AN OFFPRINT FROM SMALL FINDS AND ANCIENT SOCIAL PRACTICES IN THE NORTH-WEST PROVINCES OF THE ROMAN EMPIRE Edited by STEFANIE HOSS & ALISSA WHITMORE Paperback Edition: ISBN 978-1-78570-256-3 Digital Edition: ISBN 978-1-78570-257-0 (epub) © Oxford & Philadelphia 2016 www.oxbowbooks.com Published in the United Kingdom in 2016 by OXBOW BOOKS 10 Hythe Bridge Street, Oxford OX1 2EW and in the United States by OXBOW BOOKS 1950 Lawrence Road, Havertown, PA 19083 © Oxbow Books and the individual authors 2016 Paperback Edition: ISBN 978-1-78570-256-3 Digital Edition: ISBN 978-1-78570-257-0 (epub) A CIP record for this book is available from the British Library All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means, electronic or mechanical including photocopying, recording or by any information storage and retrieval system, without permission from the publisher in writing. Printed in Southampton by Hobbs the Printers Ltd For a complete list of Oxbow titles, please contact: UNITED KINGDOM Oxbow Books Telephone (01865) 241249, Fax (01865) 794449 Email: oxbow@oxbowbooks.com www.oxbowbooks.com UNITED STATES OF AMERICA Oxbow Books Telephone (800) 791-9354, Fax (610) 853-9146 Email: queries@casemateacademic.com www.casemateacademic.com/oxbow Oxbow Books is part of the Casemate Group Front cover: Gladiator glass © The Vindolanda Trust; York gorgoneion (no. 7) © York Museums Trust, Yorkshire Museum; “R”-shaped brooch from recent excavations in Forum Hadriani, © Restaura.nl. Contents 1. Introduction: Small finds and ancient social practices Stefanie Hoss and Alissa Whitmore 1 PART 1: SMALL FINDS, THE BODY AND IDENTITY 2. Iron footed – hobnail patterns under Roman shoes and their functional meaning Boris Burandt 3. Wearing socks in sandals: The height of Roman fashion? Barbara Köstner 4. Laying it on thick – makeup in the Roman Empire Gisela Michel 5. Of brooches and men Stefanie Hoss 9 16 28 35 PART 2: RELIGION AND RITUAL IN THE ROMAN NORTH-WEST PROVINCES 6. Ordinary objects transformed: the compound natures of material culture Mara Vejby 7. A Mars with breasts from Weißenburg in Bavaria Nicole Birkle 8. Metropolitan styling: metal figurines from London and Colchester Emma Durham 9. Staring at death: the jet gorgoneia of Roman Britain Adam Parker 57 68 75 98 PART 3: ARTEFACTS, BEHAVIOURS AND SPACES 10. Dining with Mithras – functional aspects of pottery ensembles from Roman Mithraea Ines Klenner 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman bathhouse contexts Alissa M. Whitmore 12. The complexity of intramural and extramural relationships on the northern frontier of Roman Britain – a Vindolanda case study Andrew R. Birley 117 128 146 11 Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman bathhouse contexts Alissa M. Whitmore Keywords: Site formation; deposition; bathhouse; cloth working; military; gender; class The recovery of needles, spindle whorls, loom weights and other cloth working instruments from Roman public and military bathhouses raises the question of whether sewing, spinning and weaving – activities otherwise unattested in the baths – took place in these spaces. This paper evaluates the possibility of cloth working in Roman bathhouses by examining site formation processes, focusing on drains as the most probable context for artefacts related to the normal use of the baths. A review of the alternative functions and possible owners of textile implements discovered in bathhouse drains suggests that while these activities were uncommon, sewing and spinning occurred in a few baths and these needles, spindle whorls and related objects likely offer physical evidence for the activities of lower class bathers and bathhouse attendants. Introduction The last three decades have witnessed an increase in artefact studies in a wide range of ancient spaces. This research has greatly enriched our understanding of how people actually used built space, particularly when comparing artefacts to information provided in ancient texts. Lisa Nevett’s study (1999) of small finds in ancient Greek houses indicates that there was little gender segregation in the use or storage of men’s and women’s objects, overturning decades of scholarship which tried to locate women’s quarters in ancient Greek houses. Penelope Allison’s research (2004) on artefacts and furniture in Roman houses illustrates the multifunctional use of many spaces and the dangers of depending upon ancient texts alone. Studies of small finds have also suggested the presence of women and children in Roman forts and barracks (van Driel Murray 1994; AllasonJones 1999; Allison 2006). Caution is necessary, however, when interpreting artefacts for the presence and activities of ancient peoples, since object findspots do not always indicate places of use. Significant questions, for example, have been raised about the contexts of women’s and children’s items in barracks, since relatively few were recovered in clear stratigraphic levels linked to the occupation of barrack buildings, rather than construction deposits or later dumps (Hodgson 2014; Becker 2006). Careful attention to deposition practices is crucial when evaluating small finds and offers the potential for a better understanding of ancient spaces and behaviour. This is especially the case when evaluating unique finds in settings without previous studies of small finds assemblages, such as cloth working items in Roman bathhouses. With this in mind, I seek to critically interpret the discovery of needles, spindle whorls, weaving tablets and loom weights found in Roman public and military bathhouses. While cloth working artefacts make up a very small percentage of bathhouse finds assemblages (typically less than 3%, see Whitmore 2013), and they are absent entirely from many sites, these objects have been found in a number of baths and provide potential evidence for an activity that is otherwise unattested in these spaces. To evaluate whether these needles, spindle whorls and weaving tools provide evidence for a new activity in 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman Roman bathhouses, or if the objects are simply intrusive, this paper focuses upon the Roman construction, demolition and deposition practices and archaeological site formation processes behind the stratigraphic contexts in which these objects were found. First, however, an introduction to Roman bathing is useful. A brief overview of Roman bathing Only the wealthiest individuals could afford spaces in their homes that were dedicated to bathing, which for the Romans entailed more than simply cleaning the body. Instead, the vast majority of the population bathed at the public baths: large, multi-room structures which offered a variety of bathing options, including hot and cold pools, dry and wet sweat rooms and rooms devoted to massaging, cleaning and grooming the body. Ancient texts suggest that the majority of bathhouses were visited by people of all genders, ages and classes, and the presence of bathhouses in ancient literature, inscriptions and the archaeological record attests to their popularity in large and small towns and military communities throughout the Roman Empire (Fagan 1999; Yegül 2010). In the Northwest provinces, public baths were associated with military forts and large urban centres and are often viewed as spaces involved in the creation and maintenance of Roman identity (Nielsen 1993, 73–84; Revell 2009, 172–179). Written sources also attest to the important function of Roman baths as social centres. It was here that the Romans met with friends and lovers, caught up on the latest news and occasionally struck business deals and forged political alliances (Ovid, Ars amatoria 3.639–640; CIL IV.10677; Martial 3.36; Digest 50.1.27.1). Passages from Seneca (Epistles 56), Martial (11.82) and Juvenal (Saturae 6.419– 425) mention a number of activities which occurred in these lively spaces, including eating and drinking, exercise, massage, depilation and sex. Many rooms in public and military bathhouses lack pools or other obvious architectural markers of hygienic activities, which may underline the importance of the socialization, and spaces for it, in the baths (DeLaine 1992; Revell 2007). Artefacts recovered from bathhouse contexts (Whitmore 2013; forthcoming) provide material evidence for many of these social activities, especially grooming, eating, drinking and status displays, and also indicate others, such as gaming, medical procedures and, perhaps, cloth working. Artefact deposition and site formation processes in Roman baths Before turning to the cloth working artefacts, it is first necessary to consider how the architecture of Roman baths, ancient behaviour and site formation processes, affected the 129 ways in which objects became part of the archaeological record. Like any other site, the artefact assemblages recovered in baths provide only a partial picture of what once occurred in these spaces, and several factors assuredly limited or prevented many objects from becoming interred. The majority of bathhouse floors are flagstone or mosaic, and these solid surfaces would have made it easy for bathers to retrieve lost possessions and for attendants to clean the floors, an activity which ancient sources and modern scholars suggest occurred with some frequency (Seneca, Epistulae 86.10; Pliny, Epistulae 10.32.2; CIL II.5181; Zienkiewicz 1986a, 243). Thus, lost items and debris would not accumulate or be trod into floors and many objects used in the baths will never be recovered in these spaces. Furthermore, not all objects recovered at bathhouse sites reflect the activities of ancient bathers. The most common findspots for bathhouse artefacts are discussed below, many of which do not diverge strongly from contexts at other Roman period sites. While each bath must be evaluated individually, I offer general guidelines for interpreting these types of contexts, drawing upon scholarship on Roman construction and demolition practices, rubbish disposal and archaeological site formation processes. Topsoil, post-abandonment and demolition accumulations: the afterlife of baths Roman period artefacts are often recovered in the upper, poorly stratified layers of excavations, as at the Ribchester military baths (UK), where ancient and medieval robbing, antiquarian excavations and backfill trenches have resulted in mixed deposits (Godwin n.d.). At some sites, there is clear evidence of artefacts being introduced into abandoned baths via natural processes, as at Mirobriga (near Santiago do Cacém, P), where soil, bone, Roman ceramics and a glass bracelet have washed into the courtyard and service yard of the West Baths from a nearby hill (Biers & Biers 1988a, 73; 1988b, 172, 175; cf. Schiffer 1985, 30). While these poorly stratified finds offer a general picture of life in the neighbourhood of the baths, they cannot be used to populate the baths with activities and people, even if the debris includes bathing instruments. While the floors of buildings are often considered occupation contexts (e.g. Gardner 2007, 70), it seems unlikely that we can interpret assemblages in the majority of bathhouses in this way. If left standing, abandoned baths were often reused by squatters or became industrial workshops during Roman and later periods (e.g. Ellis 2000, 75–78; Gregory 2010, 15), and some bathhouse features and accumulations can be tied to these activities. A large quantity of rubbish, mostly cooking pots, pottery and animal bones dating to the 4th century, was found in the Vindolanda 3rd century baths (UK), and are believed 130 Alissa M. Whitmore to have been dumped there by squatters or stone robbers after the baths were abandoned (Birley 2001, 6–7). While these contexts and finds provide fascinating insight into the afterlife of bath buildings and the individuals living and working in these repurposed spaces, they do not provide any evidence relevant to the bath’s period of use. Aside from sites with rapid abandonment and interment, it is unlikely that objects used in the baths would have remained in these spaces, since salvagers or squatters presumably would have collected any valuable, functional or recyclable materials (cf. Schiffer 1985, 26–28; Zienkiewicz 1986a, 42–45; Philp 2012, 75–84; Keller 2005; Peña 2007, 250–271). Many bathhouses were demolished during the Roman period for their architectural materials or land, and these activities can clearly affect our interpretations of these spaces. Workshops were sometimes set up within buildings to recycle construction materials (cf. Munro 2010), which occurred in a late 3rd century abandonment phase at the Caerleon legionary baths (UK), when a metalworking furnace was built over the frigidarium floor drain, introducing slag and architectural materials into the drainage system (Zienkiewicz 1986a, 253–254). Robber trenches, both ancient and modern, often cut through numerous bathhouse features and layers to access stone foundations and drains, mixing different strata and artefacts (e.g. Gillam et al. 1993, 1; Barker et al. 1997, 6; Philp 2012, 75–76; Proctor 2012, 68–69). Before, during and after demolition, abandoned baths could also become local dumps (cf. Schiffer 1985, 29). The abandoned Baths of the Swimmers in Ostia (IT) were infilled with ceramics and small finds which are interpreted as neighbourhood rubbish (Carandini et al. 1968, 8), and this may also explain the medical instruments found below a layer of destruction rubble in a room off the palaestra of the Xanten baths (DE, Hoss forthcoming). During demolition, bathhouse spaces, especially hypocausts, were often filled with building rubble (e.g. Daniels 1959, 92–93; Perkins 2004, 42–43; Philp 2012, 62, 67). Interpreting these infills is particularly challenging, as they may represent rubbish generated in the baths, waste produced by construction or demolition crews, or materials which were imported from outside areas as fill. While some finds might relate to activities in the baths, they are not from primary occupation layers and could easily originate from elsewhere. This is also true of objects recovered from the spoil heaps of earlier excavations. Occupation floors in rapidly abandoned baths: the baths of Pompeii and Herculaneum While most artefacts recovered directly on bathhouse floors are more likely related to post-abandonment activities, a possible counter-example are baths which were rapidly abandoned and buried, since these sites may have preserved occupation floors and avoided the squatting or demolition activities which introduce outside artefacts. The best known rapidly abandoned baths are in Pompeii and Herculaneum (IT), but as scholars of Campanian finds assemblages have warned (Allison 1992; Wallace-Hadrill 2011, 272–280), the notion of in situ artefacts at these sites can be misleading. The baths at Pompeii and Herculaneum are greatly affected by their early excavation dates. Data on excavated artefacts must be recovered from early published sources (e.g. Fiorelli 1862) and unpublished inventories (Libretti d’Inventario) and excavation books (Giornali degli Scavi), and oftentimes, information in these sources leaves much to be desired. Descriptions of artefact contexts are woefully uneven, with findspots ranging from very general (‘the baths’) to more specific (‘between the columns on the north side of the peristyle’). Rarely is the vertical position of artefacts from Pompeian bathhouses described, but on occasion, finds were noted as recovered from a meter or more above the pavement, presumably in ash and rubble layers. Whether these objects fell from shelves or terraces when roofs collapsed, or were swept into rooms from elsewhere on ash, pumice or lava flows, is unknown, but this does call into question where precisely many of these objects originated (cf. Allison 1992, 50). A particularly significant point when interpreting artefact assemblages in Pompeii’s baths, however, is that few of these buildings were open at the time of the eruption. Possible interruptions to the water supply and evidence for incomplete repairs following the earthquake of AD 62 suggest that part of Pompeii’s Forum Baths (Eschebach 1982, 319; Koloski-Ostrow 2008, 231) and potentially all of the Stabian Baths (Maiuri 1931, 574; Eschebach 1979, 70) were closed. The Sarno Baths (Koloski-Ostrow 2008, 240) and Baths of Julia Felix (Parslow 2000) may also have been under renovation, the Central Baths were not yet open (Koloski-Ostrow 2008, 224) and the Republican Baths had been demolished decades prior to the eruption (Maiuri 1950, 113–116). Recovering a true occupation floor that relates to the use of Roman bathhouses is thus difficult, if not impossible. While the Stabian Baths have a particularly rich artefact assemblage, the fact that these baths were closed and still under reconstruction at the time of the eruption means that these objects might have arrived into these spaces through any number of ways. The artefacts might relate to the normal use of the baths, and were stored here awaiting their reopening, but they could easily be associated with squatters, an altered use of the structure, or construction fill or workers. At Pompeii and Herculaneum, we must be more, rather than less, observant of context and potential disruptions to the dayto-day functioning of these spaces. 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman Construction deposits: levelling layers and foundation trenches Many bathhouses were in use for several centuries, and the long lives of these buildings resulted in numerous construction phases and deposits (Zienkiewicz 1986a, 46–50; Nielsen 1993, 82; Rook 1992, 6; Fagan 1999, 180). In the later centuries of the High Empire, pools in the North-west Provinces were regularly closed and infilled with rubbish while the rest of the baths remained in use (e.g. Zienkiewicz 1986a, 253–255; Ellis 2000, 68). Waste was also used to fill up no longer used hypocausts while the rest of the bath was still in use. Many of these deposits include artefacts. Ceramic sherds and broken tile are common ingredients in Roman levelling layers and floor make-up (Peña 2007, 250; Vitruvius, De Architectura 7.1.1–5), and construction fills from the Baths of Caracalla (IT) contained numerous ceramic fragments, including amphorae, terra sigillata and coarse wares, mixed with pozzolana, clay, mortar and fragments of brick and marble (DeLaine 1997, 138). Glass sherds, coins and small finds are also occasionally recovered in these fill layers. Beads, fibulae, pins and toilet instruments were recovered in and around the foundation trenches of the Caerleon Castle baths (Lee 1850, 17) and intaglios, beads, gaming counters and a finger ring have been recovered from various make-up layers at the Caerleon Legionary baths (Zienkiewicz 1986b, 133, 142, 149, 156). Objects recovered from these layers often provide a date for the construction or renovation of baths, but since many bathhouses have multiple construction phases, it is tempting to interpret finds in these contexts as potentially providing hints about life during the earlier periods of the baths. We must, however, be cautious when interpreting the assemblages from such contexts. Kevin Dicus has presented a compelling argument for the potential origins of artefacts found in Pompeian construction trenches and layers. Large-scale construction projects, such as the creation of levelling layers or infilling pools, would have required large quantities of sediments and materials, likely originating from outside the construction area. At Pompeii, the probable source for such fill appears to be rubbish dumps outside town walls, which included construction and demolition materials and refuse originating from numerous sources dating to various occupation periods. As a result, artefacts found in levelling layers or infilled features likely came from the community’s garbage dumps, rather than within the construction area itself (Dicus 2014, 70–76; cf. Liebeschuetz 2000, 51–54). Undoubtedly, similar construction practices occurred outside of Pompeii as well, and this suggests that finds from the vast majority of bathhouse construction layers cannot provide evidence for activities taking place during the normal use of these spaces, even if these objects fit our conception of bathhouse material culture. Since strigils and bath flasks, the quintessential 131 bathing artefacts, are frequently recovered in settings other than baths (Wardle 2008, 207–211; Whitmore forthcoming), presumably because people carried them to the baths from home (Shelton 1981, 26–28; Carandini et al. 1982, 334; Dickey 2012, 121–125, 201; Juvenal, Saturae 6.419–420), any notion of ‘exclusive’ bathhouse material culture seems overly simplistic. There are some scenarios, however, which might allow for materials from inside or around the baths to be interred in construction layers during small-scale projects. The Romans may have set aside soil removed when digging foundation or water pipe trenches for later backfilling. Any finds in such soil, however, are still unlikely to have originated from occupation layers, and subsequently would be mixed with objects from other strata (Dicus 2014, 72–76). Two levelling layers and a foundation trench at the Caerleon Legionary Baths might provide a possible exception (Zienkiewicz 1986a, 244–249). In this case, similarities in sediment colour and consistency, as well as finds assemblages, suggest that the materials in these construction deposits originated from a cleaning of the drainage system, a context which Zienkiewicz interprets as related to the daily operation of the baths. Floor and pool drains: traps for accidental losses and small rubbish disposal Although not every bathhouse drain contains an artefact assemblage, material culture, occasionally in very large quantities, has been recovered in the sediments of some bathhouse drains. Objects could enter drainage systems through bathhouse pools or floor drains, the latter of which were often covered with slotted stone grates to allow water (and small objects) to enter (cf. Zienkiewicz 1986a, 35–36, 196–199). At least some of these drain artefacts likely represent items lost by bathers or small rubbish swept toward floor drains by bath attendants, thus this context may represent the closest thing to an occupation or primary rubbish layer in the baths (cf. Schiffer 1985, 24–25). When interpreting drainage contexts, it is critical to isolate the strata and objects that represent the normal functioning of the baths. Material culture related to construction and post-abandonment activities can enter drainage systems, and a mass of lead, iron nails, glass and coins attributed to renovation activities was found under the Phase IV frigidarium floor drain in the Silchester Public Baths (UK; Hope & Fox 1905, 351–352). The infilling of pools surely could have introduced some material culture into adjacent drain segments, and rubbish and debris were dumped into sections of the Caerleon drain on at least two occasions after the closure of the baths (Zienkiewicz 1986a, 249, 253–255). The robbing of drains can also result in the removal of sediments and the introduction of artefacts, which occurred in the Dover Shore Fort Baths (UK), where 4th century coins 132 Alissa M. Whitmore are believed to have entered sediments when the drains were robbed in antiquity (Philp 2012, 76). At sites where detailed stratigraphic excavation of drain sediments occurred, as at Caerleon, it is possible to isolate drain strata associated with construction, dumping and robbing, and eliminate objects that are likely unrelated to bathers. Using the Caerleon drain stratigraphy as a model, the sediments and materials most likely associated with the normal use of the baths are the horizontal layers found throughout the length of the drain, often consisting of fine, silty sand that was presumably gradually laid down while the baths were in operation (Zienkiewicz 1986b, 13). Since waterborne materials could also flow through the drain after the closure of the baths, objects found in lower drain strata are more likely tied to the use of the structure, though distinctions between drain layers cannot always be made, and it is always possible that heavier items could sink through the fine, viscous drain sediments (Zienkiewicz 1986a, 244–249). Detailed stratigraphic drain excavation and documentation, however, is absent from many sites, and it is often possible to only associate artefacts with specific segments of the drain, or merely with the drain itself. In such cases, especially when there is evidence that demolition or salvage work may have affected sediments, drain assemblages can only be viewed as possible evidence for bathhouse activities. Latrines, pits and furnaces: bathhouse rubbish deposits? No ancient texts reveal what happened to the broken ceramics, bath flasks, toilet instruments and animal bones that were inevitably produced during a typical day in the baths. Ancient sources and scholarship on Greco-Roman rubbish practices, however, can provide some suggestions. Salvagers collected and purchased recyclable broken metals, glass and ceramics (Keller 2005; Peña 2007, 253; Martial 1.41, 10.3; Statius, Silvae 1.6.73–74; Juvenal, Saturae 5.48), and surely these individuals would have regularly stopped at the baths. Other sources indicate that carts may have removed various types of rubbish from town (Thüry 2001, 7–8; Tactius Annales 11.32; Cicero De Divinatione 1.57; CIL I2 593). Studies of artefact assemblages and refuse disposal at South Shields and York, in contrast, suggest that there may have been relatively little movement of objects from their original places of use, perhaps indicating that materials from ditches and dumps may illustrate activities occurring in spaces nearby (Gardner 2007, 85–87). Contexts in or around bathhouses may have functioned as disposal sites for garbage produced in the baths. Artefacts are often recovered in bathhouse latrines, which were frequently located at the end of a bath’s drainage system. These objects may have entered the latrine from bathing spaces further upstream in the drainage system or were lost or thrown in as rubbish via the seat openings (Van Vaerenbergh 2011a). While serving as a rubbish dump wasn’t the primary function of latrines, excavations of Campanian latrines and sewers have illustrated that the Romans threw kitchen waste and broken items into latrines and presumably lost intact objects down household and commercial toilets (Jansen 2000, 38; Wallace-Hadrill 2011, 282–285; Camardo 2011). Public latrines could not have been permanent or official garbage dumps, however, as large objects would remain and eventually compromise the latrine’s functionality, requiring that it be cleaned out with greater regularity. Instead, it is more probable that the majority of rubbish was inserted into latrines after they fell out of use, and only the lowest latrine layers might contain objects that entered through bathhouse drains or were lost down a still functioning toilet (Daniels 1959, 145–157). In addition, bathers were not the only individuals using bathhouse latrines. Many were located near entrances or courtyard areas, somewhat separate from the main bathing block, suggesting that latrines could be accessed without entering bathing spaces that required an entrance fee (Van Vaerenbergh 2011b; Koloski-Ostrow 2015, 11, 21). Since anyone potentially had access to these public bathhouse latrines, material culture from this context might better exemplify items that the Romans lost or threw down the toilet, rather than activities taking place in baths. Coins, ceramics and occasionally small finds have been recovered in furnaces and hypocausts, begging the question of whether bathhouse rubbish was thrown into furnaces. Since hypocausts were often infilled with rubbish during demolition or renovations, and sometimes flooding introduced sediments and materials into hypocaust systems (Philp 2012, 67), careful attention must be paid to stratigraphy in hypocaust layers. Ceramics and coins have been recovered in ash layers thought to reflect the use periods of hypocausts and stokeholes at the Bewcastle (UK) and Dover Shore Fort baths (Gillam et al. 1993, 41; Philp 2012, 59). In the Late Roman baths at Corinth (GR), fragments of glass cups, a broken bronze finger ring, a marble mortarium, ceramics and a coin were recovered in ash attributed to the use of the tepidarium’s hypocaust, though these layers also contained building materials (Sanders 1999, 460–461), suggesting perhaps that there is some contamination from the demolition layer above. While the vast majority of hypocaust and stokehole finds can likely be attributed to infill or destruction layers, it remains to be explained how objects might enter a stokehole or hypocaust and become part of an ash layer during normal use. Broken ceramics and other finds certainly wouldn’t be suitable as fuel, but perhaps they entered stokeholes with the wood, chaff and/or charcoal that was used to heat hypocausts (Nielsen 1993, 19–20; McParland et al. 2009). The long term disposal of solid debris, such as ceramic sherds or building materials, in hypocausts would not be practical, since they 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman would simply accumulate, requiring that the hypocaust be swept out more often and the rubbish disposed of elsewhere. The majority of ash layers, in fact, contain little material culture, suggesting that hypocausts were not regularly used for the burning or disposal of rubbish. It is possible that hypocausts simply may have served as a convenient garbage for bath attendants on occasion, though it seems unlikely that these objects necessarily represent the possessions or on site eating and drinking activities of these attendants, as the heat of the hypocaust furnace would not encourage workers to linger unnecessarily. Rubbish produced in the baths may also have been disposed of in pits nearby. At Wroxeter (UK), excavations in the portico, courtyard and precinct near the baths recovered several pits containing jewellery, bath flasks, broken glass and ceramics and other small finds which may represent garbage from the baths. It is equally plausible, however, that these pits contain rubbish from the nearby market, houses or other buildings (Cool 2000, 162, 185; Mould 2000, 138), and it is impossible to discern to which structures this garbage belongs. While pits may have provided a means for bathhouse rubbish disposal, the urban setting of most baths makes it difficult to associate pits and rubbish exclusively with bathhouses. Cloth working items in Roman baths The artefact data in this paper come from a larger study of published and unpublished small finds assemblages from Roman public and military bathhouses, the aim of which is to reconstruct social life in the baths using these objects (Whitmore 2013; forthcoming). As discussed above, drains seem to be the least problematic context for objects which were actually used by bathers, since the artefacts found in drain layers definitely entered from within the baths, likely during the normal functioning of the building, and this context is less affected by construction or demolition activities than any other bathhouse space. Thus, I have concentrated much of my research on sites with excavated drainage systems. This method of site selection assuredly has some bias since only a small fraction of bathhouses have excavated drains with finds assemblages. Large imperial thermae, for instance, are entirely absent from this study, since few have excavated drains and most are published without finds catalogues (cf. Hoss forthcoming). As a result, the artefacts in my study are best suited to illustrate social life in specific bathhouses, regions and time periods, and only tentatively can these data be extrapolated elsewhere. The majority of cloth working tools in Roman baths were recovered from drains, destruction or demolition layers and unstratified or unspecified contexts (Table 11.1). Needles are by far the most commonly recovered cloth working artefact, followed by loom weights and spindle whorls. The majority of the objects from demolition contexts were 133 recovered from generalized destruction layers or infilled rooms and pools. Most of the unstratified or unspecified finds can only be traced to certain rooms or spaces. In Pompeii’s Stabian Baths, for instance, spindle whorls, awls and a sacking needle were found in the palaestra and nearby rooms, likely those containing hip baths (Fiorelli 1862, 638, 646–647, 650; Libretti d’Inventario). These finds, however, lack a vertical context and the stratigraphic details necessary to discern how they might have entered the archaeological record. The vast majority of needles, spindle whorls and loom weights in Table 11.1 are not from contexts that can be interpreted as bathhouse occupation layers. A few needles and a loom weight were recovered in the Mirobriga and Vindolanda bathhouse latrines, but it is impossible to determine whether these items were used and lost in the baths. While the description of the silty fill in Mirobriga’s latrine seems comparable to other bathhouse drain sediments which reflect normal use, the latrine fill was not excavated stratigraphically (Biers and Biers 1988b, 179–180), making it difficult to distinguish which materials may be from the more promising lower layers. Details on the Vindolanda latrine drain’s stratigraphy have not yet been published. Both of these bathhouse latrines, like many others, are also located at the end of their respective drainage systems (Figs. 11.1 & 11.2). One of the main entrances to the Mirobriga East Baths is located quite close to the latrine, and the latrine was entered through the courtyard, rather than directly from a bathing room. The Vindolanda latrine is detached from the main bathing block and robbing has obscured the location of its entrance. While it is probable that this latrine could have been entered from within the changing room of the baths (Birley 2001, 7), it is also likely that it could have been entered from the street. As a result, the objects recovered in these latrines could have plausibly been introduced by passersby rather than bathers, and with the currently available data, it is impossible to determine which latrine finds, if any, may have originated from the baths. The only probable bathhouse occupation context that produced cloth working tools is the drains. In looking closer at these drainage contexts, however, some finds cannot be definitively tied to the use of baths. The Cataractonium (Catterick, UK) needle was found in the rubble fill of a drain dating to the last period of the baths (Wilson 2002, pt 1, 68; pt 2, 182), making it a likely demolition context, and four of the bone needles from Caerleon were recovered directly under or near the open frigidarium floor drain, a section which also may have been impacted by demolition activities (Zienkiewicz 1986b, 173, 201). While the York Sewer finds (UK) are commonly identified as a bath assemblage (MacGregor 1976; Cool 2006, 207–210), no sewer segments have been directly connected with bathhouse spaces. Furthermore, while the sewer may have drained a bath, it likely also served other nearby buildings (Whitwell 1976, 134 Table 11.1: Cloth working artefacts in Roman bathhouses, by context. Unstratified / Uncertain Site Caerleon Legionary 1st – 3rd c AD Caerleon, UK Finds Needles Weaving Tablet Spindle Whorl Stabian Baths 2nd c BC – 1st c AD Pompei, Italy Needle Spindle Whorls Loom Weights Awls ‘Women’s Tool’ 1 4 Frauenthermen 1st – 4th c AD Augst, Switzerland York Church Street 2nd – 4th c AD York, UK Baths of the Swimmers 1st – 3rd c AD Ostia, Italy Needles 5 Cataractonium Baths Late 1st – 4th c AD Catterick, UK Needles Spindle Whorls Awls Republican Baths 2nd – 1st c BC Pompei, Italy Vindolanda 3rd – 4th c AD Bardon Mill, UK Wroxeter 1st – 4th c AD Shrewsbury, UK Loom Weights Contexts Near Baths Topsoil / Post-Baths Destruction / Demolition 2 Construction Pits in Courtyards Latrine Drains 13 2 1 Total 18 4 4 ‘many’ 3 6 3 17+ 1 1 16 Needles 12 3 1 2 2 1 1 10 1 5 2 1 1 10 7 7 Needles Spindle Whorls 1 1 2 1 5 Needles Weaving Tablet Spindle Whorl 2 1 1 1 5 Mirobriga East 2nd – 3rd c AD Santiago do Cacém, Portugal Needles Loom Weights 1 2 1 4 (Continued) Alissa M. Whitmore 12 Needles Weaving Tablet Shuttle Table 11.1: Cloth working artefacts in Roman bathhouses, by context. (Continued) Total Finds Needle Unstratified / Uncertain Contexts Near Baths Topsoil / Post-Baths 1 Destruction / Demolition Construction Pits in Courtyards Latrine Drains Total 1 Needles 2 2 Needles 1 1 2 Spindle Whorls 1 1 2 Loom Weight 1 1 Needle 1 1 Needle 1 1 Needle 1 1 Needle 1 Needles Weaving Tablets Spindle Whorls Loom Weights Shuttle Other 11 5 3 19 4 1 3 8 5 7 5 2 11 2 12 4 24 11 1 2 2+ 1 15+ 2 3 1 28 2 1 1 4 32 1 71 4 16 15+ 1 9 116 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman Site Mirobriga West 2nd – 3rd ? c AD Santiago do Cacém, Portugal Caerleon Castle ? Caerleon, UK Ribchester 2nd – 3rd c AD Ribchester, UK Dover Shore Fort 2nd – 4th c AD Dover, UK Herculaneum Forum 1st c BC – 1st c AD Ercolano, Italy Carsulae Baths 3rd c BC – 4th c AD San Gemini, Italy Abbey Villa Baths 1st – 3rd c AD Minster-in-Thanet, UK Cramond Baths 2nd – 3rd c AD Cramond, UK Bewcastle Baths 2nd – 4th c AD Bewcastle, UK Subtotal 135 136 Alissa M. Whitmore Fig. 11.1: Mirobriga east baths. Note location of latrine (La) at bottom center. A: apodyterium, F: frigidarium, C: caldarium, T: tepidarium, S: service area/praefurnium, Co: courtyard, e: entrance (after Biers 1988, fig. 75. image: William R. Biers and Jane C. Biers) Fig. 11.2: Vindolanda 3rd–4th century baths. Note location of latrine (La) at top left. A: apodyterium, F: frigidarium, C: caldarium, T: tepidarium; S: service area/praefurnium, L: laconicum (sweat room), Po: porch (image: Andrew Birley) 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman 24, 32). As a result, we cannot be certain that these finds only represent activities in the baths. Although the York sewer artefacts resemble expected finds from bathhouse drains (MacGregor 1976, 18; Cool and Baxter 2002), it is possible – given the rarity of well published Roman sewer assemblages – that Roman drains, whether connected with baths or not, simply have similar types of material culture. After eliminating the problematic drain artefacts, ten needles, two weaving tablets, one spindle whorl and one loom weight remain from drainage contexts in three baths, with the vast majority stemming from the Caerleon legionary bathhouse (Table 11.2 and Figs 11.3–11.5). The majority of these finds and contexts are strongly associated with the normal functioning of the baths. Cloth working artefacts were spread throughout an 18-metre stretch of Caerleon’s main drain (from contexts N-1 to 4), with an additional weaving tablet from Inlet 3, a separate but adjacent section of the drainage system (Figs 11.6 & 11.7) and several objects (Caerleon Needles #13 & 36, Spindle Whorl #40 and possibly Ribchester Needle #445) were recovered from the lower layers of their respective drains. While these 14 artefacts represent the best evidence for cloth working in Roman public baths, this is only a small fraction – 12% – of the 116 bathhouse cloth working finds from Table 1. The vast majority of needles, spindle whorls and loom weights come from strata that cannot be strongly related to activities taking place in a functioning bathhouse. Furthermore, only 3 of the 19 baths in this sample have cloth working artefacts from use contexts and they typically make up a very minor percentage – 1–2% – of the total small finds drain assemblage (Whitmore 2013). The relatively low occurrence of these artefacts suggests that cloth working activities in bathhouse spaces were even less common than first appeared. Interpreting cloth working tools from bathhouse drains Since artefact evidence for cloth working is so rare in the baths, and this activity is otherwise unattested in bathing spaces, it is worth considering whether these objects were actually used for sewing, spinning and weaving. The most problematic cloth working object found in a bathhouse drain is the very fragmentary loom weight recovered from the Herculaneum Forum Baths. While needles and spindle whorls are tools for small scale, portable cloth working activities, loom weights require a large, stationary loom. While it is hard to imagine this activity in the baths, loom weaving is more efficient in humid environments (Kemp & Vogelsang-Eastwood 2001, 327). It is possible that the moist, warm bathhouse air may have facilitated this activity, which could have been performed by bath attendants in side rooms or after hours. The context of the Herculaneum loom weight sheds little light on its use, as 137 it was uncovered not during a stratigraphic excavation, but a cleaning of the drain, whose entry points are uncertain but likely include the palaestra, men’s apodyterium and the latrine (Pagano 1999, 181). More problematic than its context, however, is that only one loom weight was found. The Roman warp-weighted loom most likely had two separate rows of loom weights, with a number of threads tied to each weight (Wild 1973, 62–64). The total number of weights used on a loom at a given time would depend upon the diameter of the yarn, the weight and thickness of the loom weights, and width and desired thread count of the fabric being woven (Mårtensson et al. 2009). Archaeological finds of loom weights in rows, as well as hoards of weights presumably in storage, suggest that all but the narrowest fabrics would require anywhere from 30 to 70 weights (Hoffmann 1964, 311–315; WallaceHadrill 1996, 112; Allison 1999, 70). In an experiment using loom weights based upon Bronze Age examples, one weaver used as many as 100 light, narrow loom weights (each weighing 177 g and 2 cm thick) to successfully produce a 1 sq. m fabric (Mårtensson et al. 2009, 394–396). It seems probable that isolated finds of loom weights, such as this one from the Herculaneum Forum Baths, might reflect chance loss or use as simple weights, rather than on-site weaving (Wild 1970, 63 n.1). Weights may have been used in the sale of foods, drinks and perfumes in the baths, though admittedly selling these products in containers would be easier. Heavier loom weights may have been used to automatically close bathhouse doors, a practice known from the more recent 19th century (T. Fischer 2015, pers. comm.). While loom weights could have been used for other functions at any time, the fact that warp-weighted looms were beginning to wane in popularity when all these baths were in operation (Seneca, Epistulae 90.20; Wild 1970, 67), might make it even more probable that loom weights had additional functions. The weaving tablets are also challenging to interpret. Only two broken weaving tablets were recovered at Caerleon, and while this small number is characteristic for the recovery of these objects at other types of sites, additional tablets, perhaps as many as 16–150 more, would have been required to weave a border (Wild 1970, 73; Knudsen 2010, 153–156). While tablet weaving can be more mobile than loom weaving, it still requires a fixed warp of some kind to stabilize the threads, and smaller tablets, such as those from Caerleon, may have been used with a warp-weighted loom, rather than individually (Knudsen 2010, 153). Regardless, while in use, the tablets would have been attached to threads, likely making them more difficult to lose. The fact that the broken Caerleon tablets were found in a drainage section that only received materials from pool and floor drains may suggest that these objects were damaged in the baths, perhaps while someone was weaving a band or border. It is equally plausible, 138 Alissa M. Whitmore Table 11.2: Cloth working artefacts from probable use contexts in Roman bathhouses. Description Context Date Source Caerleon Legionary Baths Needles Broken bronze needle L 62 mm; T 1–1.5 mm Bronze Obj. #13 Lowest layer of early drain silts between bath building and drain curve (1302:C) AD 75–90 Zienkiewicz 1986b, 173 Bronze shaft of possible sacking needle L 94 mm; T 0.5–2 mm Bronze Obj. #116 Upper drain silts between bath building and drain curve (DG 4 [-5]) AD 160–230 Zienkiewicz 1986b, 181 Complete bone needle, rectangular eye L 112 mm; T 1–4.5 mm Bone Needle #32 Upper drain silts under frigidarium floor between inlets from the north central pool and the NW pool (DG 4 [4]) AD 160–230 Zienkiewicz 1986b, 201 Head of bone needle, rectangular eye L 22 mm; T 4 mm Bone Needle #35 Upper drain silts between bath building and drain curve (DG 4 [-4]) AD 160–230 Zienkiewicz 1986b, 201 Head of bone needle, rectangular eye L 17 mm; T 4 mm Bone Needle #36 Middle layer of upper drain silts, just outside bath building (DG 4 [1] / 936:1) AD 160–230 Zienkiewicz 1986b, 201 Complete bone needle with figure 8 eye L 68 mm; T 1–4 mm Bone Needle #39 Upper drain silts between bath building and drain curve (DG 4 [-4]) AD 160–230 Zienkiewicz 1986b, 201 Complete bone needle with figure 8 eye L 69 mm; T 0.5–4 mm Bone Needle #40 Upper drain silts between bath building and drain curve (DG 4 [N-1]) AD 160–230 Zienkiewicz 1986b, 201 Broken bone needle with figure 8 eye L 59 mm; T 4.5–5 mm Bone Needle #41 Upper drain silts between bath building and drain curve (DG 4 [N-8]) AD 160–230 Zienkiewicz 1986b, 201 Head of needle with prob. figure 8 eye L 19 mm; T 4.5 – 5 mm Bone Needle #42 Upper drain silts, near drain curve (DG 4 [N-5]) AD 160–230 Zienkiewicz 1986b, 201 Broken triangular bone weaving plate T 1 mm Bone Obj. #12 Fill of Drain Inlet 3 from NE Apodyterium or Basilica (1182) AD 160–230 Zienkiewicz 1986b, 207 Near complete triangular bone weaving plate. L 39 mm; T 1.5 mm Bone Obj. #13 Upper drain silts between bath building and drain curve (DG 4 [N-7]) AD 160–230 Zienkiewicz 1986b, 207 Lower layer of upper drain silts, near drain’s exit from bath building (DG 4 [2] / 939:2) AD 160–230 Zienkiewicz 1986b, 214–215 Clay in unrobbed portion of Drain 157 (687) 2nd–3rd century AD Ribchester Inventory Drain fill near entrance to Men’s Baths 1st century BC–1st century AD Herculaneum Depot Weaving plates Spindle whorl Stone spindle whorl D 34 mm; T 5 mm; Perf. D 6.8 mm Stone Obj. #40 Ribchester Military Baths Needle Broken bone needle Inv. #445 Herculaneum Forum Baths Loom Weight Broken loom weight with suspension hole L 40 mm; W 25 mm; T 20 mm; Hole D 5 mm n.b. For the Caerleon finds, the bracketed number indicates which main drain segment the object was found in (see Fig. 11.7) 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman Fig. 11.3: Bone needles from Caerleon frigidarium drain (© National Museum of Wales. Photo: A. Whitmore) Fig. 11.4: Bone weaving tablets from Caerleon drains (© National Museum of Wales. Photo: A. Whitmore) 139 140 Alissa M. Whitmore Fig. 11.5: Stone spindle whorl from Caerleon frigidarium drain (© National Museum of Wales. Photo: A. Whitmore) Fig. 11.6: Location of Caerleon frigidarium drain. Smaller drain segments and inlets are numbered, with1 representing the position of the floor drain. 7 indicates the position of a a small inlet or manhole of uncertain purpose in the drain roof, which has been obscured by robbing. A: apodyterium, F: frigidarium, T: tepidarium, B: basilica, e: entrance (after Zienkiewicz 1986a, fig. 70, 196–199. © National Museum of Wales) 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman 141 Fig. 11.7: Segments of Caerleon frigidarium main drain . The drain segments are divided in two ways. The labels A–C divide the early drain sediments into three large, differently oriented segments. The later drain sediments (Drain Group 4 or DG 4) were excavated in 27 smaller, roughly metre-long segments, labeled n–12. (Zienkiewicz 1986b, fig. 1, 15. © National Museum of Wales) however, that these tablets could have been broken while stored in the apodyterium with their owner’s clothing, and subsequently thrown away or abandoned in the baths. The limited evidence for this activity makes tablet weaving in the baths only a possibility. Needles and spindle whorls may also have had alternative functions. Needles were used in toilet and medical activities (Juvenal, Saturae 2.93−9, Celsus, De Medicina 7.7.4, 8), though it is more probable that metal, especially iron, needles longer than 150 mm were preferred for suturing or medical activities (Jackson 1994; 2011, 250). Since the majority of needles from bathhouse drains are bone, with the exception of two broken bronze needles from Caerleon (Table 11.2), it seems improbable that most were used for medical activities. Janet Stephens has suggested that some bone needles, especially those longer than 90 mm, may have been used to create women’s hairstyles (Stephens 2008). While one needle from Caerleon fits this criterion, and could potentially have been used in hair styling (#32, 112 mm), other complete examples fall short of the required length (Caerleon #39, 68 mm; #40, 69 mm). Spindle whorls may also have functioned as items of adornment. In the Romano-British cemetery at Poundbury, shale spindle whorls were recovered near and under the heads of three female skeletons, which may suggest that they were used as hair ornaments (Farwell & Molleson 1993, 85–86, 100). However, since needles and spindle whorls would be interpreted as evidence of cloth working if found in any other context (cf. Crummy 1983, 65; Manning 2011, 85–86), seeking alternative, non-cloth working explanations for these objects may be overly conservative. These artefacts can be reasonably interpreted as evidence for sewing, spinning and perhaps tablet weaving in the Caerleon, and possibly Ribchester, military baths (Zienkiewicz 1986b, 20; Allason-Jones 2011, 240). Sewing, spinning and weaving in Roman military baths If we conclude that sewing and spinning occasionally took place in a few Roman military bathhouses, what additional information does this provide on life in the baths? One of the challenges with interpreting drain assemblages as occupation ‘floors’ is that the majority of drain contexts cannot be associated with specific bathhouse rooms, thus drain artefacts reflect only activities occurring somewhere in the baths. Robbing and modern disturbances have eliminated direct connections between drains and bathing spaces at Ribchester, so we can only speculate where materials may have entered. The Caerleon baths and drainage system, however, is better preserved. The majority of the cloth working finds – all of the needles, the spindle whorl and one of the weaving tablets – were recovered in the main frigidarium drain (Fig. 11.6). Inlets to this drain include a floor drain in the frigidarium, the cold plunges in the frigidarium and presumed floor drains in the north-east apodyterium and the basilica. A more specific point of origin can be given to a broken weaving tablet (Caerleon #12), which was found in a drain segment (Inlet 3) coming from the north-east apodyterium or basilica. While only part of the Caerleon baths have been excavated, leaving us without comparable data from heated rooms, these drain entry points permit the location of sewing, spinning and perhaps tablet weaving at Caerleon in unheated bathing spaces, some of which included pools (the frigidarium) and some without (the apodyterium and basilica). Notably, Caerleon’s basilica and frigidarium are both very large rooms, and the pools take up only a fraction of the total available space in the frigidarium, suggesting that these spaces would be well suited, and perhaps designed, to accommodate socialization and other activities, as well as bathing (DeLaine 1997; Revell 2007). 142 Alissa M. Whitmore Who would have been sewing, spinning or tablet weaving in the Caerleon and Ribchester military baths? Roman men and women were involved in sewing and weaving (Moeller 1969; Allason-Jones 1999), but spinning was more strongly associated with women alone (Allison et al. 2005). While both sites have military associations – the Caerleon legionary bathhouse was located within the walls of the fort, while the extramural Ribchester baths were situated in the surrounding settlement – the presence of women in military spaces is not unexpected (van Driel Murray 1997; AllasonJones 1999; Allison 2013), and material culture associated with Roman women, such as beads, hairpins and bracelets, have been recovered in the drains of several military bathhouses (Whitmore forthcoming). While the gender of the individuals sewing and weaving in the baths cannot be plausibly surmised, the spindle whorl from Caerleon suggests that at least one woman brought a spindle whorl into, and perhaps spun in, the baths (Allason-Jones 1989, 179). Beyond gender, however, who were these individuals? At Caerleon, where the largest and most diverse assemblage of cloth working artefacts was found, it is suggested that bathers (Allason-Jones 2011, 240), soldiers or craftspeople might have been involved with cloth working activities (Zienkiewicz 1986b, 20). Since needles were regularly part of a soldier’s equipment (van Driel-Murray 1994, 357; Allason-Jones 1999, 1), it is possible that soldiers were sewing in military baths. But would bathers – soldiers or otherwise – have brought their chores into the baths, a space predominantly associated with leisure? Lower class bathers – a designation which may fit many soldiers – might multi-task by mending in the more enjoyable atmosphere of the baths, but it seems improbable that upper class individuals were engaged in cloth working activities. It is possible that these needles, weaving tablets and spindle whorl provide evidence for the activities of slaves. In addition to slaves that were owned by the baths, domestic slaves came as bathers and attendants (Fagan 1999, 199– 206). While some slaves actively assisted their owners with bathing (Martial 7.35, 11.75; Ammianus Marcellinus 28.4.6–9), others may have been left guarding their clothes elsewhere in the baths (Fagan 1999, 200; Lucian Hippias 5), and these individuals may have been expected to complete various tasks, such as spinning wool or mending the household’s clothing. While ancient texts focus on the experiences that were offered to bathers, it is important to recall that there were labourers, slaves and free, who provided these services. In addition to those who kept the baths in operation (Nielsen 1993, 125–130), sausage makers, confectioners and other peddlers advertised food, hair pluckers removed unwanted body hair, and masseurs offered massages (Seneca, Epistulae 56.1–2). Other commercial services may also have been offered in the baths, and a darning amenity might have been especially well suited for this space, since bathers would not need their clothing during their stay. While Roman soldiers were entirely capable of mending their own clothes, some likely had the resources to take advantage of such a service. Conclusions It often takes an unusual find in an unexpected place for scholars to critically evaluate the impact of deposition practices and site formation on archaeological contexts, processes that affect every site regardless of how typical and mundane the assemblages appear. The necessity of critically appraising the impact of these forces is evidenced here: very few cloth working tools in Roman baths likely represent activities that actually occurred in these spaces. Caerleon, with its ten needles, two weaving tablets and a spindle whorl from the drain, provides the strongest case for this practice. The virtual absence of cloth working artefacts from other bathhouse drain assemblages is interesting, but difficult to interpret. The Caerleon legionary bathhouse is one of the largest baths in this sample, and has – by far – the most plentiful and diverse small finds assemblage recovered in any bath. This makes comparisons with other assemblages, especially those from smaller baths, problematic, as it is difficult to judge the significance of the absence of artefacts from an already small assemblage. Given the current evidence, we can only speculate whether cloth working was a unique occurrence only at Caerleon and Ribchester, or a more common, but invisible, activity within other bathhouses and bathing spaces. Cloth working is absent from ancient texts and artistic depictions of life in the baths. When comparing these sources with artefact assemblages, significant areas of overlap exist, particularly the presence of jewellery, status displays and eating and drinking. Other activities that are never mentioned in ancient texts appear either regularly (gaming) and occasionally (medicine) in the archaeological record of baths. Conversely, the quintessential Roman bathing instrument, the strigil, is absent from the vast majority of bathhouse artefact assemblages, most likely due to a combination of ancient behaviours and taphonomic processes (Whitmore forthcoming). If sewing and spinning did occur in some baths, and was performed by slaves, the lower classes and women, its absence from ancient texts wouldn’t be surprising, given the relatively few sources that offer detailed information on daily life in the baths (Fagan 1999, 7–8). Furthermore, the probable low status of those sewing and spinning, and the fact that these individuals – especially women – were engaged in socially acceptable pursuits, would make these individuals and activities unlikely to attract the attention of male authors. Though cloth working doesn’t neatly fit with our conception of the Roman baths as a leisure space, these sewers, spinners and tablet weavers may have been craftsmen, servants or slaves, rather than paying customers, and these finds may 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman provide material evidence of the countless, largely invisible, individuals for whom the baths were a place of work. Acknowledgements I could not have completed this research without the access granted to unpublished reports and finds at several of these sites, and I am especially grateful for the assistance of Stephen Bull (Lancashire County Museum Service), Andrew Birley and Barbara Birley (Vindolanda Trust), Tamara Tännler (Augusta Raurica Research Archives), the Soprintendenza Speciale per i Beni Archeologici di Napoli e Pompei, Jane Whitehead (Valdosta University) and Mark Lewis (National Roman Legion Museum). This research was funded in part through the Hugh Last and Donald Atkinson funds, the assistance of which I greatly appreciate, and I would also like the thank the staff at the Des Moines Area Community College and Perry Public libraries. My thoughts on bathhouse contexts have benefited greatly from discussions with the participants of the Thermae in Context Congress (Luxembourg 2013), and the current paper is much improved by comments from Stefanie Hoss, Barbara Köstner, and Molly Swetnam-Burland. All errors remain mine alone. Ancient sources Ammianus Marcellinus, History, vol. 3. Trans. J. C. Rolfe (1939). Loeb Classical Library 331. Cambridge MA, Harvard University Press. Celsus, De Medicina (On Medicine), vol. 3. Trans. W. G. Spencer (1938). Loeb Classical Library 336. Cambridge MA, Harvard University Press. Cicero, De Divinatione. In On Old Age. On Friendship. On Divination. Trans. W. A. Falconer (1923). Loeb Classical Library 154. Cambridge MA, Harvard University Press. Corpus Inscriptionum Latinarum, vol. 1 2, pars II, fasc. 1: Inscriptiones Latinae antiquissimae. Ed. A. T. Mommsen & E. Lommatzsch (1918). Berolini apud G. Reimerum. Corpus Inscriptionum Latinarum, vol. 2: Inscriptiones Hispaniae Latinae. Ed. E. W. E. Hübner (1869). Berolini apud Georgium Reimerum. Corpus Inscriptionum Latinarum, vol. 4: Inscriptiones parietariae Pompeianae Herculanenses Stabianae. Ed. C. Zangemeister and R. Schoene (1871). Berolini apud Georgium Reimerum. The Digest of Justinian, vol. 4. Trans. A. Watson (1985). Philadelphia, University of Pennsylvania Press. Juvenal, Saturae. In Juvenal and Persius. Ed. and Trans. S. M. Braund (2004). Loeb Classical Library 91. Cambridge MA, Harvard University Press. Lucan, Hippias. Trans. A. M. Harmon (1913). Loeb Classical Library 14. Cambridge MA, Harvard University Press. Martial, Epigrams, vol. 1 and 2. Ed. & Trans. D. R. Shackleton Bailey (1993). Loeb Classical Library 94 & 95. Cambridge MA, Harvard University Press. Ovid, Ars amatoria (The Art of Love). Trans. James Michie (2002). New York, Penguin Random House. 143 Pliny, Epistulae (Letters), vol. 2. Trans. Betty Radice (1969). Loeb Classical Library 59. Cambridge MA, Harvard University Press. Seneca, Epistulae (Epistles), vols 1–2. Trans. R. M. Gummere (1919/1920). Loeb Classical Library 75 & 76. Cambridge MA, Harvard University Press. Statius, Silvae. Ed. & Trans. D. R. Shackleton Bailey (2015). Loeb Classical Library 206. Cambridge MA, Harvard University Press. Tacitus, Annals, vol. 4. Trans. J. E. Jackson (1937). Loeb Classical Library 312. Cambridge MA, Harvard University Press. Vitruvius, De Architectura (Ten Books on Architecture). Trans. I. D. Rowland (1999). Cambridge MA, Cambridge University Press. Bibliography Allason-Jones, L. (1989) Women in Roman Britain. London, British Museum. Allason-Jones, L. (1999) Women and the Roman army in Britain. In A. Goldsworthy & I. Haynes (eds) The Roman Army as a Community, 41–51. Journal of Roman Archaeology Supplementary Series 34. Portsmouth RI, Journal of Roman Archaeology. Allason-Jones, L. (2011) Recreation. In L. Allason-Jones (ed.) Artefacts in Roman Britain: their Purpose and Use, 219–242. Cambridge, Cambridge University Press. Allison, P. M. (1992) Artefact assemblages: not the ‘Pompeii Premise.’ In E. Herring, R. Whitehouse & J. Wilkins (eds) Papers of the Fourth Conference of Italian Archaeology 3: New Developments in Italian Archaeology, 49–56. London, Accordia Research Centre. Allison, P. M. (1999) Labels for ladles: interpreting the material culture of Roman households. In P. M. Allison (ed.) The Archaeology of Household Activities, 57–77. New York, Routledge. Allison, P. M. (2004) Pompeian Households: an Analysis of the Material Culture. Cotsen Institute of Archaeology Monograph 42. Los Angeles, University of California. Allison, P. M. (2006) Mapping for gender: interpreting artefact distribution inside 1st and 2nd century AD forts in Roman Germany. Archaeological Dialogues 13(1), 1–20. Allison, P. M. (2013) People and Spaces in Roman Military Bases. Cambridge, Cambridge University Press. Allison, P. M., Fairbairn, A. S., Ellis, S. J. R. & Blackall, C. W. (2005). Extracting the social relevance of artefact distribution in Roman military forts. Internet Archaeology 17. http://dx.doi. org/10.11141/ia.17.4. Barker, P., White, R., Pretty, K., Bird, H. & Corbishley, M. (1997) The Baths Basilica Wroxeter. London, English Heritage. Becker, T. (2006) Women in Roman forts – lack of knowledge or a social claim? Archaeological Dialogues 13(1), 36–38. Biers, W. R. (1988) Mirobriga. British Archaeological Report S451. Oxford, British Archaeological Reports. Biers, W. R. & Biers, J. C. (1988a) The baths. In W. R. Biers (ed.) Mirobriga, 48–117. British Archaeological Report S451. Oxford, British Archaeological Reports. Biers, W. R. & Biers, J. C. (1988b) The baths, locus summaries and catalogue of finds, catalogue of architectural blocks. In W. R. 144 Alissa M. Whitmore Biers (ed.) Mirobriga, 164–205. British Archaeological Report S451. Oxford, British Archaeological Reports. Birley, A. (2001) Vindolanda’s Military Bath Houses. Northumberland, Vindolanda Trust and Roman Army Museum. Camardo, D. (2011) Case study: Ercolano: la ricostruzione dei sistemi fognari. In G. C. M. Jansen, A. O. Koloski-Ostrow & E. M. Moormann (eds) Roman Toilets: their Archaeology and Cultural History, 90–91. Leuven, Peeters. Carandini, A., Fabbricotti, E., Gasparri, C., Gasparri Tatti, M., Giannelli, M., Moriconi, M., Palma, B., Panella, C., Polia, M. & Ricci, A. (1968) Le terme del nuotatore: scavo dell’ambiente IV. Rome, De Luca Editore. Carandini, A., Ricci, A., De Vos, M. & Medri, M. (1982) Filosofiana: The Villa of Piazza Armerina. The Image of a Roman Aristocrat at the Time of Constantine. Palermo, Flaccovio. Cool, H. E. M. (2000) The Roman vessel glass. In P. Ellis (ed.) The Roman Baths and Macellum at Wroxeter, 162–185. London: English Heritage. Cool, H. E. M. (2006) Eating and Drinking in Roman Britain. Cambridge, Cambridge University Press. Cool, H. E. M. & Baxter, M. J. (2002) Exploring Romano-British finds assemblages. Oxford Journal of Archaeology 21(4), 365–380. Crummy, N. (1983) The Roman Small Finds from Excavations in Colchester 1971–9. Colchester, Colchester Archaeological Trust. Daniels, C. M. (1959) The Roman bath house at Red House, Beaufront, near Corbridge. Archaeologia Aeliana 37, 85–176. DeLaine, J. (1992) New models, old modes: continuity and change in the design of public baths. In H. J. Schalles & P. Zanker (eds) Die römische stadt im 2. Jahrhundert n. chr.: der funktionswandel des öffentlichen raumes, 257–275. Bonn, Rheinland-Verlag. DeLaine, J. (1997) The Baths of Caracalla: a Study in the Design, Construction, and Economics of Large-scale Building Projects in Imperial Rome. Journal of Roman Archaeology Supplementary Series 25. Portsmouth RI, Journal of Roman Archaeology. Dickey, E. (2012) The Colloquia of the Hermeneumata Pseudodositheana, vol. 1. Colloquia Monacensia-Einsidlensia, Leidense-Stephani, and Stephani. Cambridge, Cambridge University Press. Dicus, K. (2014) Resurrecting refuse at Pompeii: the use-value of urban refuse and its implications for interpreting archaeological assemblages. In H. Platts, J. Pearce, C. Barron, J. Lundock & J. Yoo (eds) TRAC 2013: Proceedings of the 23rd Theoretical Roman Archaeology Conference, 65–78. Oxford, Oxbow Books. Ellis, P. (2000) The Roman Baths and Macellum at Wroxeter. London, English Heritage. Eschebach, H. (1979) Die Stabianer Thermen in Pompeji. Berlin, Walter DeGruyter. Eschebach, H. (1982) La documentazione delle terme del foro a Pompei. In La Regione sotterrata dal Vesuvio: studi e prospettive, 313–328. Naples, Università degli Studi di Napoli. Fagan, G. (1999) Bathing in Public in the Roman World. Ann Arbor, University of Michigan Press. Farwell, D. E. & Molleson, T. I. (1993) Excavations at Poundbury, 1966–80, vol. 2: the cemeteries. Dorchester, Dorset Natural History and Archaeological Society. Fiorelli, I. (1862) Pompeianarum antiquitatum historia. Naples. Gardner, A. (2007) An Archaeology of Identity: Soldiers and Society in Late Roman Britain. Walnut Creek, Left Coast Press. Gillam, J. P., Jobey, I. M. & Welsby, D. A. (1993) The Roman Bath-house at Bewcastle, Cumbria. Kendal, Cumberland and Westmorland Archaeological and Antiquarian Society. Godwin, E. G. (n.d.) Unpublished preliminary report following 1979 excavation of Ribchester baths. Preston, Lancashire Museum Service. Gregory, R. (2010) Ribchester Roman Bathhouse: Desk-based research and archive assessment. Unpublished Site Assessment. Oxford Archaeology North and Lancashire Museums. Hodgson, N. (2014) The accommodation of soldiers’ wives in Roman fort barracks – on Hadrian’s Wall and beyond. In R. Collins & F. McIntosh (eds) Life in the Limes: Studies of the People and Objects of the Roman Frontiers, 18–28. Oxford, Oxbow Books. Hoffmann, M. (1964) The Warp-Weighted Loom. Norway, Universitetsforlaget. Holmes, N. (2003) Excavation of Roman Sites at Cramond, Edinburgh. Edinburgh, Society of Antiquaries of Scotland. Hope, W. H. S. J. & Fox, G. E. (1905) Excavations on the site of the Roman city at Silchester, Hants, in 1903 and 1904. Archaeologia 59, 333–370. Hoss, S. (forthcoming) Small finds in Roman thermae: an introduction. In A. Binsfeld, S. Hoss, & H. Pösche (eds) Thermae in context: Roman bathhouses in the town and in daily life. Archaeologia Mosellana 10. Luxembourg, Musée national d’histoire et d’art. Jackson, R. (1994) Medical instruments in the ‘Antiquarium’ at Pompeii. In L. J. Bliquez, Roman Surgical Instruments and Other Minor Objects in the National Archaeological Museum of Naples, 189–207. Mainz, Philipp von Zabern. Jackson, R. (2011) Medicine and hygiene. In L. Allason-Jones (ed.) Artefacts in Roman Britain: Their Purpose and Use, 243–268. Cambridge, Cambridge University Press. Jansen, G. C. M. (2000) Systems for the Disposal of Waste and Excreta in Roman Cities. The Situation in Pompeii, Herculaneum, and Ostia. In X. D. Raventós & J. A. Remolà (eds) Sordes urbis: la eliminación de residuos en la ciudad Romana, 37–49. Rome, L’Erma di Bretschneider. Keller, D. (2005) Social and economic aspects of glass recycling. In J. Bruhn, B. Croxford, & D. Grigoropoulos (eds) TRAC 2004: Proceedings of the 14th Theoretical Roman Archaeology Conference, 65–78. Oxford, Oxbow Books. Kemp, B. J. and Vogelsang-Eastwood, G. (2001) The Ancient Textile Industry at Amarna. London, Egypt Exploration Society. Koloski-Ostrow, A. O. (2008) The city baths of Pompeii and Herculaneum. In J. J. Dobbins & P. W. Foss (eds) World of Pompeii, 224–256. New York, Routledge. Koloski-Ostrow, A. O. (2015) The Archaeology of Sanitation in Roman Italy: Toilets, Sewers, and Water Systems. Chapel Hill, University of North Carolina Press. 11. Cloth working in the baths? Site formation processes, needles and spindle whorls in Roman Knudsen, L. R. (2010) Tiny weaving tablets, rectangular weaving tablets. In E. B. A. Strand, M. Gleba, U. Mannering, C. Munkholt, & M. Ringgard (eds) NESAT X: The North European Symposium for Archaeological Textiles, 150–156. Oxford, Oxbow Books. Lee, J. E. (1850) Description of a Roman Building and Other Remains Lately Discovered at Caerleon. London, J. R. Smith. Liebeschuetz, W. (2000) Rubbish disposal in Greek and Roman cities. In X. D. Raventós & J. A. Remolà (eds) Sordes urbis: la eliminación de residuos en la ciudad Romana, 51–61. Rome, L’Erma di Bretschneider. MacGregor, A. (1976) Finds from a Roman Sewer System and an Adjacent Building in Church Street. Archaeology of York 17/1. York, Council for British Archaeology. McParland, L. C., Hazell, Z., Campbell, G., Collinson, M. E. & Scott, A. C. (2009) How the Romans got themselves into hot water: temperatures and fuel types used in firing a hypocaust. Environmental Archaeology 14(2), 176–183. Maiuri, A. (1931) Pompei: pozzi e condotture d’acqua nell’antica città. Scoperta di un antico pozzo presso “Porta Vesuvio.” Notizie degli Scavi di Antichità Ser. 6, Vol. 7, 546–576. Maiuri, A. (1950) Scoperta di un edificio termale nella Regio VIII, Insula 5, nr. 36. Notizie degli Scavi di Antichità, Ser. 8, Vol. 4, 116–136. Manning, W. H. (2011) Industry. In L. Allason-Jones (ed.) Artefacts in Roman Britain: Their Purpose and Use, 68–89. Cambridge, Cambridge University Press. Mårtensson, L., Nosch, M.-L. and Andersson Strand, E. (2009) Shape of things: understanding a loom weight. Oxford Journal of Archaeology 28(4), 373–398. Moeller, W. O. (1969) The male weavers at Pompeii. Technology and Culture 10(4), 561–566. Mould, Q. (2000) Small finds from the portico pits. In P. Ellis (ed.) The Roman Baths and Macellum at Wroxeter, 137–141. London, English Heritage. Munro, B. (2010) Recycling in late Roman villas in southern Italy: reappraising hearths and kilns in final occupation phases. Mouseion Ser. 3, 10(2), 217–242. Nevett, L. (1999) House and Society in the Ancient Greek world. Cambridge, Cambridge University Press. Nielsen, I. (1993) Thermae et balnea: the Architecture and Cultural History of Roman Public Baths. Aarhus, Aarhus University Press. Pagano, M. (1999) Ufficio Scavi di Ercolano. Rivista di Studi Pompeiani 8 – 1997, 180–183. Parslow, C. (2000). The hydraulic system in the balneum venerium et nongentum of the Praedia Iuliae Felicis in Pompeii. In G. C. M. Jansen (ed.) Cura Aquarum in Sicilia, 201–214. Leiden, Stichting BABESCH. Peña, J. T. (2007) Roman Pottery in the Archaeological Record. New York, Cambridge University Press. Perkins, D. R. J. (2004) The Roman villa at Minster-in-Thanet. Part 1: introduction and report on the bath-house. Archaeologia Cantiana 124, 25–49. Philp, B. (2012) Discovery and Excavation of the Roman Shorefort at Dover, Kent. Dover, Kent Archaeological Rescue Unit. Proctor, J. (2012) Faverdale, Darlington: Excavations at a Major Settlement in the Northern Frontier Zone of Roman Britain. London, Pre-Construct Archaeology Monograph 15. 145 Revell, L. (2007) Military bath-houses in Britain: a comment. Britannia 38, 230–237. Revell, L. (2009) Roman Imperialism and Local Identities. Cambridge, Cambridge University Press. Rook, T. (1992) Roman Baths in Britain. Princes Risboro, Shire. Sanders, G. D. R. (1999) A Late Roman bath at Corinth: excavations in the Panayia Field, 1995–1996. Hesperia 68(4), 441–480. Schiffer, M. (1985) Is there a “Pompeii Premise” in archaeology? Journal of Anthropological Research 41(1), 18–41. Shelton, K. J. (1981) The Esquiline Treasure. London, British Museum. Stephens, J. (2008) Ancient Roman hairdressing: on (hair)pins and needles. Journal of Roman Archaeology 21, 111–132. Thüry, G. E. (2001) Müll und marmorsäulen: siedlungshygiene in der Römischen antike. Mainz am Rhein, Philipp von Zabern. van Driel Murray, C. (1994) A question of gender in a military context. Helinium 34(2), 342–362. Van Vaerenbergh, J. (2011a). Flush water for toilets in and near the baths. In G. C. M. Jansen, A. O. Koloski-Ostrow & E. M. Moormann (eds) Roman Toilets: their Archaeology and Cultural History, 78–86. Leuven, Peeters. Van Vaerenbergh, J. (2011b). Location of toilets within baths. In G. C. M. Jansen, A. O. Koloski-Ostrow & E. M. Moormann (eds) Roman Toilets: their Archaeology and Cultural History, 115–119. Leuven, Peeters. Wallace-Hadrill, A. (1996) Engendering the Roman house. In D. E. E. Kleiner & S. B. Matheson (eds) I Claudia: Women in Ancient Rome, 104–115. Austin, University of Texas Press. Wallace-Hadrill, A. (2011) Herculaneum: Past and Future. London, Frances Lincoln. Wardle, A. (2008) Bene lava: bathing in Roman London. In J. Clark, J. Cotton, J. Hall, R. Sherris & H. Swain (eds) Londinium and Beyond, 201–211. York, Council for British Archaeology. Whitehead, J. K. (2006) Excavation of the baths at Carsulae 2006. Etruscan News 7, 9–10. Whitmore, A. (2013) Small Finds and the Social Environment of Roman Public Baths. Unpublished doctoral thesis, University of Iowa. Whitmore, A. (forthcoming) Artefact assemblages from Roman baths: expected, typical and rare finds. In A. Binsfeld, S. Hoss, & H. Pösche (eds) Thermae in context: Roman bathhouses in the town and in daily life. Archaeologia Mosellana 10. Luxembourg, Musée national d’histoire et d’art. Whitwell, J. B. (1976) The Church Street Sewer and an Adjacent Building. Archaeology of York 3/1. York, Council for British Archaeology. Wild, J. P. (1970) Textile Manufacture in the Northern Roman Provinces. Cambridge, Cambridge University Press. Wilson, P. R. (2002) Cataractonium: Roman Catterick and its Hinterland. York, Council for British Archaeology. Yegül, F. (2010) Bathing in the Roman World. New York, Cambridge University Press. Zienkiewicz, J. D. (1986a) The Legionary Fortress Baths at Caerleon: the Buildings. Cardiff, National Museum of Wales and Welsh Historic Monuments. Zienkiewicz, J. D. (1986b) The Legionary Fortress Baths at Caerleon: the Finds. Cardiff, National Museum of Wales and Welsh Historic Monuments.