4.3 Faunal Remains Assessment

by David Berg *

4.3.1 Background

Analysis of the large quantity of animal bone recovered from excavations at West Heslerton Anglo-Saxon settlement was considered and agreed during the excavation phase of the project and, consequently, recording was begun by Trevor Ashwin at the project headquarters in June 1987 and continued until his departure in the spring of 1989. At that time, the West Yorkshire Archaeology Service was invited to prepare an assessment of the animal bone and a project design for the completion of an analysis, which was submitted in May 1989, and subsequently took over the identification and recording in February 1990. What follows is an updated assessment of the quantity of animal bone recovered and a re-evaluation and reiteration of the potential for analysis.

4.3.2 The assemblage

4.3.2.1 Quantity of bone

Between February 1990 and February 1993 work has taken place recording animal bone recovered from the 1987 and 1988 seasons (previously recorded by Trevor Ashwin), the backlog material from the summer of 1989 and material recovered from the excavation seasons 1990, 1991 and 1992.

The number of animal bones for each area of Sites 2 and 11 is shown in Table 4.3. A total of 236,873 fragments have been recorded on computer databases with an additional 18,924 measurements and 4,395 records of mandibles or single teeth attributable to an age stage. A further 9,800 individual measurements on paper records remain to be added to the computer metrical database. The data-set includes a large body of fragmentary and not readily identifiable material described as 'non-zone'. These offer important research potential with a bearing on recovery and excavation methodology as well as studies of land-use, deposition and disposal practices, and require only minimal data entry of context and fragment type to enable the full spatial analysis to proceed.

The methodology adopted for recording animal bone has divided the assemblage into element zones and non-zones. The element zones archive is complete, except for a small quantity of material still on site. However, the larger part of the non-zone assemblage remains to be added to the computer database (see also 4.3.4.6 Recording of remaining non-zones).

Each bone fragment - except for bones lifted at context level only - has been allocated a three-dimensional tag number and, after washing, was bagged individually and boxed by area code. This has been a disadvantage in that most bones could only be assessed as individual fragments and not as context groups. The number of tagged fragments comprises 78% of recorded element zones (see also 2.6.8 Sampling and 2.6.9 Recording systems).

4.3.2.2 Provenance of bone

Animal bone was recovered from all seven areas of Site 2 and all thirteen areas of Site 11 (Tables 4.3 and 4.4). From the data currently available, 31.2% of the 52,831 mammal zones can be assigned to a period, and of these 92% are Saxon, the majority Early Saxon (Table 4.4). It is assumed that a large proportion of the contexts still to be phased will be of Saxon date.

Sixty-seven percent of mammal zones derived from contexts for which feature information was available at the time of writing (Figure 47). Within this group of 33,3049 fragments, the majority of bone comes from four feature types: Grubenhaüs fills (32.4%), spits (28.7%), layers (14.3%) and pit fills (7.4%) but these figures are likely to change when final period information allows the exclusion of bone from contexts that pre- or post-date the Anglo-Saxon settlement phase.

Further information is required before the amount of bone contained in contaminated contexts can be quantified. Some deposits do contain intrusive material, i.e. bones of rabbit and brown rat, and some exceptionally large cattle and sheep bones have been measured that may be Anglo-Saxon but post-medieval or prehistoric dates cannot be discounted at this stage (see also 3.8 Environmental Data).

4.3.2.3 Composition of bone

All the main domestic species are represented in the assemblage including cattle, sheep, horse, pig, goat, dog and cat. These species form the vast majority of the assemblage, with comparatively few wild species that include ?wild pig, red deer, roe deer, fox, badger, hare, rabbit, black rat, brown rat, house mouse, wood mouse, water vole, field vole, mole, hedgehog, stoat, weasel and whale. In addition, many wild/domestic bird bones and a small number of fish bones were recovered.

The bulk of the bone sample was retrieved by hand-trowelling and recovery appears to have been excellent compared to many handpicked bone assemblages; this may be due to the practice of three-dimensionally recording all artefacts and ecofacts. An analysis was conducted, at an earlier stage of the recording, on the size distribution of fragments recovered by hand. A random 10% sample of non-zone fragments, totalling 16,027 bones, revealed the following statistics:

Despite good recovery of small bone fragments by hand-trowelling, it is inevitable that some very small fragments, such as ovicaprid incisors, small bird and fish bones, will have been missed. This can be tested, to a degree, on the bones recovered from the programme of wet sieving and flotation conducted on site. Animal bone has been collected from a large number of bulk soil samples floated for botanical remains. Also, the contents or part contents of several Grubenhäuser were dry sieved for small bone through a 5mm mesh or wet sieved and floated.

Flot samples from a few selected areas have been examined and compared to hand-recovered bones. The recovery of two small elements, ovicaprid third phalanges and lower third molar teeth (LM3s), was compared. The results are reproduced in Table 4.6. Combined data from the six areas examined show 10.8% of all phalanges were recovered from flots and 1.3% of all LM3s, although the figures show a marked degree of variation. For example, in Area 2DC 27.9% of third phalanges was recovered in flots but none of the 15 molar teeth.

4.3.2.4 Condition of bone

The site consists of chalk gravels and mixed sand and gravel deposits. Bone from surface layers, in particular, shows a great variety of preservation from good to heavily eroded and weathered. The poorest preservation occurs on Areas 2DA-2DD and parts of 2CA and 11CE. The bone from areas 11AB, 11BD and 11CD display the best preservation. Very little bone was recovered from the large Area 11BB due to severe plough damage. The machine stripping of topsoil has resulted in many freshly broken fragments, ranging across the site from 30.1% to 59.5% of identified element zones (Figure 4.1).

Fig 4.1 Element zones (%) by excavation area

Preservation conditions have biased the assemblage in favour of the more robust and solid parts of the skeleton. In Figure 4.2 the frequency of over 22,000 cattle zones is plotted against relative bulk density for each element. The bulk density figures were compiled by Lyman (1982) from the bones of deer, Odocileus hemionus, and are related to the potential of a bone to survive taphonomic destruction. While there are obvious problems comparing these figures to the skeleton of domestic Bos, the scatterplot clearly demonstrates the greater survival of denser bones and indicates the parts of the skeleton likely to be under-represented.

Fig. 4.2 Scatterplot of cattle element zone frequency plotted against Lyman's (1982) bulk density values

In addition to below ground taphonomic destruction the gnawing and chewing of bones by animals such as pigs, dogs, cats and rodents have also biased the survival and condition of certain elements. Figure 4.3 plots the percentage of cattle and ovicaprid bone gnawed and chewed for each area. Cattle bones always show a higher percentage of damage, possibly because they survive total destruction better than ovicaprid bones. It is interesting that the greatest degree of gnawing and chewing occurs in areas 2CB and 2CC - the 'housing' area of the site.

At this stage of the analysis of such a large assemblage of bone it is difficult to quantify the true extent of information loss due to differential preservation or recovery biases or to distinguish either effect from differential deposition, which, from the patterns of spatial distribution of elements, was clearly taking place. Taphonomic biases may be better understood as bone from individual contexts and features are analysed in more detail. Nevertheless, the majority of material that has survived is in good condition: bone is solid and retains details of even the finest butchery knife marks.

Fig. 4.3 Percentage of cattle ovicaprid gnawed/chewed

4.3.2.5 Methodology

The bone assemblage recovered during excavations has been sampled to reduce the time required for analysis and therefore the costs to the project. The sample for detailed analysis is acquired by selecting a predetermined range of skeletal element zones. This will reduce the quantity of bones recorded in detail from almost half a million to c.53,000.

The zones chosen for analysis are those that hold the maximum amount of information for species identification, age, size, sex, pathology, and butchery. The numbers of non-zones, such as fragments of skull, limb bone splinters and unidentified pieces, are stored on a computer database with attributes such as context, tag number, fragment size, butchery, carnivore damage, etc.

It is considered that very little significant information, for the aims of reconstructing the site economy, has been missed as a result of this sub-sampling. For example, the contents of a Saxon pit 2DB 008 produced 1,205 zones and 7,510 non-zones. The number of fragments with butchery evidence in the assemblage of zones amounted to 28.4%. The bones with butchery evidence in the non-zone assemblage totalled only 0.7%.

The analysis of the animal bone data will be undertaken using a variety of analytical computer programmes written by the author running on commercial software such as dBase IV 1.5 (database), Quattro Pro for Windows (spreadsheet), Minitab (statistics) and the Bonn Archaeological Statistics Package (multivariate statistics) as well as G-Sys (geographical data management software).

The analysis and presentation of the descriptive data will follow the most widely used and excepted methodologies; for example, measurements follow the guidelines of von den Driesch (1976), tooth eruption and wear follow the procedures of Grant (1982), Halstead (1985) and Payne (1973, 1987).

4.3.3 Statement of potential

The importance of the West Heslerton bone assemblage has been discussed in the Project Design of May 1989 and in the Interim Report of March 1991. Several excavations over recent decades have provided valuable information leading to a better understanding of Middle Saxon urban development but little quality data have been added to our sparse knowledge of Early Saxon rural economy and society. The period AD 350-700 in Britain is poorly understood and is, to all intents and purposes, prehistoric (Addyman 1979). The archaeology of the period is the only means of understanding the processes of change from Late Roman to Early Anglo-Saxon rural society, Early Anglo-Saxon rural economy and the development of Middle Saxon urban centres and relationships with their hinterland, which is no doubt why these matters have been flagged as a national research priority by English Heritage (1991a).

The West Heslerton excavations offer an extremely rare opportunity to study a large bone assemblage from a rural Early Anglo-Saxon settlement and cemetery. The potential of the data is unequalled both regionally and nationally. The assemblage is well preserved, unlike the animal bone from Early Anglo-Saxon Mucking (Done 1993) and has been recovered from an almost completely excavated settlement, unlike the Early Anglo-Saxon bone from West Stow (Crabtree 1989). The extent of the excavations and the three-dimensional recording of all finds using geographical software makes the potential for intra-site spatial analysis unique.

Because so little is known of the Early Anglo-Saxon economy in Britain the importance of this assemblage is not limited to a purely local enhancement of the archaeology of the Vale of Pickering or the Yorkshire Wolds The data also have an intrinsic value in understanding the transition from Late Roman society and the background to the development of Middle and Late Saxon urban centres at a regional and national level.

In brief, the assemblage has the potential to provide the maximum level of information for many of the analysis priorities set out below because:

• it is generally well preserved
• it is large, and therefore has a high degree of statistical validity
• it is well stratified, the vast majority of bone belonging to the Early and Middle Saxon periods
• it includes many primary, undisturbed, deposits

Descriptive data and statistics

A priority of the analysis will be to separate and distinguish descriptive and interpretative data (O'Connor 1985) in the final report. While this has been a long-standing request of archaeozoologists (cf. Clason 1972), published reports are not always presented in this manner. A separation of descriptive from interpretative data will enable examination and reassessment by other specialists and clarify the basis and justification of interpretations.

4.3.3.1 Site economy

A reconstruction of the type and level of animal husbandry practised at West Heslerton is a major aim of the post-excavation project. To this end an analysis of each domestic species based on: relative proportions, age and sex distributions, size, patterns of butchery and carcass division, and evidence for disease and injury will be undertaken (see also 6.5.4 Agriculture, economy, production and exchange and 7.1.4.2 Reconstruct the environment and economy of the settlement.)

4.3.3.2 Major and minor species

What species were being utilised? In what proportions were they present on site? What would their relative numbers have contributed in meat weight? What was the role of ducks and goats? - two species that are rare on many Anglo-Saxon sites. Are wild resources, such as deer, hare, and wild pig, of minor importance in the rural economy, as is suggested by their relatively low numbers on many sites of this period? How important was the horse in Anglo-Saxon society or domestic 'pets', such as cats and dogs? What can the range of small mammals recovered reveal about environmental conditions on and around the settlement?

4.3.3.3 Birds

Almost 3,000 bird element zones have been recorded. What was the role of domestic and wild birds in the economy? What age and sex profiles can be deduced from the domestic bird bones? What size were Anglo-Saxon domestic fowl and geese?

4.3.3.4 Age and sex of stock species

Ageing data is available from some 4,395 mandibles and loose lower third molars or deciduous fourth premolars in addition to longbone epiphysial fusion. The construction of age profiles for the major domestic species and the calculation of age groups - for comparison with other Saxon ageing data such as Bourdillon and Coy (1980) and Crabtree (1989) - will enable the following points to be addressed: can age at death and sex profiles indicate economic regimes, e.g. were sheep important as providers of prime meat or wool, were cattle kept as part of a dairy economy, as traction animals or primarily as providers of prime meat? Was meat a regular part of the diet? Are specific milk, meat or wool regimes applicable to Early Anglo-Saxon stock animals? Is there evidence for trade in livestock, i.e. an absence of selected age groups? (See also 7.2.1.1 To explore the evidence for agriculture and diet.)

4.3.3.5 Metrical analysis

The large number of bone measurements (28,724) will significantly expand the existing data on Early and Middle Saxon domestic animals. Trends that can be tested include changes in size from Roman to the Early and Middle Saxon periods, the size of domestic Saxon dogs, a species displaying considerable size variation during the Roman period, and the introduction of different breeds comparable with continental data. Measurements will be used to examine clusters or groups that may be due to sexual dimorphism, different age groups or different breeds. The determination of sex by metrical analysis will need to be used.

4.3.3.6 Butchery

It is acknowledged that the quantification of butchery marks has inherent problems as not all dismemberment will leave visible evidence on the bone; nevertheless, there may be patterns to suggest repeated systematic methods of animal slaughter and butchery, which may vary in different parts of the site or over time. The distribution of elements may indicate concentrations of primary 'butcher's waste' or 'quality' meat bones. Horse bones outnumber those of pig and a large proportion are butchered in a similar fashion to cattle bones; the butchery evidence for this species therefore requires careful analysis to establish if dismemberment was to facilitate easier disposal or if horse flesh was a regular part of the diet (see also 7.2.1.1 To explore the evidence for agriculture and diet).

4.3.3.7 Pathology

The number of pathological bones identified in the initial recording totals 1,324. This is expected to give a good indication of the general health of domestic species, the standard of animal husbandry and the frequency of injury and disease such as congenital defects, task-related arthropathies and traumatic injuries.

4.3.3.8 Animal burials

One hundred and forty-three 'skeleton' numbers have been allocated in the WYAS bone laboratory to partial or complete skeletons found in articulation. The material ranges from articulated feet to complete bovine burials. Animal sacrifices are well recorded during the Iron Age and Roman periods. Is there evidence that animals were slaughtered for reasons other than economic, i.e. religious or ritual, during the Saxon period?

4.3.3.9 Intra-site temporal variation

It is suggested that during the Middle Saxon occupation the settlement contracted to the southern part of the site. If the bone assemblages from both phases of Saxon occupation can be confidently separated, a priority would be the examination of changes or continuities in the site economy. For example, pig bones do not form a significant part of the total assemblage, yet they are widely held to have been of major importance in the Anglo-Saxon economy. Does their role change between the Early and Middle Saxon periods? Is there evidence at West Heslerton to support an increasing reliance on wool production from the Early to Middle Saxon period? (See also 2.6.1 Chronological summary of the excavation results and 7.1.4.2 Reconstruct the environment and economy of the settlement.)

4.3.3.10 Intra-site spatial variation

Intra-site variation in the bone assemblage is particularly important, and this part of the analysis should concentrate on assisting in the interpretation of site function, i.e. areas of housing, industrial/craft activity, agricultural processing and slaughter or meat processing. Patterns of bone discard, specifically from negative features containing primary dumps such as the Grubenhäuser, may suggest divisions in activity or status.

A superficial analysis of spatial variation, based on the arbitrary area site divisions, has been completed already and crude patterns have been observed. The results indicate that cattle form over 50% and horse over 5% of identified zones in the north-west of the site; Site 2 Areas DA, DB, DC and CA. Sheep zones outnumber those of cattle only in the south-west and southern parts of the site; Areas 11AA, 11AE, 11BA, 11BB, 11BD, 11CE and 11CD. The central eastern areas, 11BB and 11BC, produced over 10% pig zones, whilst only in the north-east section, Areas 2CB and 2CC, do dog and cat zones represent more than 5% of the major species assemblage.

Fig. 4.4 Recovery of bird and small mammal bones

Variations in assemblage composition across the site can also be seen in the distribution of bird and small mammal bones, which show no correlation (Fig. 4.4). Areas 11BA, 11BD and 11CD have the highest number of bird bones, ranging between 12-18% of all bird bones, but have low numbers of small mammal bones, 4-8% of all small mammal bones. The highest small mammal retrieval rate, between 17-19%, was from Areas 11AB, 2CA and 2CB - whose juxtaposition to the relict stream channel may have some significance - but these produced relatively low numbers of bird remains, 3-4%.

At this stage of the analysis of such a large assemblage of bone, it is difficult to quantify the true extent of information loss due to differential preservation or recovery biases, or to distinguish either effect from differential deposition which, from the patterns of spatial distribution of elements, was clearly taking place. Taphonomic biases may be better understood as bone from individual contexts and features are studied in more detail and with the aid of geographical software to analyse bone distributions and associated finds. Examination of preservation differences across the site and the survival of small mammal, bird and fish bone may assist in separating the effects of soil type, human activity and site function. The higher recovery of sheep and bird bones in the southern part of the site may be due to site function. However, any conclusions on the character of the bone assemblage in this part of the site are obviously weakened by the partial excavation of this area (see also 3.2 Methodology and 6.5.6 Settlement morphology and planning).

4.3.3.11 Inter-site comparisons

The evidence from West Heslerton should not be looked at in isolation but should be compared and contrasted with other data from rural and urban sites. The rarity of large rural Saxon faunal assemblages means that data from a variety of site types and geographical locations must be used. Comparisons with earlier, contemporary and later sites are all valid, especially the use of Late Roman data to throw light on the continuation of agrarian economy, the question posted by Fineberg (1972), i.e. Saxon evolution or revolution? Two examples of the sort of questions the intra-site comparison aims to address are:

• The large proportion of mature sheep at sites such as Portchester (Grant 1976), Hamwic (Bourdillon and Coy 1980) and North Elmham Park (Noddle 1980) suggests that wool production was more important towards the latter part of the Anglo-Saxon period. However, the data from West Stow (Crabtree 1989) suggest a higher proportion of juvenile sheep at this site. Is there evidence at West Heslerton (from 4.3.3.10 above) to suggest that the West Stow results may be a function of intra-site spatial variation rather than reflecting regional variation?
• Differences in butchery techniques have been identified between urban and rural settlements during the Romano-British period in southern Britain (Maltby 1989). Are Anglo-Saxon butchery techniques the same as those used by the Romans? Do differences in methods of butchery exist between Anglo-Saxon rural and urban settlements?

The West Heslerton results will be compared to bone data from the following sites as a starting point: West Stow, Suffolk (Crabtree 1982, 1985, 1989); North Elmham Park, Norfolk (Noddle 1980); Hamwic, Southampton (Bourdillon 1986a, 1986b, 1987; Bourdillon and Coy 1980); Portchester Castle (Grant 1976); Garton and Wetwang Slacks (Noddle 1981) and Mucking (Done 1993). Urban assemblages from the York sites of Coppergate and Fishergate will also be used and it is hoped that access will be available to material that is unpublished or currently under analysis such as Flixborough, Humberside, and Wharram Percy and Thwing, North Yorkshire (see also 7.2.1.3 To explore the transition from Roman to Saxon and that from Early to Middle Saxon, both periods of large-scale settlement shift and re-organisation, in the context of larger re-organisation of the landscape).

4.3.3.12 External Research

It is expected that visits to other organisations will be necessary, such as local universities and the British Museum (NH), for use of library facilities and for confirmation of the identification of some specimens.

4.3.3.13 Archive and publication report

The archive will consist of all primary recording sheets, computerised data and a full description of sampling and analysis methodology. The publication report will consist of text, tables, figures and plates. All will be supplied on computer disk as software-dependent files and/or ASCII files and as hard copy.

4.3.3.14 Interdisciplinary co-operation

It is important that any interpretations of the bone assemblage are not made in isolation from other specialist's results and full co-operation and exchange of data and interpretations should be encouraged in order to maintain and achieve to the full the major project aims.

4.3.3.15 Software training and application

Initial training in the use of the project's Geographic Data Management System software will be necessary to explore bone distribution patterns and fully exploit and evaluate the three-dimensional tagging of bone fragments.

4.3.3.16 Photography

It may be desirable to photograph some specimens, particularly pathologies. Photography and processing will be carried out in-house by a WYAS photographer.

4.3.4 Wider research questions

It has been emphasised in interim reports by the Project Director that the project has been funded as a rescue excavation and not as a research project. The faunal assemblage has enormous scope for research, not only into aspects of Anglo-Saxon economy but also as a pure biological data bank. Many subjects will be outside the scope of the archive and publication reports and must be left to future grant-aided research. Nevertheless, it is hoped that time will be available to study some aspects of greater relevance to the Heslerton report.

4.3.4.1 Spatial analysis using geographical software

The ability to plot graphically in three dimensions the spatial distribution of animal bone, with all other finds and environmental data, is extremely rare. The facility to plot bone by species, element type, age group, etc. may have distinct advantages in understanding site formation processes, domestic activities or new aspects of animal husbandry. Interpretations of site function from settlement patterns of discard could be related to interpretations of social organisation using cemetery data. The aim of this project in spatial analysis would be to develop predictive models of Anglo-Saxon settlement and social organisation which could be tested on existing data from other sites and applied to future Anglo-Saxon project designs in the region and nationally. The tagging system requires thorough consideration and evaluation, which, in turn, will require a degree of exploratory research on the GIS software and data.

4.3.4.2 Skeleton analysis

A total of 143 articulated skeletons and limbs has been identified in the assemblage. Unlike human remains which, unless severely disturbed, are usually recognisable as discrete individuals, bone from domestic food animals cannot be related to an individual. Context information on all of these burials is not available at the time of writing but it is expected that some, if not most, will be from primary deposits and will therefore offer a rare opportunity for research, albeit not directly related to the reconstruction of the Anglo-Saxon economy. For example:

• to compare patterns of dental attrition with epiphysial fusion evidence in the same individual
• to examine current methodologies for estimating stature from different limb bone measurements
• to examine morphological and metrical methods of sex determination from different elements in the same individual
• to examine how pathologies, usually identified and interpreted from individual bones in isolation, might spread to adjacent elements or to different locations in the skeleton.

4.3.4.3 Control sampling

The number of zones recorded amounts to an average of 13% of the total bone in any context. This may seem low but the identified percentage is comparable to many published assemblages. For example, animal bone from phase 2 of the West Stow settlement (Crabtree 1989, table 5) produced 50,366 total fragments, 47.3% of which are unidentified and a further 24.6% are classed as large or small mammal and excluded from further analysis. This leaves 28.1% identified to species and element. However, a large proportion of identified bone comprises elements not recorded in the Heslerton assemblage, either because of the bias they can introduce into species proportions or because they lack useful information, e.g. small fragments of skull and maxillae, all loose teeth, carpals, long-bone shaft fragments, etc. Such element fragments account for over 72% of the West Stow identified cattle and, if excluded, reduce the identified proportion of this species from 9.5% to 2.6%. However, limiting identification to selected element zones is an aspect of the methodology that may be open to academic criticism and therefore to establish the type and degree of information lost by this method would be of considerable importance.

It is suggested that the total recovered bone from at least two contexts should be identified and recorded comprehensively as a control to predict statistically the loss of information across the whole site. One such context is a Saxon pit from Area 2DB context 0008. A total of 1,205 zones has been recorded and 7,510 non-zones have been quantified. Combined, the pit accounts for 16.6% of the bone from Area 2DB. It is recommended that the 7,510 non-zones be re-examined and identified to species and element. At least one other context should be treated similarly, preferably from a different part of the site and after consultation with the Project Director. This task, including sorting the bone, identification, the recording of measurements, butchery, computerisation and analysis of the results is estimated to take a faunal analyst five weeks for each sample.

4.3.4.4 Archive completion

At the time that the animal bone post-excavation project was halted the following archive work remained to be completed:

4.3.4.5 Metrical database

All bone measurements have been recorded on paper records and, including work completed after this report was first submitted, approximately 7,600 measurements remain to be computerised.

4.3.4.6 Recording of remaining non-zones

This part of the assemblage has had an agreed low priority and has been recorded on computer by an assistant as time allowed. Since Table 4.3 was constructed, further work has been completed and at present 222,190 non-zone fragments contained in 99 boxes have been recorded. A further 98 boxes remain to be computerised, estimated to contain in the region of 219,000 fragments. This clearly represents the bulk of further archive work. It is, however, important to complete this aspect of the post-excavation.

Apart from the time, effort and cost already invested, the recording of the non-zone fragments is the only method of quantifying the sample size of the identified zones (because tagged fragments are individually bagged and have not been stored by context group). The total quantity and distribution of bone across different areas of the site can only be calculated and examined after all non-zones are recorded on computer database. Additionally, the archive would be incomplete without the recording of the remaining bone.

4.3.5 Storage and curation

All bone fragments are contained in their original bags, retaining context and tag number information. A total of 522 boxes is known at present, requiring a storage area of 19.8 cubic metres. No discard policy has been discussed with the Project Manager or a case could be made for the discard of post-medieval, modern and contaminated material, but this is likely to form only a minor part of the total bone volume.

All bone has been received boxed by site area. During the identification process some bone has been stored by species and element, combining material from several areas although provenance for each fragment has been retained. The organisation of the material for archive deposition is a decision for the recipient museum and English Heritage. Several options are available:

• to store the bone by site area. To box the bone by area would take approximately six days to complete
• to sort the identified bone into context groups. This would be a considerable sorting task and is estimated to take 40 person-days, to include collating individually bagged fragments into new context bags, re-labelling and numbering boxes and producing a computer archive listing of the contexts contained in each box.
• to sort the total assemblage into context groups. This would be as the second option but because the non-zone fragments tend to be far smaller, many more bags are contained in each non-zone box giving an estimated completion time of 90 person-days.

4.3.6 Appendix to assessment of February 1994: Additional animal bone from the 1995 season

4.3.6.1 Background

The fieldwork undertaken during 1995 produced 213 boxes of stratified animal bone from the southern extension of the site, an area 'substantially richer and denser in features that those previously excavated and including a major Late Roman complex' (D. Powlesland, pers. comm.). The quantity of bone boxes represents an increase of 41% over the 522 boxes already received and forms 29% of the whole faunal assemblage, by volume.

4.3.6.2 Assessment of potential for analysis

This assessment is based on one site visit during the 1995 excavation season and subsequent discussions with the Project Director, Dominic Powlesland. While a laboratory assessment of the assemblage would have been recommended to justify the exclusion of small, unstratified, or badly preserved assemblages of bone from further detailed analysis, the quantity and quality of the 1995 assemblage, and the similarity in composition and deposition to that recovered in previous seasons, suggested this was not the most advantageous use of time or financial resources.

The 1995 assemblage is seen not in isolation but as a continuation and addition to stratified material already recovered and partly analysed. A detailed assessment of the quality and potential for further analysis would have been expensive and resulted in repetition of the case already discussed at some length in the original Project Design for Faunal Analysis of May 1989, in the Interim Report of March 1991, and in the Assessment of Potential for Analysis of March 1993 and February 1994, parts of which were reproduced in Dominic Powlesland's full assessment report of November 1994. Further, given the current availability of detailed phasing data, no appraisal of Late Roman, post-Roman or phased Saxon material would have been possible at this stage.

In brief, the 1995 assemblage has the potential to provide the maximum level of information for many of the analysis priorities set out in the documents above because:

• it is large, and therefore has a high degree of statistical validity
• despite a larger number of fresh breaks than expected, due to the exceptionally hot and dry weather conditions, it is generally well-preserved and stable
• it is well stratified, containing bone from Late Roman, early post-Roman and phased Saxon occupation.
• it includes many primary, undisturbed deposits.

4.3.6.3 Importance of analysis

This new material will enhance the results of all the analysis and research questions presented in Sections 3 and 4 of the 1994 assessment report by providing a larger corpus of information. The total bone assemblage can be examined and interpreted almost as a population rather than a sample, a rare opportunity for any finds analysis. A priority in the analysis of the 1995 material would be the potential for more detailed and meaningful interpretations of intra-site temporal and spatial variation. Spatial patterns have already emerged from the bone examined to date, with cattle and horse more dominant in the north-west of the site (?industrial area), dogs and cats most frequent in the north-east (?housing area) and sheep increasing in importance the farther south the deposits. These variations are not solely the result of preservational biases but must reflect some element of site function, either social or economic. The interpretation of animal bone - and other environmental and artefact remains - from contemporary features found on other parts of the settlement, such as Middle Saxon Grubenhäuser, would be one of the most important aspects of the analysis.

Another major objective of the faunal analysis is a temporal comparison of bone characteristics across the site and here the southern extension is perhaps the most important, with apparent evidence for continuation of occupation. Land-use, economy and social organisation of Late and post-Roman rural society, the evolution of Saxon occupation and the growth of the settlement will require access to a full and detailed archive of 1995 material.

4.3.6.4 Recommendations

It is recommended that the 1995 faunal assemblage be recorded to archive level using existing and established methodologies thereby ensuring the integrity of the descriptive data and providing a complete archive for future consultation and reassessment. A proportion of element zones should be selected for interpretative analysis and the results incorporated into the full site report. The quantity and spatial origin of this material should be based on informed context and phasing information.

4.3.6.5 Archive implications

If all the faunal material from the southern extension were to be kept, which is the recommended option, an estimated 8m³ of further storage space will be required.

© Internet Archaeology URL: http://intarch.ac.uk/journal/issue5/westhes/4-3.htm
Last updated: Tue Dec 15 1998