Sophie Lissonnet, School of Indigenous Australian Studies, James Cook University [HREF1], Sophie.Lissonnet@jcu.edu.au.
Liddy Nevile, Applied Computing Research Institute [HREF2], La Trobe University [HREF3], Liddy.Nevile@motile.net
The Quinkan Matchbox Project is developing a cataloguing system suitable for small, generally-skilled cultural communities. A metadata application profile (MAP) is a catalogue record of an on-line cataloguing system and is used to classify on-line distributed resources. It is the recommended name for a record based on the Dublin Core structure. Given the aim of the Matchbox Project, it is important that Matchbox can be customised suitably and the methodology for this customisation can be made explicit. This paper reports progress towards a MAP with which to describe Quinkan Culture. Specifically, it reports the first part of this process that aims to make the local catalogue interoperable, sharable with others, and anticipates the local cultural customisation.
Distributed resources that have been subject to on-line cataloguing include books, Web sites and pages, digital image collections, collections of framed paintings, and more. In the case of Quinkan Culture, as described below, there are many objects in many formats. Dublin Core records are catalogue records that cover the main classification criteria for such objects, such as the title, the author, the identity of the object and a description of it [HREF4]. These catalogue records are valued for the information they contain, which is often more useful in discovery activities than a search of the text in the resources. The resource does not always describe itself but rather lets its description emerge, which is not sufficient for someone searching for it.
Authoritative records, especially those made by cultural representatives such as Elders, or those authorised by them, currently exist in very low numbers. They are usually records that are remembered by Elders and transferred orally. There are, however, thousands of records of Quinkan Culture that have been made by researchers, tourists, and others. The Elders wish to preserve their culture and gain some control over its use by others who are not authorised or competent to talk about it. Matchbox was proposed as a tool that would assist in this process [HREF5].
Cultural artefacts and practices are quintessentially culture. It is important in the context of an Aboriginal culture that those who develop the technology do not prejudice in any way the use of that technology. The Quinkan Matchbox will not pre-set a record structure that cannot be altered or used in a variety of ways. Matchbox, as delivered, will not be a content collection, a set of records, but rather the structure into which such content can be placed by and for Quinkan Elders.
‘Quinkan Country’ is located in Tropical North Queensland, in the south eastern region of Cape York at the very north-eastern corner of Australia. It is in the basin formed by the Laura River and its tributaries. It is home to a number of family and language groups. Its lush savanna and high escarpments contain one of the world's largest collections of cave paintings, dating back as many as 36,000 years, in hundreds of scattered shelters and galleries.
The main community centre, Laura, is situated approximately 290 km to the north west of Cairns, in the centre of Quinkan Country on the main north-south and east-west road junction in Cape York. Today, Laura is a community of about 60 residents, including the descendants of the original inhabitants. It is renown for its bi-annual Aboriginal dance festivals. Currently, the Laura Community is working to realise the benefits of Federal Government funding for a substantial Quinkan Region and Resource Centre building.
European exploration of SE Cape York area began in the mid-1850s. By 1873, gold fields were established along the Palmer River, as were large cattle grazing properties. At that time, Indigenous people's contact with European and Chinese immigrants was often extremely violent, resulting in massacres and displacement of local Aboriginal groups from ancestral land to the fringes of mining camps and newly built towns (Maytown, Cooktown, etc). This violent disruption of traditional life meant the loss of most of many communities’ oral histories and interpretive links with their heritage artefacts. Some of the very few people who survived painted records of their clashes with settlers and immigrant officials, as can be seen at the Giant Horse site (George et al., 1995). Today, only a small proportion of the descendants of the original community still lives in Laura.
The Rock Art (cave paintings and petroglyphs) was "re-discovered" in the 1960s by bush pilot Percy J. Trezise (see photo) who started documenting paintings and recording local stories with his Aboriginal friend Dick Roughsey (Trezise, 1993).
Today, the local and remotely located Elders are very active in cultural heritage and conservation issues. They established the Ang-gnarra Aboriginal Corporation that has, at times, run Ranger programs designed to care for precious sites and guide visitors. Individual Elders and remotely-based tourist companies have also been active in guiding tourists.
If management implies a large degree of control, then this community is faced with a very arduous task. Many of the resources and mobile objects relating to Quinkan Culture (apart from the Rock Art) are dispersed in museums, libraries, universities, and government agencies in Australia and often overseas. There is also a proliferation of unauthorised publications and cultural representations of Quinkan culture in all sorts of media: site locations are published on the Internet using global positioning systems (GPS) technology, Rock Art motifs appear on tee-shirts and postcards. Very few of these instantiations are authorised by the Traditional Owners and the commercial ventures rarely benefit the traditional Indigenous cultural community in any way.
The Quinkan Matchbox project is taking place in a context of great enthusiasm for heritage promotion, Indigenous and otherwise, and equally strong anxiety regarding the production and management of Aboriginal cultural content for public consumption.
In 1999, a computer workshop was held at James Cook University in Cairns following a number of visits to Laura. Quinkan Elders and others accessed the Web and some local Web pages containing a recent video of an Elder talking about a honey tree. The viewing of this video prompted participants to tell other stories or their own version of a similar story. They also realised that a multimedia catalogue could act as a prompt for new stories and explanations and perhaps save the need to travel to far away story places (Nevile, 2003).
Following the workshop, a number of assumptions started to emerge with regard to the relationship between Indigenous culture(s) and the Internet. Perhaps the Internet could be harnessed as a prompt for cultural activity and assist communities in documenting their stories locally and privately, as well as incite new narratives. It may thus become a form of "back up" system to knowledge traditionally stored with people and also become the unlikely catalyst of renewed interaction between young people and their Elders. The same technology may also assist in the control of unauthorised publishing and management of authorised publications and contribute in some useful ways to Traditional Owners’ ability to convert, as they choose, their culture into their own, manageable, capital assets.
In the context of a small and remote community, with few financial resources and low connectivity, would a light-weight cataloguing system that is simple to operate and maintain, yet powerful enough to manage remote resources, be a sustainable option? And could Dublin Core based metadata play a role as the base standard for multimedia cataloguing? At a time when many communities are approaching cultural institutions to request the effective repatriation of objects and artefacts (or at least negotiate easy access and safe-keeping), the research team envisaged that a DC-based catalogue could be the vehicle for a more "virtual" form of repatriation. DC could provide the corner-stone on which the Quinkan Community could build its own distributed virtual cultural institution and manage its cultural resources both "out there" and locally.
The Quinkan Matchbox Project is a two-year ARC SPIRT Project originally established in partnership with the Ang-gnarra Aboriginal Corporation, James Cook University (JCU), and others. It is now operated in a partnership between JCU, La Trobe University and Motile Research Pty. Ltd., with the support of the Quinkan Indigenous Community The relevant research questions of the project are:
The (multi-disciplinary) research team includes four post-graduate students with a combined background in Computer Science, Software Engineering, Information Management and Indigenous Studies, each addressing a specific research question contributing to the overall aim and design of the project. The industry partners are engaged in the development of a metadata repository and content rendering system that will be given to the Quinkan Community when it is completed. This paper reports the work of the authors within this team.
The Matchbox project is not unique. Indigenous communities in Australia and around the world are building systems with a variety of names: virtual keeping place, interpetive centre, Indigenous knowledge centre, virtual camp fire, digital archives, cultural networks and more. They share the common goal to store and preserve, in a computer system, intellectual capital that has been previously held by people and to repatriate, somehow, resources (about the culture and community) held in museums, agencies, private collections, and other non-traditional ways.
Studies of the genesis of modern (anthropological) museums reveal that the prime motivation was as antidote against forgetfulness and destruction of living culture (Sculthorpe, 1985; MEN, 2002; Ames, 1992). Native American activist G. Horse (cited by Sculthorpe, 1985) reminds us that self-perpetuating cultures have no need for museums. For Horse, museums are an expression of guilt or even complicity of destruction of cultures operating in a different mode of transmission, such as oral tradition.
Most Indigenous communities in Australia have experienced some destruction of their traditional knowledge transmission system. Collectively, they have endured being objectified, classified and judged. While this experience of vulnerability may have prompted the desire to preserve culture using all means available, there is a strong desire expressed by communities to also exercise control over their heritage and its representation. For Maori scholar Linda Smith (2001),
"The struggle for the validity of indigenous knowledges may no longer be over the recognition that indigenous peoples have ways of viewing the world which are unique, but over proving the authenticity of, and control over, our own forms of knowledge."
Some authors have noted the congruence of Indigenous worldviews and multi-media systems and, in particular, the non-linear, hypertext nature of digital materials:
"The configuration of an on-line interactive text with its mapping patterns and hyper-linkages resembles the more native experience and logic patterns of individuals and groups who live within the communities. They understand the semantic similarities between these systems and those of the intersecting and coded tracks of the land." (Tafler, 2000)
But for others, the same medium is fraught with danger and is yet another potential threat to the traditional Indigenous information economy. Eric Michaels (1986) studied the introduction of broadcast television in Central Australia and its subsequent ‘re-invention’ by the Warlpiri. Michaels asserts that mass media always carry the risk of subverting the culture in which they are introduced: "Mass media are logically and practically the inverse of the personal Aboriginal information exchange system". While traditional knowledge systems are usually characterised as localised, personal and restricted, Michaels notes that broadcast systems make information equally accessible to audiences everywhere, instantaneously, and at no apparent cost. To paraphrase Mary Mortimer (2000), the Internet is about maximum access for a maximum number of people; Indigenous knowledge is about restricted access for authorised people.
Another danger attached to storage in a permanent medium is the fixity of meaning. Cultures live and stories and meaning are constantly evolving.
"Researchers who work with Indigenous communities for long periods also develop an understanding of the potential incongruities between efforts to produce publicly consumable information and the extent to which the vitality of tradition depends upon a degree of instability and constant readjustment of it in living practice." (Merlan, 2000)
In his analysis of Warlpiri Television, Michaels (1986) also noted the risks associated with capturing stories and ceremonies on a storage medium. For Michaels, any mediated storage or retrieval system is at risk of being construed as an external, archival, impersonal authority that may challenge the final authority of Elders or the Law. The "authoritative" if not "definitive" version of a story is held by Elders and may evolve as generations re-tell it and adapt it. The medium can only capture a version in a moment in time. To an extent, storing is like freezing culture and this runs against the principle of living, self-perpetuating cultures.
Assembling the MAP is only the one step in the overall Matchbox design. The literature review acquainted us with the various standards and practices of information description in the disciplines and communities intersecting with the Quinkan 'domain', such as museums, libraries and government agencies. Descriptions of culture used by these communities need to be integrated into the Matchbox design if there is to be interoperability between the Quinkan catalogue and those of such other communities. Many of these communities use Dublin Core based record profiles, and they were examined and used as the base for the proposed Quinkan records. This process is described in detail below. Once a base record structure was in place, what are called 'cross-walks' and testing with sample records followed to ensure that the Quinkan MAP retained semantic, as well as syntactic, inter-operability with these other Dublin Core records. Finally, an extensive and ongoing consultation with the representatives of the Quinkan Community has been sought and should assist us in further customising and refining the MAP to suit community needs and local specificities.
In many Indigenous knowledge projects, the focus is on organising resources for local use but little attention is paid to conformance to global standards. Frequently, the information architecture is based on defined collections of documents and what are seen as current Indigenous ‘worldviews’. This is often done with what seems like little regard for future possibilities of exchange. In the case of the Matchbox Project, and without pre-empting the willingness of the community to share information outside the group, we felt we could adopt Dublin Core (DC) architecture (and perhaps other standards) as a base then customise it to suit local Quinkan needs without compromising future inter-operability.
DC as an information description standard, for information about resources, is outstanding for its modularity and flexibility. Essentially, it consists of fifteen categories that have been selected in an extensive process of consensus between representatives of many cultural institutions and intellectual communities. Additional elements have been added to the list and qualifications of the recommended elements are supported. DC classifications are encoded in standard formats so that, in theory, other elements from different schemas can be combined with DC elements. Thus we should be able to mix and match elements from various application profiles, adopt (or adapt) various schemes and qualifiers from non-DC application profiles (treating them as extensions to DC elements) and still retain DC-compliance and inter-operability.
We were interested in taking the original fifteen or so elements, making them more specific in some cases. This is a DC compliant process where, for instance, an element such as dc:author is made more specific by being sub-divided into dc:author:namePersonal where the expected value is in the format 'lastname,firstname'. All DC elements can nowadays be qualified in three ways: the element made more specific (as shown already), the values limited to a particular format or encoding, or the values selected from a controlled vocabulary. DC used this way is known as 'qualified Dublin Core'. Interpretation of the qualifiers is made possible by the provision of information that describes the qualification, using the original (simple DC) element qualifier that allows for the specification of the schema being used to interpret the qualified element.
We had, of course, analysed the problem from our traditional western cultural point of view, first, working from what we knew about the documents, information needs of users and variety of domains: archaeology, museum, environmental protection, etc.. We started thinking from a Rock Art centric position, basing our ideas on classification of paintings, artefacts. We assumed that the adoption of Quinkan specific qualifiers and schemes would allow us to retain, internally, the granularity required for sample resources supplied by subject specialists.
When we examined the work of larger information communities, we realised that they also tested the DC offering and made local adjustments. For instance, in 1998, the Consortium for Computer Interchange of Museum Information (now CIMI Consortium) set up a catalogue test-bed to see how well DC elements described museum information. While the overall results were encouraging, CIMI made a number of propositions pertaining to qualifiers and also proposed a set of best practices and cataloguing rules [HREF7]. Significantly, they decided that simple Dublin Core, as it was at the time, was inadequate for all their purposes and they started to add their own elements to a museum-specific CIMI profile.
"Above all else, we’ve learned that Dublin Core, in its "simple" manifestation, is necessary for cross-domain searching. We recognize that there is a loss of richness in a DC Simple record. Without qualification, the necessary contextualization of values within an element, cannot occur. However, for the purposes of interoperability, this loss of richness is acceptable." [HREF8]
CIMI then tested qualified DC and decided that additional element refinements were less useful than controlled vocabularies and other value encoding schemes. They felt element refinements were a threat to inter-operability [HREF8]:
"A useful addition to the DCMES is the idea of Value Encoding Schemes. CIMI endorses the idea of Value Encoding Schemes as useful for both interoperability and resource discovery.
If the DCMES cannot be extended semantically, then a domain specific solution must be sought to address the issues related to interchange of high level, complex museum information. CIMI will undertake this initiative in its program of work through 2001."
It was important to us to review as many profiles as may be relevant in the future, as we are looking forward to sharing cataloguing records. Our current Quinkan MAP includes ideas drawn from the following research.
"The CDWA categories describe the content of arts databases by articulating a conceptual framework for describing and accessing information about art objects, architecture, collections and images. These categories map or are the basis for other metatdata standards for describing works of art and material culture, such as the Art Museum Image Consortium (AMICO) [HREF10], The International Committee for Documentation of the International Council of Museums (ICOM-CIDOC) [HREF11], CIMI Consortium and others." (Baca and Harpring, 2000)
"The Core Data Index to Historic Buildings and Monuments of the Architectural Heritage (1992) was created to identify the categories of information necessary to record buildings of historic and architectural interest, and the International Core Data Standard for Archaeological Sites and Monuments (1995) to identify the categories necessary for documenting the immovable archaeological heritage. Object ID (1997) was developed to provide an international standard for the information needed to identify cultural objects, in response to the threat posed by the illicit trade in the movable heritage." (Thornes and Bold, 1998)
Mixing and matching different ways of describing a given object was always going to be a challenge, especially while retaining compliance to DC. In this difficult exercise, the use of the ‘cross-walk’ method has been invaluable.
This method assumes that each domain has its own data structure, field names and meanings suiting its local purposes but can they be related back to a base standard through mapping. Cross-walks are not about enforcing a single standard but working out rules for transformation (or translation) from one to another. They have been used extensively to deal with legacy information systems.
"Until recently, these legacy systems existed in isolation, serving only specialized knowledge communities. Each system maintained its own structural design for fields of access, description and vocabulary control" (Woodley, 2000).
This is sometimes referred to as metadata mapping or semantic mapping. Woodley describes cross-walks as the visual representations or maps that show the relationships between elements of disparate databases or application profiles. For us, cross-walk mapping supports (or helps visualise) semantic inter-operability.
Woodley warns that:
"Inherently, there are limitations to cross-walks; there is rarely a one-to-one correspondence between the fields or data elements in different information systems".
The problem with cross-walks is the obvious logical problem that arises when there is not a clear one-to-one mapping from one element to another. This multiplicity is just as problematic for cross-walks whichever way it is happening. Wherever strict semantic mapping could not be done within the Matchbox Project, a temporary solution has been to treat the "misfit" elements either as Quinkan-specific qualifiers or as encoding schemes for DC elements. The justification is that they will be retained and contribute to improved granularity, precision and semantic understanding in the Quinkan context.
The final Quinkan MAP should be DC compliant at a "coarse level" and retain sufficient granularity to describe the resources in a meaningful way to any user performing a search or browse in the Quinkan catalogue. (To achieve this, RDF technology will hopefully be deployed in the final version.)
"Mapping the structural components of metadata systems is only one part of the picture. Cross-walks of search commands and controlled terms or thesauri will further enhance searchers' ability to retrieve the most precise search results."(Woodley, 2000)
Further work will focus on taxonomies for both content management and discovery.
Our original analysis of the problem was that the Quinkan Matchbox was going to be mostly a catalogue of Rock Art paintings, as objects. Our most developed field of ‘subject specialisation’ is derived from extensive consultations with an archaeologist who has specialised in Quinkan Rock Art. For research purposes, archaeologists often devise specialised classification schemas and taxonomies of Rock Art. Some of these are ‘universal’ and suitable for use across the discipline, others are devised by a single individual for the purpose of answering a single research question. Site structures and site materials give crucial information regarding Rock Art site use to anthropologists and motifs (their orientation, frequency and types) are also of importance in Rock Art research. Motifs can be organised according to colour schemes and production techniques.
We also considered the definition of information held in the Queensland Department of Environmental Protection (EPA) site index forms, as reproduced in the 1995 issue of Tempus (Morwood and Hobbs, 1995). These forms were used by the agency to describe large natural sites and assess their condition. They describe sites according to their general environment and use option lists to describe parent rock, soil quality, vegetation, presence of and distance to water sources. They also keep a history of the assessment work and information regarding the legal and protection status of the site. Each index form comes with a number of attachment cards, each describing the site with greater precision according to its type (ie: art site, scarred tree, burial, etc…). For example, the information contained on the Art Site card is mainly concerned with dimension, damage and exposure.
Our current MAP replicates some of the categories found on the Site Index form. Our art census categories (list of motifs) have been supplied by our consulting archaeologist (Cole, 1995). Further consultation with the Queensland Department’s representatives is expected to lead to the inclusion of more site-specific categories. This part of the MAP is not fixed: it has been designed so it can be further developed, should the community decide to keep more detailed information, for example about weathering and depredation of large sites or specific galleries or motifs.
We decided that the best way to integrate the EPA structured description was to treat these lists of terms as extensions of DC.Description. They can be regarded as either a way to structure DC.Description.Abstract through the use of a Quinkan-specific encoding scheme and its value lists or as another qualifier for this element, namely DC:Description:QuinkanObject.
Each of the lists supplied by subject specialists is a value list. It contains the vocabularies and structures familiar to those specialists and provides good predictive information about how they will classify and search for resources. We hope to derive more of these lists of terms as our consultation process extends to other academic fields. The Quinkan MAP is still a work in progress.
By the end of 2002, a draft MAP had been developed and prepared for testing. Subject-specific experts have been asked to try it with some of their resources and to comment on how useful it is to their purpose. It is anticipated that as a result of this work, the final MAP will be a refinement of the current draft. It is known to be too long, too detailed, and possibly not biased sufficiently towards what users will want.
Work must still be done on designing one (or more) taxonomies for discovery and content management. This work is being done in consultation with representatives of the Quinkan Indigenous Community during visits in the Community. This field work will also involve working out more collaboration protocols and customising the Catalogue system to suit local requirements. We want to ensure that our MAP and its choices of taxonomies, vocabularies and values truly support the Quinkan Community’s needs.
We have chosen to build a prototype of the system in order to provide the community with a concrete ‘sounding board’. Cataloguing systems are very abstract and difficult to conceptualise. We hope the prototype will reassure the Community that the proposed technology solution is ‘soft’ enough to show how it can be modified quickly and on demand. We also trust that this is a valid method to gather vital requirements for the Community from the Community.
In fact, as proposed in the original research application, working with HTML and XML technologies does not provide the kinds of freedoms that already have been identified as likely to be necessary. Using RDF to build a flexible ontology is much more likely to yield behaviours in the system that will be able to respond to the needs of all the overlapping communities interested in Quinkan culture. (This aspect is not relevant here but described elsewhere (Nevile et al, 2003).
Linda Smith (2000) advances that "Networking has become an efficient medium for stimulating information flows, educating people quickly about issues and creating extensive international talking circles. Building networks is about building knowledge databases which are based on the principles of relationships and connections."
When reflecting on the introduction of new forms of communications in Indigenous Communities, Michaels (1986) stressed that "the literature in communication and media development for indigenous people recommends that the introduction of new communications services take into account traditional forms". Michaels went on to explain that Warlpiri maintained the strict kinship system as the base for the organisation of every aspect of their production chain, from video production to video tape storage.
For the Matchbox Project, understanding traditional forms may lead to a better understanding of the information needs of the Quinkan Community and may help us by suggesting appropriate solutions. Understanding traditional Quinkan forms may help us build an information architecture that respects (and replicates) the local information economy.
In July 1998, Buchtman (2000) returned to Warlpiri lands to report on how the communities had adopted modern communication technologies and see if there had been social changes as a result. She noted that "the Warlpiri placed the new technologies into existing cultural practices" in order to "re-establish some of their traditional communications or songlines and to establish a link with non-Aboriginal Australia." She concluded that "the use of modern media could undermine the social structure of Warlpiri society yet there is strong evidence the Elders ultimately still control the broadcasting through the Warlpiri Media Association even though younger adults broadcast".
The notion of protocols reaches farther than gathering technical requirements through whatever method. It reaches into a deeper and more personal level. Protocols are also about understanding and building trust. For this, there is no technique, just time and personal qualities. It is our hope that we can develop a system that is adaptable by the Quinkan Community for their benefit.
Ames, M. (1992). Cannibal tours and glass boxes: the anthropology of museums, 2nd ed. Vancouver: UBC Press.
Buchtman, L. (2000). "Digital songlines: the use of modern communication technology by an Aboriginal community in remote Australia". Prometheus (Vol. 18, No. 1), pp 59-74.
Baca, M. and Harpring, P. (2000) Categories for the Description of Works of Art. California: The J. Paul Getty Trust. Online: [HREF14]
Cole, N.(1995). "Rock Art in the Laura-Cooktown Region, S.E. Cape York Peninsula", Tempus Vol. 3. St. Lucia, Qld: Anthropology Museum, University of Queensland.
George, T. and Musgrave, G. (1995). Our country, our art, our Quinkans. Laura: Ang-gnarra Aboriginal Corporation.
MEN (2002). The cannibal museum (Le Musee Cannibale). Musee d’Ethnographie (Neuchatel, Switzerland). Online: [HREF15]
Merlan, F. (2000). "Representing the rainbow: Aboriginal culture in an interconnected world" in Australian Aboriginal Studies (issue 1& 2), pp. 20-26..
Michaels, E. (1986). The Aboriginal Invention of Television in Central Australia 1982-1986. Canberra: Australian Institute of Aboriginal and Torres Straits Islanders Studies
Mortimer, M. (2000). "Australian Indigenous communities online". In Information literacy around the world: advances in programs and research, edited by Bruce, C. and Candy, P. pp. 99-107. Wagga Wagga (NSW): Charles Sturt University. Centre for Information Studies.
Morwood, M. J. and Hobbs, D. R. (ed.). "Quinkan prehistory: the archaeology of Aboriginal art in S.E. Cape York Peninsula, Australia" in Tempus (vol. 3, 1995). St. Lucia, Qld : Anthropology Museum, University of Queensland.
Nevile, L. and Lissonnet, S., (2003). "Dublin Core: the base for an Indigenous culture environment?" in Museums and the Web 2003, Ed. D. Bearman and J Trant. Toronto, Ontario: Museum and Informatics. Online: [HREF16]
Object-ID: international standard for describing art, antiques and antiquities (1999). Council for the Prevention of Art Theft. Online: [HREF17]
Sculthorpe, G. (1985). "Aborigines and museums: the issues from an Aboriginal perspective". In Papers presented at the ANZAAS Festival of Science (August 1985), pp. 21-31. Clayton, Vic.: Monash University. Aboriginal Research Centre
Smith, L. T. (2001) Decolonizing methodologies: research and Indigenous peoples. Dunedin: University of Otago Press.
Tafler, D. (2000) "The use of electronic media in remote communities" in Australian Aboriginal Studies (Spring-Autumn 2000), p27.
Thornes, R. and Bold, J. (1998). Documenting the cultural heritage. Getty Information Institute. Online: [HREF18]
Trezise, P. J. (1993) Dream Road: a journey of discovery (2nd ed.). St. Leonards, N.S.W.: Allen and Unwin.
Woodley, M (2000). "Crosswalks: the path to universal access?" in Introduction to Metadata: pathways to digital information. Online: [HREF19]