The Work

Specialist Report 2005 - The Mammal Bone Assemblage (Methodology and Analysis of Bone by Context)

Posted by steven on 10/02/2006 at 04:07 PM

The report on the mammal bone that follows includes sections of text and images taken from Carrie Drew’s MA Dissertation submission to the University of Durham. Carrie, who has been completing the work over the last two years under supervision from Peter Rowley-Conwy, has provided excellent analysis and discussion regarding the bone assemblage recovered from Bone Passage between 2003 and 2004. Overall, this is an excellent piece of research that provides a lively debate focusing on the ‘unusual’ bone deposits recovered from the cave. The research and analysis in the report builds on Carrie’s interim findings that were published on this website in 2005 (see Specialist Report 2004 - Animal Bone, in ‘The Work’ section). Carrie will be continuing her analysis of the animal bone from the site in 2006, primarily on the assemblage of material from Trench 2 and Trench 6 (Bone Passage). Work for her PhD will focus on the pig-rich deposits including animal husbandry and diet in the Iron Age.

REFUSE OR RITUAL? - THE MAMMAL BONES FROM HIGH PASTURE CAVE, SKYE

Carrie Drew - University of Durham

Chapter 1: High Pasture Cave- A Brief Introduction

Section 1: Brief Summary

This dissertation examines an Iron Age faunal assemblage recovered from the High Pasture Cave complex on the Isle of Skye. The assemblage has been identified as having some very unusual features, which would profit from further investigation (Rowley-Conwy, 2003:36).

The site provides an unique opportunity to investigate faunal remains of a type with few parallels in archaeology, both in terms of the composition of the assemblage and its location. Archaeological deposits identified within active limestone caves are rare in Scotland (Birch et al., 2003:10) and in most cases where faunal remains have been discovered in caves evidence of human working or deliberate deposition has been lacking (Birch et al., 2003:11), in contrast to the clearly archaeological nature of this assemblage.

Section 2: The faunal assemblage at High Pasture Cave

The faunal remains examined represent those collected from High Pasture Cave between 2002 and 2004. They consist of collections totalling 9,473 mammal bone fragments gathered from the surface of High Pasture Cave (Context C001); the Trench 1 excavation (Contexts C002-C011), and Feature F.001 (see chapter 2 for details). The assemblage was recognised as being in excellent condition at the time of excavation (Birch, 2004: personal comment) and as far as possible every fragment was identified to the level of taxa and/or element. It is intended that the further evidence from 2005 and any future excavations will be incorporated with these findings at a later date. This study examines mammal bones from the assemblage, which is dated to the mid Iron Age, with fish, rodent and antler remains being examined separately (Ceron-Carrasco, McLaren and Hunter, 2004: Web reference 1). The range of animals represented comprises both wild and domestic species including elements of sheep/goat (Ovis/Capra), cow (Bos sp.), pig (Sus sp.), red deer (Cervus elaphus), wolf (Canis Lupus) and a medium-size dog (Canis sp.).

Section 3: Reason for investigation, and aims of this study

The bone remains were recognised as being of unusual composition at the time of the preliminary recovery of the disturbed deposits in 2002 (Birch, 2003:13) with the assemblage dominated by pig (estimated as high as representing 90% of the main context) to an extent that is unprecedented in any other Hebridean site, where the average in most assemblages is around 3-4% (Birch, 2003:22) and the largest concentrations are around 15%.  It was noted also that a complete range of skeletal elements appeared to exist, rather than the distribution of skeletal parts normally expected from a domestic assemblage and that many of the pig bones were unusually complete with little breakage for marrow (Birch, 2003:15). This suggested that the animals were deposited relatively intact in the cave and raised questions about the purpose of their deposition (Birch, 2003:15). The morphology of the cave suggested that the bones placement could not have been by natural water-borne deposition and that they must have entered either through deposit from the surface or deliberate placement within the cave. The evidence of contemporary human presence in the cave, such as the paved layer excavated in Trench 1 (between contexts C003 and C004) (Birch, 2005:3) seemed to confirm that deliberate placement was likely.

This study aims to provide an in-depth examination of this unusual assemblage to answer the questions posed within the recent Data Structure report for the excavation (Birch, 2005:11- see figure 1). In particular it will seek to clarify what the faunal remains tell us about the site, specifically whether the assemblage simply represents refuse deposit or whether it displays evidence of ritual deposition. It will also seek to determine how results from this site relate to other Scottish faunal assemblages and also our understanding of ritual deposition. Additionally, few organic remains have ever been recovered from excavations on Skye (Birch, 2005:12) and any information will add to the understanding of the prehistoric society and economy of the island.

The archaeological deposits in the cave presented several intriguing questions.  How was the cave utilised by the prehistoric inhabitants?  Were they merely using a hole in the ground as a place to dump unwanted midden material, a process that has continued to the present day at other cave entrances in the region; or could the abandoned, high-level passage have been used as a natural souterrain, where people stored their produce and game in relatively stable temperatures throughout the year?  Or could there be other interpretations for the use of the cave? Locations in the landscape such as springs, wells and deep natural shafts are known to have had a special significance with early Celtic societies, such liminal sites providing access to the ?Otherworld? or ?Underworld? (Armit, 2003: 108-111)?.

(Passage from High Pasture Cave Data Structure Report (Birch et al, 2005:11)

Section 4: Summary of the aims of this study

To quantify the assemblage producing a high-quality bone report and data report, considering butchery techniques, season of death and domestication, as well as quantifying the species present.

To consider the similarities between the trench contexts (Contexts 002-010) and the unusual deposits of Context 001.

To consider the faunal assemblage in the context of the rest of the High Pasture Cave excavation.

To evaluate the findings of the analysis in the wider context of their implications for Scotland, particularly where ?High Pasture? sits in the Scottish patterns of assemblages and how atypical it is.

To consider the assemblage within the setting of other ritual faunal deposits, in an attempt to understand what High Pasture Cave appears to represent, in terms of Iron Age society and its intentions at the site.

Chapter 3: The methodology used to study the High Pasture Cave faunal assemblage.

Section 1: Brief details of assemblage

Analysis was conducted on a total of 9,473 bone fragments, comprising the complete bone assemblage hand collected from the disturbed deposits (Context 1; Zones 1-6) and all faunal remains recovered by sieving from Trench 1 (Context 2-10) within the cave. The recovery technique was as thorough as possible considering the difficult and restrictive conditions as well as the disturbance created by visiting cavers before the site was recognised.

Section 2: Methodology

The methods and measurements chosen for the analysis of the assemblage have two primary objectives; to produce accurate results enabling the most confident possible interpretation of the assemblage and, and to produce results which can be considered alongside other relevant Scottish sites.

All bone fragments were counted and recorded. ?Large Mammal? is used throughout to describe bones and bone fragments of cow or red deer size which can not be identified to species. ?Medium Mammal? is used throughout to describe bones/ bone fragments of pig/ sheep/goat size which are not identifiable to species, following the methodology of Dobney et al. (1995). Ribs and vertebrae were, unless otherwise specified, recorded as ?Large Mammal? or ?Medium Mammal? rather than to species, as these elements can not be readily identified to species (Levin, 1999:258). In recognition of the great difficulty in morphologically separating sheep and goat (O?Connor, 2000:42) it was decided that these would be identified to a single sheep/goat (Ovis/Capra) category, although it is likely at this time and location that these would represent sheep (McCormick, not dated:13).

The bones were examined within their contexts and zones due to the different features to which they relate and the need to investigate the possibility of significantly differing composition of the assemblage within each context. A standardised bone proforma was used, with a separate sheet to record each bone fragment (see Appendix 1).

The bones were examined in spacious well-lit and clean laboratory conditions at Durham University Archaeological Department. A suitable environment and good lighting minimises errors of calliper reading, aiding accurate identification of bones and allowing any distinguishing or unusual features, such as obvious pathology or working which would require recording to be easily noticed. When examining the assemblage each bone was first identified to element, element side and species of each bone before working and pathology was examined. Accurate identification to element, element side and species was aided by consulting the large zooarchaeological reference collection at Durham University. Reference books, chosen for clarity of diagrams and pictures and personal notes and diagrams were also consulted. Diagrams from Amorosi (1989) were used as the basis for recording instances of working, pathology, burning or gnawing. Reference books consulted included:

o Von den Driesch, A. (1976) A guide to the measurement of Animal Bones
from Archaeological Sites. Harvard, Peabody Museum Bulletin 1.
o Hillson, S. (1996) Mammal bones and teeth: An introductory guide to methods
of identification. London, University College Institution of Archaeology.
o Amorosi, T. (1989) A Postcranial Guide to Domestic Neo-Natal and Juvenile
Mammals: The Identification and Aging of Old World Species. Oxford, British
Archaeological Reports (International Series 533).

Section 3: The measurements taken

Any whole measurements on bones were measured with 150mm span Dial callipers of a ? 0.1 industry-standard measure of accuracy, using Von Den Driesch?s published series of standardised measurements (Von Den Driesch, 1976) with modifications as recommended by Davis (1992). This model was used as the dimensions are clear both to measure and record and is the standard model for most British site reports (Reitz and Wing, 1999:169) , allowing ready comparisons to other sites. Results quoted throughout this study are in millimeters unless otherwise stated. No readings were recorded where complete measurements were not possible (due to breakage or extreme wear). One set of Dial callipers were used throughout to eliminate instrument error, although for measurements exceeding the maximum span Vernier callipers of 300mm span were used. It was felt that error induced by changing instruments was far less than that of ?estimating? the length using inappropriately small callipers. Zooarchaeologists rarely measure immature bones when analysing faunal assemblages, as animal bones are continually growing and developing dramatically until maturation (Davis, 1996:599). It is difficult to compare changes in size across the assemblage as being significant where they may simply be due to the incorporation of differing age ranges. Within this study the decision was taken to record measurements on both mature and immature bones in order to examine pairing of bones and size distributions. It was also noted whether the epiphysis was fused or not fused at this time. Where bones, in particular epiphyses, could later be refit they were additionally measured as a refit bone.

Section 4: Working, Gnawing and Pathology Recording

The bones were studied visually for evidence of working, gnawing and pathology. Working was recorded on diagrams of bones, adapted from diagrams by Amorosi (1989). Notes were made alongside these diagrams recording information about the type of working- defining areas missing (removed by working), cutmarks (shallow and fine working), chops (wide, deep, major working), and areas which may have broken off naturally. This allowed a consideration of whether a standardised strategy of working was occurring.

The level of weathering of the bone was recorded using the Behrensmeyer scale (O?Connor, 2000:44) as a way of simplifying the condition of the bone, identifiying if preservation varied between different contexts as well as giving an overall impression of the condition of the assemblage.

Section 5: The extraction of bones from Calcite

Within Zone 2 but also notable on several bones of Zone 3 there were large calcitic encrustations related to the conditions within the cave. For Zone 2 a particularly large calcitic encrustation encorporated numerous bones (Feature 001) , and it was impossible to identify even the number of bones contained within this layer without removing the calcite. Following consulation with the archaeologists on site and experts at the University of Durham (Birch, Rowley-Conwy and Gidney, 2005) it was decided to remove the calcite from the bones by hand, using dental tools, which enabled the calcite to be removed without damaging the bones. The calcite was mostly easily separated from the bone and calcite was only left on the few bones or teeth where it was deemed that further attempts to remove it would damage the specimen. The removal of calcite allowed the bones and teeth to be recorded, the surface of the bone to be examined for working evidence and the matching of epiphyses to diaphyses to occur, in common with the rest of the assemblage.

Section 6: Technique for quantifying the assemblage: Minimum Number of Individuals

It is important when examining the composition of any assemblage to determine whether complete animals or only portions of animals are represented, to evaluate whether the deposited assemblage represents the death assemblage (as defined in Banning, 2000:188) or only selected animal parts.

In order to examine this a Minimum Number of Individual (MNI) count, a common measurement used by zooarchaeologists (O?Connor, 2001:705, Klein and Cruz-Uribe, 1984:164, Grigson, 1999:165) was calculated for the appendicular and skull elements of the body (the axial parts being too fragmented to confidently identify to use) for each context. While preservation issues have led to some criticism of MNI?s it was deemed that they were the most suitable calculation to use here because of the excellent preservation of High Pasture Cave. MNI calculations are sometimes criticised with concerns that they have little value as a parameter for estimating the size of original populations (Gautier, 1984:242) based on doubts about the relationship between the minimum number of individuals, the real number of animals present in an assemblage and the initial number of animals (Gautier, 1984:242). Because of the juvenile age of much of the High Pasture Cave assemblage, in particular in the pig-rich Zone 3 assemblage, it is clear that a number of parts are present for many of the bones as many of the bones are represented by both diaphyses and unfused epiphyses because of their young age. This would mean that any alternative Number of Identified Specimens (NISP)-based calculation would be artificially inflated for these juvenile individuals. Fragmentation can be seen to differ between contexts which would also cause major distortions. Because of the problems of NISP-based measurements, MNI is seen by many as a more suitable measurement (Klein and Cruz-Uribe, 1984:164), particularly as in the case at High Pasture Cave where fragmentation differs between contexts (O?Connor, 2001:704). At High Pasture Cave an attempt has also been made to pair bones within contexts to individuals. While White dismisses the reconstruction of pairs as ?a great deal of effort with small return? (White, 1953 in O?Connor, 2000:59), more modern archaeologists often employ attempts at pairing to get a more accurate picture of the MNI, and with the excellent preservation at High Pasture cave this is likely to be a successful technique. Pairing will lead to a more accurate MNI-count (Klein and Cruz-Uribe, 1984:172) as it takes into account whether bones could come from the same individual.

MNI is also believed to be the most appropriate estimation of number of individuals where there is good preservation (O?Connor, 2001:706). At High Pasture Cave in Zone 2 preservation is excellent, illustrated by the possible reconstruction of a complete cow spine and the presence of very fragile and rarely recovered skeletal elements such as ear bones (otoliths) and hyoid bones. The large number of pig bones surviving also indicates excellent preservation as pig bones tend to be porous and fragile (Albarella and Payne, 2005:589). The bones recovered can be confidently expected to reflect the original number of animals deposited, making the assemblage a suitable candidate for MNI calculations. This preservation together with the lack of gnawing on the bones suggests prompt burial, as it has at Durrington Walls (Albarella and Payne, 2005:590).

Section 7: Methods used to estimate age of individual bones, by bones or teeth

Teeth (Eruption and Wear):

Teeth are regarded as the most precise way of estimating the age of skeletal remains, with patterns of dental eruption and tooth wear being excellent indicators of an animals age (Banning, 2000:148, Silver, 1969:289). Deciduous teeth are replaced on a schedule known for modern animals and estimated to be the same in antiquity (Matschke, 1967:109), allowing precise identifications of age (see table 2 for eruption dates used to age the pigs in this study). The increase of wear on occlusal surfaces of teeth can also be useful as the older the animal the more wear is accumulated (Silver, 1969:289). Tooth wear can depend on a number of factors other than age, such as diet, health and individual variations in mastication although there is no variation through the sex of the animal (Matschke, 1967:111). A rough guide to age is possible particularly when tooth rows, as opposed to individual teeth, are present (Banning, 2000:198). Wear stages used to age the jaws followed the methodology of Bull and Payne (1982) and Grant (1982) for the Pigs, and Grigson (1982) for the cattle. The eruption dates used to age the teeth for this study were (Rowley-Conwy, 1993:180) for pigs, and (Grigson, 1982:20) for the cows, too few sheep/goat teeth were present for ageing to be possible.

Bones: (Epiphysial Fusion)

A consideration of the status of fusion of the bones can also be useful in ageing faunal remains. The epiphyses remain separated from the diaphysis (shaft) by cartilage for some time before fusion (Banning, 2000:198) but eventually ossification proceeds until they are fused, with bones fusing sequentially at different times during the maturation of the animal (Legge et al., 1991:56). The ages at which fusion occurs is known for modern animals and can be used to estimate age (in months) of archaeological specimens (Banning, 2000:148). Much less information is known about the timing of epiphysial fusion in the past than is known about tooth eruption and wear timings (Bull and Payne, 1982:67) which makes any ageing far less likely to be accurate (Bull and Payne, 1982:70), and is best used to corroborate other lines of evidence. Some also disagree over what constitutes ?fused?, sometimes taking fusion as early as the time the epiphysis can not be separated from the diaphysis or as late as when no fusion line can be seen which means that comparable data is often difficult to use confidently. In this study, the modern and relatively standard definitions (such as determined by Albarella and Payne, 2005:590) are employed, namely ?fused? (no fusion line visible), ?fusing- line still visible? (epiphysis and diaphysis joined but fusing line still visible), and ?unfused? (epiphysis and diaphysis separate). These definitions are becoming the norm when recording fusion data and allows the epiphysial fusion data by Bull and Payne (1982) for pigs and Grigson (1982) for cows to be used to determine the ages of the skeletal remains. Once epiphysial fusion is complete indications of age from bones are very inaccurate (Silver, 1969:284) and not considered within this study.

Section 8: Techniques to record and study the Butchery evidence

All the bones at High Pasture Cave throughout the contexts appear to have been worked exclusively by metal tools. The difference between stone/ bone and metal tools is clearly distinguishable (Greenfield, 2000:97) with metal tool marks producing consistent and characteristic marks (Greenfield, 1999:803, 804, 2000:99). Metal tools produce deep, narrow and steep-sided cutmarks with a far more uniform appearance than with stone (Binford, 1981:105) and when cut obliquely with metal a ?shelf? of bone is often left in place (Binford, 1981:106). Stone tools normally produce shallow cutmarks with wide and irregular grooves (Greenfield, 2000:100) and these stone marks are generally shorter than those made by metal and more ?ragged? in appearance (Binford, 1981:105).

There are at present few studies examining the meanings behind metal butchery marks on bones, despite recognition of the importance in examining patterns of butchery (Guilday, 1962:96). Binfords (1981) study of butchery techniques is one of the most comprehensive and widely used sources when considering working of domestic animals in archaeology. His terminology for the definition of cutmarks across bones is adopted here to record the types of cutmark present, as well as acknowledging his explanations of the processes signified by the working, alongside the study of Luff (1994) which specifically examined the working of pig bones.

For this study working is defined using the terminology ?cutmarks? or ?chopmarks? respectively, as well as deliberate breakage and points of impact being noted. The definition of the types of marks follows the descriptions of Greenfield; that a cutmark is a slice of the bone created by drawing knife across surface of bone and a chopmark is the impact on the bone with force of a knife/ sword/ axe-like blade (Greenfield 1999:798, 2000:94). Any working was identified and recorded under a strong oblique light source in the laboratory as recommended by Greenfield (2000) and Egeland (2003) to aid easy identification and no magnification was used as it is generally acknowledged that cutmarks of any importance are invariably visible to the naked eye (Greenfield, 1999:805).

Chapter 4: The analysis of the ?Pig-Rich? Zone 3 assemblage

Section 1: Overview of the Zone 3 Assemblage

The large majority (89.2%) of the bones from Zone 3 (part of context 001) are pig (Sus Scrofa). 558 bones are positively identifiable as pig bones, with another 206 having a high probability of being pig vertebrae fragments and 275 rib fragments are of suitable size for pig, giving a total of 1039. Only 132 bones from Zone 3 come from other or unidentifiable species. Of the pig bones 418 (including vertebrae and rib) exhibit cut marks (40.23%) a high percentage of working.

The other animal remains found in Zone 3 consist of cow (Bos sp.), red deer (Cervus Elaphus), sheep/goat (Ovis/Capra) as well as one premolar coming from a dog (Canis sp.).

Section 2: The Minimum Number of Individuals of Pigs (MNI)

The Zone 3 assemblage had been recognised as very complete as early as the preliminary report (Birch, 2003) and seemed of suitable preservation and completeness to consider an MNI analysis (see figure 11).

Figure 11: A Zone 3 pig femur illustrating the typical preservation and completeness of the pig bones

The main criticism of the Minimum Number of Individuals (MNI) technique to quantify assemblages is bias due to poor preservation. It is evident that this is a limited problem for Zone 3, the average weathering using the Behrensmeyer Scale (See O?Connor, 2000:44) is 0-1. This limited evidence of weathering or gnawing suggests that the bones were not left on the surface for long periods of time (Levin, 1993:259) and improves their suitability for the MNI technique which relies on good identification. The relatively discrete and small sample of the Zone 3 pigs, as well as their excellent preservation meant that the potential for these techniques was ideal and so the MNI for pig elements were calculated.

To ensure that the MNI?s were not under/overestimated the fusion status of bones, their size and whether broken articular ends could potentially be from the same element was considered. To gain the MNI of mandibles the toothwear and eruption status of jaw-sides was compared, as sides of very different tooth eruption or wear are unlikely to be a pair, a technique recommended by Bourdillion and Coy (1980:83).

The calculation of MNI of Zone 3 pigs by element shows that the largely meat bearing bones of the humerus, femur and tibia are relatively well represented and the phalanges the least represented as would be expected when some transportation had occurred (Rowley-Conwy, 1993:185) and that this was a site of consumption rather than of killing the animals. However, the differences in number are relatively small and taking into account the greater effects of taphonomy on the smaller elements and the hand-collection of this zone it may be that the pattern is due to the greater recovery of the larger and more robust elements. In contrast the presence of large numbers of cranial/ head elements suggests primary butchery was occurring nearby, as it is unlikely this element would have been transported back to site (O?Connor, 2000:135).

Comparing the MNI for elements by side also allows us to also consider whether there are any differential patterns in side distributions, particularly important as ritual/ status sites elsewhere in Scotland, such as Dalenun, Lorn (Hingley, 1998) have shown a prevalence for left-side elements. Despite the known separation into left and right of the pigs the sides at High Pasture Cave appear very similar in distribution and when the MNI numbers for individual skeletons are examined there is no predominance of left over right elements. The only exception to this pattern is the right metatarsal where there are three times the numbers of right elements to left, but this may be a quirk of preservation and is unlikely to be significant.

Figure 12: Diagram reflecting the MNI of Zone 3 pigs by element, illustrating the MNI calculated for different sides.

Excluding the skull, the appendicular elements give an overall MNI of 15 (see figure 13), of which one individual is probably perinatal. The mandibles and cranial elements paired by tooth wear data and fit (see next section for details) give an MNI of 23, suggesting an additional deposition of single skulls as well as complete animals.

Figure 13: Diagram illustrating the overall MNI of Zone 3 Pigs by element

All parts of the appendicular skeleton are present in the assemblage suggesting whole animals were deposited with few differences in numbers of parts across the skeleton (see figure 13), for example phalanges are present suggesting there is no removal of ?meat-rich? joints for deposition.

The examination of the elements therefore suggests a minimum of 23 individuals have been deposited in the cave, however the MNI suggest that these were not all deposited as whole animals, but rather around 15 ?whole? animals, and 8 additional skulls.

Section 3: Age profile of the assemblage

o Age estimation from dental evidence

At High Pasture Cave there were a good sample of juvenile mandibles and they were matched into potential pairs of sides containing similar tooth eruption and wear stages. Attempts to match possible Mandible and Maxilla partners occurred and this process examined eruption, tooth wear and how well the tooth rows fit together. These were then counted as evidence for ?one? individual, in order to prevent over-estimation of the animals present.

Once an age order had been established, the jaws were examined and aged by wear stage according to Bull and Payne (1982) and Grant (1982), considering the degree of tooth eruption and wear each showed. Eruption stages of pig dental development were aged using the model put forward by Rowley-Conwy (1993), allowing ageing even when the mandible was fragmentary. Eruption status was recorded followed the criteria proposed by Payne and Bull (1988) as the original designations by Higham of ?primary, secondary, tertiary? reproduced by Rowley-Conwy in this model were felt too vague to be of use in this case. From mandible ageing there appears to be a major group of mandibles of age 7-8 months, as well as older mandibles of varying ages.

This grouping of ageing is replicated at other Scottish sites, for example Howe, Orkney where 52.5% of the pigs died before 1 year old, and 32.7% died before 2 years old (Smith, 1994:145). Similarly, at Crosskirk, 61% of the pigs were juvenile. Pigs do not reach their full size and weight until 2/3 years old (Maltby, 1996:23) and this pattern of killing younger at Scottish sites, replicated at High Pasture Cave again suggests that the pigs were not being utilised to their full nutritional value. 

Studies on dental data in wild pigs from three differing areas (Former East Germany, United States and Poland) show close similarity in eruption age, casting doubt on suggestions that nutrition (which might be expected to be different in Skye to elsewhere) has an effect of eruption (Rowley-Conwy,1993:180). This appears to confirm Payne and Bulls belief that eruption sequences in both domestic and wild pigs are very stable (Bull and Payne, 1982:55) indeed stable to such an extent that they are believed to be able to be extrapolated onto the past (Bull and Payne, 1982:65).

Pigs are born in a relatively short season, between March and May (Lauwerier, 1983:486-7). The time of reproduction depends on the availability of food (Noe-Nygaard and Richter, 1990:175) and so it is likely that in marginal areas only one breeding season would occur per year although two farrowings may occur in more temperate climates (Smith, 2000:712). Skye is considered to be a marginal area for pigs (Armit, 1996:135) and it is unlikely that more than one litter was produced each year due to limitations of climate and diet. It is hypothesised in this study that the time of farrowing is probably restricted to around the month of April, as the conditions would be most favourable at this time (Smith, 2000:711). By examining the ages of animals represented in the assemblage using an April farrowing date it is possible to determine the season of death (Banning, 2000:200) and to examine whether there was a specific ?season? of deposition, or whether the remains were placed into the cave throughout the year. This is a common technique used to examine the season of site use (Legge et al., 1992:49) and used successfully with pigs at other sites such as Ringkloster and Lunby (Rowley-Conwy, 2001).

Figure 14: Season of Death diagram, as determined by ageing of mandibles for Zone 3 pigs, following design of Rowley-Conwy (2001:135).

It is clear from this season of death examination (see figure 14) that there is a concentration at the age of 7-8 months, with one 9-10 months old jaw (potentially a birth unusually early in the season; for example in March, falling into the same kill-range as the 7-8 month old). There are additionally kills of animals 12 months older than this 7-8 month age group- i.e of animals of 19-21 and 21-23 months of age so in their second winter, emphasising that the majority of killings are from pigs of winter age. From this data it is evident that the majority of the pigs were killed in the winter, conceivably mostly within a December-January period. It is impossible to say that this represents seasonal occupation of the site, however, as it may instead be representing seasonal cull at the site which is occupied all year round (Legge et al., 1991:50). The exceptions to this are the two mandibles which relate to an August-September kill, and these may represent mandibles put in separately to the others at a different time. It can not be proven that there were only two deposition events (one in December-January and one in August-September), instead there may have been incremental deposition over many years. This study can conclude that it is likely that on at least two separate occasions there were deposits into High Pasture Cave, although the August-September deposition may simply be animals born outside the usual period or who erupted their teeth at unusual ages, and that the predominant pattern is of pigs killed in Winter although not all of the same age. The killing of pigs of less than a year old at this time of year is unlikely to be to attain their meat for storage over winter, their size makes them ineffectual for preservation (Smith, 2000:712) and so it is unlikely that their death represents kill to preserve feeding of unnecessary animals.

o Age estimations from skeletal evidence

It has already been noted that many of the remains from High Pasture Cave Zone 3 layer are juvenile animals and it is important to consider, with the recognition of the varying number of depositions between post-cranial elements and cranial, whether these juvenile post-cranial skeletal elements correspond in age to the 7-8 month large group of mandibles identified from the dental records -as would be expected, or whether they confirm a difference in deposition between the skulls and skeletons.

1. Age from Epiphysial fusion:

At High Pasture Cave many of the diaphyses and epiphyses can be refit onto their respective parts as each has a unique articular surface and so can not be refit incorrectly. This has been undertaken with the Zone 3 pig remains and a high percentage of them do refit, again emphasising the good preservation and recovery of the assemblage. This completeness of the assemblages allows an examination of epiphysial fusion for the elements, and may provide indicators as to the ages of the animals.

An examination of the fusion of the bones confirms the age patterns already recognised, that the majority of the individuals are under 1 year old (for example seen clearly in the humerus, pelvis and radius ageings) but that there are also some older individuals, mostly between 1-2 years old. However, also visible from the fusion of the Femur are 2 individuals of c. 3?  years of age, which are not identified through tooth data. These individuals may represent pigs kept for breeding stock (Smith, 1994:145). Therefore the epiphysial fusion information largely confirms the tooth ageing information, but with some additional individuals also identified.

2: Age from Biometry of the bones

An examination of the size of the bones can be useful in determining whether there are particular clusters in size of bones from within elements and thereby if there are groups of ages, as bones of similar ages should equate to approximately similar sizes (Albarella and Payne, 2005:589). Bones increase rapidly in size with age in young pigs as the diaphysis lengthens by the deposition of new bone (Payne and Bull, 1988:29, Legge et al., 1991:49), although the increase occurs at varying speeds in differing bones (Albarella and Payne, 2005:595). Post-fusion there is less difference in growth, although some elements such as the humerus and radius do show some evidence of post-fusion growth (Albarella and Payne, 2005:595) and not all bones reach a stable plateux (Payne and Bull, 1988:29). As the High Pasture Cave pigs are predominantly juvenile and unfused/ fusing at their epiphyses, an investigation may determine whether they fall into varying groups dependent on their age of death. It would be expected that if, as the tooth eruption and wear stages suggest, there are two major groups, the measurements should also separate into two clear groups relating to each of the ages. This follows the theory used by Legge et al., to successfully show discontinuous seasons of culling in sheep at Chesterford Roman Temple (Legge et al., 1991:56).

In bone samples where a good sample of sub-adult bones survive, all sizes of juvenile bones will be present when a site was occupied all year round, while only certain size classes will be found when there was a ?.seasonal cull?
(Legge, 1991:49-50)

When considering bone measurements as a means of determining differing age groups within an assemblage, the scapula SLC, long bone shaft widths (SD), and lengths of the radius and humerus (GL) are known to show the greatest increase with age (Albarella, 2005:596, 598, Payne and Bull, 1988:30) and should be particularly useful to determine if measurements group into age. Other measurements show less marked size change with age (Albarella and Payne, 2005:596) and so the patterns from all bone measurements have not been considered, instead only those recommended by Albarella and Payne (2005:596) and those which appear to show significant ?age grouping? have been reproduced.

There appear from a consideration of the measurements to be three ?groups? of pig sizes reflected in most bones, illustrated by the red highlighting on the graphs. This suggests that the skeletons fall into three age groups (including the perinatal skeleton), and where the perinatal skeleton has no element measurements (for example the tibia SD and radius SD), two groups are evident. This means that rather than representing a complete distribution of ages, as might be expected if the cave was being utilised all year round, the graphs instead show that there are age groups. The ages unrepresented in the measurements probably correlate to times when the cave was not used for disposal and related to the months represented in the season of death analysis (See figure 14) if it is assumed that the pigs farrowed only once a year.

Figure 16: Diagram illustrating the remains of a perinatal pig from Zone 3

The graphs also confirm that the middle age-group contains the largest number of individuals, for example in the ulna GL, tibia SD and radius SD graphs. This group could correlate with the 7-8 month mandible-aged group. There is also a smaller group of larger measurements visible in the graphs, representing a smaller number of older individuals, which may correlate with the 19-23 month old group.  Only the humerus SD does not show any grouping and this may be due to vagarities of the measurement type. The skeletal measurements therefore confirm the information from the tooth data, that the age distribution does not suggest disposal across the year but at specific times.

Section 4: The Zone 3 Pigs: Wild/ Domestic?

There are various lines of evidence which can be used to determine whether the High Pasture pigs are hunted wild boar or the domestic pigs which are expected considering the climate (Grigsom, 1999:213).

It is important to consider the age distribution of the pigs and at High Pasture Cave the majority of the remains are from pigs in the winter of their first year. At Durrington Walls, a very similar age distribution (of a killing peak of pigs in their first winter, and a smaller number of specimens killed in their second and third years) has been used to illustrate the remains as those of a domestic population (Albarella and Payne, 2005:591).

Domestication can be difficult to identify through animal skeletons (B?yi, 1989:25) although pigs do show some physical changes as they are domesticated, in particular a decrease in size (Grigsom, 1999:221). At High Pasture Cave the size of bones is not easy to use for this purpose, as not only are the bones juvenile and most of the comparative data is from adult pigs, but the location of High Pasture Cave on Skye would predispose the size of the pigs to be smaller than usual as they adapt for the poor conditions (Grigsom, 1999:213).  One of the more profitable ways of examining whether pigs are wild or domestic is to study the size of their teeth; both in length, which has been studied for many years, and in width, which suffers less than length from wear in older animals (Payne and Bull, 1988:31). Teeth measurements examined follow the recommendations of Albarella and Payne (2005) and Payne and Bull (1988), as suitable and successful measurements for this purpose.

Teeth measurements, unlike skeletal evidence, can also be compared to data from other sites as morphology is likely to vary far more between individual populations in bones than in teeth, bones being more influenced by genetics, nutrition and other environmental variations (Payne and Bull, 1988:37). Teeth are therefore particularly suitable for comparison between collections from different times and areas.

The High Pasture Cave teeth measurements have for comparative purposes been averaged to reduce bias due to the herd sex ratio (sex being unable to be ascertained and studied separately due to the incomplete nature and juvenile age of the jaws). The average measurements have then been compared to two sets of data; a modern reference sample of wild boar from Kizilcahamam (Payne and Bull, 1988) and a sample of domesticated pigs from an archaeological assemblage at the Neolithic site of Durrington Walls (Albarella and Payne, 2005). The pig tooth measurements of Durrington Walls, a Neolithic settlement with known domesticated pigs (Albarella and Payne, 2005:590), has the benefit of predominantly coming from animals believed to have been slaughtered in their first winter (Albarella and Payne, 2005:590), similar to hypothesised for High Pasture Cave.

The co-efficient of variation of the High Pasture Cave pig teeth is very low for many of the teeth, helping suggest that the pigs are part of a single population, rather than representing individual wild boar from different locations. 
The High Pasture pigs teeth in comparison to other examples relate far more closely to the Durrington Walls domesticated pigs and generally the High Pasture teeth are even smaller than at Durrington Walls.

Figure 17: Comparison of the average tooth measurements of the High Pasture Cave Zone 3 pigs to two other samples

This evidence suggests that the teeth reflect a ?domestic? rather than a ?wild? population of pigs at High Pasture Cave. Tooth widths in particular are likely to better discriminate between the larger wild and smaller domestic pigs, as tooth lengths encounter problems because of wear creating greater age-related variation (Bull and Payne, 1988:36, 37). When comparing the widths it is clear that High Pasture Cave is always the smallest example, and thus the most ?domestic? following the criteria out forward by Bull and Payne (Bull and Payne, 1988:37).

Further support for the domesticated hypothesis also comes with the interpretation that the Zone 3 pigs seem to represent animals killed close to site and so are more likely to be domestic, rather than hunted animals which would have had to be transported to site. Where animals are killed off site, often only the desired nutritional or useful elements, usually the larger post-cranial bones, are transported to site (Rowley-Conwy et al., 2002:78) and the waste, such as the head and feet which are of little useful or nutritional value, are discarded where the animals are killed. At High Pasture cave all elements from the skeleton are represented within the faunal assemblage and the skeletons appear very complete.

The other aspect of the Zone 3 Pigs which may provide supporting evidence for the status of the pigs is any instances of pathology or disease. In the Zone 3 pigs there are several bones reflecting disease and three ribs which show healed breaks as well as one mandible with a misaligned tooth growing through the jaw. There is little modern clinical literature available (O?Connor, 2000:98) to aid the determination of the diseases, however, some information can be gained from considering the pathologies present in this assemblage.

The pig bones presented several examples of healed ribs with previous simple fractures (defined following Baker and Brothwell, 1980:86), which can tell us little other than an injury healed prior to killing. However the presence of two deformed phalanges, with massive presence of cavitation pitting and new bone formation does provide significant information. The degree of lameness, which would have been caused by such bone growth, would have been severe and permanent, with the bone growth taking a long time to accumulate. This suggests the pig(s) were lame for a relatively long period of time before death. It is unlikely, given the extent of lameness, that these pigs would have been able to forage for themselves (Rowley-Conwy, 2005: Pers. comm), and is further evidence that these pigs were sustained by humans. The evidence of care taken even with animals which were ailing is clearly suggestive that the pigs were of some importance to the inhabitants, and confirms their domesticated status.

Several diseased ribs were also present at High Pasture Cave. The marked flaring of the bones is highly suggestive of rickets (for comparison see Baker and Brothwell, 1980:50), a relatively common disease in the archaeological record (Baker and Brothwell, 1980:49). This suggests an inadequate diet, with respect to phosphorous or vitamin D or lack of sunlight. It has already been noted that the diet on Skye would be poor and this appears to confirm this.  From other contexts there are no examples of disease amongst the bones of any species, except for a cow tarsal in Zone 2 presenting fusion of tarsal bones as is commonly caused by spavin (Baker and Brothwell, 1980:119)- a common condition in cattle and unlikely to cause more than a mild degree of lameness.

Chapter 5: Other contexts at High Pasture Cave

a) Other ‘unusual’ deposits

Other unusual faunal assemblages exist within the cave as well as the Zone 3 pig-rich layer, in particular two cows appear to have been placed in discrete deposits.

Section 1: Zone 2 (Feature F.001)

The Zone 2 feature consists of a large number of cow bones (158), along with 241 unidentifiable fragments and 39 small-animal fragments placed in a recess in the main Bone Chamber. This feature was discovered on removal of a stalactite for dating purposes, some of the bones being held within a calcite concretion while others were held below this layer.

Dental instruments were used to remove the bones from the calcite to enable examination. The calcite and bones separated very easily and the bones remained in excellent condition for examination (see figure 19), with the majority identified as 0-1 in preservation using the Behrensmeyer scale (O?Connor: 2000:44). There is no evidence of gnawing on the bones, again suggesting swift deposition into High Pasture Cave.

Figure 19: Picture of remains of Zone 2 cow, illustrating the excellent preservation

A study of the bones indicates that almost all the constituents of an individual cow skeleton are present suggesting killing near to or on site(O?Connor, 2000:135) (see figure 20) with only the axis, tail and sacrum completely absent.

Figure 20: Elements of the Zone 2 cow present

Indeed the individual is so complete that the spine can be reconstructed almost completely providing persuasive evidence that the vertebrae and ribs (which are not identifiable to species) almost certainly come from the same animal.

Information from this deposit can be determined using similar methods to those used to analyse the Zone 3 pigs. As there is only one animal no information can be inferred about herding strategies or deliberate choice of age for deposition, however information can be determined about the age of the animal as well as the butchery methods used (see chapter 6 for butchery analysis).

To age the animal the status of fusion of the bones as well as tooth wear data was examined, as with the Zone 3 pigs, although different ageing data will be used (Grigson, 1982), as cows age in different ways and times to pigs.

The epiphysial fusion status of the bones gives an age range of 12 to 18 months for the cow. The pelvis Ilium-Ischium is fused, meaning the cow had to be older than 7-10 months, and both humerus distal epiphyses were fusing with the fusing lines still clear giving a minimum age of 12 months (fusion occurring between 12-20 months) but both the proximal and middle phalanges had not fused at their proximal epiphyses, meaning the cow had to be younger than 18 months.

Figure 22: Refit of the Zone 2 cow phalanges (on the left of this image), the proximal epiphyses of which are unfused.

Tooth data was also used and as with the Zone 3 pigs it is recognised that this is a more secure method to age the individual. Examination of the mandible of the Zone 2 cow suggests an age of c.15 months, as the M2 is half erupted, with one cusp in early wear stage e- giving a date of 15-16 months of age and so the individual is judged to be of around 15-16 months in age. It is impossible to sex the individual, as the sexually dimorphic elements of the crania (which differs between males and females) is incomplete and the pelvis is also fragmented.

No burning is noted on the individual although there is some suggestion of heating on the right astragalus. This deposit therefore appears to represent the remains of an individual cow deliberately placed into the discrete location of a recess of High Pasture Cave which later became covered by calcite concretions.

Section 2: Other Faunal remains in Zone 2

1.  Within the F.001 cow discrete deposit:

Other faunal remains were discovered in Zone 2 both within the calcite concretion layer and above it, so the cow, although the significant find of the zone, was not the only animal placed in this area. The other remains within the concretion-capped feature consist of 241 unidentifiable fragments and 39 medium mammal sized bones of which one is identified as pig vertebrae and one as a sheep/goat radial carpal. The majority of these items are vertebral plates (unfused) and rib fragments. It is possible that these fragments may have entered the feature through either slipping into the deposit or being in the sediment onto which the cow remains were placed, rather than through deliberate deposit. There are no limb bones and only two indistinct cutmarks on fragments so no further information can be gained.

2.  Zone 2: Floor layer within the Cave (Context 001)

The rest of Zone 2 (unassociated with the Feature F.001 cow deposit), appears very similar to the Zone 3 deposits and indeed is likely to be related to those deposits due to post-depositional disturbances. It consists of 376 animal bone fragments, of which 76 are identifiable to species. Species identified include sheep/goat (represented by a femur unfused distal epiphysis, tibia fragment, and radial proximal epiphysis), cow (represented by a left metatarsal, carpal, pelvis part, scapula and humerus bones, mandible fragment and loose teeth and probably ribs), red deer (represented by a pelvis part and astragalus) and pig.

Pig is by far the dominant species in this layer, as would be expected with its hypothesised association to Zone 3, with 56 of the 76 fragments which can be positively identified to species being pig. Of the rest of the bone elements 101 are identified to ?medium mammal? and consist predominantly of cranial, vertebral and rib fragments, as well as some shaft fragments. Within the identified pig bones all elements of the skeleton are present; including femur, humerus, calcaneum, phalanges, carpals, scapula, tibia and multiple vertebrae and tooth remains. The majority of the bones contain unfused epiphyses, similar to the Zone 3 juvenile pig remains and several exhibit cut marks, with one pelvis fragment showing a chop mark.

It was decided that this zone did not warrant the detailed examination of the Zone 3 assemblage. The context of certain elements was uncertain due to post-depositional disturbance and so it was unclear whether they relate to the large cluster of pig bones discovered in Zone 3.  There are also very few diaphyses present in this zone, only four femurs (one incomplete) and two humeri. This layer can be concluded to be ?pig-rich? similar to Zone 3, containing elements from across the skeleton but also including other species. None of the bones show evidence of burning, apart from the red deer astragalus and a pig accessory carpal, which is heavily burnt.
There are also indications of working by cutmarks on several bones, but little other significant information can be gained from this Zone.

Section 3: Zone 5

The bones from this deposit were recovered from a ledge in the main stream passageway close to the entrance to High Pasture Cave and formed a discrete deposit above the streamway. The faunal remains in this location consist of 192 bones, all of which are identifiable as cow (Bos sp), found within the matrix on the rocky ledge (4 bones were found above this matrix, lying on the rocky ledge). These remains appear to be another deposit of a single butchered cow. The bones are more poorly preserved than those from within the Bone Passage, appearing white, cracked and powdery, with many of the bones in the 1-3 scale of preservation (using the Behrensmeyer scale) compared with the excellent 0-1 preservation of the Bone Passage assemblage.

It is hypothesised that the more exposed location of a ledge above the active streamway may have contributed to this poorer condition, however, while the bones were perceptibly more brittle and cracked than those in the Bone Passage, 100 of the 192 fragments from Zone 5 were still identifiable either to element or species with only the very smallest fragments so badly abraded as to be unidentifiable.  There is no evidence on any of the bones that burning or heating occurred. There is evidence of working on 75 of the 100 identifiable bones (see chapter 6 for butchery details). Of these bones all of the identifiable limb bones and skull (the ribs and vertebrae impossible to identify to species) have been identified by morphology to cow (Bos sp.), and it is highly probable that the vertebrae and ribs found in this deposit are from the same individual. 

This zone appears to be the remains of one cow (Bos sp.) (see figure 28), carefully placed on a ledge of the cave above the active streamway outside of, but accessible from, the area where the majority of the deposits were placed. No head or phalanges are present which may indicate that we have the ?meat? portions of the cow rather than the whole skeleton and that the cow may not have been killed on site. The absence of the femur is harder to explain as this would be one of the most significant meat-bearing joints and would be expected to be present. It is highly unlikely that this absence is due to taphonomy and may instead suggest deliberate deposition elsewhere. 

Figure 28: Elements of the Zone 5 cow present.

Following the Grigson (1982:22) ages for epiphysial fusion, the cow can only be aged as over 3 ?-4 years (i.e. adult), as all elements of this date such as the Humerus proximal epiphysis, Femur distal epiphysis and Tibia are fused and there are no ?later? fusion dates than this. There is no skull present so a more accurate age cannot be established through tooth eruption and wear.

Zone 5 can therefore be concluded to represent the remains of one cow of indeterminate sex but of an adult age. It can be stated with confidence that it has been deliberately placed in the location and has certainly undergone extensive working (see chapter 6). Not all of the skeleton is present and the absence of the parts that are missing, including the Humerus, cannot be easily explained by taphonomy and suggests deliberate disposal elsewhere. With such deliberate placement of the bones in this location, it is unlikely that these bones would be accidentally excluded.  There are no other faunal remains of any other species in this location and so Zone 5 appears as a deposit of the cow only. The location of Zone 5 on a narrow ledge above the active streamway of the High Pasture Cave complex makes it impossible that this is an accidental deposit, it is inconceivable that the bones can have arrived here except through deliberate placement.

b) Contexts not identified as containing “unusual” deposits

Section 4: Zone 1

Only 6 fragments of bone were recovered from this location and all are of little note, probably relating to the other zones within High Pasture Cave and moved here by post-depositional activity within the cave. The bone fragments include part of a cow tooth and some fragments of bone identified as Medium Mammal. No cutmarks exist on the bones and little further information can be obtained.

Section 5: Zone 3/4

The Zone 3/4 deposits consist mainly of sieved earth recovered later than the majority of the Context 001 faunal remains, associated with the excavation of Trench 1.  2,001 fragments were recovered, although the majority of these were not identifiable. Species identified include pig, cow, sheep/goat, red deer, and one dog canine. Preservation was again excellent and even very fragile bones such as pig ear-bones survived. Due to the sieved nature of this group, much of the assemblage consists of very small elements not recovered as frequently in the hand collection of the assemblages from Zone 3 and 4; such as sesamoids, carpals and phalanges. There are very few limb bones (only 1 pig femur) and little evidence of working, although some of the vertebrae plates displayed similar saggital working to that identified from the Zone 3 assemblage (see chapter 6). There are 153 small burnt fragments none of which can be identified to species and are mostly under 2cm in size. Compared to the Trench 1 assemblage deposits this is a very low amount of burning. Little further information can be gained from this zone, the results of sieving meaning that it consists mostly of very fragmentary elements.

Section 6:  Zone 4

Zone 4 consists of 324 bone fragments, although most of these are not identifiable to species. The preservation is excellent, and there is little evidence of working. The majority of identifiable bones consist of pig bones, although no major limb bones are present (phalanges and metacarpal/tarsals producing the concentration). Sheep/goat bones are also present; including cranial elements, humerus, and metapodials, suggesting all elements of the body were represented, rather than just butchery waste. Additionally in this layer, the humerus of a perinatal calf was discovered, along with phalanges and naviculo-cuboid of an older cow. None of the bones contained significant cutmarks beyond that expected for processing. It is anticipated that this Zone, similar to Zone 2, consists of elements related to the Zone 3 pig-rich assemblage, however due to the post-depositional disturbance it is unclear whether this is the case. In particular, the greater number of sheep/goat bones may suggest some difference in deposition between the areas, and so it was decided not to incorporate the data from these areas into Zone 3. 

Section 7: Zone 6 (Lake Passage Choke):

Only 9 bones were recovered from Zone 6. All of the bones are cow (Bos sp.) and show evidence of working, with cutmarks and chopmarks. The elements present, which includes 3 cranial fragments, 1 right metatarsal, 2 humerus fragments, and 1 left ulna shaft, are all elements missing from or only partially present in the Zone 5 cow (see figure 28), which was deposited nearby. These elements could all be associated with the Zone 5 cow and so probably represent a secondary deposit or wash of bones from the primary deposit site of Zone 5. The elements are all fragmentary and worn (ranging in Behrensmeyer wear scale from 1-3) and little information can be gained from them beyond the note of their likely association with Zone 5.

Section 8: The Trench 1 Deposit

The rest of the assemblage from High Pasture Cave was recovered from below Context 001 (the Zones examined so far all being divisions of this context), being excavated from within Trench 1, a c.1x2 metre trench placed within the Bone Passage.

The Trench 1 deposits are far more fragmentary than any of the remains within Context 001. Unlike the majority of Context 001 the sediments were sieved and it is possible that the greater number of unidentifiable and very small fragments are a product of this improved selection. When examining the Context 001 in-situ layer overlying the Trench 1 deposits and which was dealt with by sieving as a Trench 1 context, it is apparent that Context 001 does have a different character to the Trench 1 deposits, exhibited even in the sieve residue.

Figure 31: Graph illustrating the relationship between identifiable and burnt fragments in Trench 1

Specifically, the Context 001 sieved deposits have a far greater percentage of identifiable pieces and a far smaller amount of burning than the Trench 1 contexts (see figure 31). Considering that this comparison does not include the earlier hand collected Context 001 assemblage (which as we have seen consists of numerous identifiable complete and unburnt bones) but only the disturbed top on Trench 1, it is evident that there is an extreme difference between the surface layers and the trench deposits. This extreme difference becomes even more apparent when the hand-collected assemblage is included. A consideration of Context 001 to the Trench 1 average however does not overtly reflect the differing amounts of burning, with the average of trench 1 being very similar to the Context 001 statistics, despite some contexts being much higher (See figure 31). It is clear that the amount of burning varies dramatically in Trench 1, and this may be seen as evidence that the nature of the deposits varies within Trench 1, rather than being deposits of very similar “domestic deposits” followed by a very unusual context -Context 001.

The fragmentation of the Trench 1 deposits and the greater amount of burning in some of the contexts, in particular contexts 7-10, does suggest domestic refuse throughout the contexts, with greater utilisation of the bones and breaking for marrow and consumption resulting in far more fragmentation. Similarly, the species representation in Trench 1 shows a faunal range more associated with domestic refuse in Iron Age Skye, for example similar to that at Buckquoy, Orkney (Ritchie, 1976) in particular with a greater percentage of cow and sheep/goat.

Figure 32: The distribution of species within contexts in High Pasture Cave

The graphs of species distribution confirm (see figure 32) that the Trench 1 deposits are not homogenous deposits, but exhibit individual variation visible between the layers. It is apparent that cow is in general the most prevalent species throughout the Trench 1 deposits, with sheep/goat and pig also proving significant.

The level of pig present within the Trench 1 contexts would be regarded as unusual within any normal domestic assemblage with the average for Iron Age Scotland being around 4% (for example Wemyss, Fifeshire (Ritchie, 1988) ) and at highest c.15% (for example Dun Vulan (Pearson, 1999) ). However, in comparison to the Context 001 Zone 3 deposit (pig-rich layer), with 89.2% pig and indeed Context 001 overall these deposits do show markedly less pig and fragmentation is also much greater with the contexts appearing far more “normal”. The Trench 1 deposits, because of their highly fragmented nature, exhibited few examples of working, although one artefact made of polished bone was recovered from Context 008.

It is important not to place too much interpretation on the Trench level deposits as so few were identifiable (the largest number of bones identifiable to species from a context being 51 from Context 003), however it is clear that there is an unusual frequency of pig in Context 006. Context 006 also shows a far greater variety of skeletal elements than the other Trench 1 contexts with carpals, calcanei, astragali and many more phalanges joining the predominate shaft fragments and vertebrae of other contexts. Because of the limited number of identifiable pieces it is impossible to interpret confidently what the concentration of pig elements here signifies, especially as it is not obviously related to either of the “floors” within Trench 1, but it may suggest an earlier deposit similar to the Zone 3 event. The bones within this section are not complete as they were in the Zone 3 assemblage, and either clearance of the “complete” bones occurred before Context 005 was deposited, or a different explanation must be considered. 

From Trench 1 we can conclude that there are variations in the composition of differing contexts, suggesting that the use of the cave changed over time. There is little dating evidence from within Trench 1, making it impossible to determine the length of time over which these contexts have been laid and how they relate to each other. All of the contexts other than Context 001 and 006, despite their variations, do appear to represent normal domestic refuse linked to the preparation of food, and all are clearly very different in character to Context 001 (the Zoned deposits). The Trench 1 assemblage provides a nice comparative, emphasising the unusual aspects of Context 001 and indicating that the use of the cave changed drastically from a place of domestic refuse disposal and normal consumption to a place where more thoughtful and intentional deposits were made.


Next entry: Specialist Report 2005 - The Mammal Bone Assemblage (Butchery at High Pasture Cave)

Previous entry: Specialist Report 2005 - Report on the Human Remains