Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation

Tina Milavec

U radu je prezentirano istraživanje koje se provodi metodom terenskog pregleda. Istraživanje je usmjereno na dinarski krš, a provodi se na području Jasenica i Obrovca. Definirani su formacijski procesi i arheološki potencijal područja istraživanja. Posebna rasprava posvećena je metodologiji istraživanja i njezinoj primjerenosti za krški teren. Predstavljene su osnovne karakteristike arheoloških izvora na krškom terenu te je razrađena kategorizacija arheoloških izvora koja valorizira materijalna svojstva i procese u dinarskom kršu. Empirijski podatci prikupljeni istraživanjem prezentirani su prema kronološkom kriteriju.

Opera Instituti Archaeologici Sloveniae 23 Zbirka / Series Uredniki zbirke / Editors of the series OPERA INSTITUTI ARCHAEOLOGICI SLOVENIAE 23 Jana Horvat, Andrej Pleterski, Anton Velušček Slavko Ciglenečki Zvezdana Modrijan Tina Milavec Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Naselbinski ostanki in interpretacija Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation Recenzenta / Reviewed by Urednica / Editor Prevod / Translation Jezikovni pregled / Language Editors Tehnična ureditev / Technical Editor Oblikovanje ovitka / Front cover design Fotografije / Photographs Računalniški prelom / DTP Priprava slikovnega gradiva / Preparation of illustrations Izdala in založila / Published by Zanju / Represented by Glavni urednik / Editor-in-Chief Tisk / Printed by Izid knjige sta podprla / Published with the support of Fotografija na ovitku / Front Cover photo Jana Horvat, Mitja Guštin Zvezdana Modrijan Sunčan Patrick Stone, Tina Milavec Sonja Likar, Alan McConnell-Duff Mateja Belak Tamara Korošec Slavko Ciglenečki Mateja Belak Drago Valoh, Lucija Lavrenčič, Mateja Belak, Tamara Korošec Inštitut za arheologijo ZRC SAZU, Založba ZRC Oto Luthar, Jana Horvat Aleš Pogačnik Present d. o. o., Ljubljana Javna agencija za knjigo RS, Znanstvenoraziskovalni center SAZU Slavko Ciglenečki CIP - Kataložni zapis o publikaciji Narodna in univerzitetna knjižnica, Ljubljana 904(497.4Tonovcov grad)«652« CIGLENEČKI, Slavko, 1949Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu : naselbinski ostanki in interpretacija = Late antique fortified settlement Tonovcov grad near kobarid : settlement remains and interpretation / Slavko Ciglenečki, Zvezdana Modrijan, Tina Milavec; sodelavci Benjamin Štular, Saša Čaval in Ivan Šprajc = with contributions of Benjamin Štular, Saša Čaval and Ivan Šprajc; [prevod Sunčan Patrick Stone, Tina Milavec ; fotografije Slavko Ciglenečki]. - Ljubljana : Inštitut za arheologijo ZRC SAZU, Založba ZRC, 2011. - (Opera Instituti Archaeologici Sloveniae ; 23, ISSN 1408-5208) ISBN 978-961-254-331-0 1. Modrijan, Zvezdana 2. Milavec, Tina, 1979258819584 © 2011, ZRC SAZU, Inštitut za arheologijo, Založba ZRC Vse pravice pridržane. Noben del te knjige ne sme biti reproduciran, shranjen ali prepisan v kateri koli obliki oz. na kateri koli način, bodisi elektronsko, mehansko, s fotokopiranjem, snemanjem ali kako drugače, brez predhodnega pisnega dovoljenja lastnikov avtorskih pravic. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the publisher. Slavko Ciglenečki Zvezdana Modrijan Tina Milavec Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu Naselbinski ostanki in interpretacija Sodelavci: Benjamin Štular, Saša Čaval in Ivan Šprajc Late Antique fortified settlement Tonovcov grad near kobarid Settlement remains and interpretation With contributions of Benjamin Štular, Saša Čaval and Ivan Šprajc Ljubljana 2011 Vsebina Predgovor (Slavko CIGLENEČKI, Zvezdana MODRIJAN in Tina MILAVEC) ............................................................ 7 1. Uvod (Slavko CIGLENEČKI) . .......................................................................................................................................... 9 1.1 Geografski oris ........................................................................................................................................................... 11 1.2 Ime in izročilo ............................................................................................................................................................ 15 1.3 Opis najdišča . ............................................................................................................................................................. 17 1.4 Zgodovina raziskav .................................................................................................................................................... 25 1.5 Posočje v poznoantičnem času . ............................................................................................................................... 33 1.6 Mreža poti (Benjamin Štular) . ........................................................................................................................... 53 2. Terenski izvid (Zvezdana MODRIJAN, Slavko CIGLENEČKI in Tina MILAVEC) ............................................... 65 2.1 Metodologija dela . ..................................................................................................................................................... 67 2.2 Kronologija in faze ..................................................................................................................................................... 69 2.3 Stavba 1 z okolico ....................................................................................................................................................... 73 2.4 Stavbi 2 in 3 . ............................................................................................................................................................... 97 2.5 Sklop cerkva .............................................................................................................................................................. 111 2.6 Cisterna (vodni zbiralnik) . ..................................................................................................................................... 155 3. Stavbni razvoj in kronologija (Zvezdana MODRIJAN in Tina MILAVEC)............................................................ 163 3.1 Stavba 1 z okolico ..................................................................................................................................................... 165 3.2 Stavbi 2 in 3 . ............................................................................................................................................................. 193 3.3 Sklop cerkva .............................................................................................................................................................. 201 3.4 Cisterna (vodni zbiralnik)........................................................................................................................................ 211 4. Primerjave (Slavko CIGLENEČKI) .............................................................................................................................. 215 4.1 Stavba 1 . .................................................................................................................................................................... 217 4.2 Interpretacija cerkvenega sklopa . .......................................................................................................................... 225 4.3 Astronomska orientacija cerkva na Tonovcovem gradu (Saša Čaval in Ivan Šprajc) . ............................ 247 5. Vloga in pomen naselbine Tonovcov grad (Slavko CIGLENEČKI) . ....................................................................... 257 5.1 Utrdba Tonovcov grad - pomemben člen poznorimske obrambe Italije . ........................................................ 259 5.2 Tonovcov grad v poznoantični poselitveni sliki vzhodnoalpskega in zahodnobalkanskega prostora ......... 273 6. Sklep (Slavko CIGLENEČKI, Zvezdana MODRIJAN in Tina MILAVEC) ............................................................ 289 7. Literatura (Zvezdana MODRIJAN)............................................................................................................................... 295 contents Foreword (Slavko CIGLENEČKI, Zvezdana MODRIJAN and Tina MILAVEC) .......................................................... 7 1. Introduction (Slavko CIGLENEČKI) . ............................................................................................................................. 9 1.1 Geographic description ............................................................................................................................................. 11 1.2 Name and tradition . .................................................................................................................................................. 15 1.3 Site description ........................................................................................................................................................... 17 1.4 Research history ......................................................................................................................................................... 25 1.5 The Posočje area during Late Antiquity . ........................................................................................................................ 33 1.6 Path network (Benjamin Štular) ........................................................................................................................ 53 2. Field report (Zvezdana MODRIJAN, Slavko CIGLENEČKI and Tina MILAVEC) ................................................ 65 2.1 Methodology . ............................................................................................................................................................. 67 2.2 Chronology and phases ............................................................................................................................................. 69 2.3 Building 1 and its surroundings . ............................................................................................................................. 73 2.4 Buildings 2 and 3 . ..................................................................................................................................................... 97 2.5 The ecclesiastical complex ...................................................................................................................................... 111 2.6 Water cistern (reservoir).......................................................................................................................................... 155 3. Building development and chronology (Zvezdana MODRIJAN and Tina MILAVEC)......................................... 163 3.1 Building 1 and its surroundings . ........................................................................................................................... 165 3.2 Buildings 2 and 3 . .................................................................................................................................................... 193 3.3 The ecclesiastical complex ...................................................................................................................................... 201 3.4 Water cistern (reservoir) ......................................................................................................................................... 211 4. Comparisons (Slavko CIGLENEČKI) .......................................................................................................................... 215 4.1 Building 1 .................................................................................................................................................................. 217 4.2 Interpretation of the ecclesiastical complex ........................................................................................................ 225 4.3 Astronomical orientation of churches at Tonovcov grad (Saša Čaval and Ivan Šprajc) ......................... 247 5. The role and importance of the settlement Tonovcev grad (Slavko CIGLENEČKI).............................................. 257 5.1 The fort at Tonovcov grad – an important part of the Late Roman defence system of Italy . ........................ 259 5.2 Tonovcov grad in the broader Late Antique settlement pattern of the East Alpine and Western Balkans .............................................................................................................................................................. 273 6. Conclusion (Slavko CIGLENEČKI, Zvezdana MODRIJAN and Tina MILAVEC) . ............................................. 289 7. Bibliography (Zvezdana MODRIJAN) ......................................................................................................................... 295 predgovor foreword Splet okoliščin je pripeljal v zadnjem desetletju prejšnjega stoletja do odkritja pomembne poznoantične naselbine Tonovcov grad pri Kobaridu, v Zgornjem Posočju, v tem prelepem koščku slovenske dežele. Takrat so bile v Sloveniji delno že raziskane prve postojanke iz obdobja prehoda antike v srednji vek, ki so odstrle vpogled v čas obstoja poslednjih bivališč romaniziranih staroselcev. Žal pa jih je bila večina že v preteklosti delno uničena ali slabo raziskana, zato se je zdel Tonovcov grad s svojo kompleksno podobo in izjemno ohranjenostjo pravi izziv za raziskovalno potrditev nakazane, a slabo poznane problematike prehoda antike v srednji vek. Prva sondiranja na najdišču leta 1993 so sprožila zamisel o sistematični raziskavi, saj so tudi kobariški domačini pokazali izredno zanimanje in posluh za naše delo. Raziskava najdišča, ki smo jo vodili sodelavci Inštituta za arheologijo Znanstvenoraziskovalnega centra SAZU iz Ljubljane, se je odlično vključila v sočasne raziskave, ki potekajo tudi v bližnjih območjih Avstrije in Italije: razkrivajo slabo poznano preživetje staroselcev in njihove prve stike z manjšimi skupinami Germanov kot tudi poznejšo slovansko naselitev. Sodobne arheološke raziskave zahtevajo veliko sredstev, saj je znanstveno delo vedno bolj kompleksno in uporablja metode številnih drugih znanstvenih disciplin. Da bi uspešno opravili raziskave in konservacijo nadpovprečno ohranjenih ostankov arhitekture ter jih ustrezno predstavili na terenu, je bila potrebna pomoč številnih institucij in ljudi. Veliko podporo smo imeli v predstojnikih inštituta Janezu Dularju in pozneje Jani Horvat, ki sta omogočila sodelovanje številnih članov inštituta na terenu in velikokrat pomagala pri reševanju nelahkih finančnih težav. Omeniti je treba vsestransko pomoč, ki smo jo dobili od načelnika Upravne enote Tolmin Zdravka Likarja in direktorja Kobariškega muzeja Jožeta Šerbca, ki sta s številnimi spodbudami in pozitivno energijo pripomogla k uspešnemu začetku in razvoju raziskav. Veliko je prispevala Občina Kobarid, predvsem z odkupi parcel in z organizacijo javnih del, za kar gre zahvala županoma Pavlu Gregorčiču in Robertu Kavčiču. Začetne raziskave je omogočilo Ministrstvo za znanost, pomemben pa je bil predvsem prispevek In the last decade of the last century, the turn of events led to the discovery of the important Late Antique settlement of Tonovcov grad near Kobarid in the beautiful Upper Soča Valley. At that time some posts from the period of transition between the Antiquity into the Middle Ages in Slovenia had already been partly researched. They offered an insight into the time of the existence of the last settlements of the autochthonous Romanized people, but unfortunately most of them had either been partly destroyed or poorly investigated in the past. Therefore Tonovcov grad with its complexity and extraordinary state of preservation appeared as a great challenge which could help us solve the indicated but poorly known issues of the transition from the Antiquity to the Middle Ages. The first trial trenches on the site in 1993 led us to consider systematic research, as the locals of Kobarid showed great interest in our work. The excavations, led by the members of the Institute of Archaeology at the Scientific Research Centre of SAZU in Ljubljana, fit very well into the contemporary investigations of the neighbouring areas of Austria and Italy; they uncover the poorly known survival of the autochthonous inhabitants and their first contacts with the groups of Germanic and later Slavic newcomers. Modern archaeological research is very costly as scientific work grows in complexity and uses the methods of various other scientific disciplines. To be able to finish the research and conservation of the extraordinarily well preserved architectural remains and suitably present them on the site, the help of many institutions and people was necessary. We were given great support by the Heads of the Institute: Janez Dular and later Jana Horvat secured us the help of many members of the Institute during fieldwork and often provided assistance in difficult financial situations. We are most grateful to the chief of the Administrative unit of Tolmin, Zdravko Likar and the director of the Kobariški muzej in Kobarid, Jože Šerbec. Their encouragement and positive energy added considerably to the successful beginning and continuation of our research. The municipality of Kobarid contributed greatly by buying off the land and organising public works, for which we wish to thank the mayors Pavel Gregorčič and Robert Kavčič. The initial research was made possible by 7 predgovor foreword the Ministry of Science; an important contribution was also made by the Ministry of Culture. We are also grateful to the then director of the Directorate for Cultural Heritage, Stane Mrvič, and the secretary of the Ministry of Culture, Silvester Gabršček. The excavations included archaeologists Lucija Lavrenčič, Andreja Dolenc Vičič, Primož Pavlin, Janez Dirjec, Anton Velušček, Dragan Božič, Srečko Firšt and (then) archaeology students Veronika Maček, Matija Črešnar, Lucija Šoberl, Dunja Černic, Bojana Rozman, Mateja Ravnik, Miha Mlinar, Rok Klasinc, Barbara Nadbath, Nataša Gomilšček, Bernarda Županek, Samo Hvalec, Samo Sankovič and Januš Jerončič. Technical work in the field was done by Andreja Dolenc Vičič, Lucija Lavrenčič, Dragica Knific Lunder and Tamara Korošec; the latter two also made drawings of the excavated small finds. The preparations for publishing the site as a monograph began already during the excavations. We invited several colleagues to cooperate and very soon it was evident that the number of contributions will exceed the limits of a single volume. Therefore the descriptions of the architecture, stratigraphic situations and the interpretation of the site are included in the first volume (Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation), while the analyses of the small finds as well as the anthropological and zoological remains form the second one (Late Antique fortified settlement Tonovcov grad near Kobarid. Finds). Both volumes are interconnected and form an integral whole. Jana Horvat and Andreja Dolenc Vičič carefully read the text of the monographs and suggested improvements. In the volumes we are presenting the results of the archaeological research. However, in order to be able to preserve the excavated structures a group of highly trained experts and workers who performed the conservation work was necessary. The site belongs to the area covered by the ZVNKD Gorica, and the responsible conservator Nada Osmuk carefully supervised the conservation works and operationalised the resolutions of the conservation commission. After her retirement in 2007 she was succeeded by Patricija Bratina. The team of the Restoration centre of Slovenia carried out the conservation and presentation of the architecture remains under the supervision of France Vardijan (†) and later Jernej Hudolin. Of course, none of the above would have been possible without the contractors who were logistically supported and organised by the colleagues from the Kobariški muzej, among whom Jože Šerbec must be especially mentioned for his untiring help and the organisation and supervision of the works. Many metal finds were excellently treated and preserved by Jana Šubic Prislan in the Goriški muzej in Nova Gorica. Most finds are kept in the Tolminski muzej in Tolmin, where the site is already presented within the permanent exhibition. To all the above mentioned, and many others who contributed to the research and later conservation and presentation, we are sincerely grateful. Ministrstva za kulturo, ki je zagotovilo finančna sredstva in vključilo raziskave v akcijo “kulturni tolar”. Posebna zahvala velja tudi tedanjemu direktorju Direktorata za kulturno dediščino Stanetu Mrviču in sekretarju Ministrstva za kulturo Silvestru Gabrščku, ki sta vseskozi podpirala naša prizadevanja in se veselila napredka del. Izkopavanj so se udeležili arheologi Lucija Lavrenčič, Andreja Dolenc Vičič, Primož Pavlin, Janez Dirjec, Anton Velušček, Dragan Božič, Srečko Firšt ter takratni študentje arheologije Veronika Maček, Matija Črešnar, Lucija Šoberl, Dunja Černic, Bojana Rozman, Mateja Ravnik, Miha Mlinar, Rok Klasinc, Barbara Nadbath, Nataša Gomilšček, Bernarda Županek, Samo Hvalec, Samo Sankovič in Januš Jerončič. Tehnično delo so na terenu opravljale Andreja Dolenc Vičič, Lucija Lavrenčič, Dragica Knific Lunder in Tamara Korošec, zadnji dve sta izkopano drobno arheološko gradivo tudi natančno izrisali. Priprave za monografsko objavo najdišča so se začele že med samim izkopavanjem, k sodelovanju pa smo povabili številne sodelavce. Že med delom se je pokazalo, da bo prispevkov preveč za eno knjigo, zato smo opise arhitekture, stratigrafske situacije in izvrednotenje najdišča uvrstili v prvo knjigo (Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Naselbinski ostanki in interpretacija), obdelavo drobnih najdb in antropoloških in zooloških ostankov pa v drugo (Poznoantična utrjena naselbina Tonovcov grad pri Kobaridu. Najdbe). Knjigi sta med seboj povezani in tvorita celoto. Besedilo monografij sta pozorno prebrali in predlagali koristne izboljšave Jana Horvat in Andreja Dolenc Vičič. V knjigah so predstavljeni rezultati arheoloških raziskav. Da pa smo lahko zavarovali izkopane objekte, je bila potrebna vrhunsko usposobljena skupina strokovnjakov in delavcev, ki so izvajali konservatorska dela. Najdišče sodi v območje ZVNKD Gorica, zato je odgovorna konservatorka Nada Osmuk, pozorno nadzirala konservatorsko delo in operacionalizirala sklepe konservatorske komisije. Po njeni upokojitvi leta 2007 jo je nadomestila Patricija Bratina. Ekipa Restavratorskega centra Slovenije pod vodstvom Franceta Vardijana (†) in kasneje Jerneja Hudolina je izvajala strokovna dela konservacije in prezentacije arhitekturnih ostankov. Seveda pa vsega omenjenega ne bi bilo brez izvajalcev, ki so jih logistično podpirali in organizacijsko vodili sodelavci muzeja, med katerimi moramo posebej izpostaviti direktorja Jožeta Šerbca, ki je neutrudno pomagal, organiziral in bedel nad potekom del. Številne kovinske najdbe je v Goriškem muzeju odlično konservirala Jana Šubic Prislan. Večino najdb hrani Tolminski muzej, kjer je najdišče že predstavljeno v sklopu stalne zbirke Naplavine obsoške zgodovine. Vsem omenjenim in številnim drugim, ki so po svoje prispevali k raziskavam in kasnejši konservaciji in prezentaciji najdišča, iskrena hvala. 8 1. Uvod 1. Introduction 1.1 Geografski oris 1.1 Geographic description 1.2 Ime in izročilo 1.2 Name and tradition 1.3 Opis najdišča 1.3 Site description 1.4 Zgodovina raziskav 1.4 Research history 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity 1.6 Mreža poti (Benjamin Štular) 1.6 Path network (Benjamin Štular) 9 1.1 Geografski oris 1.1 Geographic description Poznoantična višinska utrjena naselbina Tonovcov grad leži v goratem predelu zahodne Slovenije, ob reki Soči severno od Kobarida (sl. 1.1). Zgornje Posočje, ki sega od izvira reke Soče pod najvišjimi vrhovi Julijskih Alp do Mosta na Soči oziroma Tolmina, spada med najslikovitejše slovenske pokrajine. Soča in njeni pritoki so tu vrezali globoke struge v večinoma karbonatno podlago. Prevladuje dachsteinski The Late Antique fortified hilltop settlement of Tonovcov grad is situated in the mountainous part of the western Slovenia by the river Soča to the north of Kobarid (Fig. 1.1). The landscape of the Upper Soča Valley, which extends from the spring of the river Soča under the highest peaks of the Alps to Most na Soči or Tolmin, is one of the most picturesque parts of Slovenia. The Soča river and its tributaries have carved deep river Sl. 1.1: Karta obravnavanega prostora z glavnimi poznoantičnimi najdišči. Fig. 1.1: Map of the discussed area with the main Late Antique sites. 11 1.1 geografski oris 1.1 geographic description apnenec, iz katerega je zgrajena večina Julijskih Alp. Vmes so pasovi dolomita, rdečega laporja in fliša, na dnu doline pa prevladujejo aluvialni nanosi. Soča do naselja Žaga teče v smeri prečnih dinarskih prelomov, tam pa se obrne v jugozahodno dinarsko smer ter teče ob robu krnskega pokrova. Dolina se močno zoži povsod, kjer reka prečka dinarsko slemenitev, npr. pri Klužah in pri Žagi nad Kobaridom. Pod Kobaridom se dolina razširi (sl. 1.2, 1.3). Med Kobaridom in Mostom na Soči jo je izoblikoval ledenik beds into the predominantly carbonate base. Dachstein limestone – which most of the Julian Alps consist of – prevails. In between we can find strips of dolomite, red marl and flysch, while the bottom of the valley is covered mostly by alluvial deposits. As far as the settlement of Žaga the Soča river runs parallel to the transverse Dinaric fractures, then it turns southwest and runs in the ‘Dinaric’ direction along the edge of the mountains of Krn. The valley narrows down whenever the river crosses a Dinaric ridge, e.g. at Kluže, or at Žaga above Kobarid. Sl. 1.2: Kobarid s Polovnikom in Kaninom v ozadju. Nad kostnico na Gradiču je Tonovcov grad. Pogled z juga. Fig. 1.2: Kobarid with Mt. Polovnik and Mt. Kanin in the background. Tonovcov grad stands above the ossuary on Gradič. 12 1.1 geografski oris 1.1 geographic description na manj odporni geološki podlagi. Sledi ledeniškega delovanja so vidne čelne morene pri Mostu na Soči. Dno doline predstavljajo ledeniško-rečne prodne terase, ki jih na zahodu omejujejo predalpski vrhovi Stol, Mija, Matajur in Kolovrat (Kunaver 1998, 54–70; Buser et al. 1989, 195–203). V Zgornjem Posočju se meša več fitogeografskih območij: alpsko, predalpsko in submediteransko-pred alpsko. V dolinah prevladujejo bukovi gozdovi, pogosto pomešani s toploljubnimi listavci (črni gaber, mali jesen). Smreka raste predvsem na mestih kotanjastih površinskih oblik z zastajajočim hladnim zrakom. Višje na gorskih pobočjih se bukvi pridruži macesen, še višje pa ruševje (Dakskobler 1995, 167; Zupančič, Wraber 1989, 118–120). Delež gozdov v Zgornjem Posočju se v zadnjem stoletju povečuje predvsem zaradi zaraščanja pašniških površin. Arheološko najdišče Tonovcov grad je nizka (416 m) skalna gmota, ki je zadnji odrastek masiva 772 m visoke Babe v pogorju Kobariškega Stola (1300 m). Hrib leži na desnem bregu Soče tik nad reko, tako da popolnoma zapira prehod skozi sotesko. Vrh Tonovcovega gradu sega približno 200 m nad Sočo. Soteska je težko prehodna tudi na levem bregu, saj se tu nad Sočo dvigujejo pobočja Drežniške planote in Krna (sl. 1.2). Once it passes Kobarid the valley stretches out (Figs. 1.2, 1.3). Between Kobarid and Most na Soči the valley was formed by a glacier that slid across a less resistant geological base. The traces of the glacier can be seen in the lateral moraines at Most na Soči. The bottom of the valley is represented by terraces formed by glaciofluvial gravel. In the west these terraces are bordered by the pre-alpine peaks of Stol, Mija, Matajur and Kolovrat (Kunaver 1998, 54-70; Buser et al. 1989, 195-203). A number of different phytogeographic landscapes meet in the Upper Soča Valley: Alpine, Pre-Alpine and Sub-Mediterranean-Pre-Alpine. Beech forests, often mixed with thermophilous deciduous trees (Hop Hornbeam, Manna Ash) predominate in the valleys. Spruce trees grow mainly in surface depressions where the air remains cold. Higher on the mountain slopes larch trees start appearing amongst the beech trees, and even higher up the dwarf pine (Dakskobler 1995, 167; Zupančič, Wraber 1989, 118-120). As a consequence of the grazing lands being abandoned and overgrown over the last century, the share of forests in the Upper Soča Valley has been on the increase. The archaeological site of Tonovcov grad is a low (416 m) rock formation, the last part of the 772 m high massif of Baba located in the mountain range of the Sl. 1.3: Položaj Tonovcovega gradu v soteski za Kobaridom tik nad Sočo. Pogled s severovzhoda. Fig. 1.3: The position of Tonovcov grad in the gorge behind Kobarid, above the Soča river. A view from the northeast. 13 1.1 geografski oris 1.1 geographic description Sl. 1.4: Tonovcov grad. Pogled na najdišče z južne strani. Fig. 1.4: Tonovcov grad. A view from the south. Pobočja hriba, ki se z vseh strani strmo dvigujejo proti vrhu, so danes večinoma prerasla z gozdom, v katerem prevladujeta bukev in gaber (sl. 1.3, 1.4). Kobariški Stol (1300 m). The hill lies on the right bank of the river Soča, right above the river, so that it completely blocks the way through the gorge. The top of Tonovcov grad stands approximately 200 m above the river. On the left bank where the slopes of the Drežnica plateau and Krn rise above the Soča river the gorge is also difficult to pass (Fig. 1.2). The hill slopes that rise steeply towards the peak are nowadays mainly overgrown by forests, predominantly beech and hornbeam (Figs. 1.3, 1.4). 14 1.2 Ime in izročilo 1.2 Name and tradition Poleg v literaturi uveljavljenega in med ljudmi najbolje poznanega imena Tonovcov grad (domačini izgovarjajo Tonecov oziroma Tu n cov) je poznan tudi toponim Cekinov grad, ki je – kot zgovorno kaže navedba v zapuščini notarja Miroslava Premrova – odraz najdbe (najdb?) zlatnikov v preteklosti.1 Na nekaterih specialnih kartah je označen tudi kot Stari grad (npr. Atlas Slovenije 1986, 77). Prevladuje prvo ime, ki smo ga tako, kot ga je v prvi tiskani omembi zapisal zgodovinar Simon Rutar, sprejeli tudi arheologi, saj povsem jasno razlikuje to arheološko najdišče od številnih “Cekinovih” in “Starih” gradov po Sloveniji. Tudi domačini danes obeh zadnjih imen ne uporabljajo, oziroma zanje v večini ne vedo več.2 Da je imel Tonovcov grad v zavesti okoličanov pomembno mesto, opozarja bogato izročilo: razširjena je pripoved o gradu na njem (glej označbo Stari grad na specialnih zemljevidih), o njegovih zakladih in roparskih vitezih, ki so ropali potnike na cesti pod gradom. V pesniško oblikovani pripovedi o Livškem jezeru je Anton Klodič - Sabladoski zapustil celo dokaj podroben opis roparskega gradu nad Sočo, kjer omenja močno obzidje in štiri stolpe (Klodič - Sabladoski 1912). Pripoved o treh bratih, roparskih vitezih, ki so prežali na popotnike in tovornike, pa je zapisana v zbirki pripovedk (Dolenc 1992, 159). V ustnem izročilu so znane tudi zgodbe o prostorni jami pod najdiščem, na kar se morda nanaša izročilo o “treh tinah zlata v tretji kleti” pod Tonovcovim gradom.3 Rimsko cesto pod Tonovcovim gradom omenja že Simon Rutar (Rutar 1890, 129). Cesta naj bi bila speljana čez hribček Sv. Antona in od tam vodila polagoma navzdol v globel za Tonovcovim gradom, kjer je domneval obstoj srednjeveškega gradiča Pottenstein. Tu naj bi se še dobro videl cestni tlak in kolesnice. Njen potek naj bi se kmalu za Tonovcovim gradom združil z današnjo Apart from the name Tonovcov grad (pronounced by the locals Tonecov) which is used in literature and best known amongst the locals, the settlement is also known as Cekinov grad (Sequin castle), which is – as indicated by a reference in the legacy of the notary Miroslav Premrov – a result of the find (finds?) of gold coins in the past.1 On some survey maps the site is also marked under the name of Stari grad (Old Castle, e.g. Atlas Slovenije 1986, 77). The first name is the most common, and archaeologists have adopted it in the way it was used by historian Simon Rutar when it first appeared in print, for it clearly separates this archaeological site from the numerous ‘Gold’ and ‘Old’ castles around Slovenia. The locals have stopped using the latter two names, and in most cases they do not even know about their existence.2 The rich oral tradition indicates that Tonovcov grad always held an important position in the consciousness of the locals: there is the story of a castle on the hill (see the name Stari grad on survey maps), and the stories about treasures and robber knights who robbed the travellers on the road beneath the castle. In the poem by Anton Klodič - Sabladoski about the Livek lake the author even left behind quite a detailed description of the robber castle above the Soča river, in which he mentions strong fortification walls and four towers (Klodič - Sabladoski 1912). A collection of stories (Dolenc 1992, 159) includes A tale of three brothers – robber knights – who wait for travellers below the castle. The oral tradition also includes stories describing a large cave under the castle. Perhaps they are the source of the ‘three units of gold in the third basement’ under Tonovcov grad.3 The Roman road below Tonovcov grad was first mentioned by Simon Rutar (Rutar 1890, 129). Supposedly it led over the small hill of Sv. Anton from where it descended slowly into the hollow behind Tonovcov grad, 1 Zapuščina notarja Premrova. Hrani arhiv Narodnega muzeja Slovenije v Ljubljani. 2 Pismo Meri Koren z dne 5. 6. 1993. Hrani Inštitut za arheologijo ZRC SAZU. 3 Podatek v pismu Meri Koren z dne 27. 6. 1994. Hrani Inštitut za arheologijo ZRC SAZU. 1 The legacy of notary Premrov. Kept in the archive of the Narodni muzej Slovenije in Ljubljana. 2 Letter from Meri Koren, dated 5th June 1993. Kept at the Institute of Archaeology at ZRC SAZU. 3 Information from the letter by Meri Koren, dated 27th June 1994. Kept at the Institute of Archaeology at ZRC SAZU. 15 1.2 Ime in izročilo 1.2 Name and tradition cesto proti Trnovem. Arheološke raziskave ceste še niso bile opravljene, je pa glede na konfiguracijo terena zelo verjetno tekla po trasi, ki jo je opisal že Rutar. Nedvom no cesta zaradi ugodne lege na sicer zahtevnem terenu utegne biti naslednica v večjem delu starih poti, če že ne “prave” rimske ceste (glej tudi pogl. 1.6). Izročilo o “rimski cesti”, ki je vodila od Gradiča mimo Tonovcovega gradu naprej proti Trnovem je še vedno živo. Domačin Zdravko Likar se spominja tudi večjega kamnitega valja, ki ga je imenoval “miljnik” in naj bi še pred desetletji ležal ob cesti v predelu takoj za Gradičem. Kamna danes na terenu ni več mogoče izslediti. where he assumed the medieval castle of Pottenstein existed. The road pavement and the wheel ruts were supposed to be still visible. After it passed Tonovcov grad it supposedly joined the present-day road towards Trnovo. So far archaeological research of the road has not been carried out, but according to the configuration of the terrain it is highly likely that it did run along the route described by Rutar. Regarding its favourable position on an otherwise difficult terrain it is probable that the major part of the current road runs along the old routes if not the ‘real’ Roman road (see also chapter 1.6). The oral tradition about the ‘Roman road’ that lead from Gradič past Tonovcov grad and towards Trnovo is still alive. The local Zdravko Likar remembers that a large cylindrical stone that he took for a milestone lay at the roadside just behind Gradič even a few decades ago. Nowadays the stone could not be located anymore. 16 1.3 Opis najdišča 1.3 Site description Hrib Tonovcov grad z večplastnim arheološkim najdiščem na vrhu meri v dolžino približno tristo metrov. Za naselbino je bil uporabljen le najvišji, naravno najugodneje zavarovan zahodni del hriba, medtem ko je vzhodna polovica v zgornjem delu nekoliko zložnejša in se šele v spodnjem delu prevesi v strmo skalnato steno. Najlažji je dostop s severne strani. Sedanja pot je lepo speljana in se enakomerno vzpenja po severni strani po- The hill of Tonovcov grad and the multilayer archaeological site on its top measure approximately 300 meters in length. Only the highest, naturally best protected west side of the hill was used for the settlement. The east side is less sloping at the top and it only turns into a steep rock face in its lower part. The easiest access is from the north. The current path runs smoothly and climbs at an equal pace along the north slope. On a smal saddle it turns Sl. 1.5: Tonovcov grad z okolico. Lidarski posnetek (© ZRC SAZU). Označena današnja (domnevno poznoantična) pot, današnja cesta in reka Soča (skrajno desno). Fig. 1.5: Tonovcov grad and its surroundings on a lidar data visualization (© ZRC SAZU). The modern (supposed Late Antique) path and the modern road are marked. The Soča river is to the far right. 17 1.3 Opis najdišča 1.3 Site description Sl. 1.6: Tonovcov grad. Na površini strmega pobočja so vidni ostanki ruševine prečnega zidu. Fig. 1.6: Tonovcov grad. The remains of the transverse wall as seen on the surface of the steep slope. bočja. Na manjšem sedlu zavije proti vzhodu ter pripelje do manjšega, deloma umetno zravnanega prostora (sl. 1.5, 1.7). Tega je v poznoantičnem času branil prečni zid na zahodni strani severnega pobočja, katerega ruševino je mogoče slediti po strmini vse do vrha hriba, kjer se je priključil k obrambnemu zidu zgornjega dela naselbine (sl. 1.6, 1.7). Prečni zid je predstavljal prvo obrambno linijo na najlažje dostopni strani. Tako omejen prostor ima na severni strani dokaj ostro zaključen rob, ki utegne towards the east and leads to a smaller, partly artificially levelled area (Figs. 1.5, 1.7). In Late Antiquity this area was defended by a transverse wall on the west side of the northern slope, the ruins of which can be traced all the way to the top of the hill, where it joined the defensive wall of the upper part of the settlement (Figs. 1.6, 1.7). This transverse wall represented the first line of defence on the most accessible side. This limited area ends on the north side with a relatively sharp edge, which could be 18 1.3 Opis najdišča 1.3 Site description Sl. 1.7: Tonovcov grad. Načrt najdišča z raziskanimi (št. 1–5) in v reliefu površine nakazanimi stavbami (št. 6–31). Fig. 1.7: Tonovcov grad. Site plan of the settlement with the researched buildings (Nos. 1-5) and the buildings that can be seen in the surface (Nos. 6-31). hiding a defensive wall. On the edge traces of a building are visible, but without further research they cannot be reliably dated (possibly World War I remains?). The path rises gradually from the saddle towards the settlement. It was made in such a way that the eventual attackers would have to approach the defensive wall with their undefended right side exposed. This and the clearly recognisable gap in the otherwise solid line of buildings along the wall led us to assume that today’s path exactly follows the original line of approach. The settlement location was carefully selected: the majority of the buildings lie in the slightly lower part of the ridge, which rises gradually (by a few meters in the east and in the west). The terrain also rises gradually to the south, thus the core of the settlement was well protected from the winds on three sides (Fig. 1.8). The fortified settlement on the top of the hill covers above all the western part of the rock. The area is surrounded by steep slopes and the defensive wall and measures 150 x 90 m. The 80 cm thick defensive wall skrivati obrambni zid: na njegovem robu je videti sledove stavbe, ki je brez raziskav ni mogoče zanesljivo datirati (morda tudi ostanki iz časa prve svetovne vojne?). Pot se od sedla čez pobočje zlagoma dviga proti naselbini. Speljana je bila tako, da so morebitni napadalci prihajali po njej obrnjeni proti obrambnemu zidu z desno stranjo, ki ni bila zavarovana. Prav zaradi tako speljane poti in jasno razpoznavne vrzeli v sicer strnjeni vrsti stavb ob obzidju sklepamo, da današnja pot v celoti vodi po trasi prvotnega pristopa. Mesto naselbine je bilo izbrano izredno premišljeno: večji del stavb leži v nekoliko uleknjenem delu temena hriba, ki se na vzhodnem in zahodnem robu postopno dvigne za nekaj metrov. Teren se postopoma dviga tudi na južni strani, tako da je bilo jedro naselbine pred vetrovi odlično zaščiteno kar s treh strani (sl. 1.8). Utrjena naselbina na vrhu hriba obsega predvsem zahodni del skalne kope. S strminami in obzidjem obdani prostor meri 150 x 90 m. Obrambni zid debeline 80 cm je viden na vzhodni strani, kjer ločuje naselbino od 19 1.3 Opis najdišča 1.3 Site description Sl. 1.8: Tonovcov grad. Pogled s severozahodne strani na osrednji del naselbine z izkopano stavbo št. 1 v ospredju. Fig. 1.8: Tonovcov grad. A view of the central part of the settlement from the northwest, with the excavated building No. 1 in the foreground. nekoliko nižjega vzhodnega dela hriba. Brez izkopavanj ni mogoče zanesljivo reči, ali je bil zid postavljen na kakšnem zemljenem – morda prazgodovinskem – nasipu ali pa je na terenu vidni 1–3 m visok nasip le ruševina nekdaj veliko višjega obzidja (sl. 1.9). Nadaljevanje zidu proti jugu ni ohranjeno ali pa obzidje ni bilo zgrajeno povsem do roba južne skalne stene. Morda bi lahko tu domnevali drugi vhod v naselbino, saj je bila potrebna can be seen on the east side, where it separates the settlement from the slightly lower eastern part of the hill. Without excavations it is impossible to say whether the wall stood on a – maybe prehistoric – rampart or whether the 1-3 m high dike is merely the ruin of the once much higher wall (Fig. 1.9). Its continuation towards the south is not preserved or the wall was not built all the way to the edge of the south rock face. Perhaps this was the second Sl. 1.9: Tonovcov grad. Ruševinski nasip z delno vidnim obzidjem na vrhu in ruševino stolpa (št. 31) na vzhodni strani naselbine. Fig. 1.9: Tonovcov grad. A dike with partially visible wall at the top and the tower ruins (No. 31) on the east side of the settlement. 20 1.3 Opis najdišča 1.3 Site description Sl. 1.10: Tonovcov grad. Pogled z vzhodne strani na ruševino stavbe št. 28 po viharju pozimi leta 2008. Fig. 1.10: Tonovcov grad. A view of the ruins of building No. 28 soon after the storm in winter 2008, from the east. komunikacija s preostalim, manj zavarovanim delom hriba. Obzidje je slabše vidno na severni strani, saj so bile tu v večjem delu zgrajene manjše stavbe, tesno razvrščene druga poleg druge in prislonjene na obzidje. Ni mogoče zanesljivo razločiti, ali gre za sklenjeno vrsto stavb ali pa so bile te na posameznih mestih ločene z vmesnimi prostori. V notranjosti je bilo mogoče po temeljitem čiščenju podrasti opaziti ruševine stavb, katerih tlorisi so v površini tako dobro vidni, da je mogoče dokaj natančno rekonstruirati nekdanjo naselbino, pri večini stavb pa celo ločiti število prostorov in razbrati njihov razpored (sl. 1.7). Kot smo že omenili, so najpreprostejše stavbe, ki ležijo ob severnem obrambnem zidu. Izrazitejše so tiste znotraj zavarovanega območja. Izstopa zgradba, ki leži povsem blizu vhoda in nadzira vstop v naselbino: zavzemala je osrednji prostor v naselbini in hkrati ležala na dokaj ravnem terenu – pri izkopavanju smo jo označili kot stavbo 1 (sl. 1.7; glej pogl. 2.3). Dve triprostorni stavbi (objekta št. 22 in 26, sl. 1.7) ležita v neposredni bližini: pri eni od njiju (objekt št. 22) se zdi, da je bil osrednji prostor močneje grajen ali vsaj višji, saj kot ruševina izraziteje izstopa iz talnega reliefa. Še ena triprostorna stavba (v kolikor ne gre za tri samostojne objekte) je stala tik pod vrhom zahodnega zaključka naselbine, na manjši, lepo položeni terasi, vendar je entrance to the settlement as communication with the other, less protected part of the hill, was also necessary. The defensive wall is less clearly visible on the north side where small buildings were built close together and leaning upon the wall. It is difficult to distinguish whether this was a line of attached buildings or were they separated by empty spaces. After the undergrowth in the interior was cleared we discovered the ruins of buildings, the ground plans of which were so clearly visible on the surface that we could relatively precisely reconstruct the former settlement, and in most buildings we could even distinguish the number of rooms and their layout (Fig. 1.7). As already mentioned, the simplest buildings are positioned alongside the north defensive wall. The buildings within the protected area are more complex. Very conspicuous is a a building that lies very close to the entrance to the settlement and must have controlled it. This building occupied the central area within the settlement and stood on relatively level terrain – during the excavations it was named building 1 (Fig. 1.7; see chapter 2.3). Two three-room buildings (buildings No. 22 and 26, Fig. 1.7) can be found in the immediate vicinity. One of them (building No. 22) seems to have had a better built or at least a taller central space for its ruin is visibly higher than others. Another threeroom building (or perhaps three individual buildings) 21 1.3 Opis najdišča 1.3 Site description Sl. 1.11: Tonovcov grad. Pogled z južne strani na v celoti izkopan sklop zgodnjekrščanskih cerkva. Fig. 1.11: Tonovcov grad. A view of the three excavated Early Christian churches, from the south. bila prilagojena terenu, saj si prostori v njej ne sledijo v povsem ravni črti (objekt št. 14, sl. 1.7). V severozahodnem vogalu naselbine je moral stati neki pomembnejši objekt, saj je bilo prav to mesto eno izmed ključnih delov obrambe celotne naselbine. Vogal obzidja je še prav dobro viden, nista pa jasni izoblikovanost in velikost stavbe oziroma stolpa, ki je stal na tem mestu (objekt št. 18, sl. 1.7). Zaradi nekoliko nagnjene lege se je ruševina nedvomno večinoma posula po pobočju. Kot posebnost izpostavljamo ruševino stavbe s petimi prostori (objekt št. 28, sl. 1.7), ki so razporejeni po vzhodnem pobočju na dveh višinskih nivojih (sl. 1.10). Na nekaj metrov visokem skalnem platoju, ki se na južni strani navezuje na južno skalno steno naselbine, je bil že pred izkopavanjem dobro viden največji ruševinski sklop (objekt št. 4, sl. 1.7). To so sledovi zgodnjekrščanskih cerkva, ki smo jih kot take prepoznali že pred sistematičnim izkopom (glej pogl. 2.5). Izpostavljeno lego na robu naselbine ima tudi stavba (objekt št. 8, sl. 1.7), ki leži nekako v sredini južnega roba, tik nad prepadno steno. Glede na dolžino in orientacijo se sklada s cerkvenimi zgradbami na nekoliko višje ležečem platoju (sl. 1.12). Tik pod njenim vzhodnim zaključkom je v manjšem sedlu vidna enoprostorna stavba, ki je veliko nižja in nekoliko poglobljena (objekt št. 7, sl. 1.7). V njej domnevamo prej omenjeni večji stavbi pripadajoči vodni zbiralnik. stood right under the top of the west end of the settlement on a small, well positioned terrace. However it was still adjusted to the terrain, for the rooms were not built in a straight line (building No. 14, Fig. 1.7). An important building had to stand in the northwest corner of the settlement, for this spot was one of the key points in the defence of the entire settlement. The corner of the fortification wall is still clearly visible, but the size and shape of the building or tower that stood there is not entirely clear (building No. 18, Fig. 1.7). Due to the slightly sloping terrain it can be assumed that most of the ruin slid down the slope. An interesting feature among the house layouts is the ruin of a building with five rooms (building No. 28, Fig. 1.7) which are positioned on two different height levels along the east slope (Fig. 1.10). A slightly elevated rock plateau joins the rock face in the southern part of the settlement. There the largest complex of ruins was visible (building No. 4, Fig. 1.7) even before the excavations started. These were remains of Early Christian churches that were recognised as such even before systematic research (see chapter 2.5). Another building (building No. 8, Fig. 1.7) situated roughly in the middle of the south edge of the settlement, just above the precipice, was also very exposed. In length and orientation it resembles the church buildings on the plateau (Fig. 1.12). In a small saddle just under its east 22 1.3 Opis najdišča 1.3 Site description Sl. 1.12: Tonovcov grad. Visoko ohranjena ruševina stavbe št. 8 na robu naselbine (domnevno cerkev) in pred njo globlji prostor stavbe št. 7 (domnevni vodni zbiralnik). Fig. 1.12: Tonovcov grad. The high preserved ruins of building No. 8 at the edge of the settlement (assumed to be a church) and the depressed area of building No. 7 (presumably a water reservoir) in front of it. Še en vodni zbiralnik (objekt št. 5, sl. 1.7) je bil raziskan na najvišjem, vzhodnem platoju naselbine. Vmes med omenjenimi večjimi zgradbami je še nekaj manjših, najverjetneje enoprostornih, ki dopolnjujejo podobo naselbine. Po neurju, ki je pozimi 2008 izruvalo in polomilo nekaj dreves, se je ob čiščenju terena pokazala še ena tlorisna zasnova z več prostori: leži povsem na robu platoja z vodnim zbiralnikom in je bila na prvotnem načrtu nakazana le kot manjša dvoprostorna zgradba zahodno od zbiralnika (objekt št. 30). Njen tloris pa bo v celoti mogoče vrisati šele z nadaljnjimi raziskovanji. Nedvomno je bil njen pomen precejšen, saj gre za najvišje ležečo veliko stavbo z dokaj razčlenjeno tlorisno zasnovo. Na hribu so tudi sledovi iz prve in druge svetovne vojne. Poleg stopnic in bunkerja na vzhodnem pobočju hriba so na vrhu še jarki, kaverna in večji vkop iz druge svetovne vojne, ki so deloma poškodovali poznoantično plast. Nemški bunker so v celoti vgradili v poznoantični vodni zbiralnik (objekt št. 5) in tako izkoristili dokaj dobro ohranjeno antično strukturo (glej pogl. 2.6). end a single room building, much lower and slightly deepened, is visible (building No. 7, Fig. 1.7). Presumably this is a water reservoir that supplied the larger building. Another water reservoir (building No. 5, Fig. 1.7) was researched on the highest, eastern plateau. In between the mentioned larger buildings there are a few smaller, most likely single-room buildings that complete the layout of the settlement. After the terrain was cleared from the storm that toppled and snapped a number of trees in the winter of 2008 another building with several rooms was revealed. It is situated on the very edge of the plateau with the reservoir and was previously thought to be merely a smaller two-room building to the west of the water reservoir (building No. 30). Further research will be needed to reconstruct its ground plan. It was undoubtedly important, for is situated on the highest part of the settlement and has a relatively complex ground plan. The hill also shows traces from the two world wars. Apart from the stairs and bunker on the east slope of the hill, some ditches, a cavern and a large trench from World War II on the top of it partially damaged the Late Antique layer. Making use of the relatively well preserved Antique structures the German bunker was built within the Late Antique water reservoir (building No. 5; see chapter 2.6). 23 1.4 Zgodovina raziskav 1.4 Research history Posočje v celoti, posebej pa še njegov severni del, je bilo arheološko dolgo časa slabo znano. Pozornost raziskovalcev je bila usmerjena predvsem v veliko in izjemno pomembno prazgodovinsko postojanko v Mostu na Soči (Sv. Lucija), kjer so že v 19. st. izkopali grobišče, po drugi svetovni vojni pa so pri zaščitnih raziskavah v samem jedru kraja odkrili odlično ohranjene sledove prazgodovinskih stavb (pregled raziskav pri Mlinar 2002). Vzporedno s temi deli so potekala zgolj manjša zaščitna izkopavanja, opravljeni so bili redki topografski pregledi ter odkrite posamezne naključne najdbe arheoloških predmetov (ANSl 1975, 115–119; 124–127). Tudi ožje kobariško območje je bilo dolgo deležno enake usode raziskanosti kot celotno Posočje. Kljub izkopu druge največje halštatske nekropole na pobočjih in ravnici med Sočo in Gradičem že ob koncu 19. st. je velika večina tukajšnjega gradiva obležala neobjavljena v depojih, nekropoli pripadajoče gradišče pa je bilo povsem pozabljeno (ANSl 1975, 116). Večje zaščitne raziskave grobišča je tu leta 1979 opravil Drago Svoljšak (Svoljšak 1981). Šele z odkritjem antičnih bronastih plastik spodbujena zaščitna izkopavanja Zavoda za varstvo naravne in kulturne dediščine pod vodstvom Nade Osmuk na pobočju Gradiča so znova opozorila na velik arheološki pomen ravnice med Sočo in Nadižo, na izjemen prometni in strateški pomen tega prostora v prazgodovinskem in rimskem obdobju kot tudi na staro kultno središče na tem območju (Osmuk 1987a; 1997a). Poznoantične in zgodnjesrednjeveške najdbe so bile neznane, še tistih nekaj elementov, ki so kazali na obstoj najdišč, je bilo spregledanih ali pozabljenih. Delno je to vrzel leta 1991 zapolnil Matej Župančič, ki je opozoril na starejšo najdbo zlatega solida iz sredine 5. st. z Molide, in nakazal možnost obstoja poznoantične postojanke (Župančič 1991). Arheološko najdišče Tonovcov grad pa je imelo povsem samosvojo, nenavadno usodo in bilo kljub nekaterim zgodnejšim omembam in starejšim naključnim najdbam zelo pozno prepoznano kot eno izmed najcelovitejših večplastnih arheoloških najdišč z izjemno ohranjeno poznoantično naselbino. The archaeology of the Soča Valley, especially its north part, was for a long time very poorly known. The attention of the researchers was predominantly directed towards the large and extremely important prehistoric settlement at Most na Soči (Santa Lucia), where a cemetery was excavated as early as the 19th century. Rescue excavations after World War II revealed excellently preserved traces of prehistoric buildings in the very centre of the present-day settlement (for an overview of the various researches see Mlinar 2002). Apart from that there were only some minor protective excavations, rare surveys and individual archaeological finds (ANSl 1975, 115-119; 124-127). The Kobarid area suffered similar fate. Even though the second largest Hallstatt necropolis in Slovenia was discovered on the slopes between the Soča river and Gradič as early as at the end of the 19th century a great portion of the excavated material remained unpublished, and the hillfort that belonged to the necropolis was completely forgotten (ANSl 1975, 116). A protective excavation of the necropolis took place in 1979 and was led by Drago Svoljšak (Svoljšak 1981). It was only with the discovery of the Antique bronze statues that rescue excavations on the slope of Gradič were started by Nada Osmuk from the Institute for the Protection of the Natural and Cultural Heritage. This discovery brought back the attention to the great archaeological importance of the straights between the Soča and Nadiža rivers, its exceptional strategic and transport importance during the prehistoric and Roman period as well as to the old cult centre in this area (Osmuk 1987a; 1997a). Late Antique and Early Medieval finds were unknown, even the few elements that indicated the existence of sites, were overlooked or forgotten. This gap was partially filled in 1991 by Matej Župančič, who drew attention to an old find of a mid 5th century gold solidus from Molida, that indicated a possibility of the existence of a Late Antique post (Župančič 1991). The archaeological site of Tonovcov grad had a unique and unusual fate. Despite some early references and accidental finds it was very late to be recognised as one of the most complex multi-layered archaeologi25 1.4 Zgodovina raziskav 1.4 Research history Najdišče je bilo v arheološki literaturi prvič omenjeno leta 1882, ko ga je zgodovinar Simon Rutar, tolminski rojak, v svoji Zgodovini Tolminskega navedel kot možno arheološko lokacijo (Rutar 1882, 12; glej pogl. 1.2). Sam si ga sicer ni ogledal, navaja pa pripovedi, ki omenjajo razvaline gradu. Očitno tem pripovedim ni zaupal, saj nakazuje tudi možnost, da je “stari grad” dobil ime po obliki hriba, kar naj bi se pogosto dogajalo. V poročilu o prazgodovinskih in rimskih izkopavanjih na Slovenskem se je leta 1890 ponovno dotaknil tega najdišča (Rutar 1890, 129). Omenja ga v zvezi z rimsko cesto, ki naj bi tekla od hribčka Sv. Antona polagoma navzdol v jarek za “Tonovčevim gradom”. Pri tem domneva, da naj bi tu v srednjem veku stal gradič “Pottenstein”. Nekaj podatkov o Tonovcovem gradu se je ohranilo tudi v rokopisni zapuščini M. Premrova, notarja v Kobaridu, ki v svoji “Gedenkbuch der Pfarr Karfreit” z dne 24. 2. 1915 omenja kopanje na hribu “Tonovcev gradc”, kjer na bi pred pribl. 50–60 leti (torej okoli 1860) šest domačinov, z očetom Tonovcom vred, kopalo jamo, v kateri so zadeli na zidove.4 Omenja tudi govorico, po kateri naj bi fant iz Ladrij našel na krtini cekin. Kljub tem navedbam najdišče med strokovnjaki ni izzvalo večjega zanimanja, zato ga pri pregledu Kobarida in okolice ne upošteva niti temeljni pregled slovenskih arheoloških najdišč Arheološka najdišča Slovenije iz leta 1975 (ANSl 1975, 116). Leta 1985 je hrib Tonovcov grad obiskala arheologinja goriškega Zavoda za varstvo naravne in kulturne dediščine Nada Osmuk, ki je opazila pomembno strateško lego in odlično obrambno naravo najdišča (Osmuk 1985c). Opozorila je na sledove obrambnega okopa na jugovzhodni in južni strani, sledov kulturne plasti pa na površini ni ugotovila. Časovno opredelitev je zaradi pomanjkanja izpovednejših najdb (omenja le opeke) pustila odprto. Izrazila je mnenje, da je domnevna utrjena postojanka zapirala dostop v kobariško kotlino s severa ob cesti na desnem bregu Soče, in ugotovila, da ima postojanka ugodnejši strateški položaj od Gradiča. Na časovno uvrstitev utrjene naselbine pa so opozorile najdbe nekaterih izpovednih arheoloških predmetov iz vrhnje plasti Tonovcovega gradu, ki sta jih našla Žiga Šmit in Tomislav Drčar.5 Prve najdbe je leta cal sites with an exceptionally preserved Late Antique settlement. The settlement was first mentioned in archaeological literature in 1882, when the historian Simon Rutar from Tolmin, mentioned it as a possible archaeological location in his book Zgodovina Tolminskega [History of the Tolmin area] (Rutar 1882, 12; see chapter 1.2). He did not visit it himself, but writes about stories that mention castle ruins. He obviously did not find these stories reliable, for he also indicated the possibility that the ‘old castle’ obtained its name from the shape of the hill (which was apparently a rather common occurrence). He mentioned the site again in 1890 in his report on the Prehistoric and Roman excavations in Slovenia (Rutar 1890, 129): a Roman road was supposed to have ran from the hill of Sv. Anton down to the hollow behind ‘Tonovčev grad’ and the small castle of ‘Pottenstein’ was supposed to have stood on the hill in the Middle Ages. Some data on Tonovcov grad was preserved in the manuscripts of M. Premrov, a notary in Kobarid, who in his ‘Gedenkbuch der Pfarr Karfreit’ dated on 24th February 1915 mentions digging that took place on the hill ‘Tonovcev gradc’, when approximately 50-60 years ago (i.e. around 1860) six locals, including the owner (‘father Tonovc’), discovered walls while digging a pit.4 He also mentions a rumour according to which a boy from the village Ladra found a gold coin in a molehill. Regardless of this the site did not raise interest amongst experts, even the major overview of Slovenian archaeological sites Arheološka najdišča Slovenije from 1975 (ANSl 1975, 116) fails to mention it in the overview of Kobarid and its surroundings. In 1985 the hill Tonovcov grad was visited by Nada Osmuk, an archaeologist from the Gorica Institute for the Protection of National and Cultural Heritage. She noticed the important strategic position and the excellent defensive nature of the site (Osmuk 1985c). She drew attention to the traces of the rampart on the southeast and south side, however she failed to discover any traces of the cultural layer on the surface. Due to the lack of any revealing finds (she only mentions tegulae) she could not date the site. She noted that the assumed fortified post closed off the north entrance into the Kobarid basin on the right bank of the river Soča, and that it had a better strategic position than Gradič. The chronology of the fortified settlement was implied by some revealing artefacts discovered by Žiga Šmit and Tomislav Drčar in the topmost layer of Tonovcov grad.5 The first finds were gathered in 1991 by Ž. Šmit 4 Hrani arhiv Narodnega muzeja Slovenije v Ljubljani. Podatek mi je prijazno posredoval Silvo Torkar z Inštituta za slovenski jezik Frana Ramovša ZRC SAZU. 5 Takratne najdbe sestavljajo skupino predmetov, odkritih po celotni površini najdišča in po pobočjih zunaj njega, ki so bili izkopani v zgornji plasti, v ruši, in so, kot smo ugotovili pozneje pri sistematičnih izkopavanjih, ležali v plasti, ki je bila poškodovana z gostim koreninskim prepletom vegetacije, ki je v zadnjih sto letih zarasla prej neporaščeni Tonovcov grad. Ti predmeti so brez stratigrafskih kontekstov, ki jih v gozdni površinski plasti – pretežno v območjih med stavbami – tudi ni mogoče pričakovati. Izjema so le ostanki velike bronaste posode (Tonovcov grad. Najdbe, t. 50: 1), ki je bila najdena v ruševinski plasti narteksa južne cerkve, kot 4 Kept in the archive of the Narodni muzej Slovenije in Ljubljana. This information was kindly passed onto me by Silvo Torkar from the Fran Ramovš Institute of the Slovenian Language at ZRC SAZU. 5 The finds include a group of objects that were discovered across the entire surface of the site and along the slopes outside of it, all of which were discovered in the top turf layer, 26 1.4 Zgodovina raziskav 1.4 Research history 1991 pridobil Ž. Šmit z iskalcem kovin in jih 20. 9. 1991 prinesel na Inštitut za arheologijo ZRC SAZU. Sodelavci Inštituta smo se zato takoj odpravili na najdišče, da bi preverili mesto najdb. Po analogiji z že prej odkritim Gradcem pri Prapretnem (Ciglenečki 1975, 259–267) smo odkrili številne na površini v reliefu dobro vidne stavbne ostaline iz poznoantičnega obdobja. Datacijo stavb sta potrjevali predvsem gradnja zidov in sestava malte v zidu velike stavbe v bližini vhoda, ki je bil prekopan že pred davnimi leti – najverjetneje, kot omenja Premrov, so tu kopali lastnik Tonovc in njegovi tovariši. Po številnih terenskih obhodih najdišča in okolice smo spomladi 1992 pripravili tudi natančen geodetski posnetek, v katerega smo vrisali vse na površini vidne strukture in vnesli vanj mesta izpovednejših najdb, ki sta jih našla Ž. Šmit in pozneje T. Drčar.6 Najdišče je bilo v smeri S–J razdeljeno na kvadratno mrežo z velikostjo kvadrantov 4 x 4 m, obenem pa so Ž. Šmit, T. Drčar in Zoran Božič natančno preiskali površino. Pri tem smo ravnali skladno z navodili, ki jih je objavila spomeniška služba (Curk 1994/1995). Tako nam je uspelo vse do takrat odkrite najdbe vnesti v geodetski posnetek najdišča in okolice, s čimer smo omogočili poznejšo rekonstrukcijo eventualnih zaključenih sklopov. Ob tem je treba poudariti, da je bila velika večina tako pridobljenih arheoloških predmetov odkrita do globine 15 cm in v arheološko ne več intaktnih plasteh. Zaradi pomembnosti izjemno ohranjenih struktur poznoantične naselbine smo pripravili načrt raziskovanj, da bi odlično ohranjeno, a slabo znano naselbino čim prej vključili v sistematična raziskovanja poznoantičnega obdobja Slovenije. Glede na do tedaj znano tipološko klasifikacijo poznoantičnih višinskih postojank je prav ta naselbina kazala najkompleksnejši značaj in posrečeno združevala kar nekaj najbolj tipičnih elementov poznoantičnih utrdb: veliko in dobro ohranjeno naselbino, katere obrisi so bili dobro vidni na površini, pomembno strateško lego, na zgornjem platoju naselbine pa se je v reliefu nakazovalo veliko zgodnjekrščansko središče z več cerkvami. Nekoliko senzacionalističen opis (“slovenski Pompeji”) navdušenega obiskovalca raziskav na Tonovcovem gradu in ljubitelja arheologije Zorana Božiča v Primorskih novicah je sprožil veliko zanimanje za najdišče, hkrati pa izzval odgovor tedanjega kobariškega dekana Franca Rupnika (Božič 1993; Rupnik 1993). Čeprav so bile raziskave šele v začetni fazi, je zato moral slediti še odgovor pristojnega arheologa, da se “novi kobariški čudež” in “slovenski Pompeji” ne bi preveč neverodo- with the use of a metal detector and brought to the Institute of Archaeology at ZRC SAZU on 20. 9. 1991. With the co-workers from the Institute we immediately set off to the site in order to check the location of the finds. Comparing the surface shape of the site with the recently discovered Gradec near Prapretno (Ciglenečki 1975, 259-267) numerous building remains from the Late Antique period were found out. The dating of the buildings was confirmed by the construction of the walls and mortar composition on the wall of the big building close to the entrance. This building has been previously damaged by a robber trench – probably this was the location where the owner Tonovc and his companions had dug as mentioned by Premrov. In the spring of 1992, after numerous field inspections of the site and its surroundings we prepared a precise land survey record, into which we entered all structures that could be seen on the surface and the locations of the more revealing finds discovered by Ž. Šmit and later T. Drčar.6 The site was divided into a square grid running in the north south direction, with the quadrants measuring 4 x 4 m. Ž. Šmit, T. Drčar and Zoran Božič inspected the topsoil in great detail. They proceeded in accordance with the instructions published by the office for protection of cultural heritage (Curk 1994/1995). We managed to include all of the discovered finds in the survey map of the site and its surroundings, so as to provide a good basis for eventual later reconstructions. Here it should be emphasised that the majority of the artefacts obtained in this manner were discovered up to 15 cm deep in layers that were not archeologically intact. Because of the importance of the exceptionally well preserved structures of Late Antique settlement we prepared an excavation plan that would enable us to include the excellently preserved, but poorly known site into the systematic research of Late Antiquity in Slovenia as soon as possible. In the typology of the Late Antique hilltop posts this settlement showed an extremely complex character and combined some of the most typical elements of Late Antique forts: a large and well preserved settlement, clearly visible in the surface shape, a strategically that was, as we have ascertained during the systematic excavations, damaged by the roots of the dense vegetation which has overgrown the previously bare Tonovcov grad during the last one hundred years. These objects have no stratigraphical contexts, which is of course to be expected in a forest surface layer – mainly in the areas between the buildings. The exceptions are the remains of the large bronze vessel (Tonovcov grad. Finds, Pl. 50: 1) that was discovered in the destruction layer of the narthex of the south church as was established during the excavations in 1996. Together with the two finders we entered the precise locations of the objects into the geodetic map of the settlement, hoping that this would help explain the site during the later excavations. 6 The geodetic measurements were made by the land surveyor Bogo Žontar († 2003), who followed our excavations with great dilligence and enthusiasm. se je izkazalo pri izkopavanju leta 1996. Lego predmetov smo skupaj z najditeljema natančno vnesli v geodetski posnetek naselbine, da bi ob poznejšem izkopavanju pomagali pri tolmačenju najdiščnih sklopov. 6 Geodetsko izmero je opravil geodet Bogo Žontar († 2003), ki je z veliko prizadevnostjo in veseljem spremljal naša izkopavanja. 27 1.4 Zgodovina raziskav 1.4 Research history stojno zasidrali v zavesti slovenskih ljubiteljev antične zgodovine. Slavko Ciglenečki je na podlagi do tedaj znanih izsledkov pojasnil osnovne značilnosti novoodkrite poznoantične postojanke in tako odgovoril na dileme, ki so se o značaju najdišča porodile v časopisnih objavah (Ciglenečki 1993). Leta 1993 smo opravili prvi raziskovalni poseg v naselbini. Osredotočili smo se na saniranje že izkopane večje jame na zgornjem platoju naselbine (med osrednjo in južno cerkvijo), očistili že močno porušene robove izkopa in tako pridobili prvi uvid v kulturne plasti najdišča.7 Poleg ruševin gradbenih ostankov smo izkopali večje število odlomkov poznoantične keramike (grobe kuhinjske lončenine, amfor in afriške sigilate) in antične kritine (imbreksov in tegul). Prve podrobnejše strokovne rezultate preliminarnih raziskav smo objavili leta 1993 v reviji Kronika, kjer je bila tematska številka posvečena prav zgodovini Tolminskega (Ciglenečki 1994b). Nekoliko pozneje smo neznatno dopolnjeno verzijo prevedenega besedila objavili tudi za italijansko strokovno javnost (Ciglenečki 1994c). Začetne raziskave so tako okvirno določile razpon poselitev na naselbini vse od prazgodovinskih dob do srednjeveškega obdobja. Izkazalo se je, da se postojanka po velikosti, ohranjenosti arhitekture in kompleksnem značaju uvršča med najpomembnejše poznoantične naselbine v vzhodnoalpskem prostoru. Njena strateška lega in nadzor nad prehodom čez občutljivo območje tik pred vrati Italije pa sta opozorila tudi na verjeten vojaški značaj postojanke v poznorimskem obdobju in njegovo pomembno vlogo v sistemu poznoantične obrambe Italije. Zdelo se je, da je s tem naše delo na najdišču začasno končano. Vendar so zanimanje javnosti, velik interes kolegov in njihov odziv na prva poročila spodbudili tudi drugačna razmišljanja. Tako se je dilema, ali important position, and large Early Christian complex with a number of churches clearly showing in the relief on the upper plateau of the settlement. A somewhat sensational description (‘The Pompeii of Slovenia’) that was published in Primorske novice by Zoran Božič, an enthusiastic visitor of the excavations at Tonovcov grad and archaeology aficionado, stirred great interest in the site, as well as provoked a response from the Kobarid dean Franc Rupnik (Božič 1993; Rupnik 1993). Even though the research was in a preliminary phase, a response from an archaeologist had to follow, for it was necessary to prevent the ‘new Kobarid miracle’ and the ‘Slovenian Pompeii’ from getting needlessly embedded into the minds of the Slovenian Antiquity enthusiasts. Summing up the existing research results Slavko Ciglenečki explained the basic characteristics of the newly discovered Late Antique post and thus replied to the dilemmas that have arisen in the newspapers as regards the character of the site (Ciglenečki 1993). In 1993 we performed the first research intervention in the settlement. We focused on an earlier unauthorised trench on the upper plateau of the settlement (between the central and south church) and cleared the severely collapsed edges of the pit which gave us an insight into the cultural layers.7 Apart from ruins of masonry structures we excavated a large number of Late Antique pottery fragments (coarse kitchenware, amphorae and African Red Slip Ware) and Antique rooftiles (imbreces and tegulae). The first more detailed results of the preliminary research were published in 1993 in the journal Kronika, a thematic issue of which was dedicated to the history of the Tolmin area (Ciglenečki 1994b). Somewhat later 7 We should also correct the statement regarding the beginning of the interest for the site that was published in 1995 in Varstvo spomenikov 35, 105. N. Osmuk, the conservator from ZVNKD Nova Gorica wrote that some land owners were so appalled by the fact that the colleagues from the Institute of Archaeology at ZRC SAZU took away the objects they found, that they decided to dig a small trial trench. The chronological order of the events was quite different. The excavation on the plateau with the churches (as it was proven later) was performed by the locals at the end of the 1980s and they took samples of Roman bricks from the site to the Goriški muzej, however these did not raise the desired interest amongst the archaeologists. In their enthusiasm the locals dug a larger hole measuring 2 x 2 m, with the desire to obtain additional data on the site. It was this overgrown hole covered by a thick layer of leaves and branches that we cleared and documented in 1993. Ž. Šmit who discovered the first characteristic metal finds with a metal detector did not work for the Institute of Archaeology (as stated by the conservator), but independently, without the knowledge of the members of the Institute of Archaeology. He performed an overview of the site that was at the time still uncategorised (as regards its age) and about which he learned from the very article written by N. Osmuk and published in Varstvo spomenikov. 7 Treba je korigirati tudi navedbe o začetkih zanimanja za najdišče, ki so bile objavljene leta 1995 v Varstvu spomenikov 35, 105. Konservatorka ZVNKD Nova Gorica N. Osmuk je v njem zapisala, da so nekateri lastniki zemljišč iz ogorčenja zaradi odnesenih predmetov, ki so jih našli sodelavci Inštituta za arheologijo ZRC SAZU, poskusno izkopali manjšo sondo. Kronološki potek dogodkov pa je bil v resnici povsem drugačen. Izkop na platoju s cerkvami (kot se je izkazalo kasneje) so domačini opravili že ob koncu osemdesetih let in v Goriški muzej odnesli primerke rimskih opek z najdišča, a pri arheologih niso zbudili želenega zanimanja. V svoji vnemi so zato izkopali jamo v velikosti 2 x 2 m, da bi pridobili več podatkov o najdišču. Prav to jamo, ki je bila že povsem zaraščena in prekrita z debelo plastjo listja in vejevja, smo leta 1993 očistili in dokumentirali. Ž. Šmit, ki je z iskalcem kovin odkril prve značilnejše kovinske najdbe, ni bil sodelavec Inštituta za arheologijo, kot se je konservatorki zapisalo, ampak je povsem neodvisno, brez vednosti članov Inštituta za arheologijo, opravil pregled takrat še časovno neopredeljenega najdišča, za katerega je izvedel prav iz notice N. Osmuk v Varstvu spomenikov. 28 1.4 Zgodovina raziskav 1.4 Research history sondažno izkopavanje spremeniti v sistematično, pokazala že kmalu po opravljenih začetnih delih. V Sloveniji je bilo že od sedemdesetih let 20. stoletja dalje odkritih nekaj poznoantičnih višinskih naselbin, od katerih so bile posamezne zaščitno, druge pa tudi sistematično raziskane. Vendar pa se je pri teh kmalu izkazalo, da izkopavanjem večinoma niso sledile v enaki meri tudi sistematične in strokovno ovrednotene objave najdišč. Zato smo se seveda spraševali o smiselnosti odpiranja novega poznoantičnega najdišča. Inštitutska ekipa je prav v tem času raziskala in objavila dve pomembni arheološki poznoantični najdišči Kučar in Tinje (J. Dular, Ciglenečki, A. Dular 1995; Ciglenečki 2000), ki sta bili izkopani zaradi zaščitnih razlogov, tako da smo bili tako kadrovsko kot tudi strokovno dobro pripravljeni na sistematične raziskave. Te so pomemben segment dejavnosti Inštituta za arheologijo, ki naj razrešujejo posebej pomembne arheološko-historične probleme z usmerjenimi ciljnimi raziskavami. Iz dotedanjih izkušenj pa smo tudi vedeli, da raziskana najdišča z izjemo Rifnika in delno Ajdne niso bila tako močno vpeta v turistične tokove, da bi v slovenski zavesti spremenila pogled na izjemno pomembno obdobje transformacije antičnega sveta in na obstoj staroselcev. Temu botruje predvsem dejstvo, da leži večina poznoantičnih višinskih naselbin na odmaknjenih območjih, težko dostopnih krajih in zunaj sodobnih turističnih poti. Tonovcov grad pa vse te predpogoje v celoti izpolnjuje (lep, neokrnjen naraven ambient v najlepšem delu soške doline, turistična razvitost zgornjega Posočja, bližina Kobarida in magistralne ceste, bližina turistične poti, izredno aktivno Turistično društvo Kobarid itd.). Vse našteto, kot tudi izdatna pomoč članov Muzejskega društva in Občine Kobarid, so napeljevali na misel, da bi bilo na Tonovcovem gradu mogoče predstaviti pomembno in z arhitekturo zelo bogato poznoantično naselbino, ki bi jo bilo mogoče v celoti vključiti v turistično ponudbo soške regije in s tem njene ostaline ter problematiko, ki jo ponazarja, predstaviti širšemu krogu prebivalstva, ne samo arheologom in zainteresiranim laikom. Tehtanje za in proti je na koncu odločilo v prid sistematični raziskavi najdišča, hkrati pa smo se začeli pogovarjati tudi z vsemi pristojnimi ustanovami o njegovi konservaciji in učinkoviti predstavitvi izsledkov naših raziskav. Skratka, najdišče se je zaradi vsega navedenega zdelo najprimernejše za sistematično raziskavo, hkrati pa je ponujalo možnost prezentacije v samem središču turističnih tokov, s čimer bi lahko prispevalo k uveljavitvi tovrstnih najdišč in posredno popularizaciji prej slabo znanega obdobja zatona antike ter njegove kompleksne etnične situacije. Zato smo pripravili predlog načrta sistematične raziskave in ga predložili Ministrstvu za znanost RS, ki nam je odobrilo denar za začetne raziskave. Finančno in we published a slightly edited version of the translated text in Italian (Ciglenečki 1994c). The preliminary research thus roughly defined the time span of the settlement from Prehistory to the Middle Ages. We realized that with its size, state of preservation and its complex character the post belongs amongst the most important Late Antique settlements in the East Alpine area. Its strategic position that controls the pass through a vulnerable area at the very gateway to Italy pointed to the possible military character of the post during the Late Roman period and its important role in the defence of Italy. It seemed that our work at the site temporarily ended. However, the interest of the general public, the great interest of our colleagues and their response to the first reports encouraged a different line of thought. Thus soon after the initial research we faced the dilemma whether to change the trial trenching into systematic excavations. From the 1970s on quite a number of Late Antique hilltop settlements were discovered in Slovenia. On some of them protective excavations were carried out, while others were systematically researched. However, with these it was soon noticed that the excavations were not necessarily followed by systematic and expertly evaluated publications. This of course made us consider whether it was at all sensible to open a new Late Antique site. At that time the team from the Institute had just finished researching and publishing two important archaeological sites (Kučar and Tinje, both of which were excavated due to protective reasons; J. Dular, Ciglenečki, A. Dular 1995; Ciglenečki 2000), thus we had experts and knowledge to start systematic research. They represents an important segment of the activities carried out by the Institute of Archaeology, which should solve the particularily important archeohistorical problems through targeted research. We were aware that the previously researched sites (with the exception of Rifnik and partially Ajdna) were not enough incorporated into the tourist routes to change the general views on the extremely important period of the transformation of the Antique world and the existence of the autochthonous population. These views are the consequence of the fact that most Late Antique hilltop settlements are situated in remote areas, hard to access places and outside of the tourist paths. Tonovcov grad, on the contrary, fulfils all the conditions for a popular tourist spot (unspoilt landscape in the most beautiful part of the Soča Valley, already popular with tourists, proximity to Kobarid, the main road and the tourist paths, is under the jurisdiction of an exceptionally active Tourist association of Kobarid, etc.). All of this, along with the substantial help provided by the members of the Museum Society and the Municipality of Kobarid, lead to the idea that the important and architecturally extremely rich settlement on Tonovcev grad could be properly presented and included into the tourist offer 29 1.4 Zgodovina raziskav 1.4 Research history preko sistema javnih del sta nam pri prvih sistematičnih raziskavah pomagali tudi Turistično društvo Kobarid (predsedniki Pavel Sivec, Vojko Hobič, Željko Cimprič) in Upravna enota Tolmin (načelnik Zdravko Likar), pozneje pa še občina Kobarid (župana Pavel Gregorčič in Robert Kavčič), Fundacija Poti miru v Posočju (predsednik uprave Zdravko Likar) in predvsem Kobariški muzej (direktor Jože Šerbec). V letu 1994 je bilo sistematično izkopano območje velike stanovanjske stavbe, ki je bila edina že poškodovana, da bi se tako kar najbolje pripravili na nadaljnja raziskovanja in dodobra spoznali stratigrafijo najdišča. Tega leta je bila ustanovljena tudi posebna konservatorska komisija, ki je bedela nad izvajanjem konservatorskih del in pripravila načrt prezentacije najdišča.8 Prvič se je sestala 2. junija 1994 ter uspešno spremljala dela in sproti pripravljala smernice za konservacijo in prezentacijo. Restavratorski center iz Ljubljane (odgovorna konservatorja Franc Vardijan († 2006) in Jernej Hudolin) je leta 1996 v celoti prevzel vsa konservatorsko-restavratorska dela na tem arheološkem spomeniku. Leta 1995 je bil izdelan dolgoročni program sistematičnih raziskav in konservatorskih posegov na najdišču, ki je okvirno določal obseg in tempo raziskovanja. V tem letu je bil izkopan, v skladu z že prej zastavljeno mrežo kvadrantov, vmesni del med že raziskano stanovanjsko stavbo ter platojem s cerkvami. Sakralni sklop kot najzahtevnejši del dosedanjih raziskav je bil izkopan v letih 1996 in 1997, vmesni del med osrednjo in južno cerkvijo pa v letih 2003 in 2004. Leta 1999 smo skušali locirati grobišče, zato smo naredili tri sonde na terasi na jugovzhodnem pobočju Tonovcovega gradu, ob nekdanjem Tonovcovem seniku. Vse sonde so bile negativne. V letih 2002, 2003 in 2006 je bil izkopan velik zbiralnik za vodo, ki je bil poškodovan že med drugo svetovno vojno, ko je bil v njem zgrajen bunker in je vanj vodil strelski rov. Raziskave cerkvenega kompleksa smo končali v letih 2002 in 2005, ko so bili raziskani ostanki arhitekture tik pod severno cerkvijo, ki jih je bilo zaradi bližine cerkva hipotetično mogoče povezati z njimi. Objekta sta bila označena s številkama 2 in 3 (sl. 1.7) in sta tako zaokrožila podobo poselitve na skalnem platoju in tik pod njim. Raziskovanja so z manjšim ali večjim obsegom tako tekla pravzaprav kontinuirano od leta 1993 dalje in so bila prekinjena le s sezonami, v katerih je bilo treba zidane stavbe konsolidirati in – v primeru severne in osrednje cerkve – zavarovati z zaščitno zgradbo (arhitektka Veronika Ščetinin; sl. 1.13, 1.14) of the Soča region. Thus the remains and the issues they represent would be available to the general public and not only to the archaeologists and enthusiasts. In the end we decided in favour of a systematic excavation, and at the same time we started a dialogue with relevant institutions regarding the preservation of the site and an a good visual presentation of our research findings. Due to all this the site seemed most appropriate for a systematic research and at the same time it offered the possibility of presentation in the very heart of tourist routes, which could help popularise such sites and consequently the previously poorly known period of the decline of Antiquity and its complex ethnical situation. So we prepared a proposal for systematic investigation and presented it to the Ministry of Science of the Republic of Slovenia which funded the preliminary research. In the first systematic research we were financially (as well as through the public works system) helped by the Kobarid Tourist Association (presidents Pavel Sivec, Vojko Hobič, Željko Cimprič) and the Tolmin Administrative Unit (chief Zdravko Likar), and later on also by the municipality of Kobarid (the mayors Pavel Gregorčič and Robert Kavčič), Foundation Poti miru v Posočju (president of the board Zdravko Likar) and especially the Kobariški muzej (director Jože Šerbec). In 1994 we systematically excavated the area of the large building which had been previously damaged and thus seemed the best place to get acquainted with the stratigraphy of the terrain and prepare for future research. A special conservation committee was also established in the same year to supervise the conservation works and prepare the plan for the presentation of the site.8 It assembled for the first time on 2nd June 1994 and successfully monitored the work as well as prepared the guidelines for the preservation and presentation. In 1996 the Restoration Centre from Ljubljana (the responsible conservators Franc Vardijan († 2006) and Jernej Hudolin) took over all preservation and restoration works on this archaeological monument. In 1995 the long term plan for the systematic excavations and preservation interventions was elaborated which roughly defined the scope and tempo of the excavations. In that year the area between the previously invstigated building and the plateau with the churches was excavated according to the previously planned grid. The ecclesiastical complex, which represented the most demanding part of the research so far, was excavated in 1996 and 1997, while the part between the central and south church was explored in 2003 and 2004. 8 At the beginning the members of the committee were the following: N. Osmuk from ZVNKD Nova Gorica as the conservator in charge, M. Slabe, F. Vardijan, I. Bogovčič, M. Sagadin, I. Curk, D. Vuga and S. Ciglenečki, and later on also J. Hudolin and P. Bratina. 8 Člani komisije so bili v začetku N. Osmuk iz ZVNKD Nova Gorica kot pristojna konservatorka, M. Slabe, F. Vardijan, I. Bogovčič, M. Sagadin, I. Curk, D. Vuga in S. Ciglenečki, pozneje pa še J. Hudolin in P. Bratina. 30 1.4 Zgodovina raziskav 1.4 Research history Sl. 1.13: Tonovcov grad. Pogled z vzhoda na zaščitni zgradbi nad severno in osrednjo cerkvijo. Fig. 1.13: Tonovcov grad. A view of the protective constructions above the north and central church, from the east. In 1999 we tried to locate the cemetery, thus three trial trenches were made along the terrace on the southeast slope of Tonovcov grad, alongside the former hayloft. All trial trenches were negative. In 2002, 2003 and 2006 a large water reservoir was investigated. It had been damaged during World War II, when a bunker and a military trench were constructed within the preserved structure. The research of the ecclesiastical complex was concluded in 2005, when the architectural remains under the plateau to the north of the north church were excavated. Due to the proximity of the churches these remains could at least hypothetically be linked to the churches. The buildings were given numbers 2 and 3 (Fig. 1.7) and rounded up the image of the structures on the rock plateau and immediately below it. Thus the research continued (more or less intensively) since 1993 with the exceptions of the seasons in which the masonry buildings needed to be consolidated and – in the case of the north and central church – protected with a protective construction (architect Veronika Ščetinin; Figs. 1.13, 1.14). The site was equipped with signposts and informational panels for visitors (all of which were prepared together with our co-workers from the Kobariški muzej in Kobarid, especially J. Šerbec) during the excavations. Že med potekom večjih raziskav smo najdišče skupaj s sodelavci Kobariškega muzeja (predvsem J. Šerbec) opremili s kažipoti, ki so usmerjali obiskovalce, in pripravili informacijske panoje. Leta 2007 sta bili v celoti dokončani zaščitni zgradbi nad bolje ohranjenima cerkvenima stavbama, urejeni so bili zaščitni hodniki, ki omogočajo ogled notranjosti, leta 2008 pa postavljena še zaščitna streha nad vodnim zbiralnikom. Sakralne zgradbe in ostala že izkopana poslopja so bila zavarovana in urejena, tako da omogočajo obiskovalcem prvi uvid v arhitekturo in ambient, v Tolminskem in v Kobariškem muzeju pa so predstavljeni tudi nekateri izbrani predmeti, ki dovolj nazorno ilustrirajo življenje v poznoantični naselbini. Dosedanji obisk, ki se iz leta v leto veča, kaže številke, kakršne druga arheološka najdišča na našem območju redko dosegajo (pribl. 20.000 obiskovalcev na leto), in dokazuje, da je bila odločitev za raziskavo Tonovcovega gradu pravilna ter da v celoti izpolnjuje v izhodišču zadane cilje. 31 1.4 Zgodovina raziskav 1.4 Research history Sl. 1.14: Tonovcov grad. Odlično ohranjena in z moderno zaščitno zgradbo zavarovana notranjost severne cerkve. Fig. 1.14: Tonovcov grad. The excellently preserved interior of the north church protected by the modern construction. In 2007 the protective buildings that covered the better preserved churches were completed, protective hallways that give the view of the interior were created, and in 2008 a protective roof was raised above the water reservoir. The churches and the remaining excavated buildings were protected and presented to offer the visitor an insight into the architecture and the interior. In the Tolminski muzej in Tolmin and in the Kobariški muzej in Kobarid some artefacts were put on display in order to illustrate the life in the Late Antique settlement. As the number of visits has been growing each year, and the site has been achieving numbers that other archaeological sites in Slovenia find hard to reach (approximately 20.000 visitors per annum), we are convinced that the decision to explore Tonovcov grad was correct and that it fulfils all the goals set at the beginning. 32 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Raziskave poznorimskega zapornega sistema Claustra Alpium Iuliarum so za dolgo obdobje določale raziskovalne prioritete slovenskih in tudi nekaterih italijanskih in avstrijskih arheologov. Ostala arheološka dejavnost v zahodni Sloveniji je bila tako dolgo v senci enega najpomembnejših arheoloških spomenikov v alpskem območju. Raziskovati so ga začeli že v drugi polovici 19. st., vrh pa so raziskave doživele s sistematičnimi izkopavanji v 60. in 70. letih 20. st. (pregled pri Šašel, Petru 1971). Z raziskovanjem utrdb zapornega sistema se je povečalo tudi zanimanje za ostale poznoantične postojanke v njihovem zaledju kot tudi pred njim. Zastavljene so bile raziskave na nekaj ključnih točkah, ki omogočajo razumevanje preživetja staroselcev v nemirnem času pozne antike in preseljevanja ljudstev (Petru, Ulbert 1975; Bolta 1981). Ta raziskovalna usmerjenost se je Posočja dotaknila le bežno, saj je bila večina raziskav v zahodni Sloveniji osredotočena vzdolž glavne rimske vpadnice iz Emone v Akvilejo in na območju Gorenjske. Prve skromne poskuse sistematičnega arheološkega raziskovanja poznoantičnega obdobja v Posočju z namenom razumeti poselitvene razmere, je mogoče zaznati šele s sondiranjem na Sv. Katarini (Kekec) pri Novi Gorici v letih 1971 in 1976, vzporedno z velikimi izkopavanji, ki so takrat potekala na Hrušici (Svoljšak 1990). Nenavaden kultni prostor na pobočju Gradiča nad Kobaridom, ki so ga uporabljali vse do začetka 5. st. in odkritje izjemno dobro ohranjene utrjene poznoantične naselbine na Tonovcovem gradu so žarišče antičnih raziskav v zadnjih dveh desetletjih postavili prav v to, prej tako slabo raziskano območje (Osmuk 1987a; Ciglenečki 1994b). Poglejmo, kakšne rezultate nam ob izhodiščni navezavi na Tonovcov grad kot referenčno točko par excellence ponuja Posočje z množico skromnih podatkov, ki jih je mogoče bolje razumeti prav v luči raziskav tu obravnavanega najdišča. Naselitvena podoba poznoantičnega obdobja v Posočju je dokaj pestra in razkriva podobno kot v večjem delu slovenskega prostora široko paleto naselbinskih tipov, vse od še obstoječih nezavarovanih naselij v 4. st. pa do močno utrjenih bivalnih središč na naravno zavarovanih položajih v 6. st. (sl. 1.15). Ob tem je treba že For a long period the research priorities of Slovenian as well as certain Italian and Austrian archaeologists were focused on the Late Roman defence system known as Claustra Alpium Iuliarum. The remaining archaeological activities in western Slovenia thus remained in the shade of one of the most important archaeological monuments in the Alpine area. This system was researched as early as the second half of the 19th century, however the research peaked with the systematic excavations carried out in the 1960s and 1970s (see Šašel, Petru 1971). With the research of the fortifications found within the defence system the interest in other Late Antique posts in the system’s hinterland as well as in front of it started growing. Research was initiated into the key strongholds that enable the understanding of the survival of the autochthonous population during an extremely factious period of the Late Antiquity and the Migration Period (Petru, Ulbert 1975; Bolta 1981). This research only briefly dealt with the Posočje area, for most of the research in western Slovenia was focused on the area around the main Roman road leading from Emona towards Aquileia, and the Gorenjska region. The first modest attempts of systematic archaeological research of the Late Antiquity in the Posočje area – with the intent of understanding the settlement conditions – took place as late as 1971 and 1976 when trenching was carried out on Sv. Katarina (Kekec) above Nova Gorica at the same time as extensive excavations took place on Hrušica (Svoljšak 1990). The unusual cult place on the slope of Gradič above Kobarid that was used until the beginning of the 5th century and the extremely well preserved fortified Late Antique settlement at Tonovcov grad set the focus of the Antiquity research during the last two decades in this previously poorly researched area (Osmuk 1987a; Ciglenečki 1994b). Let’s take a look at what sort of results the Posočje area with its links to Tonovcov grad as a referential point par excellence offers us with its abundant yet modest data that can be better understood through the research of the here discussed site. The settlement image of the Late Antiquity in the Posočje region is relatively diverse and – similar to many parts in Slovenia – reveals a wide palette of settlement 33 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.15: Zemljevid Posočja s poznoantičnimi arheološkimi najdišči. Fig. 1.15: Map of the Posočje area with Late Antique sites. 34 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity uvodoma poudariti, da je ta podoba znana predvsem z nekaj bolje raziskanih območij (Kobarida, Mosta na Soči in Nove Gorice), medtem ko imamo od drugod na razpolago skromne in večinoma naključne podatke, ki ne omogočajo celovitega razumevanja tedanje poselitve. V vzhodnoalpskem območju prevladuje – gledano z vidika poselitve – okvirna omejitev poznoantičnega obdobja z letnicama 300–600 (Demandt 1989, XIX– XXI; Ciglenečki 1999, 289). Novejše raziskave pa so pokazale, da je do prvih pomembnejših naselbinskih premikov prišlo že v zadnji tretjini 3. st., in sicer zaradi vedno močnejšega pritiska barbarov na meje imperija. To je odsevalo v kratkotrajni obiskanosti starejših prazgodovinskih gradišč, gradnji številnih mestnih obzidij, hkrati pa je botrovalo nastanku novih vojaških utrdb (Ciglenečki 1990). V Posočju smo zanesljivo sled teh prvih premikov na višinske položaje ugotovili le na Tonovcovem gradu. Zaradi pretežno hribovitega sveta je bilo v Posočju ugotovljenih le nekaj vil rustik, ki bi jim bilo mogoče dokazati obstoj ali celo nastanek v poznorimskem času. Marsikje podatki sicer kažejo na njihov obstoj, raziskane so bile redke. Med bolj znanimi omenimo le rimski vili v Bukovici in Pavlinih pri Loki v zahodnem delu Vipavske doline (Osmuk 1981; Žbona Trkman 1987). Bolje raziskana je bila le slednja, ki pa je bila opuščena ali uničena prav v drugi polovici 3. st., torej že v času prvih usodnejših naselbinskih premikov. Zidove, ki jih je mogoče opredeliti kot stanovanjsko zgradbo iz 4. st., so odkrili na Ledinah v Novi Gorici (Osmuk 1987b). Več je arheoloških sledov, ki posredno nakazujejo obstoj poznorimskih nezavarovanih naselbin. Takšni sta predvsem manjša nekropola iz 4. st. Na lajšču v Avčah in podobna v Anhovem (ANSl 1975, 124; Kos 1988, 66, št. 27). Med zaščitnim izkopavanjem nekropole Na steni v Solkanu so odkrili nekaj poznorimskih grobnih celot, ki pa še niso objavljene (Knific, Svoljšak 1984, 277). Glede na velikost in časovno opredelitev omenjenih grobov je mogoče sklepati na skromne pripadajoče neutrjene naselbine iz 4. st. Obstoj bivališč ob Idrijci potrjujeta posamezna poznorimska grobova iz sicer večjih nekropol, katerih začetki segajo že v latensko obdobje: grob iz Idrije ob Bači in iz Slapa ob Idrijci (Kos 1988, 23). Skratka, podoba starejših, nezavarovanih naselbin iz 4. st. je povsem neznana, posredno potrjena le z nekaj grobišči, ki nakazujejo njihovo prisotnost. Arheološke indikacije potrjujejo obstoj dveh pomembnih naselbin, pri katerih pa ni mogoče povsem zanesljivo opredeliti značaja in obsega poselitve: prva je Gradič nad Kobaridom in druga Most na Soči. Večletna raziskovanja na Gradiču nad Kobaridom že omogočajo okvirno kronološko zamejitev trajanja poselitve. Rimska naselbina na pobočjih in temenu Gradiča je nastala v območju nekdanjega velikega prazgodovinskega gradišča in se nadaljevala – verjetno celo brez večje časovne cezure – v 4. st. in delno še v začetek 5. st. (s1. types ranging from the still existing unprotected settlements from the 4th century to the strongly fortified settlements on naturally protected positions dated to the 6th century (Fig. 1.15). Of course, it should be emphasised that this image is known from the slightly better researched areas (Kobarid, Most na Soči and Nova Gorica), while elsewhere only modest and mainly accidental data is at our disposal, and this does not allow for a wholesome understanding of the settlement of the time. As regards settlement in the Eastern Alpine area the Late Antiquity is most commonly marked by the period between 300 and 600 AD (Demandt 1989, XIX-XXI; Ciglenečki 1999, 289). Newer research has shown that the first important settlement changes occurred already during the last third of the 3rd century. These changes occurred due to the increasing pressure of the Barbarians upon the Empire borders. This was reflected in the short visits of the older prehistoric settlements, the construction of numerous town walls, as well as the appearance of new military fortifications (Ciglenečki 1990). In the Posočje area only Tonovcov grad represented a reliable proof of these first moves towards the highland positions. Due to the predominantly mountainous world of the Posočje area only a few Roman villas were discovered that existed or were built during the Late Roman period. Even though there is data that indicates the existence of villas only a handful of them have been researched so far. Amongst the best known are the Roman villas in Bukovica and Pavlini near Loka in the western part of the Vipava Valley (Osmuk 1981; Žbona Trkman 1987). Only the latter one, which was abandoned or destroyed in the second half of the 3rd century (i.e. during the period of the first important moves), has been researched in greater detail. Walls that can be defined as belonging to 4th century living quarters were discovered at Ledine in Nova Gorica (Osmuk 1987b). There are a number of archaeological indicators that indicate the existence of unprotected Late Roman settlements. Examples of such are the small 4th century necropolis at Na lajšču in Avče and a similar one in Anhovo (ANSl 1975, 124; Kos 1988, 66, No. 27). A few Late Roman graves were discovered during the protective trenching at the necropolis Na steni in Sol kan, however these finds remain unpublished (Knific, Svoljšak 1984, 277). Taking into account the size and dating of the aforementioned graves it can be concluded that they were accompanied by modest unfortified settlements dating to the 4th century. The existence of a settlement close to the Idrijca river was confirmed by two individual Late Roman graves from two otherwise larger cemeteries, the beginnings of which reach into the La Tène period: i.e. the graves from Idrija pri Bači and from Slap ob Idrijci (Kos 1988, 23). Thus the layout of the older and unprotected 4th century settlements is entirely unknown and only circumstantially confirmed by a few cemeteries that indicate their presence. 35 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.16: Gradič nad Kobaridom. Pogled na gradišče s severovzhodne strani. Fig. 1.16: Gradič above Kobarid. A view of the hillfort from the northeast. Sl. 1.17: Gradič nad Kobaridom. Pogled na gradišče z južne strani. V ozadju Tonovcov grad. Fig. 1.17: Gradič above Kobarid. A view of the hillfort from the south. Tonovcov grad in the background. 1.16–1.18). Poznoantične zgradbe doslej sicer niso bile raziskane, vendar množica novcev in drugih drobnih najdb v območju domnevnega svetišča posredno priča o njihovem obstoju (Osmuk 1987a; 1997a). Tukajšnji izsledki dokazujejo, da je naselbina obstajala (morda v zmanjšanem obsegu?) še v času, ko na Tonovcovem gradu ugotavljamo prisotnost močnejše vojaške posadke in trajnejše sledove bivanja. V Mostu na Soči so ob izkopavanjih velike in izjemno ohranjene prazgodovinske naselbine naleteli Archaeological traces confirm the existence of two important settlements, however their character and scope cannot be precisely defined: the first is Gradič above Kobarid and the other is Most na Soči. The long lasting research at Gradič above Kobarid allows for a partial chronological determination of the duration of the settlement. The Roman settlement on the slopes and ridge of Gradič emerged in the area of a former large prehistoric site and continued – most likely without a longer time gap – into the 4th and partially into the be36 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.18: Gradič nad Kobaridom. Poskus zarisa obsega gradišča po lidarskem posnetku (© ZRC SAZU). Fig. 1.18: Gradič above Kobarid. Mapping the hillfort on a lidar data visualization (© ZRC SAZU). ginning of the 5th century (Figs. 1.16-1.18). So far Late Antique buildings have not been researched, however the high numbers of coins and other small finds in the area covered by an assumed cult place indicates their existence (Osmuk 1987a; 1997a). The data from this site indicates that a settlement still existed (maybe on a smaller scope?) at a time when a stronger garrison and traces of more permanent settlement were discovered at Tonovcev grad. In Most na Soči the excavations of the large and extremely well preserved prehistoric settlement also revealed numerous Roman traces (Fig. 1.19). The preliminary data indicates the existence of eight Roman buildings, however only one of them has been researched into any sort of detail. In this context it is important to mention that a number of Roman coins and three Late Roman hoards were discovered amongst numerous other traces. The first two hoards can be dated to the second half of the 4th century and were discovered in tudi na številne rimske sledove (sl. 1.19). Iz preliminarnih podatkov je mogoče izluščiti omembo osmih rimskih stavb, bolje znana je le ena. V obravnavanem kontekstu je pomembno, da so ob številnih drugih sledovih odkrili več rimskih novcev in tri poznorimske zakladne najdbe. Prvi dve, ki sodita v drugo polovico 4. st., sta bili najdeni v ruševinski plasti rimske stavbe, izkopane leta 1979, tretja, katere zakop je natančno datiran v november leta 401, pa je bila odkrita na ognjišču rimske stavbe (Kos 1988, 31–32, št. 9/3 in 9/4; 33, št. 9/5). Njihov zakop odlično nakazuje trajanje rimske naselbine v začetek 5. st., kar je omogočila naravno odlično zavarovana lega na sotočju Soče in Idrijce, v celoti pa tej dataciji pritrjuje grobišče s 149 grobovi, od 1. do 5. st. (Horvat 1999, 253). Pri obeh zadnjih obravnavanih naselbinah ni mogoče zanesljivo ugotoviti, ali gre za nezavarovani naselbini ali pa so prebivalci vsaj deloma že izkoristili njuno ugodno obrambno lego: na Gradiču delno 37 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.19: Most na Soči. Načrt najdišča z rdeče označenimi rimskimi stavbami (po Maggi, Žbona Trkman 2007, sl. 7). Fig. 1.19: Most na Soči. Plan of the site with Roman structures marked in red (after Maggi, Žbona Trkman 2007, Fig. 7). the destruction layer of the Roman building excavated in 1979, while the third, for which it has been successfully ascertained that it was buried in November 401, was discovered on the fireplace of a Roman building (Kos 1988, 31-32, Nos. 9/3 and 9/4; 33, No. 9/5). Their burial indicates that the Roman settlement continued into the beginning of the 5th century, which was enabled by the naturally excellently protected position at the confluence of the Soča and Idrijca rivers. This has also been confirmed by the graveyard with 149 graves that can be dated between the 1st and 5th centuries (Horvat 1999, 253). At both of the settlements mentioned above it is impossible to precisely determine whether they were unsecured settlements or whether the settlers made at least partial use of their favourable defensive position: on Gradič partially with the old prehistoric ramparts and on Most na Soči by the two rivers that offer an excellent protection of the location that could be successfully defended at the point with the easiest access. Much clearer is the settlement pattern of fortified hilltops that was already fully established at the time (Ciglenečki 1999, 291-294). Recognising these settlements is made easier by their position on naturally well znotraj starih prazgodovinskih okopov in na Mostu na Soči v prostoru, odlično zavarovanem z rekama, ki ga je bilo mogoče na najlažje dostopnem delu tudi uspešno braniti. Veliko bolje je razvidna poselitev na utrjenih višinskih položajih, ki se je polno uveljavila v tem času (Ciglenečki 1999, 291–294). Prepoznavanje teh naselbin je olajšano z lego na že naravno dobro zavarovanih krajih, hkrati pa so bile zaradi svoje po navadi odmaknjene hribovske lege pozneje manj poškodovane kot sočasne ravninske naselbine. Sistematične raziskave na Tonovcovem gradu so dokazale, da je bila tam v drugi polovici 4. st. nameščena močna vojaška posadka, katere naloga je bila predvsem kontrola prehoda proti Italiji in varstvo rimske ceste (glej pogl. 5.1). Podobne postojanke so takrat nastale tudi na nekaterih drugih naravno zavarovanih hribih, kjer pa manjkajo zanesljivi elementi za podrobnejšo označitev njihove funkcije. Vse označujejo naravno zavarovana lega, ponekod odkrite poznorimske ali tudi že starejše obrambne naprave, hkrati pa arheološki predmeti, ki dovoljujejo podrobnejšo datacijo. Takšno je pred nedavnim odkrito najdišče Selce pri Zatolminu (sl. 1.20), ozek gozdnat pomol, ki se strmo dviga nad ravnico 38 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.20: Selce pri Zatolminu. Pogled na poznoantično postojanko z južne strani Fig. 1.20: Selce near Zatolmin. A view of the Late Antique post from the south. Sl. 1.21: Sv. Katarina nad Novo Gorico. Pogled na poznoantično utrdbo z južne strani. Fig. 1.21: Sv. Katarina above Nova Gorica. A view of the Late Antique fortification from the south. (Cvitkovič 1999, 56; Mlinar, Knavs 2004). Nekaj kovinskih najdb (predvsem fibula s čebulastimi gumbi) datira najdišče v poznorimsko obdobje. Ali gre za strateško nameščeno opazovalnico, pribežališče v dolini prebivajočih staroselcev ali celo kombinacijo obojega? Poznorimska naseljenost in pripadnost obrambnih naprav sta vprašljivi tudi na velikem kompleksnem najdišču Sv. Katarina nad Novo Gorico. Velik kopast hrib, na katerem je nekdaj stala cerkev sv. Katarine, z n. m. v. 328 m, je prislonjen na južno pobočje Škabri- protected locations and due to the fact that their usually remote highland position did not make them as prone to being destroyed later on as the settlements in the lowland from the same period. The systematic research at Tonovcov grad has shown that in the second half of the 4th century this was the home to a garrison, the task of which was to control the access route towards Italy and protect the Roman road (see chapter 5.1). At the time similar posts emerged on other naturally protected hills, however in 39 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity the other cases no reliable elements that would enable us to precisely define their function were discovered. All are marked by the naturally protected position, and some have also revealed Late Roman or even older defence elements, as well as archaeological objects that lead to a precise dating. Such was the recently discovered site Selce near Zatolmin (Fig. 1.20), a narrow forest ridge that rises steeply above the plain (Cvitkovič 1999, 56; Mlinar, Knavs 2004). A few metal finds (especially the crossbow fibula) date the site into the Late Roman period. Was this a strategically placed observation point, a refuge of the autochthonous population living in the valley or a combination of both? Late Roman settlement and the defensive elements are questionable also on the large and complex site of Sv. Katarina above Nova Gorica. With an altitude of 328 metres above sea level, the large rounded hill that used to be the home to the church of St. Catherine and that leans upon the southern slope of Škabrijel is strategically excellently placed at a point where the river Soča runs from the mountains into the plains. Steep slopes and rocks on three sides offer excellent protection, and it is only on the northern side that the few metres lower saddle divides it from the higher hills behind it. The approximate size of the protected area measures 250 x 160 m (Fig. 1.22). The large prehistoric site was refortified in the Late Antiquity. Older authors mention Late Antique walls and four clearly visible defensive towers. With his trenching in 1971 and 1976 the excavator Drago Svoljšak ascertained two phases of later buildings on the prehistoric rampart: a 1.4 m thick wall and a defensive tower, the size of which is unknown (Svoljšak 1990; Osmuk 1997b, 176). The discovery of the well used 4th century coin cannot represent a sufficient reason to date the first phase of the wall into the Late Roman period, however keeping in mind the contemporary examples from the Eastern Alpine region we could consider this to be a 4th or early 5th century wall as this has proven to be the case elsewhere (Rifnik, Duel, Zbelovska gora, Šumenje, etc.). This is also confirmed by the thickness of the wall that was matched only at the older Late Roman fortifications (Gradišče in Vrhnika, Velike Malence, Hrušica), while it never surpassed 0.80 m at younger fortifications (Ciglenečki 1987a, 129). The chronological placement of the protruding defence towers is questionable and their construction should be assumed to belong into the last stages of the Late Antiquity (similarly Rifnik). The link between the Late Roman graveyard in Ko soveli and the settlement Tabor above Črniče 200 m away cannot be questioned. As indicated by individual Roman ceramics and especially a jug with burnished decoration the settlement was built on the location of a Late Roman fortification (Zavrtanik 1984; Knific 2004, 9). As a certain addition to the settlement pattern we have to take into account the prehistoric site Kozmac near Gojače that lies in the Vipava Valley and is situated Sl. 1.22: Sv. Katarina nad Novo Gorico. Načrt naselbine (Svoljšak 1990, sl. 3). M. = 1:5000. Fig. 1.22: Sv. Katarina above Nova Gorica. Settlement plan (Svoljšak 1990, Fig. 3). Scale = 1:5000. jela in strateško zelo ugodno umeščen pri izteku Soče s hribovja v ravnico (sl. 1.21). Na treh straneh je odlično zavarovan s strmimi pobočji in deloma s skalami, le na severni strani ga nekaj metrov nižje sedlo loči od višjega hribovja za njim. Približna velikost zavarovanega prostora znaša 250 x 160 m (sl. 1.22). Veliko prazgodovinsko gradišče je bilo v poznoantičnem času ponovno utrjeno. Starejši avtorji omenjajo poznoantično obzidje in kar štiri dobro vidne obrambne stolpe. Izkopavalec Drago Svoljšak je s sondiranji leta 1971 in 1976 ugotovil dve fazi poznejših gradenj na prazgodovinskem okopu, 1,4 m debelo obzidje in obrambni stolp, katerega velikost ni znana (Svoljšak 1990; Osmuk 1997b, 176). Najdba zlizanega novca iz 4. st. ne more biti dovolj tehten razlog za datacijo prve faze obzidja v poznorimski čas, a bi smeli glede na sočasne primerjave z vzhodnoalpskega območja v starejši fazi vendarle pomisliti na obzidje iz 4. ali začetka 5. st., kot se je to izkazalo drugod (Rifnik, Duel, Zbelovska gora, Šumenje itd.). To bi potrjevala tudi debelina zidu, ki je bila tako močna le pri starejših poznorimskih utrdbah (Gradišče v Vrhniki, Velike Malence, Hrušica), medtem ko pri poznejših postojankah ne presega širine 0,80 m (Ciglenečki 1987a, 129). Časovna pripadnost izstopajočih obrambnih stolpov je vprašljiva, njihovo izgradnjo bi smeli domnevati šele v zadnjem poznoantičnem obdobju (podobno Rifnik). 40 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity on a low, but vast, today partially ruined elevation (approximately 20 metres high; Fig. 1.23). The finds indicate a modest Late Antique settlement and allow for the post to be dated into the 4th and also into the late 6th or even 7th century (Harej 1989). Its position on a low elevation in the immediate vicinity of the Roman state road could indicate a small provisional military post or a refuge that was based on old prehistoric ramparts. The site Sv. Marija na Jezeru in the vicinity of Golo Brdo, where two Late Roman coins were discovered on a strategically placed prehistoric hillfort seems similar. The coins indicate the possibility that the naturally and in prehistory artificially fortified hill was for a short time used also in the second half of the 4th and the beginning of the 5th century (Šemrov 2004, 35, No. 12; Bratina 1999). Even though our overview is limited to the Slo venian side of the Posočje region we should mention the important nearby post Castelazzo near Doberdò, for this will shed additional light on the conditions of the time. This post has been partially researched, however the large quantities of discovered coins enable for a more reliable chronological frame (Furlani 1969; Maselli Scotti 1992). The naturally protected fortification at an elevation of 150 m a.s.l. stood alongside the road that connected the two important roads towards Italy: the northern that lead along the Vipava Valley and the southern coastal road (Maselli Scotti 1992, 372). The prehistoric rampart was reinforced in Late Antiquity (Figs. 1.24, 1.25). The trenches have shown that the wall was between 3 and 3.5 metres thick and that it could be dated into the beginning of the 5th century. The fortification was inhabited from the 3rd century onwards but the finds are more abundant from the transition between the 4th and 5th century on. 800 Late Roman coins from that period – maybe remains of a hoard – were discovered. The assumption that the find belongs to the time of Attila’s invasion in 452 cannot be excluded (Maselli Scotti 1992, 372). Similar to the arrowheads the ceramic finds could also be dated after the mid 5th century. Maselli Scotti even assumed that this could be the remains of a fortification with a Byzantine garrison (Maselli Scotti 1992, 373). Late Roman settlements dated with the use of coins in the Posočje area as well as in the broader Eastern Alpine area go up to the third decade of the 5th century. The discontinuation of the coin circulation and the lack of other chronologically definable objects do currently not permit us to precisely define the time in which these settlements and fortifications were abandoned. Thus special attention is raised by the site in the vicinity of Robič, where a 5th century post is indicated. On Molida east of village Robič in the stone deposit of a landslide that broke from the northeast slope of Matajur, a solidus of Valentinian III was found by chance in 1892. The coin was minted between 425 and 455 (Župančič 1991, 167). During the railway construction a wall and some pottery fragments were supposedly found on this location Sl. 1.23: Kozmac. Načrt poškodovane poznorimske postojanke (po Harej 1988–1989, sl. 1). M. = 1:5000. Fig. 1.23: Kozmac. Site plan of the partly destroyed Late Roman post (after Harej 1988-1989, Fig. 1). Scale = 1:5000. Povsem zanesljiva je povezava med poznorimskim grobiščem v Kosovelih in 200 m oddaljeno naselbino Tabor nad Črničami, ki je bila zgrajena na mestu poznorimske utrdbe, kot nakazujejo posamezne najdbe rimske keramike in še posebej vrč z zglajenim okrasom (Zavrtanik 1984; Knific 2004, 9). Kot svojevrstno dopolnitev poselitvene slike je treba pritegniti tudi že v Vipavski dolini ležeče prazgodovinsko gradišče Kozmac pri Gojačah, situirano na povsem nizkem, a prostranem, danes delno uničenem griču (višina griča pribl. 20 m; sl. 1.23). Izsledki raziskav pričajo o skromni poznoantični poselitvi in dovoljujejo datacijo postojanke v čas 4. in tudi poznega 6. oz. celo 7. st. (Harej 1989). Njegova lega na nizki vzpetini povsem v bližini državne rimske ceste bi utegnila kazati na manjšo zasilno vojaško postojanko ali pribežališče, ki se je naslonilo na stare prazgodovinske okope. Zelo podobna se zdi problematika najdišča Sv. Marije na Jezeru pri Golem Brdu, kjer so na strateško umeščenem gradišču odkrili dva poznorimska novca: nakazujeta možnost, da je bil naravno zavarovan in v prazgodovini umetno utrjen hrib za kratek čas uporabljan tudi v drugi polovici 4. in na začetku 5. st. (Šemrov 2004, 35, št. 12; Bratina 1999). Čeprav smo se v svojem pregledu omejili na slovenski del Posočja, pa je zaradi boljšega razumevanja tedanjih razmer na tem območju koristno pritegniti tudi pomembno bližnjo postojanko Gradina pri Doberdobu (Castelazzo di Doberdò), ki je bila le delno raziskana, vendar velike količine novcev z nje omogočajo zanesljivejši kronološki okvir (Furlani 1969; Maselli Scotti 41 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.24: Gradina pri Doberdobu (Castellazzo di Doberdò). Pogled na utrdbo z južne strani. Fig. 1.24: Gradina near Doberdob (Castellazzo di Doberdò). A view of the fort from the south. Sl. 1.25: Gradina pri Doberdobu (Castellazzo di Doberdò). Načrt poznoantične utrdbe (Furlani 1969, sl. 65). Fig. 1.25: Gradina near Doberdob (Castellazzo di Doberdò). Site plan of the Late Antique fort (Furlani 1969, Fig. 65). (Osmuk 1985e, 297). The poem Molida by J. Lavrenčič mentions a Roman sword and the story told by the locals speaks about a town that was covered by a landslide from Matajur (because of the sins of its inhabitants). This data indicates a potential Late Antiquity site that cannot be precisely ascertained. 1992). Naravno zavarovana utrdba na n. m. v. 150 m je ležala ob cesti, ki je povezovala obe pomembni vpadnici proti Italiji: severno, ki je tekla po Vipavski dolini, in južno obalno cesto (Maselli Scotti 1992, 372). Prazgodovinski okop so ponovno utrdili v pozni antiki (sl. 1.24, 1.25). Sonde so pokazale, da je obzidje debelo 3–3,5 m in 42 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity ga je mogoče datirati v začetek 5. st. Utrdba je bila obljudena od 3. st. dalje, zgoščene najdbe pa so s prehoda med 4. in 5. st. Našli so 800 poznorimskih novcev – morda depo – iz istega časa. Ni izključena domneva, da sodi najdba v čas Atilovega vpada leta 452 (Maselli Scotti 1992, 372). Keramične najdbe je podobno kot puščične osti mogoče datirati tudi v čas po sredini 5. st. Maselli Scottijeva domneva v utrdbi celo bizantinsko posadko (Maselli Scotti 1992, 373). Vse z novci dobro datirane poznorimske naselbine v Posočju in tudi na širšem območju vzhodnih Alp segajo v tretje desetletje 5. st. Prenehanje novčnega obtoka in pomanjkanje drugih kronološko občutljivih predmetov nam za sedaj onemogočata natančneje opredeliti čas opustitve teh naselbin in utrdb. Zato zbuja posebno pozornost najdišče v bližini Robiča, kjer se nakazuje postojanka iz 5. st. Leta 1892 so namreč na Molidi, ledini vzhodno od naselja Robič, v kamnitem nanosu plazu, ki se je utrgal s severovzhodnega pobočja Matajurja, po naključju našli solidus Valentinijana III., ki je bil kovan med letoma 425–455 (Župančič 1991, 167). Pri gradnji železnice naj bi na tem mestu naleteli na zidovje in lončenino (Osmuk 1985e, 297). V pesmi J. Lavrenčiča Molida pa je omenjen rimski meč in tudi pripoved domačinov pozna mesto, ki ga je zaradi grešnosti njegovih prebivalcev zasul plaz z Matajurja. Ti podatki nakazujejo potencialno poznoantično najdišče, ki pa ga podrobneje ni mogoče označiti. Podobno kot opustitve poznorimskih postojank v prvi polovici 5. st., doslej ni bilo mogoče v vzhodnoalpskem svetu natančneje datirati niti nastanka zadnjih oblik poznoantičnih utrjenih višinskih naselbin, ki so najznačilnejši izraz transformacije poznoantičnega sveta (Ciglenečki 1987a, 114–116). Raziskave na Tonovcovem gradu so sicer dokazale obstoj velike utrjene naselbine s številnimi stanovanjskimi zgradbami in cerkvami po sredini 5. st., pri čemer pa je verjetnejša zadnja tretjina 5. st. (glej pogl. 2.2). To datacijo posredno potrjujejo primerjave z drugimi sorodnimi naselbinami v vzhodnoalpskem prostoru, ki so bolje datirane z že raziskanimi grobišči (Rifnik, Kranj, Teurnia idr.). V zadnjo skupino poznoantičnih utrjenih naselbin v Posočju bi poleg Tonovcovega gradu smeli uvrstiti še dve manjši postojanki, ki pomembno dopolnjujeta poselitveno sliko. Površinsko zelo izrazit je Gradec pri Drežnici, ki leži v neposredni bližini Tonovcovega gradu na levem bregu Soče na severnem odrastku hriba Ozben na robu Drežniškega kota (Osmuk 1985b; Ciglenečki 1997b, 25). Dviguje se strmo nad kanjonom potoka Kozjeka, na drugih straneh pa sega le nekaj metrov nad okolico (sl. 1.26). Približno 80 x 20 m velika površina hriba je bila na mestih, ki niso bila naravno zavarovana, to je predvsem na južni in vzhodni strani, dobro utrjena z okopom. Ker arheološke raziskave na hribu niso bile opravljene, ni mogoče reči, ali se v dobro vidnem okopu skriva kombi- It has been so far impossible to precisely date the abandonment of the Late Roman sites during the first half of the 5th century as well as the foundation of the Late Antique hilltop settlements that represent the most characteristic expression of the transformation of the Late Antique world in the Eastern Alps (Ciglenečki 1987a, 114-116). The research at Tonovcov grad has proven the existence of a large fortified settlement with numerous living quarters and churches after the middle of the 5th century, but it is more likely that the settlement was founded in the last third of the 5th century (see chapter 2.2). This data is indirectly confirmed by the comparisons to other similar settlements in the Eastern Alpine area that have been more precisely dated with the use of the explored cemeteries (Rifnik, Kranj, Teurnia, etc.). Alongside Tonovcov grad two smaller posts that importantly add to the settlement pattern should be added to the last group of Late Antique fortified settlements in the Posočje region. Gradec near Drežnica lies in the vicinity of Tonovcov grad – on the left side of the river Soča and on the northern slope of the hill Ozben which stands on the edge of Drežniški kot (Osmuk 1985b; Ciglenečki 1997b, 25). On one side it rises sharply above the stream of Kozjek, while on the remaining sides it raises above its surroundings by merely a few metres (Fig. 1.26). In the spots where it was not naturally protected, i.e. especially on the southern and eastern side, the area of the hill measuring approximately 80 x 20 m was fortified with a rampart. Because no archaeological research was performed on the hill, it is impossible to say whether a combination of earthwork and quarry stone or even a wall construction with mortar is hiding beneath it. As much as it was possible to ascertain merely by observing the surface, the first option seems to be the likeliest. The surface in the interior is artificially levelled out and shows a number of variously wide terraces that gradually descend towards the northwest (Fig. 1.27). Some modest surface finds (especially pottery) place the settlement into the Late Antique period. The lack of surface traces that would indicate the presence of permanent building forms in the extremely small area of the post as well as the modest surface finds indicate a naturally well protected short term refuge. We should consider the possibility of a fort that, in the case of need, protected or closed the route from the Soča Valley towards Drežnica, for this was one of the most important entries into Drežniški kot. Taking into account the vastness of the area under the mountain of Krn we could assume scattered settlements or individual homesteads of the autochthonous population who retreated to this position during the Late Antiquity. Undoubtedly the inhabitants who used this fortified settlement saw their centre in the nearby settlement of Tonovcov grad, which is just over 1 kilometre away. This fortified settlement indicates an autochthonous population’s enclave in Drežniški kot, 43 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.26: Gradec pri Drežnici. Pogled na najdišče z zahodne strani. Fig. 1.26: Gradec near Drežnica. A view of the site from the west. which has been additionally confirmed by the finds discovered in the studies of the folklore, names and other research (Kuret 1972; 1989; Volarič 2008). Gradec near Logje – the third settlement that can be reliably dated into Late Antiquity – is located in Breginjski kot next to the Italian border (Osmuk 1985d; Ciglenečki 1997b, 25). The hill (446 m a.s.l.), which is naturally well protected by slopes and on the eastern side by a vertical rock face, is located at the confluence of the river Nadiža and the stream Legrado (Fig. 1.28). It can be accessed only from the southern side, and even then a long saddle and a steep slope need to be passed. The width at the top of the exposed ridge rarely exceeds 20 metres and the ridge measures approximately 120 metres in length. Minute traces where the terrain has been levelled out are visible, and for two points it can be assumed that they created a space for improvised wooden constructions. Due to its naturally protected position it remains an excellent refuge in a region which is sparsely inhabited even today. There are no traces of a rampart at the top, however it was most likely also not needed (Fig. 1.29). A recently discovered fibula (Osmuk 1999, 64-66; 2001, 47) proves that it was inhabited in the 6th century, thus it could be assumed to be a Late Antique refuge, similar to the one at Gradec near Drežnica. This purpose is indirectly indicated by extremely modest surface finds. It seems that the inhabitants were – due to the remoteness of the area from the transport route connecting Cividale del Friuli and Carinthia – safe and that the refuge in Gradec was used only in extreme cases. There are no settlements in the central part of the Posočje region that could be reliably dated into the final phase of the late Late Antiquity. There is a high Sl. 1.27: Gradec pri Drežnici. Poskus zarisa obsega naselbine po lidarskem posnetku (© ZRC SAZU). Fig. 1.27: Gradec near Drežnica. Mapping the settlement on a lidar data visualization (© ZRC SAZU). nacija ilovnatega nasipa in kamnitih lomljencev ali celo z malto zidana konstrukcija. Kolikor je mogoče soditi zgolj po površinskem ogledu, se zdi najverjetnejša prva možnost. Površina v notranjosti je umetno zravnana in kaže več različno širokih teras, ki postopoma padajo proti severozahodu (sl. 1.27). Nekatere skromne površinske najdbe (predvsem keramike) naselje opredeljujejo v poznoantično obdobje. Odsotnost sledov na površini, ki bi nakazovali prisotnost trajnih oblik zgradb na izredno majhni površini postojanke, kot tudi skromne površinske najdbe nakazujejo predvsem naravno dobro 44 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity zavarovano kratkotrajno pribežališče. Pomisliti bi smeli tudi na utrdbo, ki je v primeru potrebe varovala oziroma zapirala pot iz doline Soče proti Drežnici, saj je tu eden pomembnejših dohodov v Drežniški kot. Glede na veliko prostranost območja pod Krnom bi smeli domnevati tu raztresene zaselke ali posamezne domačije staroselskega prebivalstva, ki se je sem umaknilo v času pozne antike. Nedvomno pa so prebivalci, ki so uporabljali to utrjeno naselje, svoje središče videli na bližnjem Tonovcovem gradu, ki je v zračni črti oddaljen le dober kilometer. Ta utrjena naselbina posredno nakazuje staroselsko enklavo v Drežniškem kotu, kar dodatno potrjujejo tudi izsledki folklorističnih, imenskih in drugih raziskav (Kuret 1972; 1989; Volarič 2008). Tretja zanesljiva poznoantična naselbina Gradec pri Logjeh leži v Breginjskem kotu tik ob italijanski meji (Osmuk 1985d; Ciglenečki 1997b, 25). Hrib (višine 446 m n. m.), ki je že po naravi odlično zavarovan s strminami in na vzhodni strani celo z navpično skalno steno, leži v sotočju Nadiže in potoka Legrado (sl. 1.28). Dostop nanj je mogoč le z južne smeri čez daljše sedlo in strmo pobočje. Širina na vrhu zelo izpostavljenega grebena je le redko kje več kot 20 m, v dolžino pa meri okoli 120 m. Vidni so neznatni sledovi izravnav terena, na dveh mestih je mogoče slutiti prostor za improvizirane lesene konstrukcije. Zaradi svoje naravno zavarovane lege je v še danes redko poseljenem svetu odlično zatočišče. Okopa na vrhu ni videti, v večjem delu tudi ni bil potreben (sl. 1.29). Pred nedavnim najdena fibula (Osmuk 1999, 64–66; 2001, 47) dokazuje poseljenost v 6. st., zato bi, podobno kot na Gradcu pri Drežnici, tudi tu smeli domnevati pribežališče iz poznoantičnega obdobja. To namembnost posredno nakazujejo izredno skromne površinske najdbe. Zdi se, da so bili prebivalci zaradi odmaknjenosti celotnega območja od prometne žile Čedad–Koroška dovolj varni in so zato pribežališče na Gradcu uporabljali le v izjemnih primerih. V srednjem delu Posočja dokazanih naselbin iz sklepne faze poznoantičnega obdobja ne poznamo. Zelo verjetna se zdi naselbina na že prej opisani Sv. Katarini nad Novo Gorico (sl. 1.21, 1.22), kjer pa sondiranja niso nedvoumno potrdila obstoja te faze (Svoljšak 1990, 43; Osmuk 1997b). Vendar bi ga smeli upravičeno pričakovati: takrat že obstoječe močno obzidje, ki je bilo utrjeno z vsaj štirimi izstopajočimi stolpi, je pravzaprav značilnost prav posebnega tipa poznih postojank, kakršne smo dobro spoznali pri raziskovanju Korinjskega hriba nad Velikim Korinjem (Ciglenečki 1985). Tudi Rutarjeva omemba s “cementom zidanih grobov” v povezavi s postojanko na Sv. Katarini bi utegnila kazati na naslednji značilen element poznih strateško umeščenih utrdb: v doslej raziskanih so bile kot največja značilnost poleg močno utrjenega obzidja s stolpi tudi manjše zgodnjekrščanske cerkve, ob katerih so zelo pogoste zidane grobnice (Korinjski hrib, Zidani gaber). Zanimiv je tudi pri Bavčarju objavljen podatek o leta 1600 najdenih zlatih Sl. 1.28: Gradec pri Logjeh. Pogled na najdišče z južne strani. Fig. 1.28: Gradec near Logje. A view of the site from the south. Sl. 1.29: Gradec pri Logjeh. Poskus zarisa obsega naselbine po lidarskem posnetku (© ZRC SAZU). Fig. 1.29: Gradec near Logje. Mapping the settlement on a lidar data visualization (© ZRC SAZU). 45 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity kovancih, ki se nanaša na Sv. Katarino (Svoljšak 1990, 35–37). Ti bi, podobno kot smo opozorili pri Molidi, morda kazali na poznoantično zakladno najdbo, kakršne so bile odkrite v mnogih utrdbah na strateških mestih. S sondiranji pridobljene skope najdbe teh domnev še ne potrjujejo, čeprav je med grobo keramiko nekaj fragmentov, ki imajo dobre analogije prav v gradivu najpoznejše faze poznoantičnih višinskih postojank v Sloveniji (Tonovcov grad, Ajdna). Težava pri tovrstni določitvi je predvsem v tem, da je bila utrdba poseljena še pozneje v slovanskem obdobju, ko so uporabljali zelo sorodno keramično gradivo. Nekoliko zunaj obravnavanega območja, že globoko v Vipavski dolini, pa leži še ena izredno dobro ohranjena poznoantična višinska postojanka – Sv. Pavel nad Vrto vinom, ki odlično dopolnjuje podobo zadnjih antičnih naselbin (Svoljšak 1985). Njene značilnosti so daleč po dolini vidna, izpostavljena lega, še danes visoko ohranjeni poznoantični zidovi in delitev naselbine v dva dela (sl. 1.30, 1.31). Mogočna skalna utrdba leži na južnem pobočju Čavna, na n. m. v. 525 m in je nad okolico dvignjena 10–30 m. Zgrajena je bila na prepadnem skalnem osamelcu, na vrhu katerega je pribl. 500 m dolg in 70–130 m širok plato. Na treh straneh je naravno zavarovana s skalnimi stenami. Nekoliko bolj dostopna je severna stran, kjer je bil med skalami prirejen tudi vhod v naselbino. Manjša sondiranja žal niso razkrila zidane arhitekture v takšni meri, da bi jo lahko primerjali z drugimi sočasnimi objekti. Vendar že njena nenavadna ureditev teras kaže, da je bila naselbina zasnovana drugače. Postojanka je z 2,5 m širokim prečnim zidom razdeljena na dva dela. Nekoliko slabše zavarovani severni del je bil intenzivneje poseljen, južni pa kaže le občasno poselitev, zato je najbrž rabil za pribežališče. Ob robu skal je še dobro ohranjen obrambni zid, ki je bil speljan tudi nad prepadnimi stenami. Širok je 1 m, ob njem je vodila cesta. Pri sondiranjih leta 1966 sta D. Svoljšak in P. Petru ugotovila stavbne ostaline na treh mestih. Južno od cerkve sta našla manjšo stavbo trapezoidnega tlorisa, ki je bila zgrajena iz slabega zidu, debelega 55 cm. Del večje in boljše grajene stavbe je bil najden tik za današnjo cerkvijo. Ostanek zidu tretje stavbe so izkopali na terasah, ki ležijo na skrajnem severu utrdbe in so bile – sodeč po vidni izoblikovanosti terena na površini – osrednji prostor poselitve. Posebno zanimiv je vodni stolp, ki leži tik pod naselbino in je velik 6,7 x 10 m (Petru 1972, 359–361). S severa se v njegov bazen spuščajo stopnice, ki so deloma vsekane v živo skalo. Med najdbami je najštevilneje zastopana groba kuhinjska keramika, ki najdišče časovno dobro umešča v 5. in 6. st. Arheološko doslej nepreverjeni so ostanki domnevne naselbine Na kapli v Podmelcu (sl. 1.32). V neposredni bližini manjšega grobišča, ki ga pridatki datirajo v 6. st., so namreč vidni obrambni okopi, ki jih je v skico najdišča vrisal V. Šribar (Šribar 1967). Teh Sl. 1.30: Sv. Pavel nad Vrtovinom. Pogled na skalni plato z južne strani. Fig. 1.30: Sv. Pavel above Vrtovin. A view of the rocky plateau from the south. probability that a settlement existed on the previously mentioned Sv. Katarina above Nova Gorica (Figs. 1.21, 1.22), even though the trenching did not confirm its existence in this phase (Svoljšak 1990, 43; Osmuk 1997b). However, this phase is to be expected: there was a strong wall fortified by at least four protruding towers at the time, and this is a characteristic of a special type of late posts that have been well researched and documented during the research of Korinjski hrib above Veliki Korinj (Ciglenečki 1985). Even Rutar’s mention of ‘graves built with the use of concrete’ in relation to the post on Sv. Katarina could indicate the next typical element of the late strategically placed forts: in the ones researched so far the most common characteristic (apart from the strongly fortified walls and towers) is represented by small Early Christian churches that were often accompanied by tombs (Korinjski hrib, Zidani gaber). Interesting is also the data that was published in Bavčar’s book and deals with the gold coins linked to Sv. Katarina that were discovered in 1600 (Svoljšak 1990, 35-37). Similar to the ones found at Molida these coins could indicate a Late Antique hoard as is the case at numerous other fortifications on strategic positions. The modest finds discovered through trenching do not confirm these assumptions, even though a few pottery fragments that can be compared to the coarse wares from the latest phase 46 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity of Slovenian Late Antique hilltop posts (Tonovcov grad, Ajdna) were discovered. The problem with this definition is that the fort was inhabited also later in the Slav period when very similar pottery was used. Somewhat outside the discussed area, deep in the Vipava Valley, lies another extremely well preserved Late Antique hilltop post – Sv. Pavel above Vrtovin – which provides an excellent addition to the image of the last settlements of the Antiquity (Svoljšak 1985). Its characteristics are an exposed and far seen position, even today high up preserved Late Antique walls and the division of the settlement into two parts (Figs. 1.30; 1.31). The mighty rock fort stands on the south slope of Čaven, at 525 m a.s.l. and is about 10-30 m above its surroundings. It was built on an isolated rock hill with an approximately 500 m long and a 70-130 m wide plateau. It is naturally protected by rock faces on three sides. Somewhat more accessible is the northern side, where the entrance into the settlement was created amongst the rocks. Unfortunately the modest trench interventions did not reveal any stone architecture to the extent where it could be compared to other buildings from the time. However, already the unusual organisation of its terraces shows that the settlement was conceived differently. The post was divided into two parts by a 2.5 metre wide wall. The somewhat poorer protected northern part was more intensively inhabited, while the southern part shows only occasional settlement, thus it was most likely used as a refuge. At the edge of the rocks stands a well preserved defensive wall that ran also above the precipitous rock face. It is 1 metre wide and a road used to run alongside it. During the 1966 trenching D. Svoljšak and P. Petru ascertained building remains in three locations. South of the today’s church they found a smaller building with a trapezoid ground plan that was built with a weak wall, only 55 cm thick. A part of the larger and better constructed building was found just behind today’s church. The wall remains of the third building were excavated on the terraces that can be found on the northernmost part of the fort and represented – according to the layout of the terrain visible on the surface – the central space of the settlement. Especially interesting is the water tower that stood below the settlement and measured 6.7 x 10 m (Petru 1972, 359-361). Stairs, some of which are carved into the bedrock, lead into its pool from the north. Coarse kitchenware is the most common amongst the finds and it dates the site into the 5th and 6th centuries. Archeologically still unverified are the remains of the assumed settlement Na Kapli in Podmelec (Fig. 1.32). In the vicinity of a smaller graveyard, dated into the 6th century according to the grave goods, ramparts can be seen. These ramparts were drawn into the sketch of the site by V. Šribar (Šribar 1967). Without further archaeological research these ramparts cannot be reliably dated to the Late Antique period, for they differ from Sl. 1.31: Sv. Pavel nad Vrtovinom. Načrt utrdbe (po Svoljšak, Knific 1976, sl. 7). Fig. 1.31: Sv. Pavel above Vrtovin. Site plan of the fortification (after Svoljšak, Knific 1976, Fig. 7). okopov ni mogoče prepričljivo umestiti v poznoantično obdobje brez arheoloških raziskav, saj se v marsičem razlikujejo od drugih doslej zanesljivo znanih najdišč. Nerazjasnjen je tudi pomen dveh grobelj, v katerih je Šribar opazil suhi zid, domačini pa so na ta dva objekta navezovali pripoved o obstoju kapele, ki naj bi bila razlog za poimenovanje najdišča. Bolj izpovedno je grobišče Na Bukovčevem brdu nedaleč vstran, kjer je bilo pri sondiranjih leta 1963 in 1965 izkopanih 5 skeletnih grobov (Šribar 1967, 379–386). Pridane pasne spone so takrat dovoljevale datacijo grobišča v 6. st. Na večji obseg grobišča kažejo omembe grobov, ki so jih izkopali 47 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.32: Podmelec. Skica domnevne naselbine (Šribar 1967, sl. 1). Fig. 1.32: Podmelec. A sketch of the presumed settlement (Šribar 1967, Fig. 1). že pred prvo svetovno vojno (Šribar 1967, 379; Svoljšak 1975, 210; Vuga 1979). Pasni sponi iz grobov sta bili pred kratkim konservirani in kažeta značilnosti poznega oblikovanja iz časa okoli leta 600 (Knific 2004, 13). Ostale poznoantične naselbine lahko, podobno kot v Podmelcu, posredno nakažemo le s po naključju znanimi manjšimi grobišči. Takšna je nekropola Na Mirih v Ljubinju neposredno ob stari poti, ki je vodila iz Tolmina čez Podmelec v dolino Bače. Manjša sondiranja Timoteja Knifica na mestu, kjer so že prej našli grobove (omenjajo zlat prstan), so pokazala na obstoj manjšega grobišča, ki ga je mogoče umestiti v drugo polovico 6. st. (Cvitkovič 1999, 40; Knific 2010). Naslednje grobišče na Ledinah v Novi Gorici je znano z območja, ki je bilo gosteje obljudeno v pozni other sites that have been reliably dated. The meaning of the two cairns in which Šribar noticed a dry wall is also unclear. The local inhabitants linked to them the story of the chapel, and supposedly this was the reason behind the name of the site. More telling is the nearby graveyard on Bukovčevo brdo, where five skeletal graves were discovered during the 1963 and 1965 trenching (Šribar 1967, 379-386). The belt buckles found in the graves dated the cemetery into the 6th century. The mentions of graves that were excavated prior to World War I indicate that the cemetery was probably larger (Šribar 1967, 379; Svoljšak 1975, 210; Vuga 1979). The two belt buckles from the graves have been recently conserved and show form characteristics that can be dated to around the year 600 (Knific 2004, 13). 48 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity antiki (glej karto najdišč pri Knific, Svoljšak 1984, 279, sl. 1). V ostanke poznorimskih zgradb je bilo vkopanih pet skeletnih grobov z značilnimi, s kamni obdanimi grobnimi jamami. Pridatki skeletnih grobov so bili skromni, vendar kažejo izrazit staroselski značaj (Osmuk 1987b; Svoljšak 1987). Pripadajoče naselje ni bilo ugotovljeno, glede na lego grobišča na bolje zavarovanem območju pa bi smeli pomisliti na manjši ravninski zaselek v neposredni bližini, morda prav tako umeščen v ostanke starejših rimskih zgradb. Izredno pomembno je, žal v veliki meri uničeno, langobardsko grobišče iz časa langobardske selitve v Italijo v Biljah. Oprema, ki nakazuje pokop več langobardskih vojščakov, opozarja na bližnjo neodkrito postojanko tik ob pomembni vpadnici v Italijo, ki bi jo morda smeli iskati celo v bližnjih podrobneje neraziskanih rimskih ruševinah (Osmuk 1978). Po temeljiti konservaciji takrat najdenih predmetov so na orožju ugotovili še nekatere dodatne podrobnosti (Knific 2004, 13). Nenavadna je najdba žganega germanskega groba z najdišča Rodne pri Tolminu. Iz pisma Karla Moserja z dne 26. 7. 1895 Naravoslovnemu muzeju na Dunaju je mogoče razbrati, da je bil na omenjeni ledini leta 1890 odkopan žgan grob, ki je ležal kar 1,80 m globoko (sl. 1.33). 9 V njem so našli bronast umbo in železen meč z – kot je mogoče razbrati z Moserjeve skice – bronasto nožnico. Zraven je bil najden tudi zlat prstan, ki pa se je izgubil. Moser je ob ogledu najdišča sklepal na posamični grob. Usoda najdb je neznana, Moser je v pismu nakazal pripravljenost, da najdbe pošlje v dunajski muzej. Moserjevo pisno poročilo zastavlja nekaj problemov: grob naj bi ležal na desni strani Soče, pol ure hoda proti Mostu na Soči. Jan Cvitkovič je najdbo lociral na rob Similar as is the case in Podmelec, other Late Antique settlements can only be indirectly indicated with the accidentally discovered small cemeteries. Such is for instance the cemetery Na Mirih in Ljubinj which is located on the old route that led from Tolmin across Podmelec and into the valley of Bača. Small trenching interventions carried out by Timotej Knific on the location at which graves have been previously found (a gold ring is mentioned), have shown the existence of a small cemetery that can be dated into the second half of the 6th century (Cvitkovič 1999, 40; Knific 2010). The next cemetery on Ledine in Nova Gorica is known from the area that was more densely populated during the Late Antiquity (see map of sites in Knific, Svoljšak 1984, 279, Fig. 1). Five skeletal graves with characteristic grave pits encircled with stones were dug into the remains of Late Roman buildings. The grave goods were modest; however they are characteristic of the autochthonous population (Osmuk 1987b; Svoljšak 1987). No accompanying settlement was ascertained, however taking into account that the cemetery was located in a secure Italic area we could consider the existence of a smaller unprotected plain settlement in the vicinity, possibly positioned on the remains of older Roman buildings. Extremely important, however unfortunately mainly destroyed, is the Lombard cemetery in Bilje that can be dated to the time the Lombards were moving to Italy. The grave goods indicate the burials of a number of Lombard soldiers and lead us to believe into the existence of a yet undiscovered post in the vicinity. This post stood alongside the important route to Italy and could have been located on the nearby not so well researched Roman ruins (Osmuk 1978). Following thorough conservation efforts of the objects discovered at the time additional details were discovered on the weapons (Knific 2004, 13). Unusual is the discovery of the Germanic cremation burial at the site Rodne near Tolmin. In Karl Moser’s letter to the Naturhistorisches Museum in Vienna, dated 26th July 1895, he mentions that a 1.80 metre deep cremation burial was excavated there in 18909 (Fig. 1.33). The grave revealed a bronze shield boss and an iron sword with – as depicted in Moser's sketch – a bronze sheath. A gold ring was discovered in the grave, however it was lost through time. When viewing the site Moser assumed it was an isolated grave. The fate of the finds is unknown, however in the letter Moser indicated his willingness to send the finds to the museum in Vienna. Moser’s report poses a few problems: the grave was supposedly positioned on the right bank of the Soča river, half an hour by foot towards Most na Soči. Jan Cvitkovič located the site on the edge of Tolmin, where the settlement mentioned in the letter is located (Cvitkovič 1999, 42). Contrary to Moser's data Rodne is situated on the left bank of Soča as well as on 9 Na grob nas je že pred leti prijazno opozoril Dragan Božič, za kar se mu na tem mestu zahvaljujemo. Sl. 1.33: Rodne pri Tolminu. Moserjeva risba predmetov iz germanskega groba (pismo K. Moserja z dne 26. 7. 1895. Hrani Naravoslovni muzej Dunaj). Fig. 1.33: Rodne near Tolmin. Moser's drawing of the objects from the Germanic grave (Karl Moser’s letter, dated 26th July 1895. Kept by the Naturhistorisches Museum in Vienna). 9 Dragan Božič told us about this grave years ago, and we would like to take this opportunity to thank him. 49 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Tolmina, kjer leži v pismu omenjeni zaselek. V nasprotju z Moserjevimi podatki pa ta ledina dejansko leži na levi strani Soče in celo na levi strani Tolminščice, ki se nekoliko južneje od tod izliva v Sočo (Cvitkovič 1999, 42). Najdišče tudi ni oddaljeno pol ure hoje od Tolmina, ampak le slab kilometer od njegovega središča. Zato pa se sklada podatek, da leži najdišče ob cesti za Sveto Lucijo (Most na Soči) in da je blizu obrežja reke, kjer so lahko pridobivali v pismu omenjeni gramoz. Cvitkovič datira grob glede na ustno mnenje A. Pleterskega v 7. st. Podrobnejši ogled Moserjeve skice pokaže, da gre nedvomno za moški grob s pridano spatho in umbom ter zlatim prstanom. Zaradi približno skiciranih predmetov ni možna njuna natančnejša opredelitev, analogije imata v širšem geografskem prostoru Barbarika kot tudi v mejah imperija na današnjem madžarskem in avstrijskem prostoru in ju je mogoče datirati v 6. st. Pomenljiva je tudi globina groba: podobna je bila ugotovljena pri najmanjši skupini najpomembnejših germanskih grobov v Dravljah (Slabe 1975, 34–36) in na Lajhu v Kranju (Stare 1980, 89; Odar 2006, 246). Nenavadno je predvsem dejstvo, da je grob žgan, kar ga razlikuje od doslej odkritih germanskih grobov na slovenskem ozemlju. Najbližje žgane grobove je mogoče zaslediti v Kajdacsu na Madžarskem, kjer so bili po mnenju I. Bóne poleg langobardskih vojščakov pokopani tudi zavezniški Sasi, ki so ohranili star običaj pokopavanja (Bóna 1970–1971, 49). Možnost, da je bil v Rodnah pokopan langobardski zaveznik, bi utegnilo dati pomemben namig glede smeri poti, po kateri so se Langobardi s svojimi zavezniki selili v Italijo. Če pritegnemo Moserjevo opažanje, da gre za posamičen grob in dejstvo, da so stare poti potekale po levi strani Soče med Tolminom in Mostom na Soči, bi smeli upravičeno pomisliti na bojevnika, ki je omagal na dolgi poti iz Panonije v Italijo, tik pred obljubljeno deželo. To je seveda le ena izmed možnosti, vendar ne povsem neverjetna glede na dejstvo, da žgani germanski grobovi v Italiji niso bili odkriti. Zato bi smeli upravičeno domnevati pokop v času pohoda Langobardov leta 568 ali pa kmalu zatem, ko so se Langobardi v tem prostoru že ustalili in za zavarovanje območja postavili manjše posadke, kot se je to potrdilo že na Tonovcovem gradu (primerjaj najdbo spathe in držaja ščita: Tonovcov grad. Najdbe, t. 11: 10 in 46: 9). Gosto razpredena mreža poznorimskih cest in tovornih poti, ki jo podrobneje predstavljamo v poglavju o poselitvi v poznorimskem obdobju, je bila uporabljena pri prehodu Langobardov iz Panonije v Italijo (glej pogl. 5.1). Med redkimi sledovi teh premikov bi utegnil biti prav ta grob eden izmed odločilnih kazalnikov (Werner 1962, 121–130). Skromno, a pestro poselitveno podobo poznoantičnega obdobja dopolnjujejo posamezne drobne poznoantične najdbe, ki jih je težko povezati s konkretnimi naselbinskimi sledovi. Takšni sta predvsem zgodnjekrščanska oljenka na ledini Pod Čahlnami in poznoantična fibula z ledine Varda v Novi Gorici, ki the left bank of Tolminščica, which runs into Soča slightly further south from here. The site is also not half an hour by foot from Tolmin, but less than 1 kilometre from its centre. On the other hand the mention in the letter that the site is positioned alongside the road towards Sveta Lucija (Most na Soči) and that it is close to the river bank where gravel was dug holds true. Taking into account the opinion passed on by A. Pleterski Cvitkovič dates the grave into the 7th century. A detailed view of Moser’s sketch shows that it was undoubtedly a male grave with a spatha, shield boss and a gold ring. Due to the rough sketches of the first two objects it is impossible for them to be precisely defined, however they have analogies in the broader geographical area of Barbaricum as well as within the Empire (on present day Hungarian and Austrian grounds) that can be dated into the 6th century. The depth of the grave is also important: a similar depth was ascertained in the smallest group of the most important Germanic graves in Dravlje (Slabe 1975, 34-36) and on Lajh in Kranj (Stare 1980, 89; Odar 2006, 246). Unusual is the fact that this is a cremation grave, which makes it different from the other Germanic graves that have been discovered in Slovenia. The closest cremation graves were discovered in Kajdacs, Hungary, where – according to I. Bóna – not only the Lombard soldiers but also allied Saxons were buried, and they preserved the old burial ritual (Bóna 1970-1971, 49). The possibility that a Lombard ally was buried in Rodne could provide an important hint as regards the direction of the route along which the Lombards and their allies moved towards Italy. If we agree with Moser that this was an isolated grave and take into account the fact that the old routes between Tolmin and Most na Soči ran on the left side of the Soča river, we could rightly assume that this was a warrior who failed to make the long journey from Pannonia to Italy, and died just prior to reaching the promised land. Of course, this is merely one of the possibilities; however, it becomes more probable if we take into account the fact that Germanic cremation graves have so far not yet been discovered in Italy. We could assume that the burial took place during the Lombard march in 568 or soon afterwards, when the Lombards settled in this area and positioned small garrisons that protected the area, as has been confirmed for Tonovcov grad (compare the spatha and the shield handle finds: Tonovcov grad. Finds, Pls. 11: 10 and 46: 9). When the Lombards were making their way from Pannonia to Italy they used the dense network of Late Roman roads and transport routes (see chapter 4.1). Amongst the rare traces of these movements this grave could represent a decisive indicator (Werner 1962, 121-130). In addition to the modest and diverse Late Antique settlement image we have the individual Late Antique finds that are hard to link with actual settlement traces. Such are for instance the early Christian oil lamp from Pod Čahlnami and the Late Antique fibula from Varda in Nova Gorica, both of which enrich the knowledge of the 50 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity Sl. 1.34: Sv. Helena v Podbeli. Pogled na naselbino z južne strani Fig. 1.34: Sv. Helena in Podbela. A view of the settlement from the south. Sl. 1.35: Sv. Helena v Podbeli. Lidarski posnetek površja (© ZRC SAZU) z opaznimi sledovi stavb (Kokalj, Zakšek, Oštir 2011, 265, sl. 1). Fig. 1.35: Sv. Helena in Podbela. Remains of structures visible on a lidar data visualization (© ZRC SAZU; Kokalj, Zakšek, Oštir 2011, 265, Fig. 1). bogatita vedenje o poselitvenem prostoru pod utrjeno postojanko na Sv. Katarini (Zavrtanik 1982). Lepo okrašena dvoramna fibula je bila najdena tudi v območju Lokavca na Banjšicah, na območju torej, kjer je že F. Truhlar glede na odmaknjeno lego in predslovanske toponime domneval poznoantično enklavo (Knific 2004, 12; Truhlar 1976). Prav tako je nekaj najdišč, ki jih z dosedanjimi raziskavami in obhodi ni bilo mogoče povsem zanesljivo settlement area under the fortified post on Sv. Katarina (Zavrtanik 1982). A nicely decorated equal-arm fibula was found also in the area of Lokavec on Banjšice, i.e. in an area for which F. Truhlar assumed a Late Antique enclave – due to the remote position and the pre-Slav toponyms (Knific 2004, 12; Truhlar 1976). The research carried out so far failed to provide precise data that would place certain sites into this period: however, they should be mentioned because of certain 51 1.5 Posočje v poznoantičnem času 1.5 The Posočje area during Late Antiquity umestiti v to obdobje: zaradi nekaterih najdb in njihove lege ter oblikovanosti površja pa se vendarle zdi smiselno opozoriti nanje. Takšna je Sv. Helena v Podbeli, od koder je bila pred kratkim pridobljena tudi železna triroba puščična ost (Knific 2004, 9). Zravnan plato na naravno odlično zavarovanem hribu nakazuje na površju sledove stavb. Ti so se posebej dobro pokazali po uporabi metode zračnega laserskega skeniranja (lidar), ki smo ga izvedli leta 2007 (sl. 1.34, 1.35): pri tem je bilo mogoče na površini zaznati več obrisov stavb, ki bi utegnile soditi v poznoantično obdobje (Kokalj et al. 2011). Povsem vprašljiv je obstoj poznoantične postojanke na Kozlovem robu nad Tolminom, kjer naj bi pri izkopavanjih leta 1964 našli poznoantično žlico (Vuga 1970, 172) oziroma “nekaj elementov, ki kažejo na poznoantično poselitev gradu” (Svoljšak 1970–1971, 154). Novejše raziskave srednjeveškega gradu tega doslej niso potrdile. Fragmenti poznoantične oziroma zgodnjesrednjeveške keramike naj bi bili najdeni tudi v Selih nad Podmelcem (Osmuk 1985f, 297), a je zaradi majhnih in slabo določljivih delcev keramike težko potrditi datacijo. Drugačna je situacija v Kanalu ob Soči, kjer so na obrežju Soče našli značilno poznoantično trirobo puščično ost, že prej pa je bil v Gorenji vasi, ki leži v neposredni bližini, najden Konstantinov zlatnik (Žbona Trkman 1981; Kos 1988, 66, št. 29). Terenski pregledi doslej niso pojasnili lege naselbine. Glede na zgodnejšo varianto najdene puščične konice bi obe najdbi utegnili kazati na obstoj poznorimske postojanke ali tudi na naselje, ki je bilo zaradi pomembne strateške lege ob prehodu čez reko obljudeno ali celo utrjeno za daljši čas v obdobju pozne antike. Povsem neznatni in neizpovedni so arheološki sledovi s konca 6. stoletja in naslednjih dveh stoletij, ko naj bi – glede na sporočila pisnih virov – postopoma naseljevali Posočje slovanski prišleki (Štih 1999). Prav to obdobje je doslej najskrivnostnejše in si o njem ni mogoče ustvariti zanesljive podobe. Domnevati je mogoče, da so ponekod staroselci vztrajali še v 7. st., o čemer pričajo posamezni kosi nakita in orožja na Sv. Pavlu nad Vrtovinom, Sv. Katarini nad Novo Gorico, v kobariškem območju pa na Tonovcovem gradu. Kdaj natanko je življenje v njih zamrlo, ni mogoče natančneje ugotoviti. Ker je bilo to območje sestavni del langobardske države, omenjene postojanke niso bile tako izpostavljene kot tiste v vzhodni in osrednji Sloveniji, nedvomno pa opuščene kmalu po koncu antičnega obdobja (Svoljšak, Knific 1976, 80–81). O tem zgovorno priča skromna prisotnost ostalin zgodnjesrednjeveških prebivalcev na Tonovcovem gradu, ki so izkoristili ruševine deloma že podrtih antičnih zgradb, da so si v njih uredili zasilna bivališča. finds, their position and terrain. Such is Sv. Helena in Podbela, where an iron trefoil arrowhead has been recently discovered (Knific 2004, 9). The levelled plateau on a naturally excellently protected hill indicates traces of buildings on its surface. These were especially visible when the laser scanning method – introduced in 2007 – was used (Figs. 1.34, 1.35): with this method a number of building outlines could be seen and they could belong to the Late Antique period (Kokalj et al. 2011). The Late Antique post on Kozlov rob above Tol min is debatable, even though a Late Antique spoon was supposed to have been found at the site during the 1964 excavations (Vuga 1970, 172) as were ‘certain elements that indicate a Late Antique settlement under the castle’ (Svoljšak 1970-1971, 154). Until now this has not been confirmed by the newer research of the medieval castle. The fragments of the Late Antique or Early Medieval pottery were supposedly also found in Sela above Podmelec (Osmuk 1985f, 297), but due to the small and poorly definable fragments it is hard to confirm its date. In Kanal ob Soči the situation is different, for here a typical Late Antique trefoil arrowhead was found on the Soča riverbank, and even before that a gold coin of Constantine I was found in Gorenja vas, which lies in the immediate vicinity (Žbona Trkman 1981; Kos 1988, 66, No. 29). So far the terrain inspections have not explained the position of the settlement. Taking into account the earlier variant of the discovered arrowhead both finds could indicate the existence of a Late Roman post or even a settlement that was – due to its important strategic position at the river crossing – populated or even fortified for a longer time in Late Antiquity. Extremely minute and non-explanatory are the archaeological traces from the end of the 6th century and the following two centuries, when – as gathered from written sources – the Posočje area was gradually inhabited by the Slav newcomers (Štih 1999). So far this period remains the most mysterious and it is impossible to establish a clear image of it. It can be assumed that in certain places the autochthonous population persisted into the 7th century, which is indicated by the individual pieces of jewellery and weapons discovered at Sv. Pavel above Vrtovin, Sv. Katarina above Nova Gorica, and at Tonovcov grad in the Kobarid area. It is impossible to ascertain the precise time in which these settlements were abandoned. Because this area was a constituent part of the Lombard state, the aforementioned posts were not as exposed as those in eastern and central Slovenia and were undoubtedly abandoned soon after the end of Antiquity (Svoljšak, Knific 1976, 80-81). This is clearly shown by the modest presence of the remains from the later Early Medieval settlers who settled at Tonovcov grad and made best use of the ruins of the partially demolished Antique buildings in which they created their temporary dwellings. 52 1.6 Mreža poti 1.6 Path network Benjamin Štular 1.6.1 Uvod 1.6.1 Introduction V arheologiji se vedno znova srečujemo z vprašanji, povezanimi z vlogo posameznega najdišča v prostoru. Med najpogostejšimi je vprašanje poti, ki so določeno najdišče povezovale s pomembnimi kraji v prostoru. Pri tem se v arheologiji najpogosteje spopadamo s pomanjkanjem podatkov. V Sloveniji tako arheološko dokumentiranih poznorimskih ali poznoantičnih poti skorajda ne poznamo. Poznamo le primera poti v naselbini (Ajdovski gradec nad Vranjem: Petru, Ulbert 1975) oziroma na grobišču (Bled - Pristava: Pleterski 2008a, 75–95) in študijo o zgodnjesrednjeveških gospodarskih poteh v Blejskem kotu (Štular 2006a). Na drugi strani poznamo za to obdobje regionalne ali nadregionalne študije, pri katerih sklepamo o trasah poti na podlagi bližine naselbin in najdb (Ciglenečki 1985; 2007). Na tem mestu nas zanima vmesen, mikroregionalni pogled. Natančneje, gre za poti, ki so naselbino na Tonovcovem gradu povezovale z okolico, predvsem s sočasnimi centralnimi kraji (npr. Kranj, Čedad). Študijsko območje smo izbrali tako, da smo zajeli vse tri verjetne dostope do Tonovcovega gradu z juga in zahoda, torej dolino reke Soče južno od Kobarida ter dva kraka doline Nadiže od izvira do prehoda v Furlansko nižino. Na severu smo območje zamejili s križiščem, na katerem se razcepijo poti na različne alpske prelaze (sl. 1.36). Ker neposrednih arheoloških podatkov o poteh nimamo, se lahko opremo le na historično izpričane poti in na prostorske analize z daljinskim zaznavanjem ter geografskimi informacijskimi sistemi (GIS). In archaeology we constantly deal with issues that are linked to the role of a site in its surroundings. Amongst the most common is the issue of paths that linked a certain site to the important places in its surroundings. However, while performing this archaeology (as a scientific field) usually does not have sufficient data at its disposal. For instance, there are almost no documented Late Roman or Late Antique paths in Slovenia. The only excavated ones are the paths within the settlement (Ajdovski Gradec above Vranje: Petru, Ulbert 1975) or the burial site (Bled - Pristava: Pleterski 2008a, 75-95). There is also the study on the Early Medieval economic paths in Blejski kot (Štular 2006a). On the other hand certain regional and interregional studies assume that the paths and roads ran close to the settlements (Ciglenečki 1985; 2007). In this chapter we are interested in the intermediary, micro-regional aspect. To be more precise we are interested in the paths that linked the settlement at Tonovcov grad with its surroundings, especially with the main settlements of the time (Kranj, Cividale del Friuli). The study area was selected so that it included all three likely southern and western accesses to Tonovcov grad, i.e. the valley of the river Soča to the south of Kobarid and the two branches of the Nadiža Valley from the spring to the point where it flows into the Friuli plain. In the north the area is limited by the crossroads at which the various paths leading from the Alpine passes meet (Fig. 1.36). As there are no archaeologically documented paths, the historically mentioned paths and the spatial analysis performed with remote sensing and the use of Geographical Information Systems (GIS) were the only sources at the disposal for this study. 1.6.2 Zgodovinski viri Najstarejši relevanten podatek, izpričan v zgodovinskih virih, se nanaša na t. i. Bovško pot, ki je v srednjem veku povezovala Furlanijo in Koroško. Približen potek poti je znan: čez Čedad je pot vodila proti severu po dolini Nadiže. Pri kraju Robič je zapustila Nadižo in se usmerila proti vzhodu, pri Kobaridu dosegla Sočo in ob njej potekala proti Bovcu ter dalje čez Predel in 1.6.2 Historic sources The oldest relevant data mentioned in historic sources is linked to the so-called Bovec route, a medieval route 53 1.6 Mreža poti 1.6 Path network Sl. 1.36: Tonovcov grad, študijsko območje. Fig. 1.36: Tonovcov grad, the study area. Trbiž. Na to pot so se z južne strani skozi Tolmin in po dolini reke Bače navezovale tudi poti proti osrednji Kranjski, bodisi čez prelaz Sorica v Bohinj, bodisi po dolini Selške Sore proti Škofji Loki (Gestrin 1987, 97; glej tam navedeno literaturo). Bovška pot je prvič omenjena v pisnem viru leta 1326. Ohranil se je tudi zapis, da je med letoma 1399 in 1404 mesto Čedad začelo graditi cesto čez Predel. Na odseku Trbiž–Bovec so po trasi predhodne tovorniške poti, primerne za ljudi in tovorne živali, zgradili cesto, po kateri so lahko potovali vozovi (Gestrin 1987, 97; Rajšp 1994, 46–47; glej tam navedeno literaturo in vire). Za neposredno okolico Tonovcovega gradu je ta podatek posredno lahko zanimiv. Gradnja ceste namesto tovorniške poti od Bovca dalje bi ne bila smiselna, če v tem času že ne bi bilo ceste mimo Kobarida do Bovca. Ta podatek torej posredno lahko uporabimo kot terminus ante quem za datacijo ceste, ki vodi od Kobarida proti that connected Friuli plain and Carinthia. A rough outline of the route is known: it led from Cividale del Friuli towards the north following the valley of the Nadiža river, through the settlements of Staro selo and Kobarid into the valley of the river Soča to Bovec and further across the Predel/Predil pass and to Treviso. This route was joined by routes that led towards central Carniola via Tolmin and the valley of the river Bača, either across Sorica pass in Bohinj, or along the Selška Sora Valley towards Škofja Loka (Gestrin 1987, 97; see cited literature). The first written record of the Bovec route dates back to 1326. Also preserved is the record that states that between 1399 and 1404 the town of Cividale del Friuli started to build a road across Predel. A road suitable for horse drawn carriages was built in the section between Treviso and Bovec, that was previously used only as a track for people and transport animals (Gestrin 1987, 97; Rajšp 1994, 46-47; see cited literature and sources). 54 1.6 Mreža poti 1.6 Path network Sl. 1.37: Tonovcov grad, lidarski posnetek (© ZRC SAZU) prikazan s senčenjem. Puščici kažeta mesti, kjer je bila srednjeveška obsoška pot vsekana v skalo. Fig. 1.37: Tonovcov grad, the lidar data visualization (© ZRC SAZU) shown with hillshading. Arrows are pointing towards the places where the Medieval road was cut into rock. severu neposredno ob Soči in je na dveh mestih vsekana v skalno pobočje (sl. 1.37). Zanimiv je tudi politični kontekst izgradnje t. i. obsoške ceste, ki so jo vzpostavili v 16. stoletju z namenom neposredne povezave med Gorico in Trbižem. Po več kot pol stoletja načrtovanj so po posredovanju nadvojvode Karla zahtevno gradnjo končali leta 1587. Regionalno je bila izgradnja ceste povezana s povečevanjem prometa čez Gorico na račun Čedada, predvsem pa je s tem habsburška monarhija želela povezati Trst z Notranjo Avstrijo v celoti čez svoje ozemlje, saj je krajša in manj zahtevna Bovška pot deloma potekala po ozemlju Beneške republike (Rajšp 1994, 46–48; glej tam navedeno literaturo in vire). Kontekst dogodkov v 16. stoletju jasno kaže, da je bilo za izgradnjo obsoške ceste od Gorice do Kobarida potrebno posredovanje močne centralne oblasti, ki so jo pri tem vodili strateški interesi. Hkrati je bila neizogibna posledica izgradnje ceste slabitev gospodarskega položaja Čedada. Ali, skozi poznoantično prizmo, Čedad kot močno središče negira obstoj pomembne regionalne povezave This information could be interesting for the case of Tonovcov grad. The construction of the road instead of merely a track from Bovec onwards would not make any sense if the road from Kobarid to Bovec would not have already existed. This data can thus be used as a terminus ante quem for dating the road that leads from Kobarid towards the north along the river Soča and that cuts into the rocky slopes at two points (Fig. 1.37). Also interesting is the political context behind the 16th century construction of the so-called Soča bypass road. This ran along the valley of the river Soča from Gorica to Kobarid from where it continued along the previously described Bovec path across Bovec and Predel all the way to Treviso. Following more than half a century of planning and the intervention by Archduke Charles, the demanding construction was completed in 1587. The road was a result of the desire of the Habsburg monarchy to link Trieste and Inner Austria across its own territory, for a part of the shorter and less demanding Bovec path ran across territories belonging to the Venetian Republic. As a result of this 55 1.6 Mreža poti 1.6 Path network na trasi Gorica–Kobarid. Nasprotno velja za povezavo Kranj–Škofja Loka–Tolmin–Kobarid, ki je bila nujna ob soobstoju Kranja in Čedada kot centralnih krajev. road construction the traffic through Gorica increased at the expense of Cividale del Friuli (Rajšp 1994, 46-48; see cited literature and sources). The events in the 16th century clearly show that the construction of the Soča bypass road from Gorica to Kobarid could only take place as a result of the mediation by a strong centralised power that was lead by strategic interests. At the same time the economic weakening of Cividale del Friuli was an unavoidable consequence of the road construction. Or, to look at it through the prism of Late Antiquity, Cividale del Friuli could only exist as a strong centre if there was no important regional link between Gorica and Kobarid. The contrary holds true for the connection Kranj - Škofja Loka - Tolmin - Kobarid, which was a necessity in order for Kranj and Cividale del Friuli to coexist as powerful centres. 1.6.3 Analiza optimalnih poti V arheologiji je iskanje idealnih poti z GIS-i uveljavljena metoda (za pregled glej npr. Connolly, Lake 2006, 252–256). Za analizo morajo biti izpolnjeni vsaj trije pogoji: dovolj kakovostne kartografske podlage (predvsem relief), znana izhodišče in cilj ter znano transportno sredstvo. Prednost uporabe te metode v arheologiji je, da analizo na dovolj razgibanem reliefu lahko opravimo brez kakršnih koli arheoloških podatkov. Šibkost pa je v natančnosti. Za našo analizo to pomeni, da rezultat lahko razumemo kot nekaj deset metrov širok koridor in ne kot dejansko traso poti. Kakovost kartografskih podlag je za analizo optimalnih poti ključen dejavnik (Podobnikar 2009). V primeru pričujoče analize smo razpolagali s podatki nadpovprečne kakovosti. Za osrednji del izbranega študijskega območja smo imeli na voljo izjemno natančen digitalni model reliefa z osnovno celico 0,5 metra in absolutno višinsko in položajno natančnostjo med 0,1 in 0,2 metra (dalje DMR 0,5). Ta je bil izdelan iz podatkov, zajetih z namenskim zračnim laserskim skeniranjem (ang. Light Detection And Ranging; lidar), z algoritmom REIN (Kobler et al. 2007; Kokalj, Oštir, Zakšek 2008). Za celotno študijsko območje smo uporabljali kakovosten digitalni model višin z osnovno celico 12,5 metra (DMV 12,5; Podobnikar 2003). Kot podlage smo uporabljali tudi historične zemljevide, Jožefinske vojaške zemljevide (Rajšp 1997, sekcije 132, 133, 154) in zemljevide Franciscejskega katastra (katastrske občine Kobarid, Staro selo, Ladra, Drežnica in Trnovo). Prvi so bili v merilu 1:28800 izdelani v letih 1763–1787, drugi pa večinoma v drugi četrtini 19. stoletja (arhivirani so dokumenti od 1811 do 1880) v merilu 1:2880. Na podlagi podatkov v pisnem delu omenjenih vojaških zemljevidov iz druge polovice 18. stoletja smo pridobili pomembne podatke o premikanju po pokrajini pred modernimi infrastrukturnimi posegi. Reka Idrija in potoka Učja ter Boka (sl. 1.36) so opisani kot neprehodni, enako Soča na območju soteske med Kobaridom in Trnovim. Območja prodnih plitvin reke Soče južno in severno od soteske so opisana kot prehodna, vendar niso primerna za potovanje vzdolž toka. Ta pogoj smo v analizi simulirali tako, da smo ta območja označili za 20-krat težje prehodna kot kopno z enakim reliefom. Drugače pa smo ravnali v primeru struge Nadiže, po kateri je bilo “ob nizki vodi … moč iti večinoma po vodi iz Kreda in Robiča proti Podbeli, po potoku Nadiža pa tudi na Beneško” (transkripcija in prevod vira v Rajšp 1997, 11). 1.6.3 Analysis of optimal routes In archaeology it is common practice to search for ideal routes with the aid of GIS (for an overview see e.g. Connolly, Lake 2006, 252-256). At least three preconditions have to be fulfilled before an analysis can be performed: the maps need to be of sufficient quality (especially the digital terrain model), the starting point and the final destination of the route have to be known as does the form of transport. The advantage of using this analysis is that it can be performed on a rough relief without any direct archaeological data. The weakness lies in its precision. In our analysis this means that the result can be understood as a corridor a few tens of metres wide rather than the actual route. The quality of the digital terrain model represents a key factor in the analysis of the optimal routes (Podob nikar 2009, 25-29). We had above average quality data, an extremely precise digital terrain model with a basic cell of 0.5 metre and an absolute altitude and positional precision between 0.1 and 0.2 metre (henceforth DTMDTM 0.5) at our disposal. This was produced from the data obtained from a Light Detection And Ranging (LiDAR) system with an REIN algorithm (Kobler et al. 2007; Kokalj, Oštir, Zakšek 2008). For the entire study area we also used the digital terrain model with a basic cell of 12.5 metres (DTM 12.5; Podobnikar 2003). Historical maps – the first military mapping also known as Joseph II military maps (Rajšp 1997, sections 132, 133, 154) and the maps of the Franciscan cadastre (cadastre municipalities of Kobarid, Staro selo, Ladra, Drežnica and Trnovo) provided an additional source of data. The first were produced between 1763 and 1787 at a scale 1:28800, while the second were mainly produced in the second quarter of the 19th century (the documents were archived between 1811 and 1880) and were mapped at a scale 1:2880. 56 1.6 Mreža poti 1.6 Path network Tok reke Idrije in vseh njenih pritokov je bil v začetku 20. stoletja zaradi melioracije povsem spremenjen. Prvotno strugo smo rekonstruirali na podlagi kartografskega dela Franciscejskega katastra in z vizualno analizo DMR 0,5. Z analizo DMR 0,5, v katerem smo iskali konkavna in konveksna območja, nam je uspelo določiti tudi poti “okrog Sužida (po katerih) je ob mokrem vremenu moč voziti le z lahkimi vozovi”, kot jih je opisal podporočnik liškega regimenta Bodoky, kartograf tega območja (transkripcija in prevod vira v Rajšp 1997, 7). Podatke smo analizirali z algoritmom, ki je bil razvit za iskanje optimalnih poti med arheološkimi najdišči ob predpostavki potovanja peš ali s tovornimi živalmi (Podobnikar, Tecco Hvala, Dular 2004). Ta algoritem se je izkazal kot zelo uspešen pri uporabi z DMV 12,5 na podobno razgibanem terenu (Štular 2006b). Poleg DMV 12,5 smo na t. i. karti trenja kot vhodni podatek upoštevali tudi zgoraj opisane dejavnike, ki so omejevali gibanje po pokrajini: neprehodne in prehodne vode, zelo strma skalna pobočja soteske Soče ter območje močvirnih tal. Pri natančnem umeščanju izhodiščnih in končnih točk smo upoštevali načela dobre prakse (Podobnikar 2009). Točke smo umestili na rahlo konkavne mikrolokacije, saj v nasprotnem primeru lahko pride do velikih odstopanj v končnem rezultatu zaradi poteka na prvih 50 do 100 metrih. Dodaten ukrep proti tovrstnim napakam in hkrati preverjanje kakovosti smo dosegli tako, da smo vsako pot izračunali v obe smeri. Razhajanja poti “tja in nazaj” so zaradi anizotropnosti algoritma, ki razlikuje med vzponom in spustom, pričakovana (Podobnikar 2009). Vendar koherentnost našega rezultata dokazuje, da niti specifičnost algoritma niti izbira izhodiščnih točk nista odločilno vplivali na rezultat. Rezultati se ujemajo s pričakovanji (sl. 1.38). Vse poti po dolini Nadiže potekajo po samem rečnem koritu, enako kot še v 18. stoletju. Od struge Nadiže do Kobarida se pot še pred vasjo Staro selo utiri na stik med ravnino in pobočjem Starijskega vrha. V samem Kobaridu se pot povzpne do zahodnega roba prazgodovinske in rimskodobne naselbine Gradič. Od tam se po manjših dolinah in pobočjih nadaljuje mimo Tonovcovega gradu in se spusti do Soče šele pri Srpenici. Pot prečka Sočo na istem kraju, pri naselju Žaga, kot v 18. stoletju. From the written data accompanying the aforementioned military maps we obtained important data as regards the movements through the landscape prior to the modern infrastructure interventions. The river Idrija and the streams Učja and Boka (Fig. 1.36) were described as impassable, as was the river Soča in the gorge between Kobarid and Trnovo. The gravel shallow areas of the Soča river (to the south and north of the gorge) were described as passable but inappropriate for travelling. In our attempt to simulate this condition we marked these areas as 20 times harder to pass than land with the same relief. We applied a different approach in the case of the Nadiža river, where it was ‘at low waters... possible to go mainly on water from Kred and Robič towards Podbela, and along the stream all the way to the Venetian lands’ (transcription and Slovenian translation of the source in Rajšp 1997, 11). At the beginning of the 20th century the Idrija river bed (as well as all its contributories) was completely changed due to melioration. The course of the original river-bed was reconstructed with the maps from the Franciscan cadastre and a visual DTM 0.5 analysis. With the DTM 0.5 analysis, in which we sought for concave and convex areas, we managed to define the roads ‘around Sužid (on which) it was possible to travel only with light wagons at wet weather ‘, as was described by Bodoky, the cartographer and second lieutenant of the Lika regiment (transcription and Slovenian translation of the source in Rajšp 1997, 7). We analysed the above described data with an algorithm that was developed especially for searching for optimal routes between archaeological sites, assuming that the travel was performed on foot or by transport animals (Podobnikar, Tecco Hvala, Dular 2004). This algorithm has proven itself extremely successful on a similarly diverse terrain when combined with the use of a DTM 12.5 (Štular 2006b). Apart from DTM 12.5 we also took into account (on the so-called friction map) the previously described factors that restricted movement across the landscape: impassable and passable waters, very steep rocky slopes of the Soča gorge and wetlands. Following the precise positioning of the starting and arrival points we took into account the principles of good practice (Podobnikar 2009, 27). The points were placed into slightly concave micro-locations, for otherwise great differences in the final result would occur due to the first 50 or 100 metres of the route. An additional measure performed in order to avoid mistakes (as well as a way to validate the results) was achieved by calculating every route in both directions. Due to the anisotropity of the algorithm, which takes into account the difference between an incline and a decline, the differences in the routes ‘there and back’ are to be expected (Podobnikar 2009, 27-19). However, the coherence of our result proves that neither the specifics of the algorithm, nor the selection of the starting points had a decisive influence upon the result. 1.6.4 Mrežna analiza Za razliko od analize optimalnih poti so mrežne analize (ang. network analysis) poti v arheologiji redke (npr. Štular 2008; za metodo glej Conolly, Lake 2006, 234–252). Razlog je zelo preprost. Mrežne analize so zasnovane za reševanje problemov, kjer iščemo rešitev med mnogimi možnimi rešitvami. Mrežna analiza, kakršno smo uporabili, med znanimi potmi išče najkrajšo povezavo med izbranim izhodiščem in ciljem ter more57 1.6 Mreža poti 1.6 Path network Sl. 1.38: Tonovcov grad, rezultati analize optimalnih poti. Fig. 1.38: Tonovcov grad, optimal pathway analysis. bitnimi postanki. V arheoloških raziskavah pa, vsaj kar se tiče gibanja v prostoru, največkrat nimamo nobene vnaprej znane rešitve. Prednost mrežne analize v arheološki raziskavi je natančnost rezultata, saj kot rezultat dobimo natančno traso poti. Slabost pa je, da v analizi ne moremo upoštevati morebitnih neprepoznanih poti. Vendar v primeru tako kvalitetnega vira, kot je DMR 0,5, pričakujemo, da so potencialno neprepoznane le poti, ki so bile trasirane brez posega ali le z minimalnim posegom v relief. Potencialne tovorniške poti pa si na tako razgibanem terenu ne moremo predstavljati brez določenih posegov in vplivov na relief, predvsem pri prečenju pobočij. V primeru okolice Tonovcovega gradu smo z analizo omenjenega DMR 0,5 prepoznali več kot sto različnih poti v skupni dolžini 305 kilometrov (sl. 1.39). Podatek je bil torej kot nalašč za t. i. mrežno analizo. V mrežo poti smo vstavili izhodišča in cilje, tako da smo točkam, The results proved to be in line with expectations (Fig. 1.38). All routes along the Nadiža Valley run along the river bed, which remained the case as late as the 18th century. Between the Nadiža river bed and Kobarid the route sticks to the border between the plain and the slope of Starijski vrh. In Kobarid the route climbs to the western edge of the prehistoric and Roman settlement of Gradič. From there it continues along small valleys and slopes past Tonovcov grad and descends to the Soča river at Srpenica. It crosses the Soča river at the same point, near Žaga, as it did in the 18th century. 1.6.4 Network analysis In opposition to the analysis of optimal routes, the network analysis of routes is rare in archaeology (e.g. Štular 2008; for the method see Conolly, Lake 2006, 58 1.6 Mreža poti 1.6 Path network Sl. 1.39: Poti, ki so prepoznavne na lidarskem posnetku. Fig. 1.39: Paths recognised on the lidar data visualization. uporabljenim v zgornji analizi, poiskali najbližjo točko na mreži poti. Načrtovali smo večstopenjsko analizo. V prvem koraku smo želeli izvesti enostavno mrežno analizo brez kakršnih koli omejitev. V drugem koraku smo nameravali izločiti poti, ki so bile zaradi strmin neprimerne za tovorništvo (za kriterije glej Štular 2006a, 205), in v tretjem koraku še ceste (glej nadaljevanje). V zadnjem koraku smo želeli dodati še postanek pri Tonovcovem gradu. Že po izvedenem prvem koraku analize je pot z izhodiščem ob Soči vodila po podobni trasi kot zgoraj opisana optimalna pot. Po drugem koraku, ko smo izločili za tovorništvo neprimerno pot čez Stol, so vse poti vodile po trasah, skoraj identičnih optimalni poti. To pomeni, da je rezultat tudi po tretjem in četrtem koraku ostal nespremenjen, saj rešitev mrežne analize v nobenem koraku ni vodila po cestah, hkrati pa je že vodila mimo Tonovcovega gradu. Slednje pomeni, da se pot ni prilagajala lokaciji Tonovcovega gradu, temveč obratno. Če bi se trasa poti prilagajala lokaciji Tonovcovega gradu, bi idealna pot vodila vstran in bi se približala lokaciji šele v četrtem koraku, ko smo upoštevali obvezen postanek pri naselbini. 234-252). The reason behind this is extremely simple. Network analysis is used to find the shortest connection between the starting and ending point (and the possible stops in between) amongst the known routes. In archaeological research – at least when dealing with movement in space – no solutions are given in advance. In archaeological research the advantage of network analysis lies in the precision of its result, for the result is reflected in an exact route. The disadvantage lies in the fact that we cannot take into account the possible previously unrecognised routes. However, when we have such a high quality source at our disposal as a DTM 0.5, we expect that merely routes that were created without any intervention into the relief could be potentially unrecognised. On such a rough terrain as the one analysed any potential transport route could not be imagined without certain interventions into the relief, especially when crossing the slopes. The analysis of the aforementioned DTM 0.5 in the surroundings of Tonovcov grad has shown over one hundred possible routes in a total length of 305 kilometres (Fig. 1.39). This data is thus perfect for a network analysis. We entered the starting and ending points of 59 1.6 Mreža poti 1.6 Path network 1.6.5 Historična analiza the routes into the network by pinpointing the points used in the previous optimal path analysis. We planned a multilevel analysis. In the first step a simple network analysis without any restrictions was performed. In the second step the routes that were inappropriate for animal transport due to the steep slopes (for criteria see Štular 2006a, 205) were excluded and in the third step the roads were excluded (see below). In the final step the stop at Tonovcov grad was added. Following the first step the route starting alongside the Soča led along a similar route as the previously described optimal path. After the second step (in which we eliminated the route across Stol as it was inappropriate for transport) all routes led along routes that were almost identical to the optimal path. This means that the result remained unchanged also after the third and fourth step, for the solution provided by the network analysis did not lead along the road at any stage, and it already lead past Tonovcov grad. In other words, steps three and four were not necessary. The interpretation of the latter leads us to the conclusion that the path did not adjust to the location of Tonovcov grad, but the other way round. If the route were adjusted to pass Tonovcov grad, the ideal route would lead to the side and would come close to the location only in stage four, when we took into account the obligatory stop at the settlement. As the results show, the location of Tonovcov grad has been chosen in the proximity to the optimal route. Zgoraj opisano mrežo vseh poti smo analizirali tudi s stališča konteksta. Poglavje smo sicer naslovili historična analiza, saj smo najpomembnejše podatke pridobili na podlagi pisnih virov in historičnih zemljevidov. Vendar smo večino poti interpretirali na podlagi konteksta. Postopek najlaže opišemo kar na konkretnih primerih (sl. 1.40). Najlaže je bilo izločiti poti, nastale v 1. svetovni vojni, saj vodijo neposredno do položajev, kjer se tudi končajo. Podobno smo na podlagi različnih kriterijev interpretirali tudi večino preostalih poti: Pohodniške poti – imajo izhodišče in/ali cilj na modernih cestah (stratigrafski odnos mlajše od) – sekajo in/ali delno izkoriščajo poti iz 1. svetovne vojne (stratigrafski odnos mlajše od) – pogosto gre za najožje izmed dokumentiranih poti, poseg v pobočje pri prečenju je zanemarljiv – pogost cilj je vrh ali razgledna točka – pogoste so krožne poti – pogoste so številne variante na kratkih odsekih. Gospodarske (vaške) poti – povezujejo naselbine s pripadajočimi gospodarskimi objekti, torej vasi s polji, senožetmi in planinami – gostota mreže teh poti je sorazmerna z intenzivnostjo gospodarskega objekta, tj. najgostejša mreža na poljih, najredkejša v strmih pobočjih s senožetmi – praviloma so vrisane v Franciscejski kataster. 1.6.5 Historical analysis The above described network of all routes was also analysed from the contextual aspect. This chapter was entitled Historical analysis, for the most important data was obtained from written sources and the contextual interpretation of historical maps. The context in question has been obtained from DTM 0.5. The procedure can be best described as a regression analysis and is best explained on actual examples (Fig. 1.40). The easiest step was to remove all routes that were created during World War I, for they led to military trenches at which they ended abruptly. Ceste – velik poseg v okolje in posledično velik vložek dela (useki v skalna pobočja, poti s številnimi mostovi) – (lahko prepoznavnih modernih cest v analizo nismo vključili). Stare poti – smo poimenovali poti, ki ne sodijo v nobeno izmed zgornjih kategorij. Sklepamo namreč, da smo z zgoraj opisano erudicijo vira izluščili vse poti razen najstarejših na srednje ali dolge razdalje. Vendar je ta korak v določeni meri tudi subjektiven. Najpogosteje so problem gospodarske poti, ki se povezujejo v poti na dolge razdalje. Tak primer so na primer gospodarske poti Kobarida in Idrskega. Na odseku poti severno od Tonovcovega gradu smo v analizo skušali pritegniti tudi topografski podatek o rimski cesti oziroma ledinsko ime Pod staro potjo južno od naselja Trnovo ob Soči (sl. 1.40). Edina pot v bližini je moderna cesta, ki na tem delu poteka po ne- Most of the remaining routes were similarly interpreted on the basis of various criteria: Hiking paths – Have a starting and/or ending point on modern roads (stratigraphic relationship younger than); – They cross or partially make use of the World War I routes (stratigraphic relationship younger than); – Often they are the narrowest of the documented routes, and they create a minimum impact when crossing a slope; 60 1.6 Mreža poti 1.6 Path network Sl. 1.40: Poti, retrogradna analiza. Fig. 1.40: Paths, retrograde analysis. spremenjeni (vsaj) poznosrednjeveški trasi. Pridevnik “stara” oziroma v varianti “rimska” se torej nanaša na predhodnico te poti. Poleg omenjene ceste so na tem območju le poti, ki jih je podporočnik Bodoky (transkripcija in prevod vira v Rajšp 1997, 7) opisal kot “strme ozke poti, ki vodijo v senožeti”. Slednje imajo tudi priključek na pobočno pot. Opisan položaj lahko razložimo le na dva načina. Verjetneje je, da podatek kaže na lokalni priključek na staro pot, saj gre za povsem nelogičen ovinek. Druga možna je razlaga, da predpona pod označuje odnos ledine pod staro potjo. – The path often ends on top of a peak or at a vista point; – Circular paths are common; – Numerous options can be found on the steep sections. 1.6.6 Vrednotenje Roads – Show great interventions into the environment, are the product of large investment of labour (cuts into the rocky slopes, routes with numerous bridges); – (the easily recognisable modern roads were not included in the analysis). Economic (settlement) paths – Link settlements with the economic objects belonging to it, i.e. villages with fields, meadows and pastures; – The network density of these paths is in accordance to the intensity of the economic object, i.e. it is the densest in the fields and the least dense in the steep slopes with meadows and pastures. – The majority of them have been mapped for the Franciscan cadastre. Rezultati analiz so si zelo podobni. Ko jih prikažemo na istem zemljevidu (sl. 1.41), vidimo, da v neposredni bližini Tonovcovega gradu prikazujejo isti koridor. S kombinirano uporabo analize optimalnih poti in mrežne analize smo dosegli združevanje prednosti in izničenje slabosti obeh metod. 61 1.6 Mreža poti 1.6 Path network Sl. 1.41: Tonovcov grad, rezultat analize optimalnih poti, mrežne analize in vidljivosti. Fig. 1.41: Tonovcov grad, the results of optimal pathway analysis, network analysis and visibility. Old routes – The routes that did not belong into any of the above categories were considered old routes. We concluded that with the above described erudition of the source we eliminated all routes except for the oldest middle and long distance ones. However, to a certain extent this step is subjective. The most common problem is represented by the economic routes that partially overlap with the long distance routes. Such an example is for instance represented by the economic routes in the vicinity of Kobarid and Idrsko. On the section of the route north of Tonovcov grad we tried to include the data stemming from archaeological topography on the place name Roman road or the old Slav name Pod staro potjo (Under the old path), south of the settlement of Trnovo ob Soči (Fig. 1.40). The only existing route in the vicinity of this two place name path is the modern road that has in this section not changed since the Late Medieval times. The adjective ‘old’ or in one variant ‘Roman’ thus denominates a predecessor to this route. Apart from the aforementioned road only tracks described by second lieutenant Bodoky as ‘steep narrow paths that lead to meadows’ (transcription and Slovenian Podobno lahko ugotovimo za celoten odsek razgibanega reliefa z minimalnimi modernimi infrastrukturnimi posegi od Kobarida mimo Tonovcovega gradu do Srpenice. Prikaz rezultata analize optimalne poti na podlagi DMR 0,5 namreč jasno pokaže, da so odstopanja posledica nenatančnosti DMV 12,5. Tovorniška pot je torej vodila po trasi, ki jo kaže rezultat mrežne analize. Natančnost zagotavlja mrežna analiza, optimalnost trase ne glede na danes ohranjene poti pa analiza optimalnih poti. Samo dejstvo, da smo do istega rezultata prišli s tremi neodvisnimi metodami in dvema neodvisnima viroma, potrjuje pravilnost rezultata. Predvsem možnost, da bi bilo ujemanje prve analize z drugima dvema naključno, je zaradi uporabe različnih virov (DMR 0,5 in DMV 12,5) skorajda nična. Natančneje, možnost naključnosti je 0,2 %1. Nenaključnost rezultata sama po sebi seveda ne priča o starosti poti, temveč zgolj o primernosti te poti 1 Verjetnost naključnega ujemanja dveh binarnih rastrov v eni celici je 50 % (kombinacije so 0 0, 0 1, 1 0, 1 1). Verjetnost naključnega ujemanja 9 celic je torej ½9, kar je približno 0,2 %. 62 1.6 Mreža poti 1.6 Path network translation of the source in Rajšp 1997, 7) can be found in this part. These paths lead to the ridge-path that has been identified as the optimal path in the previous analyses. The described situation can be interpreted in two ways. It is more likely that the place names describe the local path leading to the old route. The actual digression of the old route from the ridge path seems rather unlikely. The other possible interpretation is that the prefix ‘under’ denotes the relation of the place name location as lying directly under the old path. This would be a good description of the actual topographic position if the place name location is situated in the valley under the ridge path that has been identified as the optimal path within the previous analyses. 1.6.6 Evaluation Sl. 1.42: Prerez poti med Kobaridom in Srpenico, primerjava stare poti (zgoraj) in v srednjem veku zgrajene ceste (spodaj). Fig. 1.42: Profile of the path between Kobarid and Srpenica; comparison of the old track (above) with the road built in the Middle Ages (bellow). The various analyses yield very similar results. When depicted on the same map (Fig. 1.41) we can see that they run along the same corridor in the direct vicinity of Tonovcov grad. With a combined use of the analysis of optimal routes and the network analysis we merged the advantages and annulled the disadvantages of the two methods. Similar can be ascertained for the entire section from Kobarid past Tonovcov grad to Srpenica, which has a diverse relief and shows minimal modern infrastructural interventions. The depiction of the result obtained by the analysis of the optimum route on the basis of DTM 0.5 clearly shows that the deviations of the optimal path calculated on the basis of DTM 12.5 are indeed a consequence of the DTM 12.5 inaccuracies. The track used by transport animals (but not wagons) thus ran along the route as shown in the network analysis result. The precision of the result is ensured by the network analysis, while the validity irrespective of the modern preservation is ensured by the optimal routes analysis. The result is verified by the fact that we achieved the same result with the use of three independent methods and two independent sources. It is almost impossible for the first analysis to yield the same results as the second. To be precise the possibility for this being a coincidence is 0.2 %1. Of course, the non-randomness of the result does not indicate the age of the route, but merely indicates the appropriateness of this route for the needs of transport. It is merely a statistical confirmation that there are only a few possible solutions in areas with such a diversified za potrebe tovorniškega prometa. Gre le za statistično potrditev, da na območjih s tako razgibanim reliefom možnih rešitev ni veliko. Zato tudi pravilno ovrednotene srednjeveške vire lahko uporabljamo kot analogijo za starejša obdobja. O starosti poti pa lahko sklepamo na podlagi dejstva, da te poti povezujejo železnodobne naselbine Tolmin, Kobarid in Bovec s sočasnima naselbinama ob Nadiži – (verjetno) Sveti Volar nad Robičem in Špeter (sl. 1.36; prim. Mlinar 2004; Mlinar, Pettarin 2007). Pri tem velja poudariti, da lokacije slednjih pri analizah nismo upoštevali in gre torej za še en neodvisen podatek, ki potrjuje pravilnost trase. S stališča najdišča Tonovcov grad je najpomembnejši rezultat te analize določitev trase poti med Kobaridom in Srpenico pred izgradnjo ceste, ki je mestoma vklesana v skalna pobočja soteske Soče. Stara pot je na tem odseku krajša za 1,39 kilometra oziroma 15 %, vendar premaga 390 metrov višinske razlike namesto 180 (sl. 1.42). Slednje jasno kaže na prednosti trase ceste. Po vsem sodeč je bil torej nastanek naselbine na Tonovcovem gradu povezan tudi s traso tovorniške poti, ki je povezovala Čedad čez Nadižo, Bovec in alpske prelaze s Koroško. Je pa ta naselbina v primerjavi s starejšo lokacijo Gradič nad Kobaridom umaknjena od poti, ki je povezovala Kranj s Čedadom čez Bačo, Tolmin, Kobarid in Nadižo. Kljub temu so imeli prebivalci Tonovcovega gradu pregled vsaj nad ključnim delom te poti, prehodom čez Sočo. Prehod zgornje analize sicer postavljajo nekoliko južneje od Kobarida, a glede na rezultate avtorjevih terenskih ogledov se zdi lokacija ob samem vznožju današnjega Kobarida najverjetnejša. 1 The possibility of a coincidental match of two binary rasters in a single cell is 50 % (combinations are 0 0, 0 1, 1 0, 1 1). The results of the optimal path analysis and network analysis match in 9 cells. The possibility of a coincidental match of nine cells would therefore be ½9, which is roughly 0.2 %. 63 1.6 Mreža poti 1.6 Path network Analiza vidnosti potrjuje pregled nad prehodom reke in hkrati razkriva, da s Tonovcovega gradu ni možno opazovati poti Kobarid–Nadiža v smeri iz Čedada ali proti njemu (sl. 1.41). relief. Thus, the correctly evaluated medieval sources can be used as an analogy for the older periods. We can ascertain the age of the path on the basis of the fact that these paths connected the Iron Age settlements of Tolmin, Kobarid and Bovec with the settlements alongside the river Natisone, i.e. (probably) Sveti Volar nad Robičem and San Pietro al Natisone [Slovenian Špeter] (Fig. 1.36 ; cf. Mlinar 2004; Mlinar, Pettarin 2007). At this it should be emphasised that the locations of the Iron age hillforts were not taken into account in the analysis and that this therefore represents additional independent verification for the validity of the route. From the aspect of the site at Tonovcov grad the most important result of the analysis lay in the definition of the route between Kobarid and Srpenica prior to the road (in some places carved into the rocky slopes of the Soča gorge) construction. In this section the path that predated the medieval road was shorter by 1.39 kilometres (15 %), however it has to overcome 390 altitude metres instead of the 180 covered by the road (Fig. 1.42). The latter clearly shows the advantages of the road, i.e. it is passable by wagons. Taking everything into account the origin of the settlement at Tonovcov grad was obviously also linked to its position adjacent to the transport route that linked Cividale del Friuli with Carinthia (across the Nadiža, Bovec and the Alpine passes). When compared to the older settlement location Gradič nad Kobaridom, Tonovcov grad was removed from the route that linked Kranj with Cividale del Friuli (across Bača, Tolmin, Kobarid and the Nadiža). Regardless of this the inhabitants of Tonovcov grad had a clear view of one of the key sections of the route, i.e. the crossing of the Soča river at Kobarid. The above analyses are showing the river crossing somewhat south of the present day Kobarid, but the author’s ground truthing suggests the location just beneath the Kobarid. This is revealed by the visibility analysis (Fig. 1.41), which also revealed that it was impossible to see the route Kobarid - the Nadiža (in the direction to or from Cividale del Friuli) from Tonovcov grad. 64 2. Terenski izvid 2. Field report 2.1 Metodologija dela 2.1 Methodology 2.2 Kronologija in faze 2.2 Chronology and phases 2.3 Stavba 1 z okolico 2.3. Building 1 and its surroundings 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 2.5 Sklop cerkva 2.5 The ecclesiastical complex 2.6 Cisterna (vodni zbiralnik) 2.6 Water cistern (reservior) Sl. 2.1: Načrt najdišča. M = 1:1000. Fig. 2.1: Map of the site. Scale = 1: 1000. 65 2.1 Metodologija dela 2.1 Methodology Pred začetkom izkopavanja je bila na Tonovcovem gradu narejena geodetska izmera. Postavljenih je bilo več fiksnih geodetskih točk, ki so določale koordinatni sistem z linijama y (smer sever-jug) in x (smer vzhod-zahod). Na osnovne linije je bila potem postavljena mreža kvadrantov s stranicami 4 x 4 m (sl. 2.1), znotraj izkopnih polj pa so bili kvadranti razdeljeni še na mikrokvadrante s stranicami 1 x 1 m. Višine so bile do leta 2002 merjene relativno od dveh absolutno izmerjenih geodetskih točk (399,55 m n. m. na območju izkopnega polja stavbe 1 ter 410,25 m n. m. na območju cerkvenega sklopa), po letu 2002 pa absolutno. Tudi relativno merjene višine so v knjigi preračunane v absolutne nadmorske višine. Vrh Tonovcovega gradu s poznoantično naselbino je zelo neraven. V osnovi se deli na tri dele. Najnižji je severozahodni del naselbine. Teren se proti jugu nato strmo dvigne za približno 10 m in se zravna v raven, izpostavljen plato, na katerem je bil postavljen sklop cerkva, na vzhodnem delu pa se tik nad Sočo dviga najvišji plato (412 m. n. m, glej tudi pogl. 1.1 in 1.3). Na strmih pobočjih je bil humusni pokrov zelo tanek, tako da je bila na nekaterih mestih že pred arheološkimi posegi vidna skalna osnova, medtem ko so bile drugje med izkopavanji odkrite tudi več metrov globoke skalne kotanje. Nekatere so bile zapolnjene s sterilnimi plastmi gline in gruščev, druge pa so vsebovale bogate kulturne plasti. Hrib je bil pred začetkom raziskav prekrit z redkim gozdom (bukev, gaber, hrast) in grmičevjem. Plast humusa, ki je prekrivala ruševine poznoantičnih stavb, je bila zelo tanka, na nekaterih mestih pa je sploh ni bilo. Osnovne oblike zidanih objektov so bile zato ponekod na površini še jasno razvidne, tako da se je v grobem dalo določiti njihove obrise (glej pogl. 1.3, sl. 1.6, 1.9, 1.10, 1.12). Na podlagi tako vidnih objektov so bila potem izbrana območja raziskav. Zaradi strmega terena so bile posebno v ruševinskih plasteh najdbe pomešane, saj so predmeti z višje ležečih predelov padli ali spolzeli na nižje ležeče. Tako lahko za veliko število novcev in nekaj razbitih tegul na sicer neposeljenem pobočju nad stavbo 1 domnevamo, da so prišli sem z višje ležečega cerkvenega platoja. Prior to the beginning of the excavations a land survey was conducted at Tonovcov grad. A number of fixed geodetic points were positioned and they were used to define a coordinate system with the y (north - south direction) and x (east - west direction) axes. A grid of quadrants measuring 4 x 4 metres was placed along the basic lines (Fig. 2.1), and within the excavation area the quadrants were divided into micro quadrants measuring 1 x 1 m. Until 2002 all heights were measured relatively by using two previously absolutely measured geodetic points (399.55 m a.s.l. in the excavation area of building 1 and 410.25 m a.s.l. in the area of the ecclesiastical complex), while after 2002 all heights were measured absolutely. For the purpose of this book all relatively measured altitudes were recalculated into absolute altitudes. The peak of Tonovcov grad with the Late Antique settlement on it is very uneven. The top of the hill can be divided into three parts, the lowest of which was occupied by the northwest part of the settlement. From here the terrain rises sharply towards the south (by approximately 10 m) and levels out into a straight and exposed plateau, the top of which was covered by ecclesiastical complex. The highest plateau on the east rises above the Soča river (412 m.a.s.l., see also chapters 1.1 and 1.3). The steep slopes were covered by a very thin layer of humus, in some spots revealing the rock base already prior to the archaeological excavations, while elsewhere up to a few meters deep sinkholes were discovered during the excavations. Some of them were filled with sterile layers of clay and gravel, while others included rich cultural layers. Prior to the beginning of the research the hill was covered in trees (beech, hornbeam and oak trees) and bushes. The humus layer that covered the ruins of the Late Antique buildings was very thin or even nonexistent. Thus the basic forms of some of the structures were visible on the surface and their outlines could be at least roughly defined (see chapter 1.3, Figs. 1.6, 1.9, 1.10, 1.12). The research areas were selected on the basis of the visible structures. Due to the steepness of the terrain the finds were very mixed, especially in the destruction layers, for 67 2.1 Metodologija dela 2.1 Methodology Korenine dreves in grmovja, ki so preraščali ruševine objektov, so bile zelo razvejene in so zato na nekaterih mestih močno poškodovale arheološke plasti. Pri odstranjevanju koreninskih sistemov je bilo velikokrat poleg same ruševine poškodovanih tudi nekaj plasti pod njo. Problem je predstavljala tudi ločljivost plasti. Posebno zgornje, humusno-ruševinske, so si bile po barvi in strukturi zelo podobne, razlikovanje med njimi pa so oteževale tudi poškodbe zaradi korenin in kasnejših posegov. Ti so bili dokumentirani v stavbi 1, na več mestih v cerkvenem sklopu in v vodnem zbiralniku (glej pogl. 2.3, 2.5, 2.6). Metodologija izkopavanj se je zato morala prilagoditi danemu stanju. Kjer so bile plasti nepoškodovane in jasno razvidne, je izkopavanje potekalo po plasteh. Najdbe znotraj plasti so bile dokumentirane tudi v okviru mreže mikrokvadrantov z metrsko natančnostjo, posebne najdbe pa so dobile še koordinate x, y in z. Najdbe iz nejasnih ali premešanih plasti pa so bile prostorsko umeščene samo znotraj mreže mikrokvadrantov, brez določitve plasti, dokumentirana je bila tudi višina. Tudi v teh primerih so posebne najdbe dobile še koordinate x, y in z. Pri izkopavanjih je bil kot pomoč pri iskanju kovinskih najdb uporabljen detektor kovin. Plasti so bile pregledane z detektorjem že pred odstranjevanjem, vzorec pa je bil nato, ko je bil že odstranjen, pregledan še enkrat. Ta način se je izkazal kot najuporabnejši pri zgornjih, ruševinskih plasteh, v katerih so prevladovali kamni. Sejanje ali spiranje se je pri takih vzorcih izkazalo za neekonomično. Pri plasteh, kjer je njihova sestava to omogočala, pa je bilo uporabljeno tudi suho sejanje.1 Plasti v izkopnih poljih so bile v večini primerov raziskane do geološke osnove. Izjema je bilo samo izkopavanje cerkvenega sklopa, kjer je bila zaradi odlično ohranjenih estrihov v severni in osrednji cerkvi sprejeta odločitev, da se v čim večji meri ohranijo in zaščitijo (glej pogl. 2.5). Globino in sestavo plasti pod estrihi smo lahko tako raziskali le z manjšimi kontrolnimi sondami v ladjah severne in osrednje cerkve. Terenska dokumentacija je obsegala terenske dnevnike, opise plasti, sezname najdb, risbe planumov in presekov ter fotodokumentacijo. the objects from the top of the slope fell or slid to the lower lying parts. Thus it can be assumed that the large number of coins and some of the broken tegulae on the slope over building 1 where there are no structures fell from the church plateau. The roots of the trees and bushes that grew over the ruins spread out and caused severe damage to the archaeological layers in some spots. As the roots were removed it was often revealed that not only the ruins but also several layers underneath were destroyed. The distinction between various layers represented an additional problem. Especially the upper destruction layers proved to be very similar in colour and texture, and distinguishing between them was made harder by the damage created by the roots and the later interventions. These were documented in building 1, on a number of places within the ecclesiastical complex and alongside the water reservoir (see chapters 2.3, 2.5, 2.6). The methodology of the excavations thus had to be adjusted accordingly. Where the layers were undamaged and clearly visible the excavation proceeded by layers. The finds in a layer were documented within a microquadrant grid with metre preciseness, and special finds obtained coordinates x, y and z. If the layers could not be precisely defined the finds were recorded merely within the micro-quadrant grid, and their z-coordinate was documented. The special finds again obtained their x, y and z coordinates. A metal detector was used for collecting metal finds. The layers were checked with a metal detector prior to their removal, and the sample was checked once again after it was removed. This has proven to be useful in the upper, destruction layers, in which stones dominate. Sieving and flotation have proven to be inefficient with such samples. Dry sieving was used wherever the composition of the layer permitted.1 In most cases the layers in the excavation area were researched all the way to the geologic base. An exception to this rule was the excavation of the ecclesiastical complex in which the mortar floor in the north and main church was excellently preserved, thus a decision was reached that the floors are to be preserved and protected as well as possible (see chapter 2.5). The depth and composition of the layer under the mortar floor were thus defined on the basis of smaller control trenches in the naves of the north and central church. The documentation consisted of field diaries, layer descriptions, lists of finds, drawings of planums and sections as well as photo documentation. 1 To so predvsem spodnje plasti na izkopnem polju stavbe 1, medtem ko pri izkopavanju cerkvenega sklopa in objektov 2 in 3 ta način ni bil uporabljen. 1 These are mainly the lower layers on the excavation area of building 1, while at the excavation of the ecclesiastical complex and buildings 2 and 3 this method was not used. 68 2.2 Kronologija in faze 2.2 Chronology and phases Na podlagi dosedanjih raziskav je bilo mogoče na Tonovcovem gradu določiti pet obdobij poselitve (prazgodovina, antika, pozna antika, zgodnji srednji vek, srednji vek). Nekatere izmed njih je bilo mogoče razdeliti tudi na faze (predvsem velja to za najbogatejše poznoantično obdobje). Pri tem je treba poudariti, da nekatera obdobja za zdaj nimajo jasne potrditve tudi v sami stratigrafski situaciji na najdišču. Tako prazgodovinsko (železnodobno), antično in srednjeveško poselitev predvidevamo predvsem na podlagi posamičnih najdb iz teh obdobij, najdenih v plasteh obeh poznoantičnih faz, ter še neraziskanih ostankov arhitekture (domnevni srednjeveški stolp v zahodnem delu naselbine). Pri časovni določitvi obdobij in faz so bile uporabljene samo najdbe, ki izvirajo iz zanesljivih plasti. Vse ostale najdbe, kot so npr. detektorske, pobrane pred začetkom izkopavanj zunaj izkopnih polj (glej pogl. 1.4), in tiste iz poškodovanih ali nezanesljivo razbranih plasti, so bile sicer obdelane v okviru analize gradiva in kot pomoč pri interpretaciji, niso pa bile uporabljene pri datacijah faz, kartah razprostranjenosti predmetov po objektih in statističnih analizah. The research carried out at Tonovcov grad has revealed five settlement periods: Prehistory, Antiquity, Late Antiquity, Early Middle Ages and Middle Ages. Some of them could be divided into various phases (especially the Late Antique period, which revealed the most finds). Here it should be stressed that some periods have so far not been confirmed in the stratigraphic situation at the site. Thus the prehistoric (Iron age), Antique and High Medieval settlement was assumed on the basis of individual finds belonging to these periods, which were found in the layers of the two Late Antique phases and the as yet uninvestigated architectural remains (the assumed Medieval tower in the west of the settlement). Only finds that originate from reliable layers were used to define the time scales of the various periods and phases. All other finds - such as the finds discovered with the use of a metal detector that were gathered prior to the excavations and outside the excavation areas (see chapter 1.4) or finds from damaged or unreliably dated layers - were treated within the material analysis and provided help in the interpretation, however they were not used to date the phases, nor were they shown in the distribution maps of finds or in statistical analyses. 2.2.1 Prazgodovina 2.2.1 Prehistory Nad plastmi geološke osnove (skalna osnova, skalna preperina, sterilne gline) so bile na več mestih dokumentirane plasti, ki so vsebovale skromne prazgodovinske ostanke. Take plasti so bile na območju stavb 1, 2, 3 in cisterne. V vseh primerih gre za plasti trde, oranžno-rdeče gline, ki je bila zelo težko ločljiva od sterilne glinene osnove. Najdbe iz plasti predstavljajo sileksi, nekaj kamnitih orodij in nekaj zelo fragmentirane in slabo ohranjene prazgodovinske keramike (pogl. 2.3.1, 2.4.1; glej tudi Tonovcov grad. Najdbe, pogl. 6). Že na vrhu prazgodovinskih plasti pa so se začenjale pojavljati tudi antične najdbe (včasih celo neposredno skupaj). Poleg tega so bila nekatera kamnita orodja najdena tudi kot rezidualne najdbe v mlajših, antičnih plasteh. Ta prazgodovinska poselitev je datirana v dolgo časovno obdobje med paleolitikom in bronasto dobo. Znotraj The layers above the geological base (bedrock, decayed bedrock, sterile clays) included modest prehistoric remains. Such layers were found in the areas of buildings 1, 2, 3 and the water cistern. In all cases these were layers of compact orange-red clay that were very hard to separate from the sterile clay base. The finds from these layers are represented by a combination of silex, stone tools and extremely fragmented and poorly preserved prehistoric pottery (chapters 2.3.1, 2.4.1; see also Tonovcov grad. Finds, chapter 6). Antique finds appear already on the top of these prehistoric layers, sometimes together with the prehistoric ones. Stone tools also appear as residual finds in the later Antique layers. This prehistoric settlement is dated into the long period between the Paleolithic and Bronze Age. There 69 2.2 Kronologija in faze 2.2 Chronology and phases tega časa obstaja več poselitvenih faz, ki pa jih zaradi slabe ohranjenosti plasti in najdb za sedaj ne moremo natančneje opredeliti. Verjetna je poselitev v mezolitiku in bronasti dobi (glej Tonovcov grad. Najdbe, pogl. 6.5). Najdeno je bilo tudi nekaj predmetov, ki so datirani v starejšo in mlajšo železno dobo (glej Tonovcov grad. Najdbe, pogl. 6.2, 6.3; sl. 6.2). To so dokaj reprezentativne najdbe, ki pa so bile vse odkrite v mlajših, že poznoantičnih plasteh. Zanesljivih poselitvenih plasti iz železne dobe na do sedaj raziskanih območjih nismo odkrili. were a number of settlement phases within this period, however, due to the poor preservation of the layers and finds it is impossible to establish this more precisely. Settlement in the Mesolitic period and in the Bronze Age is probable (see Tonovcov grad. Finds, chapter 6.5). A few objects that could be placed into the early or late Iron Age were also discovered. These are relatively representative finds that were all discovered in the later, Late Antique layers (see also Tonovcov grad. Finds, chapters 6.2, 6.3; Fig. 6.2). In the areas researched so far we have not found any reliable settlement layers from this period. 2.2.2 Antika 2.2.2 Antiquity Poselitev iz časa 1. in 2. st. ni zanesljivo potrjena z naselbinskimi ostanki. V ta čas bi po stratigrafskih odnosih sicer lahko sodile najstarejše antične plasti pod stavbo 3 (SE 177, 179, glej pogl. 2.4.1, 3.2.2), vendar je potrebna previdnost pri opredeljevanju zgodnjih antičnih najdb, posebno fibul, ki se na Tonovcovem gradu pogosto pojavljajo v mlajših plasteh. V poznoantičnih (predvsem ruševinskih) plasteh stavb 1 in 3 je bilo najdenih nekaj novcev in fibul iz tega obdobja. Prva poselitvena faza višinskih naselbin v Sloveniji, čas zadnje tretjine 3. stoletja (Ciglenečki 1990, 154–156; 1999, 292), je na Tonovcovem gradu zastopana predvsem s posamičnimi najdbami. Edine plasti, ki bi lahko sodile v ta čas, so bile ohranjene pod stavbo 3 (SE 172, 175). V SE 172 je bil najden novec, datiran v leto 272 (glej pogl. 2.4.1 in Tonovcov grad. Najdbe, pogl. 5.1, kat. št. 22). Ostalo dobro opredeljivo gradivo tega časa je bilo najdeno v plasteh prve (večina novcev) in druge poznoantične faze na območju stavbe 1 (obročaste fibule, dva ključa v obliki prstana, pasna spona ter ščitnik noža; Tonovcov grad. Najdbe, pogl. 2.1). The discovered remains are not enough to reliably confirm settlement on the hill in the 1st and 2nd centuries. Due to their stratigraphic relations the oldest Antique layers under building 3 (SU 177, 179, see chapters 2.4.1, 3.2.2) could be dated into this period, however great care is needed when defining layers on the basis of early Antique finds, especially fibulae, which represent a common find in the younger layers at Tonovcov grad. Finds from this period were also discovered in the Late Antique (especially destruction) layers of buildings 1 and 3. At Tonovcov grad the first settlement phase of the hilltop settlements in Slovenia, i.e. the last third of the 3rd century (Ciglenečki 1990, 154-156; 1999, 292), is also represented mostly by individual finds. The only layers that could belong into this period were preserved under building 3 (SU 172, 175). A coin dated to 272 AD (see chapter 2.4.1 and Tonovcov grad. Finds, chapter 5.1, Cat. No. 22) was found in SU 172. The remaining material that can be reliably dated to this period was found in the layers of the first (mainly coins) and second (ring fibulae, two rings-keys, a belt buckle and a knife hilt plate) Late Antique phases in the area of building 1 (Tonovcov grad. Finds, chapter 2.1). 2.2.3 Pozna antika Prva poznoantična faza (PA 1) 2.2.3 Late Antiquity Čas pozne antike pomeni vrhunec poselitve Tonovcovega gradu. Razdelimo ga lahko na dve fazi, ki sta obe jasno vidni tudi iz stratigrafskih odnosov. Starejša, prva poznoantična faza (PA 1) sodi v čas druge polovice 4. in v začetek 5. st., njen konec pa časovno ni povsem dobro omejen. To obdobje (imenovano tudi poznorimsko; o problematičnosti terminologije glej Ciglenečki 1999, 290–292) je v jugovzhodnih Alpah čas nastanka številnih utrjenih vojaških postojank, katerih namen je bil varovanje strateško pomembnega območja ob vstopu v Italijo (Ciglenečki 1999, 291–292). Na Tonovcovem gradu so bili ostanki iz tega časa najdeni pod stavbama 1 in 2. Faza je dobro dokumentirana tudi s keramičnimi, kovinskimi in novčnimi najdbami iz mlajših plasti. Na Late Antiquity phase 1 (LA 1) The Late Antique period represents the pinnacle of the settlement at Tonovcov grad. It can be divided into two phases, both of which can be clearly seen in the stratigraphy. The earlier Late Antiquity phase (LA 1) is dated to the second half of the 4th and beginning of the 5th century. The end of this phase could not be clearly defined. This period (also known as Late Roman; for the terminology issues see Ciglenečki 1999, 290–292) was a period in which numerous fortified military posts appeared in the Southeastern Alps in order to guard the strategically important entrance into Italy (Ciglenečki 70 2.2 Kronologija in faze 2.2 Chronology and phases območju cerkvenega sklopa ter vodnega zbiralnika plasti te faze niso bile najdene. 1999, 291–292). At Tonovcov grad the remains from this period were found under buildings 1 and 2. The phase is also well represented by pottery, metal and coin finds discovered in the later layers. No Late Antiquity 1 layers were found in the area of the ecclesiastical complex and the water reservoir (cistern). Druga poznoantična faza (PA 2) Mlajša, druga poznoantična faza (PA 2) je še bogatejša in bolje ohranjena od prve. Trajala je predvidoma med koncem 5. in začetkom 7. stoletja. V to obdobje lahko umestimo večino raziskanih objektov v naselbini (stavbi 1 in 2, cerkveni sklop, vodni zbiralnik) ter veliko večino kovinskega in keramičnega drobnega gradiva. Late Antiquity phase 2 (LA 2) The Late Antiquity phase 2 (LA 2) is even richer and better preserved than the first one. It is assumed that it spanned from the end of the 5th to the beginning of the 7th century. Most of the researched buildings in the settlement (buildings 1 and 2, the ecclesiastical complex, water reservoir) can be dated into this period, as can most of the metal and ceramic small finds. 2.2.4 Zgodnji srednji vek (ZSV) Kot trenutno še ne dobro razumljiva se kaže poselitvena faza, ki jo lahko grobo datiramo v zgodnjesrednjeveško obdobje. Vanjo lahko uvrstimo nekatere drobne najdbe, nekaj plasti v stavbi 1, kurišče s keramiko v glavni cerkvi, grob 21 iz cerkvenega sklopa, skupino grobov ali vsaj enega od njih (grob 18) za južnim zidom stavbe 2 ter kurišče s keramiko v poznoantični cisterni. Časovna opredelitev teh najdb za zdaj kaže na čas med koncem 7. in začetkom 9. stoletja. Možni sta tudi dve krajši zgodnjesrednjeveški fazi, prva nedolgo po opustitvi poznoantične naselbine ter druga v času okrog leta 800. Na prvo kažejo nekaj keramičnega gradiva ter grobova 18 in 21, na drugo pa nekaj kovinskih predmetov (Tonovcov grad. Najdbe, t. 8: 22; 26: 1, 6–8). Podobne poznamo tudi z drugih slovenskih poznoantičnih višinskih naselbin, kjer predstavljajo še ne dobro pojasnjen zgodnjekarolinški horizont (Knific 2007). 2.2.4 Early Medieval period (EMP) So far the Early Medieval settlement period is not very well understood. Some small finds can be dated into it, as can some of the layers in building 1, the fireplace with some pottery in the main church, grave 21 in the area between the main and south church, the group of graves or at least one of them (grave 18) behind the south wall of building 2 and the fireplace with pottery within the Late Antique water reservoir. It seems that these finds originated from between the end of the 7th and beginning of the 9th century. Two shorter Early Medi eval phases are also possible, the first shortly after the Late Antique settlement was abandoned and the second sometime around the year 800. The first is indicated by a few pieces of pottery and graves 18 and 21, while the second is indicated by several metal objects (Tonovcov grad. Finds, Pls. 8: 22; 26: 1, 6–8). Similar finds were discovered at other Late Antique hilltop settlements in Slovenia, where they formed the as yet poorly understood Early Carolingian horizon (Knific 2007). 2.2.5 Srednji vek Tudi iz časa po 9. stoletju je bilo na hribu najdenih nekaj predmetov (Tonovcov grad. Najdbe, t. 4: 25), ki morda nakazujejo poselitev Tonovcovega gradu tudi v srednjem veku. V ta čas morda sodijo tudi še neraziskani ostanki arhitekture na zahodnem robu naselbine (objekt št. 13), ki bi lahko predstavljali ruševine srednjeveškega stolpa. 2.2.5 Medieval period Certain objects (Tonovcov grad. Finds, Pl. 4: 25) found on the hill can be dated after the 9th century and they might indicate that some form of settlement existed at Tonovcov grad also in the Medieval period. The uninvestigated architectural remains on the west of the settlement (building 13), which could represent the ruins of a Late Medieval tower, could belong into this period. 71 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Za izkopavanje leta 1994 je bilo izbrano območje zahodno od vhoda v naselje, na razmeroma ravnem, le rahlo uleknjenem delu pod strmo skalno steno, ki se dviguje proti osrednjemu platoju Tonovcovega gradu. Tu sta bila že pred začetkom izkopavanj na površini vidna ruševina in delno tudi tloris velike zidane stavbe (stavba 1). V njen jugovzhodni vogal je bil v preteklosti (glej pogl. 1.1.4) narejen vkop, ki je razkril tudi del južnega zidu. Geološko osnovo v večini izkopnega polja predsta vlja skala, katere površina je močno razgibana. Skala je na nekaterih mestih zelo preperela, vmesne prostore pa zapolnjuje plast trde oranžne gline. Teren pada od juga proti severu in od zahoda proti vzhodu. Izkopavanja, ki so se nadaljevala tudi v letu 1995, so zajela 480 m2 velik prostor (sl. 2.2). Poleg ostankov zidane stanovanjske stavbe (stavba 1) je bil raziskan tudi obsežen prostor okrog nje. Izkopno polje je segalo do ostankov sosednjih objektov (stavbe 10, 22, 26; sl. 1.7), zajelo pa je tudi strmo pobočje brez zidanih objektov nad stavbo 1, vse do skalne stene, ki se dviguje do platoja s cerkvami. Pod ostanki stavbe 1 je bilo dokumentiranih več faz po selitve iz časa pred njeno izgradnjo, v ruševinskih plasteh pa aktivnosti iz časa, ko je bila stavba že delno porušena. Najstarejša poselitev tega območja sodi v pra zgodovino. Zanesljivi naselbinski sledovi iz rimskega obdobja na območju izkopnega polja stavbe 1 niso During the 1994 excavations the area west of the settlement entrance was explored. This area lies on a rela tively flat, only faintly curved part under the steep rock face that rises towards the central plateau of Tonovcov grad. Even before the excavations commenced the ruin and the contour of the ground plan of a large stone build ing (building 1) could be seen. In the past the southeast corner was damaged and this intervention revealed a part of the south wall (see chapter 1.1.4). Throughout most of the excavation area the geo logical base is represented by bedrock, the surface of which is extremely diverse. In certain spots the rock has disintegrated strongly, and the spaces that appeared in between became filled by layers of compact orange clay. The terrain drops from the south to the north and from the west to the east. The excavations that continued in 1995 covered an area measuring 480 m2 in size (Fig. 2.2). Alongside the remains of the stone building that was used as the living quarters (building 1) a large area surrounding it was also researched. The excavation area reached all the way to the remains of the neighbouring structures (buildings 10, 22 and 26; Fig. 1.7) and included the steep slope (without any stone structures) above building 1 – all the way to the rock face that reaches the plateau with the churches. Under the remains of building 1 a number of settlement phases from the time before the building was Tab. 2.1: Stavba 1. Preglednica stratigrafskih enot (SE). Tab. 2.1: Building 1. Table of startigraphic units (SU). Opredelitev / Definition Sterilna / Sterile Prazgodovina / Prehistory PA 1 / LA 1 PA 1/ PA 2 / LA 1/ LA 2 PA 2 / LA 2 Rušenje / Destruction ZSV / EMP Premešano / Mixed SE / SU 40 33, 39 04, 16, 17, 20, 29a, 30, 31, 32, 36=68, 36a, 36b, 51, 53, 54, 57, 66, 68, 75, 76, 77, 78; zidovi /walls 13, 14, 15 18, 21, 22, 24, 63, 74 03, 05, 12, 14, 23=26, 25, 28, 29, 35a, 50, 55, 56, 62, 64, 67; zidovi / walls 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 01, 08, 11, 13 09, 10 02, 06, 34 73 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.2: Tloris stavbe 1 z mejo izkopnega polja in mrežo kvadrantov. M. = 1:200. Fig. 2.2: Ground plan of building 1 with the excavation area and quadrant grid. Scale = 1:200. bili ohranjeni (glej pogl. 2.2.2). Ponovno je območje intenzivno poseljeno v pozni antiki, ko lahko ločimo dve poselitveni fazi, to je prvo in drugo poznoantično fazo (PA 1 in PA 2). Najbolje ohranjena in najbogatejša je druga poznoantična faza, v katero sodijo tudi ostanki zidane stavbe. V stavbnih ruševinah so bili ohranjeni tudi sledovi zgodnjesrednjeveškega obdobja (tab. 2.1). erected were documented, and the destruction layers show activities from the time when the building was already partially demolished. The earliest settlement on this area can be dated to the prehistoric times. No reliable settlement traces from the Roman period were found within the excavated area (see chapter 2.2.2). The area was settled again the Late 74 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings 2.3.1 Prazgodovina Antiquity. This period could be divided in two settlement phases: Late Antiquity phase 1 (LA 1) and Late Antiquity phase 2 (LA 2) . The best preserved and richest in finds is Late Antiquity phase 2, which also includes the remains of the stone building. Traces of the Early Medieval period were preserved within the destruction layers (Tab. 2.1). Nad geološko osnovo je bila na nekaterih mestih izkopnega polja najdena plast s prazgodovinskimi ostanki. Najizrazitejša je bila na severni strani izkopnega polja, v kvadrantih 617, 618, 619, 668 in 669, kjer je bila skalna osnova najnižja in so bile kulturne plasti najdebelejše. Skalno osnovo je prekrivala plast trde oranžne, kulturno sterilne gline (SE 40). Nad njo je ležala tanka (10–20 cm) plast trde oranžno-rjave gline (SE 33), ki je vsebovala ži valske kosti, drobce oglja in kamnite artefakte (sl. 2.3). Na njeni zgornji površini, na meji s SE 32, so se že pojavljale antične najdbe. Kamniti artefakti so bili najdeni tudi še v SE 32, ki pa je bila glede na ostale najdbe že antična. 2.3.1 Prehistory In some spots a layer with prehistoric remains was found above the geological base. This was most pronounced on the north side of the excavation area in quadrants 617, 618, 619, 668 and 669, where the bed rock was the lowest and the cultural layers the thickest. The bedrock was covered by a layer of compact orange clay that did not show any cultural traces (SU 40). This was covered by a thin (10-20 cm) layer of compact, orange-brown clay (SU 33) that contained animal bones, charcoal particles and stone artefacts (Fig. 2.3). On its upper surface, on the border with SU 32, finds dating to the Roman period were discovered. Stone artefacts were also found in SU 32, however the remaining finds date this layer into the Roman period. In the central part of the excavation area (qus. 717, 718, 719, 767, 768) the few prehistoric remains were preserved merely in the rock crevices that were filled with compact orange clay (SU 39). No prehistoric remains were discovered in the southern or western part of the excavation area where the bedrock rises sharply. 2.3.2 Late Antiquity phase 1 (LA 1) On top of the prehistoric layer or on top of the geological base lay cultural layers that represent the remnants of Late Antiquity phase 1, i.e. the phase that pre-existed building 1. During the construction of building 1 (at the beginning of Late Antiquity phase 2) the layers were severely damaged, however we can still ascertain that these remains – similar to the prehistoric ones – were also based in the northern and eastern part of the excavation area. Remains were documented in the interior of the later building as well as on the outside of it (and under its walls), however they were severely damaged during construction. In the eastern part of the excavation area a number of cultural layers (SU 30, 31 and 32) that included finds from the Antiquity were discovered above the prehis toric layer SU 33. SU 32 ran under wall 4 also in the interior of the later object (Figs. 2.3, 2.4). SU 29a – a layer of loose soil with a strong presence of cultural finds and larger stones (Fig. 2.3) – was located on top of SU 32. This layer also ran under wall 4 (Figs. 2.3, 2.4), however it was not preserved in the interior of the later building. Sl. 2.3: Stavba 1, presek 4 pravokotno na zid 4. M. = 1:50. Fig. 2.3: Building 1, section 4 at a right angle to wall 4. Scale = 1:50. V osrednjem delu izkopnega polja (kv. 717, 718, 719, 767, 768) je bilo nekaj prazgodovinskih ostankov ohranjenih le v skalnih razpokah, zapolnjenih s trdo oranžno glino (SE 39). V južnem in zahodnem delu izkopnega polja, kjer se skalna osnova močno dvigne, prazgodovinski ostanki niso bili najdeni. 2.3.2 Prva poznoantična faza (PA 1) Nad prazgodovinsko plastjo oziroma nad geolo ško osnovo so ležale kulturne plasti, ki predstavljajo ostanek prve poznoantične faze, to je faze pred gradnjo 75 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.4: Stavba 1, zid 4 - zunanje lice. M. = 1:50. Fig. 2.4: Building 1, wall 4 - exterior. Scale = 1:50. stavbe 1. Plasti so bile ob gradnji stavbe 1 na začetku druge poznoantične faze zelo uničene, kljub temu pa lahko ugotovimo, da so bili tudi ti ostanki – podobno kot prazgodovinski – strnjeni v severnem in vzhodnem delu izkopnega polja. Dokumentirani so bili tako v no tranjosti kot zunaj kasnejše stavbe, pa tudi pod njenimi zidovi, vendar so bili ob gradnji zelo poškodovani. V vzhodnem delu izkopnega polja je nad pra zgodovinsko SE 33 ležalo več kulturnih plasti (SE 30, 31 in 32), ki so že vsebovale antične najdbe. SE 32 se je nadaljevala pod zidom 4 tudi v notranjost kasnejšega objekta (sl. 2.3, 2.4). Nad SE 32 je ležala SE 29a, to je rahla, močno kul turna prst, v kateri so bili tudi večji kamni (sl. 2.3). Tudi ta plast se je nadaljevala pod zid 4 (sl. 2.4), v notranjosti kasnejše stavbe pa ni bila več ohranjena. Ob zunanjem vogalu zidov 3 in 4 so bili ohranjeni ostanki dveh zelo uničenih zidov, to sta zidova 13 in 14 (sl. 2.2), od katerih je ostala samo spodnja linija v vrsto postavljenih kamnov brez sledov malte ali ometa. Zid 13 je potekal približno pravokotno na zid 4, vendar z njim ni bil neposredno povezan, saj je bila med njima približno 20 cm široka reža. Bil je dolg približno 3 m, nanj se je pravokotno navezoval zid 14, ki je bil v dolžino ohranjen samo še približno 2 m. Spodnja linija zidu 13 je bila 30–40 cm niže kot spodnja linija zidu 4. V osrednjem delu izkopnega polja (kv. 767 in delno kv. 717) je plasti prve poznoantične faze zelo poškodovala gradnja stavbe 1. Tu so nad glineno osnovo SE 40 ležale zaplate rjave gline (SE 36; sl. 2.5, 2.6). Plast je bila močno kulturna, v mkv. 718/D4 sta bili vanjo vkopani dve jami nepravilne oblike (SE 36a in SE 36b). Jama 36a je bila zapolnjena s črno mehko zemljo, vsebovala pa je tudi opeko, keramiko in živalske kosti. Tik ob njej je ležala strnjena plast opeke, poleg nje pa keramika in kosti. Jama 36b je bila zapolnjena z zemljo, vendar je bila brez najdb. The remains of two severely destroyed walls (wall 13 and wall 14) were discovered on the exterior of the corner where walls 3 and 4 meet (Fig. 2.2). Only the lower line of stones was preserved, without any traces of mortar or plaster. Wall 13 ran roughly at a right angle to wall 4, however it had no direct contact with it, for there was an approximately 20 cm wide gap between the two. It was approximately 3 metres long and wall 14 (of which roughly a mere 2 metres in length was preserved) was attached to it at a right angle. The bot tom row of wall 13 was 30-40 cm lower than the lower line of wall 4. The Late Antiquity 1 layers in the central part of the excavation area (qu. 767 and partially qu. 717) were severely damaged during the construction of building 1. Patches of brown clay (SU 36) were discovered above the clay base (SU 40; Figs. 2.5, 2.6). This layer had a strong cultural presence, and mqu. 718/D4 revealed two irregular shaped pits (SU 36a and SU 36b). Pit 36a was filled with loose black soil that included bricks, pot tery and animal bones. A composite layer of bricks was discovered next to it and alongside this lay pottery and animal bones. Pit 36b was filled with soil but was void of any finds. SU 36 ran under wall 1 of the later building (Figs. 2.5, 2.6) also on the north of the excavation area, in quadrants 716 and 666 (Fig. 2.7). Under wall 1 another two cultural pits (SU 77 and SU 78) could be observed, both of which were dug all the way through the cultural layers right to the bedrock (Fig. 2.5). Wall 15 stood in the northern part of the excavation area, on SU 36 (Figs. 2.2, 2.7). The wall was 5.4 metres long and 0.5 metres wide, and merely a single row was preserved in height. Its width was covered by one large or two slightly smaller stones, without an intermediate gravel layer. Modest mortar remains were discovered amongst the stones. 76 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.5: Stavba 1, zid 1 – notranje lice in vhod. M. = 1:50 Fig. 2.5: Building 1, wall 1 - interior and threshold. Scale = 1:50. Sl. 2.6: Stavba 1, zid 1 – zunanje lice. M. = 1:50 Fig. 2.6: Building 1, wall 1 - exterior. Scale = 1:50. SE 36 je potekala pod zidom 1 kasnejše stavbe (sl. 2.5, 2.6) tudi na severni del izkopnega polja v kvadrante 716 in 666 (sl. 2.7). Pod zidom 1 sta bili vidni še dve kulturni jami, SE 77 in SE 78, ki sta bili obe vkopani skozi kulturne plasti do skalne osnove (sl. 2.5). V severnem delu izkopnega polja je na SE 36 ležal zid 15 (sl. 2.2, 2.7). Bil je 5,4 m dolg in 0,5 m širok, v višino je bil ohranjen samo še v eno vrsto. V širino je bil zgrajen iz dveh ali celo samo iz enega večjega kamna, brez vmesne plasti drobirja. Med kamni so bili ohranjeni skromni ostanki malte. Alongside walls 1, 4 and 5 – on the outer side of the later building 1 (above SU 36) – lay a thick cultural layer (SU 24), into which walls 1 and 4 as well as a part of wall 5 were dug. The SU 24 remains were also visible at some other locations under the building walls (wall 1: Figs. 2.5, 2.6; wall 7: Fig. 2.8), while in the interior the entire layer was removed when the construction started. In some places alongside the wall SU 24 reached as high as 0.5 metres. It also covered the remains of wall 15. The layer declined from the west to the east, in the same 77 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.7: Stavba 1 – zunaj, kv. 666, 667 in 716, planum 3. M. = 1:50 Fig. 2.7: Building 1- exterior, qus. 666, 667 and 716, planum 3. Scale = 1:50. Ob zidovih 1, 4 in 5 na zunanji strani kasnejše stavbe je nad SE 36 ležala debela kulturna plast SE 24, v katero sta bila vkopana tudi zidova 1 in 4 ter del zidu 5. Ostanki SE 24 so bili vidni tudi na nekaterih mestih pod zidovi stavbe (zid 1: sl. 2.5, 2.6; zid 7: sl. 2.8), med tem ko je bila v notranjosti ob izkopu gradbene jame popolnoma odstranjena. SE 24 je ob zidovih kasnejše stavbe na nekaterih mestih segala tudi do 0,5 m visoko. Prekrivala je tudi ostanke zidu 15. Plast je padala od zahoda proti vzhodu, tako kot pada tudi geološka osnova. Zgornji nivo SE 24 ob severozahodnem vogalu je bil na pribl. 397,35 m n. m., ob vhodu v stavbo pa že 396,55 m n. m. Na delu, kjer se ob zidu 5 skalna osnova močno dvigne, se je SE 24 izklinila in se spet pojavila ob severovzhodnem vogalu. Ob zidu 4, kjer je prekrivala SE 29a, je bil njen zgornji nivo približno na višini 396,5 m n. m. (sl. 2.3). Sl. 2.8: Stavba 1, zid 7 – notranje lice in stik z zidom 1. M. = 1:50 Fig. 2.8: Building 1, wall 7 - interior and contact with wall 1. Scale = 1:50. 78 Sl. 2.9: Stavba 1, pre sek 1. M. = 1:50. Fig. 2.9: Building 1, section 1. Scale = 1:50. 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Za zidovoma 3 in 11 se skalna osnova strmo dvigne. Tu je kulturno plast prve poznoantične faze predstavljala oranžna ilovica SE 04, ki je ležala neposredno na skalni osnovi (sl. 2.9). Na zahodnem delu izkopnega polja in na območju kasnejšega prizidka stavbe 1 (zidovi 8–11) plasti, ki bi pripadale prvi poznoantični fazi, ni bilo. Tudi tu je skalna osnova precej višja kot na osrednjem delu in ruševinske plasti stavbe 1 so ležale neposredno na skali. way as the geological base. In the northwest corner the upper level of SU 24 was approximately at 397.35 m a.s.l., while at the wind shield it was already at 396.55 m a.s.l. In the part where the bedrock rises sharply alongside wall 5, SU 24 disappears only to reappear in the northeast corner. Alongside wall 4, where it covered SU 29a, its upper level was approximately at an altitude of 396.5 m a.s.l. (Fig. 2.3). The bedrock rises sharply behind walls 3 and 11. Here the cultural layer belonging to Late Antiquity phase 1 was represented by orange clay (SU 04) which lay directly on the bedrock (Fig. 2.9). Not a single layer belonging to Late Antiquity phase 1 was discovered in the western part of the excavation area and in the area that was later covered by the out house. At this location the bedrock is also much higher than in the central part and the building 1 destruction layers were discovered directly on the rock. 2.3.3 Druga poznoantična faza (PA 2) Nad ostanki prve poznoantične faze je bil postav ljen zidan bivalni objekt (stavba 1), delno vkopan vanje. Sestavljala sta ga glavni prostor pravokotnega tlorisa in nanj prizidan manjši prostor (pril. 1; sl. 2.2, 2.10). Stavba je bila usmerjena približno jugozahodnoseverovzhodno. Vhod v osrednji prostor je bil s seve rozahodne strani. Bil je zaščiten z dvema, pravokotno na zidova 1 in 5 postavljenima zidcema (zidova 6 in 7), dolgima 1,8 m in širokima 0,6 m, ki sta služila kot vetrolov. Zidova 1 in 5 je povezoval kvalitetno izdelan prag (sl. 2.10, 2.11), katerega zgornja površina je ležala na višini približno 396,6 m n. m. Na jugozahodni zid glavnega prostora (zid 2) je bil prizidan stranski prostor (prizidek), ki je bil pribl. 1,2 m krajši od glavnega prostora (pril. 1; sl. 2.12). Tudi vhod v prizidek je vodil s severozahodne strani, ostanki praga pa tu niso bili najdeni. 2.3.3 Late Antiquity phase 2 (LA 2) Above the Late Antiquity 1 remains, partially dug into them, stood a stone building (building 1). It con sisted of the main room with a rectangular ground plan and a smaller outhouse (Insert 1; Figs. 2.2, 2.10). The orientation of the building was approximately southwest-northeast. The entrance into the main room was located on the northwest and was protected by two small walls (walls 6 and 7). The walls (1.8 m long and 0.6 m wide) were set at a right angle to walls 1 and 5 and Sl. 2.10: Stavba 1, glavni prostor in del prizidka. Pogled s severozahoda. Fig. 2.10: Building 1, main room and part of the outhouse. A view from the northwest. 80 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings served as a wind shield. Walls 1 and 5 were connected by a high quality threshold (Figs. 2.10, 2.11), the upper surface of which stood at approximately 396.6 m a.s.l. The adjacent room (the outhouse) was added to wall 2 of the main space, and it was approximately 1.2 m shorter than the main room (Insert 1; Fig. 2.12). The entrance into the outhouse was also located on the northwest, however no threshold remains were found. Walls (Insert 1) The walls of the main room of building 1 were constructed from limestone quarry stones. They were mostly constructed from two rows of stones, and were between 55 and 60 cm thick. The space between the rows was filled with gravel and mortar. Wherever the walls were not placed onto the bedrock, but onto looser layers, the foundations were constructed from larger stones, set in a single row. The side wall surfaces were covered in mortar so that only the top of the stones could be seen. The walls of the outhouse were built almost entirely without mortar and are narrower (approx. 50 cm), con structed from two or sometimes even a single row of stones, without a layer of gravel in between. Sl. 2.11: Stavba 1, vhod s pragom in vetrolovom. Fig. 2.11: Building 1, entrance with threshold and wind shield. Zidovi (pril. 1) Zidovi glavnega prostora stavbe 1 so bili zidani iz apnenčevih lomljencev. Večinoma so bili grajeni v dveh vrstah, široki med 55 in 60 cm. Prostor med vrstami je bil zapolnjen z gruščem in malto. Na mestih, kjer zidovi niso bili postavljeni na skalno osnovo, ampak na mehkejše plasti, so bili temelji grajeni iz večjih kamnov, postavljenih v eni vrsti. Lica so bila zamazana z malto, tako da so bili vidni samo vrhovi kamnov. Sl. 2.12: Stavba 1, prizidek. Pogled s severovzhoda. Fig. 2.12: Building 1, outhouse. A view from the northeast. 81 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Wall 1 (Figs. 2.5, 2.6, 2.13) Zidovi stranskega prostora so bili zidani skoraj brez malte in so ožji (okrog 50 cm), grajeni v dveh ali pa tudi le v eni vrsti, brez vmesne plasti grušča. Wall 1 is 4.4 m long and is connected to wall 5 with a threshold. On the inner side a diagonal crack runs from the top to the very foundations. At the contact with wall 2 it was placed directly onto the bedrock, while at the entrance it stood on SU 24 which represents the remains of an older settlement into which two pits (SU 77 and SU 78) were dug in, both with a strong presence of cultural remains and both of which can be seen only in the building interior (Fig. 2.5). At the entrance wall 1 is based on a series of large, slightly protruding quarry stones (Fig. 2.13), while wall 7 (the wind shield) is at tached at a right angle to the outer side. Zid 1 (sl. 2.5, 2.6, 2.13) Zid 1 je dolg 4,4 m in s pragom povezan z zidom 5. Na notranji strani poteka od vrha do temeljev poševna razpoka. Ob stiku z zidom 2 je bil postavljen na skal no osnovo, ob vhodu pa na plast SE 24, ki predstavlja ostanek starejše poselitve, v katero sta bili vkopani še dve močno kulturni jami SE 77 in SE 78, vidni samo na notranji strani stavbe (sl. 2.5). Ob vhodu zid 1 temelji na vrsti večjih, nekoliko izstopajočih kamnitih lomljen cev (sl. 2.13), na njegovo zunanje lice pa je pravokotno prizidan zid 7 (vetrolov). Wall 2 (Figs. 2.14, 2.15) This wall stands on Late Antiquity 1 layers SU 36 and SU 20 almost along its entire length. The base is repre sented by a line of large non-plastered quarry stones, the upper surface of which is levelled by a layer of mortar, onto which smaller stones were placed. In the south a series of flat stones formed a 1.4 metre long and roughly 10 cm wide fissure between the base and the wall (Fig. 2.15). The stones above the base were covered by a thick layer of plaster, thus only the tops of the stones could be seen in some places. At the point where the outhouse was attached to it the outer side of wall 1 was also cov ered in plaster. Sl. 2.13: Stavba 1, zid 1 – notranje lice. Fig. 2.13: Building 1, wall 1 - interior. Zid 2 (sl. 2.14, 2.15) Zid skoraj v celi dolžini stoji na plasteh prve po znoantične faze SE 36 in SE 20. Temelje sestavlja vrsta večjih neometanih kamnitih lomljencev, katerih zgornja površina je zravnana s plastjo malte, nanje pa so postav ljeni manjši kamni zidu. V južnem delu je v dolžini 1,4 m med temeljem in samim zidom s ploščatimi kamni izo blikovana približno 10 cm široka reža (sl. 2.15). Wall 3 (Figs. 2.16-2.18) This is the longest (10.8 m) single wall of building 1. It was built on a very uneven terrain, for in the central part it stands on a rock that rises approximately 1.2 m Sl. 2.14: Stavba 1, zid 2 – notranje lice. M. = 1:50. Fig. 2.14: Building 1, wall 2 - interior. Scale = 1:50. 82 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.15: Stavba 1, stik zidov 2 in 3. Odprtina med kamni v zidu 2 in iz malte oblikovana polička v kotu. Fig. 2.15: Building 1, contact between walls 2 and 3. The fissure between the stones in wall 2 and the mortar shelf in the corner. Kamni zidu nad temelji so bili ometani z veliko količino malte, tako da so na nekaterih mestih vidni le vrhovi kamnov. Ometana je bila tudi zunanja stran zidu 1, kjer je bil nanj prizidan stranski prostor. Zid 3 (sl. 2.16-2.18) Je najdaljši (10,8 m) sklenjeni zid stavbe. Zgrajen je bil na zelo neravnem terenu, saj na osrednjem delu stoji na skali, ki se pribl. 1,2 m visoko dviga nad siceršnjo osnovo terena, takšna pa je tudi razlika med spodnjim nivojem zidu na osrednjem in severovzhodnem delu. Zid je bil zgrajen dokaj neenakomerno. Ob stiku z zidom 4 je v dolžini približno 3,5 m temeljen na eni vrsti večjih kamnov, ki stojijo na sterilni glini (SE 40). Ob stiku z zidom 2 stoji na kulturni plasti prve poznoantične faze SE 04. Tu je spodnja vrsta kamnov porušena navzdol. Na osrednjem delu, kjer sega nivo skale najvišje, posebnega temeljenja ni, zid je na tem mestu zgrajen iz dokaj pravilno izoblikovanih kamnov v poravnanih vrstah (sl. 2.16, 2.17). Na severovzhodnem delu, kjer nivo skale močno pade, je v dolžini približno 2,4 m v treh vrstah opazno močno temeljenje (sl. 2.16). Blizu vogala z zidom 4 je nad nivojem temeljnih kamnov izoblikovana polička iz malte (sl. 2.18). Prostori med kamni zidu na notranji strani so bili ometani precej manj kot pri zidu 2, vendar ni povsem jasno, ali je to prvotno stanje ali stanje ohranjenosti ometa. Zunanja stran zidu 3, ki gleda proti strmemu pobočju, ni bila ometana. Sl. 2.16: Stavba 1, zid 3 – notranje lice. M. = 1:50. Fig. 2.16: Building 1, wall 3 - interior. Scale = 1:50. 83 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.17: Stavba 1, zid 3 – notranje lice. Detajl osrednjega dela zidu. Fig. 2.17: Building 1, wall 3 - interior. Detail of the central part. Sl. 2.19: Stavba 1, zid 4 – zunanje lice. Fig. 2.19: Building 1, wall 4 - exterior. Sl. 2.18: Stavba 1, stik zidov 3 in 4. Iz malte oblikovana polička v kotu. Fig. 2.18: Building 1, contact between walls 3 and 4. The mortar shelf in the corner. Sl. 2.20: Stavba 1, zid 5 – zunanje lice. Fig. 2.20: Building 1, wall 5 - exterior. Zid 4 (sl. 2.4, 2.19) above the clay base, which also represents the distance between the lower level of the wall in the central and northeastern part. The wall was built rather unevenly. At the contact with wall 4, approximately 3.5 m in length, the foundations are represented by a single line of larger stones that are positioned on sterile clay (SU 40). At the contact with wall 2 it stands on the cultural layer SU 04 that belongs to Late Antiquity phase 1. At this point the lower line of stones has crumbled downwards. In the central part of the wall, where the rock level is at its highest, there are no special foundations, and the wall is built from relatively well formed stones that run in a straight line (Figs. 2.16, 2.17). In the northeast, where the base drops considerably, strong foundations measuring approximately 2.4 metres in length and made of three rows of stones were discovered (Fig. 2.16). A small mortar shelf was created close to the corner with wall 4, above the level of the base stones (Fig. 2.18). On the inner side of wall 3 the spaces between the stones were plastered by a much thinner layer of plaster Zid je bil vkopan v mehko kulturno plast SE 29a skoraj do pod njo ležeče rjave peščene plasti SE 32 (sl. 2.3). Temelji zidu so bili grajeni iz velikih kamnitih lomljencev, kar je še posebej izrazito ob stiku z zidom 5 (sl. 2.19). Tudi pri zidu 4 je na notranji strani (podobno kot pri zidu 2) v dolžini približno 1,5 m med temelji in sa mim zidom opazna vodoravna reža, ki pa je tukaj manj izrazita. Prostor med kamni na zunanji in notranji strani zidu je bil obilno ometan. Zid 5 (sl. 2.20) Zid je bil skoraj v celoti postavljen na skalo, razen ob vhodu v stavbo, kjer je v dolžini pribl. 1 m temeljen na glineno osnovo. Nivo skale ob stiku z zidom 4 močno pade in na tem mestu je bil zgrajen močan temelj iz velikih kamnitih blokov (glej tudi opis zidu 4). Ob vhodu v stavbo je bil pravokotno na zid 5 prizidan zid vetrolova (zid 6). 84 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings than at wall 2, however it is not absolutely clear whether this was the original state or is this merely the state of preservation. The outer side of wall 3 (looking towards the steep slope) was not covered in plaster. Wall 4 (Figs. 2.4, 2.19) The wall was dug into the loose cultural layer (SU 29a) almost all the way to the brown sand layer (SU 32) that lies underneath (Fig. 2.3). The wall foundations were constructed from large quarry stones, which is especially noticeable at the contact with wall 5 (Fig. 2.19). On the inner side of wall 4 (similar as is the case with wall 2) a horizontal fissure was discovered between the foundation and the wall itself. The fissure measures approximately 1.5 m in length, however it is not as no ticeable as the fissure in wall 2. The space between the stones on the outer as well as on the inner side of the wall was covered by a thick layer of plaster. Sl. 2.21: Stavba 1, detajl vhodnega dela, zid 6 – notranje lice in stik z zidom 5 in pragom. Fig. 2.21: Building 1, detail of the entrance, wall 6 - interior and contact with wall 5 and threshold. Wall 5 (Fig. 2.20) Almost the entire wall was positioned on the bed rock, except for the entrance into the building that was positioned on a clay base (approximately 1 metre in length). At the contact with wall 4 the bedrock declines sharply and thus a strong foundation from large stone blocks was constructed (see the description of wall 4). At the entrance into the building the wall of the wind shield (wall 6) was built at a right angle to wall 5. Sl. 2.22: Stavba 1, zid 7 – notranje lice in stik z zidom 1 in pragom. Fig. 2.22: Building 1, wall 7 - interior and contact with wall 1 and threshold. Wall 6 (Figs. 2.2, 2.11, 2.21) Zid 6 (sl. 2.2, 2.11, 2.21) The wall belonging to the wind shield was built at a right angle to wall 5 in such a way that it covered approximately 20 cm of the inner side of the entrance (Insert 1; Figs. 2.2, 2.11, 2.21). It was based on the rock approximately 20 cm below the wall of the building itself. Only two to three rows were preserved in height. Severovzhodni zid vetrolova je bil prizidan pravo kotno na zid 5, tako da je bil zamaknjen pribl. 20 cm proti notranjosti vhoda (pril. 1; sl. 2.2, 2.11, 2.21). Temeljen je na skali oziroma skalni preperini, in sicer pribl. 20 cm globlje kot zid same stavbe. V višino so bile ohranjene le dve do tri vrste. Wall 7 (Figs. 2.8, 2.11, 2.22, 2.29) Zid 7 (sl. 2.8, 2.11, 2.22, 2.29) This wall was built at a right angle to wall 1, ap proximately 20 cm towards the entrance interior. The foundations of wall 7 were also approximately 20 cm lower than the wall of the building and the threshold, and stood on the cultural layer SU 24, which belonged to the first Late Antiquity phase (Fig. 2.8). The foundation was constructed from larger, and the wall from smaller, relatively well worked flat stones. The contacts between them were covered by a thick layer of plaster. Prizidan je bil pravokotno na zid 1 tako, da je bil prib. 20 cm zamaknjen proti notranjosti vhoda. Tudi zid 7 je bil temeljen pribl. 20 cm nižje kot sam zid stavbe in prag, postavljen pa je bil na SE 24, to je kulturno plast prve poznoantične faze (sl. 2.8). Temelj je bil zgrajen iz večjih, sam zid pa iz manjših, dokaj pravilno oblikovanih ploščatih kamnov. Stiki med njimi so bili obilno prekriti z ometom. 85 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.23: Stavba 1, zid 8 in stik z zidom 2. Fig. 2.23: Building 1, wall 8 and contact with wall 2. Sl. 2.24: Stavba 1, del zidu 10 – notranje lice. Fig. 2.24: Building 1, part of wall 10 - interior. Zid 8 (sl. 2.23) Wall 8 (Fig. 2.23) Zid je bil prizidan na zid 2 glavnega prostora pribl. 1 m od vogala zidov 1 in 2 (sl. 2.2). Bil je zgrajen potem, ko je bil zid glavnega prostora že ometan. 1 m po stiku z zidom 2 je prekinjen z vhodom. Brez posebnega temeljenja je bil postavljen na SE 14 približno 30 cm višje kot zid 2 glavnega prostora. Zid 8 ni bil ometan, tudi v njegovi notranjosti ni bilo opaziti nobenih sledov malte. The wall was attached to wall 2 of the main room roughly 1 metre away from the corner between walls 1 and 2 (Fig. 2.2). It was constructed once wall 2 was al ready covered by a layer of plaster. 1 metre after the con nection with wall 2 the wall was cut off by the entrance. The wall was built on SU 14 approximately 30 cm higher than wall 2 and did not have any special founda tions. Wall 8 was not covered in plaster, and no traces of mortar were found on the inner side. Zid 9 Wall 9 Zid je dolg 1,4 m in podobno kot zid 8 neometan, brez sledov malte in posebnega temeljenja. V kotu, ki ga tvori z zidom 10, je bilo ohranjeno ognjišče (ognjišče 1). The wall is 1.4 m long and similar to wall 8 it is not covered in plaster, it shows no traces of mortar and has no explicit foundations. A fireplace was preserved in the corner with wall 10 (fireplace 1). Zid 10 (sl. 2.24) Wall 10 (Fig. 2.24) Zid v večini leži na skalni osnovi, ki močno pada od jugovzhoda proti severozahodu, le ob stiku z zidom 9, kjer se teren nekoliko zravna, leži na kulturni plasti SE 12 (sl. 2.24). Tudi pri zidu 10 ni bilo opaziti nobenega posebnega temeljenja, zid tudi ni bil ometan. Med kamni zidu je bilo ohranjenih nekaj skromnih ostankov malte. Na zunanji strani tik ob zidu je ležal skelet, položen neposredno na skalno osnovo in prekrit z ruševino zidu (grob 1). Most of the wall lies on the bedrock that sharply declines from the southeast to the northwest, for only at the point where it touches upon wall 9, i.e. where the terrain levels out slightly, is it positioned on the cultural layer SU 12 (Fig. 2.24). No explicit foundations or traces of plaster were noticed. A few traces of mortar were found amongst the stones in the wall. A skeleton was discovered on the outer side, right next to the wall. The skeleton was placed directly onto the bedrock and was covered by the wall ruins (grave 1). Zid 11 (sl. 2.25) Wall 11 (Fig. 2.25) Zid je prislonjen na zid 2 na vogalu z zidom 3 (sl. 2.25). Temeljen je pribl. 40 cm plitveje kot zid glavne stavbe. Tudi na tem mestu skalna osnova močno pada, in sicer od jugozahoda proti severovzhodu. Zid je tako v The wall reached wall 2 at the corner with wall 3 (Fig. 2.25). The foundations were approximately 40 cm 86 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.25: Stavba 1, del zidu 11 in stik z zidom 2. Fig. 2.25: Building 1, part of wall 11 and contact with wall 2. Sl. 2.26: Stavba 1, postament (zid 12). Pogled z jugovzhoda. Fig. 2.26: Building 1, pedestal (wall 12). A view from the southeast. večini postavljen na skalo, razen ob stiku z zidom 2, kjer je bil močan skalni padec zravnan s peščenim nasutjem. Samo na tem mestu je opaziti nekaj večjih kamnov, ki so služili za temelj. Zid ni bil ometan. shallower than those of the wall in the main building. At this point the bedrock also declines sharply in the northeast direction. The wall is thus mainly positioned directly on the bedrock, except where it reaches wall 2, where the great decline of the bedrock was levelled out with gravel. This is the only location where a few larger stones that used to be a part of the foundations were noticed. The wall was not covered in plaster. Zid 12 – postament (sl. 2.26, 2.27) Na zunanji strani zidu 1 je bil pribl. 1 m oddaljen od njega postavljen zidan podstavek dimenzij 70 x 70 cm, vkopan v SE 24 (sl. 2.27). Njegovi temelji segajo v globino približno toliko kot temelji zidu 1, v višino so ohranjene štiri vrste kamnov, ki jih povezuje velika količina malte. Z malto je bil močno premazan tudi na vrhu. Wall 12 - pedestal (Figs. 2.26, 2.27) On the outer side of wall 1, approximately 1 m from the wall, a 70 x 70 cm stone pedestal was dug into SU 24 Sl. 2.27: Stavba 1 – zunaj, kv. 716, 766, planum 2. M. = 1:50. Fig. 2.27: Building 1- exterior, qus. 716, 766, planum 2. Scale = 1:50. 87 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Prag (pril. 1; sl. 2.5, 2.11, 2.21, 2.22) (Fig. 2.27). The foundations are roughly the same depth as the foundations of wall 1, and four stone rows joined by large quantities of mortar remain in height. The top of the wall was also covered with large quantities of mortar. Prag je povezoval zidova 1 in 5. Zgrajen je bil iz 10–15 cm debele plasti kamenja in kakovostne malte, s katero so bili kamni na delu, obrnjenem proti notranjosti hiše, tudi premazani. Po vsej dolžini je bil v malti izdelan približno 50 cm širok in 2 cm globok utor. Prag je v več jem delu temeljil na skalni osnovi. Na strani, obrnjeni proti notranjosti stavbe, je podenj ob stiku z zidom 1 segal vkop prve poznoantične faze (SE 78), ob stiku z zidom 5 pa skromni ostanki SE 24 (sl. 2.5). Threshold (Insert 1; Figs. 2.5, 2.11, 2.21, 2.22) The threshold joined walls 1 and 5. It was con structed from a 10 to 15 cm thick layer of stones and high quality mortar, with which the stones that were turned towards the house interior were also covered. Along the entire length ran a roughly 50 cm wide and 2 cm deep groove into the mortar. To a greater extent the threshold was based on the bedrock. On the side turned towards the interior of the building, a pit (SU 78) belonging to Late Antiquity phase 1 reached under the threshold at the point where it touched wall 1. At the contact with wall 5 modest remains of SU 24 were found (Fig. 2.5). Ognjišče 1 (sl. 2.28, 2.29) V jugovzhodnem delu glavnega prostora je bilo približno 1,2 m oddaljeno od zidu 3 postavljeno zidano ognjišče, veliko približno 1,2 x 1 m. Zgrajeno je bilo iz več jih, vodoravno položenih ploščatih kamnov, ki so bili ob robu postavljeni navpično, tako da so tvorili nizko ograjo. Pri gradnji so uporabili tudi kose opeke in del zgornjega žrmljnega kamna. Na nekaterih mestih je bil še ohranjen glinen premaz, ki je nekoč verjetno pokri val celo površino. Ognjišče je bilo delno postavljeno na skalno osnovo, delno pa na glineno plast prve pozno antične faze SE 18. Fireplace 1 (Figs. 2.28, 2.29) A stone fireplace, approximately 1.2 x 1 m in size, was positioned in the southeast part of the main room, approximately 1.2 m from wall 3. The fireplace was constructed from large, horizontally placed flat stones, and on the edge the stones were placed vertically so that they formed a low fence. Brick parts and a part of the upper stone of a rotaryquern were also used in the construction. In some places a clay coating was preserved. It is highly likely that it covered the entire surface. The fireplace was partially positioned on the bedrock and partially on the clay layer SU 18 that belonged to the first Late Antiquity phase. Fireplace 2 (Insert 1) Sl. 2.28: Stavba 1, ognjišče 1, pogled z zahoda. Fig. 2.28: Building 1, fireplace 1, a view from the west. Next to the entrance in the outhouse, in the corner created by walls 9 and 10, stood a 10 cm high fireplace created from flat, heavily cracked stones and brick par ticles. It was constructed in a much poorer fashion than fireplace 1 in the main room. Sl. 2.29: Stavba 1 – notranjost, planum 2. M. = 1:50. Fig. 2.29: Building 1- interior, planum 2. Scale = 1:50. 88 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings 89 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings OGnjišče 2 (pril. 1) Remaining stratigraphic units V prizidku, v vogalu zidov 9 in 10, je bilo tik ob vhodu postavljeno ognjišče, zgrajeno iz ploščatih, močno razpokanih kamnov in delov opek, visoko do 10 cm. Bilo je grajeno precej slabše kot ognjišče 1 v glavnem prostoru. At the time the building was used a larger part of the western and central part of the main room was covered by SU 21, i.e. a layer of hard, compact clay, into which rather sharp stones, approximately the size of a fist, were positioned more or less closely together. In the southwest part of the building the layer lays at an approximate altitude of 396.8 m a.s.l., and it descends by approximately 30 cm towards the northeast. In some places parts of mortar or pieces of brick were found amongst the stones (Figs. 2.9, 2.29). The upper level SU 21 fits the level of the threshold and the height at which wall 1 comes into the rough, non-plastered foundations. The eastern part of the building is covered by SU 18. The layers are hard to distinguish, however SU 21 contained a slightly higher number of stones than SU 18. Both lay above SU 36, a Late Antiquity 1 layer, and in some places also directly on the bedrock. SU 22, a compact gravel layer was located in the centre of the space, above the bedrock. SU 18 and 21 were covered by layers of fallen off plaster (SU 08) and stones of the destruction layer. SU 21 was not preserved alongside wall 2. Along this wall the layers that emerged when the building was destroyed lay directly on SU 04, SU 20 and SU 36 (Fig. 2.29), all of which are a remnant of the older settlement (Late Antiquity phase 1), for they reached under walls 2 and 3 (Figs. 2.14, 2.16). The floor surface on the outer side of walls 1, 4 and 5 was represented by SU 24, which belongs to Late Antiquity phase 1 (see chapter 2.3.2). Walls 1 (Fig. 2.6), 4 (Fig. 2.3) and 5 of the main room as well as walls 6, 7 (Fig. 2.8, wind shield) and 12 (Fig. 2.27, pedestal) were dug into this layer. The top level of SU 24 declines sharply from the southwest to the northeast. The only place where SU 24 was preserved also within the building interior was in the outhouse. Here the layer reached under the wall 9 and into the outhouse interior. It was covered by some sort of levelling created from hard compact soil (SU 12), which represented the floor surface alongside wall 10, and fireplace 2 was placed onto it. SU 24 outside of building 1 was covered by thick cultural layers originating from the time the building was in use. On the outer side of wall 1, around the ped estal (wall 12), an approximately 4 m long and 1 m wide compact layer of mortar (SU 35) stood on top of SU 24 (Fig. 2.27). The layer was loose and measured between 5 and 40 cm in thickness. Between the pedestal and the wall an irregular shaped pit measuring approximately 1 m2 in size (SU 35a), and filled with soil, mortar and stones was discovered. Above SU 24 outside the building, alongside walls 1, 4 and 5 lay an up to 40 cm thick loose cultural layer (SU 23; Fig. 2.3). Similar to the underlying level SU 24 its level also steeply declined from the southwest towards the Ostale stratigrafske enote Večji del zahodnega in osrednjega dela glavnega prostora je v času uporabe stavbe zavzemala SE 21, to je plast trde, zbite gline, v katero so bili bolj ali manj strnjeno položeni približno pest debeli, precej ostrorobi kamni. Nivo plasti je pribl. 396,8 m n. m. v jugozahodnem delu stavbe, proti severovzhodu pa pade za približno 30 cm. Na nekaterih mestih so bili med kamni ohranjeni zaplate malte in kosi opek (sl. 2.9, 2.29). Zgornji nivo SE 21 (približno 396,60 m n. m.) ustreza nivoju praga ter višini, kjer zid 1 prehaja v grobo zidan, neometan temelj. Vzhodni del stavbe zavzema SE 18, ki je bila slabo ločljiva od SE 21, vsebovala je le nekoliko manj kamnov kot SE 21. Obe plasti sta prekrivali SE 36, ki pripada prvi poznoantični fazi, na nekaterih mestih pa sta ležali neposredno na skalni osnovi. V osrednjem delu prostora nad skalno osnovo je ležala SE 22, trda plast grušča. Nad SE 18 in 21 so ležale plasti odpadlega ometa (SE 08) in ruševina stavbe 1. Ob zidu 2 SE 21 ni bila ohranjena, tu so plasti, ki so nastale ob rušenju hiše, ležale neposredno na SE 04, SE 20 in SE 36 (sl. 2.29), ki so še ostanek starejše poselitve (prve poznoantične faze), saj segajo pod zidova 2 in 3 (sl. 2.14, 2.16). Hodno površino na zunanji strani zidov 1, 4 in 5 je predstavljala SE 24, ki pripada še prvi poznoantični fazi (glej pogl. 2.3.2). Vanjo so bili vkopani zidovi 1 (sl. 2.6), 4 (sl. 2.3) in 5 glavnega prostora ter 6, 7 (sl. 2.8, vetrolov) in 12 (sl. 2.27, postament). Zgornji nivo SE 24 močno pada od jugozahoda proti severovzhodu. Edini primer, ko je bila SE 24 ohranjena tudi v notranjosti stavbe, je bil v prizidku. Tu je segala pod zid 9 v severozahodni vogal prizidka. Nad njo je ležala nekakšna izravnava iz trde zbite zemlje (SE 12), ki je predstavljala hodni nivo ob zidu 10, nanjo pa je bilo postavljeno ognjišče 2. SE 24 zunaj stavbe 1 so prekrivale debele kulturne plasti iz časa poselitve stavbe. Na zunanji strani zidu 1 okrog postamenta (zid 12) je nad SE 24 ležala pribl. 4 m dolga in 1 m široka strnjena plast malte SE 35 (sl. 2.27). Plast je bila mehka in debela od 5 do 40 cm. Med po stamentom in zidom je bila prebita s pribl. 1 m2 velikim vkopom nepravilne oblike (SE 35a), ki ga je zapolnjevala zemlja, pomešana z malto in kamni. Nad SE 24 je ob zidovih 1, 4 in 5 ležala do 40 cm debela, mehka, močno kulturna plast SE 23 (sl. 2.3). 90 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.30: Stavba 1 – zunaj, kv. 666, 667 in 716, planum 2. M. = 1:50. Fig. 2.30: Building 1- exterior, qus. 666, 667 and 716, planum 2. Scale = 1:50. Tudi njen nivo je, podobno kot nivo spodaj ležeče SE 24, močno padal od jugozahoda proti severovzhodu. Plast je vsebovala veliko najdb (keramika, opeka, kosti, drobci žganine), večina pa jih je ležala v njenem spodnjem delu, na meji s SE 24. V kvadrantu 666 severno od zidu 15 je na SE 57, ki predstavlja poselitev prve poznoantične faze, ležalo več močno kulturnih plasti (SE 56, 50), ki so bile po strukturi zelo mehke. V te plasti so bile postavljene skupine ravnih kamnitih plošč (SE 55) ali posamezne ravne plošče (sl. 2.30). Nekatere so bile ožgane od ognja in močno razpokane. Plošče je na jugozahodnem delu kvadranta 666 prekrivala SE 23, ki je proti severu in vzhodu prehajala v sorodno SE 29. Na južni zunanji strani stavbe, tik pod skalno steno, je hodno površino iz časa poselitve objekta predstavljal zgornji nivo SE 04, ki pripada še prvi poznoantični fazi. Nad SE 04 je na stiku zidov 2 in 11 ležala intenzivna žganinska, močno kulturna plast SE 25. Območje je nato prekrivala plast rjave, z gruščem mešane prsti SE 03 (sl. 2.9), podobne SE 23 na severni in vzhodni strani objekta. Kulturne plasti druge poznoantične faze (SE 62, 69) so se northeast. The layer included numerous finds (pottery, bricks, bones, parts of charred material), most of which were found in its lower part, on the border with SU 24. In quadrant 666, to the north of wall 15, on top of SU 57 which represents the Late Antiquity phase 1, some layers high in cultural finds (SU 56, 50) were discovered and all of them had a loose structure. Groups of flat stone slabs (SU 55) or individual flat slabs were positioned into these layers (Fig. 2.30). Some slabs were charred by the fire and had cracked. The slabs in the southwest part of quadrant 666 were covered by SU 23, which transformed into the similar SU 29 towards the north and east. On the south exterior of the building, just below the rock wall, the floor surface of the Late Antiquity Phase 2 was represented by the upper level of SU 04 which belongs to Late Antiquity 1. Above SU 04, on the connection of walls 2 and 11, the highly charred layer SU 25, full of finds, was discovered. The area over SU 25 was covered by a layer of brown soil mixed with gravel (SU 03), similar to SU 23 on the north and east of the building (Fig. 2.9). Culturally rich Late Antiquity 2 layers (SU 62, 69) continued further upwards along the steep 91 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings 92 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings rocky slope. Here they lay directly on the bedrock, and they were covered by the humus layer SU 34. The floor surface in the main room (SU 18, 21) was covered by a thick layer of charred debris that emerged once the building was no longer in use. The northwest part of the building (SU 18, 21) with the fireplace was covered by a strongly charred layer SU 11 (Figs. 2.29, 2.31). Along all four walls of building 1 lay a thick (in some places up to 60 cm) layer of fallen off plaster (SU 08; Figs. 2.29, 2.31). In the central part of wall 3 and along wall 5 the layer was positioned directly onto the bedrock, at wall 4 it could be found on SU 18 and alongside wall 2 on SU 36 and SU 21. In the outhouse a layer of fallen off plaster (SU 08) was discovered merely along wall 2 (shared with the main room). The entire area of the outhouse and its walls was covered by a layer of large stones (SU 01) that stood directly on the bedrock in the centre of the room, and on SU 09 in the northwest corner. The layer of plaster that has fallen off the walls (SU 08) can be found also in an approximately 0.5 m wide strip on the outer sides of walls 1, 2, 4 and 5 (Fig. 2.3), however it was not to be found on the outer side of wall 3. Most of it lay on SU 23 and SU 29, however alongside wall 5, where the bedrock rises sharply it was located directly on the bedrock (Fig. 2.9). Sl. 2.31: Stavba 1 – notranjost, planum 1. M. = 1:50. Fig. 2.31: Building 1- interior, planum 1. Scale = 1:50. nadaljevale tudi še navzgor po strmem skalnem pobočju. Tu so ležale neposredno na skalni osnovi, prekrite pa so bile s humusom SE 34. Hodno površino v osrednjem prostoru (SE 18, 21) so prekrivale močne ruševinsko-žganinske plasti, ki so nastale, ko hiša ni bila več v uporabi. Severovzhodni del stavbe (SE 18, 21) z ognjiščem je prekrivala močna žganinska plast SE 11 (sl. 2.9, 2.31). Ob vseh štirih zido vih stavbe 1 je ležala debela (na nekaterih mestih tudi do 60 cm) plast odpadlega ometa SE 08 (sl. 2.29, 2.31). Plast je ob osrednjem delu zidu 3 in ob zidu 5 ležala neposredno na skali, ob zidu 4 nad SE 18 in ob zidu 2 nad SE 36 in SE 21. V prizidku je bila plast odpadlega ometa SE 08 vidna le ob zidu 2, ki je bil skupen z glavnim prosto rom. Celotno območje prizidka in njegovih zidov je prekrivala plast velikih ruševinskih kamnov SE 01, ki je v osrednjem delu prostora ležala neposredno na skalni osnovi, v severozahodnem vogalu pa nad SE 09. Plast odpadlega ometa SE 08 je ležala tudi v pribl. 0,5 m širokem pasu na zunanjih straneh zidov 1, 2, 4 in 5 (sl. 2.3), ni pa je bilo na zunanji strani zidu 03. Večinoma je ležala na SE 23 in SE 29, ob zidu 5, kjer se skalna osnova močno dvigne, pa neposredno na njej (sl. 2.9). 2.3.4 Early Middle Ages Traces of Early Medieval settlement were preserved in the destruction layers of building 1. In the interior of the main room SU 08 and SU 11 were covered by a layer of loose black soil mixed with individual large stones (SU 10), which covered SU 21 towards the centre of the room. The layer was rich in cultural finds and contained a few Early Medieval finds. Above SU 10 lay a layer of large stones and humus (SU 01). This layer covered the entire interior of the building and its walls except for the southeast corner, where it was removed during recent trenching SU 02 (Fig. 2.32) that also penetrated the underlying layers (Figs. 2.29, 2.31). A strongly charred layer (SU 09) was also docu mented in the outhouse. Here it covered the fireplace and SU 12. In the southeast corner of the outhouse, directly on the bedrock, another strongly charred layer (SU 13) was discovered. A skeleton was discovered behind wall 10, just outside of the outhouse (grave 1). It was positioned parallel to wall 10 so that the corpse lay with its head pointing in the northwest direction (Fig. 2.33). The skeleton was placed directly onto the untreated bedrock, and was covered by a thin layer of soil, that was in turn covered by the destruction layer SU 01. No grave goods accompanied the skeleton. Its time definition therefore isn’t possible (see chapter 3.1.1). 2.3.4 Zgodnji srednji vek V ruševinskih plasteh stavbe 1 so bili ohranjeni tudi sledovi zgodnjesrednjeveške poselitve. SE 08 in 11 je v notranjosti glavnega prostora prekrivala plast črne mehke zemlje, pomešane s po sameznimi velikimi kamni (SE 10), ki je v osrednjem prostoru ležala na SE 21. Plast je bila močno kulturna in je vsebovala tudi nekaj zgodnjesrednjeveških najdb. Nad SE 10 je ležala še plast velikih ruševinskih kamnov in humusa (SE 01), ki je prekrivala vso notranjost hiše in njene zidove razen jugovzhodnega vogala, kjer je bila odstranjena z recentnim vkopom SE 02 (sl. 2.32), ki je prebil tudi pod njo ležeče plasti (sl. 2.29, 2.31). Močno žganinska plast SE 09 je bila dokumentirana tudi v prizidku. Tu je prekrivala ognjišče in SE 12. Še ena 93 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings Sl. 2.32: Stavba 1, izkopno polje z očiščeno ruševino. Pogled s severozahoda. Fig. 2.32: Building 1, excavation area with cleared ruins. A view from the northwest. The destruction layer SU 01 covered almost the entire area of building 1 and its surroundings. The destruction layer declined from the south towards the north and from the west towards the east. The ruins were clearly limited and reached ap proximately 2 metres from the walls of the building, except at the outhouse, where the ruins were not as dense (Fig. 2.32) and lay directly on the bedrock. skala / bedrock 398,53 398,55 ZID 10 / WALL 10 x = 62 y = 60 398,76 398,80 398,86 398,72 398,81 398,85 398,82 398,85 398,85 398,9 398,99 399 398,95 398,94 398,93 398,41 0 Sl. 2.33: Stavba 1, grob 1 za zidom 10. M. = 1:20 Fig. 2.33: Building 1, grave 1 behind wall 10. Scale = 1:20. 1m 94 2.3 Stavba 1 z okolico 2.3 Building 1 and its surroundings močno žganinska plast (SE 13) je ležala v jugozahodnem vogalu prizidka, neposredno na skalni osnovi. Na zunanji strani prizidka, za zidom 10, je bil odkrit skelet (grob 1), ki je ležal vzporedno z zidom 10, tako da je bil z glavo usmerjen proti severozahodu (sl. 2.33). Skelet je bil položen neposredno na skalno osnovo, ki ni bila posebej obdelana, prekrit pa je bil le s tanko plastjo zemlje, čez katero je ležala ruševinska plast SE 01. Bil je brez pridatkov in ga zato časovno ni mogoče zanesljivo uvrstiti (glej pogl. 3.1.1). Plast ruševine SE 01 je prekrivala skoraj vse obmo čje stavbe 1 in njegovo okolico. Nivo ruševine je padal od juga proti severu ter od zahoda proti vzhodu. Venec ruševine je bil jasno omejen in je segal približno 2 m od zidov stavbe razen ob prizidku, kjer je bilo ruševine nekoliko manj (sl. 2.32), na tem mestu je ležala nepo sredno na skalni osnovi. 95 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 V bližini severne cerkve sta ležali dve med seboj povezani zidani stavbi, ki smo ju poimenovali stavba 2 in stavba 3 (sl. 2.34). Bili sta umeščeni v sedlo med platojem s cerkvami in najvišjim vrhom Tonovcovega gradu (sl. 1.7, 3.22). Pobočje sedla se na vzhodni in zahodni strani strmo dviga tik za zidovi samih stavb, tako da je nekoliko bolj raven vmesen prostor širok približno 5 m. V smeri proti severu in jugu je teren ravnejši. Na južni strani stavbe 2 se skalna terasa razširi tudi proti vzhodu, pod vzhodne stranice cerkvenega sklopa, na severni strani stavbe 3 pa vodi naraven prehod v spodnji del naselja. Pred izkopavanjem je bil v terenu viden samo obris stavbe 2, ki ga je prekrivala velika gmota ruševine. Na Two connected masonry buildings stood in the vicinity of the north church, i.e. buildings 2 and 3 (Fig. 2.34). They were positioned in the saddle between the plateau with the churches and the highest peak of Tonovcov grad (Figs. 1.7, 3.22). The sides of the saddle rise sharply towards the east and the west right behind the walls of the buildings. The slightly levelled out area between them measures approximately 5 m in width. In the north and south directions the terrain is slightly more levelled out. On the south side of building 2 the rocky terrace spreads towards the east (under the east walls of the ecclesiastical complex), while on the north side of building 3 a natural path leads towards the lower part of the settlement. Prior to the excavations merely an outline of building 2 could be seen, covered by a large pile of ruins. No ruins covered the area of building 3, where the terrain was only slightly deepened and covered by humus. Building 2, which was visible prior to the start of the excavations, was researched in 2002. At this it was ascertained that the north wall of building 2 (SU 121), which was built with mortar, also represented the south wall of an older building (building 3) and that the remaining three walls of building 2 were attached to this wall. Four graves were discovered south of building 2 and the excavations stopped at that even though there is a possibility that there are more graves towards the south. The steep rocks towards the east and west prevented any graves in these directions (Fig. 1.7). The area covered by building 3 was researched in 2005. This building had been severely destroyed, and only wall SU 121, which was shared with building 2, was completely preserved. The west wall (SU 126) was preserved approximately 3 m in length, while the east wall (SU 127) was preserved approximately 1 m in length. The researched area failed to reveal any wall remains to the north, thus it is unclear where and how this building ends (Fig. 2.34). Sl. 2.34: Stavbi 2 in 3 z mejo izkopnega polja in mrežo kvadrantov. M. = 1:200 Fig. 2.34: Ground plan of building 1 with the excavation area and quadrant grid. Scale = 1:200. 97 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 Tab. 2.2: Stavbi 2 in 3. Preglednica stratigrafskih enot (SE). Tab. 2.2: Buildings 2 and 3. Table of stratigraphic units (SU). Opredelitev / Definition Sterilna / Sterile Prazgodovina / Prehistory Antika / Antiquity PA 1 / LA 1 PA 1/PA 2 / LA 1/LA 2 PA 2 / LA 2 ZSV / EM Premešano / Mixed SE / SU 117, 139, 143, 151, 157, 173, 176, 178, 183 180=140 171, 172=165, 177, 179 121, 126, 127, 160, 170, 175 142, 174 103, 106, 108, 110, 111, 112, 113, 120, 122, 123, 124, 147, 150=144, 152, 155, 158, 161, 162, 164, 166, 169 grob / grave 18, grobovi / graves 15?, 19?, 20? 135, 137 območju stavbe 3 ruševine ni bilo, tu je bil teren le nekoliko vglobljen in prekrit s humusom. V letu 2002 je bila najprej raziskana stavba 2, ki je bila vidna že pred začetkom izkopavanj. Pri tem je bilo ugotovljeno, da je bil z malto zidan severni zid stavbe 2 (SE 121) hkrati tudi južni zid neke starejše stavbe (stavba 3) in da so bili preostali trije zidovi stavbe 2 na ta zid prizidani. Južno od stavbe 2 so bili odkriti štirje grobovi. Izkop je bil tu končan navkljub možnosti, da se grobovi nadaljujejo tudi še proti jugu. Proti vzhodu in zahodu je njihovo nadaljevanje zaradi strmih skalnih sten izključeno (sl. 1.7). Leta 2005 je bilo raziskano še območje stavbe 3. Izkazalo se je, da je ta zelo uničena, v celoti je bil ohranjen samo zid SE 121, ki je tudi skupni zid s stavbo 2. Zahodni zid (SE 126) je bil ohranjen v dolžino približno 3 m, vzhodni (SE 127) pa v dolžino enega metra. Na severnem delu na raziskanem prostoru ni bilo najdenih nobenih ostankov zidu, tako da ni jasno, kje in kako se je objekt zaključil (sl. 2.34). Na območju izkopnega polja stavb 2 in 3 so bili z gotovostjo ugotovljeni prazgodovinski poselitveni ostanki ter poselitev prve in druge poznoantične faze (PA 1 in PA 2). Nekaj skromnih naselbinskih ostankov pod plastmi prve poznoantične faze kaže tudi na možnost poselitve že v antičnem obdobju, ki pa je časovno ne moremo natančneje opredeliti (glej pogl. 2.2). Eden izmed štirih grobov za stavbo 2 (grob 18) vsebuje tudi pridatke, datirane v zgodnji srednji vek (tab. 2.2). Prehistoric settlement remains as well as settlement remains from the Late Antiquity phases 1 and 2 (LA 1 and LA 2) were confirmed in the excavation area of buildings 2 and 3. An earlier settlement is indicated by certain modest settlement remains found underneath the Late Antiquity 1 layers, however it cannot be precisely dated (see chapter 2.2). One of the four graves behind building 2 (grave 18) includes grave goods that were dated into the Early Medieval period (tab. 2.2). 2.4.1 Settlement prior to the buildings Throughout the majority of the excavation area the bedrock was covered by a culturally sterile compact orange clay layer (SU 183; Fig. 2.35), which descended from the south towards the north. In the area of building 3 the upper level descended by approximately 1 m over the 9 m in length. A similar drop was established in the cultural layers accumulated upon this base. In the south of the excavation area, in the area covered by building 2, SU 183 levels out, as does the prehistoric layer above it. The rock represented geological base in the east and west edge of the south part of the excavation area (Fig. 2.43). Traces of a prehistoric settlement were documented within a narrow zone in the west part of the excavation area (in the area of the later building 3), and towards the south, to the area of the later building 2. Prehistoric remains were discovered in the layer of brown-red clay SU 180 (SU 140 in the area of building 2), which included large quantities of sandstone and pieces of large limestone (Figs. 2.35-2.36, 2.39, 2.42-2.44; see also Tonovcov grad. Finds, chapter 6.5). The majority of the eastern part of the excavation area, under the later building 3, the geological base SU 183 and partially the prehistoric layer SU 180 was covered by an up to 30 cm thick sand layer (SU 176), which is most likely of natural (erosion) origin (Figs. 2.35-2.38). In the northwest part of the excavation area a layer of grey soil SU 179 (Fig. 2.37), which included modest 2.4.1 Poselitev pred izgradnjo stavb Skalno osnovo na večini izkopnega polja je prekrivala kulturno sterilna trda oranžna glinena plast (SE 183; sl. 2.35), katere nivo je padal od juga proti severu. Na območju stavbe 3 se je tako njen zgornji nivo v dolžini 9 m znižal za približno meter. Podobno so padale tudi kulturne plasti, ki so se nalagale na to osnovo. Na južnem delu izkopnega polja, na območju stavbe 2, se nivo SE 183 zravna, prav tako nivo nad njo ležeče prazgodovinske 98 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 Sl. 2.35: Stavba 3, presek 2. M. = 1:50. Fig. 2.35: Building 3, section 2. Scale = 1:50. plasti. Ob vzhodnem in zahodnem robu južnega dela izkopnega polja je geološko osnovo predstavljala skala (sl. 2.43) in kulturne plasti so tukaj ležale neposredno na njej. Sledovi prazgodovinske poselitve so bili dokumentirani v ozkem pasu na zahodnem delu izkopnega polja na območju stavbe 3, poselitev pa je segala tudi proti jugu na prostor stavbe 2. Prazgodovinski ostanki so bili najdeni v plasti rjavo-rdeče gline SE 180 (na območju stavbe 2 kot SE 140), v kateri je bilo tudi precej peščenjaka in nekaj večjih apnenčevih kamnov (sl. 2.35–2.36, 2.39, 2.42–2.44; glej tudi Tonovcov grad. Najdbe, pogl. 6.5). Na večini vzhodnega dela izkopnega polja na območju stavbe 3 je geološko osnovo SE 183 in delno tudi prazgodovinsko plast SE 180 prekrivala do 30 cm debela peščena plast (SE 176), ki je verjetno naravnega (erozijskega) nastanka (sl. 2.35–2.38). Na severozahodnem delu izkopnega polja je nad SE 180 ležala plast sive prsti SE 179 (sl. 2.37), ki je vsebovala skromne antične najdbe. V njej je bil viden vkop nepravilne oblike (SE 177) z bogatim kulturnim gradivom. Zapolnjen je bil s črno prstjo s precej oglja, ostanki keramike in kovine. Na vzhodnem delu izkopnega polja je ležala debela plast kulturno sterilnega rumenega peska (SE 176). SE 179 in vkop SE 177 je prekrivala plast rjave gline SE 173 (sl. 2.38). Na meji med SE 179 in SE 176 je bilo dokumentirano dno jame dokaj pravilne oblike (SE 175), ki je segala še v SE 173 (sl. 2.37, 2.38). V vogalu med zidovoma SE 121 in SE 126 je nad SE 180 ležala tanka (5–10 cm) plast oranžne gline SE 178 (sl. 2.37), ki je segala pod zid SE 121 (sl. 2.35), vendar je še bolj proti jugu, na območju stavbe 2, nismo več zasledili. Jama SE 175 je bila vkopana v SE 173, ki je obsegala severni in osrednji del izkopnega polja (sl. 2.38–2.39). V vogalu med zidovoma SE 121 in SE 126 je nad SE 178 ležala siva kulturna plast (SE 172), ki je segala pod zidom SE 121 proti jugu na območje stavbe 2 (sl. 2.35, 2.38). Antique finds, covered SU 180. An irregularly shaped pit (SU 177) was dug into this layer. A strong cultural presence was discovered in the pit that was filled with black soil which contained large quantities of charcoal, pottery and metal finds. The east part of the excavation area was covered by a thick layer of culturally sterile yellow sand (SU 176). SU 179 and pit SU 177 were covered by a layer of brown clay SU 173 (Fig. 2.38). The bottom of a pit with a relatively regular shape (SU 175) was documented on the border between SU 179 and SU 176, it also reached into SU 173 (Figs. 2.37, 2.38). A thin (5-10 cm) layer of orange clay SU 178 (Fig. 2.37) lay in the corner between walls SU 121 and SU 126, above SU 180. The layer reached under wall SU 121 (Fig. 2.35) however all of its traces were lost further to the south, in the area of the later building 2. Pit SU 175 was dug into SU 173, which covered the north and central part of the excavation area (Figs. 2.38, 2.39). In the corner between walls SU 121 and SU 126 and above SU 178 lay a grey cultural layer (SU 172) that run under wall SU 121 towards the south to the area covered by the later building 2 (Figs. 2.35, 2.38). On the far north edge of the excavation area the section revealed a group of stones (SU 171) positioned in relatively regular rows that were placed on top of SU 173 (Fig. 2.38). Brown soil, mixed with large stones (SU 174; Fig. 2.38), most likely represents the ruin of SU 171. A thin layer of soil mixed with charcoal was discovered (SU 170; Figs. 2.38-2.39) at the far north edge of the excavation area, just under the rock wall but above SU 173 and 174. SU 170 was covered by an approximately 30 cm thick layer of culturally sterile brown sandy soil (SU 168, Fig. 2.39). The northwest corner of the excavation area was covered by the gravel layer SU 169 (Fig. 2.39), which most likely eroded from the western rocky slope. 99 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 Sl. 2.36: Stavba 3, planum 6. M. = 1:50. Fig. 2.36: Building 3, planum 6. Scale = 1: 50. Na skrajnem severnem robu izkopnega polja je bila v profilu vidna skupina dokaj pravilno zloženih kamnov (SE 171), ki so bili postavljeni na SE 173 (sl. 2.38). Rjava prst, pomešana z velikimi kamni (SE 174; 100 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 Sl. 2.37: Stavba 3, planum 5. M. = 1:50. Fig. 2.37: Building 3, planum 5. Scale = 1:50. sl. 2.38, 2.39), je verjetno ruševina SE 171. Nad SE 173 in 174 je ob skrajnem severnem robu izkopa, tik pod skalno steno, ležala še tanka plast z ogljem pomešane prsti (SE 170; sl. 2.38, 2.39). SE 170 je prekrivala okrog 101 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 Sl. 2.38: Stavba 3, planum 4. M. =1:50. Fig. 2.38: Building 3, planum 4. Scale = 1:50. 102 2.4 Stavbi 2 in 3 2.4 Buildings 2 and 3 Sl. 2.39: Stavba 3, presek 3. M. = 1:50. Fig. 2.39: Building 3, section 3. Scale = 1:50. 2.4.2 Building 3 30 cm debela plast kulturno sterilne peščene prsti (SE 168; sl. 2.39). Severozahodni vogal izkopnega polja je prekrivala plast grušča SE 169 (sl. 2.39), verjetno posledica erozije materiala z zahodnega skalnega pobočja. The wall SU 121 was the only wall belonging to building 3 preserved in its entire length (SU 121; Fig. 3.22). It was 6.3 m long, 0.6 m wide and approximately 0.8 m high, constructed from two rows of lime quarry stones joined by mortar. It was placed on SU 172 (Fig. 2.35), with no additional foundations. Wall SU 126, which represented the west wall of building 3, is attached to it. Approximately 4.5 m in length and 0.5 m in width is preserved. For the most part of the wall (except at the connection with the south wall) only the lower row of stones was preserved. This wall was also built from two rows of lime quarry stones and mortar, and it was erected on SU 172. The east wall of building 3 (SU 127) was preserved approximately 1 metre in length and a single line in height. It was placed on SU 172 right alongside the rock face. On the level of the lower line of the wall, there was an approximately 3 m long and 40 cm wide dark patch (SU 142) where it is believed that the wall continued. The narrow area between wall SU 127 and the rock face was filled with gravel (SU 157; Figs. 2.38, 2.40). The north wall of the building was not discovered. Within the three walls, the older layers were covered by a few cm thick layer of hard, compact soil, mixed with fine sand SU 160 (Figs. 2.39, 2.40). It reached its greatest depth (approx. 15 cm) at the north and south edge, and was at its thinnest in the central part, where it hardly covered the layers underneath (SU 176 and SU 173). Its upper level was roughly at the same depth as the lower level of walls SU 121, SU 126 and SU 127 (approx. 2.4.2 Stavba 3 Edini v celotni dolžini ohranjeni zid stavbe 3 (SE 121; sl. 3.22) je bil dolg 6,3 m, širok 0,6 m in ohranjen v višino približno 0,8 m. Sezidan je bil iz dveh vrst apnenčevih lomljencev, med sabo povezanih z malto. Brez posebnega temeljenja je bil postavljen na SE 172 (sl. 2.35). Nanj se navezuje zid SE 126, tj. zahodni zid stavbe 3. V dolžino je bil ohranjen pribl. 4,5 m, širok 0,5 m. Večinoma (razen ob stiku z južnim zidom) je bila ohranjena samo spodnja vrsta kamnov. Tudi ta zid je bil izdelan iz dveh vrst apnenčevih lomljencev z uporabo malte, postavljen pa na SE 172. Vzhodni zid stavbe 3 (SE 127) je bil ohranjen le pribl. 1 m v dolžino in eno vrsto v višino. Postavljen je bil na SE 172 tik ob skalno steno. Na nivoju spodnje linije zidu je bila na mestu, kjer bi se zid moral nadaljevati, v dolžini pribl. 3 m vidna pribl. 40 cm široka temna lisa (SE 142), ozek prostor med zidom SE 127 in skalno steno je bil zapolnjen z gruščem (SE 157; sl. 2.38, 2.40). Severni zid stavbe ni bil najden. V notranjosti prostora, ki so ga omejevali opisani trije zidovi, je starejše plasti prekrivala nekaj cm d

RELATED PAPERS

RELATED TOPICS

Log In

Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation

Late Antique fortified settlement Tonovcov grad near Kobarid. Settlement remains and interpretation

Related Papers

RELATED PAPERS

RELATED TOPICS