The following is concerned with the identification, quantification and analysis of the pottery assemblages recovered in three interventions at the Cottam B site; fieldwalking in 1993 (COT93F) followed by excavations in 1993 (COT93E) and 1995 (COT95E). It builds on and adds to earlier investigations undertaken by metal detectorists (Haldenby 1990, 1992, 1994) and to an episode of structured fieldwalking in 1989 (Didsbury 1990).
In total 637 sherds were examined, weighing 6308 grams with an average sherd weight of 10 grams. Activity as represented by the pottery was assignable to all the main period descriptors from the Bronze Age onwards. In order to provide a general overview of the assemblage identified fabrics were grouped accordingly (specific sherds are referred to by context followed by small find number, eg 1234.100, or grid reference in the case of fieldwalking, eg A3)
The Medieval and Post Medieval assemblages have not been broken down into specific fabric types here. The former is represented by a very small group of sherds and the latter largely beyond the interest of the current project. Fuller details are available through referencing the project archive. Also a small number of sherds could not be assigned to specific fabrics and have been grouped together (0) for the purpose of analysis.
Quantification and analysis has been conducted in terms of sherd counts and by weight. A full breakdown of the assemblages from each intervention is given in appendix A - tables 3-5. Modern pottery studies usually see estimated vessel equivalents (eve's) as the most reliable means of quantification (eg Orton et al, 1993, 168-73); however, the dispersed and limited nature of the Cottam B assemblage precludes such an approach. When fabrics are grouped by period (fig 2) it is clear that late Roman pottery dominates accounting for over 70% of the assemblage by weight with the Anglian and Post-Medieval periods at a little over 10% each whilst other periods and unidentified sherds account for around 5%; however, it should be remembered that the scale of pottery production and hence usage has varied greatly in the past, for example, an apparent decline in the post-Roman period.
During January of 1993 fieldwalking was undertaken at the Cottam B site by students within the Department of Archaeology of the University of York under experienced supervision (Vyner, 1993). Divided into 30 metre squares just under 12 hectares (29 acres) were walked in the large arable field to the west of Burrow House Farm. Recovery levels seem to have been good judging from the size of some of the sherds in the assemblage with a number weighing under one gram. One difficult area may relate to Crambeck parchment ware (5) in that only rims were recovered. Presumably this reflects the difficulty of differentiating essentially white parchment ware body sherds from material in the topsoil brought up by ploughing from the underlying natural chalk of the Wolds; however, this is a relatively minor problem in that this ware is a minimal part of the major assemblage by period (see fig 1). Distributions in the fieldwalked area are intriguing (see also Appendix B)
Consists of a thin scattering of small abraded sherds, eight in total with an average weight of 4.0 grams. It is difficult to suggest that such an assemblage even represents manuring. The Iron Age material was concentrated to towards the west and north of the fieldwalked area broadly reflecting the later distribution of Roman sherds and it may be that the former became incorporated into midden in the latter period. In other words the Iron Age and Roman settlement foci correspond
There appears to be an ubiquitous spread of Roman material across the fieldwalked area which is mostly late in date. Subsequent excavations (see below) recovered a few sherds of later nd century Samian (3) and the forms of some of the non-specific grey wares (6) suggest they may date to this period or the earlier 3rd century but the assemblage is overwhelmingly later 3rd century through to the 5th consisting mainly of East Yorkshire grey wares (9), late forms of calcite gritted pottery (7) and Crambeck wares (5 & 8). Thus the pottery suggests a massive increase of interest in the area of the Cottam B site late within the Roman period. Peter Didsbury notes of the 1989 fieldwalking that the low average sherd weight (under 12 grams - see appendix A for COT93F) is unlikely to represent more than the process of manuring using midden (1990, 66). A few sherds appeared fairly fresh but densities within the fieldwalked area (an average of 27.5 sherds per hectare without significant clustering) do not suggest anything other than general background agricultural activity. Any settlement focus must lie outside the fieldwalked area. A close study of the distribution of Roman sherds reveals a tapering off towards the west and east of the fieldwalked area which possibly suggests boundaries or changes in land use in these directions.
In terms of the pottery assemblage there is very little evidence for activity within the earlier parts of this period. A sherd of organic tempered ware (11) was identified (H8) which was similar to a small number of sherds subsequently recovered during the 1993 excavations. It may be, as noted above, that this softish fabric breaks up fairly rapidly upon having its equilibrium disturbed through transition into a plough soil which could explain an albeit limited excavation presence against a fieldwalking paucity in an area where other data (metal detector finds) clearly demonstrate Middle Saxon activity. Apart from this and a occasional outlier the 1993 fieldwalking identified a discrete and exciting distribution of Late Saxon pottery in the extreme east of the fieldwalked area (Appendix B). Concentrated in an area approximately 200 by 100 metres the Late Saxon sherds fade towards the south but activity as represented by these sherds may well continue to the north and east outside of the area investigated. Consisting mainly of Torksey type ware (12) and some York Ware (13) this assemblage (see appendix A) corresponds with the distribution of a small group of 10th century metalwork recovered by metal detectorists and was subsequently related to settlement activity (see 1995 excavations below).
The Medieval component of the fieldwalking assemblage at the Cottam B site could be seen as insignificant; however, it does appear to support arguments for major landscape reorganization during the transition to Medieval England. The assemblage consists of 14 small abraded sherds widely spread within an area approximately 150 metres square to the west and south of the distribution of Late Saxon material (above) and interestingly corresponding with a major concentration of Middle Saxon metalwork finds, the area of the 1993 excavations. Whilst this appears to demonstrate a clear shift of focus it is likely that the Medieval sherds represent an occasional, perhaps seasonal, focus of activity. Alternatively there may have been an increase in the amount of land under pasture with the distribution of sherds representing a small arable area. Whichever the sparseness of material may suggest a remote source possibly reflecting the establishment of a new settlement foci as represented by the nearby Deserted Medieval Villages (DMV's) of Cottam and Cowlam.
There was a light spread of post-Medieval material over much of the fieldwalked area totalling 84 sherds with an average of seven per hectare although there was a distinct decrease in density of sherds to the west and south. The sherds presumably represent manuring and the observed patterning may reflect no-longer extant boundaries and changing land use. As noted above there was a relatively small amount of modern material in the assemblage which may perhaps reflects changing agricultural practice with domestic waste no longer being used for large-scale manuring.
Three 10 * 20 metre trenches were opened within the main crop-mark area at Cottam B and within area of the main concentration of Middle Saxon metalwork finds. In the event resources only allowed the detailed examination of two of these trenches. The third produced a solitary sherd of Crambeck grey ware (8) during clearance. Sampling within the combined 400 square metres of the other two trenches produced a limited but possibly illuminating assemblage.
In total 88 sherds were recovered (fig 3) 42 of which were associated with clearance or modern activity such as deep ploughing. The remaining material mostly dated late within the Roman period. The vast majority of these sherds were abraded and small with an average weight of 5.5 grams. Despite being the major component of the assemblage their number and condition suggest they are most likely to be redeposited from the general background scatter of Roman material noted during fieldwalking (see above - similarly for the three Iron Age sherds identified in this assemblage). Certainly residuality is confirmed for a number of contexts through the presence of artefacts datable to later periods including a very small group of organic tempered material (11) including fragments from loomweights which have been assigned a Middle Saxon date as discussed in the fabric descriptions above. As already noted this softish fabric may disintegrate easily and consequently only material that quickly reached a buried equilibrium survives.
Activity, as represented by excavated contexts, appears to cease in the Anglian period. The one later sherd identified, the base of a Medieval jug (3002.19), was in the destruction layer associated with a corn-drying oven. Other finds are suggestive of an Anglian date for this feature and it is not inconceivable that Medieval material became incorporated into the final destruction deposit. It is worth noting that the 1993 excavations were within the area of a limited distribution of Medieval pottery identified during fieldwalking and suggested as representing intermittent or seasonal activity (see above). At the other extreme the earliest activity is represented by part of a Bronze Age collared urn (1004.96) which was recorded as being within the upper fill of a major north-south ditch which is datable to the Anglian period; however, the sherd is more specifically noted as a surface find during clearance and should more correctly be included with the latter. Peter Didsbury notes a small Bronze Age component within the 1989 fieldwalking assemblage (1990, fig. 9) with this and the excavated sherd probably representing the destruction of Bronze Age features by ploughing. Certainly, barrows known and referenced in the SMR as within the vicinity of the Cottam B site have become all but invisible today.
The excavation assemblage is summarised in table 1 which show fabric types by context grouped by phase. Although limited it can be seen that the pottery assemblage supports this phasing. The bulk of the material is late Roman in date; calcite gritted (7), Crambeck grey (8) and, to a lesser degree, East Yorkshire grey wares (9); however, as noted, this mostly consisted of small abraded sherds that should be seen as residual. At best re-deposition could have occurred very late within the Roman period but is perhaps more likely to be post-Roman which is confirmed in two cases; 1003, the backfill of a large circular straight-sided pit, and 1004, upper fill of a recut of a major north-south ditch, both of which contained material in an tempered fabric (11) assigned to the Middle Saxon period. With the primary fill of the initial cut of the same ditch containing several sherds of abraded late Roman pottery it suggests a use-span for this feature restricted to the earlier parts of the Anglian period. Similarly in the other contexts containing pottery the evidence that supports an Anglian date is essentially negative in that the latest fabrics present (with one exception as already noted) are sherds of abraded late Roman pottery although some can be dated more precisely by the presence of other Anglian material or stratigraphically (ibid). It is of interest that organic tempered material was only found in the later of the two Anglian phases, IIB.
An area 20 * 50 metres was opened in 1995 to the north of the 1993 excavations on the edge of a concentration of metal detected 10th century finds and within an area of Late Saxon pottery identified during fieldwalking (above). Geophysics had also confirmed a complex series of features in this area that were not visible in aerial photographs (Garner-Lahire, this vol.). A total of 65 sherds were recovered although 15 were associated with clearance or modern activity such as deep ploughing.
When examined by period (fig. 4) the assemblage is immediately striking in comparison to the overall assemblage from Cottam B (fig. 2) or to that from the 1993 excavations (fig. 3) in that the total dominance of Roman fabrics is clearly challenged and even physically surpassed in terms of weight by a group of mostly late Anglian material. The Roman sherds, as elsewhere at the Cottam B site, are mostly small and abraded and similarly representative of manuring spreads redeposited in to later features. The average weight of these late Roman sherds is 9.5 grams which, while still low, is far in excess of the average 5.5 from 1993; however one example from 1995 weighed just under 100 grams and when removed from any calculation gives an average weight of under 6 grams. The unusually heavy sherd was a pedestal base from a Crambeck grey ware beaker (4092.58) which showed evidence of burning and appeared to have been cut down for possible re-use as a lamp. It was recovered from one of a series of east-west ditches which have been assigned a Middle Saxon date, phase IIB.
The earliest identifiable Anglian material is represented by three sherds in a Maxey type fabric (10). That this fabric was not identified within the 1993 assemblage and that the latest material from the this area was in an organic tempered fabric (11) which has been assigned a Middle Saxon date it suggests that the Maxey type ware is later in date, probably later within the Middle Saxon period, especially in consideration of its association with mostly Late Saxon material (see seriation - tables 1-2). As discussed above the fabric of the Maxey type sherds is closest to Peter Addyman's group III, type G which has been assigned a date range within the later 8th - 9th centuries, Thus a shift in activity foci is suggested late within the Middle Saxon period. The Maxey type sherds were identified within a quarry pit (4091) also containing Late Saxon Torksey type ware and within one of a number of associated enclosure ditches and gullies some of which contained Late Saxon material; however, the bulk of the Saxon material from the 1995 excavations is late suggesting activity was concentrated within the 10-11th centuries including pottery recovered from features representing structural activity (ie 4279). The Late Saxon material, representing nearly 50% of the excavation assemblage, was mostly in a Torksey type fabric (12) but also included a small number of sherds in a fabric similar to York ware (13) and a number of sherds, fabric C (14), which are almost certainly Late Saxon as they fall within the tradition of Late Saxon sandy wares and were recovered from a post-hole (4279) containing York Ware. In short, the pottery assemblage supports other data (this volume) in assigning a generally Late Saxon date, but probably starting late within the Middle Saxon period, to settlement activity identified during the 1995 excavation.
Settlement activity also appears to cease within the Late Saxon period. The latest material within layers interpreted as representing general silting and in part sealing excavated features is Late Saxon apart from one (intrusive?) post-Medieval sherd (Table 2 - 4002, 4004, 4116 and 4178) which suggests another shift in settlement focus at this time. Thus the pottery assemblage helps to categorise a discrete settlement episode starting late within the Middle Saxon period and appearing to end within the Late Saxon period. The latest recognizable forms with in the Late Saxon assemblage were consistently 10th century as represented by Torksey type rims.
Fieldwalking was on a large enough scale to allow the identification of distinct patterning in the distribution of the assemblage by period particularly in relation to the overall research aims of the investigations being undertaken at Cottam. Generally this appears to represent changing land use but discrete spreads of both Late Saxon and Medieval pottery where noted. The former, in the extreme east of the fieldwalked area, correlated with 10th century metalwork finds and is suggestive of an activity focus as subsequently confirmed by excavation (see below - COT95E). The latter, to the south and west of the earlier Late Saxon material, is suggestive of a restricted area of occasional activity with any significant Medieval foci elsewhere.
Despite its smallness the pottery assemblage recovered during the 1993 excavations from within an area of cropmarks and and within the area of a concentation of 8th and 9th century metal detected finds is supportive that excavated features largely relate to activity in this period. The evidence is largely negative in that the bulk of the assemblage consisted of small abraded sherds of late Roman date. These are interpreted as having been re-deposited into later features from late Roman manuring spreads identified during fieldwalking; however, (with one exception noted above) the latest material in the assemblage is in an organic tempered fabric assigned a Middle Saxon date. Thus activity appears bracketted between post-Roman and the latter period with a lack of early Anglian material emphasising the latter.
The 1995 excavations confirmed that material recovered during fieldwalking in 1993 represented underlying settlement activity which, from the excavated assemblage, can be reasonably suggested as starting at some time within the 9th and lasting at least into the 10th century but probably ceasing therein or soon after as indicated by the lack of sherds representing later fabrics and forms. Whilst the excavated and fieldwalked assemblages are small (67 sherds in total) they are not only significant in helping to identify a Late Saxon or Viking period focus of rural settlement activity but also in providing an important rural comparative for assemblages from this period in urban contexts such as York and more generally in consideration of town - hinterland relationships at this time.
Table 3 - 1993 fieldwalking assemblage
Table 4 - 1993 excavation assemblage
Table 5 - 1995 excavation assemblage
Addyman, P. & Whitwell, J. 1970. ‘Some Middle Saxon pottery types in Lincolnshire', Antiq J 50: 96-102
Andrews, D. & Milne, G. (eds). 1979. Wharram: A Study of Settlement on the Yorkshire Wolds, vol 1: Domestic Settlement, 1, Soc Medieval Archaeol Monograph Ser 8
Blake, H. & Davey, P. 1983. Guidelines for the Processing and Publication of Medieval Pottery from Excavations, DAMHB Occasional Paper 5 (HMSO London)
Brewster, T. 1958. ‘Staxton ware: an interim report', Yorkshire Archaeol J 39: 445-6
Buckland, P. et al. 1990. ‘The petrology of the Iron Age pottery', in Wrathmell & Nicholson 1990: 131-4
Corder, P. 1934. ‘The Roman pottery at Throlam, Holme-on-Spalding Moor, East Yorks.', Trans East Riding Antiq Soc 27: 6-35
Corder, P. 1950. ‘The pottery made at the kilns', in Hayes & Whitley 1950: (pages)
Corder, P. 1989. ‘The Roman pottery at Crambeck, Castle Howard', in Wilson 1989: 3-24
Crossley, D. 1990. Post-Medieval Archaeology in Britain, Leicester University Press
Darling, M. & Davies, B. 1993. ‘The Roman pottery from Cottam - COT93', unpub. notes
Didsbury, P. 1990. ‘Fieldwork in Cottam and Cowlam parish', Yorkshire Archaeol J 62: 63-7
Evans, J. 1989. ‘Crambeck: the development of a major northern pottery industry', in Wilson 1989: 43-90
Gibson, A. 1986. Neolithic and Early Bronze Age Pottery, Shire (Aylesbury)
Gibson, A. & Woods, A. 1990. Prehistoric Pottery for the Archaeologist, Leicester University Press
Green, S. 1992. ‘The pottery', in Milne & Richards 1992: 26-39
Hayes, R. 1988. ‘Roman Norton. excavations and discoveries', in Wilson 1988: 66-89
Hayes, R. & Whitley, E. 1950. The Roman Pottery at Norton, East Yorkshire, Roman Malton & District Report 7
Hayfield, C. 1985. Humberside Medieval Pottery, Parts 1 & 2, BAR bs 140 (Oxford)
Jennings, S. 1992. Medieval Pottery in the Yorkshire Museum, The Yorkshire Museum
Johns, C. 1971. Arretine and Samian Pottery, British Museum (London)
Le Patourel, H. 1979. ‘Medieval pottery', in Andrews & Milne 1979: 74-107
Mainman, A. 1990. Anglo-Scandinavian Pottery from 16-22 Coppergate, Archaeology of York 16/5 CBA (London)
Mainman, A. 1993. Pottery from 46-54 Fishergate, Archaeology of York 16/6 CBA (York)
McCarthy, M. & Brooks, C. 1989. Medieval Pottery in Britain AD 900-1600, Leicester University Press
Milne, G. & Richards, J. 1992. Wharram: A Study of Settlement on the Yorkshire Wolds, VII. Two Anglo-Saxon Buildings and Associated Finds, York University Archaeological Publications 9
Monaghan, J. 1993. Roman Pottery from the Fortress, Archaeology of York 16/7 CBA (York)
Orton, C., Tyers, P. & Vince, A. 1993. Pottery in Archaeology, Cambridge University Press (Cambridge)
Sumpter, A. 1990. ‘Iron Age pottery', in Wrathmell & Nicholson 1990: 128-30
Swan, V. 1988 ed. Pottery in Roman Britain, Shire (Aylesbury)
Vyner, B. 1993. ‘Fieldwalking at Cottam, 1993', unpub. notes (University of York)
Wilson, P. (ed.) 1988. Yorkshire Studies: Archaeological Papers. Yorkshire Archaeological Society
Wilson, P. (ed.) 1989. Crambeck Roman Pottery Industry, Roman Antiquities Section Yorkshire Archaeological Society
Wrathmell, S. & Nicholson, A. (eds) 1990. Dalton Parlours: Iron Age Settlement and Roman Villa, Yorkshire Archaeology 3 (West Yorkshire Archaeological Service)