) .........................................226
Figure60.Areasincludedinthetownscapefromhorizon2throughhorizon5 ....................231
Figure61.DetailoftheareabetweenBontelaboandVeisan...................................252
Figure62.Sourcesforthepre-urbantopography‘thenaturaltopography’.........................253
Figure62a.Sourcesforthepre-urbantopography‘thenaturaltopography’........................254
Figure62b.Sourcesforthepre-urbantopography‘thenaturaltopography’. ......................255
Figure62c.Sourcesforthepre-urbantopography‘thenaturaltopography’........................256
Figure62d.Sourcesforthepre-urbantopography‘thenaturaltopography’. ......................257
Figure62e.Sourcesforthepre-urbantopography‘thenaturaltopography’........................258
Figure62f.Sourcesforthepre-urbantopography‘thenaturaltopography’........................259
’after1144’-sometreeringsmaybemissing,thefellingyearforthetreecannotbeestablished.’1128/1129’
-thepreservedoutertreeringrepresentsthelastyearofgrowth.’j’-yes,’n’-no.’surfacework?
-surfaceworkonthespotwherethesamp......................................... 260
276
LISTOFTABLES
Table1.Site1,Koengen(1986)..........................................................57
Table2.Site6,Bryggen(1955-1979)BRM0 ...............................................68
Table3.Site7,ØvreDreggsalmenningen(1989)BRM298....................................70
Table4.Site8,Dreggsalmenningen14-16(1986/90)BRM237.................................71
Table5.Site9,Sandbrugaten5(1967)BRM3..............................................76
Table6.Site10,Sandbrugaten3(1953) ...................................................77
Table7.Site11,Dreggsalmenningen20(1969)BRM4.......................................80
Table8.Site12,Dreggsalmenningen10-12(1972)BRM42 ...................................81
Table9.Site14,Dreggsalmenningen(1979)BRM83.........................................82
Table10.Site15,Stallen,Svensgården(1981)BRM90 .......................................83
Table11.Site20,Øvregaten39(1981)BRM94.............................................87
Table12.Site21,Klingesmauet(1990)BRM299............................................88
Table13.Site22,Kroken3(1985)BRM223 ..............................................89
Table14.Site26,Finnegården6a(1981)BRM104 ..........................................92
Table15.Site27,Finnegården3a(1982)BRM110 ..........................................93
Table16.Site28,Rosenkrantsgaten4(1978/79)BRM76 .....................................95
Table17.Site29,Vetrlidsalmenningen2,Kjøttbasaren(1996and1997)BRM490(NIKUprojekt22321) 95
Table18.Site30,Vetrlidsalmenningen(1991/92)BRM342 ...................................98
Table19.Site37,NedreKorskirkealmenning/Vågsalmenningt(1998)BRM544 ..................101
Table20.Site38,Domkirkegaten6(1987)BRM245 .......................................102
Table21.Sitenumber,streetaddress/monument,museumnumber .............................104
Table22.Thetemporaldistributionofsourcesforhorizons1-5................................106
Table23.Thenumberofartefactsassignedtohorizons2-5(N=9798)...........................109
Table24.Relevantbotanicalsources.....................................................109
Table25.Horizon2(c1020/30-c1070)approximatenumberofartefactsperexcavatedm2atthe
artefact-yieldingunit(N=1).................................................... 110
Table26.Horizon3(c1070-c1100),approximatenumberofartefactsperexcavatedm2atthe
artefact-yieldingplot/unit(N=21)............................................... 110
Table27.Horizon4(c1100-1120s),approximatenumberofartefactsperexcavatedm2atthe
7artefact-yieldingplots/units(N=252)........................................... 111
Table28.Horizon5(1120s-c1170),approximatenumberofartefactsperexcavatedm2atthe
24artefact-yieldingplots/units(N=9100)......................................... 111
Table29.Plotswherethelengthorwidthcanbemeasured,horizons2-5.........................142
Table30.Producertypesandhowtheymaybereflectedinthesources........................... 161
Table31.Horizon4(c1100-1120s),miscellaneousantler,bone,hornandwhale/walrusbone
working(N=25).............................................................162
277
Table32.Horizon5(1120s-c1170),combmakingandmiscellaneousantler,bone,hornand
whale/walrusboneworking(N=254) ............................................164
Table33.Horizon3(c1070-c1100),shoemakingand‘otherleatherworking’(N=3)................ 165
Table34.Horizon4(c1100-1120s),shoemakingand‘otherleatherworking’(N=47)...............166
Table35.Horizon5(1120s-c1170),shoemakingand‘otherleatherworking’(N=643) ..............167
Table36.Horizon4(c1100-1120s),metalworking(N=8)....................................169
Table37.Horizon5(1120s-c1170),metalworking(N=42)...................................170
Table38.Horizon2(c1020/30-c1070),stoneworking(N=1).................................171
Table39.Horizon3(c1070-c1100),stoneworking(N=1).................................... 171
Table40.Horizon4(c1100-1120s),stoneworking(N=5) ....................................171
Table41.Horizon5(1120s-c1170),stoneworking(N=33)....................................172
Table42.Horizon3(c1070-c1100),woodworking(N=1) ...................................173
Table43.Horizon4(c1100-1120s),woodworking(N=1) ....................................173
Table44.Horizon5(1120s-c1170),woodworking(N=10) ...................................173
Table45.Horizon5(1120s-c1170),tracesofskinning:skullsofcatsanddogs(N=12).............. 174
Table46.PossibletextiletoolsinHorizon3(c1070-c1100)(N=1)............................. 175
Table47.PossibletextiletoolsinHorizon4(c1100-1120s)(N=4).............................. 175
Table49.Horizon3(c1070-c1100),fishingtackleandpossiblefishingtackle(N=1)............... 176
Table50.Horizon4(c1100-1120s),fishingtackleandpossiblefishingtackle(N=3)................177
Table48.TextiletoolsandpossibletextiletoolsinHorizon5(1120s-c1170)(N=177) ..............177
Table51.Horizon5(1120s-c1170),fishingtackleandpossiblefishingtackle(N=128)..............178
Table52.Horizon5(1120s-c1170),weaponsofwar,huntingandgame(N=9)....................178
Table53.Horizon5(1120s-c1170),agriculturesources(N=2) ................................179
Table54.Horizon2(c1020/30-c1070),foodandbeverageprocessing(N=1) ....................180
Table55.Horizon3(c1070-c1100),basiccooking,foodandbeverageprocessing(N=4)............180
Table56.Horizon4(c1100-1120s),basiccooking,foodandbeverageprocessing(N=36) ........... 181
Table57.Horizon5(1120s-c1170),basiccooking,foodandbeverageprocessing(N=1265).......... 181
Table58.Productiveactivitiesdocumentedfromhorizon2throughhorizon5.....................182
Table59.Combsassignedtohorizons4andhorizon5andaccordingtocombtype(N=81)..........183
Table60.Horizons4and5productsofantler,bone,hornandwhale/walrusbone(N=45)...........187
Table61.Buildingswiththepresenceofmorethanoneproductiveactivitycarriedoutbyprofessional
ambulatingartisans,indicatedbyartefactsofcategoryI...............................203
Table62.ThenatureandorganisationofproductiveactivitiesindicatedinBergenbeforec1170....... 206
Table63.Trade-indicatingsources.......................................................209
Table64.Horizon2(c1020/30-c1070),trade-indicatingsources.............................. 211
Table65.Horizon3(c1070-c1100),trade-indicatingsources.................................212
Table67.Horizon5(1120s-c1170),toolsoftrade........................................... 217
Table68.Horizon5(1120s-c1170),thefunctionofbuildingscontainingtoolsoftradeor
interpretedasstoragerooms.................................................... 219
Table69.Horizon5(1120s-c1170),trade-indicatingsources.................................. 219
Table70.Horizon3(c1070-c1100),sourcesforthecharacterofthesettlements...................221
Table71.Horizon4(c1100-1120s),sourcesforthecharacterofthesettlements ...................221
Table72.Horizon5(1120s-c1170),sourcesforthecharacterofthesettlements ...................222
Table73.Potteryandcombsinassemblages1-7fromDreggsalmenningen20(1967)BRM4.........270
278
REFERENCES
Theoriginalsitedocumentationforexcavations Andersson,Hans1990.Sjuttiosexmedeltidsstäder
carried out from 1955 is found in the archives
-aspekterpåstadsarkeologiochmedeltida
(Topografiskarkiv,‘TopArk’)ofBergenUniverurbaniseringsprocessiSverigeoch
sityMuseum,TheMedievalCollectionslocated
Finland.Vol.73,Medeltidsstaden.
inBryggensMuseum.
Stockholm.
Andrén,Anders1980.Lund.Vol.26,
Medeltidsstaden.Stockholm.
Maps
Andrén,Anders1985.Denurbanascenen.
Generalkart 1879-80, 1:1000 i XXVIII Blade.
Städerochsamhälleidetmedeltida
Den Private Oppmaalings Lith: Anstalt. KrisDanmark.Vol.13,ActaArchaeologia
tiania.
Lundensia.Malmö.
Bergen 1913-30. Map of Bergen 1:1000 plate Andrén,Anders1989.StateandTownsin
NOXVserie1913-30.
theMiddleAges.TheScandinavian
Grunnkart Bergen, 1:500. Bergen kommune,
Experience.TheoryandSociety.Vol.18,
tekniskutbygging,oppmålingsseksjonen.Ajour585-609.
førtjuni1992.
Armstrong,Peter,DavidTomlinson,andD.H.
Evans1991.ExcavationsatLurkLane,
Beverly1979-82.SheffieldExcavation
Publishedandunpublishedtitles
Reports.Vol.1.
Ambrosiani,Björn,andHelenClarke1995
Bagge,Sverre2002.Mellomkildekritikk
(1991).TownsintheVikingAge.
oghistoriskantropologi.Olavden
London.
Hellige,aristokratietogrikssamlingen.
Ambrosiani,Kristina1981.VikingAgeCombs,
Historisktidsskrift.Vol81,173-212.
CombMakingandCombMakersinthe
Universitetsforlaget.
LightofFindsfromBirkaandRibe.Vol. Baug,Irene2002.KvernsteinsbrotaiHyllestad.
2,StockholmStudiesinArchaeology.
Arkeologiskepunktundersøkingar
Stockholm.
isteinbrotområdetiHyllestadi
Andersen,PerSveaas1977.SamlingenavNorge
SognogFjordane.Vol.22,Norsk
ogkristningenavlandet800-1130.
Bergverksmuseumsskriftserie.Kongsberg.
Ed.K.Mykland,A.Holmsen,A.
Bencard,Mogens1973.Problematiskpølsepind.
Kaartvedt,J.A.SeipandM.Skodvin.
Skalk.Vol.1,29-30.Højbjerg.
Vol.2,Handbokinorgeshistorie.
Bendixen,B.E.1896.Udgravninger
Bergen,Olso,Tromsø.
paaNikolaskirkenstomti
Andersson,Hans1977.Sverige.
Bergen.ForeningentilNorske
UrbaniseringsprocesseniNorden.
FortidsminnesmærkersBevaring.
Middelaldersteder.Ed.G.A.Blom.
Aarsberetningfor1895.Kristiania.38Oslo.Vol.1.
43.
279
Bennike,Pia1993.Menneskene.Daklingerimuld,25årsarkæologiiDanmark.Ed.S.Hvassand
B.Storgaard.Århus,34-39.
Berglund,Joel2001.OmkringdagliglivetpåGårdenunderSandet.TidsskriftetGrønland.Vol.7,
267-278.
Bergman,Kjell,andIngmarBillberg1976.Metallhantverk.UppgrävtförflutetförPKbankeniLund.
ArchaeologiaLundensia.Ed.A.W.Mårtensson.Lund.Vol.VII,199-212.
Bergquist,Ulla1989.Gjutningochsmide.Vol.16,Meddelelser.FortideniTrondheimbygrunn.
Trondheim.
Bertelsen,Reidar,andArneJ.Larsen1971.BRM15BugårdenN.Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Bertelsen,Reidar,andPrzemyslawUrbanczyk1988.TwoperspectivesonVåganinLofoten.Acta
Borealia.Vol.1-21988,98-110.
Biddle,Martin,Ed.1990.ObjectandEconomyinMedievalWinchester.Red.M.Biddle.Vol.7,
WinchesterStudies.Oxford.
Bjørgo,Narve1971a.BergensframvekstsettfråBryggen.BergensTidende,14/09/1971.
Bjørgo,Narve1971b.DeteldsteBergen.SjøfartshistoriskÅrbok1970.Bergen.53-130.
Bjørgo,Narve1971c.DeteldsteBergen.BergensTidende,17-18/08/1971.
Bjørndal,DanutaD.B.,andA.RoryDunlop1992.BRM345Torgalmenningenm/tillstøtende
gater.Archivereport,Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Bl1923.MagnusLagabøtersBylov,Ed.K.Robbestad.Kristiania.
Blackmore,Lyn,andAlanVince1994.MedievalPotteryfromSoutheastEnglandfoundinthe
BryggenExcavations1955-68.TheBryggenPapers.SupplementarySeries.Ed.A.E.Herteig.
Bergen.Vol.5,9-161.
Blindheim,Charlotte,BirgitHeyerdal-Larsen,andAnneStineIngstad1999.Kaupang-funnene
bindII.Ed.D.Skre.Vol.XIX,Norskeoldfunn.Oslo.
Blomquist,Ragnar1942.KammarfrånLundsmedeltid.Kulturen.133-162.
Brendalsmo,A.Jan1994.Tønsbergførår1000.Fragårdtilby.Vol.28,Varia.Oslo.
Brendalsmo,A.Jan2001.Kirkebyggogkirkebyggere.ByggherreriTrøndelagca.1000-1600.Dr.
art.avhandling,UniversitetetiTromsø,Tromsø.
Bryant,ChristopherG.A.,andDavidJary1991.Giddens’TheoryofStructuration.Acritical
appreciation.London.
Bugge,Lars1999.AnthonyGiddens’strukturellesosiologi.Hovedoppgave,Sosiologi,Universitetet
iOslo,Oslo.
Callmer,Johan1991.PlatsermedanknytningtilhandelochhantverkiyngreJärnålder.Eksempel
frånsödraSverige.FrastammetilstatiDanmark.2.Høvdingesamfundogkongemakt.Ed.P.
MortensenandB.M.Rasmussen.Højbjerg.Vol.22,29-47.
Carelli,Peter1999.ExchangeofCommoditiesinMedievalLund-PatternsofTradeor
Consumption?LübeckerKolloquiumzurstadarchaeologieimHanseraumII:DerHandel.Ed.
M.Gläser.Lübeck.469-492.
Carelli,Peter2001.Enkapitalistiskanda.Kulturelleforandringari1100-taletsDanmark.Vol.26,
LundStudiesinMedievalArchaeology.Stockholm.
Carelli,Peter,andPeterKresten1997.GiveUsThisDayOurDailyBread.AStudyofLateViking
AgeandMedievalQuernstonesinSouthScandinavia.ActaArchaeolgica.Vol.68,109-137.
Cassell,Philip1993.TheGiddensreader.Basingstoke.
Christensen,ArneEmil1985.BoatfindsfromBryggen.BryggenPapers,MainSeries.Ed.A.E.
Herteig.Vol.1,47-280.
Christensen,ArneEmil1986.Reinjegerogkammaker,enforhistoriskyrkeskombinasjon?Viking.
Vol.XLIX1985/86,113-133.
Christensson,Ann1980a.BRM144ØvregatenvedNikolaikirkealmenning.Archivereport,
280
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Christensson,Ann1980b.BRM252Øvregaten43.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Christensson,Ann1980c.IndberetningomudgravningernevedFinnegårdsgaten/Øvregaten.
Archivereport,Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Christensson,Ann1980d.Lungegårdsgaten2(smågravning43).Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Christensson,Ann1981.BRM146ØstreMuralmenningen.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Christensson,Ann1985.BRM226Walchendorfsgate5.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Christensson,E.AnnS.1988.BrandeogkronologiiBergen-belystvedtremindre
udgravningsfelter.Magistergradsavhandling,ArkeologiskInstitutt,UniversitetetiBergen,
Bergen.
Christensson,E.AnnS.,A.RoryDunlop,andHansGöthberg1982.BRM90Indberetningom
Svensgårdensstallbygning.Archivereport,Topografiskarkiv.Bergen,Hordaland.Bergen
Museum,Bergen.
Christophersen,Axel1980.Håndverketiforandring,studierihornogbenhåndverketsutviklingi
Lund1000-1350.Vol.13,ActaArchaeologicaLundensia.Lund.
Christophersen,Axel1982.Denurbanevaruproduktionensoppkomstogbetydningforden
tidigmiddelalderskebyutviklingen.Bebyggelseshistorisktidskrift.Denmedeltidastaden.Vol.
1982:3,104-122.
Christophersen,Axel1987.Trondheim-enbyimiddelalderen.Trondheim.
Christophersen,Axel1989.Kjøpe,selge,bytte,gi.VareutvekslingogbyoppkomstiNorgeca.8001100:Enmodel.Medeltidensfødelse.Ed.A.Andrén.Lund.109-145.
Christophersen,Axel1991.PortsandtradeinNorwayduringthetransitiontohistoricaltime.
AspectsofMaritimeScandinaviaAD200-1200.Ed.O.Crumlin-Pedersen.Roskilde.159-170.
Christophersen,Axel1994.Powerandimpotence:Politicalbackgroundofurbanisationin
Trøndelag900-1100.ArcheologiaPolona.Vol.32,95-108.
Christophersen,Axel1997.Ibrygge,bodogstrete.Havnoghandeli1000år.Karmøyseminaret.
Ed.KarmøyKommune.40-68.
Christophersen,Axel,WolfgangCramer,andMichaelJones1989.NaturlandskapetpåNidarnes
iYngreJernalder,enterrengmodell.Vol.21,Meddelelser.FortideniTrondheimbygrunn.
Trondheim.
Christophersen,Axel,andSæbjørgWalakerNordeide1994.KaupangenvedNidelva.Vol.7,
Riksantikvarensskrifter.
Clarke,H.,andA.Carter1977.ExcavationsinKingsLynn1963-1970.Vol.7,TheSocietyfor
MedievalArchaeologyMonographSeries.London.
Crumlin-Pedersen,Ole1985.CargoShipsofNorthernEuropeAD800-1300.Conferenceon
WaterfrontArchaeologyinNorthEuropeanTownsBergen1983.Ed.A.E.Herteig.Bergen.
83-93.
Crumlin-Pedersen,Ole1991.ShipTypesandSizesAD800-1400.AspectsofMaritimeScandinavia
AD200-1200.Ed.O.Crumlin-Pedersen.Roskilde.69-82.
Dahlbäck,Göran,ed.1983.Helgeandsholmen1000åriStockholmsström,Stockholmsmonografier.
Stockholm.
Dunlop,A.Rory1980.BRM140DetGamleRådhuset.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1981a.BRM138Koengen.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
281
Dunlop,A.Rory1981b.BRM139NedreHamburgersmauet5.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1982.BRM94Øvregaten39.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1982(1998).BRM104Finnegården6A,withsupplementsfrom1983and1998.
report,Riksantikvaren,Bergen.
Dunlop,A.Rory1983a.BRM153Nikolaikirkealmenning,østforØvregaten.Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1983b.BRM157Nordnes33.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984a.BRM90Stallen,Svensgården.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984b.BRM173Marken,Tverrgaten.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984c.BRM174Tverrgaten4-6.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984d.BRM175Badstustredet2.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984e.BRM180Marken3.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984f.BRM184Enhjørningen.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984g.BRM188Grønnevollen2.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984h.BRM190Nikolaismauet7.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984i.BRM193AlleHelgensgate7(Magistratsbygningen).Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984j.BRM200Korskirken.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984k.Kaigaten1c-5(smågravning22).Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984l.Lungegårdsgaten,Marken(smågravning42).Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1984m.StrømgatenopptilVestreStrømkaien(smågravning59).Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1985a.BRM202Nikolaikirkealmenning.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1985b.BRM221DetGamleRådhuset.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1986a.BRM236Strandgaten55-57.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1986b.BRM242Dreggsalmenningen.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1987.BRM223Kroken3.Archivereport,Topografiskarkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Dunlop,A.Rory1988a.BRM280Strandgaten80.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1988b.BRM282BekketomtenNordnes.Archivereport,Topografiskarkiv.
282
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1988c.BRM283Tollbualmenningenv/Tollboden.Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,ARory1989a.BRM287Bryggeparken.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1989b.BRM294SchøtstuenetilKlingesmauet.Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1989c.BRM295Nikolaikirkealmenning.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1989d.BRM297Wesenbergsmauet.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1989e.BRM298ØvreDreggsalmenning.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1989f.BRM299Klingesmauet.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1989g.BRM322Bergenhus-Bontelabo.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1990.BRM327LodinLeppsGate.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1991a.BRM331Forstandersmauet4.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1991b.BRM332Nordnesgatenv/nr47.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1991c.BRM333Nygaten2.Fielddocumentation,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1991d.BRM334Knøsesmauet.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1991e.BRM336Rådstuplassen.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1993.BRM462HalfdanKjerulfsgate.Fielddocumentation,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1996a.Arkeologiskeundersøkelser.HøringsutkasttilBergenhusFestnings
Verneplan.Vedlegg2.ForsvaretsKulturminneprosjekt.Forsvaretsbygningstjeneste.
Dunlop,A.Rory1996b.BRM488Øvregaten25-29.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory1998.AnarchaeologicalsurveyofBergen’smedievalfires.MedievalFiresin
Bergen-Revisited.BryggenPapersSupplementarySeries.Ed.I.Øye.Bergen.Vol.6,129-156.
Dunlop,A.Rory1999.ArkeologiskeundersøkelseriogomkringKjøttbasaren,Vetrlidsalmenning
2,Bergen,1996-97.Oppdragsmelding,NIKU,Bergen.
Dunlop,ARoryundated-a.Klosteret.Archivereport,Topografiskarkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Dunlop,A.Roryundated-b.KongOscarsgate67.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory,andAndrzejGolembnikinprep.TheFinnegårdenProject.Manuscript.
Dunlop,A.Rory,andGitteHansen1993.BRM464AlleHelgensgate3-5.Archivereport,
Topografiskarkiv.BergenHordaland.BergenMuseum,Bergen.
Dunlop,A.Rory,andGitteHansen1994a.BRM472OlavKyrresgate.Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory,andGitteHansen1994b.Klostergaten16.Archivereport,Topografiskarkiv.
283
Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory,andGitteHansen1994c.Nordnesparken2(Akvariet).Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Roryinprep.BRM342Vetrlidsalmenningen.Manuscript,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory,KariLoeHjelle,JochenKomber,andJonVidarSigurdsson1994.BRM245
Domkirkegaten6.Bergen.
Dunlop,A.Rory,andHanneDahlerupKoch1985.Manufakturhuset.Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Dunlop,A.Rory,andJonVidarSigurdsson1995.AnInterdisciplinaryInvestigationofBergen’s
Forgottenfire:ConfrontationandRenconciliation.NorwegianArchaeologicalReview.Vol.
28(2),73-92.
Dyvik,Helge1988.Addendarunicalatina.RecentlyFoundRunicInscriptionsinLatinfrom
Bryggen.TheBryggenPapersSupplementarySeries.Ed.A.E.Herteig.Bergen.Vol.2,1-9.
Edvardsen,Edvard1951(1630-95).BergenI.BergenHistoriskeForening.Skrifter.Ed.O.
Brattegard.Bergen.Vol.55/56.
Edvardsen,Edvard1952(1630-95).BergenII.BergenHistoriskeForening.Skrifter.Ed.O.
Brattegård.Bergen.Vol.57/58.
Egan,Geoff1998.TheMedievalHousehold.DailyLivingc.1150-c.1450.Vol.6,MedievalFinds
fromExcavationsinLondon.London.
Eide,OleEgill1974.DetoskipedekirkeriOslo.Magistergradsavhandling,ArkeologiskInstitutt,
UniversitetetiBergen,Bergen.
Eide,Oleegil1986.Hovedøyakloster-bidragtilbygningshistorien.Hikuin.Vol.12,73-78.
Ekroll,Øystein1981.BRM76Rosenkrantzgate4.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Ekroll,Øystein1997.Medkleberogkalk.Norsksteinbyggingimellomalderen.Gjøvik.
Emmelin,Lars1984.Visuellkonsekvensanalys-Enmetodförattbeskrivaochanalysera
förändringarikulturlandskapochsamhällsplanering.Kulturlandskapochsamhällsplanering.
Ed.L.Emmelin.213-242.
Enger,Cato1953.Hoteltomtenv/Sandbrugaten.Fishermaterialet,RiksantikvarensarkivOslo.
Enger,Cato1957.BRM48Slottsgaten3a/Sandbrugaten1(Sildesalslaget).Fielddocumentation,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Ersland,GeirAtle1988.JohanKristianKorenWibergbyhistorieogkulturminnevern.Bergen
HistoriskeForening.Skrifter.Vol.85/86,51-77.
Ersland,GeirAtle1989.EitforsøkpårekonstruksjonavgrunneigefordelingaiBergenved
utgangenavseinmellomalderen.Hovedoppgåve,Historie,UniversitetetiBergen,Bergen.
Ersland,GeirAtle1994.Kveneigdebyen?Dr.art.avhandling,HistoriskInstitutt,Universiteteti
Bergen,Bergen.
F:DenældreFrostathings-lov,NgLI,119-258.
Farmer,P.G.,andN.C.Farmer1982.ThedatingoftheScarboroughWarePotteryIndustry.
MedievalCeramics.BulletinoftheMedievalPotteryResearchGroup.Vol.6,66-88.
Ferveile,Claus1994.TheLatestNewsfromVikingAgeRibe:ArchaeologicalExcavations1993.
TheTwelfthVikingCongress.DevelopmentsAroundtheBalticandtheNorthSeaintheViking
Age.Ed.B.AmbrosianiandH.Clarke.BirkaStudies.Vol.3,91-99,Stockholm.
Fischer,GerhardUndated.KassettA2:Fishermaterialet‘Bergenhus’.Riksantikvarensarkiv,Oslo.
KopiiKassettB5iFishermaterialet‘Bergenhus’.IRiksantikvarensarkiv,Bergen.
Fischer,Gerhard,andDorotheaFischer1980.NorskekongeborgerII.Vol.2,Norskeminnesmerker.
Oslo.
Floderus,Erik1941.Sigtuna.Sverigesäldstemedeltidsstad.Stockholm.
284
Flodin,Lena1989.KammakerietiTrondheimca.1000-1600.Vol.14,Meddelelser.Fortideni
Trondheimbygrunn.Trondheim.
Frandsen,LeneB.,PerKristianMadsen,andHansMikkelsen1988.Byudgravningerog
bygningsarkæologiskeundersøgelseriRibe1983-89.By,marskoggeest.Vol.1,3-29.
Fritzner,Johan1973(1867).OrdbokoverDetgamlenorskesprog.4.utg.Oslo,Bergen,Tromsø.
Fritzvold,H.K1976.StrandlinjeniBergenår1000-1100.NOTEBYNorsktekniskbyggekontroll
A/S.Grunnundersøkelseogvurdering22/12/1976.
Fsk:1902-3.Fagrskinna,Ed.F.Jónsson.København.
Fuglesang,SigneHorn1981.Woodcarvers-ProfessionalsandAmateurs-inEleventh-century
Trondheim.EconomicAspectsoftheVikingAge.Ed.D.M.WilsonandM.L.Caygill.
London.Vol.30.
Fuglesang,SigneHorn1984.WoodcarvingfromOsloandTrondheimandsomeReflectionson
PeriodStyles.FestskrifttilThorleifSjøvoldpå70-årsdagen.Ed.J.H.Larsen,M.Høgestøl,
E.StraumeandB.Weber.Oslo.Vol.5.
Fuglesang,SigneHorn1991a.Ornament.DagliglivetsgjenstanderdelII.Dearkeologiskeundersøkelseri
Gamlebyen,Oslo.Ed.P.B.MolaugandE.Schia.ØvreErvik.Vol.8,159-222.
Fuglesang,SigneHorn1991b.Spoons.DagliglivetsgjenstanderdelII.Dearkeologiskeundersøkelseri
Gamlebyen,Oslo.Ed.P.B.MolaugandE.Schia.ØvreErvik.Vol.8,223-250.
Furnes,Harald2001.Resultatavanalyserav5spinnehjulogettbryne.DepartmentofEarth
Science,UniverstityofBergen.Archivereport.Topografiskarkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Fægri,Knut1979.Etnobotaniskbidragtilfunksjonsanalyse.EksemplerframiddelalderundersøkelseriBergen.Arkeo.Vol.1,36-39.
Færden,Gerd1990.Metallgjenstander.DagliglivetsgjenstanderdelI.Dearkeologiskeundersøkelseri
Gamlebyen,Oslo.Ed.P.B.MolaugandE.Schia.ØvreErvik.Vol.7,181-292.
Giddens,Anthony1979.CentralProblemsinSocialTheory.London.
Giddens,Antony1984.TheConstitutionofSociety.OutlineoftheTheoryofStructuration.Berkley.
Giddens,Anthony1995(1981).AContemporaryCritiqueofHistoricalMaterialism.Basingstoke.
Gilje,Nils,andHaraldGrimen1992.Samfunnsvitenskapenesforutsetninger.2.utg.Bergen.
Golembnik,Andrzej1993.ReportontheExcavationsinFinnegården3A,1982,Riksantikvarens
UtgravningskontorforBergen,Bergen.
Golembnik,Andrezj1995.StratigraphicReconstructionoftheUrbanDepositsattheSitesof
Finnegården3A,Dreggsalmenning14-16andSkostredet10inBergen.AcquisitionofField
DataatMulti-StrataSites.Ed.P.Urbanczyk.Warszawa.Vol.2,301-328.
Golembnik,Andrzejinprep-a.BRM237Dreggsalmenning14-16(1986).Manuscript,
TopografiskArkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Golembnik,Andrzejinprep-b.BRM237Dreggsalmenning14-16(1990).Manuscript,
TopografiskArkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Golembnik,Andrzejinprep-c.BRM346Skostredet10.Fielddocumentation,TopografiskArkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Grandell,Axel1988.FindsfromBryggenindicatingBussinessTransactions.TheBryggenPapers
SupplementarySeries.Ed.A.E.Herteig.Bergen.Vol.2,66-72.
Grew,Francis,andMargrethedeNeergaard1988.ShoesandPattens.Vol.2,MedievalFindsfrom
ExcavationsinLondon.London.
Grieg,S1933.MiddelalderskeByfundfraBergenogOslo.Oslo.
Grimnes,Ø.W1937.BergensTopografiiMiddelalderen.BergenHistoriskeForeningsSkrifter.Vol.43.
Groenman-vanWaateringe,W1978.ShoeSizesandPaleodemography.Helinium.Vol.XVIII,184189.
Göthberg,Hans1982.BRM108Kaigaten4-6.Archivereport,TopografiskArkiv.Bergen,
285
Hordaland.BergenMuseum,Bergen.
Hagen,KarinGjøl(1988)1994.Profesjonalismeogurbanisering.2.utg.Vol.16,Universitetets
OldsaksamlingSkrifter.Nyrekke.Oslo.
Hagland,JanRagnar1986.Runefunna.Eikjeldetilhandelshistoria.Vol.8,Meddelelser.Trondheim.
Hansen,Gitte1991.BRM340ChristianMichelsensgatev/Tinghuset.Archivereport,Topografisk
Arkiv.Bergen,Hordaland.BergenMuseum,Bergen
Hansen,Gitte1992.VetrlidsalmenningenBRM342,V3feltetogProfil56og57.Archivereport,
TopografiskArkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Hansen,Gitte1994a.BRM464Allehelgensgate3-5.Archivereport,TopografiskArkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Hansen,Gitte1994b.Denoverordnedebebyggelsestopografiomkring1190iBergen.Riksantikvaren
UtgravningskontoretforBergen.Bergen.
Hansen,Gitte1994c.St.Hansstredetfriområde.Archivereport,ToppografiskArkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Hansen,Gitte1995a.BRM474Håkonshallenssprinkleranlegg.Archivereport,TopografiskArkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Hansen,Gitte1995b.LilleØvregatenFriområdeBRM465,RiksantikvarenDistriktskontorVest,
Bergen.
Hansen,Gitte1998.TheBryggenChronology.NewLightupontheDatingofthePeriodsbefore
FireV.MedievalFiresinBergen-Revisited.BryggenPapersSupplementarySeries.Ed.I.Øye.
Bergen.Vol.6,81-127.
Hansen,Gitte2000.Bydannelseogforklaringafsocialefænomener.Individualisme,kollektivisme
ogGiddensstrukturationsteori.META.Vol.2000/4,2-16.
Harris,ChristopherJohn1991.Bergenikartfra1646tilvårtårhundre.Bergen.
Hartvedt,GunnarHagen1994.Bergenbyleksikon.Oslo.
Helle,Knut1982.Bergenbyshistorie.KongsseteogKjøpstad.Fraopphavettil1536.Vol.1,Bergenbys
historie.Bergen.
Helle,Knut1992.TidligbyutviklingiVestnorge.Våreførstebyer.OnsdagskvelderiBryggens
Museum.Ed.I.Øye.Bergen.Vol.7,7-30.
Helle,Knut1995.Underkirkeogkongemakt1130-1350.Ed.K.Helle.Vol.3,Aschehougs
norgeshistorie.Oslo.
Helle,Knut1998.MedievalFiresinBergenaccordingtoWrittenSources.MedievalFiresinBergen
-Revisited.BryggenPapersSupplementarySeries.Ed.I.Øye.Bergen.Vol.6,15-80.
Helle,Knut,andArnvedNedkvitne1977.Norge.UrbaniseringsprocesseniNorden.
Middelaldersteder.Ed.G.A.Blom.Oslo.Vol.1,189-272.
Herteig,AsbjørnE1957.KaupangeniBorgund.BorgundogGiske.Bergen.Vol.1,421-474.
Herteig,AsbjørnE1969.Kongershavnoghandelssete.Oslo.
Herteig,AsbjørnE1970.Bergensopprindelse.BergenHistoriskeForening.Skrifter.Vol.69/70,9-26.
Herteig,AsbjørnE1971a.DeteldsteBergen.BergensTidende,28/08/1971.
Herteig,AsbjørnE1971b.DeteldsteBergenendaengang.BergensTidende,27/10/1971.
Herteig,AsbjørnE1975.BorgundinSunnmøre.Topography,HistoryofConstruction,State
ofResearch.ArchaeologicalContributionstotheEarlyHistoryofUrbanCommunitiesin
Norway.Oslo,23-48.
Herteig,AsbjørnE1985.TheArchaeologicalExcavationsatBryggen,“TheGermanWharf”,in
Bergen,1955-68.BryggenPapers.MainSeries.Bergen.Vol.1,9-49.
Herteig,AsbjørnE1990.TheBuildingsatBryggentheirTopographicalandChronological
Development.Ed.A.E.Herteig.Vol.3,part1,TheBryggenPapers.MainSeries.Bergen.
Herteig,AsbjørnE1991.TheBuildingsatBryggentheirTopographicalandChronological
development.Ed.A.E.Herteig.Vol.3,part2+plates,TheBryggenPapers.MainSeries.
286
Bergen.
Herteig,AsbjørnE1992.The‘CellarBuildings’andPriviesatBryggen.TheBryggenPapers,
SupplementarySeries.Ed.A.E.Herteig.Bergen.Vol.5,287-320.
Herteig,AsbjørnE2000.NyttlysoverBergenstidligstehistorie.SærligOlavKyrreskongsgjerning.
BergensHistoriskeForening.Skrifter.Vol.95,3-11.
Hill,David1994.AnUrbanPolicyforCnut?TheReignofCnut:KingofEngland,Denmarkand
Norway.Ed.A.R.Rumble.London.101-105.
Hjelle,KariLoe1986.PaleobotaniskundersøkelseavmarinesedimentogavfallslagiVeisan-et
bidragtilbosetningshistorieniBergen.Cand.scient.-opgaveispeciellbotanik,Botanisk
Institut,UniversitetetiBergen,Bergen.
Hjelle,KariLoe1987.Rapport,Kroken3,botaniskeundersøgelser.BRM223Kroken3.Ed.A.R.
Dunlop.Bergen.61-69.
Hjelle,KariLoe1989.BotaniskeundersøkelseriforbindelsemedgrøftegravningeriBergen
sentrum,1989.Archivereport,TopografiskArkiv.Bergen,Hordaland.BergenMuseum,
Bergen.
Hjelle,KariLoe1994.Pollenanalyse.InnberetningomutgravningerneiBRM245Domkirkegaten6,
1987.Ed.J.Komber,A.R.Dunlop,J.V.SigurdssonandK.L.Hjelle.Bergen.158-171.
Hjelle,KariLoe1998.PaleobotaniskeundersøkelserNedreKorskirkealmenning/Vågsalmenning,
VågsbunneniBergen1998.Archivereport,NIKU,TopografiskArkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Hjelle,KariLoeundated.Resultsofthepalynologicalinvestigationsofdepositsat
Dreggsalmenningen14-16.Archivereport,TopografiskArkiv.Bergen,Hordaland.Bergen
Museum,Bergen.
Hkr:1893-1901.Heimskringla,Ed.F.Jónsson.København.
Hkr:1911.Heimskringla,Ed.F.Jónsson.København.
Hkr:1982(1979).Noregskongesoger.TranslatedbySchjøtt,SteinarandMagerøy,Hallvard.Ed.F.
HødnebøandH.Magerøy.Gjøvik.
Hodges,Richard(1982)1989.DarkAgeEconomics.TheOriginofTownsandTradeAD600-1000.
Ed.C.Renfrew.2.ed,NewApprochesinArchaeology.
Hodges,Richard1999.DarkAgeEconomicsRevisited.InDiscussionwiththePast.Archaeological
StudiespresentedtoW.A.vanEs.Ed.H.Sarfatij,W.J.H.VerwersandP.J.Woltering.
Zwolle.227-231.
Hodges,Richard2000.TownsandTradeintheAgeofCharlemagne.Ed.R.Hodges,Duckworth
DebatesinArchaeology.London.
Holtsmark,A1961.Sverressaga.Entalemotbiskopene.Oslo.
Holtsmark,Anne1970.Orkenøyingenessaga.Oslo.
Holtsmark,Anne,andDidrikArupSeip1975.SnorreSturlassonKongesagaer.Stavanger.
Hommedal,AlfTore1987.OlavsklosteretiOslo.EitDominikaneranleggfråhøgmellomalderen.
Foreningentilnorskefortidsminnersbevaring.Årbok.Vol.141,1987,129-154.
Hommedal,AlfTore1999.ArkeologiskundersøkingpåØvreDreggsalmenning/Slottsgaten,
Bergen(BRM553).Archivereport,NIKU,TopografiskArkiv.Bergen,Hordaland.Bergen
Museum,Bergen.
Hufthammer,AnneKarin1987.BeinmaterialetfraBryggenogerhvervslivetimiddelalderen.
Kystliv.Ed.I.Øye.Bergen.Vol.3,59-71.
Hufthammer,AnneKarin1994.TheDogBonesfromBryggen.TheBryggenPapersSupplementary
Series.Ed.A.E.Herteig.Bergen.Vol.5,209-286.
Hurley,MauriceF.,OrlaM.B.Scully,andSarahW.J.McCutcheon1997.LateVikingAgeand
MedievalWaterford.Excavations1986-1992.Ed.T.Barry,R.M.ClearyandM.F.Hurley.
Waterford.
287
Høeg,OveArbo1976.Planterogtradisjon.Oslo.
Iversen,Frode2004.Eiendom,maktogstatsdannelse.KongsgårderoggodsiHordalandiyngre
jernalderogmiddelalder.Dr.art.avhandling,ArkeologiskInstitutt,UniversitetetiBergen,
Bergen.
Jennings,Sarah,andAndrewRogerson1994.ThedistributionofGrimstonWareinEastAnglia
andbeyond.TheLateSaxonandMedievalPotteryIndustryofGrimston,Norfolk:Excavations
1962-92.Ed.M.Leah.Norfolk.Vol.64,116-120.
Jensen,Stig1990.Handelmeddagligvareriivikingetiden.HandelogudvekslingiDanmarksoldtid.
Ed.A.B.GebauerandS.Jensen.Højbjerg.Vol.16,119-138.
Jensen,Stig1992.RibesVikinger.DenantikvariskeSamling.Ribe.
Johnsen,IngridSanness1990.BryggeniBergen.Ed.J.Knirk.Vol.6,andethefte,Norgesinnskrifter
meddeyngreruner.Oslo.
Johnson,Trine1988.BRM273Nygaten5.Archivereport,TopografiskArkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Kaland,PeterEmil1979.LandskapsutviklingogbosetningshistorieiNordhordlands
lyngheiområde.Påletingetterdeneldstegarden.Ed.R.FladbyandJ.Sandnes.Oslo.41-70.
Kellmer,IngerKammer,Bryggeprosjektet.Archivemanuscriptandnotes.Topografiskarkiv.
Bergen,Hordaland.BergenMuseum.Bergen.
Kjersgaard,Erik1978.MadogøliDanmarksmiddelalder.København.
KLNM1956-78.Kulturhistoriskleksikonfornordiskmiddelalder.København.
Koch,HanneDahlerup1982a.BRM106Heggebakken/Sentrum.Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Koch,HanneDahlerup1982b.BRM136Kaigaten(v/Narvesen).Archivereport,Topografisk
arkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Koch,HanneDahlerupUndated.Torggaten1c-1d.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Komber,Jochen,A.RoryDunlop,JonVidarSigurdsson,andKariL.Hjelle1994.Innberetningom
utgravningerneiBRM245Domkirkegaten6,1987.Bergen.
Koren-Wiberg,Christian1900.JournalførtunderKontoretsnedrivning.BergeniJanuar1900.
Bergenbyarkivsskrifter.RekkeBnr.3Nordlandshandelarkivet.Bergen1984.Kat.nr.
646.SeriesignaturHd.5.Bergen.
Koren-Wiberg,Christian1908a.BidragtilBergensKulturhistorie.Vol.2,DetHanseatiskeMuseums
Skrifter.Bergen.
Koren-Wiberg,Christian1908b.Protokol12/6-5-6/81908.Bergenbyarkivsskrifter.RekkeB,
nr.3.Nordlandshandelarkivet.Bergen1984.Kat.nr.646.SeriesignaturHd.1-7.Bergen.
Koren-Wiberg,Christian1921.BergenskKulturhistorie.Vol.3,DetHanseatiskMuseumsSkrifter.
Bergen.
Krag,Claus1995.Vikingtidogrikssamling.Ed.K.Helle.Vol.2,Aschehougnorgeshistorie.Oslo.
Krzywinski,K1991.Botanikkibyarkeologisksammenhæng.Norskbyarkeologiinni1990-årene.
NyttfraUtgravningskontoretiBergen.Bergen.Vol.1,137-154.
Krzywinski,Knut,andKariLoeHjelle1985.BRM220Kroken7-9.Projektskisse.Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Krzywinski,Knut,andPeterEmilKaland1984.Bergen-fromFarmtoTown.Ed.A.E.Herteig.
BryggenPapersSupplementarySeries.Bergen.Vol.1,1-39.
Larsen,ArneJ1967a.InnberetningomutgravningenavNeoFabrikkerstomtSandbrugaten5.
Archivereport,Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Larsen,ArneJ1967b.InnberetningomutgravningeniDreggsalmenningen20.Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Larsen,ArneJ1970.SkomaterialetfrautgravningeneiBorgundpåSunnmøre1954-1962.Vol.1,
288
ÅrbokforUniversitetetiBergen.Humanistiskserie.Bergen.
Larsen,ArneJ.1975.BRM64Øvregaten41.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Larsen,ArneJ1978.Øvregaten25.Archivereport,Topografiskarkiv.Bergen,Hordaland.Bergen
Museum,Bergen.
Larsen,ArneJ1992.FootwearfromtheGullskoenAreaofBryggen.Ed.A.E.Herteig.Vol.4,The
BryggenPapers.MainSeries.Bergen.
Larsen,ArneJ,andEgillReimers1978.Øvregaten37-39.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Lidén,HansEmil1993.DeeldstekirkeriBergen.BergenHistoriskeForening.Skrifter.Ed.G.A.
Ersland,K.EngelsenandA.Haaland.Bergen.Vol.89,71-82.
Lidén,HansEmil,andEllenMarieMagerøy1980.NorgesKirker,Bergen.3vols.Vol.1,Norske
Minnesmerker.Oslo.
Lidén,HansEmil,andEllenMarieMagerøy1983.NorgesKirker,Bergen.3vols.Vol.2,Norske
Minnesmerker.Oslo.
Lidén,HansEmil,andEllenMarieMagerøy1990.NorgesKirker,Bergen.3vols.Vol.3,Norske
Minnesmerker.Oslo.
Lind,KethE1991.Skosommateriellkultur.Vàgarsamfunnetimiddelalderen.
Magistergradsavhandling,Arkeologi,UniversittetiTromsø,Tromsø.
Lindh,Jan1979.BRM76Rosenkrantzgate4,(19978-79).Innberetning.Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Lindholm,F1916.Registreringafgrundforhold.HotelRosenkrantz.Archivereport,
Byggesaksarkivet,BergenRådhus.BergenKommune,Bergen.
Long,CliffordD,andLyderMarstrander1980.DreggsalmenningeniBergen.(BRM83).
Rapportfradearkeologiskeundersøkelser1979.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum.
Lorentzen,Bernt1952.GårdoggrunniBergenimiddelalderen.Vol.16,DetHanseatiskemuseums
skrifter.Bergen.
Lorentzen,Bernt1971a.Omkringdetførstebyanlegget.BergensTidende,07/08/1971.
Lorentzen,Bernt1971b.Omkringdetførstebyanlegget.VågsbotnenogStrandsiden.Bergens
Tidende,14/08/1971.
Lund,Niels1994.Cnut’sDanishKingdom.TheReignofCnut:KingofEngland,Denmarkand
Norway.Ed.A.R.Rumble.London.27-42.
Lüdtke,Hartwig1989.TheBryggenPotteryI.IntroductionandPingsdorfWare.TheBryggen
Papers.SupplementarySeries.Bergen.Vol.4.
Madsen,PerKristian1985.TheEarliestDatedFindsofGlazedPotteryinRibe.Medieval
Ceramics,BulletinoftheMedievalPotteryResearchGroup.Vol.9,57-63.
Madsen,PerKristian1988.OntheDatingofMedievalPottery-intheLightofRecentFindsfrom
Ribe.JournalofDanishArchaeology.Vol.6,1987,190-197.
Madsen,PerKristian1996.BleiglasierteHochmittelalterlicheIrdenwareinNordeuropa.Töpferei-
undKeramikforschung.Ed.H.LüdtkeandR.Vossen.Bonn.Vol.3,15-29.
Marstein,Oddlaug1989.Skoogandregjenstanderilær-entypologiskanalyse.Vol.23,Meddelelser.
FortideniTrondheimbygrunn.Trondheim.
Marstrander,Lyder1983.EnalmenningiBergen.PaperreadatHus,gårdochbebyggelse.Det
XVInordiskaarkeologmøtet,atIsland.
Matthiessen,H1927.MiddelalderligeByer.BeliggenhedogBaggrund.København.
Mejdahl,Vagn1988.Nordisklaboratoriumfortermoluminiscens-datering(vejledningforbrugere).
Roskilde.
Meyer,Hildebrand1904(1764).SamlingertildenBerømmeligeogNavnkundigeNorskeHandelStad
289
BergensBeskrivelsevedBorgemesterMeyer.BergenHistoriskeForening.Ed.B.E.Bendixen.
Bergen.
MHN:1880.MonumentaHistorica,Norvegiæ,Ed.G.Storm.Kristiania.
Mitchell,J.G.,H.Askvik,andH.G.Resi1984.Potassium-argonAgesofSchistHonestonesfrom
VikingAgeSitesatKaupang(Norway),Aggersborg(Denmark),Hedeby(WestGermany)
andWolin(Poland),andtheirArchaeologicalImplications.JournalofArchaeological
Science.Vol.11,171-176.
Molaug,PetterB1977.Leirkarmaterialetfra‘MindetsTomt’.FeltetMindetsTomt.Stratigrafi-
Topografi-Daterendefunngrupper.DearkeologiskeutgravningeriGamlebyen,Oslo.Ed.H.I.
Høeg,H.E.Lidén,A.Liestøl,P.B.Molaug,E.SchiaandC.Wiberg.Oslo.Vol.1,72-120.
Molaug,PetterB1987.Leirkarmaterialet.SøndreFelt.Stratigrafi,bebyggelsesresterogdaterende
funngrupper.DearkeologiskeutgravningeriGamlebyen,Oslo.Ed.E.Schia.Vol.3,229-328.
Molaug,PetterB1991.Gamlebyprojektet-forskningsfrontellerrepetitionsøvelse?Norsk
byarkeologiinni1990-årene.NyttfraUtgravningskontoretiBergen.Ed.S.Myrvoll,A.
ChristenssonandÅ.Bergset.Bergen.Vol.1.
Moldung,HanneMereteR.2000.Etstedimiddelalderbyen.Hovedfagsoppgave,Arkeologisk
Institutt,UniversitetetiBergen,Bergen.
Msk:1932.Morkinskinna,Ed.F.Jónsson.København.
Munch,PA1849.Historisk-geographiskBeskrivelseoverKongerigetNorgeiMiddelalderen.Moss.
Myhre,Bjørn1985.TrendsinNorwegianarchaeology.JournalofDanishArchaeology.Vol.4,179185.
Myrvoll,Siri1980.BRM143Koren-WibergsPlass.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Myrvoll,Siri1986.SkienogTelemark-naturressurser,produkterogkontakterisenvikingtidog
tidligmiddelalder.Viking.Vol.XLIX1985/86,161-180.
Myrvoll,Siri1987.ArchaeologicalInvestigationsinBergen1980-86:SomeNewAspectsonthe
DevelopmentoftheTown.NorwegianArchaeologicalReview.Vol.20/2,100-109.
Myrvoll,Siri1991.Forvaltningsstyrtforskning?ByundersøkelseriBergen1979-1989.Norsk
byarkeologiinni1990-årene.NyttfraUtgravningskontoretiBergen.Ed.S.Myrvoll,A.
ChristenssonandÅ.Bergset.Bergen.Vol.1.
Myrvoll,Siri1992.HandelstorgetiSkien-aStudyofActivityonanEarlyMedievalSite.Ed.S.
Myrvoll.Vol.2,NyttfraUtgravningskontoretiBergen.Bergen.
Myrvoll,Siri1993.SiriMyrvollisamtalemedKirstenEngelsen.Grøftegravningogarkeologii
Vågsbunnen.BergenHistoriskeForeningsSkrifter.Vol.89,83-99.
Mårtensson,AndersW.,ed.1976.UppgrävtförflutetförPKbankeniLund.Vol.VII,Archaeologica
Lundensia.Lund.
Narmo,LarsErik1997.Jernvinne,smieogkulproduksjoniØsterdalen.Vol.43,Varia.Oslo.
Narmo,LarsErik(1991)1996.JernvinnaiValdresogGausdal-etfragmentavmiddelalderens
økonomi.Vol.38,Varia.Oslo.
NgL:1846-95.NorgesGamleLove,Ed.P.A.M.R.Keyser,G.Storm,E.Hertzberg.Christiania.
Nicolaysen,Nicolay1861.OmMunkelifsklosteretiBergenogdetslevninger.Aarsberetning1860.
Chrisitiania.59-79.
Nielsen,ErikLevin1969.PederstrædeiViborg.Købstadarkæologiskeundersøgelser1966/67.
KUML.Vol.1968,23-81.
Nielsen,Solveig1997.ByudviklingiSkandinavieniperioden700-1100e.Kr.LAG6.Ed.N.Haue
andM.Runge.Højbjerg.Vol.6,179-227.
Nielsen,Yngvar1877.BergenfradeældsteTiderindtilNutiden.Enhistorisk-topografiskSkildring.
Christiania.
Nordeide,SæbjørgWalaker1989.”...Debestebønderikiøbstæden...”Enfunksjons-ogaktivitetsanalyse
290
basertpågjenstandsmaterialet.Vol.20,Meddelelser.FortideniTrondheimbygrunn.
Trondheim.
Nordeide,SæbjørgWalaker1999.Urbaniseringsprosessen-påkvinnersvilkår?NIKU1994-1999
Kulturminneforskningensmangfold.Ed.G.Gundhus,E.SeipandE.Ulriksen.Oslo.Vol.
31,44-48.
NOTEBY1978.KabeltrasseHeggen-Kaigatensak,12273.Archivereport,NorskTeknisk
byggekontroll,Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Nybø,Marit2000.AlbanuskirkenpåSelja.Klosterkirkeellerbispekirke?Dr.art.avhandling,
Historisk-filosofiskefakultet,UniversitetetiBergen,Bergen.
Näsman,Ulf1990.OmfjärrhandeliSydskandinaviensyngrejärnålder.Handelmedglasunder
germanskjärnålderochvikingetid.HandelogudvekslingiDanmarksoldtid.Ed.A.B.
GebauerandS.Jensen.Højbjerg.Vol.16,89-118.
Næss,JennyRita1963.InnberetningomUtgravningenepåRådstuplass1-2.Archivereport,
Topografiskarkiv.Bergen,Hordaland.BergenMuseum,Bergen.
Nøttveit,Ole-Magne2000.MiddelalderskevåpenfunnfraVestlandet.Hovedfagsoppgave,
ArkeologiskInstitutt,UniversitetetiBergen,Bergen.
Ohlsson,Tom1973.VikingatidochmedeltidiLöddeköpinge.ALEHistorisktidsskriftför
Skåneland.Vol.1,27-42.
Olsen,Olaf1975.NogletankerianledningafRibesuventethøjealder.FraRibeAmt1975,224258.
Olsen,OleMikal1998.EianalyseavfiskereiskapfråmellomalderenfunneiBergen.
Hovedfagsoppgave,ArkeologiskInstitutt,UniversitetetiBergen,Bergen.
Olsen,OleMikal2004.MedievalFishingTacklefromBergen.TheBryggenPapersMainSeries.Ed.
I.Øye.Fagbokforlaget.Bergen.Vol5,7-106.
OrdericusVitalis:1972-80.TheEcclesiasticHistoryofOrdericVitalis.OxfordMedievalTexts.Ed.
M.Chibnall.Oxford.Vol.I,III-V.
Orkn:1913-16.Orkneyingasaga,Ed.S.Nordal.København.
Pedersen,Unn2001.Vektlodd-sikrevitnesbyrdomhandelsvirksomhet?Primitivetider.19-36.
Reed,Ian1990.1000YearsofPottery.AnAnalysisofPottery,TradeandUse.Vol.25,Meddelelser.
FortideniTrondheimbygrunn.Trondheim.
Reimers,Egill1965.Mariakirkensomgivelser.Bergen.
Reimers,Egill1971a.BRM19Koren-WibergsPlass.Archivereport,Topopgrafiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Reimers,Egill1971b.KongOscarsgate/Heggebakken.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Reimers,Egill1972a.BRM11Øvregaten9.Archivereport,Topografiskarkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Reimers,Egil1972b.BRM42Dreggsalmenningen10-12.Archivereport,Topografiskarkiv.
Bergen,Hordaland.BergenMuseum,Bergen.
Reimers,Egill1973a.BellgårdenSteinkjeller.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Reimers,Egill1973b.BRM44BugårdenSogBredsgårdenN.Archivereport,Topografiskarkiv.
Bergen,Hordalaland.BergenMuseum,Bergen.
Reimers,Egill1973c.Hollendergaten9.Archivereport,Topografiskarkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Reimers,Egil1974.FleretopografiskedatamellomHolmenogBryggeniBergen.ARKEO.Vol.
1974:1,21-24.
Reimers,Egill1977.RegistreringØvregaten.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum.
291
Reimers,Egill1979.Peterskirkenskirkegårdsmur.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Reimers,Egillinprep.StaveriBryggenmaterialet.
Riisøen,KirstiHauge2001.Kvalitativogkvantitativanalyseav10gjenstanderfra
Middelaldersamlingen,BergenMuseum.Konserveringsavdelingen,BergenMuseum
UniversitetetiBergen.Archivereport,Topografiskarkiv.Bergen,Hordaland.Bergen
Museum,Bergen.
Robinson,David,HansKrongaardKristensen,andIdaBoldsen1992.BotanicalAnalysesfrom
ViborgSøndersø.AWaterloggedUrbanSitefromtheVikingPeriod.ActaArchaeologica.
Vol.62-1991,59-87.
Roesdal,Else1980.Danmarksvikingetid.Viborg.
Roesdal,Else1993.OnKeys.Twenty-eightpaperspresentedtoHansBekker-NielsenontheOccation
ofhissixtiethBirthday28April1993.North-westernEuropeanLanguageEvolution.
NOWELE.Vol.21/221993217-224.OdenseUniversitypress.Odense.
Ros,Jonas2001.Sigtuna.Staden,kyrkornaochdenkyrkligaorganisationen.Ed.O.KyhlbergandF.
Herschend,OccasionalPapersinArchaeology.Uppsala.
Roslund,Mats1997.Pådriftintidochrum?Ominformationspotentialenikomposita
dokumentationsmaterial.META.Vol.1997/3,37-53.
Rui,LivMarit1993.Etarkeologiskmaterialeframiddelalder-Osloifeministiskperspektiv.
Magistergradsafhandlinginordiskarkeologi,Instituttforarkeologi,UniversitetetiOslo,
Oslo.
Rumble,AlexanderR,ed.1994.TheReignofCnut:KingofEngland,DenmarkandNorway.Ed.N.
Brooks,StudiesintheearlyHistoryofBritain:MakersofEngland.London.
Ruth,LindaCain2000.DesignStandardsforChildren’sEnvironments.NewYork.
Rytter,Jens1991.KammefraKunghälla.Kungahällaarkeologi1989.Ed.K.Carlsson.Göteborg.
Vol.22.
Rytter,Jens1997.GevirhåndverketiKonghelleca1140-1300.Hovedoppgave,IAKN,Det
Historisk-filosofiskeFakultet,UniversitetetiOslo,Oslo.
Sarfatij,Herbert1999.TielinSuccessiontoDorestad.InDiscussionwiththePast.Archaeological
StudiespresentedtoW.A.vanEs.Ed.H.Sarfatij,W.J.H.VerwersandP.J.Woltering.
Zwolle.267-278.
Sawyer,Peter1994.Cnut’sScandinavianEmpire.TheReignofCnut:KingofEngland,Denmark
andNorway.Ed.A.R.Rumble.London.10-22.
Schia,Erik1975.SkomaterialetfraMindetstomt.MagistergradsavhandlingiNordiskArkeologi,
UniversitetetsOldsakssamling,UniversitetetiOslo,Oslo.
Schia,Erik1977.Skomaterialetfra”MindetsTomt”.Feltet’MindetsTomt’.Stratigrafi-Topografi-
Daterendefunngrupper.DearkeologiskeutgravningeriGamlebyen,Oslo.Ed.H.I.Høeg,H.
E.Lidén,A.Liestøl,P.B.Molaugh,E.SchiaandC.Wiberg.Oslo.Vol.1,121-199.
Schia,Erik1987a.ReconstructingTownyardsonthePeripheryofTheEuropeanUrbanCulture.
NorwegianArchaeologicalReview.Vol.20/2,81-96.
Schia,Erik1987b.Skoogstøvler.‘SøndreFelt’Stratigrafi,bebyggelsesresterogdaterendefunngrupper.
DearkeologiskeutgravningeriGamlebyen,Oslo.Ed.E.Schia.ØvreErvik.Vol.3,329-412.
Schia,Erik1991.OsloinnerstiViken.Livogvirkeimiddelalderbyen.Oslo.
Schia,Erik1992.DenførsteurbaniseringeniOslofjord-regionen.Våreførstebyer.Onsdagskvelderi
BryggensMuseum.Ed.I.Øye.Vol.7,31-58.
Schreiner,Johan1953.GårdoggrunniBergenimiddelalderen.HistoriskTidsskrift.Vol.36,429439.
Schück,A1926.Studierrörandedetsvenskastadsväsendetsuppkomstochäldstautveckling.Uppsala.
Seim,KarinFjellhammer1988a.AReviewoftheRunicMaterial.TheBryggenPapers
292
SupplementarySeries.Ed.A.E.Herteig.Bergen.Vol.2,10-23.
Seim,KarinFjellhammer1988b.RunicInscriptionsinLatin.AsummaryofAslakLiestøl’sFascicle
(Vol.VI,1)ofNorgesInnskriftermeddeyngreRuner.TheBryggenPapersSupplementary
Series.Ed.A.E.Herteig.Bergen.Vol.2,24-65.
Shahar,Shulamith(1990)1992.ChildhoodintheMiddleAges.2.ed.LondonandNewYork.
Skovgaard-Petersen,Inge1977.Oldtidogvikingetid.DanmarkshistorieI,tidenindtil1340.Ed.A.
E.Christensen,H.P.Clausen,S.EllehøjandS.Mørch.København.Vol.I,15-209.
Skre,Dagfinn1995.Kirkenførsognet.DentidligstekirkeordningeniNorge.Møtetmellom
hedendomogkristendomiNorge.Ed.H.-E.Lidén.Oslo.170-233.
Skre,Dagfinn,LarsPilø,andUnnPedersen2001.TheKaupangexcavationProject.Annual
Report.Oslo.
Skaare,Kolbjørn1984.CoinFindsfromBryggen.TheBryggenPapersSupplementarySeries.Ed.A.
E.Herteig.Bergen.Vol.1,52-72.
Sletten,Birte1984.BRM206Nøstegaten65a-91.Archivereport,Topografiskarkiv.Bergen,
Hordaland.BergenMuseum,Bergen.
Smedstad,Ingrid1991.Oslogate2-8,1987-1988.Grøftegravninger.Dearkeologiskeutgravningeri
Gamlebyen,Oslo.Ed.E.SchiaandT.Wiberg.ØvreErvik.7-62.
Sognnes,Kalle1974.BRM59NyeSandviksveifraMariakirkentilSkuteviken.Archivereport,
TopografiskarkivBergenHordaland.BergenMuseum,Bergen.
Solberg,Per1969.BRM11Øvregaten9.Archivereport,Topografiskarkiv.Bergen,Hordaland.
BergenMuseum,Bergen.
Solli,Brit1996.NarrativesofVeøy.AnInvestigationintothePoeticsandScientificsofArchaeology.
Vol.19,UniversitetetsOldsaksamlingsSkrifter.Oslo.
Ss:1920.SverrisSagaetterCod.AM3274,Ed.G.Indrebø.Kristiania.
Storm,Gustav,ed.1880.MonumentahistoricaNorvegia.Kristiania.
Storm,Gustav1899.DeKongeligeByanlægiNorgeiMiddelalderen.HistoriskTidsskrift.3.række,
femtebind,433-436.
Strømmen,T1969.Profiler,dybdertilfjell,fundamentplan.Øvregaten9.Archivereport.
Byggesaksarkivet,BergenRådhus.BergenKommune,Bergen.
Stuiver,Minze,andJohannesvanderPlicht1998.INTCAL98.Radiocarbon.Vol.40(3).
Sørlie,Mikjel1957(1559/60).BergensFundas.Bergen.
Tesch,S1990.Stadochstadplan.MaktochmänniskorikungensSigtuna.Sigtunagrävningen198890.Ed.S.Tesch.Sigtuna.23-37.
Tesch,Steen1992.TheTownPlan-aKeytoUrbanizationandFormationofStateinSweden.
MedievalEurope1,Urbanism.York.
Tilley,Christopher1989.InterpretingMaterialCulture.TheMeaningsofThings.Ed.I.Hodder.
London/NewYork.Vol.6,185-194.
Trumpy,Caspar1954.CasparTrumpysprosjekt260,tegning9,fundamentplan:coterforfjell
ogfastgrunn.Bradbenken1,BergenskeDampskipsselskapsadministrajonsbygg.Archive
report,InstanesA/SRådgivendeingeniøreroglandskapsarkitekter.Bergen.
Tørhaug,Vanja1998.SkomakerhåndverketiOslo1050-1300.Hovedfagsoppgaveinordisk
arkeologi,IAKN,UniversitetetiOslo,Oslo.
Ucko,P.J1989.Foreword.TheMeaningsofThings.Ed.I.Hodder.London/NewYork.Vol.6ixxvii.
Ulbricht,Ingrid1984.DieVerarbeitungvonKnochen,GeweirhundHornimmittelalterlichen
Schleswig.Ed.V.Vogel.Vol.3,AusgrabungeniSchleswig.BerichteundStudien.
Neumünster.
Ulriksen,Eli1996.Utkantenshåndværkereogarbeidere.Enaktivitetsanalyseav“Nordrebydel”i
middelalderensTønsberg,NIKUTemahefte,Tønsberg.
293
Ulriksen,Jens1998.Anløbspladser.BesejlingogbebyggelseiDanmarkmellem200og1100e.kr.
Odense.
Vince,Alan1991.EarlymedievalLondon:refiningthethechronology.TheLondonArchaeologist.
Vol.6(10),263-271.
Weber,Birthe1990.Tregjenstander.DagliglivetsgjenstanderdelI.Dearkeologiskeutgravningeri
Gamlebyen,Oslo.Ed.P.B.MolaugandE.Schia.ØvreErvik.Vol.7,11-180.
Wiberg,Christina1977.Hornogbenmaterialetfra‘MindetsTomt’.FeltetMindetsTomt.Stratigrafi
-Topografi-Daterendefunngrupper.DearkeologiskeutgravningeriGamlebyen,Oslo.Ed.H.
I.Høeg,H.E.Lidén,A.Liestøl,P.B.Molaug,E.SchiaandC.Wiberg.Oslo.Vol.1,202213.
Wiberg,Tina1987.Kammer.SøndreFelt.Stratigrafi,bebyggelsesresterogdaterendefunngrupper.De
arkeologiskeutgravningeriGamlebyen,Oslo.Ed.E.Schia.Vol.3,413-422.
Øye,Ingvild1988.TextileEquipmentanditsWorkingEnvironment,BryggeninBergen1150-1500.
Ed.A.E.Herteig.Vol.2,TheBryggenPapersMainSeries.Bergen.
Øye,Ingvild1989.KaupangeniSognikomparativbelysning.Viking,144-165.
Øye,Ingvild1997.State,TasksandOutlookforArchaeologyinBergen.LübeckerKolloquium
zurStadtarchäeologieimHanseraumI:Stand,AufgabenundPerspektiven.Ed.M.Gläser.
Lübeck.441-454.
Øye,Ingvild1998.Middelalderbyensagraretrekk.Bergen.
Ågotnes,Anneinprep.HusgerådpåBryggenc1150-1700.
Ågr:1936.Ågrip,eilitinorskkongssoge.Norrønebokverk,Ed.G.Indrebø.Oslo.
294
INDEX
A
Agriculture157,176,180,200,238
Alfiva(Ælfgifu)222
AllSaints(AlleHelgensKirke),site45103,104,154,
171,227
Alrekstad20,23,25,26,130,131,230,231,232,238
Ambulatingartisans184,189,191,193,194,201,202,
203,204,235,236,237,238,240
Animalhead(ornament)196
Antleroffcut160,162
Arrowhead176
B
Basiccooking157,177,178,179,180,200,201,202,
203,204,238
Beerbrewing177,178,180,200,203,204,219
Bjorgvin23,24,130
Boneoffcut160,162
Borgund21,23,61,62
Bottom-upinitiative32,33,221,223,227,230
Bow176
Bryggen(1955-1979)BRM000,site642,46-49,54,
58-63,66-68,72,74-77,79-83,90,92,94,95,104106,109-111,131,132,134-136,138-141,143-145,
150,153,159,160,163,178,201,208,209,212,262,
264-266,268-271
Bryggeparken(1989)BRM287,site1683,104,111,136
Byre87,177
C
Chalk163,164,168,169,170
Children218,219,272
ChristchurchCathedral(StoreKristkirke),site257,77,
81,87,88,89,92,93,95,104,109,110,111,134,
136,139,147,150,152,154,169,170,171,194,225,
226,227,232,235
Christchurchminor(LilleKristkirke),site357,101,
102,104,109,111,147,225,227
ChurchoftheApostles(Apostelkirken),site457,58,
104,108,150,152,154,169,170,171,225,226,233
Combmaking157,159,160,162,180,181,184,189,
201,203,204,271
Combblank159,160,162
CombD1181
CombD2181
CombE1181
CombE3181
CombE3-b181
CombE4181
CombE5-1181
CombE5-2181,184
CombE5-3onerowofrivets,noprofile181,182
CombE5-3onerowofrivets,oneprofile181,182
CombE5-3tworowsofrivets181,182
CombE5-5181
CombE6-1181
Comboffcut159,160,162
Conspicuousconsumption234,235,236,237
Copperalloy165,167,168,180,191,193,271
Craniumcat172
Craniumdog172
Crucible167,168
D
DarkGreySchisthoneswithawesternNorwegianorigin
206
Domkirkegaten6(1987)BRM245,site3842,47,101,
102,104,109,111,154,213,227
Dorestad33
Doubletenement25,26,33,131,132,134,136,138,
268
Dreggsalmenningen(1979)BRM83,site1481,82,104,
134,136,139,212,225
Dreggsalmenningen10-12(1972)BRM42,site1280,
295
81,104,136
Dreggsalmenningen10-16(1986)BRM242,site1380,
81,104
Dreggsalmenningen14-16(1986and1990)BRM237,
site869,70,71,72,75,76,104,111,134,142,143,
150,152,166,167
Dreggsalmenningen20(1967)BRM4,site1168,77,
78,79,80,104,105,106,132,134,136,138,143,
145,150,265,273
Drop-spindle173,175
Horizon3plotsystem142,232
Horn/antleroffcut160,162
Hornoffcut160,162
Householdproducers159,185,186,189,191,193,195
Householdwaste40,70,100,101,127,128,130
Hunting157,176,180,200,203,204,238,272
I
Iceland/Greenland209,210,213
Innkeeping200,201,203,204,236,238
Interrelatedsources44,48,53,106,128,147,148,154
E
England52,206,209,210,212,213,217,223,231
F
Finemetal165,167,168,191,236
Finnegården6a(1981)BRM104,site2642,47,48,49,
77,91-95,100,104,106,111,136,138,139,148,153,
210,212,216
Fire-layerchronology37,39,40
Fishing24,29,52,157,172,173,174,175,176,180,
194,195,200,203,204,216,217,234,238,291
Flax-beater173,175
Flax-comb173,175,271
Float174,175,176
Foodandbeverageprocessing157,177-180,196,200204,216-220,236,238,271
Foundedtownhypothesis24,25,30,31
France206,209,210,213,217
Furnacelining167,168
G
Gamingpiece160,162,185,186,196
Germany206,209,210,212,213,217,234
Grindingslab177-179
H
Hamar21
HaraldGille(Gilchrist)(1103-1136)91,226
HaraldHårfagre(HaraldFairhair)(-932)20,23,131,
271
HaraldSigurdssonHardråde(HaraldHardruler)(10461066)23
Harbourconditions205,207-211,213-215,217,239
Holmenarea38,53,56,127,130,131,248
HonesfromEidsborg35,206,207
Horizontallink44,45,47,48,62,68,72,76,80,105107,141,199,200,270
Horizon2plotsystem143,144,224,225,230
296
K
Kaupanger21
KaupanginTjølling(Sciringsheal)21,33,218
Klingesmauet(1989)BRM299,site2187,88,104,109,
111,136,153,154,177
Knifebeater173,175
KnutdenMektige(KnutthePowerful)19,27,222,223,
231,237,239,286
Koengen(BotanicalinvestigationinVeisanbyKariLoe
Hjelle)(1986),site142,47,56,77,80,81,82,104,
109-111,127,128,129,130
Konghelle21,158,184,271
Koren-WibergsPlass(1980)BRM143,site1884,104
Korskirken(1984)BRM200,site3599,102,104,154
Kriterienbündel19
Kroken3(1984)BRM223,site2288,89,104,111,153
L
Landskyld20
Last163-165,191
Lathe-turnedcore170,171,196,219,220
Lidforcontainer196
LilleØvregatenfriområde(1994)BRM465,site3499,
104,130,154
Lime-slaking164,168-170,194,236
Linen-smoother173,175
Linerunner160,162,174-176,185
Linesinker174-176
Long-distancetrade34,35,205-207,217,218,231-236,
239
Long-toothedcomb173,175,271
LowCountries83,206,209,210,212,213,217,234,
236,265
Lund33,48,158,181,183,184,193,194,223
M
MagnusOlavsson(1034-46)222,231
Metalworking34,157,165-168,180,191,193,194,201,
203,204,206,215,236
Methodologicalcollectivistapproach30
Methodologicalindividualistapproach30
Middletownarea18,24,33,36,37,40,53,55,91,106,
108,128,134,136,139-141,143-145,147,148,150,
152,154,156,160,162,163,167,169,170,173-176,
179,208-210,212,215,218-221,223,228,232,233,
235,236,238-240,270
Milk-processingcross178,179
Miscellaneousantler,bone,hornandwhale/walrusbone
working157,159,160,162,180,185,186,203,204,
216,238
Mould166,167,168
MunkelivBenedictineAbbeywiththeChurchofSt
Michael(Munkelivkloster),site4336,103,104,152,
154,169,225,233
Myricagale85,178,179,219
N
NedreKorskirkealmenning/Vågsalmenning(1998)BRM
544,site3747,100,101,104,109,128,130
Needle/pin160,162,171,173,175,185
Nettingneedle174,175,176
Nikolaikirkealmenningen(1985)BRM202,site1783,
104,148,150,153,156,227
Nonneseterarea18,36,103,108,145,148,152,154,
169,226-228,235,239
Nonneseterconvent(NonneseterKloster),site4636,
104,105,154,170,226,235
Nordnesarea18,36-38,55,103,108,145,148,152,154,
169,225,226,228,233,239,245,247
Northerntownarea18,24,27,29,37,41,53,58,64,
106,108,127,132,134,136,138-141,144,145,147,
148,152-154,156,157,160,163,167,169-180,207210,212,215,217,219,221-228,230,232,235,236,
238,239,270
Nygaten2(1991)BRM333,site42103,104
O
OlavHaraldsson(laterSaintOlaf)(1015-1028)20,21,
222,231,237,239
OlavKyrre(theGentle,’thepeaceful’)(1066-1093)17,
21,23-27,34,43,57,147,148,194,225,227,228,
231-233,237,239,240,246,247
OlavMagnusson(1003-1115)21
OlavTryggvason(994/995-999/1000)20,21
Organictownhypothesis24,27,30,31
Oslo21,24,29,33,36,37,181,188,189,193,266,269,
271,273
Otherleatherworking157,162,163,164,165,180,191,
203,238
P
Palisade-builtfence62-65,68,72-76,78,79,105,131,
132,134,138,140,141,143-145,230,269
Plan-unit27,228,240
Primarylayer39
Professionalambulatingproducers158,159,184,189,
193,201,203,204
Professionalsedentaryfull-timeproducers159,204
Professionalsedentarypart-timeproducers159,204
Punch159-162
R
Rabbet171
Rådstuplass2-3,’Vestlandsbanken’(1963)BRM20,site
41103,104
Reed-hook173,175
Ribe19,33,218,271
Rosenkrantzgaten4(1978/79and1981)BRM76,site28
77,94,95,104,110,111,138,153
Runicinscription29,39,185,207
S
Sandbrugaten5(1967)BRM3site972-77,79,80,104106,110,132,134,138,141,143,145,147,167,261,
262,273
Sarpsborg(Borg)21,222
Sausagemaking177,201,203,204
Sausagepin177-179,196,201,202,216,271
Secondarylayer39
Setjakaupstad23,26,221
Shears173,175
Shoemaking34,157,162-165,180,186,189,191,201,
203,204,216,217,238
Sickle177
Sigtuna33,48,181
SigurdMagnussonJorsalfar(SigurthJerusalemfarer)
(1103-1130)21,234
Skewer178,179,271
Skien21
Skinning157,172,180,199,203,204
Skostredet10(1992)BRM346,site3699-101,104,
128,154
Slag110,165-168,191,193
Slateoffcut169,170
Sling176
297
Socialchange30,31,230,237
Southerntownarea18,27,29,36,37,41,53-55,99,108,
128,143,145,147,148,152,154,156,157,169,179,
213,219,226-228,233,235,239
Spearhead176
Spindlewhorl169,170,173,175,185,194,195,271
Spoon196
Stable87,177
Stallen,Svensgården(1980/82)BRM90,site1582,83,
104,110,111,136,143,145,153
Stavanger21,54,
Stavebeaker206,207
Steatiteoffcut169,170
Steatitevessel178,179,194,195
Steinkjer21
Stoneworking157,168-170,180,194,196,203,204,
212,216,217,232,238
Storagebuilding205,207,209,210,213,217,239
Stratigraphicalexcavation38,39
StColumba’s(Steinkirken),site33,37,99,104,154,169,
170,227
StCross(Korskirken),site4029,37,38,99,102,104,
108,131,154,169,170,226,233,244
StJohn’sAugustinianAbbey(Jonskloster),site44103,
104,154,170,226,235
StMary’s(Mariakirken),site2329,37,38,53,81,82,
89,90,99,102,104,110,134,136,143,152,154,
169,170,194,225,226,232,233,236,241,270
StNicholas’(Nikolaikirken),site3229,36,98,104,
134,136,150,152,154,156,169,225-227,233,243
StOlav’sinVågsbotn(OlavskirkeniVågsbotn),site39
99,102,104,154,170,227
StOlav’sontheHill(OlavskirkenpåBakkene),site25
91,104,154,171,226,233
StPeter’s(Peterskirken),site2437,90,104,136,139,
141,143,154,169,170,227,242
SveinKnutsson(Swein)(1030-34)222
T
Tallystick205,207,209,210,212,213,215,217
Temporaryworkshop201,204,238
Textileproduction29,34,52,157,172-175,180,199,
200,203,204,216-220
Theoryofstructuration31
Thedoublenucleussituation28,29,240
Tønsberg21,33
Top-downinitiative32,33,221,227
Touchstone166-168,270
Townplan143,145
298
Toy170,219,220
Trialpiece160,162
Trondheim(Nidaros),21,24,29,33,158,181,184,185,
188,189,192,193,196,199,205,206,212,217,218,
231,236,262,264,266,271
Twinproducts158,159,181,184,188,189,193,196,
205,271
V
Vågan21,35,234
Veitsler20,231
Veøy21
Verticallink44,45,47,48,68,72,76,105,106,107,
269
Vetrlidsalmenningen(1991/92)BRM342,site3042,
50,95,96,98,104,106,111,128,130,131,145,147,
148,150,153,154,156,168,209,227,232,270
Vetrlidsalmenningen2,Kjøttbasaren(1996and1997)
(BRM490),site2995,104,138,153
Visitor32,34,48,50,200,201,203,221,229,236
VisualImpactMethod43
W
Warpingpaddle173,175
Wastetype2(otherleatherworking)163,164,165
Wastetype3(shoemaking)163,164,165
Weight/balancearm168
WeightT/F(fortextileproductionorfishing)31,35,
147,158,167,168,172-176,206,207,209,210,213,
215,223
Wesenbergsmauet(1989)BRM297,site1984,104
Whale/walrusboneoffcut160,162
Windingpin173,175
Wiredrawer159-162
Women218,219,272
Woodworking157,158,170,171,180,196,199,203,
204,232,238
Ø
Øvregaten/Finnegårdsgaten(1979and1980)BRM86,
site3197,104,130
Øvregaten39(1981)BRM94,site2084-87,104,110,
111,148,150,153
ØvreDreggsalmenningen(1989)BRM298,site767-70,
80,104,105,109,110,127,130,132,145,178
ØysteinMagnusson’shallatHolmen,site558,104,150,
152,225,228,233,239
ØysteinMagnusson(EysteinMagnusson)(1103-1123)
21,43,57,58,99,103,104,150,152,225,228,233,
235,237,239
299