The Next Generation — 'Knowledge Environments' and Digital Collections

Fiona Cameron, University of Sydney, Australia

www.usyd.edu.au

Abstract

Despite recent technical advances in collections access and interpretation, most notably on the World Wide Web, a number of key issues still remain. A sustained consideration, beyond technological innovation, to pragmatic concerns arising from current acquisition and documentation practices and data quality is necessary. Coupled with this, a consolidated response is required to recent post-structuralist, postmodern epistemological, disciplinary and museological debates about knowledge making and meaning construction and how current collections information relates to, and revises these themes.

Keywords: collections databases, knowledge models, user research, postmodernism, post-structuralism, documentation, interpretation

Preface

Paul Marty (1999) in a recent paper argued that information infrastructures can be viewed as organic, evolving with society or the organization they support, defining it as much as they are defined by it. Transposing this argument more specifically to a collections context, and in the framework of the aforementioned issues, one may ask the following questions. What are the potential futures for the next generation of digital collections? How will documentation systems and practices evolve to meet the needs of contemporary knowledge paradigms and users? In what ways will they transform institutions and the way objects are understood and accessed?

This paper poses some answers to these questions drawing on the results of two research projects. First, a discussion of findings from the Themescaping Virtual Collections project (an Australian Research Council funded project with partners the University of Sydney History Department, the Powerhouse Museum and Vernon Systems Ltd.) elucidates some potential futures for online collections interfaces, navigational, searching, browsing concepts and content while profiling and detailing the requirements of an emerging and diverse community of users.

Second, preliminary results from the Knowledge Objects project (an Australian Research Council funded Sesqui grant with partners the University of Sydney History Department and the Powerhouse Museum) offer a sustained consideration, beyond technological innovation, to pragmatic concerns arising from current acquisition and documentation practices and data quality. But more significantly, the discussion offers fresh insights into ways collections documentation and interpretation can be reconceptualized to form new knowledge models in line with contemporary theoretical, pedagogic and public access concerns.

The benefits of this discussion extend beyond how the substantial resources allocated to tasks such as digitization and technical development may reap maximum returns for museums and users. Rethinking the nature of collections information and public access offers new insights into key areas for change in museum policy, practice, knowledge creation and management.

Introduction

With the challenge of digital technologies (most notably the WWW), new knowledge paradigms and public access concerns, it has become necessary to revise museum collections databases, the current epistemological status of data and established documentation practices. The scale and speed at which many museums are responding to some of these new challenges is remarkable. In institutions traditionally viewed as retrospective, where strict conventions rooted in the empirical tradition govern collections management and documentation, opportunities offered by digital technologies radically reform what has been previously possible in terms of knowledge environments.

Most significantly, over the last decade responses to these challenges have led to considerable technological advances in the field of collections access in the context of the World Wide Web as well as investments by museums and providers in collections automation systems. Here research has focused on the development of sophisticated browsing and searching mechanisms, architecture and descriptive concepts such as metadata for retrieval purposes.

The first generation of online collections is typified by a stories/thematic solution to narrative formation as well as searching interfaces. More recently there is a trend towards tailored but also more adaptive responses to different usage situations as well as rethinking of collections data, and how it can be put to more productive interpretive ends. This offers alternative pathways through collections information while posing greater contextual possibilities around objects through the inclusion of additional multimedia and text based information. Above all, it empowers the user to create pathways through and new organizations of information thereby contributing to the development of the knowledge environment. Examples include the EMP Digital Collections (Experience Music Project, http://www.emplive.com), Hypermuseum (http://www.HyperMuseum.com), and History Wired (Smithsonian National Museum of American History, http://historywired.si.edu/index.html) (Gillard, 2002).

In spite of these innovative advances, two issues remain. The first, originally raised by Sarasan & Donovan (1998), is whether collections databases and exported data presented in the public realm really fulfill the information and pedagogic needs of an emerging community of online users. The second by Thomas & Mintz (1998;2) concerns the problem of data quality. That is, whether the data stored in these systems, and the way objects are interpreted by museum staff originally for the purposes of documentation and inventory control, really engage the needs of online users.

In this paper I address these enduring dilemmas in two parts drawing on the results of sustained research. The first part critically examines the current and future implications for public access to online collections databases and exported data. Selected findings from the Themescaping Virtual Collections project (an Australian Research Council funded project with partners the University of Sydney History Department, the Powerhouse Museum and Vernon Systems Ltd.) initially address the requirements of an emerging community of users. This entails the profiling of user groups, detailing their motivations and matching these to current and future access, retrieval and content requirements while revealing the diversity and breadth of potential uses for these resources. Complementing this I will also consider future concept and content options drawing on emerging technical developments and the potential applications for new knowledge paradigms in the creation of innovative knowledge environments.

The second part utilizes some preliminary research findings from the Knowledge Objects project (an Australian Research Council funded Sesqui grant with partners the University of Sydney History Department and the Powerhouse Museum) to address the important problem of documentation, the very foundations that underpin collections databases and hence public access. Here I take the argument beyond the immediate concerns of Thomas and Mintz (1998), and issues of data conversion and technical capability, to reconceptualize traditional museum models and practices for collections documentation according to current theoretical paradigms and the capabilities digital technologies have to offer. Most significantly, I offer some insights into how this transformative process may occur, the nature and form of new knowledge models, how these relate to user needs outlined in the aforementioned project, and the way they may be applied to collections documentation and information capture. To conclude, I discuss the broader implications for future museum practice. The benefits of this discussion extend beyond how the substantial resources put to tasks such as digitization and technical development may reap maximum returns for museums and users. Rethinking the nature of collections information also provides new insights into key areas for change in museum policy, practice, knowledge creation and management.

Tangible futures - user profiles, access, retrieval and content requirements

Now revisiting the first consideration. How might documentation systems and practices evolve to meet the needs of contemporary discourses and users? Here the Themescaping Virtual Collections project offers some insightful suggestions illuminated through a program of sustained research. To this end, research methods included interviews and focus groups with a range of users, Indigenous museum workers and IT specialists in Sydney and New Zealand, as well as the evaluation of existing online collections. Contingent to this was an analysis of the current state of user research, state of the art portals and online databases, search functionality, browsing, searching, narrative and personalization tools, metadata and information architecture.

Our research revealed four broad user groups. These were curators, collection managers, educators and non-specialists. Yet none of these profiles is mutually exclusive. Professional users for example may search or browse collections for free-choice learning. So how do user profiles and motivations match up with available and potential searching, browsing, navigational concepts and content?

With these questions in mind we sought to investigate the motivations and needs of the two main user groups, curators and collection managers. Interestingly, these two groups displayed similar but also differing motivations.

The curatorial user profile

Curatorial tasks for example include collections documentation and exhibition development. Therefore, curators' motivation for searching and browsing their own and other online collections is to identify objects by comparison with like examples, develop exhibition concepts and publications, plan visits to other museums, research potential acquisitions and to identify objects for loan. Furthermore, databases and exported data were viewed by many focus group participants as having the potential to enhance their own research capabilities, enabling choices to be made on the system thereby cutting down on the physical handling of collections. Such opportunities were seen as enabling democratized access to collections and efficient ways of answering collection inquiries. The more 'enlightened' curators among them were also interested in extending their interpretive functions through online collections by the experimentation in new ways of configuring and extending the value of collections information.

Their primary concerns were improving research capabilities through more intelligent searching and browsing mechanisms, and in devising combinations to mine data, as this profile required quick and reliable access to objects, information and images for collection identification, research and exhibition tasks. To this end, the preferred option for specific searches was keywords with the use of thesauri (by subject, function and classification) with mechanisms that suggest search terms, phonetic spellings and make typological corrections. Complementing this, our research revealed a need for the development of a range of logical search schemas. Suggested options included highly structured searching, intuitive browsing mechanisms and navigational tools such as well-indexed subject searches and personalization/theming search engines connected to a theme generator function like the Hypermuseum model (http://www.HyperMuseum.com) (Stuer, 2001). Furthermore, future possibilities may involve the implementation of personalization mechanisms popular in e-commerce environments such as Amazon. As one focus group participant remarked,

'I like the way Amazon.com associates things such as suggested readings based on a similar subject/author and so forthÖif you search for a book that interests you it will find other suggested readingsÖit would be good to have a clear personalization engine working within the collections environment.' (Cameron & Kenderdine, 2001; 28)

More specifically, in a collections context, this personalization concept has the potential to generate relationships between objects/topics while searching.

Other desirable features included advanced searching across databases and the comparison of related search results using the indexing of metadata or an existing metadata index. Mechanisms to elicit responses to search results and collections information from users were also high on the list of priorities.

Preferred browsing options on the other hand were specifically for supporting searching and for answering collection enquiries. Potential future options include question and answer search engines, predetermined thematic tours or collection descriptions showing highlights of the collections, followed up by an advanced search using keywords and a browser to further mine data. Collection statements connected to a hyperbolic tree or site map were also seen as a way of specifying its scope.

Most significantly, providing information in various forms to suit the conceptual maps of a range of users was seen by the curatorial profile as vitally important. These include standard schemas such as timeline/chronologies and information hierarchies by theme and sub-theme, as well as more innovative mindmaps as with the Revealing Things (http://www.si.edu/revealingthings/) and EMP Digital Collections models (http://www.emplive.com) (Adolsek 2001). Additionally the repackaging of information to suit the specific needs of each broad user group profile defined as educators, curators/researchers and general browsers (similar to the Museum of Rural Life model) (http://www.ruralhistory.org/index.html) allows for the user to chose and tailor their experience.

Enriching collection information using both narrative and object-centered histories, according to our research on the curatorial profile, has the potential to extend and enhance collections information for a range of conceptual needs. Exploiting the associative and hyperlinking abilities of the technology in the creation of object-centered histories allows for the layering and exploration of the multiple meanings and contexts of selected objects. Likewise, this method can present and deconstruct the documentation/interpretation process as a post-structuralist, post-modernist text. Furthermore, it invites the user to explore collections in more depth while allowing self-guided interpretations.

Suggested content included the juxtaposition of subject, disciplinary/cultural perspectives, contexts, information on the use and manufacture of objects and visualizations showing their creation histories. The use of first person voices and quotes, significance statements, curatorial/expert opinions, artists' comments as well as primary source material with supporting documents, images, audio-visual, sound bytes and bibliographic information were seen as ways to expand out the meanings surrounding objects. Allowing the user to put up their own interpretations of this material, according to our research, also supports a constructivist approach to learning. For others, however, it was read as a threat to curatorial authority. These concerns were expressed by one participant and confirmed by others. That is,

'A curator's authority is threatened when the user can teach and discover thematic relationships for themselves. There is an issue if audiences are invited to construct something without any professional guidance and especially if it goes out into the public domain so we need to be careful about the information they can pull out i.e. privacy and information on people.' (Cameron & Kenderdine, 2001; 29)

Most importantly, exposing the nature of museum significance and disciplinary/museum assessment procedures through links to policies and curatorial essays contributes to an understanding by the user of how meaning and value is formulated within the museum context.

Compelling narrative-centered thematic approaches not only allow a means for highlighting the strengths of the collection but, according to this profile, provide a less resource-intensive form of conceptualization. These in tandem with selected zoomable thumbnail images, 3D objects, object movies and prompts to highlight significant points, offer a greater understanding of an object's form through multiple views and its function and use through movement and sound. Acknowledging sources and the authority and the subject position of collections statements through authoring were seen as vitally important in contributing to the pedagogic usefulness of information. So too were links to current exhibitions related to search results, and theming mechanisms that trigger options to encourage people to go further with their research or visit the institution.

Registrars/Collection Managers user profile

By contrast, registrars/collection managers are mainly preoccupied with the tasks of acquisition, inventory control and collections documentation. To this end, this group uses databases and online collections on a daily basis to locate, identify, research and date objects.

Here curators and collections managers can agree on one thing — a need to increase the functionality of searching tools. Options include solutions outlined with the curator's profile. And for those in art museums additional choices comprise searches by date, collection types, media and artist and an alphabetical search by artist/subject. Given that reliable searching is a key need, devices that enhance this process necessarily require the development of tools that both inform and direct. Here informing the searching process can be achieved through the exposition of the scope and breadth of the collection via collection overviews and site maps, and established practices such as thematic highlights with additional and newer searching and browsing mechanisms. To complement this, solutions to direct searching may include graphical representations of browsing pathways and multiple entries to collections information. Furthermore, the ability to store search results via a notebook was also deemed desirable.

Interestingly, content needs specially relate to improving search functionality and hence identification and documentation tasks. And as such, desired content includes zoomable thumbnail images, links to sources, a glossary function, collection percentage statements, copyright information, statements about the exhibition status of objects and language translation facilities.

Teachers and Museum Educators user profile

Teachers and museum educators are a potentially significant user group. Yet, despite these latent possibilities, collections resources are under-utilized due to problems with computer access, bandwidth and download time. But most significantly, and in many but not all instances, a lack of enriched data precludes the use of such a resource by the profile. Marketing the availability of online collections to this group is also a key issue. Once many of these constraints are surmounted, online collections have the potential to support lesson preparation, learning and student projects about material culture, and to complement school visits.

Ideally educators favored a range of searching/browsing options from tailored thematic trails and intelligent keyword searches to customizable personalization search engines. Here the structuring and navigation of information using traditional hierarchical metaphors similar to a book paradigm and timelines/chronologies as well as mindmaps for visually connecting objects, ideas, metaphors and related themes were deemed to support a range of learning options. Likewise, graphically driven interfaces to search, browse and contextualize collections, while offering an engaging, interactive experience, serve as a departure from catalogued information.

Offering a range of interpretive opportunities and information organizational concepts for students was, according to educators, vitally important to support project work. That is, object-centered histories with links to sources and bibliographic information as outlined previously, with additional significance hierarchies and narrative-centered themes. Other suggestions included the layering of information by complexity to serve a range of information needs, and the creation of action spaces through quizzes and mechanisms to generate suggested thematic trails.

'The more realistic the better' was the feedback we received from educators. Scaled 3D simulations of objects, utilizing multiple views and object movies where students can manipulate digital objects according to their intended use, have the potential to fulfill learning objectives by allowing students to explore the form, construction, texture and use of objects using various senses. Similarly, 3D environments and browsing tools as seen with games technology, have the potential to act as a conduit in the creation of a cultural ambience around collections and to place objects in spatial relationships while creating a narrative format. A strong pedagogic case was put forward by one participant who stated that 'this technique is particularly important for younger children as they need to visually explore the connections between objects, contexts and stories' (Cameron & Kenderdine, 2001;30).

Non-specialist user profiles

Lynn Dierking and John Falk (1998) convincingly argued that online visitors to museum websites follow the psychological and demographic profiles of bricks-and mortar museum visitors. More recently the research of Kravchyna and Hastings (2002) on the informational value of museum websites supports and expands on Dierking and Falk's contention by suggesting that a large proportion of users access online collections and museum websites pre and post visiting the physical site. Yet, a lot more detailed information is needed about this group particularly in a specialized collections context. Despite this, our investigation supports these findings. That is, potential users of online collections are generally museum visitors. Most importantly however, this profile values free choice learning. Additionally, users would potentially access online collections to explore and discover new things and build on their knowledge base, as a form of entertainment if enriched information was available, and to plan visits to museums. Significantly, the belief in the inherent integrity of information a museum web site has to offer was also a motivating factor in accessing online collections.

Due to the diversity of the profile, a range of experiences from simple searching and browsing tools, and content for the inexperienced, to sophisticated challenging tools for the initiated, delivering complex rich media to support a range of learning and entertaining needs was deemed essential. Here, the delivery of these experiences as well as personalization could potentially be achieved through user/task profiles and the future use of searching agents to define profile, mood and required experience.

Younger members of the group preferred browsing options above specific searches. More specifically, their interest was in semantic structures of 3D space as an exploratory medium, whereby concepts such as personalization, theme generator tools and spatial associative concepts such as mindmaps with magazine style interfaces, visual prompts and interactivity were preferred methods to explore and convey information. Here, a younger member of the group put forward a persuasive argument for the use of 3D semantic structures as a tool to enhance the interpretive and communication potential of collections where 'you see relationships and the way they change and interact using the metaphor of 3D spaceÖonce you use the metaphor of 3D spaces the visualization could be 2D —doesn't matter you have the semantic there' (Cameron & Kenderdine, 2001; 30).

Surprisingly, this group was less interested in prescribed material, choosing to drive their own pathways through collections and to explore object-centered narratives with rich streaming media, 3D objects and visual environments as both a content and delivery tool. By contrast, older participants preferred traditional exhibition metaphors for structuring information and familiar classifications such as collection overviews with keyword searches, thematic structures, chronologies/timelines, linear browsing pathways and searches under known categories. As one contributor pointed out, 'these techniques are so simple they are like eating sponge, but there is nothing wrong with eating sponge' (Themescapes Focus Group, Non-Specialists, 2001; 5).

We now have some valuable data on the needs and preferences of diverse users and requirements for Indigenous collections. So how can content needs be realized through the acquisition and documentation process for the next generation of digital collections and knowledge environments? What new knowledge models are required, and what are the implications for existing knowledge creation and documentation practices in museums?

New Knowledge models — the transformative process

Arguably, the transformation of databases from documentation tools to effective and sustainable 'knowledge environments' starts at the documentation level. Collections management databases are the primary means in which museums document and interpret their collections. And more importantly, these tools form the starting point by which museums define and communicate the significance and heritage value of collections (Robinson & Cameron 2003:3). Prominent museological scholar Gaynor Kavangah (1990;10) acknowledges that it is at the individual object records that conventional and totalizing practices take root. Yet despite this, there needs to be a sustained consideration, beyond technological innovation, to pragmatic concerns arising from current acquisition and documentation practices and data quality. Coupled with this, a consolidated response is required to recent postmodern epistemological, disciplinary and museological debates about knowledge making and meaning construction and how current collections information relates to, and revisions these themes. At this juncture, an opportunity exists to embrace the opportunity for change (Robinson & Cameron 2003:3). That is, to reconceptualize new knowledge models and streamline the content of collections management databases in line with contemporary theoretical, pedagogic and public access concerns (Ibid). Drawing on some preliminary observations from the Knowledge Objects project, the following considerations have been enumerated as critical to the transformative process.

• Types of information required for in-house use and for diverse and targeted user groups and how these can be streamlined into the acquisition and documentation processes
• Contemporary knowledge making discourses and the possibilities they offer in meaning making around collections
• Current trends in learning theory
• New and emerging technologies - the potential for extending the interpretative potential of collections
• Ways existing data and investments in the digitization of materials can be reused and configured into new relationships
• Critical analysis of existing acquisition, interpretive and documentation/classification paradigms, practices and procedures- the way they can be adapted, modified and streamlined to meet new information needs
• Examination of the broader information resources in museums and how they may contribute to collections data
• Reconciling disciplinary documentation and interpretive practices
• Consideration of potential returns ie the promotion of museums and their collections increased accessibility and visitation to the physical site, revenue generation

Resource issues.

Most significantly, all these considerations also have implications for documentation procedures. That is, the types and ways information is gathered and documented by curators and collection managers. Leading on from this are more practical concerns about how meaning making is created around collections and the resultant implications for staff roles and tasks.

Maximizing existing and enriched data

At the most basic level, in the creation of new 'knowledge environments', there needs to be a consideration of how information such as fielded data can be structured and coded in new ways. Equally important is the manner in which museums can maximize the use of enriched data through cross-linking and filtering to different collection records. This represents a strategy to improving the interpretive potential of collections.

Beyond description

Recent documentation practices have witnessed a trend away from long descriptions to significance statements. Interestingly, these foreground a shift in documentation from the predominately empiricist tradition based on the physicality of objects. Originally written to justify acquisitions to management, there needs to be a further consideration of how these statements can be written in a compelling way for audiences. Rather, curators need to consider the writing of text in the context of constructivist approaches to learning.

But more importantly how can the information and practical needs of curators, users, and the theoretical imperatives brought about by post-structuralist, postmodern and museological debates be feasibly tethered to documentation practices, and in the creation of new knowledge models and environments?

The impact of post-structuralist and postmodern paradigms

First, the impact of post-structuralist and postmodern paradigms in knowledge creation and understanding, as seen through our research and trends in scholarly thinking, fundamentally challenges and undermines traditional concepts about the truth-value of empirically based forms of knowledge. In this discursive context, material collections are no longer given privileged standing. That is, where a definitive meaning of the past is deemed to lie dormant in material objects exposed through empirical forms of observation, description and measurement and where one narrative, an interpretation of an object usurps all others. According to the empiricist tradition, documentation as a process is viewed as the collation of self-evident data rather than a subjective form of interpretation on the part of the curator as might be argued from a post- modernist's position (Robinson & Cameron 2003:20). Current disciplinary classification, descriptions, anonymous statements of significance and provenance have their genesis in the empiricist tradition. Additionally, in documentation there is an underlying view that object descriptive data, the 'cold hard facts' such as measurements and visual form, and meaning created through subsequent interpretive text, are two distinct fields (Robinson & Cameron 2003:23). That is, the former is an objective act, and the latter, subjective in nature. Instead, these practices invite a re-examination and the exposure of the discursive nature of museum objects as 'factual' statements about events and people, the nature of descriptive and interpretive data as intrinsically subjective and biased, and the ways they are used to substantiate knowledge about the past in collections documentation.

Incorporating theoretical principles in practices and procedures

So how do we get collections documentation up to speed with current knowledge making? Here museums over recent years have been confronted with a difficult challenge. That is, to provide a conceptual structure and order to disparate collections at a documentation level, to acknowledge and utilize previous investments but to also engage a growing community of users. But more importantly how can postmodern and post-structuralist discursive principles about the role of objects as polysemic, holding plural, cross-disciplinary meanings, alternative and sometimes conflicting, be incorporated into documentation? Also how can the meaning of narratives and classification systems be acknowledged as products of cultural, disciplinary, museum and curatorial opinion? Furthermore, in a postmodern context where disciplinary privilege and hierarchies disappear, how can interpretive authority be shared with users?

Addressing the concreteness of collections

The first task is to address issues around the concreteness of objects and their interpretations. Interestingly our research suggests that collections items and information must also be considered as a growing and evolving body of knowledge rather than a definitive or quintessential set of facts. A science and technology curator at the Powerhouse Museum noted this point and stated that 'there is the realization that while writing, a curator is creating an ëartifact' of that time, place and social context' (Robinson & Cameron, 2003;26). So here we need to present collections and information as temporarily situated expert opinions through authoring and dating.

Additionally we also need to instill a reflexive consciousness among curators of the limited legitimacy and lifespan of collections information (Robinson & Cameron, 2003; 23). To this end, collections information needs to be revised from time to time within available resources.

Exposing interpretive frameworks

The second task is to expose the epistemological/disciplinary frameworks in which objects are interpreted, to explain the fact that object interpretations change and as such should no longer be presented as definitive accounts carrying the ultimate authority. For example these postmodern principles could be shown through disclaimers, authored text and linked curatorial essays about disciplinary contexts outlining the types of information privileged in each domain and how each contributes to our understanding of collections. Furthermore, exposing the nature of museum significance assessment procedures and links to museum collecting policies foregrounds the institutions' role in meaning making.

Promoting plurality of meaning

Third, museums need to capitalize on a plurality of meaning inherent in collection items, rather than favoring one interpretation. Postmodernist principles suggest that an object's meaning and classification is not self evident nor singular, but is imposed on an object by a range of influences once inducted into the museum (Robinson & Cameron, 2003;11). So how can these issues be acknowledged and dealt with in documentation? Most basically this involves the rethinking of the relationships and meanings of collections in tandem with an understanding of the limitations of current approaches to knowledge making. The specific conceptual and disciplinary frameworks of art history, social history, science, decorative arts, anthropology and the institutional tradition curators work within, as well as the individual standpoints of museum staff determine the types of questions asked of objects. This then establishes the types of information documented and privileged in significance statements, and the subsequent values and meanings ascribed to individual collection items, thereby narrowing their potential interpretative possibilities.

For example a social historian may focus on the use, context and the personal significance of an object to the owner or user. In contrast, a decorative arts curator may note its form, function and material, how it fits into a chronological stylistic framework, the artist/maker and its history of ownership. A technology curator on the other hand could look at the same object in terms of how it is manufactured, its degree of technological innovation, function and how it works.

Our research also revealed the personal and idiosyncratic predilections, goals and interests an individual curator may consciously and tacitly bring to bear on the analysis and documentation of objects (Robinson & Cameron, 2003;19). That is, the task of object interpretation is a dialogic and organic process where various influences come into play and intermingle (Ibid;19). One participant in the decorative arts focus group discussion at the Powerhouse Museum succinctly espoused the feelings of others by stating that 'curators cast interpretations from different backgrounds and some objects have more to offer' (Robinson & Cameron, 2003;36).

But although many curators we interviewed were aware of the current discursive contexts of objects, appreciated the merits of other disciplinary interpretative frameworks and experiment with integrating themes in exhibitions, this process has been slow to be applied to documentation. So how do we reconcile these information needs, revise the types of questions asked of collections, contribute to their broader understanding and create a useful resource? This could involve collaborations between departments/curators to discuss potential meanings and the significance of selected objects, and the writing of joint significance statements to expose a range of meanings and opinions across disciplinary areas. In procedural terms this involves pulling frameworks together while expanding the potential meanings of objects.

Our research also raised a very important point about the problem of "conceptual fit" between existing object meanings and classification schemes and highlighted the difficulty of prescribing categories that can be applied universally (Cameron 2000). This was particularly obvious when we investigated issues around access to and the documentation of Maori and Aboriginal collections. That is, museums currently require the user to accept the institutional way of organizing information and the meanings ascribed to objects.

Enabling a polysemic and shared authoritative approach

So how can a polysemic and shared authoritative approach be applied in documentation? Here a new knowledge model invites the opportunity to create categories and associated linkages through documentation that could potentially connect objects with a whole range of cultural, social, historical, technological, artistic and disciplinary contexts. In conceptual terms, this involves the virtual layering of meanings and contexts of objects that can subsequently be configured in different ways in a live environment. Establishing relationships between objects on the basis of extended fields and additional ontologies could form the basis of retrieval methods. However, developing concepts on how this metadata is to be handled and standardized will continue to be an important issue.

Extending thesauri, nomenclatures and glossaries

Fourth, problems of conceptual structure and naming need to be dealt with in documentation. Ensuring greater intellectual access to collections and enriched data means extending existing thesauri, nomenclatures and glossaries. Most basically, this needs to occur during the documentation process in order to create a broader range of search and naming options thus contributing to and expanding interpretive options.

Use of non-text based information

And the fifth task involves addressing the emerging interest in the contextualization of objects and their non-textual representation. There is a trend, demonstrated through the Knowledge Objects project, away from detailed descriptions and primary text based provenanced information in documentation to a greater use of non-text based evidence in supporting multiple interpretive strategies. Here we need to consider how objects can be documented at acquisition, especially those deemed the most significant, in ways that capture a range of information. For example, this could include the digital recording of significance through comments by makers, users and donors and of objects in use. The documentation of the pre-museum context of objects will become increasingly useful.

Drawing on the wider information assets to best advantage

In Donovan's formative paper (1998) he convincingly argued for the building of knowledge bases drawing together the wider information assets of a museum into a centralized resource. The next task is to consider how existing museum data can be put to more productive ends in the documentation and interpretive process. Research files, documents, recorded interviews, graphics, audio/visuals, publications, interactives, and educational materials are all currently held in separate files and collections. Furthermore, in the future all these components could be indexed, managed and delivered by database tools while contributing to the creation of relational thematic linkages, multimedia presentations and rich research resources. But the challenge of delivering multimedia within the context of constantly evolving digital platforms will continue to have major implications for the ways in which collections information can be preserved, accessed, configured and interpreted into the future.

Additionally, the emergence of more creative content environments, in particular 3D visualizations and multimodal sensory experiences to which this material will contribute, are expected to become cheaper and more persuasive. In a collections context, this could potentially lead to the creation of complex cultural interpretations such as highly detailed and dynamic visualizations and navigation environments that have strong popular appeal. However, as Scali and Tariffi (2001) argued, the rethinking of collection management and multimedia delivery systems to provide effective access to such multimedia content will have major implications on how collections information can be configured and elucidated.

To conclude, a shift from the predominant use of highly prescribed authored information, text based descriptions and significance statements to a greater inclusion of interpretive materials around selected significant objects will involve new curatorial roles. Curators may become more involved in bringing together and linking forms of evidence, for example creating relationships between information and objects similar to an exhibitions paradigm. Likewise, the tasks of collection managers may witness a greater emphasis on creating and linking digital resources. Instead, this will ensure a more open and inclusive approach to the ways these materials can be interpreted, giving greater power to the user to create their own knowledge pathways and to make their own interpretations in a kind of shared authorship. While this will undoubtedly require greater effort and resources in compiling each individual record, it will simultaneously allow a greater depth to knowledge woven around a given object, thereby maximizing its meaning potential (Robinson & Cameron, 2003; 23). One answer to the problem of workload is to institute a program of documentation where deemed significant objects get prior standing. Others however could be recorded according to a more simplistic polysemic template where hyperlinks in documentation are used as the primary tool in the creation of relationships and to link sources. Above all these concerns, moreover, such practices correspond and embrace contemporary theoretical premises, that is, the recognition that objects are legitimately interpretable in a variety of ways (Ibid). This new way of curatorial thinking was aptly expressed by a participant in the science and technology focus group discussion, 'As a curator, I would define myself as a 'knowledge broker': we have to be aware of notions of pluralism and acknowledge the lack of singular authority of the museum or curator' (Robinson & Cameron, 2003:26).

So the Themescaping Virtual Collections and the Knowledge Objects projects have revealed some possible directions for the next generation of online collections and documentation. It looks exciting but one of our greatest issues is the tension between the potential richness collections have to offer and the relatively poorly resourced institutions we work within. This issue, first raised at the Digital Libraries Research for Access to Cultural and Scientific Resources meeting in Luxembourg in March 2000 (IST European Commission, 2000) will continue to be the single biggest defining matter in the future development of digital collections and knowledge environments.

Nonetheless, developing knowledge bases and other potential futures such as interoperable libraries of collections (Besser 2002) and collaborative knowledge spaces (Fleishman et al 2002) can be viewed as an incremental process occurring through the establishment of a program of iterative development. Yet convincing management of the educational and marketing value of such initiatives, and to commit substantial resources to relevant projects will be one of our most fundamental strategic issues for the future.

Acknowledgments

This research was made possible through grant funding from the Australian Research Council. Special thanks goes to my colleagues, Sarah Kenderdine and Kevin Sumption, Powerhouse Museum and Bil Vernon, Vernon Systems Ltd, partner investigators on the Themescaping Virtual Collections project. Furthermore, I would also like to acknowledge Professor Stephen Garton, a fellow chief investigator, who has been a wonderful mentor and supporter. But most importantly I would like to express my gratitude to research assistant, Helena Robinson who made an outstanding contribution to the Knowledge Objects project.

References

Adolsek, D. & M. Freedman (2001). Artifact as Inspiration: Using Existing Collections And Management Systems To Inform and Create New Narrative Structures. Museums and the Web 2001 Conference, 14-17 March, Seattle, Washington (http://www.archimuse.com/mw2001/papers/andolsek/andolsek.html) Last accessed 1.14.03

Besser, H. (1997). Integrating Collections Management Information into Online Exhibits: The World Wide Web as a Facilitator for Linking 2 Separate Processes. Museums and the Web 1997 Conference, 16-19 March 1997, Los Angeles, CA. (http://www.archimuse.com/mw97/speak/besser.htm). Last accessed 1.15.03

(2002). The Next Stage: Moving from Isolated Digital Collections to Interoperable Digital Libraries. In First Monday Volume 7, Number 6 (June 2002), (http://firstmonday.org/issues/issue7_6/besser/index.html). Last accessed 1.14.03

Cameron, F.R. (2000). Shaping Maori Histories and Identities: Collecting and Exhibiting Maori material culture at the Auckland and Canterbury Museums, 1850s to 1920s. Ph.D. thesis, Massey University, New Zealand

(2002). Wired Collections — the Next Generation. In the International Journal of Museum Management and Curatorship, in press

Cameron F. R. & S Kenderdine, (2001). Themescaping Virtual Collections: Accessing and Interpreting Museum Collections Online, unpub. mans. University of Sydney and the Powerhouse Museum

Donovan, K. (1998). The Best of the Intentions: Public Access, the Web & the Evolution of Museum Automation. Museums and the Web 1997 Conference 16-19 March 1997, Los Angeles, CA. (http://www.archimuse.com/mw97/speak/donovan.htm). Last accessed 1.15.03

Dierking, L. D. & J. H. Falk, (1998). Understanding Free-Choice learning: A Review of the Research and its Application to Museum Web Sites. Museums and the Web Conference 1998, 22-25 April, Toronto, Canada (http://www.archimuse.com/mw98/papers/dierking/dierking_paper.html) Last accessed 7.12.02

Experience Music Project (http://www.emplive.com). Last accessed 7.12.02

Fleischmann, M., W. Strauss, G. Blome, J. Novak, S. Paal, (2002). Netzspannung,org — A Collaborative Knowledge Space for Media Art and Technology. Museums and the Web 2002 Conference, 14-17 April 2002, Boston, Massachusetts (http://www.archimuse.com/mw2002/papers/blome /blome.html) Last accessed 1.14.03

Gillard, Patricia, (2002). Cruising through History Wired. Museums and the Web 2002 Conference, 14-17 April 2002, Boston, Massachusetts http://www.archimuse.com/mw2002/papers/gillard /gillard.html

Hypermuseum (http://www.HyperMuseum.com). Last accessed 1.15.03

IST European Commission DIGICULT (http://www.cordis.lu/ist/ka3/digicult/). Last accessed 7.12.02

Kavanagh, G. (1990). History Curatorship. Leicester and London: Leicester University Press

Kravchyna, V & S.K. Hastings (2002). Informational Value of Museum Web Sites. In First Monday, Volume 7, Number 2 (February 2002), (http://firstmonday.org/issues/issue7_2/kravchyna/index.html) Last accessed 11.20.2002

Marty, P. F. (1999). Museum Informatics and Information Infrastructures: Supporting Collaboration across Intra-Museum Boundaries. Archives and Museum Informatics 13 1999, 169- 85

Museum of Rural Life http://www.ruralhistory.org/index.html. Last accessed 1.15.03

Revealing Things (http://www.si.edu/revealingthings). Last accessed 1.15.03

Robinson H & F. R. Cameron, (2003). Knowledge Objects: Multidisciplinary Approaches in Museum Collections Documentation, unpub. mans. University of Sydney

Sarasan, Lenore & Donovan, Kevin (1998). The Next Step in Museum Automation: Staging Encounters with Remarkable Things (the Capture, Management, Distribution and Presentation of Cultural Knowledge On-Line. Occasional Papers on the Value and Use of Museum Information, Willoughby Press (http://www.willo.com/cgi/content.cgi?main.html,article_nextstep.html,left,article_nextstep.html,right) Last accessed 1.15.03

Scali, Gabriele & Flavio Tariffi, (2001). Bridging the Collection Management System Multimedia Exhibition Divide: A New Architecture for Modular Museum Systems. ICHIM Conference, Milan, 3-7 September 2001 (http://www.archimuse.com/ichim2001/abstracts/prg_115000625.html). Last accessed 7.12.02

Stuer, Peter, Robert Meersman & Steven De Bruyne, (2001). The HyperMuseum Theme Generator System: Ontology-based Internet Support for the Actual use Of Digital Museum Data For Teaching And Presentation. Museums and the Web 2001 Conference 14-17 March 2001, Seattle, Washington. (http://www.archimuse.com/mw2001/papers/stuer/stuer.html) Last accessed 1.15.03

Themescapes Focus Group, Curators — Collection Managers, Te Papa. Museum of New Zealand, 2001, unpub. mans., University of Sydney

Themescapes Focus Group, Non-Specialists, 2001, unpub. mans. University of Sydney

Thomas S. & A. Mintz, (1998). The Virtual and the Real - Media in Museums. Washington D.C.: American Association of Museums