A number of questions are posed by this evidence for the local production of fine and other wares, including the origin of the potters, their relationship with other immigrant potters and why they came to Brough. Potters came to Britain from the Continent at various times from the conquest period onwards. Leaving aside the potters accompanying the army, the earliest recognised industry is represented by the Colchester kilns, producing a very similar range to that found at Brough, production possibly starting c.AD 120. The kilns producing colour-coated vessels of the type considered here are conventionally dated to the latter part of the 2nd to early 3rd century. Some of the fine ware vessel types produced at the kilns at Wilderspool suggest a similar early 2nd century start, although these types may be due to influences from production in the 1st and 2nd century at Wroxeter (Hartley and Webster 1973).
Possibly rather later in the 2nd century, potters from the Continent producing fine wares arrive in the Nene Valley, although the production of rough-cast beakers known at Great Casterton may suggest an early 2nd century date, and in the Antonine period potters similarly came to South Carlton in Lincolnshire (Webster 1944), where a range of rough-cast beakers, flagons and mortaria were made, the latter traded to the northern frontier and the Antonine Wall. There is also evidence for rough-cast beakers being made in Oxfordshire (Paul Booth, pers. comm.), and other production sites probably remain to be identified.
In examining the possible origin of potters from the Continent in the 1st century broad stylistic zones could be identified, and are discussed with great clarity by Greene (1993, 45-9). By the later 2nd century these zones have become somewhat blurred, but differences still occur between the Lower Rhine and the southern part of the Upper Rhine, the Wetterau and provinces to the east: Raetia, Noricum and Pannonia. The appearance of a Raetian mortarium and uncommon barbotine ‘hairpin' and other vessels at Brough suggest it is to this latter area that attention should be directed in the search for the origins of the potters, although the ease of movement of potters to and from East and Central Gaul has to be borne in mind.
The Colchester potters seem to have originated in East Gaul, shown most strongly in the production of Colchester samian (Hull 1963; Hartley 1977), and the close connection between samian and colour-coated wares is shown by the appearance of a stamp of the samian potter Acceptus ii on a beaker of form 391 with barbotine decoration (Hull 1963, 191, fig.50, no.1). Most of the vessels produced at Brough can be paralleled at Colchester in the pottery from the 1933 and 1959 kilns (ibid., figs 57-9), but the ‘hairpin' style of decoration, various other vessels and the Raetian mortarium do not appear. There are, however, indications of influences more likely to have come from the south Upper Rhine to the Raetia area, such as the bowls of the same type as Raetian mortaria, with scalloped ‘handle' projections, internal concavity, and the wider flange, not normally seen on British-made types (ibid., fig.73, nos.16 and 17). The use of roller-stamping is also common in the eastern provinces (ibid., fig.58, nos.16 and 18 on beakers; fig.71, nos.6 and 7 on pedestalled vessels), and colour-coated barrel-beakers (ibid., 82 and fig.79, nos.1 and 2 from grave 302 both in local colour-coated ware, cf fig.47, 11-16 in samian) occur in the same area (Szönyi, 1973, 93, abb 6, form B), alongside Raetian fine wares. The bowl with inturned rim (ibid., fig.73, no.19) appears to be from the same tradition as the Brough bowls, nos.55-6, of the range Gose (1950) 484-7, derived from one of the commonest La Tène forms in the Upper Rhine area. These all appear to be rarities at Colchester, and may therefore be the products of a single potter.
Wilderspool has a specific bearing on the kiln ‘waste' from Brough due to the production there of the same type of cornice-rimmed beakers, rough-cast and rouletted (Hartley and Webster 1973), and also Raetian-style mortaria (Hartley forthcoming). Further evidence of influence from Raetia lies in the painted fine wares and roller-stamping (Hartley 1981), with parallels at Straubing (Walke 1965, Tafs 54-5 and 82-3) and Cambodunum (Fischer 1957, Taf 13, nos.6 and 7; Taf 14, nos.2 and 3), both sites with Raetian mortaria and Raetian fine wares.
The repertoire of the potters working at Brough needs to be broken into its component parts. There are certain types which became common currency amongst potters working on the Continent in the 2nd century, such as the rough-cast beakers, the style starting in the 1st century, and best known in Britain as Lyon and Central Gaulish colour-coated wares. The cornice-rimmed beakers, folded types with both cornice- and curved-rims, plain-rimmed beakers and zones of rouletting are all types with a wide geographical spread while that for the so-called hunt cups is more confined. The unusual vessels are the beakers decorated with barbotine ‘hairpins' and ‘wish-bones' or ‘tear-drops' (nos.22, 23 and 24) and vegetal decoration (no.25), both combined with rouletted zones.
Beakers decorated with ‘hairpins' and ‘tear-drops' are well known in Central Gaulish colour-coated ware (Greene 1979, 45; Symonds 1992, 7), produced during the Flavian-Trajanic period. Such beakers were produced in other areas, including East Gaul where crossed hairpins occur, and similar hairpin beakers are known from a kiln site at Brive in Central Gaul in the 1st to early 2nd century (Tilhard et al., 1991, fig.13, nos.3-14) which also produced bowls decorated with hairpins (ibid., fig.14, nos.1-5). One of the beakers, a folded form with wish-bones between the folds, also has a basal rouletted zone. The hairpin beakers are reminiscent of fine wares made in Raetia, known as Raetian ware, where the combination of barbotine diagonal lines and rouletting is extensively used, together with crescent-like motifs. The diagonal lines are not, however, hairpins, and no motifs like the ‘wish-bone' or ‘tear-drop' seem to occur; hairpin beakers in Switzerland (as at Augst, Ettlinger 1949, Taf 22, nos.10, 13) and further east (as Fischer 1957, Taf 15, nos.15-17) are all likely to be Gallic imports. Nevertheless, the combination of hairpins with rouletted zones seems to be a feature of fine wares more in the Raetian than Gallic tradition. The vegetal barbotine motif on no.25, on the other hand, resembles the Gallic style rather than that seen further east.
The other colour-coated vessels include the rim of a hemispherical bowl, no.54, decorated with rouletting, and the possibly local bowl no.82 — a relatively common form but one which can be paralleled at Cambodunum (Fischer 1957, Taf 12, no.8). The shallowly cupped flagon rim, no.52, may be from a jug such as the Colchester-type 383, and similar Colchester parallels occur for boxes and their lids, nos.57, 58, 59 and 60. The derivation of the box form is uncertain, although it resembles Gose (1950) 497 and perhaps the lid 562, both in a white clay, Niederbieber forms 105 and 120b; the inturned rim is a common late La Tène type.
The more surprising vessels are no.53, probably a tazza, the shallow bowls/dishes, nos.55 and 56, and the face-mask, no.61, in exceptionally poor condition, but with traces of colour-coating. This was probably applied to a closed form, perhaps a flagon or jug or a beaker, and is too fragmentary for useful comment; appliqué faces occur in the samian repertoire. The only feasible parallels for the vessel no.53, particularly since fragments probably from the same vessel include part of a pedestal base, are with tazze, not all of which are notched or frilled, and it is similar to one from Cambodunum (Fischer 1957, taf 27, no.6). The closest parallels traced for the inturned bowls or dishes with their flat-topped rims, nos.55 and 56, are at a fort, Öhringen-West, in the Odenwald-Neckar-Limes, adjacent to the start of the Raetian limes (Schönberger, 1973, Abb 12, 85b and c). This 2nd century site also has a Raetian mortarium and Raetian fine wares.
The unslipped BROX vessels include relatively unexceptional types, the flagons, nos.62 and 63, a splayed-wall dish or bowl, no.64, a platter similar to the Pompeian red ware type (also made at the Colchester kilns (Hull 1963, fig.57, nos.19-21), no.66, and the curious inturned bowl or dish, no.65. Although some of the beakers have been recorded as unslipped, the non-appearance of colour-coating may be solely due to the condition of the kiln ‘waste'. The inturned bowl or dish no.65 is another late La Tène type, more common in the Upper Rhine and the provinces to the east, paralleled at the 2nd century Öhringen-West fort (Schönberger, 1973, Abb 11, 83b) but also at Augst (Ettlinger 1949, taf 9, nos.1-3).
An added complication to establishing the repertoire is no.85, a small fragment of a plain-rimmed bowl with a bead or flange, in a fabric and slip very close indeed to the BRCC, with indications of painted vertical stripes internally. Comparison with sherds of York painted ware left attribution of this single sherd to that ware in doubt. If the form is a copy of a samian form 38, these are rare at York (Monaghan 1997, 878) and the plain rim does not resemble the two published examples (Perrin 1981, no.368; 1990, no.1203), or the closed form 24/25 (Monaghan 1993, no.2855). It is possible, therefore, that this is also a local product, particularly as it came from Trench I (the fill of 76) with quantities of ‘waste'. As with the Wilderspool painted vessels, an origin for this tradition of painting vessels may lie in the eastern provinces.
Marbled and painted wares have been recently discussed in relation to Pannonia; Krekovic (1997) shows that vessels closely similar to the range known from York are dated mostly to the first half of the 2nd century (as ibid., fig.2, nos.2, 6,7, although the classic Ebor hemispherical bowl differs). The term ‘marbled' is a mis-nomer for many of these, where the decoration is more often short brush-painted lines. He concludes that the marbling technique was probably spread by military as well as civilian potters; in Pannonia production is evident on civilian settlements, while marbled pottery at Nijmegen has been found more frequently on civilian sites. Apart from Pannonia, the ware is also well known in Raetia.
Little can be said about the occasional sherds which may be associated with local production, such as the bowl, no.87, from Trench 4 (Period 4, Phase 6, group 34.4) which are in a slightly coarser fabric than seen in the main ‘waste', but with traces of red colour-coat on the interior, and the strange lid-seated vessel, no.77 of similar fabric, unslipped (Trench 1, Period 6, Phase 1, group 12.3). Neither are sufficiently distinctive to lead to useful parallels.
The mortaria, particularly the Raetian type, no.321, and the collared types nos.322, 323 and 324 all came from Trench 1, not directly associated with the kiln ‘waste' deposits, and nos.321 and 323 are worn. The earliest stratified is no.323 from Period 4, while the others occurred in Periods 6, 7 and 9. The fabric of no.321 appears to be identical to other sherds from the production ‘waste', and is of Hartley's type Aii (Hartley forthcoming), close to an example probably from Lancaster made at Wilderspool (Hartley forthcoming, fig.4, Aii). This is a form certainly made by a potter from a Continental workshop. These mortaria are distributed throughout the Rhine frontier, Raetia, Noricum and Pannonia, and clearly made in numerous potteries, including at Augst in Upper Germany (Furger 1990, fig.8, no.10, ten examples) in a kiln dated to the mid to late 2nd to 3rd century. The distribution in Britain is concentrated in the West Midlands and west coast, extending north to Bearsden and Ardoch. This belongs in Hartley's Category I, types alien to the Romano-British tradition, some of which must have been introduced by potters from the Continent, and her type A is the only one which can be paralleled fairly well on the Continent. Examples are known from Holt, Wilderspool, Wroxeter, Chester and one from York (Hartley forthcoming; York: Perrin 1975, fig.20, no.402).
No.322 is an undoubted ‘waster', both over- and under-fired in the single vessel, and its fabric is identical to that of the Raetian no.321. It has a white slip and traces of possible diagonal painted lines on the rim, probably with a painted band at the bottom edge of the rim, which would suggest it is possibly not earlier than the mid 3rd century. But this is an unusual type, and can be paralleled by one from Lorenzberg in Raetia (Ulbert 1965, taf 21, no.9) which has the same fairly wide groove on the wall. A more local example is a mortarium from Malton (Mitchelson 1964, 248, fig.14, no.161); the description, a pink fabric, grey in fracture, with a cream slip, suggests this may be the same fabric as the Brough mortarium. This came from a later 3rd or early 4th century deposit. No.324 has definite traces of white slip, and the red-brown fabric is within the range of the other BROX vessels; it also appears to be under-fired, and may dated to c.AD 180-230. No.323 appears to differ, having more quartz inclusions (dated c.AD 150-230) and, with the other fragmentary flanged example probably of 3rd century date, seems not to be associated with the production ‘waste'.
There is a wide range in the dates applied to the local mortaria, with the Raetian no.321 possibly pre-dating AD 140, and the other two fairly certainly local vessels, nos.322 and 324, ranging from the late 2nd to the mid 3rd century. This poses a problem so far as the Raetian vessel is concerned. The ‘waste' of beakers, flagons, boxes and other vessels give no reason to believe the dumps covered an extended period of production, and while rough-cast beakers could be earlier, it is difficult to see how a starting date other than mid 2nd century at the earliest can be applied to the group as a whole. This is, however, just one dump of ‘waste', and potters making mortaria may not have been working in this same area. The Raetian mortarium suggests the arrival of Continental potters in the early 2nd century, broadly contemporary with the Colchester potters.
The final factor to be taken into account is that of the grey wares, highlighted by the presence of the ‘waster' bowl no.167. There is nothing sufficiently distinctive about the local fabric to enable the positive identification of local grey wares which may have been associated with the fine ware production, and many of the grey vessels will have been locally made, using similar clays, but at different periods, unrelated to the fine ware production phase. It is, however, worth noting some of the vessels which either on the basis of fabric and form, and/or ‘second' condition, may belong to the same period. Very few reduced sherds were found with the ‘waste', although the jar no.137 is a strong candidate for local production, together with the bowl no.167.
The ‘waster' bowl no.167 (other examples nos.169 and 170), which equates with Gillam (1957) 301 from Ilkley, is a relatively common type in North Lincolnshire, being type S at the Roxby kilns (Rigby and Stead 1976, 139-1470), and already known from Brough (Corder & Romans 1937, pit 11, fig.12 nos.62-64; 1938, fig.15, no.26; Wacher 1969, fig.64, no.289), Lincoln (Darling 1984, fig.15, no.47 and many other examples unpublished), Dragonby (Gregory 1996, fig.20.14, no.1023; fig.20.25, no.1310; fig.20.33, no.1442 from Kiln 5, Hadrianic-early Antonine), York (Monaghan 1997, fig.401, nos.4010-11) and Malton (Bidwell and Croom 1997, fig.26, no.129; fig.29, no.231). This is a seemingly very localised bowl type, and does not appear to derive from British or Gallic antecedents. It is interesting that a bowl of this type occurs at the fort of Oberstimm in Raetia (Schönberger 1978, taf 87, D364), dated c.AD 90-120. The bifurcated rim, so characteristic of this form, has additional fingered decoration, but it is remarkably similar.
In stratified deposits in Lincoln, this bowl type is often associated with three other regional types all represented from these excavations, the platter derived from a Gallo-Belgic type, nos.226, 227 and 228, the carinated jar or bowl, as nos.172 and 173, and the shallow bowl nos.83, 164, 165 and 166. The platter and carinated jar or bowl are both made at the Roxby kilns (types E and H), and occur amongst the products of the Trent kilns at Newton-on-Trent and Lea (Field and Palmer-Brown 1991). There are numerous examples from Brough, Old Winteringham, Winterton (Rigby and Stead 1976), Doncaster (Buckland et al. 1980), Dragonby, probably kiln products (Gregory 1996), Malton (Bidwell and Croom 1997), Rudston (Rigby 1980) and other sites on either side of the Humber. The platter is a form often stamped as no.226 (as at Lincoln, Darling and Jones 1988, 24, fig.8, no.81; Doncaster, Buckland 1986, fig.12, no.9) and its relationship with the bowls of the type of no.167 is shown by a stamped version of one of these in the same fabric (ibid., fig.12, no.8), also occurring at Lincoln (Darling forthcoming a). It is notable that none of this type of platter had the complex internal burnished designs seen in examples from earlier excavations (Wacher 1969, nos.234, 250, 284) and also common at Dragonby (Gregory 1996, 519).
The shallow bowl-type, such as nos.83, 164, 165 and 166, has a more confined geographic spread. Examples are already known from Brough (Corder & Romans 1937, fig.10, no.3; fig.12, no.61; 1938, fig.14, no.18; fig.13, no.90; Wacher 1969, fig.63, no.274), York (Monaghan 1997, fig.400, no.4003), Malton (Bidwell and Croom 1997, fig.26, no.125; fig.36, no.409), Lincoln (Webster 1949, fig.14, no.72; Coppack 1973, no.11; Darling 1984, fig.15, nos.45-6) and Dragonby (Gregory 1996, fig.20.6, no.835; fig.20.26, no.1329; fig.20.27, no.1361; fig.20.34, nos.1459-1460 from Pit F 2567 with kiln waste; variant: fig.20.24, no.1290). The possibility of the derivation of the form from a Pannonian type has been raised by Swan (1996), who notes other British examples including some from Verulamium (Frere 1984, fig.100, nos.2383-4, deposit AD110-130), and an unpublished example from Braughing, Herts. Swan has drawn attention to the broad similarity of the form to the bowl Gose (1950) Type 494 (type example from Trier, later 1st century), but the closer parallel is in Pannonia (Bónis 1942, fig.21, no.23; fig.23, no.10). The first of these Pannonian vessels is a shallow dish, while the second is more of a bowl; both are painted on the flange, and have much larger diameters (c.20 and 34cm) than the British examples. The examples from these excavations do not share the same fabric, and no.166 is heavily burnt, as if it had been used as a lid. These two bowls types, with parallels suggesting derivation from forms appearing in the eastern provinces, are of considerable interest for a better understanding of the pottery of this region, and need further research.
Only three vessels appear to show signs of being what could be termed ‘seconds' which may, or may not, indicate production locally: the jar no.137 (found with production ‘waste' of Period 4), the possible bowl or jar no.173 and the small necked bowl no.181. There are a number of vessels which share similar micaceous fabrics, as jars 133, 140, 153 (Continental derivation), 157, bowls nos.164, 165, 166, 167, 168, 169, 170, 171, 191, 198, 199 and 200, dishes nos.225, 226 and 227, and lid no.255.
The affinities of the pottery indicate that the potters were Continentals, drawing on traditions likely to be prevalent in the southern part of the Upper Rhine, probably arriving as early as the early 2nd century. This relates them to the movement of similar potters to Colchester, Wilderspool and probably a number of other areas for which the evidence is sparse. Coarse ware types with similar antecedents in Lincolnshire may be evidence for other potters arriving at this time. Such relationships suggest the potters were civilians, and likely to be catering for both civilian and military requirements, as quickly became the case in the later 1st century. The viability of the market dictated the locations of such potters, and the buying power of the military is an important factor. Despite marketing to civilian settlements, the fortunes of the Wilderspool potters were to a large extent tied to the military market, and the movement of troops led to a similar movement of potters (Hartley and Webster 1973, 103). Kilns further south in Lincolnshire, in Lincoln and to the north at South Carlton (Webster 1944), also traded extensively with Yorkshire and the northern frontier, and the evidence from South Carlton mortaria suggests the kilns were located specifically for access to the Humber, and trans-shipment to the north.
Early recognition of the value of Brough as a base and trans-shipment point led to a stores-base there in the Flavian period, and the development of the vicus is dated to the Hadrianic to Antonine periods (Wacher 1969, 3). It is at this point in the history of the site that the potters arrive to take advantage of a developing civilian community, coupled with the chances of trading with the military to the north. The extent of their trading is likely to be difficult to establish, but close examination of rough-cast beakers on Yorkshire and other northern sites would be worthwhile. In the meantime, there is the possibility that the Gillam (1957) 301 from Ilkley derives from this influx of potters although, given the variety of fabrics and production at other kiln sites, this is not necessarily made by the Brough potters. The mortarium at Malton (Mitchelson 1964, fig.14, no.161) is a possible product, and trade between Brough and Malton is a strong possibility.
© Internet Archaeology
URL: http://intarch.ac.uk/journal/issue9/brough/potbrolo3.html
Last updated: Tue Nov 28 2000