This art icle was downloaded by: [ Universit y of Arizona] On: 29 April 2013, At : 12: 42 Publisher: Rout ledge I nform a Lt d Regist ered in England and Wales Regist ered Num ber: 1072954 Regist ered office: Mort im er House, 37- 41 Mort im er St reet , London W1T 3JH, UK

World Archaeology Publicat ion det ails, including inst ruct ions for aut hors and subscript ion informat ion: ht t p:/ / www.t andfonline.com/ loi/ rwar20

The deer hunters: Star Carr reconsidered a

a

a

John M. Andresen , Brian F. Byrd , Mark D. Elson , Randall H. a

a

McGuire , Ruben G. Mendoza , Edward St aski a

a

& J. Pet er Whit e

b

Universit y of Arizona, (Tucson)

b

Universit y of Sydney Published online: 15 Jul 2010.

To cite this article: John M. Andresen , Brian F. Byrd , Mark D. Elson , Randall H. McGuire , Ruben G. Mendoza , Edward St aski & J. Pet er Whit e (1981): The deer hunt ers: St ar Carr reconsidered, World Archaeology, 13:1, 31-46 To link to this article: ht t p:/ / dx.doi.org/ 10.1080/ 00438243.1981.9979812

PLEASE SCROLL DOWN FOR ARTI CLE Full t erm s and condit ions of use: ht t p: / / www.t andfonline.com / page/ t erm s- and- condit ions This art icle m ay be used for research, t eaching, and privat e st udy purposes. Any subst ant ial or syst em at ic reproduct ion, redist ribut ion, reselling, loan, sub- licensing, syst em at ic supply, or dist ribut ion in any form t o anyone is expressly forbidden. The publisher does not give any warrant y express or im plied or m ake any represent at ion t hat t he cont ent s will be com plet e or accurat e or up t o dat e. The accuracy of any inst ruct ions, form ulae, and drug doses should be independent ly verified wit h prim ary sources. The publisher shall not be liable for any loss, act ions, claim s, proceedings, dem and, or cost s or dam ages what soever or howsoever caused arising direct ly or indirect ly in connect ion wit h or arising out of t he use of t his m at erial.

The deer hunters: Star Carr reconsidered

Downloaded by [University of Arizona] at 12:42 29 April 2013

John M. Andresen, Brian F. Byrd, Mark D. Elson, Randall H. McGuire, Ruben G. Mendoza, Edward Staski and J. Peter White

i Excavated in 1949-51 and fully published in 1954 (Clark 1954), the British Mesolithic site of Star Carr has been seen as an archaeological classic in excavation and interpretative techniques for more than two decades, as generations of Cambridge University-trained archaeologists have testified (Sieveking et al. 1976). It has served as a cornerstone around which broad interpretations of seasonal economic patterns in the eighth millennium B.C. have been built (Clark 1972; Mellars 1976; Jacobi 1978). The excellent preservation of organic remains has rivalled that at Mesolithic sites in northern Europe and has encouraged comparisons between them. More recently, several workers have re-investigated certain aspects of the site. In particular, N. NoeNygaard (1975, 1977) has re-examined the bone material from a taphonomic viewpoint, and A. Wheeler (1978) has looked into the absence of fish remains. More ambitiously, M. Pitts (1979) has completely re-interpreted the purposes for which the site existed. This paper also re-examines the purposes for which Star Carr existed, and we here point out that it is some tribute to the director of an excavation conducted under difficult conditions thirty years ago that his reports provide so much of the data which made possible this re-interpretation. In 1954, J. G. D. Clark interpreted the site of Star Carr as a settlement consisting of several families of men, women and children who occupied the site during midwinter and spring and returned to it at least once. His interpretation was re-emphasized in 1972. Star Carr was a 'winter settlement' occupied by a 'small hunting band which supported itself... by the culling of red deer stags supplemented by hunting . . . other game, and which took advantage of a period of settled existence to replenish the equipment needed during the course of the year' (Clark 1972:31). He also suggested that the site was probably occupied for around five months in the year (1972:23). This interpretation has been widely accepted (e.g. Mellars 1976; Jacobi 1978; Thomas 1979). Recently Pitts (1979:33) has re-interpreted Star Carr as a 'specialized industrial complex' rather than a complete settlement. He claims that most of the material can be related to antler working and to skin processing and tanning, with a smaller quantity deriving from other activities carried on at a nearby settlement. He infers that the industrial aspects of the site derived from summertime use, although agreeing that some debris did come from winter use also. World Archaeology

Volume 13 No. 1 Miscellany

© R.K.P. 1981 0043-8243/81/1301-031 $1.50/1

Downloaded by [University of Arizona] at 12:42 29 April 2013

32

John M. Andresen, et al.

Our view is that both previous interpretations misconstrue some important evidence from the site. Considered in the light of site formation processes, this evidence suggests that Star Carr was primarily a hunting stand and butchering station used for very short periods at various times of the year. The total accumulation of material may have taken tens or possibly hundreds of years. This interpretation is derived from two rather different lines of argument. On the one hand, there are severe problems with viewing Star Carr as a long-term winter campsite or a specialized industrial complex. We consider the absence rather than the presence of some data, and the implied reasons for this, to be significant. On the other hand, the application of recent ethnoarchaeological, taphonomic and use-wear studies, along with consideration of site formation processes, strongly suggests that the Star Carr site resulted primarily from human behaviour associated with hunting. We recognize that some aspects of our interpretation are not as robust as we would wish. However, it has not been possible to re-examine the original documentation and collections, which could support some of our views and which could certainly test between various interpretations. We also suggest that the analytical approach taken in this paper may itself be of some interest. The recognition that archaeological sites are complex phenomena is not new. But an attempt to investigate the various processes which resulted in the accumulation, disposition, and differential destruction of the material remains at a particular site has not often been tried. We show how an approach along these lines results in a different and more probable view of the human behaviour which created Star Carr than is given by previous approaches. We look first at evidence for the seasonally and length of visits, then at major artifact and ecofact classes, and finally review the various interpretations of site use.

II Use at all times of the year. Occupation of Star Carr throughout the winter has always been stressed (e.g. Clark 1972:31). This view has been accepted by those (Mellars 1976; Jacobi 1978) who see the site as resulting from one aspect of a human seasonal cycle. But the fact that Star Carr fits so neatly into a model of seasonally patterned exploitation of the environment has led to a downplaying of other seasonal indicators. Perhaps the most important of these, since it is clearly non-cultural, is Walker and Godwin's suggestion (1954:67) that reedswamp growth at the site implies a late summer or autumn rather than a spring use. This is supported by Corner's interpretation (Clark 1950:124) of the probable collection season for the bracket fungus, and by Fraser and King's comment (Clark 1950:128) that the white stork found at the site is a summer visitor (note that the identification is once given as tentative, 1954:70). Site use earlier in the summer is indicated by the presence of young red deer and elk (Noe-Nygaard 1975), while roe deer crania with attached antlers may have been collected at any time throughout the wanner period (Fraser and King 1954:75; see also Pitts 1979:fig. 3). A summer use is possibly implied by the presence of birch bark rolls if these were collected for pitch, since this is the season when the sap flows most strongly (Clark 1954:166). A spring use may be supported by the occurrence of four red deer crania with shed antlers, though since Phragmites continued to grow on and around the platform during the occupation (Walker and Godwin 1954:58), a very short-term use is likely. An indicator of seasonality which has not to

Downloaded by [University of Arizona] at 12:42 29 April 2013

The deer hunters: Star Carr reconsidered

33

our knowledge been examined is the tooth eruption patterns of the main game animals represented. A major reason for believing in winter use of the site has always been the large number of red deer antlers present, particularly shed antlers which are best collected in April. However, only 41 (39 per cent) of 104 antlers listed were collected shed (Clark 1954:119-22), and, as Pitts (1979:37) remarks, these could have been stored. The antlers collected unshed could have been acquired at any time from late summer through until April (Fraser and King 1954:94). The argument that these unshed antlers were probably collected during winter is based especially on their use for points, and the fact that antler becomes increasingly compact and therefore presumably better raw material as autumn and winter continue (Jacobi 1978:318). We do not know, of course, that the inhabitants of Star Carr regarded compact antler more favourably for manufacturing. We do know that they utilized the most compact (i.e. shed) antler no more intensively in terms of either the numbers of splinters removed or the proportion coming from them (table 1). Indeed, if the numbers of grooves from which the splinters have not been removed (unshed 14, shed 5) are added to the numbers of splinters, the percentage of splinters from each kind of antler is almost the same as the percentage occurrence of that kind of antler in the assemblage. We therefore suggest that while shed antler may document site use in winter, the unshed antler does not do so directly. The other evidence for winter use is the two elk skulls without antlers. These must have been collected in January or later (Fraser and King 1954:93). TABLE 1 Number of splinters actually removed from red deer antlers. Source: Clark 1954:119-22 No. of splinters 0* 1 2 3 4 5 6 Total % of total

Unshed antlers

Shed antlers

11 9 13 18 4 4 4

8 7 11 12 13 — -

149

77

66

34

* Includes specimens for which the number is not deteiminable.

We conclude that both plant and animal remains definitely imply use of the site during the summer and autumn and probably imply a spring use. Animal remains may imply some winter use but the direct evidence is not strong. There is little reason to believe use in one season predominates. The duration of visits. Problems relating to the lengths of time for which discontinuously occupied sites were actually occupied have been an on-going concern of archaeologists, and no general solutions have been found. But for some hunter-gatherer sites at least, help may be at

Downloaded by [University of Arizona] at 12:42 29 April 2013

34

John M. Andresen, étal

hand from ethnoarchaeological studies which demonstrate that the spatial organization of debris bears some relation to the period of site use. Working among the San people of Namibia, who camp at each site for only a few days, Yellen (1977a) observed no spatial segregation of any class of refuse. This seems to be supported in a general way by Gould's studies among Aborigines of the Western Desert, Australia (1977:29-46). Similar behaviour has been observed in completely different environments, for instance by McKellar in his study of urban littering (Rathje 1979:9). In his analysis of a short-term hunting stand, L. Binford (1978a) does distinguish between a 'drop' zone and a 'toss' zone of debris, but the distance larger debris are 'tossed' is only a metre or two. If similar behaviour to the last occurred at Star Carr, the published report would not let it be distinguished. In striking contrast to these patterns, among the Alyawara of Central Australia who camp at sites for periods of a month or considerably longer, O'Connell (1977) observed a clear spatial patterning of refuse, with debris larger than 9 cm. being regularly removed well away from the main living and sleeping areas. The implication of these studies is that we should expect little or no re-organization of debris ('secondary refuse' (Schiffer 1972)) in sites used for a few days only, while we may expect such reorganization in sites used for a longer period (see also Murray 1980). In trying to assess the Star Carr site in the light of these studies, it has become apparent that the published data on the horizontal distribution of material is not as complete as we would wish. The published plans focus on artefacts, while we would like to look at the scattering of such debris as broken bones. We look first at available data on the relationship of debris to the birchwood platform, apparently the centre of activities at the site. We then look at areal patterning within artefact and debris classes. Clark (1954:4-23) claims that there is a general congruence between all classes of remains and the platform, but also provides hints of some variation. For example, while antler points (fig. 5), axe-sharpening flakes (fig. 9) and microliths (fig. 12) show a very marked correlation with the lakeside edge of the platform (taken as zone M), about 20% of the cores are found on the lakeward side of this. Regarding animal bones Clark says (1954:15) that 'this material came from barely two-fifths of the area of settlement as defined by such imperishables as worked flint', but in figure 2 he plots animal bones as occurring in zones N-T, well out into the swamp above the birch trunk and in front of the platform. The distribution of cores and animal bones thus suggests limited use of the swamp as a garbage dump. Most material, however, was clearly dropped within the area of major human activity, defined by the platform and the occurrence of in situ flaking. Within the area of concentrated activity, the published data show a relatively even scatter of debris. This occurs with all classes of mapped inorganic remains except beads, and with organic remains within the area of their survival (Clark 1954: figs. 3-5, 9,11-14). Some distributional variability does exist: four areas with higher densities of 'worked flints' and some artefact classes may be observed, each about 3-4 yards in diameter. In two of these 'clusters', microliths and burins are associated; in a third, microliths and scrapers; and the fourth consists of burins. These 'clusters' are, however, only areas of somewhat higher density within an overall scatter of similar artefacts. They do not imply the clear spatial segregation of specialized industrial activities such as is commonly found in single long-term occupations (O'Connell 1977). One of the clear lessons of recent ethnoarchaeological studies is that the spatial patterning of garbage can reveal temporal information. The fact that most of the sharp flints and antler points, as well as such large objects as elk-antler mattocks and red-deer antlers, were left on the

The deer hunters: Star Carr reconsidered

35

Downloaded by [University of Arizona] at 12:42 29 April 2013

platform rather than being cleared away off it into the swamp, implies a series of short visits rather than a few long ones. Some confirmation of this, we think, comes from another aspect of the site, namely the low number of hearths and low quantity of charcoal. Clark (1954:11-12) does not identify any of the charcoal patches as definite hearths, and at least three of the six are off the platform and in the swamp, implying the dumping of old ashes rather than in situ fireplaces. The other three are only 'shallow lenses' of charcoal. Jacobi (1978: fig. 6) interprets all the patches as hearths but they show none of the well-developed character, with a thickness of ash and burned earth, which would be expected of hearths kept burning for several winter months (J. O'Connell, pers. comm.). Fires known at hunting sites (Binford 1978a) tend to be on the small, short-term kind that the remains at Star Carr seem to imply. Use over a number of years. It has been already accepted that the site was occupied on at least two occasions separated by some appreciable interval (Clark 1972:21). But the problem of distinguishing between two long-term uses and two or more series of short-term ones has not been addressed. Clark recognized (1954:9) that trampling of material would undoubtedly occur and we would note that debris is likely to have been moved vertically by such natural processes as freezing and thawing cycles. There is also the possibility of sinkage within the soft lake muds. Thus we think it unlikely that stratigraphie data would allow us to distinguish between a few and numerous visits. The existence of at least two major episodes of site use was best expressed by Walker who said in 1950: 'The depth of deposit which lies between the main timbers of the site and the highest artefacts certainly suggests a long period' (Clark 1950:121). Vertical separation occurred between the 'spreads' of charcoal (1954:9) and with the antler point classes A and E (1954:10, 126; 1972:21), although only nine whole, clearly-identified specimens of these two classes exist (1954: table 17, schedule). Less clearly defined separation occurs between coarse, medium and finely barbed points (1954:127). From the presence of at least three barbed points beneath the trunk or branches of the felled birch tree (1954:4), we deduce some use of the site prior to this event, while there is even a hint in the only published stratigraphie section (1954: fig. 2; cf. 1949: plate VIII) that flints and animal remains tend to lie above the tree, while many of the points lie within and below it. This suggestion needs support from re-analysis of the field data. There is also the enigmatic 'upper birch level' (1954:2) which appears to have consisted of birch stems with their bark intact sitting horizontally in the swamp at a level above the main platform. Only two unworked bones were found above this feature (1949:56), and Clark seems therefore to have dismissed it from consideration. In the final report, Walker and Godwin do not comment directly on it, and while a natural rather than cultural explanation for its existence may have been developed, such has not so far appeared in print. It may have been another platform. We conclude that the site was used over a considerable period of time, but the available stratigraphie data do not allow us to distinguish between two long-term and multiple shortterm visits. Either interpretation seems to us to be equally possible.

m We turn now to consideration of three major classes of debris — animal bone, flaked stone and antler points — and the activities which these imply.

36

John M. Andresen, et al.

Downloaded by [University of Arizona] at 12:42 29 April 2013

Animal bone. It was noted in the original site report that different numbers of the body parts of various animals existed (Fraser and King 1954: ch 3), but no interpretation of this variability was made. Jacobi (1978:316) recently suggested that it indicated that animals were being killed and butchered elsewhere and only some parts of them were being brought back to the site. However, in 1974 N. Noe-Nygaard re-analysed the Star Carr bone and drew attention (1977) to three important characteristics. She observed that a) much of it was butchered and showed extensive evidence of marrow fracturing, b) some of the bone was chewed probably by dogs, and c) bone from this site was less well preserved than bone from other Mesolithic sites used in her analysis and was 'in a very bad state due to décalcification' (1977:220). Her observations raise the possibility that processes occurring at the site may be responsible for much of the variation in animal bone occurrence. Within the last decade, several studies (Brain 1976; Yellen 1977b; Binford and Bertram 1977; Binford 1978a, 1978b; Behrensmeyer 1978) of butchering and other processes affecting bone have shown that the faunal assemblage present at a site is the outcome of at least three sequentially operating factors, namely what is brought to the site, what happens to it in the systemic context (Schiffer 1972) of site use, and what is preserved in the archaeological context. Although the operation of each of these factors varies with local circumstances and there is no formula which will predict what a faunal assemblage should look like in any particular situation, some relevant generalizations have been made and can be applied to the Star Carr material. We will start with the archaeological assemblage and the most recently operating factor. It is clear from the work of Behrensmeyer and other taphonomists that fluctuation in temperature and moisture are the principal causes of weathering and that the density of particular bones and, importantly, parts of bones, along with the degree of fusion of epiphyses are the major determinants of their survival of weathering processes. We assume here that décalcification has the same effect as weathering in terms of differential survival of bones, as is implied by Noe-Nygaard (1975:220-1). Bones most strongly affected by weathering include ribs and vertebrae, along with small fragments, all of which are rare at Star Carr. Juvenile (unfused) bones are also at far greater risk than adult bones, and Noe-Nygaard (1975:221-2) notes very few of them. We therefore suspect that differential survival of bones may be occurring at Star Carr. One test of this proposition, suggested by D. Gifford, would be to examine the ratio of compact to cancellous bone in the skeletal elements which are preserved, since it should be higher in them than in those elements not present. We turn now to the second factor. While the site was in use, at least two attritional processes were operating — much of the bone was broken open to get at the marrow, and dogs were present (Clark 1972:9). Both of these processes also operate selectively and may be partly responsible for the configuration of the Star Carr assemblage. Regarding bone fracturing, NoeNygaard (1977) points out that this increases the number of bone fragments, reduces their size, and exposes softer parts to weathering. Softer and smaller bone fragments are thus less likely to be present in the archaeological assemblage. Long bones generally have more and better marrow than other bones, and those at Star Carr were nearly all fractured. Once this has occurred, the dominance of distal fragments of long bones (table 2) becomes partly explicable in terms of differential destruction. Concerning dogs, Brain (1976) notes that their chewing affects bone differentially, and that they destroy in particular such blood-rich parts as proximal ends of femurs and humeri. They may totally consume such small bones as phalanges. Brain's study of the effects of weathering

The deer hunters: Star Carr reconsidered

37

TABLE 2 Long bone fragments at Star Carr (number) and from a modern assemblage (survival expressed as a percentage of the original number). Sources: Noe-Nygaard (1977) and Brain (1976) Starr Carr Distal Proximal

Downloaded by [University of Arizona] at 12:42 29 April 2013

Humérus Radius Metacarpal Femur Tibia Metatarsal

21 34 41

0 17 17

very low 2 13

29 38

Namib Desert goats Proximal Distal 0 51 25 14 10 30

64 17 18 7 56 15

and dogs on a population of goat bones in the Namib Desert shows that great differences in survival chances can occur between different parts of the same bone (table 2). In a study similar to Brain's, Binford and Bertram (1977) show that the density of bones and parts of bones is a critical variable in their survival, and that where, in their studies, dogs had unrestricted access to total animal remains, the frequencies of surviving bones match well with those observed by Brain. The fact that some, although 'no great quantity' (Noe-Nygaard 1977:217), of the bones at Star Carr were gnawed, implies that the present assemblage is partly the result of canine attrition. But whether dogs were only given parts of animals, as sometimes occurs among Eskimo hunters (Binford and Bertram 1977:82-3), cannot be determined without re-examination of the fauna. The third major factor to be considered is what is brought to the site. This is difficult to determine until attrition factors have been accounted for, but the fact that all parts of the larger animals do occur at Star Carr (though in low numbers: Noe-Nygaard 1977:221-3; Fraser and King 1954:75-89 passim) suggests that butchering some distance from the site was not common. We believe that most animals were killed at, or near, the site and were first butchered there, and will now suggest how this proposition may be tested. Recent studies among the Nunamiut (Binford 1978b), Calling Lake Crée (Bonnischen 1973) and San (Yellen 1977b) suggest that where hunted animals have to be transported but must be butchered first, the hunters generally eat portions which have a low meat content and will also eat the marrow from many bones. They therefore tend not to transport such bones as vertebrae and metapodials away from the butchering location and back to camp. Selective transport of bones does not occur with smaller animals which can be transported whole. Thus if animals were being transported to Star Carr, we would expect bones of larger animals (red deer, elk, bos) to be differentially represented but the various bones of smaller animals (roe deer, beaver) to occur in approximately equal numbers. There are only hints that these differences do not occur at the site; hints which suggest that, prior to attrition, all bones did occur on the site. If so, the animals were probably killed very close by. To summarize this section, we think that nearly all bones found at Star Carr came from animals that were killed nearby and butchered there. The bones were then subject to attrition by both human and dog activity, and have subsequently been differentially preserved. Some indication of the presence of these factors can be gained from the reports, but their proportionate role cannot be assessed without restudying the actual material.

Downloaded by [University of Arizona] at 12:42 29 April 2013

38

John M. Andresen, et al.

Flaked stone tools. Both in the original and later reports, these artefacts have been assumed to indicate both that certain tasks were carried out, and that different sexes were involved in them. However, the purposes served by the flaked stone tools at Star Carr, as with most prehistoric tools of this kind, have been guessed at rather than properly inferred. For the English Mesolithic, it seems reasonably clear that backed microliths are indicators of archery (Clark 1954:103; Jacobi 1978), while at Star Carr burins were almost certainly used for the removal of splinters (point blanks) from red-deer antlers. Burins may also have been used as shavers, the side edges of the burin facet being both strong and sharp, and suitable for working wood, bone and antler (Bordaz 1970; Hayden 1977:185; Newcomer 1974). We may be able to determine the uses of Star Carr burins by examining their edges microscopically. The scrapers are a more controversial part of the assemblage to interpret. As Keeley (1977: 108) points out, the term has been used for many years without any real evidence of function. Clark (1954:21; 1972:31) suggested that these tools were 'well adapted' to preparing animal skins, which must have been commonly available, and also argues from this to the presence of women at Star Carr on the basis of ethnographic analogy. This interpretation is implicitly accepted by Jacobi (1978:315) and Pitts (1979:34). It is certainly clear from modern studies that stone scrapers are often used on skins and hides, especially for cleaning them (e.g., Nissen and Dittemore 1974; Hayden 1979). But there are also many instances of morphologically similar tools being used for woodworking (Gould, Koster and Sontz 1971 ; Hayden 1977:182). Other observations show that skins may be prepared using flat slate blades, smooth stones, stone flakes, bones and shells (Mulvaney 1975:93-4). Archaeologically, Gould; Koster and Sontz (1971) have given some reasons for believing that Quina-type scrapers were used for woodworking, while Keeley (1977:126) considers that at least some of the side scrapers from Hoxne were used for skin scraping. Given proper equipment and methods, Keeley claims to be able to distinguish between woodworking, meat cutting and skin scraping uses, though it is not clear that all workers in the microwear field would fully accept the current claims. As an illustration of the problems involved, Cantwell (1979) detects gross morphological differences which correlate with different wear patterns on scrapers in two Illinois sites, but cannot demonstrate that the wear pattern differences derive only from the raw materials being worked (see also Kamminga 1978). It is also not entirely clear that if animal skins are being processed then women must be present, or that only males are involved in woodworking (Hayden 1977). We are therefore not persuaded that the Star Carr scrapers must have been used only on skins. We suggest that scrapers may have been used on wood, antler and bone, even though not a great deal of this material has been recovered. Once again, microwear studies may assist in the resolution of this problem. Craft activities involving a range of materials are, we suggest, more likely to be practised at a hunting site than is the specialized operation of skin scraping. Turning briefly to another aspect of the scrapers, Pitts (1979:34) lays considerable stress on their high number at Star Carr, and relates this to hide-working. However, although scrapers are important at Star Carr, Pitts' Principal Co-ordinates Analysis seems to us to over-stress this. Reconsidering the data on thirty-three Mesolithic sites, reported by Mellars (1976: table 2) and used by Pitts, we note that of the thirty-three, twenty-three have a lower proportion of scrapers in the tool assemblage than does Star Carr, while nine have a higher, in several cases much higher, proportion (e.g. Blubberhouses Moor, Flixton I, and Morton). Star Carr is hardly as unique as Pitts suggests, and Mellars classifies it as a 'Type B: Balanced Assemblage' rather than 'Type C: Scraper dominated' (1976:391,394).

Downloaded by [University of Arizona] at 12:42 29 April 2013

772e deer hunters: Star Carr reconsidered

39

It is also unclear to us that the other major tool group, 'awls', is necessarily related to skin working (Pitts 1979:34). Clark (1954:106) says that 'the points are often chisel-like in character and in some instances there is a small burin facet', which leads us to suspect that some awls may well have been used in a similar manner to burins. In some other examples the extreme tip of the tool has been worn smooth, a fact consistent with skin working but not diagnostic of it (Kamminga 1979). As Clark (1954:166) points out, the polishing on awls may have resulted from drilling holes in the amber and shale beads and animal teeth. Finally we note the many ethnographic situations in which a high proportion of tools used are not shaped to set and regular patterns (e.g., Gould, Koster and Sontz 1971;Hayden 1977; White 1968; White, Modjeska and Hipuya 1977). And it may be that among the 85 per cent of the collection showing 'no signs of secondary working or of use' (Clark 1954:96), some would reveal such under the microscope. Meat cutting, however, may leave not even microscopic wear traces (Brose 1975). We conclude that the flaked tools at Star Carr do not unequivocally support the uses which have been inferred for them. Their use by both men and women cannot be regarded as substantiated either. We suggest that the flaked stone tools may well derive primarily from craft manufacturing activities, mostly carried on by men. Antler points. Of all the activities recorded at Star Carr, the removal of splinters from red deer antlers and the manufacture of barbed points from them has always seemed to be most clearly documented. The material has, however, puzzling aspects, to some of which Jacobi (1978: 318-21) has already drawn attention. Most important, there are only a very few unworked splinters, and Clark says all of these are broken (1954:118). Of 191 points listed in Clark's schedule (1954:128-36), only five (P9, P27, P81, P136, P187) may be unfinished, and certainly no partly-made but broken points are recorded. As Jacobi says, it seems likely that point manufacture took place elsewhere, although we might suggest that the remains are consistent with a very high rate of manufacturing success, coupled with a low rate of recovery of small manufacturing debris. Still unexplained, however, is the number, nature and size range of the points at the site. The 191 points are a much larger collection than from any other Mesolithic site, although this may derive simply from better conditions of preservation here. Of these points, Clark (1954:123, 128-36) originally classified 57 as intact or very nearly so, 60 as tangs with or without barbs, 37 as tips or foreparts, and the remainder as stem fragments. (Three pieces are unclassifiable.) Tangs and tips are explicable as deriving from broken tools occurring in hafts and in animals respectively, but on any current interpretation of site use, including our own, the high proportion of whole points is curious. Are these points really whole? Re-examination of Clark's data reveals that only nineteen points are completely intact, although another three in poor condition may be so. Fourteen of the nearly intact points have lost their tip, sixteen have lost part of their tang and five have broken barbs. We suggest that most of these points, including those apparently whole, were in fact regarded as broken and were discarded. Only a few points were re-worked. We note also that their even scattering across the site suggests discard rather than caching behaviour. This pattern is more consistent with short-term than with long-term site use, and the nature of the points with craft activities directly associated with hunting. The other curious feature of the points is their size range (fig. 1). Clark originally suggested (1949:61) that these points may have been hafted and used in more than one way and while

40

John M. Andresen, et al.

there is as yet no archaeological documentation of this, the shortest points hardly seem suitable for the 'final despatch of animals' proposed by Jacobi (1978:320) as the function of all these tools. On the other hand there is no bimodality in the lengths, while the lengths of clearly whole points cover the entire size range (table 3). Further, the lengths of grooves left in antlers by the removal of sprinters display the same extended distribution as do the points (fig. 1). It is noticeable, however, that the grooves are rather longer than the points. Taken in conjunction with the low amount of reworking, we believe that the surviving points provide a true picture of the original population. We assume that they were armatures for projectiles, but whether spears, arrows, or both is not determinable. TABLE 3 Downloaded by [University of Arizona] at 12:42 29 April 2013

Lengths of clearly whole points and those with only barbs broken. Source: Clark 1954:128-36 53

4-4.9 3

Length (in.)

6 4

n

NUMBER 10" 98" 7" 6" 5" 4" 3" 2'

1. 1

i

1

i

•

10l

11 2

122

132

• * •

«•

Li]•• •*

1

1"

93

SPLINTCD