The end of points

INTRODUCTION

For much of this century archaeologists working in northern Australia have perceived two types of stone points and have pondered the meaning of those types. In 1935 Davidson described two point forms, based on cross-section but incorporating a separation of bifacial from unifacial forms (Davidson 1935:160-162). The use of a bipartite division, between unifacial and bifacial points, was continued by McCarthy and Setzler (1960), Flood (1970) and others. Almost 50 years after Davidson's paper, Schrire (1982:246) expressed the question that is to be addressed in this paper, saying that,

Authors distinguish bifacial and unifacial points. The question arises as to whether this dichotomy is simply a function of the manufacturing process in that unifacial points may simply be incomplete bifacial ones.

ALTERNATIVE MODELS

Two models describing the relationship between bifacial and unifacial points are illustrated in Figure 1. The DIVERGENCE MODEL posits that the two point forms are independent and mutually exclusive end products, constructed using different procedures from a common antecedent form or unretouched flakes. This vision identifies different forms as the result of complex, branching trends analogous to the phylogenetic patterns labelled cladogenesis in organisms. In contrast, the SEQUENCE MODEL perceives a unilinear trend, with unretouched flakes manufactured into unifacial flakes that in turn are converted into bifacial points. This model identifies different forms as stages within a single lineage, analogous to a phyletic structure labelled anagenesis in organic evolution.

Figure 1. Two models of the relationship of bifacial and unifacial points.

Although both models have been considered in the literature, it is the divergence model that is favoured by most researchers. A few examples will illustrate the recent application of the divergence model to stone points in northern Australia.

In many instances the preference for the divergence model is based on analysis of point raw materials and dimensions. For example, in addressing the possibility of unifacial points in Arnhem Land rockshelters being incomplete bifacial points Schrire (1982:246) observed that,

...the Jimeri sites suggests that the dichotomy reflects discrete types, because quartzite points are usually bifacially worked, while unifacial ones are often made of quartz.

A detailed analysis of all broken points in the site showed them to divide into two main types, the unifacially and bifacially trimmed, but differentiated by other characteristics as well as the location of their secondary trimming. Thus the bifacials tended to be smaller, but more symmetrical and more fully-trimmed than the unifacials. A whole range of intermediate varieties also occurred, but these largely proved to be variants of the normal unifacial points. (Flood 1970:27).

...the only technical difference among prehistoric and recent points in Australia is between unifacial points ('points', 'unifacial points', and 'pirri points') and bifacial points ('bifacial points' and 'Kimberley points').

The attempt to change a plano-convex cross-section into a bi-convex cross-section resulted from a deliberate technological change designed to produce a 'bifacial point'. To produce a point with a bi-convex cross-section, the knapper had to establish platforms on the preform by first removing flakes from the ventral surface...and then abrading the margins. This action moved the margins of the preform toward the middle of its mass so that flakes could be removed successfully from both faces...This technique produces a true bifacial point...

POINTS IN KAKADU

Jimeri II, a rockshelter in the Arnhem Land escarpment excavated by Schrire (1982), was selected for examination. The archaeological assemblage recovered from Level I contains 187 points, dated to the last 5,000 years. This site is ideal for the analysis presented here, not only because of the large sample of points, but also because it is one of the sites used by Schrire to argue for a divergence model of point typology.

Points from this site were examined for information about the order of retouch. Both broken and complete points were included in the analysis. Since virtually all of the broken specimens were butts each specimen can be taken to represent a point, with no danger of double counting two fragments from the one specimen. Only quartzite points have been analysed, to reduce the effects of raw material properties on retouching patterns. As noted above quartzite points are commonly bifacially flaked, whereas points made on other materials such as quartz are not. Using a 5x hand lens each specimen was examined, and where the order of retouching could be reliably determined, the face that had been flaked first was recorded (Table 4). Specimens for which the ventral and dorsal faces could not be distinguished were eliminated from consideration. This usually occurred in the most fully retouched bifaces, where the entire surface of both faces was completely covered by scars and there was a lenticular cross-section.

In most specimens it is obvious that the complete history of retouching could be recorded in this fashion, and that only two phases of retouching have occurred: removal of flakes from one surface followed by the removal of flakes from the opposite surface. This was typically demonstrated by remnants of the unretouched ventral or dorsal surfaces indicating that much of the retouching process was recorded by flake scars still visible on the specimen. Exceptions may exist where both faces were extensively covered by invasive flake scars. In these cases evidence for earlier marginal retouch may have been removed by later retouching, and the observations in Table 4 would then relate to the penultimate rather than the first face to have been retouched. While it is likely that this mechanism may sometimes make my analysis inaccurate, in the vast majority of specimens it obvious that a hidden third phase does not exist. As noted above, where retouching was sufficiently extensive to make interpretations unreliable the specimen was excluded from the analysis. Consequently the data in Table 4 is considered to be a reliable indication of manufacturing patterns.

Most unifacial points were retouched on the dorsal face, although one specimen was retouched only on the ventral face. Bifacial points display a broadly similar pattern, with most having been retouched on the dorsal-face first, and only then having flakes removed from the ventral surface. Seven bifacial points display initial retouching on both faces. This occurs primarily because on those specimens one margin has a dorsal-first pattern, while the other lateral margin shows a ventral-first pattern.

Clearly, the bifacial points do not generally show the ventral-first retouching pattern predicted by Flenniken and White (1985). Indeed only on one-eighth of the bifacial points was manufacture begun retouching the ventral surface. This observation has several implications. Firstly, the archaeological data demonstrate that removing flakes from the ventral surface prior to removing them from the dorsal face it is not a technical requirement for the creation of a bifacial point; it is not even a common procedure for prehistoric knappers at Jimeri II. Secondly, a ventral-first process clearly does exist, and it coexists with a dorsal-first procedure. Hence there are several ways of retouching a flake to form a point (ventral-first, dorsal-first, and both-dorsal-and-ventral-first on different margins). These patterns may simply be variants of a flexible technology, used as when appropriate. More detailed analyses of these assemblages are being carried out by the author to establish the interpretation of this technological diversity within bifacial points.

Another implication of this data is its support for the sequence model in which unifacial points can be transformed into bifacial points. Bifacial points are frequently made by flaking the dorsal face, and then the ventral. This means that prior to the retouch of the ventral face the specimens were typical of unifacial points. Such a conclusion applies to not only specimens with dorsal-first retouch, but also to those with a ventral-first pattern of retouch. The unifacial point retouched on the ventral face has bifacial counterparts that have been retouched first on the ventral face. Moreover, the frequency of ventral-first patterns on unifacial points (9%) is replicated on bifacial points (11%). Consequently, transformations from unifacial to bifacial points could have occurred on every bifacial specimen represented in Table 4.

Reworking of existing unifacial points to form bifaces is therefore technically feasible, and on the basis of this data I would consider it likely. The amount of time points spend in a unifacial form before being transformed, and the functions for which they were used (if any), are not currently known. Specific investigations will be necessary to define the nature of the transformation. Similarly, these issues need to be examined in other regions beyond Kakadu if we are to determine the widespread applicability of the sequence model. As a contribution to these explorations a second study, located in northwest Queensland, is presented here.

POINTS AT LAWN HILL

At Lawn Hill Station, in the Gulf of Carpentaria, unifacial and bifacial points were manufactured from greywacke that cropped out as low hills on an otherwise flat plain (Hiscock 1988a). Retouching flakes to form points was an activity not carried out on the greywacke quarries, and it is likely that retouching of the points continued as they were transported around the landscape, broken, re-shaped and re-hafted. A study of 35 points collected from sites and background scatter on the plain indicates that the retouching process was generally undertaken in a particular sequence. Table 5 shows that it was common for points to be flaked on one or two margins but relatively uncommon for them to be flaked on both lateral margins and the butt. Moreover, the table shows that the more margins flaked on a point, the more likely it was that these margins would have been flaked bifacially.

All of the unifacial points have retouch scars only on the dorsal surface. If this represents the first phase of point flaking, then bifacial points manufactured from unifacial points should consistently show that the flaking of the ventral surface is superimposed upon flaking of the dorsal face. Table 6 shows that this pattern is found in the sample of points studied.

Points at Lawn Hill show clear signs of increased reduction away from the greywacke outcrops. The average size of greywacke points decreases from 8.6 cm near the quarries, to 6.5 cm in the gorge country 20-30 km south of the quarries (N=65). This trend is accompanied by an increase in the proportion of bifacial points over unifacial points, because the latter are worked down to become bifacial (Figure 2).

These data suggest that when flakes were removed from the greywacke quarries they were usually retouched by removing flakes from along one lateral margin on the dorsal surface. At a later time the second margin might have been flaked in a similar fashion, and eventually one or both of these margins might have been made bifacial by the removal of flakes from the ventral surface. Thus at Lawn Hill there is a technological continuum from unifacial to bifacial points, as the former are gradually transformed into the latter.

Figure 2. Historgram showing the increasing proportion of points with greater distance to raw material source at Lawn Hill.

There is no technological evidence to suggest that only the points flaked on both margins and both faces were used and that points with less retouch are merely preforms, the scheme developed for North American points (eg. Crabtree 1966). Since points at Lawn Hill were not invariably flaked on both margins and both faces before being carried around the landscape, it is likely that most of these point forms were functional and that the variation in form resulted from re-sharpening and re-shaping. Use-wear investigations might test the proposition that points at all stages of manufacture have been used.

DISCUSSION

One so-called 'type' merges continuously into another, and to demarcate boundaries between them seems artificial. (Macintosh 1951:200).

CONCLUSION

The end of points is best described by a sequence model. When their time span as unifacial points is finished, they may often be reworked rather than discarded. In this way many unifacial points have been transformed into bifacial points. Consequently, while some unifacial points were discarded in that state, other unifacial points were transformed into bifacial points before they were discarded. Perception of this transformational process raises the possibility that bifacial points may have been transformed into other artefact forms. This issue will be examined in a separate paper. Transformation of one form into another has implications for the explanation of assemblage variability. In western Arnhem Land it appears that intersite variability in the point forms at least partly reflects dissimilar rates for the transformation of unifacial to bifacial points in different environmental and economic contexts. Such transformations deserve wider study across northern Australia.

ACKNOWLEDGEMENTS

I thank the Museums and Art Galleries of the Northern Territory, and particularly Kim Akerman and Norma Richardson, for providing access to the Jimeri II collection. I thank Kim Akerman, Sally Brockwell, Robin Gregory, Scott Mitchell, and Ian Walters for their comments on the draft of this paper.

REFERENCES

Allen,H. and G.Barton n.d. Ngarradj Warde Djobkeng. Oceania Monograph 37.

Brockwell,C.J. 1989 Archaeological investigations of the Kakadu wetlands, northern Australia. M.A. thesis, Australian National University.

Crabtree,D.E. 1966 A stoneworker's approach to analyzing and replicating the Lindenmeier Folsom. Tebiwa 9(1):3-39.

Davidson,D.S. 1935 Archaeological problems of Northern Australia. Journal of the Royal Anthropological Institute. 65:145-184.

Flenniken,J.J. and J.P.White 1985 Australian flaked stone tools: a technological perspective. Records of the Australian Museum 36:131-151.

Flood,J.M. 1966 Archaeology of Yarar Shelter. Unpublished MA thesis. Australian National University.

Flood,J.M. 1970 A point assemblage from the Northern Territory. Archaeology and Physical Anthropology in Oceania, 5(1):27-52.

Hiscock,P.D. 1988a Prehistoric settlement patterns and artefact manufacture at Lawn Hill, Northwest Queensland. Unpublished PhD thesis, University of Queensland.

Hiscock,P. 1988b A cache of tulas from the Boulia District of Western Queensland. Archaeology in Oceania 23(2):60-70

Jones,R., and I.Johnson 1985 Deaf Adder Gorge: Lidner Site, Nauwalabila 1. In R.Jones (ed.) Archaeological Research in Kakadu National Park. Pp.165-228. Australian National Parks and Wildlife Service Publication 13. Canberra: Australian National University.

McCarthy,F.D. and F.M.Setzler 1960 The archaeology of Arnhem Land. In C.P.Mountford (ed.) Records of the American - Australian Scientific Expedition to Arnhem Land, Vol. 2. Melbourne University press. Pp.215-295.

Macintosh,N.W.G. 1951 Archaeology of Tandandjal Cave, South-west Arnhem Land. Oceania 21:178-213.

O'Connor,S. 1990 30,000 years in the Kimberley: a prehistory of the islands of the Buccaneer Archipelago and Adjacent mainland, West Kimberley, Western Australia. Phd Thesis, University of Western Australia.

Schrire,C. 1982 The Alligator Rivers prehistory and ecology in western Arnhem Land. Terra Australis 7. Department of Prehistory, Australian National University.

Smith,M.A. and B.J.Cundy 1985 Distribution maps for flaked stone points and backed blades in the Northern Territory. Australian Aboriginal Studies. 1985/2:32-37.

Information on the volume that contains this paper.

HOW TO ORDER ARCHAEOLOGY IN THE NORTH

You are visitor to this page.