Archaeology, claimant connection to sites, and native title: employment of successful categories of data with specific comments on glass artefacts.

Veth, Peter ; O'Connor, Susan

Abstract: This paper argues that Indigenous utilisation and manipulation of post-contact materials should be given more attention in both routine and native title-oriented archaeology. We analyse a range of categories of post-contact occupation, including art, habitation sites, and implements based on introduced materials. We draw particular attention to glass artefacts in that they have the advantage of being near ubiquitous in the landscape. We would argue that judiciously treated they can be highly informative about the timing and nature of settlement and subsistence patterns after contact (revealing both continuities and transformations). They may also reflect on group identity and processes of aggregation during the imposition of pastoral, mining, forestry, defence and other regimes on traditional lands. Of the numerous Federal Court archaeology expert witness reports prepared by the authors in the arid zone, in all cases glass artefacts were recorded. Equally, in all cases claimants had some connections to and knowledge of sites where such post-contact artefacts were noted.

Introduction

Archaeology as expert evidence has been employed since the Mabo decision. For example, it was solicited by the Aboriginal Legal Service of Western Australia in the Martu native title claim in 1992, with multiple reports being produced over the next ten years by one of the authors (PV), contributing towards the successful consent determination of 2002. This resulted in native title being given for 136 000 [km.sup.2] of Western Desert lands (Federal Court of Australia 2002a; Veth 2001, 2003). This was the largest portion of land accorded native title in Australia. The use of archaeological evidence and sites, and claimant connections to these, was initiated by claimants and respondents before the Native Title Act 1993, the passing and amendment of which has pushed evidence towards statutory thresholds rather than Common Law particulars.

Archaeological evidence and claimant connections to sites containing archaeological features have historically been seen as an integral part of land determinations in other settler societies, such as New Zealand, Canada and the USA (cf. Bartlett 1993). There was no reason to think it would have a different valency in the Australian context. Despite the obvious primacy of claimant testimony and contemporary (anthropological) evidence in native title proceedings, there has been a persuasive and ongoing need to locate present and past social configurations (i.e. the group or coalition) in space and time. Archaeology is uniquely equipped to do this, often in a complementary fashion to oral, historical and linguistic narratives.

In this paper we will examine the categories of archaeological data which have been used in selected archaeology expert witness reports prepared for the Federal Court. We will critique how these materials, and claimant connections to them, have been accepted as evidence and where judges have found these to address native title particulars. We will then elaborate on a single class of post-contact artefacts--glass implements--as an exemplar of the opportunities and limitations inherent in the deployment of archaeology in native title cases.

Claimant connections to archaeological sites

In an earlier paper Veth (2000) made an extended case for the relevance of archaeology to native title. He argued that the construction of an 'archaeological landscape', employing conventional methods of inference and reconstruction, without reference to a known historic or contemporary Indigenous group, would not meaningfully address a consideration of native title. Specifically, he argued (2000:80) that archaeology was relevant:

 in its ability to provide both independent and
 complementary evidence to tests for continuing and
 referable occupation, use and exploitation of land some
 time before, during and after colonisation consistent
 with the laws and customs of the original inhabitants.
 Importantly, it allows an added appreciation of the
 transitions that may have occurred to groups through
 enforced changes to social organisation, economy and
 technology such as the imposition of pastoral regimes.
 However, this evidence will need to be framed within
 the context of peoples' known association to sites,
 places and landscapes.

He further suggested that the types and landscapes that might be most meaningfully considered are those that:

* are known to claimants through oral records, use and visitation;

* are part of a domestic or totemic landscape; and

* can possibly provide information on laws and customs via standard archaeological approaches.

He concluded (2000:85) that the optimal approach was a staggered one, involving the following stages:

* intensive background preparation through inspection of cultural heritage databases, reports, relevant literature and extant collections towards a predictive model;

* survey on the claim lands, with claimants focusing on places known to them; and

* documentation of archaeological evidence at these sites (where present) and how this might address particulars of the claim. A combination of data from background research and additional survey at this stage can control for sample bias in characterising the recent archaeology of the claimed lands.

What aspects of the group can archaeology help to establish?

The requirement for the independence of the archaeologist as an expert witness to the Federal Court has been clearly outlined in Practice Directions by the Chief Justice, and essentially notes that the witness has an overriding duty to assist the court in that person's area of expertise; that the witness must not act as an advocate for a party; and that the paramount duty is to the court and not to the person retaining the expert. In this sense, the archaeology expert witness has no obligation to prove that there is a seamless trajectory between the pre-contact and contemporary society (with attendant laws and customs intact) or to make the opposite case. The Practice Directions make it clear that an archaeologist should endeavour to interpret and understand the archaeological evidence and share such knowledge with the court towards an informed decision.

Major claims which have found that native title exists, such as Miriuwung-Gajerrong (Federal Court of Australia 1998, 2003c) and Ngarluma-Yindjibarndi (Federal Court of Australia 2003a), specifically note in their judgements that archaeology has helped to establish:

* long occupational histories of the study areas;

* change and continuity in occupation consistent with the use of a homeland;

* community organisation and connection with the land at sovereignty;

* varying levels of maintenance of connection;

* consistency with historical evidence of connection; and

* occupation reasonably attributable to the claimant group.

In Ben Ward and Ors v State of Western Australia & Ors [1998] 1478 FCA (24 November 1998), Justice Lee did not require that the claimants had a seamless history with the past. There was no expectation that the claimants were in some sense a fossilised society immune from the effects of frontier conflict, pastoral activity or engagement with the modern economy. Rather, different lines of evidence attested to the continuities and transformations experienced by the claimants with the resulting conclusion that their use and occupation of their lands had been ongoing.

Although some subsequent judgements, which have taken a more 'limited' approach (cf. Strelein 2003), have provided little room for historical transformation--for example, The Members of the Yorta Yorta Aboriginal Community (Federal Court of Australia 2002b)--the fact remains that the 'standing brief' for archaeologists is to interpret the archaeological record and share such understanding with the court and not to act as an advocate per se in the case for either historic continuity or change.

Requirements of the Native Title Act 1993

Key particulars for archaeology reports in litigation/ consent determinations (seeking definition of the group, its connection to the claim land/waters, its traditional laws and customs and referable behaviours) include establishing:

* patterns of residence;

* the nature and maintenance of residential mobility patterns;

* subsistence behaviours on both lands and waters;

* the nature of economies based on hunting and gathering of foods;

* the construction and use of shelters;

* aggregation based on ceremonial activities;

* the procurement and use of stone, ochres and minerals;

* the sharing, exchange and trade of resources derived from the land;

* visitation and residence at particular places on the landscape; and

* the production and curation of art.

Although normative in their tenor, these categories immediately evoke images of the numerous bundles of archaeological information which have already been recorded and analysed from numerous parts of Indigenous Australia and which are contained in myriad sources, such as cultural heritage reports, theses, management plans, research papers and regional volumes, or which are physically housed in major holding institutions, such as the Australian Museum. These kinds of information can be readily assessed in situ at sites known to claimants and also recorded during systematic survey of claim areas.

The material spatial and chronological dimensions of these particulars should be immediately obvious, and clearly fall into the domain of a discipline which provides a diachronic perspective on the emergence of contemporary social configurations, characterised inevitably in a settler society both by cultural continuities and transformations.

Although current guidelines for the preparation and content of connection reports prepared for mediation with the state (e.g. WA Government 2004) provide no direction as to how archaeology may be employed, the categories of information required seek data on: (a) the native title claimant group; (b) the traditional system of law and custom; (c) the connection between the claimant group and the claimed area; (d) description of the claimed area; and (e) the claimant group's asserted rights and interests in the claimed area.

Examples of cases where native title was found to exist

Here we will restrict ourselves to cases that one of the authors (PV) has been directly or indirectly involved with, and which have been heard to judgements/ consent determinations, including Martu (Federal Court of Australia 2002a), Ngarluma-Yindjibarndi (Federal Court of Australia 2003a), De Rose Hill--successfully appealed (Federal Court of Australia 2003b), and Miriuwung-Gajerrong (Federal Court of Australia 1998, 2003c; Fullagar & Head 2000; Veth 1998, 1999, 2001). In all cases there were categories of archaeological sites and data which were admitted as evidence to be considered in the question of whether native title had continued to exist or not. We will use specific examples from each of the cases to illustrate how different categories of information have been mobilised. In summary, the kinds of evidence that may be addressed include:

* Aboriginal presence in the form of European materials such as glass, metals and ceramics (often located on pre-contact sites);

* the depiction of European items such as horses and guns in rock-art;

* hunting and gathering behaviours, in the form of food processing sites and dietary remains;

* shelters/caves with evidence for occupation which contain modified European materials;

* ceremonial activity, witnessed by the maintenance of regeneration ceremonies at places comprising stone arrangements/modified landscapes;

* quarrying of stone and ochres still in contemporary circulation;

* distribution of resources, as seen in the exchange of prized commodities, such as pearl shell and ochres;

* camping at specific places on the land, and in the repeated and structured occupation of a range of habitation sites on a homeland; and

* the establishment and maintenance of totemic estates, as witnessed in the production and repainting of art.

Aboriginal presence in the form of European materials

In all of the above claims there were numerous localities where pre-contact assemblages had post-contact artefacts lying on their surfaces as a palimpsest. In some cases pastoral-era outstations contained modified artefacts, and these did not overlie earlier assemblages. Spatial relocation of Aboriginal groups as part of an indentured labour system appears to be reflected in these post-contact occupation sites.

A number of government wells constructed at the end of the nineteenth century in the arid lands of Western Australia--to provide watering points for camel trains, prospectors, and the like--have large assemblages of pre-contact flaked and ground stone artefacts. The wells were often constructed near Aboriginal rockholes, soakages and spring sites. These water sources have been the focus of millennia of occupation and the assemblages visible on contemporary and older land surfaces (where exposed) may number in the hundreds of thousands of artefacts. The routine presence of hundreds (and sometimes thousands) of glass artefacts, which have unquestionably undergone secondary modification at these localities, attests to the presence of an Aboriginal population at/after contact. These glass artefacts additionally document the continuation of a tradition of production, especially where the types have some functional correlates (e.g. tula adzes and concave scrapers). The fact that bottles and objects known to have been produced/discarded during a relatively narrow timeframe (c. 1895-1914) are systematically modified and that later assemblages (say of beer bottles dating to the 1950s) are not, attests to the transferral of a 'tradition'--however, one that has likely undergone further transformations through time (see discussion of glass artefacts below and paper by Harrison, this volume).

In the Kimberley, metal shovel-nose spearheads and glass points are found at such sites; in the arid lands, glass is the most common post-contact artefact class (often close to where bottles have been discarded, thereby providing quarries), and, where near early telegraph lines and mining centres, modified ceramics from insulators, porcelain containers and crockery become more common.

The depiction of European items in rock-art

In three of the five claims, painted and engraved art sites depicting contact themes were documented by either the claimants' or respondents' archaeologists. Examples include the depiction of mounted riders, horses and bullocks, guns, axes, European dwellings and European apparel. Sometimes this contact art overlies demonstrably earlier art, such as a mounted rider painted near engravings with desert varnish (in De Rose Hill), or is found on structures which are known to date to the pastoral era, such as engravings of Europeans with details of their clothes engraved onto dry stone walling used to enclose sheep (in Ngarluma-Yindjibarndi).

Hunting and gathering behaviours

All claims had a plethora of evidence for such behaviours which could be reliably dated to both before and overlapping with sovereignty and/or contact.

For example, in Miriuwung-Gajerrong, the long-term sequences described from the excavation and analysis of cultural assemblages from Monsmont and Miriwun rockshelters in the Ord River Valley were relied on to show the antiquity and breadth of reliance on both riverine/aquatic and terrestrial resources from that catchment. In De Rose Hill, materials were collected in situ and dated. For example, a hearth at an occupation site (a 'traditional' bough shelter, or wiltje), stated by the senior claimant to have been occupied by him in the early 1960s (with his father) while residing and mustering on the station, returned a modern date which due to the presence of atmospheric 'bomb carbon' located it precisely to this period (Alan Hogg, pers. comm.).

In the Western Desert, it is common for groups to leave behind larger grindstones (as site 'furniture') and to place the working surface face down--or to bury the stone altogether. In both the Martu and De Rose Hill cases, senior claimants recalled where they had buried large basal grindstones. In Martu, some of these had been buried by senior women in family groups as early as the 1940s in the more remote portions of their claim. In these claims it was both highly persuasive and salutary that claimants could pre-describe the nature and size of these large basal stones (tjiwa) and then unerringly locate them and dig them up during survey.

Arid land economic fauna, whose presence from excavations within and adjacent to the arid land claims could be reliably dated to the pre-contact period and chronologically overlapped with the dates for sovereignty and contact, were still being procured, processed, consumed and ceremonially 'maintained' by claimants during survey. This referable resource use provided tangible connections between the past and present. That this continuing subsistence pattern was against new pastoral backdrops was not at issue.

Linear and mound middens excavated and dated from the Dampier Archipelago (Ngarluma-Yindjibarndi) attest to the procurement and consumption of marine fauna from the time the sea reached the outer islands (c. 8500 BP), to the base of the Burrup Peninsula (c. 6000 BP) and then through to the contact era (modern dates). Some of the larger mounded and linear middens on West Intercourse Island and elsewhere have contact materials lying on their surface.

Shelters/caves containing modified European materials

There are rockshelters and caves from the Kimberley, Goldfields and wider Western Desert which have evidence for repeated occupations which demonstrably extend into the contact period. The oldest Western Desert site, Serpents Glen, has glass artefacts on its surface (O'Connor et al. 1998). One of the caves excavated and reported on in Martu extends back several thousand years. One of the claimants was born in this cave and reported that he took back clothes and fruit to it, when he first encountered Europeans, in this case Len Beadell in 1963 near Rudall River, to the west of the Canning Stock Route. Needless to say, the shelter has contact materials on its surface. One of us (PV) was shown another exceptionally remote site to the east of the Canning Stock Route where a known individual camped in the early 1970s--again, there were contact items at this site, such as fence-droppers which had been ground down at one end to form edges similar to an adze.

Ceremonial activity

A persuasive case of such activity is the thalu site of the West Pilbara. These can be both stone arrangements and modified earth mounds but also natural features. These are regeneration sites designed to perpetuate a range of economic fauna (such as fish, turtle and kangaroo), natural phenomena (such as rain) and also pestilences focused against enemies (such as sand flies). They were recorded by amateur ethnographers in the late nineteenth century, professional anthropologists such as Radcliffe-Brown in the early twentieth century and researchers in the latter part of that century, and, most significantly, published on (in bilingual illustrated format) and registered with the Department of Indigenous Affairs by Ngarluma-Yindjibarndi claimants and Sites Officers in the early 1990s (Daniel 1990).

Many of these sites are still actively visited and maintained by Aboriginal people in the West Pilbara and some of these people provided first-hand accounts of prior use to the anthropologists and Peter Veth when preparing the expert witness reports for trial. While respondents in the case were quick to challenge the veracity of claimant-thalu connections under various aspects of the Evidence Act, the trial judge dismissed the majority of the objections.

Quarrying of stone and ochres

There are specific ochre and stone quarries from which materials are still being actively procured for a range of ceremonial and ritual activities in the Kimberley, Pilbara and Western Desert regions. While the economic utility of lithic materials is all but gone, particular stone materials and ochres are still procured, traded and used by Aboriginal people from all of these culture areas. The presence of these specific materials in deposits dated to before and overlapping with sovereignty and contact attests to the longevity of these procurement practices. While the functional context through time needs to be ascertained via independent means, the evidence for naming, maintenance and exploitation of quarry sites (some of which have totemic significance) clearly embeds these sites in the cultural landscape of claimants. For example, there are named chalcedony outcrops within the Martu native title area which people still visit and collect materials from. These stones are imbued with additional powers which extend well beyond the functional or technological context (cf. Veth 2001).

Distribution of resources

The presence of coastal shellfish (e.g. baler and pearl shell) and ochres at considerable distance from their known sources in both the Martu and De Rose Hill claims provides clear evidence for exchange down-the-line and gives an insight into the social systems which supported high group mobility and transferral of prized goods.

In a recent paper Smith and Veth (2004) describe baler shell fragments recovered with claimants for the Martu native title claim at sites associated with rainmaking along the Percival Lakes of the Great Sandy Desert. These fragments date from as early as c. 2500 BP right through to the recent period. It appears that at least the chains of connection to the very remote coast extend back several thousand years before European contact.

Camping at specific places

In particularly the Ngarluma-Yindjibarndi and De Rose Hill cases, it was argued that while there was clearly a new residential configuration in place following the impacts of pearling/pastoralism (and restricted access to land), the post-contact use of particular habitation sites was referable to the pre-contact mode. For example, large middens on the Abydos Plain and Dampier Archipelago had marine fauna and contact artefacts added to them after the advent of pearling, pastoralism and then mining.

Habitation sites based on dune tongues on De Rose Hill Station had clear evidence for habitation continuing during the pastoral era (twentieth century), containing modified glass and metal artefacts. There were also multiple occurrences of both 'traditional' domed bough shelters (wiltje) and 'transitional' square-framed shelters--both kinds had grinding stones and other contact materials in association. Perhaps most persuasive was the fact that apparently 'traditional' shelters were often bound with fencing wire or had the addition of flattened kerosene tins-some of these identical to the earliest European structures on the station and presumably recycled once the original structures were abandoned. Early tobacco tins and matchboxes, toys made from wire, and wooden items such as digging sticks, boomerangs and spear-throwers all combined to attest to the exclusive (or shared) Aboriginality of these early pastoral sites.

The establishment and maintenance of totemic estates

Perhaps the clearest example of ongoing maintenance of totemic sites comes from varied West Kimberley claims and the Miriuwung-Gajerrong claim, where totemic sites are retouched (repainted) with ochres procured from quarries. As with the copiously described Wandjina repainting sites of the south-west Kimberley, these are much more than solely art sites or places where markings have been made. Claimants and clans exert rights to repaint such sites at multiple levels and these connections can be elaborated on as claimants clearly over-paint earlier phases of art production. Archaeological approaches involving the documentation of superimposition, patina and rates of weathering allow reasonable inferences about phases of art production, as being either more recent or of greater antiquity, to be made. Direct dating of organic admixtures can provide absolute dates for such phases.

Glass artefacts as a contact category

One of the most commonly reported items of material culture that was adopted by Aboriginal people at contact with Europeans / settlers was glass, essentially as a replacement for the stone that was used to fashion a wide range of flaked stone artefacts (see Harrison, this volume). Many of these artefacts were used in the manufacture, decoration and maintenance of wooden implements, such as spear-throwers and shields, or for the butchering of game.

Gould (1968:177) comments on the use of glass in the Western Desert in the 1960s:

 Contact with advanced technology has caused rapid
 changes in aborigines' choice of tool materials. A piece
 of glass ... is used to sharpen a spear, while the leaf of a
 Landrover spring ... has replaced the stone adze ... New
 or old, the tools are used in identical fashion.

Further comment is made by Gould, Koster and Sontz (1971:165) from the Cundeelee Reserve:

 Sometimes an untrimmed piece of broken bottle-glass
 will serve as an impromptu spokeshave or (as on one
 occasion witnessed at the Cundeelee Reserve) as a
 circumcision knife.

A number of archaeological studies in the Australian context provide a firm basis for assessing the technology, form and age of glass artefacts and also of understanding variations that have been noted due to cultural differences at a regional level (see reviews in Birmingham 2000; Bolton 1999; Harrison 2000, 2002, 2004b and this volume; McCarron-Benson 2004; Paterson 1999; Ulm et al. 1999; Williamson 2002). Although analyses of Aboriginal glass artefacts can be found in a range of earlier theses and papers (e.g. Allen 1969; Allen & Jones 1980), a critical assessment of technological modes of production, use-wear patterns and organic residues has been a relatively recent development.

Of particular relevance are two studies that examined assemblages of glass artefacts which come from the arid zones of Western Australia and Central Australia. Harrison (2000:39) compared glass artefacts from the same kinds of late nineteenth century bottles recovered from Shark Bay and the Southwest. He noted that the majority from the arid study area were fashioned from the bases of bottles (>90%), while most from the Southwest were made on bottle body parts. Regional differences are also suggested in glass artefacts from the Kimberley and from the Pilbara (Harrison, pets.comm.).

Harrison (2000:39) divided glass pieces into four artefact types based on morphological characteristics and the presence of use-wear, residues or secondary modification:

* glass flakes with a single ventral surface, a point of applied force or intact margins;

* glass cores where at least one negative flake scar having an inverse percussive bulb and/or inverse feather or hinge termination;

* worked fragments which, while neither flakes nor cores, had use-wear, residues or apparent secondary edge modification; and

* other category to accommodate items such as burnt glass.

Many of the bottles described are the same as those recorded from the claims and include: olive-green turn-past ale bottles whose production is dated c. 1880-1920; and black (dark olive) glass ale bottles dating from the late nineteenth century.

The study of glass artefacts by Bolton (1999) focused on a functional analysis of artefacts from the Illamurta Springs Police camp, Northern Territory, which operated from 1893 to 1912. Ethnographic evidence suggested that the glass had been used by Aboriginal people when they came into contact with Europeans, and this was confirmed through the functional analysis with evidence for the scraping and engraving of wood. While no preference for bottle part was recorded, the darker olive bottles were preferentially selected.

Bolton (1999:10) considers dating in some detail and notes how amethyst glass was manufactured before 1915, as the supply of magnesium it was treated with to make it clear was lost at the outbreak of World War I. Dark olive, or black, glass was used in the 1890s. Olive bottles with ring seals were used for alcohol such as champagne until c. 1900-20, while amber bottles with ring seals were made by the Pickaxe Brand in 1911 and 1912. With reference to how dates can be bracketed, Bolton (1999:19) notes:

 Dating the glass provides the earliest date that the glass
 arrived at the site (terminus post quem). It also provides
 an approximate date of when the glass was used. For
 example, if there are two deposits of glass at a site, one
 dating from the 1890s and one from the 1950s, and it
 was found that only the glass from the 1890s had been
 used by Aborigines, it is possible that the Aborigines
 used the 1890s glass before the 1950s.

Bolton (1999:16) advocates an integrated approach in the identification of utilised glass. Features to note include presence of flaking, secondary modification (retouch flakes), microscopic evidence of use and the presence of residues, and the context in which the artefact was found.

For the purposes of field inspection, the criteria that might be employed to assess glass artefacts and which are consistent with recent critical reviews included the presence of: point of applied force, ventral surface and margins on flakes; negative flake scars; retouch flakes; and undercutting, step fractures, crushing and polishing (under 10x hand lens).

A detailed consideration of use-wear and residues is beyond the scope of most first-stage inspections of artefacts in the field. An extremely conservative approach must be employed in accepting the deliberate modification/use of a glass bottle or fragment, especially given the range of taphonomic factors that can cause accidental damage to glass on open sites (cf. Allen & Jones 1980; Cooper & Bowdler 1998). Harrison (2003) and Wolski and Loy (1999) have drawn attention to the differences between formal and 'ad hoc' glass tools in showing that many unmodified edges of glass bottles in south-eastern assemblages have been used as tools while, conversely, many formal tools have not been utilised.

Comments on glass implements from the arid lands of Western Australia

The presence of formal glass artefacts from the Kimberley region of Western Australia is well documented, with the pressure-flaked Kimberley points being well described in the literature (cf. Harrison 2004b). Here we want to profile the presence of a much wider range of formal glass artefacts from the arid zone of Western Australia which are certainly much less well known.

During the course of several native title surveys in the state's arid zone (Figure 1), and especially during more intensive inspection of contact sites where early Aboriginal groups are known to have at least temporarily co-resided with early mining, pastoral and government depots, we have identified significant numbers of modified glass artefacts. Major complexes of demonstrably pre-contact lithic artefacts have palimpsest assemblages of thousands of pieces of both modified and unmodified glass. What was unexpected was the recurrence of standardised forms in glass, mirroring those in stone. Types included tula adzes (both minimally reduced and also in slug form), burren adzes, thumbnail, notched and nosed scrapers, and specialised wood engraving forms such as the marni wadna. Some of these forms occur repeatedly at a variety of sites where aggregation is known to have occurred at contact.

[FIGURE 1 OMITTED]

The presence of many thousands of formal glass artefacts in the arid zone of Western Australia (away from ports, collectors and other external sources which might influence mode of production) raises some important questions about their intended purpose, function and perhaps even their meaning. These questions might include:

* Are these formal artefacts evidence for the continuation of a technology of production and set of cultural attributes that extend into the historic period?

* Is the ready adoption of glass for long use-life implements evidence that it is a highly prized and amenable new medium?

* Is the evidence for continuance of standardisation an ongoing technological response to mitigate risk (after Hiscock 1994)?

* Could these formal types have acted as markers of group identity and cohesion in contested spaces when disparate groups were brought into atypically close and extended proximity?

These and related issues will now be explored here. We would emphasise that the main purpose of this part of the paper is to highlight the presence of these formal artefacts and to be discursive, rather than attempt to provide absolute answers to these varied questions. All materials have been recorded in situ and we believe that a program of systematic collection and detailed analysis at key sites is now warranted.

Regional variability in glass artefacts

in their discussion of the problems of identification of Indigenously flaked glass artefacts and of distinguishing these from discarded pieces of glass which have been fortuitously flaked, Cooper and Bowdler (1998:75) argue that, with the exception of Kimberley points, formal morphological types were not made in glass, noting that 'Kimberley points ... are unusual, as they are clearly identifiable as artefactual whether hafted or not ... We are not aware of other examples of composite tools including a flaked glass component'.

Recent studies, however, demonstrate that this dearth of formal patterned glass artefacts results from the sparse attention that archaeologists have traditionally accorded to glass assemblages, rather than reflecting a low-level regional variability in this component of the archaeological record. The proliferation of research projects focused on cross-cultural engagement in the pastoral and mining industries, and claimant-led (rather than research-led) recording of known sites for native title, have produced many hundreds of sites with flaked glass assemblages and brought to light significant regional variations in Indigenously worked glass. Our preliminary observations indicate that these variations occur at both the assemblage and artefact level. Indeed, for the Kimberley, Harrison (2004a, 2004b) has noted that there are predictable assemblage differences depending on site context. He found that there are more glass artefacts in general, as well as more formal glass artefacts, on 'fringe' settlements associated with European places (e.g. pastoral camps, town camps, prisons) than on contemporary 'bush' settlements.

Over the past five years the authors have undertaken surveys throughout the eastern Goldfields of Western Australia--including contact sites at and near Leonora, Laverton, Wiluna, Kalgoorlie and Coolgardie (and to the east of here in the Great Victoria Desert and on the Nullarbor Plain)--for native title claims and other purposes. In a more purposeful search for modified glass it became apparent how numerous these artefacts were at major aggregation sites (usually associated with early contact), the glass clearly being introduced onto sites both as whole bottles which were then reduced or in already reduced forms (i.e. only part of the bottle or jar had been brought onto the site for use or further reduction).

Our greatest surprise was in the high percentage that appeared to have been manufactured into tools with formal morphology that had parallels in the lithic domain. Our overview of consultancy reports for this region indicated that while this fact had not gone entirely unrecognised, it had failed to make its way into the academic literature. For example, Builth and Draper (1999) report that they found many glass 'adzes' and retouched implements in an area they surveyed in Kalgoorlie and this was despite the fact that the area had been designated to be set aside for its bushland value and had previously been cleared of European materials. The morphological types they report include an anvil rested backed adze made from purple glass and a 'very fine example of a tula adze ... created from a glass with a thick base'. The latter artefact featured an air bubble which had been centrally positioned in the finished implement (Builth & Draper 1999:4).

Discussion of specific examples

The most commonly occurring formal artefacts in glass from the WA arid lands include tula adzes and slugs, burren adzes, steep-edged, concave and discoidal retouched/utilised flakes, geometric microliths, and engravers. The majority of these implements have been manufactured on the base, kick-up and neck/rim segments of dark ale and champagne bottles. The apparent retouch/use of body segments appears to be less than, say, in the south-west of the state and consistent with patterns noted by Harrison (this volume) for the Murchison and Pilbara. It is reasonable to note that these areas are also arid zone habitats where stone implements were made for the manufacture and maintenance of hardwood artefacts (such as shields, spear-throwers and bowls).

Numerous bottle bases have also been used as single platform cores, with the orientation of flaking towards either the interior or exterior. From our field notes we can observe several emerging patterns in artefact form and bottle portion:

* Tula adzes are largely made on (flakes/portions) from bottle bases.

* Burren adzes are made on bases/lower portions of robust bottles (Figure 2).

* Geometric microliths are made on straight body segments / flakes (Figure 3).

* (En)gravers are made on turned surfaces such as bottle lips and necks (Figure 4).

* Steep-angled scrapers are largely made on robust basal portions (Figure 5).

* Discoidal scrapers are made on gently curved body/shoulder portions (Figure 6).

[FIGURES 2-6 OMITTED]

We believe that these early observations now invite detailed recording of these sites, with sample collections subjected to technological, use-wear and residue studies. As noted above, the large size of many of these assemblages will permit this kind of analysis (thousands of pieces of segments / flakes with hundreds of formal artefacts).

These observations lead to the inevitable conclusion that formal glass types are not restricted to the more obvious and spectacular Kimberley point style and that the glass forms are morphologically and technologically referable to lithic suites found in the same areas.

What can glass artefacts tell us about contact settings?

The transferral of bifacial point production from cryptocrystalline materials (such as silcrete and chalcedony) in the Kimberley into larger glass forms after contact has been argued to be a product of its commoditisation for consumption by and for national and international museums (Harrison 2004a). These same incentives were not in place for more mundane objects, such as tula adzes, in the arid lands of Western Australia, however. The production of broad platform flakes (from bottle bases) with deeply convex dorsal surfaces, and then the subsequent use and retouch of these artefacts, initially opposite the platform, is identical to the technology of formal production and reduction witnessed in the stone analogues.

Perhaps the most parsimonious explanation for the mirroring of adzes, scrapers and specialist engravers from stone to glass forms (using apparently similar technological approaches) is that the same organisational systems for implement manufacture and maintenance, and the extraction of resources, were still in place following contact. What we might be witnessing here is simply the transfer from one optimal raw material to another even more desirable supply. The preference for the use of fine-grained highly siliceous stone materials with predictable fracturing qualities, such as chert, chalcedony and silcrete, for the manufacture of 'formal types' can be seen in the qualitative and quantitative assessments of large arid zone lithic assemblages (cf. Veth 2001). In many cases these hard, controllable raw materials were not available locally near the freshwater source which formed the focus of habitation, and their acquisition would need to have been embedded in a wide economic or social procurement round.

The consumption of alcohol and other beverages at early prospector, provisioning and watering points which were one and the same as, or close by, the watering points that were the focus of Indigenous habitation appears to have been ubiquitous across the WA arid zone and perhaps brought a new source of high-quality material to hand. That these large volumes of thick bottle glass were discarded at, or near, these settlements seems axiomatic. Significant 'quarries' of flakeable glass become present, therefore, almost immediately after European settlement in the final years of the nineteenth century. At the same time the increasing presence of European prospectors and the establishment of government wells and ration depots may have brought changes that curtailed traditional range or the frequency of procurement of fine-grained stone supplies. In this scenario, glass simply substitutes for stone as a fine-grained raw material with controllable fracturing properties.

Another reason for the continuation of these formal types into a new medium (and the technological strategies which produced them) might also be the phenomenon of the continuation of 'ideal' types, whereby isochrestic and/or aesthetic factors result in the reproduction of proscribed templates which may or may not have associated symbolic dimensions or meanings. Many of the larger contact assemblages we inspected were at places which are known historically to have served as aggregation locales (after Veth & McDonald 2002).

This appears to have been a unique historical process whereby social gatherings of speakers of different dialects and languages not only came together in possibly larger numbers than would have occurred in pre-contact times, but also for longer periods of time. Certainly ration depots provided continuous supply lines which were a novel experience for most arid land peoples. The pressure to assert group identity and possibly negotiate new forms of alliance (and enmity) would accelerate in these situations. An increase in group identifying behaviour is predicted where additional external stresses are encountered (McDonald 1994; Wiessner 1990). The production of recognisable forms may have intensified in the 'contested' spaces of these aggregation centres. Harrison (2002:372) extends this line of argument, in the context of glass Kimberley point manufacture by dislocated Aboriginal men, to include gender signalling.

Glass bifacial points have been subdivided on the basis of size, shape and margin treatment (Akerman & Bindon 1995). However, even within this relatively well-studied class of glass artefact, opinion is divided over factors responsible for resultant variability in gross morphology and margin finish. Suggestions include selective pressure on form as a result of European collector preference (Harrison 2004a); new styles or 'fashions' which flourished when a new medium allowed elaboration and experimentation within traditional types which were stimulated by wunan exchange (Akerman 1978; Akerman et al. 2002); expressions of ethnic/group identity; expressions of individual skill and style; size and availability of glass preforms; and metal indentors for pressure flaking.

While we can be certain that European collectors were not responsible for promoting or stimulating the production of glass tulas, burrens, scrapers or microliths in the regions discussed above, it does seem that the continuity of the manufacture of some of these types may extend beyond purely functional requirements. For example, the selective use of bottle base segments for tula adzes, some of which show deliberate aesthetic centring of included air bubbles and portions of the kick-up, may provide an analogue to the well-documented aesthetic of Kimberley point production described by Love (1936:75-6) and other commentators.

Why do these forms occur in such numbers in this culture bloc?

One of the most intriguing aspects of our observations is the fact that these formal artefacts are present in noticeable numbers in the Western Australian arid zone. This is an extensive region where many previous descriptions of lithic assemblages have characterised the majority of implements as being multifunctional and utilitarian. While use-wear and residue studies of a sample of bottle segments and flakes from these sites are necessary to test this assumption (after Bolton 1999), it is worth briefly revisiting some previous studies and rationales.

In an earlier study Gould et al. (1971) stressed the ad hoc nature and form of Western Desert implements. However, they did note that in the utilised category there was a bimodal distribution between those that were (thought to) represent cutting versus scraping functions. While Gould (1977) has certainly espoused the formal typological approach (see also Hiscock & Veth 1991), his contemporary observations of the use of modern materials suggest that this was impromptu.

One of the central tenets of Hayden's (1976) doctoral thesis was the homogeneity between assemblages of Western Desert habitation sites and, again, the ad hoc nature of many maintenance tools. While his ethnoarchaeological work (Hayden 1979) does examine post-contact artefact categories and attributes (e.g. form, edge-angles, fracture mechanics and wear patterns), these are focused on lithic rather than modern materials. The multifunctionality of implements and the marred lines between form and function appear real--apart from, again, the artefacts' edge-angle characteristics.

In contrast to these kinds of conclusions about the impromptu nature of Western Desert forms, however, was a study on the ethno-taxonomy and form of implements from the northern Great Sandy Desert, which led Cane (1984:172) to conclude that ethno-taxonomical and functional data indeed corresponded closely with ethnographically recorded use-categories: 'Of particular interest also is the close correlation between the Aboriginal taxonomy of stone tools and the formal stone tool typology used in this thesis and by other archaeologists working in arid Australia'. Equally, Veth (1993) found that there were repeatedly recurring implement forms in both lithic and post-contact materials from the southern Great Sandy and Little Sandy deserts.

Again, any debate about the relative frequencies of formal glass artefacts in the arid zone, their proportion of overall glass segment/flake assemblages and degree of regional formal variation (after Harrison 2000) must reply on analysis of controlled collections, including the technology of production, use-wear and residue studies.

Conclusion

We have carried out systematic survey and some in situ recording of post-contact glass assemblages and formal artefacts located at early aggregation locales, such as government wells and ration depots. Over 100 sites have been found in the vicinity of early pastoral and prospecting centres of the WA arid zone (e.g. near Leonora, Laverton, Wiluna, Kalgoorlie and Coolgardie) and on the western edge of the Nullarbor Plain. Other researchers (e.g. Builth & Draper 1999) also note the presence of formal glass artefacts at such centres. We conclude that the occurrence of formal glass artefacts, such as tula adzes, discoidal 'scrapers' and geometric microliths, is not a fortuitous product of taphonomic processes. It remains for future research to investigate these artefacts as components of total assemblages in the same way as has been undertaken for their lithic counterparts. Only then will it be possible to begin to evaluate the technological, functional and/or symbolic dimensions of their production. In the meantime, however, glass artefacts constitute powerful evidence of the continuity of traditional practice throughout the arid and semiarid regions of Western Australia.

While we would agree with Harrison (this volume) that Indigenous utilisation and manipulation of other types of post-contact materials should be given more attention, glass artefacts have the advantage that they are ubiquitous in the landscape. The nature of specific claimant connections to country (as recorded by both archaeologists and anthropologists both before Mabo and during native title surveys) included oral tradition, birth, labour history, visitation during their lifetime, or actual knowledge of historic structures and artefacts. Of the 14 Federal Court archaeology expert witness reports prepared by the authors in the arid zone, in all cases at least some claimants had such connections and knowledge. In all cases glass artefacts have been recorded.

REFERENCES

Akerman, K 1978, 'Notes on the Kimberley stone-tipped spear focusing on the point hafting mechanism', Mankind 11(4):486-90.

Akerman, K & Bindon, P 1995, "Dentate and related stone biface points from Northern Australia', The Beagle: Records of the Museums and Art Galleries of the Northern Territory 12:89-99.

Akerman, K, Fullagar, R & van Gijn, A 2002, 'Weapons and wunan: production function and exchange of Kimberley points', Australian Aboriginal Studies 2002 / 1:13-42.

Allen, FJ 1969, 'Archaeology and the history of Port Essington', PhD thesis, Australian National University.

Allen, FJ & Jones, R 1980, 'Oyster Cove: archaeological traces of the last Tasmanians and notes on the criteria for the authentication of flaked glass artefacts', Papers and Proceedings of the Royal Society of Tasmania 114:225-33.

Bartlett, RH 1993, The Mabo decision: with commentary by Richard Bartlett, Butterworths, Sydney.

Birmingham, J 2000, 'Resistance, creolization or optimal foraging at Killalpaninna Mission, South Australia', in R Torrence & A Clarke (eds), The archaeology of difference: negotiating cross-cultural engagements in Oceania, One World Archaeology 38, Routledge, London & New York, pp. 360-405.

Bolton, S 1999, 'Towards a functional analysis of glass artefacts from central Australia', BA (Honours) thesis, University of Sydney.

Builth, H & Draper, N 1999, 'Karlkurla Bushland Park Project, Kalgoorlie: results of an archaeological site inspection 2nd July, 1999', Australian Cultural Heritage Management report prepared for Goldfields Land Council, Kalgoorlie.

Cane, S 1984, 'Desert camps: a case study of stone artefacts and Aboriginal behaviour in the Western Desert', PhD thesis, Australian National University.

Cooper, Z & Bowdler, S 1998, 'Flaked glass tools from the Andaman Islands and Australia', Asian Perspectives 37(1):74-83.

Daniel, D 1990, 'Thalu sites of the West Pilbara', unpublished report, AAD, Perth.

Federal Court of Australia 1998, Ben Ward and Ors v State of Western Australia & Ors [1998] 1478 FCA (24 November 1998).

-- 2002a, James on behalf of the Martu People v State of Western Australia [2002] FCA 1208 (27 September 2002).

--- 2002b, Members of the Yorta Yorta Aboriginal Community v Victoria [2002] HCA 58 (December 2002).

-- 2003a, Daniel v State of Western Australia [2003] FCA 666 (3 July 2003).

-- 2003b, De Rose v State of South Australia [2003] FCAFC 286 (16 December 2003).

-- 2003c, State of Western Australia v Ward [2003] FCAFC 124 (3 June 2003).

Fullagar, R & Head, L 2000, 'Archaeology and native title in Australia: national and local perspectives', in I Lilley (ed.), Native title and the transformation of archaeology in the postcolonial world, Monograph 50, Oceania Publications, Sydney, pp. 24-34.

Gould, RA 1968, 'Preliminary report on excavations at Puntutjarpa rocksbelter, near the Warburton Ranges, Western Australia', Archaeology and Physical Anthropology in Oceania 3:161-85.

-- 1977, "Puntutjarpa rockshelter and the Australian desert culture', Anthropological Papers 54, American Museum of Natural History, New York.

Gould, RA, Koster, DA & Sontz, AHL 1971, 'The lithic assemblage of the Western Desert Aborigines of Australia', American Antiquity 36:149-69.

Harrison, R 2000, '"Nowadays with glass": regional variation in Aboriginal bottle glass artefacts from Western Australia', Archaeology in Oceania 35(1):34-47.

-- 2002, 'Archaeology and the colonial encounter', Journal of Social Archaeology 2(3):352-77.

-- 2003, 'The magical virtue of these sharp things: mimesis, colonialism and knapped bottle glass artefacts in Australia', Journal of Material Culture 8(3):311-36.

-- 2004a, 'Kimberley points and colonial preference: new insights into the chronology of pressure flaked point forms from the southeast Kimberley, Western Australia', Archaeology in Oceania 39(1):1-11.

-- 2004b, 'Contact archaeology and the landscapes of pastoralism in the north of Australia', in T Murray (ed.), The archaeology of culture contact in settler societies, New Directions in Archaeology, Cambridge University Press, pp. 109-43.

Harrison, R & Williamson, C (eds) 2002, After Captain Cook: the archaeology of the recent Indigenous past in Australia, Sydney University Archaeological Methods Series 8, Sydney.

Hayden, B 1976, 'Australian Western Desert lithic technology: an ethno-archaeological study of variability in material culture', PhD thesis, University of Toronto.

-- 1979, Palaeolithic reflections: lithic technology and ethnographic excavation among Australian Aborigines, Australian Institute of Aboriginal Studies, Canberra.

Hiscock, P 1994, 'Technological responses to risk in Holocene Australia', Journal of World Prehistory 8:267-91.

Hiscock, P & Veth, P 1991, 'Change in the Australian desert culture: a reanalysis of tulas from Puntutjarpa Rockshelter', World Archaeology 22(3):332-45.

Love, JRB 1936, Stone-age bushmen of today: life and adventure among the savages in north-western Australia, Blackie, London.

McCarron-Benson, A 2004, 'Native title archaeology: the role of archaeology in native title litigated determinations in Australia', BA (Honours) thesis, Australian National University.

McDonald, JJ 1994, 'Dreamtime superhighway: an analysis of Sydney Basin rock art and prehistoric information exchange', PhD thesis, Australian National University.

O'Connor, S, Veth, P & Campbell, C 1998, 'Serpent's Glen Rockshelter: report of the first Pleistocene-aged occupation sequence from the Western Desert', Australian Archaeology 46:12-21.

Paterson, A 1999, 'Confronting the sources: the archaeology of culture contact in the south-western Lake Eyre Basin, Central Australia', PhD thesis, University of Sydney.

Smith, MA & Veth, PM 2004, 'Radiocarbon dates for baler shell in the Great Sandy Desert', Australian Archaeology 58:37-8.

Strelein, L 2003, Members of the Yorta Yorta Aboriginal Community v Victoria [2002] HCA 58 (12 December)--Comment. Land, Rights, Laws: Issues of Native Title, Volume 2: Issues Paper No. 21, Native Title Research Unit, AIATSIS, Canberra.

Ulm, S, Eales, T & L'Estrange, S 1999, 'Post-European Aboriginal occupation of the southern Curtis Coast, central Queensland', Australian Archaeology 48:42-3.

Veth, PM 1993, Islands in the interior: the dynamics of prehistoric adaptations within the arid zone of Australia, International Monographs in Prehistory, Archaeological Series 3, Ann Arbor, Michigan.

-- 1998, 'Statement of archaeological evidence in the De Rose Hill native title claim', unpublished report prepared for Aboriginal Legal Rights Movement Inc., Adelaide.

-- 1999, 'Statement of archaeological evidence in the Ngarluma-Yindjibarndi native title claim', Aboriginal Legal Service of Western Australia, Perth.

-- 2000, 'The contemporary voice of archaeology and its relevance to native title', in I Lilley (ed.), Native title and the transformation of archaeology in the postcolonial world, Monograph 50, Oceania Publications, Sydney, pp. 78-87.

-- 2001, 'Statement of archaeological evidence in the Martu native title claim', unpublished report prepared for Ngaanyatjarra Council (Aboriginal Corporation), Alice Springs.

-- 2003, '"Abandonment" or maintenance of country? a critical examination of mobility patterns and implications for native title', Land, Rights, Laws: Issues of Native Title, Volume 2: Issues Paper No. 2, Native Title Research Unit, AIATSIS, Canberra.

Veth, P & McDonald, J 2002, 'Can archaeology be used to address the principle of exclusive possession in native title?', in Harrison & Williamson, After Captain Cook, pp. 121-9.

WA (Western Australian) Government 2004, Guidelines for the provision of information in support of applications for a determination of native title, Office of Native Title, Perth.

Wiessner, P 1990, 'Is there a unity to style?', in M Conkey & C Hastorf (eds), The uses of style in archaeology, New Directions in Archaeology, Cambridge University Press, pp. 105-12.

Williamson, C 2002, 'Finding meaning in the patterns: the analysis of material culture from a contact site in Tasmania', in Harrison & Williamson, After Captain Cook, pp. 75-102.

Wolski, N & Loy, TH 1999, 'On the invisibility of contact: residue analyses on Aboriginal glass artefacts from western Victoria', The Artefact 22:65-73.

Peter Veth

AIATSIS, Canberra

Susan O'Connor

Australian National University, Canberra

Peter Veth is Director of Research at AIATSIS and also holds an adjunct Senior Research Fellowship at the Australian National University. He specialises in the emergence and nature of desert societies. He has also carried out longer-term research and cultural heritage projects in most parts of Australia and more recently in the Aru Islands and East Timor. Peter has prepared numerous reports for Federal Court native title cases in Australia.

Susan O'Connor is Associate Professor of Archaeology and currently Head of the School of Archaeology and Natural History, Research School of Pacific and Asian Studies, Australian National University. She has specialised on the archaeology of the Kimberley and the emergence of maritime societies. Susan has carried out major research projects in both northern and southern Australia and also initiated long-term research within the islands of eastern Indonesia.