Weapons and wunan: production, function and exchange of Kimberley points.

Akerman, Kim ; Fullagar, Richard ; van Gijn, Annelou 等

Abstract: We describe the production, function and exchange of projectile points in the Kimberley region, northwestern Australia. Our aims are to identify problems in the interpretation of residues and use-wear, and to assess whether aspects of exchange can be inferred from microscopic study of these archaeological points. For the late prehistoric and contact periods, we identify change in production technique, artefact design and function. We also document microscopic traces (use-wear, resin, ochre, starch granules and apparent blood films) on glass, stone and ceramic points from museum collections and excavated artefacts. Although ethnography suggests that size and elaborate flaking may indicate production for exchange, we conclude that microscopic traces alone do not.

Introduction

Aborigines of the Kimberley produced delicately made stone spearheads for fighting, hunting and exchange. After the late nineteenth century, spearheads were also made of glass and ceramics. Some were produced solely for exchange. While a precise chronology for these stone, glass and ceramic points is unavailable, it seems likely that production and exchange of stone points span millennia (Bowdler and O'Connor 1991; Fullagar et al 1996; O'Connor 1999). In several places, production intensified in the twentieth century, before ceasing in recent decades (Akerman and Bindon 1995). The exchange network covered vast areas and recently incorporated tourists and other non-Aboriginal participants. In the last 100 years, large points were made for exchange, via a chain of partners called wunan or winan (Akerman 1979c; Kaberry 1939; McCarthy 1939) that connects with other networks (Stanner 1933-34). Smaller points functioned as spearheads but were also exchanged.

Thousands of Kimberley points are in museums, although their provenance is often poorly documented. To improve understanding of these collections, our aims are (1) to review the nature and context of recent point production; and (2) to assess whether we can infer function and exchange from points in museum collections. First, we review known point types, ethnographic and historic information, likely tool residues and possible indicators of exchange. Second, we analyse use-wear and residues on ethnographic and archaeological points.

Point types

There are two main classes of point: biface and uniface (see Akerman and Bindon 1995; Flood 1970; McCarthy 1967; O'Connor 1999:67ff). Dortch (1977) and McCarthy (1967) defined a further class called 'backed points', found in the Kimberley. Distinctions are based on manufacturing stages and the presence, location and techniques of marginal flaking. Biface and uniface points from the Kimberley were made with hard hammer percussion and sometimes with pressure flaking. The biface and uniface points from the Kimberley may have pressure-flaked projections with serrated or denticulated margins.

Four biface types have been proposed by Akerman and Bindon (1995): 'Wanji biface', 'Kimberley points', 'Northern Territory triangular points' and 'Kimberley dentate points' (Figure 1). The Northern Territory triangular points (known prehistorically only) and the Wanji biface (known ethnographically) occur mainly in western and southwestern Arnhem Land. Glass points modelled on the Wanji biface were made in northwestern Northern Territory. Kimberley points (known ethnographically) and Kimberley dentate points (known prehistorically only) were made in the Kimberley and widely distributed. Akerman and Bindon (1995) suggest that 'Kimberley point' should only be used for biface points that:

* are manufactured by pressure flaking, with a standard reduction procedure; and

* have pressure-flaked margins that are either denticulate or serrate.

[FIGURE 1 OMITTED]

The terms 'serrate' and 'denticulate' refer to margins with small projections or 'teeth', separated by notches of similar or smaller width (Figure 2). Serrated margins generally have about 6-8 teeth per 10 mm and denticulate margins have about 2-3 teeth per 10 mm (Akerman and Bindon 1995). Kimberley dentate points have margins where the teeth are separated by notches that are wider than the teeth.

[FIGURE 2 OMITTED]

Uniface points (Figure 3) have marginal or invasive flaking on the dorsal surface only. Flood (1970) and McCarthy (1967:38) refer to three classes: Pirri, Bi-marginal and Paroo. Pirri points, as distinguished by various criteria including invasive pressure flaking, occur in the Kimberley, where some are serrated. Margins of some Pirri points are only partially flaked, others are bifacially flaked, with some flaking to remove prominent bulbs of percussion. The term 'Pirri point' has 'become a synonym for any small unifacial point' (Flood 1970:47), so it does not constitute a special type. Nor do Bi-marginal points constitute a special type, because they include points with various degrees of marginal and invasive flaking that could reflect stages in the production of biface points (see Bradshaw 1986; Flood 1970; O'Connor 1999). It should be noted that Kamminga has demonstrated that the term 'pirri' in its original sense denoted a specialised engraving tool from the Eyre Basin, South Australia (Kamminga 1985).

[FIGURE 3 OMITTED]

McCarthy (1967:38) includes the Paroo among uniface points, and notes that it is common in the Kimberley and associated with leilira blade industries of northern and central Australia. This hardly constitutes a specialised type but is distinguished by McCarthy (1967) according to similarities with Levallois points. Following Bordes (1961), Dortch (1977:117) defined prehistoric Levallois points from Miriwun rockshelter in the east Kimberley as 'elongated triangular flakes whose forms have been predetermined by preparatory flaking of the flaking faces of special cores...In the Ord valley typical Levallois points often have three principal flake scar facets on their dorsal faces, the last produced of which removes the proximal part of the central ridge formed by the adjacent flake scar facets of the two initial preparatory flakes. Most Ord valley Levallois points are of the elongated form (Dortch and Bordes 1977), and most have plain butts though faceted butts are not uncommon' (Figure 3). These were the most common type of point in the Widgingarri shelters and O'Connor referred to them as 'laterally retouched points'. Ethnographic points of this type are not known in the Kimberley.

Dortch (1977:117) described prehistoric Kimberley backed points as thicker than bondi or other microlithic backed blades, and with backing retouch produced by direct percussion (not bipolar techniques). Often the backing retouch appears to be denticulated or notched. They are not known ethnographically but are present in excavations in the Kimberley, including Keep River (Figure 4).

[FIGURE 4 OMITTED]

While reduction stages suggest that unifacial points may be part of a transformation process in the life history of bifacial points for some regions of northern Australia (Flood 1970; Hiscock 1994), this does not appear to be the case for Kimberley points (see below). Certainly, research is required to establish the range of manufacturing processes for each type of point, and the recycling from one type to another and to other functional classes altogether. The range of functions for the ethnographic and archaeological point types found in the Kimberley has not been described in detail. Preliminary investigations suggest a variety of functions and that some were not used as spearheads at all (e.g. Dortch 1977; Wallis and O'Connor 1998; Kim Akerman, personal observations).

In archaeological contexts, distinctively pressure-flaked Kimberley points occur on the surface or in upper prehistoric levels, while other point types span roughly the last 4500 years (Bowdler and O'Connor 1991; O'Connor 1999).

We now review ethnography on point production to assess mythology, quarries, manufacture, exchange, and potential problems in interpretation of microscopic traces. Much of the information comes from unpublished notes of Kim Akerman (cited as KA); see also Goring (2000).

Mythology

Oral tradition and stories told to KA show how stone-point production is linked with social organisation and particular moieties associated with certain classes of projectile point. According to Sam Woolagoodja, an Umede-Worora man from the Yampi Sound, prior to the time of two culture heroes and moiety figures Wodoi (spotted nightjar) and Tjungkun (owlet nightjar), men used simple wooden spears without a spearthrower to hunt and fight. Tjungkun invented the first spearthrower, made from a thin, branching tree limb, trimmed to form a shaft with a small projecting peg. The proximal end of the branch formed the distal end of the finished implement, with the peg projecting along the shaft to engage the butt of a spear. Spears made by Tjungkun were straight pointed sticks. The simple hooked stick gave way to the light and slender spearthrower carved from wood of the bean tree (Erythrina vespertilio). The hook, carved from a prong of hard wood (Acacia), was fitted to the shaft, lashed with sinew and reinforced with spinifex (Triodia) or Callitris resin. Northern Kimberley spearthrowers range from 90 cm to 130 cm in length, including the longest made in Australia.

Wodoi made the first stone-tipped spears--light, reed-shafted darts that could be propelled great distances with the new spearthrower. The first stone spearheads were simple pointed flakes and blades with coarse toothed or steep edges. Tips were often thick and heavy. Even so, the new weaponry was far superior to the old, hand-thrown spears. Wodoi and Tjungkun bequeathed the art of making such weapons to the ancestors of the Worora, Wunambal and Ngarinjin people of the North Kimberley.

In the east Kimberley, the Miriuwung attribute the introduction of pressure flaking to the 'blanket lizard' (frilled neck lizard), although they believe that the nightjar, Panangka, produced Kimberley dentate points (Akerman and Bindon 1995), and the eagle, Kanbira, created uniface points. Archaeological examples of uniface points are believed to be Kanbira's talons used as spearpoints. Tindale (1985:26) suggests that these relationships refer to a time when moiety divisions predominated. It remains relevant today since moiety totemism based on pairs of 'opposites' still operates at a number of levels within the 8-subsection system of social organisation found in the region.

People believe that spearheads introduced by Wodoi were improved by humans. Invasive retouch using pressure led to refinement of symmetry. The use of bone indentors permitted the creation of denticulations or serrations that further enhanced the lacerating potential. According to legend, the first reed spears were small, about the height of a man. In some instances, the stone tip was socketed into a reed shaft and fixed with beeswax. On impact, the stone (or occasionally bone) point would enter the quarry and detach from the shaft which would fall free. 'A man doesn't know he has been killed with this sort of spear' (Wattie Nyerdu, pers. comm. 1976). Later spears were made with a stone tip fixed in resin to a foreshaft (usually a thin stem of acacia, mangrove or eucalyptus about 1 m long) that was in turn socketed into a shaft (a phragmites stem between 1.5 and 2 m long (see 'Hafting').

Raw materials, quarries, and distribution of Kimberley points

Akerman and Bindon (1995:93) describe the distribution of lithic materials used to produce points in the Kimberley. While certain materials can be sourced to specific areas, others have a much wider distribution. In some areas, knappers within a local group could access two or more materials easily, in others they were not so fortunate. All men had access through their exchange-network partners to points, made from a wide variety of materials, that had been crafted elsewhere in the region. Clendon (1999:317) suggests function (either esoteric or mundane) may be determined by the choice of material used to make a point and also the degree of refinement attained in manufacture. Clendon states that the Worora differentiated between points:

 normal, all-purpose, functional referred to by the noun
 jimara [and another type] essentially nonfunctional...
 made of shiny or semitranslucent stone such as jasper
 or agate, and of bottle-glass after European contact. It
 was beautifully crafted and often so thin as to be
 impractical for use in hunting. These points were not
 made for hunting but for their aesthetic value and for
 the magic and curative powers they derived from that
 value. The commonest general term for this type of
 spearhead is jimbeerlanya, which is feminine...The
 special properties of jimbeerlanya...meant that they
 were not only culturally highly significant items but
 were also shiny and dangerous by virtue of their
 magical properties. (Clendon 1999:317)

KA believes that the term jimbila/jimbeerlanya initially referred to quartz crystal points only, with jasper and agate points having other names dependent on their colour and/or the actual mineral or rock type utilised. Glass used for points was, after contact, termed jimbila. Jinalya, the Kimberley stone- or glass-tipped spears, were generally termed (in Kimberley creoles) 'bottle-spears' to distinguish them from the metal-tipped 'shovel-spears'. Clendon may have translated back from the creole to Worora and thus classified all highly refined points as jimbeerlanya. Clendon implies that highly refined points were not used as spear tips, and that all points made from translucent stone had magical properties. Neither assertion accords with KA's observations.

In the 1960s and 1970s, KA participated in a number of hunting trips in which spears, tipped with finely worked glass and stone, were used by men who had, from their youth, made and used such weapons. At no time did these men distinguish between the real or perceived capabilities of point materials made from locally available materials. For example, winjalalanyu, a translucent agate, orange or white in colour, was peculiar to the area immediately south of the mouth of the Prince Regent River. Nodules of the material are small and most points are shorter than those made of other materials. On the central Kimberley plateau, points are generally made of quartzites knapped from the cortical margins of boulders and outcrops. When white or grey, this material is known as kiimbu/m; if yellow or pale orange or red, it is called bilirri and is said to be named from the sound of feet shuffling through sandy ground. Black quartzite was called ngalat 'it twinkles'. The large and beautifully finished stone point described and illustrated by King (1827:68-9) was made of an opaque, dark green to black chert material termed wurumungkayi that is found along contacts between igneous intrusions and the quartzites that make up the bulk of the Kimberley massif.

Clendon's use of the term jimara refers, perhaps, to an oval flake knife used for ceremonial purposes and introduced from regions south of the Fitzroy River, or perhaps a mis-hearing of the term jilera, which was used to describe small pointed flakes and blades as well as archaic point forms. Large unhafted blades, termed jarung by the Worora, were used as knives, as were jilera and Kimberley points.

Davidson and McCarthy (1957:447-51) examined the distribution and function of Kimberley points, based on work by Davidson in 1930-31 and 1938-39, and suggest that the Jaru and Kija of the southeast Kimberley obtained points by trade from tribes occupying the central Kimberley. However, Tindale observed both Jaru and Kija knappers manufacturing Kimberley points at Moolabulla station in 1953, and Micha (1970:289) confirms that the Jaru present at Sturt Creek station had a long history of pressure flaking Kimberley points. Furthermore, KA has observed that quarries of white chalcedonic silcrete on Nicholson station were important sources of material that was heated to make exceptionally large (>8 cm) Kimberley points (Figure 5). These quarries lie on blacksoil plains, and pits last dug in about 1960 were still evident in 1982. The large white points were produced purely for exchange purposes and sent to trading partners to the south and east of Jaru country (see also Harrison 2000).

[FIGURE 5 OMITTED]

The Miriuwung and Gadjerong occupying the Ord River Valley and lands across to Keep River also moved Kimberley points and spears east to the middle reaches of the Victoria River basin. From here they were diverted into northern and southern exchange routes.

Isolated examples of Kimberley points, usually of white chert, have been recovered from the Gibson Desert and Central Australia. Points have been recovered as far east as Kilmorey station in central Queensland (Akerman with Stanton 1994:17). Traces of resin at the proximal end indicate that it had been hafted, but the Kilmorey point is unusual in that it is made of fine-grained quartzite (not chert). Quartzite was also commonly used to manufacture points in the central Kimberley, particularly to the north and west of the Ord River.

Between 1880 and 1903, Aboriginal prisoners were sent from the Kimberley to Rottnest Island prison, 19 km west of Fremantle, where they made pressure-flaked glass spearpoints (Serventy 1967: 123-4). No early records throw light on the purpose of these points, although they may have been for trade with prison officers and visitors to the island. Alternatively, the practice of this unique craft may have been an important way for the Kimberley Aborigines to express, maintain and reinforce their identity among Indigenous prisoners from all parts of Western Australia.

Manufacture of Kimberley points

Several scholars describe, sometimes inaccurately, how Kimberley points were made. For example, in an early account of pressure flaking, Clement (1903:5) confused how the indentor was held and used. The indentor was incorrectly seen as a passive tool against which the preform was pressed. Balfour (1903), and more recently Jelinek (1975:176), repeated this mistake. Hardman (1888:58) provides an equally confusing description of the flaking technique used to produce spearheads.

Excellent descriptions of pressure flaking have been given by Love (1917:25-6; 1936:74-5), Basedow (1925:367-70), Idriess (1937:59-62), Elkin (1948: 110-13) and Tindale (1985:1-33). Idriess details the use of red-hot wire to divide glass bottles into suitably sized pieces. Porteus (1931:111-13) describes the manufacture of glass points, with insights into the technological problems solved by Aboriginal craftsmen. Mahony (1924) observed how a new tip was made on a damaged glass point at Argyle in the east Kimberley. Akerman (1979a), Graham (1984) and Tindale (1985) describe heat treatment prior to the manufacture of pressure-flaked points. Moore (2000) provides a detailed technological description of Kimberley pressure flaking.

Despite some similarities (see Akerman 1979b; Crabtree 1970), Kimberley pressure-flaking techniques are different from those in the New World. The Kimberley craftsman sits on the ground with one leg tucked under the other, which is extended before him. Astone (about 15 x 15 x 12 cm) is placed in front of the knapper, to serve as a working platform, and covered with a cushion of Melaleuca bark. The knapper usually holds the piece between thumb and fore-finger, with one edge resting on the bark. He grasps the indentor across the palm of his other hand, with the tip emerging below the little finger. He places the tip of the indentor on the upper margin of the object piece and braces his wrist and arm before applying pressure by leaning forward with his body, and pushing down and outward with his hand.

The contemporary reduction sequence follows one of several possible paths (Figure 6):

(a) Production of a blank

Blanks are invariably produced by direct percussion with a hammerstone and can be made by:

* reducing a tabular piece, or flat cobble, to suitable size for pressure flaking;

* removing a flake from a core. The flake is reduced by direct percussion, usually bifacially applied;

* removing a large flake from a core. This flake is prepared into a small core from which a flake is removed and reduced by percussion flaking;

* creation of a long narrow blade which is ready for pressure flaking. This is very rare.

The blank is usually made during collection of the raw material:

 The Worora man...always keeps his eyes open for
 useful bits of stone that will make spearheads. He will
 take up a piece of broken rock...and test it by striking it
 with any convenient lump of stone lying about that will
 serve as a hammer...If it flakes nicely he will break it
 down to somewhere near the size of a spearhead, and
 put it in his paperbark wallet, to be dressed at his
 leisure in his camp. (Love 1936:74)

Spencer (1928, fig. 378) illustrates a wallet that contains at least 16 partly worked points and one finished point, made uniquely from a piece of pearlshell.

[FIGURE 6 OMITTED]

KA has observed that the original orientation of large flakes may not dictate the orientation of the finished point. From 1965 to 1978, Aboriginal knappers would usually attempt to obtain thick, oval or rectangular flakes for use as blanks, rather than blades or flakes that terminated in narrow points. The latter were considered too thin to withstand pressure flaking. Exceptions can be found, usually outside the normal boundaries of point-manufacture, where lithic experimentation has taken place (e.g. Akerman and Bindon 1983).

(b) First stage preform

The blank is subjected to pressure flaking with a hardwood indentor. This process removes broad, flat flakes with bending initiations and negligible bulbs, and levels the topography of each side. Flenniken and White (1985:149) argue that the ventral surface is flaked first. The first stage preform often appears like a miniature ovate handaxe with a bi-convex section and regular margin.

(c) Second stage preform

The first stage preform is now subjected to serial pressure flaking using an indentor made from strong bone, usually the ulna of a large macropod. The mandible of a freshwater crocodile may be used (Mathews 1901:84), as may large splinters of human bone taken from the distal tibia. Sections of dugong rib were also, on occasion, used as indentors.

The bone indentor permits controlled force and ensures that flake scars are more regular than those made by a wooden indentor. The second stage preform resembles the finished point, except that the margins do not have serrations or denticulations and the tip is not accentuated.

(d) The completed point

The knapper starts developing the distal end, and flaking is usually directed from the edges of one side only. A series of short flakes is removed from each side of the tip, transforming the section from biconvex to triangular and forming a crest down the mid-line of the side that was last flaked. The face with the crest is in Worora called the ilatu--'back' or 'spine' of the point. The other face is appropriately enough designated the ingulum--the 'belly'. A distal keel is a recent design feature that permits easy production of sharp but strong points.

Tindale notes that, with Jaru and Kija people of the east Kimberley, 'a slight keel left on one face was normal, but it was best to reduce its height to a minimum, and there was a distinct effort to make its crest a relatively straight line'. He also notes that an Aboriginal knapper had indicated to him that 'all the present-day people endeavoured to reduce the keel on the non-bulbar side of the blade, whereas earlier people had made no attempt to reduce this, even when some of them had pressure-flaked both sides of their spear blades' (Tindale 1985:9, 24).

A fine spatulate or chisel-ended bone is used to serrate the margins and create a needle-like tip. Tindale (1985:10) suggests that butt-thinning was a final stage in the manufacture of spearheads but not if they were to be hafted as knives. He also states that butt-thinning often occurred when the point was being hafted to a spear. However, this is not supported by study of museum collections and there was no need because Kimberley points were not lashed or socketed to shafts.

The Worora term for the needle tip is the inguri, and the butt or proximal end is the emalapa. Once hafted, the spearhead is termed ambul--'eye' (of the spear).

Hafting

Spears tipped with Kimberley points often originated in the central Kimberley close to stands of phragmites reed (Phragmites karka), used for the rear shaft of composite spears. Stands of phragmites were not common and the shafts themselves were important items of exchange. Love (nd:81) notes that Worora distinguish phragmites spear shafts from different sources. After the Kimberley hinterland was opened up for pastoralism, exotic bamboos, planted near settlements, were also used as spear shafts. Bamboo also replaced phragmites as shaft material when feral cattle destroyed stands of the latter, now rarely found with culm diameters more than 15 mm (Glen Wightman, pers. comm.).

The presence of small bifacial and unifacial points in archaeological contexts in Arnhem Land and adjacent parts of the Northern Territory (Jones and Johnson 1985) suggests that spears, similar to those used historically in the Kimberley, were once common in the former area. Access to large phragmites, or some similar plant, for hafting purposes would be required in the manufacture of these spears and it is possible that depletion of these sources may have led to changes in the arsenal. It may be that spears with heavy blade or wanji biface points superseded them as feral water buffalo destroyed old large-culmed stands of phragmites and, with continual cropping, maintained a smaller growth pattern.

Alternatively, the adoption of the heavier spearheads may have led to the abandonment of the lighter composite spears and the deliberate seeking out of new hafting materials. Heavy blades required strong shafts to support the heavier point and hafting materials. The indigenous bamboo Bambusa arnhemica and light strong timbers such as yellow hibiscus (Hibiscus tiliaceus) and spearbush (Macaranga tanarius) provided this strength. Shafts made of these materials permitted the use of very large iron and steel heads, which in Arnhem Land are often in excess of 300 mm in length and 180 g in weight. Metal points collected from Kimberley spears average about 200 mm in length and weigh 66 g and are more compatible with the relatively fragile shaft. Glen Wightman (pers. comm.) suggests that, in the Northern Territory, phragmites is rarely found with culm diameters in excess off 15 mm.

Kimberley points were commonly embedded in a knob of resin on the foreshaft, after the distal end was bound with sinew to prevent splitting on impact--not, as Carnegie (1898:340) suggested, to hold the resin in place. However, metal blades and large glass points produced for sale were set in a split foreshaft, lashed with string or sinew binding and then covered with wax or resin (Akerman 1978:486).

The method of hafting solely with resin permits release from the shaft at impact, minimising damage to the point. However, the weight of large points, even on a warm Kimberley day, would exacerbate slumping of the softened resin, causing the point to move. Even small points will slump from their true position, if left long enough on a museum shelf. Spears, and other tools, were continually overhauled, often on a daily basis, to ensure their reliability.

Function of points as projectile tips and knives

Unlike the stone-headed spears of Arnhem Land or northern Central Australia, the composite spear of the Kimberley was long, slender and very light. In use, the Northern Territory and Kimberley stone-tipped spears were as different from each other as the solid slug of a shotgun and the high-velocity bullets of the Japanese rifle in World War II. Blade-tipped spears depended on the shock value of their generally high mass, coupled with the massive haemorrhaging created by long cutting edges. Kimberley composite spears are, on the other hand, exceedingly light for their length, averaging about 170 g in weight (Akerman 1978:486). The stone tips of Kimberley spears are nearly all less than 6 cm long and most are under 4 cm. Hafted points rarely exceed 8 cm, except in the southeast Kimberley. Saville-Kent (1897:12) describes a glass point 16.5 cm long as being 'abnormally elongated' and states that a hafted stone point, of which 4 cm is exposed above the resin, is of more normal dimensions (1897, figs 3 and 4). Kimberley spears were thus low-mass, high-velocity hunting and fighting weapons, with a point that disengaged from the shaft to promote bleeding and infection. Longer points were also regularly made, but they were not intended to arm functional spears.

'The Worora still make and prize their beautiful stone-headed spears, but, for everyday use, they now like to have an iron-headed spear' (Love 1936:81). This situation prevailed at Kalumburu until the early 1980s and to a lesser extent at Gibb River and Tablelands stations and Fork Creek in the 1970s. Spears were made by craftsmen at Kalumburu and traded or given to men at the other communities, although some were made locally. The last knappers to regularly make glass spear points were located at Kalumburu in the 1980s. The tips of the Kalumburu spears were, at this period, invariably made of milky-white glass derived from cheap ovenware. At Kalumburu, at this time, one or two older men still hunted with iron shovel-nose spears. Their quarry were agile wallabies that fed at late afternoon and early evening, on cultivated pasture land adjacent to the community. Hunting occurred on average about twice a week, with one or two wallabies each per hunting episode being taken.

Stone-tipped spears (with Kimberley points) also served as knives for gutting prey, butchering game after cooking and for minor trimming purposes. When used in this way, the spear was held point down and the foreshaft grasped at, or adjacent to, the resin mounting (Lommel 1997:7, plate 108). The point of the spear that killed a beast often detached at impact, and was used as a knife for gutting or for dismembering cooked meat. Points would therefore acquire residues not associated with projectile use. Similar damage to the needle-like tip could occur when a point was used as projectile tip and as a knife.

Stone and glass Kimberley points were also used within the region unhafted, as knives for both ritual and secular purposes. Mitchell (1962:362) describes an artefact collected within the Kimberley that consisted of two stone Kimberley points set butt to butt in resin; 35 mm and 25 mm of the distal end of each point, respectively, is exposed. Mitchell notes that the object could serve admirably as a double-bladed knife, but cautions that its design may be due to external influences.

The large white stone points made and exported by the Jaru (see above) were used as ceremonial knives in the central and western arid zones. Tindale (1965:156) illustrates examples of such points from the Tanami and Gibson deserts, with either resin or resin and wood hafts, of the type used in Central Australian hafted-blade knives. Davidson and McCarthy (1957: 450) refer to similar knives on the Canning stock route, which traverses the Great Sandy Desert, and Tindale (1985:18) illustrates a resin-handled Kimberley point, used as a circumcision knife, collected at Balgo Hills in 1953. Spencer (1928, fig. 147) illustrates a stone Kimberley point hafted with resin to a length of hair string, used by the Kaytej around Tennant Creek in a form of sorcery similar to bone-pointing. Spencer (1928:510, fig. 376) notes, however, that the points were used as spear tips in Central Australia, rarely as knives. Indeed, whole Kimberley spears were probably traded as far south as the MacDonnell Ranges. Spears tipped with Kimberley points as well as unhafted points were also traded into the Fitzmaurice and Daly River regions of the Northern Territory. A finely serrated, bifacially flaked stone point that appears to be Kimberley in origin, but was found in the Alligator Rivers region, is on display at Cooinda, Kakadu National Park.

The context of production and exchange

The large numbers of Kimberley points in public and private collections suggest that, in this century, most were not made for Indigenous purposes but were in fact created for the curio and artefact market. It is also apparent that point production was a major occupation for most men, when they were not involved in hunting or ceremonial activities (Love 1936:74; Porteus 1931:113). From both observation and experimental work, it is known that a small (3-4 cm) point can be made in about 15 minutes. Larger points generally take longer, but glass points are an exception. The main effort in creating points from pieces of curved glass is directed at reducing the concave section. The inner face of the glass is subjected to serial flake removals from either edge until the face is relatively flat. Only then does the knapper tackle the convex face, which is already set up for optimum flake removal. He then only needs to remove a single series of flakes from each edge before serrating the edge and accentuating the tip.

A glass point about 20 cm long can be made from a rectangular piece of bottle glass in about 45 minutes. Glass points made at Kalumburu during the 1970s were leaf shaped, averaging 7 cm in length. Serrations on these points were minute, reflecting the regular and delicate trimming of the margin as the knapper adjusted plan symmetry about longitudinal axis.

Most points for everyday are usually less than 6 cm long, and six hours of knapping could produce about eighteen points suitable for hunting and fighting purposes. A skilled hunter, hunting two or three days a week, would require no more than six points and might replace or renew four points in a week. The constant production of points was as much to make goods for gifts or exchange, as it was maintain a constant supply of spear armatures. KA estimates that between five and twenty points may be held by a person at any one time.

What happens to the excess points? Where do they go? The answer lies in the Indigenous economic exchange systems that operated within and adjacent to the Kimberley region at, and subsequent to, European settlement.

Trade routes lay, in pre-contact times, like a fine mesh over the continent, linking disparate cultural groups (Akerman 1979c:250). Goods moved initially within the range of kin and personal exchange partners, to maintain complex systems of reciprocity, rights and obligations. They extracted commodities required for their own gratification and contributed others in turn. Prior to contact, the system was a very dense series of exchanges that caused most objects (sacred cult or religious objects did not move in the same way) to circulate, first of all, among adults of the local group. Objects then filtered out to adjacent groups with kinship affiliations. Centres of exchange were restricted to places where two or more groups participated in religious and economic activities. Movement of objects was diffuse through a populated landscape, rather than linear with single directional routes that linked a finite number of specific population centres that were isolated in an otherwise empty continent. The immediate exchange environment probably absorbed most of the surplus. Only a few curated items, which acquired greater value or significance as they moved from their source, actually travelled the great distances.

If exchange networks were finely meshed prehistorically, after European contact Aboriginal relocation probably created a more linear system at isolated centres, with a much broader mesh. Post-contact, goods probably moved from centre to centre. However, few collections, notably those made by Donald Thomson, can be related to extant Aboriginal landscapes and most only document the nearest named European settlement. Nevertheless, spearhead blanks, preforms, spearheads and spears themselves continued to circulate around the Kimberley, along with wooden artefacts, hairbelts, stone axes and shell objects. In the past, as the objects moved to the periphery, some were diverted out to other cultural areas, and new objects were introduced. Consequently, artefacts primarily associated with the Pilbara, the Western Desert, northern Central Australia and the Victoria-Daly River regions of the Northern Territory have all been documented within the Kimberley.

Implications for use-wear and residue studies

Wood, resin, skin, hair, feathers, blood and ochre are all likely to come into firm contact with points at various stages of manufacture, during daily transportation, maintenance of equipment and incidental use in a variety of tasks. Such traces could obscure the primary function of a hunting or fighting weapon, or falsely suggest a range of multifunctional tasks. Shatter and overshot flakes often cause small cuts and bleeding. Flakes become trapped in the Melaleuca bark cushion and can cut the hand holding the indentor. According to old craftsmen in the 1970s (and personal experience by KA), if a preform rolls under pressure, the inner wrist and fingertips can be cut. 'During the operation the native often cuts his fingers on the flake or razor-sharp splinters; the blood which follows he removes by passing his fingers through his hair' (Basedow 1925:370).

Wiping fingers through the hair may well transfer other residues to stone tools. In the Kimberley and elsewhere, men bound their long hair into a chignon to carry spare spear points, sinew and other small objects. The chignon was also used to carry fat and other tidbits home from a hunt. Love (nd:38) describes a Worora hunter enclosing the paunch of a kangaroo, filled with intestinal fat and the spleen and wrapped in green leaves, within his hair, binding the lot together with a string headband. Add grease and ochre for cosmetic purposes and we have a complex residue pot-pourri.

Particularly colourful or translucent points were often wetted in the mouth (Love 1936:75; personal observation by KA). The moistened surfaces enhanced the colour and lustre, allowing the craftsmanship to be fully appreciated. Tobacco residues and food particles could thus be also transferred to the finished point.

Points were also stored together in Melaleuca bark wallets and cushioned with bird down (particularly from raptors) or bulrushes. Larger points were also protected by wrapping soft fibre about the tip and distal end (Akerman and Bindon 1995:93). King (1827:69) describes a wallet containing several stone spearpoints, each protected by fur and separated with strips of bark. While in a wallet, points were often in contact with other objects.

 This bark wallet, the bururu, is a cigar shaped roll of soft
 bark, bound with fibre string. Inside is a nest of down,
 usually of the bulrush, in which rest finished
 spearheads, pieces of stone in different stages of
 manufacture, the tools used in their making, and,
 nowadays, quite usually a stick or two of tobacco, or
 perhaps a small bit of tobacco, according to the man's
 financial state. (Love 1936:74)

Balfour (1951:274) exhibits the contents of one such wallet, which includes small lumps of prepared resin, sinew, boab fibre string, stone, bone and wooden tools for manufacturing spearheads, as well as partly worked and completed stone and glass points. At least two of the illustrated points appear to have resin traces indicating that they had been hafted at least once. Fifteen points from a collection of 107 stone spearheads collected in the Kimberley in 1917 (Stuart 1923, plate facing p. 78) show similar resin traces. Whether they were collected unhafted or were removed from spears for the purpose of collection cannot be discerned. It is clear, however, that, within the Indigenous setting, points may be removed from spears and curated by their owners prior to being rehafted or transferred via exchange routes to other owners.

Hafting is also an occasion for the further accumulation of residues. The resins may be obtained from spinifex (Triodia sp.) or the native cypress (Callitris columellaris). Lommel states that wax obtained from a native bee was also used to secure points to shafts, and Tindale (1985:16) implies that eucalyptus resin was used. KA failed to confirm these uses of waxes and resins, but confusion could arise because beeswax was certainly used for joining hardwood foreshaft to phragmites. The resins are applied in a heated state and smears of different resins and waxes can occur all over the spears. The resin at points of junction with both head and foreshaft is often brought to a liquid state with a small burning stick and then moulded to form a perfect seal. As the glowing embers of the stick burn, the end often collapses, so that ash and crushed charcoal are wiped across resin and stone components, further adding to the accretion of extraneous material.

Kimberley spears were often decorated with a wash of red or white ochre applied to the shaft, foreshaft and resin point mounting. Red, yellow or black dots were also applied over a ground wash of colour, but this was usually restricted to the foreshaft and resin.

Finally, craftsmen's hands were rarely washed of food, grease, ochre or other residues prior to making and using spears, spearheads or other artefacts. We caution against any exclusive interpretation of residues that ignores potential contaminants from a variety of sources.

Archaeological indicators of function and exchange

The above descriptions of manufacture, maintenance and exchange suggest a range of plant and animal residues is likely, whether the points are used exclusively as functional projectile tips or made for exchange. Many may have been exchanged and subsequently used as projectiles. Specific taxonomic identification of residues was not feasible in the following study, although simple screening tests were undertaken to assess visual observation of some organic tissue. The most secure traces of tool utilisation are: use-wear consistent with projectile use (specifically impact scars at the tip, scarring, rounding, polish and striations), in possible combination with hafting resin and residues (Akerman 1978; Fullagar 1991; Fullager et al 1996; Kamminga 1982; van Gijn 1990).

Items made and used exclusively in exchange networks should have no use-wear but may have hafting traces (for exchange as elements of complete spears, some of which may never have been used). However, absence of use-wear is not, on its own, diagnostic of exchange. Traces of bark or other plant wrapping may be consistent with protection of both exchange items and functional spear tips. Nevertheless, absence of hafting traces and absence of use-wear, together with evidence of protective wrapping, would seem to indicate either exchange or, at least, delayed use. In combination with sourcing studies of raw material, and technological analysis, it may be possible to securely identify exchange of projectile tips. Our objective here is simply to document the range of use-wear and residue patterns found on ethnographic and archaeological items.

Extensive traces of fibrous plant tissue in the absence of use-polish are good indicators of wrapping, for delayed use or exchange. The best indicator of a functional projectile tip is distinctive use-wear, perhaps in combination with hafting evidence (resin) and animal traces (hair, blood, skin). Hafting is indicated by resin. We set out to identify the eight possible combinations of these three sources of evidence (wrapping traces, use traces, and hafting traces) in ethnographic and archaeological point collections:

* 1A: hafting, use, no wrapping

* 1B: hafting, use, wrapping

* 2A: hafting, no use, no wrapping

* 2B: hafting, no use, wrapping

* 3A: no hafting, no use, no wrapping

* 3B: no hafting, no use, wrapping

* 4A: no hafting, use, no wrapping

* 4B: no hafting, use, wrapping

Different coloured ochres provide a fourth variable, introducing more possible combinations of residue and use-wear traces. However, the presence of ochre does not appear to be exclusively related to exchange or use (see discussion above), although the amount and kinds of ochre can potentially provide information about decoration and, perhaps, ethnic identity. Hafted points in the Australian Museum collections are more highly decorated with a greater range of ochre colours than any non-hafted points, few of which have macroscopic traces of pigment.

The ethnographic and historic sources discussed above provide a context for analyses of points in museum and archaeological collections. Our objective is to assess whether aspects of function and exchange can be inferred from analysis of archaeological and museum collections. To do this we present a pilot study with points selected from several contexts. Our pilot study includes:

* a preliminary technological study of points collected, with precise known provenance in the Kimberley by KA (Table 1);

* a detailed study of ethnographic artefacts, classified as points and sourced to the Kimberley, in Australian Museum collections (Table 2); and

* a selection of points from archaeological artefacts excavated from sites in the Keep River area of the east Kimberley, Northern Territory (Table 3).

Museum collections

The Akerman collection (Table 1) derives from ethnographic and archaeological contexts across the Kimberley and includes bifacial, unifacial and backed points made of ceramic, glass and stone. Summary statistics for these points, including two silcrete 'blanks', are shown in Table 4. Glass points appear to be significantly larger, as anticipated. The mean lengths in order of decreasing size are glass (58.6 mm), silcrete and quartzite (51 mm), chert (50.7 mm), quartz (36.8 mm), agate (35.5 mm) and ceramic (35.0 mm). In contrast, the thickness does not vary as much and is smaller for glass than for chert, ceramic and quartz. In part, this must be due to the fixed size of available glass (mostly bottle but also window glass).

Although invasively flaked biface points (mean: 45.8 mm) as a whole (all raw materials) are shorter than all invasively flaked uniface points (mean: 47 mm), if we compare similar materials, say quartzite and silcrete, biface points (mean: 51 mm) are longer on average than uniface points (mean: 36.5 mm). Although the sample sizes are small, given the reduction sequence documented by Akerman's review of ethnographic data (see above and Figure 6), the evidence does not lend support to the argument that biface points are transformed from uniface points. Because of extensive flaking, none of the biface points had clear indicators of whether dorsal or ventral surface had been flaked first, as Hiscock (1994) was able to demonstrate for flake assemblages from Arnhem Land.

We recorded in more detail a selection (n=26) of invasively retouched bifacial points with serrate or dentate retouch. The number of pressure scars per 10 mm ranges from 3 to 6 on chert biface points and from 3 to 4 on chert dentate points. Again, the numbers are small, but they support the argument that the type of marginal retouch (serrate, denticulate or dentate) on biface points did not affect the initial reduction stages or preform preparation. The largest bifacially invasive point in this sample is from Carlton, in the east Kimberley, but no pattern correlating size with geographic location is obvious. Large points are found throughout the Kimberley, but it is only in the south and southeast that craftsmen, utilising massive white cherty material suitable for heat treatment, consistently produced points more than 80 mm long.

In a preliminary study, Kristin Schubert (University of Sydney) analysed a sample of archaeological points which RF excavated (n=12), and a sample of points (n=128) collected between 1895 and 1963 and donated to the Australian Museum. Quantitative data on 84 complete points from this sample are shown in Table 2. To provide a temporal dimension for the contact period, the points were divided into early (1895-1910), middle (1911-40) and late (1941-63) periods. Unfortunately, these divisions relate only to when the collections were donated to the museum and there is little information on when the artefacts themselves were made. No significant differences could be identified between the three time periods. However, multivariate analysis (correspondence analysis using Richard Wright's MV nutshell program) did show some interesting relationships.

Variables recorded included: length, width, thickness, percentage cortex, location of retouch (bifacial or unifacial; marginal or invasive), maximum scar length, number of scars per 20 mm, presence or absence and number of serrations per 10 mm, and the presence or absence of resin. No significant change could be demonstrated for the contact period, and this may reflect poor documentation on when the artefacts were made. However, the ethnographic points, treated as a single group, have more finely serrated edges than the archaeological stone points (a few of which have dentate or serrate margins). The stone points are generally thicker, shorter and more invasively flaked. Finally, stone artefacts overall tend to have more resin than glass or porcelain points. Summary statistics are presented in Table 5.

Unfortunately, the documentation does not give precise provenance of enough artefacts within each type to estimate geographic variations.

Use-wear and residue studies I: museum collections

From a preliminary study of 207 points in the Australian Museum's collection, 22 implements were selected for further microscopic study. The selection deliberately included stone and glass implements with various types of retouch (uniface, biface, dentate, denticulate). All tools were examined with a stereomicroscope with oblique incident light (Orient SM-1, magnifications up to x100) for residues and macroscopic traces of use. Subsequently, a metallographic microscope with vertical incident light (Olympus BH series, magnifications x100, x200, x500, and x1000) was used to examine use-polishes and striations (for a more extensive methodological outline, see Fullagar 1986, van Gijn 1990). The implements were not cleaned during the analysis and were handled with starch-free gloves.

On some points, traces which looked similar to blood or other animal tissue were screened for haemoglobin and immunoglobulins IgG and IgA (Loy and Hardy 1992; Loy and Matheson in press).

All of the implements were well preserved, with minor contaminant traces from storage. A few had post-depositional modifications, like fungi (51408), pencilmarks (33045 and 42881) or plasticine (33045). The remaining artefacts were in very good condition.

Use-wear and residue results are presented in Table 6, with selected micrographs in Figure 7. Only a few points displayed diagnostic wear traces of projectile use, including four implements with impact scars and other traces showing they were used as projectiles (85607, 85609, 85610 and 85612), and one point (54614-2) displaying traces from skin perhaps associated with possible blood residue (trace positive 1+ result for IgG and IgA, Freeblot 27; negative Hb; but also trace positive 1+ result for the negative control). Four factors lead us to have little confidence in the haemoglobin and immunoassays. First, the positive traces were mostly low (+1). Second, they rarely corresponded with positive haemoglobin tests. Third, negative controls taken on areas that appeared to lack any residues often gave positive results. Fourth, the pieces were from collections which had been frequently handled and, given the generally low positives, it would be difficult to eliminate a range of possible contaminants. This caution should not be taken as a criticism of these techniques, merely that the collections that we analysed were, on their own, unsuitable for affirming presence or absence of blood residues.

[FIGURE 7 OMITTED]

However, a large number of points displayed residue: yellow, white and red ochre; mashed plant remains; wood fragments; and resin. There were also some that displayed no traces of wear or residues at all. We present results of the functional analysis in four groups, suggested on the basis of possible hafting, use and wrapping traces (Table 7).

Group 1: Hafted and used projectile points

1A: Plant (wrapping) traces are absent. A group of smaller bifacial points, ranging in length from 3 cm to 6 cm, all display traces of use. They are made of agate or chert and have small, sometimes indistinct, serrations. Serrations are so delicate on some of these points that they must have been made by a very fine edge. (Very fine serrations on some points are described ethnographically as produced by a knife blade.) All points in Group 1A have microscopic or macroscopic traces of resin, probably related to hafting. The traces of use included impact scars, due to their former use as projectile tips. Indistinct microscopic linear streaks of polish, typical of projectile use (van Gijn 1980:45-6), were visible on E85610. In addition, all ridges were heavily abraded with abundant yellow ochre. Craftsmen had probably rubbed yellow ochre on them before use. The ochre rubbing may also have a role in exchange networks. Artefact 56414-1 is weathered with probable use traces and traces of resin.

1B: Plant (wrapping) residues are present. Three other artefacts (50739, 54614-2 and 55047-1) also have distinct traces of hafting and use, probably as projectile tips. They all have serrations but are larger than the artefacts in Group 1A, with lengths 8.5 cm, 8.9 cm and 9.5 cm, respectively. Another difference is that the serrations on 54614-2 and 55047-1 appear to have been produced after the artefacts had been used. Perhaps these were recycled from functional spearheads to exchange goods. Traces of plant tissue on 50739 and 54614-2 suggest they may have been wrapped, a possible indication of packaging for transportation perhaps as items in exchange networks.

Artefact 54614-1 lacks serrations and the tip is broken. We are not sure whether or not it is an impact scar. Residues (including plant fibres, ochre and probably resin) are impacted on the surface and immunoassays for IgG and IgA were slightly positive (+), with negative tests for haemoglobin. We include it in Group 1A because we think it was probably used. It is a similar shape to some of the archaeological points with distinct traces of use found on excavated Keep River points, and it may have been recycled, which would explain why residues survive on the relatively weathered surface.

Group 2: Hafting, wrapping, no use

A second group of chert and glass points range in length from 5 cm to 10 cm. Adhering to their surfaces are ochre and resin, indicative of a haft. None of the implements had actually been used either as projectiles or for other purposes. Their tips and lateral edges were completely fresh and no microscopic linear streaks of polish could be detected. One chert implement (Figure 7a), attached to a sawn-off haft (85332), had been extensively decorated with red and white ochre, which had also been applied on top of the resin securing the flint implement. All of these points were delicately serrated, and at their distal end displayed clumps of resin mixed with red ochre. Interestingly, they also showed a yellowish/orange (sometimes fibrous) residue associated with starch grains, indicating a plant. From ethnographic sources it is known that bark from the paperbark tree has been used to wrap objects for delayed use and exchange. We suspect that the plant residue is related to this very practice.

Group 3: No hafting, wrapping and no use

A third group, mostly glass and of variable shape and size (4 cm to 8 cm in length), did not display any traces of use, nor any application of ochre. Many larger points in the Australian Museum collections, although not analysed microscopically, are of similar appearance. Points in Group 3 have no hafting traces, but residues from the same kind of fibrous starchy plant as that on tools in Group 2 were common. Assuming that this residue is related to the wrapping in paperbark for safekeeping as well as exchange purposes, it is suggested that this group of implements still played a role in Aboriginal exchange systems. However, these tools had no practical application as projectiles, nor were they hafted or decorated. One serrated glass point (20525) is included in Group 3 (rather than Group 2) because, although there do appear to be resin traces present, the distribution of these residues could not be necessarily related to hafting. Point 57191-1 has serrations, traces of plant wrapping and is also probably unused.

Group 4: No hafting, no wrapping, no use

The absence of residues and wear traces distinguishes the fourth group. This group consists of one quartzite (E33045) and three brown glass (E33061, E4438A, E4438C) points with no traces of use, no hafting resin, no ochre and no wrapping traces. Artefact E33061 is delicately serrated, whereas both E4438A and E4438C have marginal retouch only, similar to design features of the Wanji biface except that these points are made on bottle glass which is curved. All edges of Group 4 points have fresh fracture scars. We suggest that these artefacts were made solely for European tourists in order to obtain cash and were removed from an Indigenous milieu soon after manufacture.

Use-wear and residue studies II: Keep River archaeological collections

While analyses of excavated data (currently 11 trenches from 8 sites) are incomplete, no complete biface points with serrations have yet been found in excavated deposits. Use-wear and residue results are presented in Table 8, which includes broken and whole points found in excavations from Punipunil, Marralam and Jinmium (see Atchison 2000; Fullagar et al 1996). Debate over the oldest Pleistocene dates for Jinmium does not affect the chronology of points, which all occur in upper Holocene levels. Biface and uniface stone points are found archaeologically in prehistoric contexts spanning the last few thousand years, and some have pressure-flaked margins conforming to our definition of Kimberley points. Ochre is present on a few pieces only.

Several (n=41) complete and broken points have distinctive traces of use as projectile tips. These include backed points, Kimberley points, other biface points and uniface points. Some of the uniface points have serrations. Abroken tip of one dentate point was also found. Several had plant tissue, possibly related to wrapping to protect the delicate edges. The archaeological residues match most closely those in Group 1A: they have distinct traces of use as projectile tips, with less evidence of wrapping and ochre than Groups 1B, 2 or 3. Small glass flakes were found at the Marralam site, where all levels are probably post-European, but no complete or broken glass points were found (Head and Fullagar 1997).

Discussion

Kimberley points that Aboriginal people recognise as 'early' types include both unifacial and bifacially flaked points, invariably made on flakes. Evidence of the original flake morphology can be seen on some early points in the form of dorsal ridges, striking platforms, ventral surface and dorso-ventral curvature. Retouch is usually short, rarely extending to the mid-line of the point, and only at the distal end. Retouch may or may not be invasive and stepping of flakes is common. In contrast with these early points, invasive pressure flaking on contemporary points commonly extends beyond the mid-line of each face, and it is usually impossible to tell whether the point was made originally on a flake, a tabular piece of stone, or a core that was initially reduced by bifacially applied percussion flaking. The reduction sequence is directed at producing a point that is straight about the longitudinal axis and planes, and is bi-convex or plano-convex in transverse section. Unfortunately, these criteria do not provide a secure basis for assessing the age of individual points from surface collections.

White and O'Connell (1982:112) suggested that the most elegant points with finely serrated or denticulated margins are a post-contact phenomenon. However, as Akerman and Bindon (1995:98) demonstrate, the earliest recorded ethnographic stone points (King 1827:68-9 and accompanying plate) and others collected prior to the turn of the century (Etheridge 1890, 1891:34-5; Hardman 1888:60-1, plate I) are as refined as any points of stone or glass manufactured this century on settlements and pastoral stations.

Preliminary assessment suggests that it would be difficult to establish a relative chronology of point production based on dates for registration in the museum records. However, ethnographic information does indicate that points functioned in both social and subsistence contexts, with increased production of larger Kimberley points (serrated pressure-flaked bifaces) in recent years for exchange rather than as hunting or fighting weapons. Microscope study of museum and archaeological collections has shown that both serrated and non-serrated points functioned as projectile heads, as points for exchange and delayed use, and as tourist commodities. Points in very early collections are thin, highly symmetrical and have extremely fine serrations, and refined serrated points are found in archaeological contexts likely to be pre-European. There is no reason to accept that finely worked Kimberley points were a product of post-European forces.

Four main groups occur in the museum collections. No point had traces of hafting in combination with no traces of use and no extensive plant wrapping. On the other hand, finely serrated points were included in the category of points with no hafting, no use and no wrapping. This supports our assumption that the extensive plant wrapping traces that we identified are unlikely to be residues from manufacturing stages (which we would expect on all pressure-flaked points). Impacted wood or other plant fibres were visible in some of the denticulations and are perhaps from wooden pressure flakers.

To what extent, if any, do the groups identified in the museum collections have any chronological significance? Do they represent a sequence, from functional spearheads with a trend towards increasing production for exchange? We cannot answer this without more precise age estimates of when the points in the collections were produced. The earliest are uniface points from C1/II spit 20, Unit 2, 65-75 cm below surface, where elemental radiocarbon age estimates (OZC603 and OZC538) vary from about 1020 to 1760 years before the present (Roberts et al 1998). The earliest biface point is from C1/II spit 9, with optical age estimates (COOR 9/1) from about 540 to 1200 years (Roberts et al 1998), similar in age to the earliest serrated/denticulated margins from C1/I spit 6.

Analyses of the lithic, glass and metal artefacts from the Marralam excavation (Head and Fullagar 1997) indicate that the production of points probably increased through station times at this place. However, such activities may have been more concentrated at certain localities when Aboriginal station workers took time off in the wet season, and held ceremonies and organised other activities accordingly. Points do not appear to increase in frequency during the last hundred years at rockshelter sites, suggesting that they were made elsewhere, probably close to pastoral station camps. Further archaeological work could test this suggestion.

The functional grouping of points based on microscopic traces suggests that serrated Kimberley points were not the only type of point exchanged among Aboriginal groups or in tourist markets. Marginally retouched glass points with little modification and no pressure flaking were also likely to have been wrapped, curated and exchanged.

Limitations of this study arise from the geographic focus of archaeological collections and the small samples overall. Although practically difficult to mount on a microscope stage, it would also be helpful to extend the use-wear/residue study to the large number of hafted points in the museum collections. The complete spears would also provide additional information in terms of shaft manufacture and haft decoration (cf Torrence 1993).

Conclusions

The aims of this study are to review point production, function and distribution from ethnographic and historic accounts, and to assess whether exchange can be inferred from functional analysis of archaeological and museum collections. Longer-term objectives are to understand the chronological and spatial variability of point forms in order to identify technological indicators of social change, particularly the last 100 years, spanning European contact.

Review of ethnographic and historical sources suggests that few, if any, collections have sufficient documentation to provide a clear idea of exchange networks prior to pastoral times (about AD 1880). Spencer's (1928) documentation of hafted Kimberley points in Central Australia and Tindale's (1965) record of them among Western Desert people are the sole exceptions. It is also apparent from the literature that exchange was extensive and that the manufacture of spearheads was a major occupation for most men when they were not involved in hunting or ceremonial activities. Ethnographic and historic sources provide accounts of the production and exchange of Kimberley points both as functional spearheads and as exchange goods within one of the large networks across the Kimberley, known as wunan. Discrete manufacturing stages and specialised toolkits were identified for biface production of stone, ceramic and glass points. The development of a tourist market did not of itself destroy the Indigenous exchange network.

Study of museum collections indicates four main groups, some of which have traces suggesting a degree of curation either for delayed use or exchange, and one group which appears to lack any traces and may have been made for a tourist market. Points display a range of craft skills, but no group lacks finely worked pieces. Larger points (sometimes very finely worked) are mostly associated with glass.

As indicators of exchange, microscopic traces are not definitive because, as the ethnography suggests, there are many sources of plant and mineral residues and any one combination may arise in the course of production, local contexts and protective wrappings. However, the combination of technological and functional variables can indicate whether or not an implement was used. The ethnography, together with study of museum collections, suggests that exchanged items are likely to have either no traces or a combination of ochres and plant wrapping traces.

It is not a new idea to view Kimberley points as both economic and social goods: wunan (exchange goods) and weapons (for fighting and hunting). Further study may link particular microscopic traces and technological constraints with stylistic variables related to pressure flaking and overall morphology, for example that the base shape (rounded or triangular) of biface points in plan view may correspond with social boundaries. Analysis of Kimberley points to mark social boundaries is the subject of future research.

Table 1 Points from Akerman collection (n=77)

Type Place Ref. Stone Length Width
 (mm) (mm)

BFI Derby 566 ceramic 43 17
BFI Derby 566 ceramic 27 16
BFI Kunmunya 553 fg agate 41 18
BFI Kunmunya 553 fg agate 35 16
BFI Munja 553 fg agate 40 16
BFI Munja 553 fg agate 52 19
BFI Munja 553 fg agate 42 16
BFI Munja 553 fg agate 26 11
BFI Munja 553 fg agate 33 15
BFI Munja 553 fg agate 33 15
BFI Munja 553 fg agate 38 15
BFI Munja 553 fg agate 21 11
BFI Munja 553 fg agate 29 14
BFI Tablelands 563 fg chert 57 17
BFI Rifle Range 563 fg chert 43 15
BFI Argyle 561 fg chert 53 27
 (Durack)
BFI Barramundi Y. 563 fg chert 30 20
 Fossil D.
BFI Carlton 549 fg chert 91 32
BFI Derby 566 fg chert 65 22
BFI Derby 566 fg chert 68 24
BFI Derby 566 fg chert 43 18
BFI Halls C. 549 fg chert 83 31
BFI Moolaboola 549 fg chert 29 16
BFI Rifle Range 563 fg chert 37 20
BFI Saddler C. 563 fg chert 28 21
BFI Saddler S. 563 fg chert 33 23
BFI Lamboo 907 fg silcrete 85 35
BFI Lamboo 907 fg silcrete 70 40
BFI Derby 566 fg silcrete 51 22
BFI Derby 566 fg silcrete 33 17
BFI Derby 566 glass 54 22
BFI Derby 566 glass 73 23
BFI Derby 560 glass 63 28
BFI Gibb R. 566 glass 33 17
BFI Kimberley 560 glass 70 24
 Downs
BFI Derby 566 quartz 48 20
BFI Derby 566 quartz 46 17
BFI Derby 566 quartz 32 14
BFI Luma 563 quartz 29 14
BFI Saddler S. 563 quartz 29 12
BFI Saddler S. 563 quartzite 28 24
BFI Derby 566 quartzite 45 24
BFI Forrest R. 555 quartzite 58 15
BFI Gibb R. 560 quartzite 57 19
BFI Kimberley D. 555 quartzite 70 17
 Forrest R.
BFI Kimberley D. 555 quartzite 68 21
 Forrest R.
BFI Mowanjum 555 quartzite 73 22
 Forrest R.
BFI Saddler S. 563 quartzite 27 19
BFI Halls C. 549 fg chert 66 32
BFI Moolaboola 549 fg chert 70 32
BFI Saddler S. 563 fg chert 59 29
BIM Derby 560 ceramic 66 24
BIM Rifle Point 563 fg chert 59 44
BIM Wanjina G. 560 fg chert 28 16
BIM Iminji 563 quartzite 31 19
BM Derby 560 ceramic 50 27
BM Fossil D. 563 fg chert 23 15
BM Winjana G. 560 fg chert or 47 24
 ?slate
BM Winjana G. 560 quartzite 28 15
BP Kununurra 560 fg chalcedony 42 16
BP Saddler S. 563 fg chert 42 16
BP Saddler S. 563 fg chert 30 11
BP Umbulgarri 555 quartzite 63 34
 Forrest R.
BP Winjana G. 560 quartzite 57 26
BP Rifle Range 563 silcrete 59 23
UI Jalburu 563 fg chalcedony 55 41
 Xmas C.
UI Adcock G. 560 quartzite 53 34
UI Saddler S. 563 quartzite 20 9
UIM Kununurra 560 fg chert 47 28
UIM Rifle Range 563 fg chert 62 17
UIM Liveringa 563 fg silcrete 45 17
UM Kununurra 560 fg chert 31 17
UM Saddler S. 563 quartzite 63 24

Type Place Thickness Scars/ Max. Breakage/
 (mm) 10mm scar Comments
 max. (mm)

BFI Derby 7 7 6 tip broken
BFI Derby 5 5 6 resin
BFI Kunmunya 7 5
BFI Kunmunya 6 6
BFI Munja 6 9 resin
BFI Munja 5 9
BFI Munja 7 9
BFI Munja 4 ?
BFI Munja 5 5
BFI Munja 6 7
BFI Munja 5 6
BFI Munja 5
BFI Munja 5 5 resin
BFI Tablelands 7 dentate
BFI Rifle Range 6 1 x edge
 unworked
BFI Argyle 6 6 5
 (Durack)
BFI Barramundi Y. 5 broken
 Fossil D.
BFI Carlton 6 6 4 resin
BFI Derby 8 5 7
BFI Derby 7 4 4
BFI Derby 6 4 5
BFI Halls C. 8 4 5
BFI Moolaboola 5 3 4 broken dentate
BFI Rifle Range 5
BFI Saddler C. 6
BFI Saddler S. 6
BFI Lamboo 8 blank
BFI Lamboo 10 blank
BFI Derby 7 5 3 resin
BFI Derby 7 4 ?
BFI Derby 6 -- 5
BFI Derby 5 6 5
BFI Derby 8 blank
BFI Gibb R. 5 5 5
BFI Kimberley 7
 Downs
BFI Derby 7 5 5
BFI Derby 5 ? 6
BFI Derby 7 4 ?
BFI Luma 7
BFI Saddler S. 6
BFI Saddler S. 5 reworked
BFI Derby 6 4 ? reworked
BFI Forrest R. 7
BFI Gibb R. 5
BFI Kimberley D. 3
 Forrest R.
BFI Kimberley D. 7
 Forrest R.
BFI Mowanjum 7
 Forrest R.
BFI Saddler S. 7
BFI Halls C. 8 3 5 dentate
BFI Moolaboola 6 4 4
BFI Saddler S. 9
BIM Derby 7
BIM Rifle Point 11
BIM Wanjina G. 7
BIM Iminji 6
BM Derby 6
BM Fossil D. 4
BM Winjana G. 7
BM Winjana G. 5
BP Kununurra 5
BP Saddler S. 10
BP Saddler S. 5
BP Umbulgarri 13 rounded edge
 Forrest R.
BP Winjana G. 11
BP Rifle Range 11
UI Jalburu 12 26.9.73
 Xmas C.
UI Adcock G. 8
UI Saddler S. 5
UIM Kununurra 9
UIM Rifle Range 9
UIM Liveringa 7
UM Kununurra 5
UM Saddler S. 10

Notes

(1.) Types: BFI = bifacial, invasive, serrate; BIM = bifacial,
some edges marginal, some invasive; BM = bimarginal;
BP = backed point; UI = uniface, invasive flaking; UIM = uniface,
some edges marginal, some bifacial; UM = unifacial,
marginal flaking.

(2.) Places: R. = River; G. = Gorge; D. = Downs; S. = Springs;
Y. = Yard; C. = Creek.

(3.) Other: fg = fine-grained microcrystalline siliceous rock which
includes agate, silcrete, chert; Max. = maximum; No./n = number;
n.p. = no known provenance; Ref. = collection location; Scars/10
mm = number of serrations or dentates per 10 mm, if present;
Standard dev. = standard deviation; ? = precise quantification
was not possible macroscopically.

Table 2 Complete points from Australian Museum collections (n=84)

Type Place Ref. Stone Length Width
 (mm) (mm)

?I Drysdale R. E10581 quartz 31.6 18.7
BFI Wyndham E42835 ceramic 44.6 21.7
BFI Kimberley E29702 fg chalcedony 54.9 27.6
BFI E57191 fg chalcedony 134.6 29.9
BFI E76522 fg chalcedony 71.7 29.6
BFI E85610 fg chalcedony 38 16.5
BFI E85612 fg chalcedony 29.1 12.2
BFI Kimberley E10565 fg chert 78.7 33.18
BFI E54614 fg chert 84 23.1
BFI E54614 fg chert 44.4 25.7
BFI E85332 fg chert 42.1 18.9
BFI Ord R. E05208 glass 66.6 28.5
BFI Ord R. E05211 glass 69.8 28.4
BFI Kimberley E10589 glass 55.7 25.3
BFI Kimberley E10591 glass 100.9 26.9
BFI Kimberley E10593 glass 61.8 28.3
BFI Kimberley E10596 glass 77.5 29
BFI Kimberley E10599 glass 80 31.9
BFI Kimberley E10609 glass 62.08 26.5
BFI Kimberley E10610 glass 87.2 22.2
BFI Kimberley E10613 glass 54.9 24.3
BFI Kimberley E10615 glass 67.4 29.7
BFI Kimberley E10617 glass 99.6 27.7
BFI Kimberley E10618 glass 98.2 28.7
BFI E10803 glass 11.39 33.7
BFI Kimberley E12752 glass 82.3 29.7
BFI NW Aust. E20527 glass 53.8 25.3
BFI NW Aust. E20530 glass 43.1 20.5
BFI NW Aust. E20531 glass 48.6 22.1
BFI NW Aust. E20532 glass 47.9 24.7
BFI NW Aust. E20538 glass 45.2 20.8
BFI NW Aust. E20543 glass 54.8 23.1
BFI Balmaningarra E24074 glass 70.1 24
 NW Aust.
BFI Port George IV E24076 glass 46.2 9.2
 NW Aust.
BFI Leopold R. E24077 glass 88 26
BFI Kimberley E29702 glass 66.3 22.4
BFI Forrest R. E33063 glass 174.8 33.9
BFI E35917 glass 61.6 24.3
BFI Wyndham E42827 glass 116 35.4
BFI Wyndham E42827 glass 12.8 3.4
BFI Wyndham E42828 glass 100.8 32.4
BFI Wyndham E42830 glass 72.7 31.5
BFI Wyndham E42830 glass 92 30.9
BFI Kimberley E50741 glass 37.1 19.4
BFI Kimberley E50741 glass 56.7 22.5
BFI Arnhem Land E53447 glass 87.7 25.2
BFI E58399 glass 71.6 29.6
BFI E58399 glass 97.1 35.3
BFI E58399 glass 62.6 30.1
BFI E58399 glass 62.9 27.5
BFI E58399 glass 61 30.4
BFI E59615 glass 60.3 23.6
BFI E64493 glass 57.3 19.8
BFI E76515 glass 111.18 25.33
BFI Broome E77779 glass 150.7 31.1
BFI Kimberley E34571 porcelain 84.7 24
BFI Kimberley E10625 quartz 27.3 13.1
BFI Kimberley E29669 quartzite 61 19.8
BFI Forrest R. E33045 quartzite 75.2 20.7
BFI E76509 quartzite 43.6 17.4
BFI E76510 quartzite 56.9 20.7
BFI E76511 quartzite 70.24 21.94
BFI Unarinyra E76515 quartzite 82.4 23.3
BFI E76517 quartzite 54.56 19.6
BFI E76518 quartzite 101.2 28.5
BFI Ijarinyia E76520 quartzite 99.8 26.5
BFI E76522 quartzite 74.1 29.4
BFI E76524 quartzite 50.87 20.15
BFI E77778 quartzite 76.2 24.7
BFIM Kimberley E10572 ceramic 58.59 27.27
BFIM Kimberley E50740 ceramic 62.3 37.4
BFIM Kimberley E50741 glass 43.5 24.3
BFIM E58399 glass 72.2 28.8
BFIM E58399 glass 74.2 30
BFIM E58399 glass 77.1 29.8
BFIM E64493 glass 54.7 26.5
BFIM Forrest R. E76526 quartzite 109.5 28.7
BM Balmaningarra E24075 glass 70.5 29.3
 NW Aust.
BM Fitzmaurice R. E29177 glass 94.3 27
BM Kimberley E29700 glass 69.6 40
BM Kimberley E33465 glass 122.9 31.2
BM Granites, CA E44338 glass 86.2 41.5
BM E67207 glass 200.9 43.3
UM Tandanjal E76523 quartzite 181.4 47

Type Place Thickness Serr./ Max. Purchase
 (mm) 10 mm scar (mm) or other date

?I Drysdale R. 9.1 0 10 1902
BFI Wyndham 4.9 5 12 1936
BFI Kimberley 5.8 5 11.8 1925
BFI 10.5 4 18
BFI 7.1 4 16
BFI 5.3 12 8.5
BFI 4.2 6 7
BFI Kimberley 9.56 0 23.5 1902
BFI 8.1 6 13.4
BFI 9.5 0 11 1951
BFI 5.8 7 10.4
BFI Ord R. 5.4 4 14 1895
BFI Ord R. 5.9 4 13.8 1895
BFI Kimberley 5.8 5 18.4 1902
BFI Kimberley 6.4 6 16 1902
BFI Kimberley 5.8 4 17.1 1902
BFI Kimberley 7.4 4 19.7 1902
BFI Kimberley 5.6 5 16.3 1902
BFI Kimberley 7.6 4 12.8 1902
BFI Kimberley 6.5 3 15.9 1902
BFI Kimberley 5.1 6 6.6 1902
BFI Kimberley 6.9 5 9.8 1902
BFI Kimberley 7.5 5 15.4 1902
BFI Kimberley 5.8 5 14.2 1902
BFI 6 5 18.8
BFI Kimberley 7.3 4 13.9 1904
BFI NW Aust. 6.3 4 12.6 1912
BFI NW Aust. 4.21 6 8.8 1912
BFI NW Aust. 4.9 4 12.2 1912
BFI NW Aust. 6.5 5 6.5 1912
BFI NW Aust. 5.7 4 12.3 1912
BFI NW Aust. 6.1 4 12.9 1912
BFI Balmaningarra 4.9 7 8.5 1916
 NW Aust.
BFI Port George IV 6.8 7 13.6 1916
 NW Aust.
BFI Leopold R. 5.3 7 13.9 1916
BFI Kimberley 5.2 5 12 1925
BFI Forrest R. 6 8 20.4 1930
BFI 4.4 7 10.5 1933
BFI Wyndham 6 4 17.7 1936
BFI Wyndham 6.7 3 16.5 1936
BFI Wyndham 5.8 3 17.6 1936
BFI Wyndham 6.5 4 17 1936
BFI Wyndham 5.6 3 17.6 1936
BFI Kimberley 6.3 5 11.1 1943
BFI Kimberley 6.5 5 10.6 1943
BFI Arnhem Land 5.5 3 14.5 1948
BFI 7 4 15.6
BFI 7.4 3 19
BFI 7.9 5 17.5
BFI 6.3 4 14.8
BFI 5.2 6 18.4
BFI 5 7 11.3
BFI 11.4 0 11.5
BFI 6.29 6 15.34
BFI Broome 6 0 17
BFI Kimberley 8.3 5 14.2 1931
BFI Kimberley 5.2 4 6.7 1902
BFI Kimberley 5.8 5 10.5 1925
BFI Forrest R. 7.1 5 12.9 1930
BFI 5.5 8 10
BFI 5.3 8 11.3
BFI 5.84 6 10.81
BFI Unarinyra 8.9 5 13 1928
BFI 8.1 3 8.2
BFI 7.4 5 14.2
BFI Ijarinyia 6.8 4 4 1928
BFI 9.7 0 18.4
BFI 5.44 6 9.67
BFI 5.7 4 11.4
BFIM Kimberley 9.86 8 13.6 1902
BFIM Kimberley 3.3 3 13.9 1943
BFIM Kimberley 3.7 6 10.8 1943
BFIM 6 4 13.9
BFIM 5.8 4 16.9
BFIM 6.2 4 15.9
BFIM 7.8 0 11.9
BFIM Forrest R. 8.2 1 14.1 1946
BM Balmaningarra 6.3 6 10.5 1916
 NW Aust.
BM Fitzmaurice R. 6 4 13.5 1925
BM Kimberley 6.9 3 19.9 1925
BM Kimberley 8.3 4 8 1930
BM Granites, CA 5 3 4.4 1938
BM 5.4 0 12.9
UM Tandanjal 14 2 6.2 1949

Note: See Table 1 notes.

Table 3 All archaeological points (broken and whole) from
excavated sites in Keep River area, east Kimberley (n=41)

Type Site Spit No. Stone Completeness Length
 (mm)

 C1\II 2 1 whole 22.83
 C1\II 16C 7 tip 26.44
BFI BOAB 2 9 silcrete whole 37
BFI BOAB 6 16 silcrete whole 30.2
BFI BOAB 0 2 silcrete mesial 24
BFI C1\I 4 1 quartztite tip 27
BFI C1\I 7 8 quartztite tip 26.21
BFI C1\I 7 10 silcrete whole? 31.22
BFI C1\I 7 11 silcrete tip 16.39
BFI C1\I 9 21 quartzite base 30.95
BFI C1\II 9 10 quartzite whole 34.96
BFI PP2 3 4 chert base 29.06
BFI PP2 3 5 quartzite whole 28.5
BIM BOAB 0 1 quartzite base 31.6
BIM C1\I 6 4 quartztite tip 16.1
BIM C1\I 7 13 silcrete base 26.98
BIM C1\I 7 14 silcrete centre 29.93
BIM C1\I 9 19 quartzite base 23.46
BIM PP2 2 3 silcrete centre 26.53
BP C1\I 7 6 quartztite tip 45.85
BP C1\I 7 7 quartztite tip 40.26
BP C1\I 7 12 silcrete tip 24.75
BP C1\I 9 22 quartzite base 20.28
UI BOAB 1 5 silcrete tip 18.6
UI BOAB 2 10 chert whole 24.9
UI C1/IV 9 1 quartzite tip 18
UI C1/IV 10 2 quartzite whole 24
UI C1\II 20C 5 quartzite base 25.38
UI C1\II 20A 8 quartzite whole 32.38
UI C1\II 17A 11 quartzite tip 36.13
UIM C1\I 6 2 quatztite tip 18.38
UIM C1\I 7 9 quatztite whole 22.76
UIM C1\I 9 16 quartzite whole 30.16
UIM C1\II 16A 3 quartzite tip 27.32
UM BOAB 3 12 silcrete base 25.3
UM C1\I 6 3 quartztite whole 22.78
UM C1\I 7 5 quartztite tip 33.28
UM C1\I 9 17 silcrete base 24.42
UM C1\I 9 18 quartzite tip 16.49
UM C1\II 15 6 silcrete whole 42.2
UM C1\II 17A 12 quarztite tip 25.03

Type Site Width Thickness Serr./ Max. Comments
 (mm) (mm) 10 mm scar
 (mm)
 C1\II 11.5 9.81
 C1\II 18.2 8.34
BFI BOAB 23.9 13.1 0
BFI BOAB 13.8 6.5 0 tip broken
BFI BOAB 23 6 1-2? resin?
BFI C1\I 19 9.45 0
BFI C1\I 16 6.34 0
BFI C1\I 13.3 7.2 0 mostly
 unifacial
BFI C1\I 11.8 9.34 0
BFI C1\I 14.6 7.85 0
BFI C1\II 18.1 7.69 isoc tri plan
BFI PP2 12.7 6.16 0
BFI PP2 12 10.2 5 serrate
BIM BOAB 25.4 10.4 0
BIM C1\I 7.49 2.23 ? serrate
BIM C1\I 17.5 11.69 0 point?
BIM C1\I 18.8 8.67 0
BIM C1\I 15 9.95 0
BIM PP2 21 8.21 5 tip of a
 Kimberley
 dentate point
BP C1\I 15.9 9.67 0 plusUM
 retouch
BP C1\I 22.3 9.63 0
BP C1\I 16.7 9.18 0 marginal at
 tip
BP C1\I 14.3 5.05 0 marginal
 retouch
UI BOAB 13.1 4.5 0
UI BOAB 16.4 15.2 0 pressure
 flaking?
UI C1/IV 10 6 0
UI C1/IV 14.5 5 0
UI C1\II 15.6 6.51
UI C1\II 12.7 6.35
UI C1\II 10.1 5.83
UIM C1\I 10.5 5.03 0
UIM C1\I 14 7.64 0
UIM C1\I 17.6 5.51 0
UIM C1\II 11 7.16
UM BOAB 22.3 9.10
UM C1\I 11.9 4.08 ? serrate marginal at
 tip (cf.
 backed
 blade)
UM C1\I 22.9 6.3 0
UM C1\I 17.7 7.69 0 one margin
 UM
UM C1\I 15.1 6.09 0
UM C1\II 16.3 6.3 5
UM C1\II 13.4 8.07

Note: See Table 1 notes.

Table 4 Summary statistics for points in Akerman collection

All (n=73)

 Length (mm) Width (mm) Thickness (mm)

Mean 47.37 Mean 21.11 Mean 6.74
Standard dev. 17.06 Standard dev. 7.39 Standard dev. 1.95
Range 71.00 Range 35.00 Range 10.00

Bifacial invasive retouch (all materials) (n=46)

 Length (mm) Width (mm) Thickness (mm)

Mean 45.80 Mean 19.04 Mean 6.04
Standard dev. 17.26 Standard dev. 4.81 Standard dev. 1.09
Range 70.00 Range 21.00 Range 5.00

Bifacial invasive retouch (agate) (n=11)

 Length (mm) Width (mm) Thickness (mm)

Mean 35.45 Mean 15.09 Mean 5.55
Standard dev. 8.52 Standard dev. 2.47 Standard dev. 0.93
Range 31.00 Range 8.00 Range 3.00

Bifacial invasive retouch (ceramic) (n=2)

 Length (mm) Width (mm) Thickness (mm)

Range 27-43 Range 16-17 Range 5-7

Bifacial invasive retouch (chert) (n=13)

 Length (mm) Width (mm) Thickness (mm)

Mean 50.77 Mean 22.00 Mean 6.23
Standard dev. 20.92 Standard dev. 5.37 Standard dev. 1.01
Range 63.00 Range 17.00 Range 3.00

Bifacial invasive retouch (glass) (n=5)

 Length (mm) Width (mm) Thickness (mm)

Range 33-70 Range 17-28 Range 5-8

Bifacial invasive retouch (quartz) (n=5)

 Length (mm) Width (mm) Thickness (mm)

Range 29-48 Range 12-20 Range 5-7

Bifacial invasive retouch (quartzite and silcrete) (n=10)

 Length (mm) Width (mm) Thickness (mm)

Mean 51.00 Mean 20.00 Mean 6.10
Standard dev. 17.27 Standard dev. 3.09 Standard dev. 1.37
Range 46.00 Range 9.00 Range 4.00

Bifacial invasive retouch dentate (chert) (n=3)

 Length (mm) Width (mm) Thickness (mm)

Range 59-70 Range 29-32 Range 6-9

Note: See Table 1 notes.

Table 5 Summary statistics for analysed points in
Australian Museum collection

All

 Length (mm) Width (mm) Thickness (mm)

Mean 69.96 Mean 26.73 Mean 6.34
Standard Standard Standard
 dev. 32.34 dev. 6.80 dev. 1.58
Range 202.81 Range 60.20 Range 10.70
Count 206.00 Count 205.00 Count 205.00

Bifacial invasive retouch (all materials) (n=97)

 Length (mm) Width (mm) Thickness (mm)

Mean 67.05 Mean 25.71 Mean 6.30
Standard Standard Standard
 dev. 28.61 dev. 5.76 dev. 1.38
Range 193.01 Range 40.90 Range 7.80
Count 174.00 Count 173.00 Count 173.00

Bifacial invasive retouch (ceramic) (n=4)

 Length (mm) Width (mm) Thickness (mm)

Range 32-57.4 Range 17.9-27.7 Range 4.1-8.4

Bifacial invasive retouch (chalcedony) (n=9)

 Length (mm) Width (mm) Thickness (mm)

Range 20.9-134.6 Range 12.2-29.9 Range 4.2-10.5

Bifacial invasive retouch (chert) (n=10)

 Length (mm) Width (mm) Thickness (mm)

Mean 60.77 Mean 24.67 Mean 7.55
Standard Standard Standard
 dev. 19.23 dev. 5.07 dev. 1.36
Range 54.60 Range 15.28 Range 3.76

Bifacial invasive retouch (glass) (n=55)

 Length (mm) Width (mm) Thickness (mm)

Mean 68.87 Mean 26.70 Mean 6.08
Standard Standard Standard
 dev. 29.70 dev. 5.71 dev. 1.20
Range 193.01 Range 40.90 Range 7.50
Count 128.00 Count 127.00 Count 127.00

Bifacial invasive retouch (quartz) (n=2)

 Length (mm) Width (mm) Thickness (mm)

Range 27.3-114.43 Range 13.1-22.6 Range 5.2-8.42

Bifacial invasive retouch (porcelain) (n=2)

 Length (mm) Width (mm) Thickness (mm)

Range 30.6-84.7 Range 15.3-24 Range 3.6-8.3

Bifacial invasive retouch (quartzite and silcrete) (n=17)

 Length (mm) Width (mm) Thickness (mm)

Mean 68.87 Mean 23.18 Mean 7.21
Standard Standard Standard
 dev. 20.09 dev. 3.49 dev. 1.41
Range 64.21 Range 12.00 Range 4.40

Note: See Table 1 notes.

Table 6 Wear trace and residue analysis of complete
ethnographic points (n=22)

Type Ref. Stone Colour Size Serr./ Use Haft
 10 mm resin

Group 1A

BFI 85607 fg chert red-black 4.6 7-8 SH? ?
BFI 85609 fg chert 6.0 8 SH +
BFI 85610 fg chert orange 4.0 7-12 SH +
BFI 85612 fg chert orange 3.0 6 SH +
BFI 54614-1 fg chert green- 4.4 - ? +
 black
Group 1B

BFI 55047-1 fg chert white 9.5 + (SH) +
UFI 50739 fg chert black 8.9 + SK/SH +
BFI 54614-2 fg chert black 8.5 + SK +

Group 2

BFI 35917 glass brown 6.3 + UN +
BFI 42830 glass trans. 9.5 + UN +
BFI 51408 glass trans. 9.3 + UN +
BFI 53447 glass brown 8.9 + UN +
BFI 85332 fg chert beige 5.1 + UN +

Group 3

BFI 42881 glass green 7.5 - UN -
BFIM 50741-4 glass brown 4.8 + UN -
BFI 50741-5 glass brown 4.6 + UN -
BFI 20525 glass dark green 5.1 + UN -
BFI 57191-1 fg chert white 13.4 + UN -

Group 4

BFI 33061 glass brown 15.1 + UN -
BM 44338A glass brown 7.1 - UN -
BM 44338C glass brown 8.8 - UN -
BFI 33045 quartzite red 7.5 + UN -

Type Ref. Red Yellow White Plant Remarks
 ochre ochre ochre residue

Group 1A

BFI 85607 - + - -
BFI 85609 - + - -
BFI 85610 - + - -
BFI 85612 - + - -
BFI 54614-1 + - - - Too
 weathered
Group 1B

BFI 55047-1 + - - -
UFI 50739 - - - +
BFI 54614-2 + - - +

Group 2

BFI 35917 + - - +
BFI 42830 + - - + BL: -+-
BFI 51408 + - - +
BFI 53447 + - + +
BFI 85332 + - + +

Group 3

BFI 42881 - - - +
BFIM 50741-4 - - - +
BFI 50741-5 - - - +
BFI 20525 - - - + BL:tr++
BFI 57191-1 - - - +

Group 4

BFI 33061 - - - -
BM 44338A - - - -
BM 44338C - - - -
BFI 33045 - - - -

Notes

(1.) Types: BFI = bifacial, invasive, serrate; BIM = bifacial,
some edges marginal, some invasive; BM = bimarginal; BP = backed
point; UI = uniface, invasive flaking; UIM = uniface, some
edges marginal, some bifacial; UM = unifacial,
marginal flaking.

(2.) Places: R. = River; G. = Gorge; D. = Downs; S. = Springs;
Y. = Yard.

(3.) Other: fg = fine-grained microcrystalline siliceous rock which
includes agate, silcrete, chert; Max. = maximum; No./n = number;
NP = no known provenance; Ref. = collection location; Scars/10
mm = number of serrations or dentates per 10 mm, if present;
Standard dev. = standard deviation; ? = precise quantification
was not possible macroscopically; SH = spearhead; SK = traces of
skinning; UN = unknown use; trans. = translucent.

Table 7 Functional groups of ethnographic points analysed
for use-wear and residues (n=22)

Functional Plant Market Size range Use-wear
group wrap (cm)

1A: hafted and no none 3-6 present
used

1B: hafted and yes none and 8-10 present
used: recycled Aboriginal
 exchange

2: hafted and not yes Aboriginal 5-10 absent
used exchange

3: neither hafted yes Aboriginal 4-8 absent
nor used exchange

4: neither hafted no European 7-15 absent
nor used

Functional Residues Hafting Raw
group resin material

1A: hafted and ochre present stone
used

1B: hafted and ochre present stone
used: recycled

2: hafted and not ochre present stone and
used glass

3: neither hafted plant absent glass
nor used

4: neither hafted absent absent stone and
nor used glass

Note: See Table 6 notes.

Table 8 Wear trace and residue analysis of archaeological points

Type Ref. Stone Colour Size Serr./ Use Haft
 10 mm resin

UIM C1/I/2 quartzite red 18.4 0 SH +
BIM C1/I/4 quartzite brown 16.2 indistinct SH -
BIM C1/I/5 quartzite red 33.3 0 SH -
BP C1/I/7 quartzite red 40.3 0 SH -
BIM C1/I/13 quartzite brown 27 0 UN -
UM C1/I/17 silcrete brown 24.4 0 SH +
UM C1/I/23 quartzite yellow 51.6 0 plant -
UIM C1/I/9 quartzite brown 22.8 0 SK ?
BFI C1/I/10 silcrete brown 31.2 0 SH -
BFI C1/I/11 silcrete brown 16.4 0 UN -
BM C1/I/24 quartzite grey 27 3 NP -
BP C1/I/12 silcrete brown 24.8 0 SH -
UI PP2/1 quartzite red 36.8 0 NP -
BIM PP2/3 chert pink 29.1 ?5 ? -
BFI PP2/5 quartzite red 28.5 ?5 SK ?

Type Ref. Red Yellow White Plant Remarks
 ochre ochre ochre residue

UIM C1/I/2 - - - + tip
BIM C1/I/4 - - - - tip
BIM C1/I/5 - - - - tip
BP C1/I/7 - - - - tip
BIM C1/I/13 - - - - base
UM C1/I/17 - - - - base
UM C1/I/23 - - - - whole
UIM C1/I/9 - - - - tip broken
BFI C1/I/10 - - - - tip
BFI C1/I/11 - - - - tip
BM C1/I/24 - - - - whole
BP C1/I/12 - - - - tip
UI PP2/1 - - - - whole
BIM PP2/3 + - - - centre
 dentate
BFI PP2/5 - - - - whole

Note: See Table 6 notes.

ACKNOWLEDGMENTS

Funds from an Australian Research Council fellowship were awarded to Fullagar (1993-97) and an Australian Museum Fellowship was awarded to van Gijn (1998). We also acknowledge a current ARC Large Grant (No. A59905957) awarded to L. Head and R. Fullagar. For advice and assistance in making specimens available, we thank Australian Museum staff, particularly Leanne Brass, Maree Darrell, Stan Florek and Robin Torrence. Blood residue analyses were carried out with assistance from Estelle Lazar, Stephanie Garling and Judith Field. Artefacts were sorted with assistance from students at the University of Wollongong, Judith Field, Laurie Victor and Fiona Leslie. Multivariate statistical results reported here were undertaken by Kristin Schubert, University of Sydney. We also thank Jennifer Atchison and Ken Mulvaney and appreciate comments from various people at seminar presentations, particularly the Australian Archaeological Association Annual Meeting in Mandurah, Western Australia, in 1999.

NOTE

(1.) Hardman was an otherwise excellent observer and a student of the new science of prehistory. He named the Lubbock Range, in the south Kimberley, after Sir John Lubbock, British prehistorian, historian, naturalist and author of Prehistoric Times (1865). Hardman was moved to dedicate the range in Lubbock's honour after discovering a series of chert-topped hills on the flanks of the range that bore evidence of extensive quarrying as well as debitage left from the manufacture of stone tools.

REFERENCES

Akerman, K. 1978 Notes on the Kimberley Stone-Tipped Spear Focusing on the Point Hafting Mechanism, Mankind 11(4), 486-9.

-- 1979a Heat and Lithic Technology in the Kimberleys, WA, Archaeology and Physical Anthropology in Oceania 14(2), 144-51.

-- 1979b Flaking Stone with Wooden Tools, The Artefact 4(3&4), 79-80.

-- 1979c Material Culture and Trade in the Kimberley Today. In R.M. Berndt and C.H. Berndt (eds), Aborigines of the West: Their Past and Present, University of Western Australia Press, 243-51.

Akerman, K. and P. Bindon 1983 Evidence of Aboriginal Lithic Experimentation on the Dampierland Peninsula. In M. Smith (ed.), Archaeology at ANZAAS 1983, Western Australian Museum, 75-80.

-- 1995 Dentate and Related Stone Biface Points from Northern Australia, The Beagle, Records of the Museum and Art Gallery of the Northern Territory 12, 89-99.

Akerman, K. with J. Stanton 1994 Riji and Jakoli: Kimberley Pearlshell in Aboriginal Australia, Monograph Series 4, Northern Territory Museum of Arts and Sciences.

Atchison, J. 2000 Continuity and Change: ALate Holocene and Post Contact History of Aboriginal Environmental Interaction and Vegetation Process from the Keep River Region, Northern Territory, PhD thesis, University of Wollongong.

Balfour, H. 1903 On the Method Employed by the Natives of NW Australia in the Manufacture of Glass Spearheads, Man 35, 65.

-- 1951 A Native Tool Kit from the Kimberley District, Western Australia, Mankind 4(7), 273-4.

Basedow, H. 1925 The Australian Aboriginal, Preece and Sons, Adelaide.

Bordes, F. 1961 Typologie du paleolithique ancien et moyen, 2 vols, Institut de Prehistoire, Universite de Bordeaux, Memoire, 1, Bordeaux.

Bowdler, S. and S. O'Connor 1991 The Dating of the Australian Small Tool Tradition, with New Evidence from the Kimberley, WA, Australian Aboriginal Studies 1, 53-62.

Bradshaw, E. 1986 A Typological Analysis of Five Excavated Stone Tool Assemblages, East Kimberley, Western Australia, BA (Hons) thesis, University of Western Australia.

Carnegie, D.W. 1898 Spinifex and Sand: A Narrative of Five Years Pioneering and Exploration in Western Australia, Arthur Pearson, London.

Clement, E. 1903 Ethnographical Notes on the Western-Australian Aborigines with a Descriptive Catalogue of a Collection of Ethnographical Objects from Western Australia by J.D.E. Shmeltz, Internationales Archiv fur Ethnographie 16, 1-29.

Clendon, M. 1999 Worora Gender Metaphors and Australian Prehistory, Anthropological Linguistics 41(3), 308-55.

Crabtree, D.E. 1970 Flaking Stone with Wooden Implements, Science 169, 146-53.

Davidson, D.S. and F.D. McCarthy 1957 The Distribution and Chronology of Some Important Types of Stone Implements in Western Australia, Anthropos 52, 390-458.

Dortch, C.E. 1977 Early and Late Stone Industrial Phases in Western Australia. In R.V.S. Wright (ed.), Stone Tools as Cultural Markers, Australian Institute of Aboriginal Studies, Canberra, 104-32.

Dortch, C.E. and F. Bordes 1977 Blade and Levallois Technology in Western Australian Prehistory, Quartar Band 27/28, 1-19. Elkin, A.P. 1948 Pressure Flaking in the Northern Kimberley, Australia, Man 130, 110-13.

Etheridge, R. 1890 On Some Beautifully Formed Stone Spearheads, from the Kimberley, Northwest Australia, Records of the New South Wales Geological Society 11, 61-5.

-- 1891 Notes on Australian Aboriginal Stone Weapons and Implements (xi) Spearheads from the Kimberley, Proceedings of the Linnaean Society 6(1), 34-7.

Flenniken, J.J. and J.P. White 1985 Australian Flaked Stone Tools: A Technological Perspective, Records of the Australian Museum 36, 131-51.

Flood, J. 1970 A Point Assemblage from Northern Australia, Archaeology and Physical Anthropology in Oceania 5, 27-52.

Fullagar, R. 1986 Use-wear and Residues on Stone Tools: Functional Analysis and Its Application to Two Southeastern Australian Archaeological Assemblages, PhD thesis, La Trobe University.

-- 1991 The Role of Silica in Polish Formation, Journal of Archaeological Science 18:1-25.

-- (ed.) 1998 A Closer Look: Recent Australian Studies of Stone Tools, Archaeological Computing Laboratory, Archaeology (Prehistoric and Historical), University of Sydney, 149-78.

Fullagar, R., D. Price and L. Head 1996 Early Human Occupation of Northern Australia: Stratigraphy and Dating of the Jinmium Rockshelter, Northern Territory, Antiquity 70, 751-73.

Goring, J. 2000 Gwion Gwion: Secret and Sacred Pathways of the Ngarinyin Aboriginal People of Australia, Konemann, Cologne.

Graham, R. 1984 Kimberley Ethno-archaeology, Records of the Anthropological Society of South Australia Newsletter 22(2), 6.

Hardman, E.T. 1888 Notes on a Collection of Native Weapons and Implements from Tropical West Australia, Proceedings of the Royal Irish Academy, Third Series 1(1), 57-69.

Harrison, R. 2000 Challenging the Authenticity of Antiquity: Contact Archaeology and Native Title in Australia. In I. Lilley (ed.), Native Title and the Transformation of Archaeology in a Postcolonial World, Oceania Publications, University of Sydney, 35-53.

Head, L. and R. Fullagar 1997 Hunter-gatherer Archaeology and Pastoral Contact: Perspectives from Northwest Northern Territory, World Archaeology 28(3), 418-28.

Hiscock, P. 1993 Interpreting the Vertical Distribution of Stone Points within Nauwalbila 1, Arnhem Land, The Beagle, Records of the Northern Territory Museum of Arts and Science 10(1), 173-8.

Hiscock, P. 1994 The End of Points. In M. Sullivan, S. Brockwell and A. Webb (eds), Archaeology in the North, Northern Australia Research Unit, Australian National University, Darwin, 72-83.

Idriess, I. 1937 Over the Range: Sunshine and Shadow in the Kimberley, Angus & Robertson, Sydney.

Jelinek, J. 1975 The Pictorial Encyclopedia of the Evolution of Man, Hamlyn, London.

Jones, R. and I. Johnson 1985 Deaf Adder Gorge: Lindner Site, Nauwilabila1. In R. Jones (ed.), Archaeological Research in Kakadu National Park, Australian National Parks and Wildlife Service, Canberra, 165-228.

Kaberry, P. 1939 Aboriginal Woman: Sacred and Profane, Routledge, London.

Kamminga, J. 1982 Over the Edge, Occasional Papers in Anthropology 12, Anthropological Museum, University of Queensland.

-- 1985 The Pirri Graver, Australian Aboriginal Studies 2, 2-25.

King, P.P. 1827 Narrative of a Survey of the Intertropical and Western Coasts of Australia, Performed between the Years 1818 and 1822, John Murray, London.

Lommel, A. 1997 The Unambal: A Tribe in Northwest Australia, trs. I. Campbell, Takarakka Nowan Kas Publications, Carnarvon Gorge, Qld.

Love, J.R.B. 1917 Notes on the Wororra Tribe of North-Western Australia, Transactions of the Royal Society of South Australia 41, 20-38.

-- 1936 Stone-Age Bushmen of Today, Blackie, London.

-- nd Notes on the Worora, 1927-1940, unpublished ms, Archives of the State Library of South Australia, 1-190.

Loy, T.H. and B.G. Hardy 1992 Blood Residue Analysis of 90,000 Year Old Stone Tools from Tabun Cave, Israel, Antiquity 66, 24-35.

Loy, T.H. and C.D. Matheson (in press) Screening for Prehistoric Blood Residues: Using the Hemastix Colourimetric Test, Journal of Archaeological Science.

Lubbock, J. 1865 Prehistoric Times as Illustrated by Ancient Remains and the Customs of Modern Savages, William and Norgate, London.

Mahony, D.J. 1924 Note on Making Spearheads in the Kimberley District, WA, Australian Association for the Advancement of Science 117, 474-5.

Mathews, R.H. 1901 Ethnological Notes on the Aboriginal Tribes of the Northern Territory, Queensland Geographical Journal 16, 69-90.

McCarthy, F.D. 1939 Trade in Aboriginal Australia and Trade Relations with Torres Strait, New Guinea and Malaya, Oceania 9(4), 405-38; 10(1), 80-104; 10(2), 171-95.

-- 1967 Australian Aboriginal Stone Implements, Australian Museum, Sydney.

Micha, F.J. 1970 Trade and Change in Australian Aboriginal Cultures: Australian Aboriginal Trade as an Expression of Close Culture Contact and as a Mediator of Culture Change. In Pilling and Waterman (eds), Diprotodon to Detribalization, 285-313.

Mitchell, S.R. 1962 Some Ethnological Notes, Victorian Naturalist 78, 361-3.

Moore, M. 2000 Kimberley Spearpoints of Northwestern Australia, Chips 12(3), 5-17.

O'Connor, S. 1999 30,000 Years of Aboriginal Occupation: Kimberley, Northwest Australia, Terra Australis 14, Department of Archaeology and Natural History and Centre for Archaeological Research, Australian National University, Canberra.

Pilling, A.R. and R.A. Waterman (eds) 1970 Diprotodon to Detribalization: Studies of Change among Australian Aborigines, Michigan State University Press, East Lansing.

Plew, M.G., J.C. Woods and M.G. Pavesic (eds) 1985 Stone Tool Analysis: Essays in Honor of Don E. Crabtree, University of New Mexico Press, Albuquerque.

Porteus, S.D. 1931 The Psychology of a Primitive People: A Study of the Australian Aborigine, Edward Arnold, London.

Roberts, R., M. Bird, J. Olley, R. Gallagher, E. Lawson, G. Laslett, H. Yoshida, R. Jones, R. Fullagar, G. Jacobsen and Q. Hua 1998 Optical and Radiocarbon Dating at Jinmium Rock Shelter, Northern Australia, Nature 393, 358-62.

Saville-Kent, W. 1897 The Naturalist in Australia, Chapman and Hall, London.

Serventy, V. 1967 Aboriginal Artefacts at Rottnest Island, Western Australian Naturalist 10(5), 123-4.

Spencer, B. 1928 Wanderings in Wild Australia, Macmillan and Co., London.

Stanner, W.E.H. 1933-34 Ceremonial Economics of the Mulluk Mulluk and Madngella Tribes of the Daly River, North Australia, Oceania XIV(2&4), 156-75; 458-71.

Stuart, E.J. 1923 A Land of Opportunities, Bodley Head, London.

Tindale, N.B. 1965 Stone Implement Making among the Nakako, Ngadadjara and Pitjandjara of the Great Western Desert, Records of the South Australian Museum 15(12), 131-64.

-- 1985 Australian Aboriginal Techniques of Pressure-flaking Stone Implements: Some Personal Observations. In Plew et al (eds), Stone Tool Analysis, 1-13.

Torrence, R. 1993. Ethnoarchaeology, Museum Collections and Prehistoric Exchange: Obsidian-tipped Artefacts from the Admiralty Islands, World Archaeology 24(3), 467-81.

van Gijn, A. 1990 The Wear and Tear of Flint, University of Leiden, Leiden (Analecta Praehistorica Leidensia 22).

Wallis, L. and S. O'Connor 1998 Residues on a Sample of Stone Points from the West Kimberley. In R. Fullagar (ed.), A Closer Look: Recent Australian Studies of Stone Tools, Archaeological Computing Laboratory, Archaeology (Prehistoric and Historical), University of Sydney, 149-78.

White, J.P. and J.F. O'Connell 1982 A Prehistory of Australia, New Guinea and Sahul, Academic Press, Sydney.

Kim Akerman

4 Dorset St, Moonah, Tas. 7009

Richard Fullagar

University of Wollongong

Annelou van Gijn

University of Leiden, Netherlands