On the Production and Distribution of the Ceramic Building Material in Vindobona

STUDIA HERCYNIA XXI/2, 81–96 On the Production and Distribution of the Ceramic Building Material in Vindobona Tomáš Janek ABSTRACT This paper deals with the technical aspects of ceramic building material production in the brickyard located near the legionary fortress of Vindobona (nowdays Vienna, Austria). First, the general procedures of brick manufacture are described, then the paper focuses on the material from Vindobona itself. Tegulae, which make up most of the preserved evidence, are also treated at some length. In comparison to other ceramic building material, tiles are more distinguishable, thus more criteria can be observed. The observed criteria were: treatment of the surface, proportions, types of lower cutaways and shape of flanges. On the basis of these criteria, it is possible to distinguish differences in working procedures, which may relate to achange of units in the fortress and an exchange of workers within one unit. Attention is also paid to the economic aspect of production, which is reconstructed on the basis of the‑ oretical calculations. The amount of material necessary for the construction of camps was calculated along with the estimated time which it took to produce this material and the necessary work ‑power. The last part deals with the distribution of bricks to the forts in the upper Pannonian Limes, with an attempt to determine if the material was transported to the construction sites from Vindobona or was produced on the sites. The results show it was more cost effective to transport the material over even long distances. KEYWORDS Roman; ceramic building material; brick production; Vindobona; Pannonia. INTRODUCTION Building ceramics form an important body of material from Roman camps and buildings in general. Since building inscriptions are rather rare, the identification of individual building phases is (except for pottery and coins) also based on stamped material. For this reason, the research focus in the past used to be solely on stamped material. In this paper, Iwill talk about the technical aspects of brick production, which were almost completely overlooked in the past. More specifically, the focus will be on the area of the Roman legionary fortress of Vindobona and the brickyard located nearby. After ageneral description of brick production, attention moves to ceramic roofing, especially tiles. There are two main reasons for such afocus. The results presented in this paper are based on museum collections, which consist mostly of teg‑ ulae. Tegulae are also more complicated in shape than other ceramic building material, which allows one to create more comparison criteria. Since the research is still ongoing, the data and the observations mentioned in this article should not be considered final. The purpose is rather to show how the material should be approached and what procedures can be used to reconstruct the production and distribution of the ceramic building material. HISTORICAL BACKGROUND Judging by the stamped material, construction works in the upper Pannonian Limes were conducted mostly by three units, garrisoned in the castra legionis of Vindobona or at Carnun‑ 82 STUDIA HERCYNIA XXI/2 Fig. 1: Scheme of ceramic building material distribution produced by the 13th legion stationed in Vindobona and the 15th legion stationed in Carnuntum in the camps on the limes (the lower line) and buildings placed in Barbaricum (the upper line). Fig. 2: Scheme of ceramic building material distribution produced by the 14th legion stationed in Vindobona and the 15th legion stationed in Carnuntum in the camps on the limes (the lower line) and buildings placed in Barbaricum (the upper line). 83TOMÁŠ JANEK tum. The relationship between the units stationed in Vindobona and Carnuntum and their building activity is not very clear. The ratio between the relevant stamps and their spread across the sites is very similar and sometimes they are even found next to each other. There is evidence of abrickyard in Vindobona (Mosser 2013, 161), but we have found none in Car‑ nuntum so far, although analysis of the clay suggests that production also took place there. The first unit stationed in Vindobona was the 13th legion, which started the construction of the fortress around the year 97/98 AD. It did not stay for long and after four years the legion was transferred to participate in the Dacian Wars (Mosser 2005, 131). During this period, the construction works on this part of the limes seem to be mostly conducted by the 15th legion, stationed in Carnuntum (Fig. 1). Around the year 101 AD the 14th legion arrived in Vindobona and continued working on the construction of the fortress (Brandl 1999, 150). It is however possible, that the unit arrived to Vindobona already in 97/98 AD and started the construction works along with 13th legion (Mosser 2014, 203). Stamps of this unit are common on sites along the limes and even in Barbaricum (Fig. 2), though it is not clear whether they made the building material during their stay in Vindobona or after moving to Carnuntum in 114 AD (Mosser 2014, 205). The last unit garrisoned in Vindobona was the 10 th legion, which remained in the fortress till Late Antiquity (Brandl 1999, 116) and participated on most construction works on the limes (Fig. 3). Fig. 3: Scheme of ceramic building material distribution produced by the 10th legion stationed in Vindobona and the 14th legion stationed in Carnuntum in the camps on the limes (the lower line) and buildings placed in Barbaricum (the upper line). 84 STUDIA HERCYNIA XXI/2 DESCRIPTION OF CERAMIC BUILDING MATERIAL PRODUCTION SOURCES The only ancient source we have about the brick production is Vitruvius (II, 3.1). Although he gives us just the description of the mud bricks manufacture in the Mediterranean, many details can also be applied for the production of fired bricks in the Roman provinces. Because this is the only source from antiquity, we need to seek further information and parallels in medieval and early modern times. In these periods, production of hand ‑made fired bricks was widely spread and can provide us with more technical information and descriptions of practical experience. By combining these sources with finds from archaeological excavations, it is possible to reconstruct the processes of the Roman provincial brick manufacture. Practical experience acquired by experimental archaeology also plays an important role in this research. PROCEDURES OF BRICK PRODUCTION If the brickmakers wanted to attain high quality products, the procedures of choosing and pre‑ paring the clay had remained the same since the time of Vitruvius. The most important step was choosing the clay: ‘They should not be made of sandy or pebbly clay, or of fine gravel, because when made of these kinds they are in the first place heavy; and, secondly, when washed by the rain as they stand in walls, they go to pieces and break up, and the straw in them does not hold together on account of the roughness of the material. They should rather be made of white and chalky or of red clay, or even of acoarse grained gravelly clay. These materials are smooth and therefore durable; they are not heavy to work with, and are readily laid’ (Vitruvius II, 3.1). If there was no suitable source near the brickyard, the clay could be transported from distant places. The raw material needed to be further processed. The clay was put in the working pits and watered to clean out undesirable objects. Heavy parts such as stones sank to the bottom, while light particles like grass, roots and wood fragments floated and could be removed. Large amounts of water were needed, therefore the brickyard had to be located near awater source. Workers helped the process by stomping the clay with their bare feet to more efficiently mix the material (Federhofer 2007, 13). This ensured that the clay was more homogenous and solid after firing. It also lowered the possibility of cracking during the firing process. After the clay was prepared, it needed to be put in moulds. The most common types of build‑ ing ceramics were lateres and tubuli for floors and walls, and tegulae and imbrices for roofing. The sizes were derived from the Roman foot. Lateres were divided into bessalis (20×20 cm), pedalis (30×30 cm), sesquipedalis (45×45 cm) and bipedalis (60×60 cm). Tegulae usually meas‑ ured one and ahalf Roman feet (Federhofer 2007, 14). To achieve the same proportions and shape, wooden moulds were used (Warry 2006, 9). The use of various types of moulds and technologies can be observed on the building ceramics. Those variations may be associated with technological change in different periods or may be related to the diverse working hab‑ its of the brickmakers. Raw building ceramics firstly needed to dry in the wooden moulds so it could preserve its shape and become detached from the wood (Warry 2006, 28). For this purpose, large drying halls were used, which could have had awidth from 8 to 18 m and alength of over 50 m (Mosser 2013, 148). The prolonged shape facilitated better air circulation. Drying had to be slow and it took approximately one week. The clay was checked using the fingers, which left marks of various shapes (Kurzmann 2006, 18). Some of the shapes appear repetitively and the marks can therefore also have asecondary meaning, which is unknown to us. If everything was all right, the brick was probably stamped, though the true meaning 85TOMÁŠ JANEK of the stamping is also unknown. The stamps could have served as asign of quality control, or marked military property. Dies for stamping were made of wood, iron and probably also clay (Kurzmann 2006, 24). After the brick was stamped and removed from the form, it needed to dry for another week before it could be fired in akiln (Warry 2006, 28). Apart from stamps and finger ‑marks, there are exceptional bricks which bear numeric inscriptions made by the workers before firing. They are potentially avaluable source of information about the amount of daily production, or the date when the bricks were made. They can be interpreted as proof of production control by officials, or they could have served as some sort of ‘delivery note’. Inscriptions are usually in the form of rows of numbers. Bricks from Lauriacum held 16 numerals separated into four rows: 1– XLII XXXII XLII XX XXXX (total 176); 2– XXX XXX XXX LXXX (total 170); 3– XXX XL XXX LXXXX (total 190); 4– XXX XXX XXXX (total 100) (Ruzicka 1919, 110). Three different people wrote these values and their total sum is 636 bricks. Similar values were inscribed on bricks from Siscia:…Ikal iunias / Candidus CCXX / Iustinus CCXX / Felicio CCXX / in uno DCLX (29th June / Candidus 220 / Iustinus 220 / Felicio 220 / altogether 660); XIII koctobr(es) / Fortis CCXXII / Candidus CCXXV / Iustinus CXXXVII / Artemas CLXXXXVIII / min XXI (20th October / Fortis 222 / Candidus 225 / Iustinus 137 / Artemas 198/…21). These numerals can be interpreted as the daily production of one worker. The edict of Di‑ ocletian only supports this theory with the prescribed norms of daily production: one man had to make 100–120 large bricks or 220 small bricks. According to the inscribed dates, which stretch from June to October, the production was seasonal. As was described by Vitruvius (II, 3.2), the climate of continental Europe was not suitable for making bricks during the winter. The low temperatures prevented most of the brick making works, starting with digging the soil to drying the formed bricks. The last, technologically most difficult step, was the firing of the bricks. Roman kilns were sunk in the ground and consisted of aworking pit, afurnace opening and afiring chamber of rectangular shape with abottom grate (Federhofer 2007, 33). Peter Warry claims on the basis of an experiment that ten people were necessary to maintain one kiln (Warry 2006, 121). Two people controlled the firing process and regulated the temperature. Eight people were required to carry the wood and to keep the fire burning. During the firing, the temperature in the kiln reached over 900 °C and the whole process took four days or more (Federhofer 2007, 18). After cooling, the kiln was opened, the bricks removed and probably stored in the warehouses, or in the part of drying hall waiting to be distributed to the construction sites. SIGNS OF DIFFERENT PRODUCTION TECHNIQUES LEFT ON TILES The technical aspect of brick production discussed in this article is based on the examina‑ tion of bricks from the Roman sites of Bratislava ‑Rusovce and Stupava (the Slovak National Museum), Mušov ‑Burgstall (the Institute of Archaeology of the Czech Academy of Sciences, Brno) and Vienna (Wien Museum). The most important is the collection deposited in the Wien Museum, consisting of more than 4,500 stamped bricks. The majority of bricks from these sites had their origin in the brickyard located in Vienna’scity district of Hernals. The main goal was to distinguish traces of technical procedures carried out during the brick making. Although various production techniques were observed also on lateres and imbrices, aspecial focus is placed on tegulae only, which form the majority of finds. The observed criteria are based on Peter Warry’sresearch of tiles in the Roman province of Britannia. The criteria were: the treatment of the surface and proportions and types of lower cutaways. The most important are the cutaways, the bottom edges of tiles, which were on one 86 STUDIA HERCYNIA XXI/2 part cut away, so they can fit better to each other. Their various forms can be used as dating criteria (Warry 2006, 43–45). In this observation, the aim was to verify if the different types of cutaways also appear on the material from the examined sites. The shape of flanges was also included in the observations. RESULTS OF OBSERVATIONS Due to the mostly fragmentary condition of the finds it was not possible to observe the com‑ plete dimensions of the tiles. The thickness was measured on the breaks and varies from 2 cm to 4 cm, however groups with the same thickness seem to appear. The upper surface of the tegulae was always carefully levelled, probably with astave, leaving small straight scratches. The bottom surface shows various kinds of treatment, based on how the tile was separated from the workbench. Four basic methods were distinguished: 1. The workbench was lined with coarse sand, which got stuck to the bottom of the tile. The clay was thus not moulded precisely, leaving an uneven surface with folds (Pl. 3/1). 2. The workbench was lined with fine sand, traces of which can be found at the bottom of the tile. The surface is now straight and the clay was moulded carefully. 3. The workbench was lined with coarse sand, which was later removed from the bottom of the tile with aknife or similar device. This kind of treatment can be distinguished by preserved cutting marks, which are often in various directions (Pl. 3/2). 4. The clay was put directly on the surface of the workbench and removed afterwards with awire, leaving long shallow scratches from grains swept away by the wire (Pl. 3/3). Fig. 4: Photographs of various cutaway types on tiles (Wien Museum). Photo by author. 87TOMÁŠ JANEK The different kinds of treatment seem to be linked to the various working habits. The difference in quality might be the result of changing demand for the material, lowering the quality of treatment in aperiod of larger building activity and higher demand for the building material. For the lower cutaways, five different types were observed in the deposits of the Wien Museum (Figs. 4 and 5): Type Ais created by inserting two rectangular blocks into the moulding form, covering the whole width of the flanges. All such tiles were stamped by the 13th legion (Vienna, Wildpret‑ markt 8–10; Wien Museum, Inv. 17200/003). Type B is created by two cuts, one diagonal along the flanges, the other vertical to the former. This type appears on tiles stamped by both the 10th and 14th legions (Vienna, Hoher Markt 9; Wien Museum, Inv. Neumann 375). Type C is created by inserting arectangular block into the form, curved on one side, covering half of the height of the flanges. This type appears on tiles stamped by the 10th legion (Wien Museum, Inv. Neumann 705). Type D represents avery interesting group, which at first was moulded as the Type C cut‑ away and afterwards adjusted with two diagonal cuts to match the Type B. Sometimes the partial imprints of the wooden blocks from the form remain and sometimes they are almost completely cut away. This type appears on tiles stamped by the 10th legion. Some finds bear the stamp ‘LEG X GPF AN’ and can thus be dated to the reign of Caracalla (Vienna, Hoher Markt 9–11; Wien Museum, Inv. Neumann 387). Type E is created by three cuts, removing asmall rectangular block from the bottom. This type appeared on tiles stamped by the 14th legion (Vienna, stray find; Wien Museum, Inv. Neumann 1492). The flanges have various profiles. After the clay was bent into the flanges, they were further shaped with the hands or other devices. The side, facing the inside of the tegula, was finished with asharp edge or with along line made with one or two fingers. The different forms are represented in the table (Fig. 6). Further research is needed to determine if the typology can be used for dating purposes. 88 STUDIA HERCYNIA XXI/2 Fig. 5: Table with ideal drawings of cutaway types on tiles, based on the finds from Vienna. Drawings by author. 89TOMÁŠ JANEK Fig. 6: Table with ideal profile drawings of tile flanges, based on the finds from Vienna. 96 STUDIA HERCYNIA XXI/2 ACKNOWLEDGEMENTS Iwould like to express my gratitude to Michaela Kronberger (Wien Museum) and Martin Mosser (Museen der stadt Wien ‑ Stadtarchäologie) for allowing me to study the material from the collection of the Wien Museum. This article is the outcome of the project ‘Roman Building Terracotta in the Pannonian Limes and Area along the Danube’ funded by the Grant Agency of Charles University (GAUK 277215). BIBLIOGRAPHY Vitruvius = Marcus Vitruvius Pollio: The Ten Books on Architecture. Transl. M.H. Morgan. Cambridge, Mass. 1914. Brandl, U. 1999: Untersuchungen zu den Ziegelstempel römischer Legionen in den nodrwestlichen Provinzen des Imperium Romanum. Katalog der Sammlung Julius B. Fritzmeier. Passauer Universitätsschriften zu Archäol‑ ogie 6. Rahden/Westf. Federhoffer, E. 2007: Der Ziegelbrennofen von Essenbach, Lkr. Landshut und Römische Ziegelöfen in Raetien un Noricum. Passauer Universitätsschriften zu Archäologie 11. Rahden/Westf. Kurzmann, R.2006: Roman Military Brick Stamps: aComparison of Methodology. Oxford. Mosser, M. 2005: Befunde im Legionslager Vindobona. Teil V: das Intervallum an der westlichen Lagermau‑ er– Vorbericht zu den Grabungen Am Hof in den Jahren 2008/09. Fundort Wien 13, 50–76. Mosser, M. 2013: Zwei römische Ziegelöfen in Wien 17, Steinergasse 16/Geblergasse 47. Fundort Wien 16, 144–161. Mosser, M. 2014: Die legio XIIII Gemina Martia Victrix in Nordwestpannonien am Ende des 1. Jhs. n. Chr. In: F. Lang – S. Traxler – E.M. Ruprechtsberger – W. Wohlmayr (eds.): Ein kräftiges Halali aus der Römer‑ zeit! Festschrift Norbert Heger. Archaeoplus. Schriften zur Archäologie und Archäometrie der Paris Lodron‑Universität Salzburg 7. Salzburg, 201–213. Mosser, M. 2015: Die Legionsziegelei von Vindobona im 17. Wiener Gemeindebezirk. Fundort Wien 18, 50–93. Rajtár, J. 2005: Na úsvite dejín: Rimania abarbari. In: H. Moravčíková (ed.): Architektúra na Slovensku. Stručné dejiny. Bratislava, 12–17. Ruzicka, F. 1919: Ziegel aus Lauriacum mit Schrift. Römisches Limes in Österreich 13. Wien/Leipzig. Visy, Z. 2003: The Roman Army in Pannonia. Pécs. Warry, P.2006: Tegulae: Manufacture, Typology and Use in Roman Britain. British Archaeological Reports 417. Oxford. Tomáš Janek Institute of Classical Archaeology Faculty of Arts, Charles University Celetná 20, CZ‑11000 Prague 1 [email protected] 188 STUDIA HERCYNIA XXI/2 Pl. 3/1: Bottom of tegula spilled with coar‑ se sand (Vienna, stray find, Vienna Museum inv. 1492). Pl. 3/2: Bottom of tegula, sand was removed with knife or similar device (Vienna, Hoher Markt 3, Vienna Museum inv. 319). 189PLATES Pl. 3/3: Wired bottom of tegula (Vienna, Hoher Markt 4, Vienna Museum inv. 378).