Return to Babel: Emergent Diversity, Digital Resources, and Local Knowledge more

Return to Babel: Emergent diversity, digital resources, and local knowledge Dr Robin Boast University of Cambridge, Museum of Archaeology and Anthropology (UK) Dr Michael Bravo University of Cambridge, Dept. of Geography & Scott Polar Research Institute (UK) Dr Ramesh Srinivasan University of California at Los Angeles, Department of Information Studies (USA) 11 December 2005 Abstract: Introduction: Information system research that recognises, supports and enables access and presence of diverse knowledge communities on-line has become a major concern recently. These issues are especially important for Indigenous groups who are creating digital resources which support and preserve local identities and attempt to invoke grassroots involvement in devolved governance. The aim of this paper is to expose some of the conceptual obstacles to programmes of devolved and local knowledge resources, and to provide an account of agency that recognises that on-line communities, and community identity, is essential for eliciting, managing and sharing local knowledges. Further, this paper will demonstrate how the implications of these issues impact institutions for managing knowledge and may be extended to all sorts of knowledge communities. DRAFT: Not to be quoted without permission of the Authors (December 2005) 1 The issues are explored in three parts. First, we will explore the background to the existing solution, that of standardisation or meta-ontologies, and consider problems with this approach. Second, we explore recent responses from grassroots and opensource programmes that have sought to directly engage with and accommodate community diversity. Finally, we examine the epistemological problems with both approaches. We conclude by arguing that a lack of consideration of community and negotiated agency (knowledge), a lack of consideration of temporality and spatiality of knowledge, a lack of a sufficient understanding of the relationship between authorship and authority, and an insufficient attention to the emergent nature of local knowledge forms the key issues through which solutions may be sought. Finally, we discuss an ongoing digital museum project that attempts to navigate the issues raised in this paper. Standarising and meta-ontologies: Universes of commensurability From the creation of the first large public libraries, to databases, to the birth of the World Wide Web, the problem of how to find what you are looking for, knowing what it is when you found it and finding associated knowledge has been a major challenge (Wiegand (1996); Neufeld, M.L. and M.Cornog (1986); Chenhall, R.G. (1975); Garshol, Lars Marius (2004)). The usual solution to this problem, from at least the mid-19th century (Wiegand, 1996), has been to create an index. The index is, of course, merely a small set of fairly standardised categories that can be cross-referenced. In the early days of the Dewey system, sold by the Library Bureau in the 19th century, libraries could buy cards which imposed the Dewey classification. From the end of the 19th 2 DRAFT: Not to be quoted without permission of the Authors (December 2005) century, libraries could purchase ready printed cards for any particular book using the U.S. Library of Congress classification system, if they wished to use that system. Museums too have created many such systems from the mid-19th century as have archives. Often the systems were ad-hoc, serving the needs of the particular collection, institution or research community, but quite often standardised systems were used as well. From the beginnings of the Web, similar indexes have been tried, but the size and diversity of the Web has generally led to these attempts being seen as unattractive by the larger public.i The problems with systems of ad-hoc indexing are well known. The first is that even within a single collection, different terms were used for 'similar' objects, or different descriptive terms were used to describe the 'same' attribute of an object. In addition, terms have changed over time so older terms persist while 'similar' objects are given new terms. To overcome this problem Thesauri have been increasingly introduced to enforce usage and standardise the indexing process (Doughty, P.S. (1977); Marcum, Deanna B. (2002); MDA Development Committee (1982)). Another dilemma indexing systems met has been the changes they have encountered over time. As new systems of indexing emerge, previous forms of categorization, which may have maintained useful insights, were lost. Moreover, a number of programmes of 'standardization' have existed for years in museums, libraries and archives to ensure that all such institutions maintain the same system (Canadian Heritage Information Network (http://www.chin.gc.ca/English/index.html); The Getty (http://www.getty.edu/research/conducting_research/standards/intrometadata/ind ex.html); The Museum Documentation Association of Great Britain DRAFT: Not to be quoted without permission of the Authors (December 2005) 3 (http://www.mda.org.uk/); CIDOC (http://www.willpowerinfo.myby.co.uk/cidoc/index.htm); MARC Standards (http://www.loc.gov/marc/); The Open Archives Initiative (http://www.diglib.org)). However, imposing a meta-indexing system has sacrificed some of the findings and value that particular institutions may have developed (or been able to develop otherwise) that are specific to their own collections, let alone across assemblages of collections. Though the vast amount of work globally that has gone into making information accessible -- for digital resources in museums, archives, libraries and on the Web generally -- undoubtedly counts as a positive trend, how often do we stop to ask what we mean by 'accessible'? Usually we are talking about opening the metaphorical doors of our repositories of information to enable new audiences to read or view our documents or exhibits, in the hope that they will create connections between what they learn from our collections and their knowledge or experience gained from other sources. What makes this digital interconnectivity possible? The general response to the challenge of linking disparate information -- of interoperability -- has been to standardise systems of classification and description. These processes of standardization are motivated by the need to catalogue ever-growing digital resources to enable users to access them more efficiently and uncover connections between information objects. In the language of database management, standardising digital resources entails making them referenceable, rapidly retrievable, interoperable, or expressible in relational terms to other objects within the database. In the drive for greater interconnectivity, a key question has emerged that DRAFT: Not to be quoted without permission of the Authors (December 2005) 4 standardization pundits must grapple with: are the organizational standards created from a deep knowledge of the context, cultural use, and meaning of the objects to which they refer? If not, does standardised classification and description restrict audiences or limit the range of ways in which they can interact with the objects? Although it is convenient to imagine standardisation as a passive and invisible set of structures that enable rather than restrict interaction, a little reflection shows that this is not so. Scholars of standardization themselves acknowledge that with the goal of rapid retrieval, the meaning and description of the object may in turn be sacrificed. In practice, what transpires is an imposition of the efficiency-driven priorities of the public institution upon its diverse publics (Bower, Jim and Andrew Roberts (2001); Doerr, Martin & Nicholas Crofts (1999); Marcum , Deanna B. (2002). For critique see: Turnbull, David (1991) and Bowker, Geoff and Star, Leigh (1999)). As a result visitors and other stakeholders are potentially exposed to objects that have been described with information that derives from data interoperability goals instead of the object's historical and cultural context of origin. We may therefore ask whether data can remain interoperable without stripping underlying meanings that are essential to an object's emergence and cultural significance in the first place. This paper focuses on the possibilities of designing a new kind of information access system: one that captures the cultural diversity of knowledge resources while still incorporating sufficient systematic information to enable effective retrieval. We recognise that there exists an inherent tension between these two aims, and we argue that certain kinds of existing digital resources go some way to satisfying them both, and hence provide us with a useful point of departure. For any such argument to have real DRAFT: Not to be quoted without permission of the Authors (December 2005) 5 purchase, we must have in mind some very preliminary ideas about the kinds of communities who would become the users of these resources. For the purpose of this paper, we shall make reference to a concomitant ongoing project, “Emergent Databasing Emergent Diversity” (ED2), in which we are seeking to develop new digital resources and implement them in collaboration with several indigenous communities from different geographical regions.ii This inquiry exceeds simply including more facts about the cultural origins and contexts of digital objects. It requires thinking through how the fundamental concept and architecture of an access system can be conceptualised and put to work in knowledge spaces and cultural settings that may not share much in common with museums or archives. We need to consider how local knowledge communities provide contextual meaning to the practices by which their objects are represented in information systems. Take for example an Inuit hunter who has carved a soapstone model of a kayak in response to a commission from a museum collector. When the carver makes the model from memory, he is selective about which features matter most. To meet the criteria of the commission, he will scale it down to fit his block of soapstone, chosen small enough to be able to make a profit from the commission and ship it out by air, but large enough to be able to convey the correct form. Some features of the kayak, such as the harpoon head, will be larger than their true proportions to make visible and emphasise the important details in its function. When visitors access a digital image of the kayak from an online museum website, for example, standard information like the length of the model may seem important, but they are more likely to be captivated by the carver's DRAFT: Not to be quoted without permission of the Authors (December 2005) 6 memories of the feelings, skills, and textures of kayak travel that inspired the model. The goal of the online system should be to expose the visitors to these deeper sets of knowledge associated with the kayak -- the objects, thoughts, and memories associated with the kayak. But how likely is the museum collector to have had the time and privilege to have paddled with the carver or to have documented his experiences through a patient oral history? A much stronger likelihood is that the carver's hunting partner, or his grandson or granddaughter may want to tell the stories that explain how the experiences out of which the model kayak emerged. We know that oral tradition is strongest in the places where it is set; not very many Inuit have the opportunity to make long journeys to museums to document the artifacts, which must often seem strangely out of context. This story of the kayak though based on one particular example, typifies the inherent tension between the richly situated life of objects in their communities and places of origin, and the loss of narrative and thick descriptions when transporting them to distant collections. In the digital world, the open source movement has, in certain respects, a good track record of helping to bring together, juxtapose, and link heterogeneous resources like oral histories, photographs, narratives, and carvings. Web-based authoring has shifted from solely following pre-defined templates, to enabling communities to sculpt the media that they utilise in their communication with one another. Often, the open source movement has worked to engage already technologically involved populations, through “virtual communities” (Rheingold, Howard (1993)) that are defined by the substrate around which they meet, exchange resources, and develop and extend specialised social networks (Gulia, Milena and Barry Wellman. (1998); Smith, Marc and DRAFT: Not to be quoted without permission of the Authors (December 2005) 7 Peter Kollock (1999)). In describing how open source ethics may be extended to other communities, it is vital to keep in mind how diverse communities start with their own assumptions about the value of different technologies, and employ different strategies in appropriating them. For example some groups of Australian aboriginal artists are extremely sophisticated and experimental in their use of museum spaces. Canadian Inuit artists and filmmakers have in general shown much less interest in exploring museum spaces, though they in many respects 'techno-philic', and some have successfully used their domestic satellite broadcasting facilities and websites as platforms for showing that performances and productions created locally can be of a standard that earn the highest international acclaim. For many indigenous artists, creating new virtual communities with people in far-off places is less enticing than employing their authorship to establish new rapports amongst their own people -- for them the focus of community-building is usually at home. They recognise that marginalization is a pervasive phenomenon that takes many forms locally and regionally within a society no less than between disparate societies on a global scale. The application of the Web to the traditionally marginalised -- a dynamic that cuts across national and ethnic bounds -- remains an important issue of research enquiry. Certainly, the Web has enabled voices to be heard and circulated that may have been suppressed by broadcast media or where the ownership of the information infrastructure lay in the hands of a single government or private entities. Amongst the critiques of traditional central-peripheral models of power and mobility has been the decentralised authorship and lateralizing power that the Web provides. This dynamic DRAFT: Not to be quoted without permission of the Authors (December 2005) 8 would only be magnified if a previously marginalised community is also being allowed to articulate its own classifications and categorizations for the information systems that serve it, and consulted as an important source of input in emerging global digital knowledge collections. Indeed, research has found that community-classified information systems can mobilise infrastructures of education and literacy (Novak, J. D. and D. B. Gowin. (1984); Kavanaugh, Andrea L. and Scott J. Patterson (2001)), assetmapping (Kretzmann, John and John McKnight (1993)), and collective decision-making (Shapiro, Michael et al (1988); Srinivasan, R. (2004)). Indeed, marginalised ethnic voices have begun to populate cyberspace (Mitra, A. (2004)), and the central-peripheral models of power that Castells for example (Castells, M. (1997)) has formulated, warrant a re-framing in an era where the search engine may maintain greater impact over information prioritization and retrieval than traditional economics. However, there are multiple means and routes by which these dynamics may be extended. We highlight two possibilities: First, how can disenfranchised communities extend the power of authorship to the power of classification? In other words, how can research address the potential of communities to author their own ontologies, database representations, and therefore the means by which information systems represent the content they submit? A small but growing body of research has pointed to the power of engaging communities to elicit and create their own classifications in the information systems they utilise and own (Boast, R. (1995); Turnbull, David (2003); Srinivasan, R., and J. Huang (2004)). This research has demonstrated the ability of a system to harmonise with existing cultural discourses and activities rather than represent cultural information around alien DRAFT: Not to be quoted without permission of the Authors (December 2005) 9 standards and paradigms. Second, how can this research integrate with digital libraries and museums, and the mission they hold to serve as information institutions for multiple publics? Cross-cultural issues remain important in digital libraries research, but tend to be solely addressed around issues of icons, language, search histories, and so on. These approaches can be extended with greater understanding of the narratives and knowledge that are specific to a community, allowing one to uncover cultural differences in information seeking (Geertz, C. (1978)). The power of a digital library or museum could therefore be extended by accounting for differences in their representations of different cultural objects. Can such a system promote access for different knowledge communities, diverse research communities, or the public in general? There is already evidence to show that new decentralised Web-based systems of access and classification have resonated with Web users -- Wikis, Tagging, Blogging and RSS to name but a few. These are exciting developments, yet none has been applied to the domain of cultural institutions and communities that may not be information technology-fluent or -exposed. This paper is therefore focused around the problematic of creating systems that accommodate the diversity, multiplicity and incommensurability of multiple knowledge communities and develop prototype digital contact spaces. A number of important steps, we argue, are critical to develop such culturally-driven information systems. First there is the need to develop a new methodology by which public institutions can access and maintain digital knowledge resources in a manner that directly consults stakeholder communities and cultures, as well as other diverse knowledge communities. This methodology must recognise all aspects of Intellectual Property DRAFT: Not to be quoted without permission of the Authors (December 2005) 10 Rights, those of the author(s) and those of the institution(s), and recognise the distributed and diverse nature of ownerships. Second, it must provide a mechanism by which new digital knowledge resources can be made to carry the supporting contextual narratives and authorship(s) of an object or knowledge statement, while still maintaining the goals of interoperability, referenceability, and retrievability that has driven the standards movement. Third, it must create a strategy to encourage publics to take up authorship, to describe, and to classify digital knowledge resources to rework or create new emergent social meanings within their communities. And, finally, it must describe an approach that public institutions can take toward maintaining more proactive relationships with constituent and diverse publics, as well as present exhibitions to the general public that recognise and respect the histories and important insights and cultural differences that characterise stakeholder communities. This highlights the greatest problem that standardising approaches have always grappled with: the fact that knowledge, authority, and modes of framing/classification have always been culturally distributed and have been frequently been epistemologically incommensurable. In other words, the knowledge of a particular topic, object, geography, or process, may not always be able to be uncovered within a single description or descriptive trope. Instead, many such insights can only be found through consultation with various experts or stakeholders, that maintain age-old forms of orality, and even then, can seldom be translated or reduced to a single framework without significant loss of understanding. The library and museum have attempted to integrate these different perspectives, and have done an admirable job. Yet, we must admit the impossibility of containing all forms of knowledge and think through the DRAFT: Not to be quoted without permission of the Authors (December 2005) 11 consequences of this for community-building. Over the past few years, standards research has begun to recognise the importance of integrating multiple perspectives and data sources, and has therefore sought alternative solutions -- solutions that consider multiple Ontologies. We define the term 'ontology' as a structure or representation of knowledge. It is a term that invokes a variety of questions, and circulates in fields including Computer Science, Philosophy, Science Studies, and Anthropology. These expressions of ontology can be as distinct as the architecture of databases (in computer science), the fundamental categories of being (philosophy), and the classification of culturally discursive knowledge (structural anthropology). In the case of this paper, ontology is understood as the way in which a certain community negotiates the conceptualisation and organisation of its knowledge and information. This may encompass the community's notions of geography, temporality, and social, political, educational, and cultural priorities. The goal of these ontology-driven enquiries is to create ever more encompassing classifications that can accommodate ever-wider communities. With this motivation, the CIDOC (International Committee for Documentation of the International Council of Museums) Ontology states that: The primary role of the CRM [Conceptual Reference Model] is to enable information exchange and integration between heterogeneous sources of cultural heritage information. It aims at providing the semantic definitions and clarifications needed to transform disparate, localized information sources into a coherent global resource, be it within a larger institution, in intranets or on the Internet. (Crofts, Nick et al (2003)) DRAFT: Not to be quoted without permission of the Authors (December 2005) 12 The assumption is that locally each collection can maintain its own level of indexing and description, but that it simply must translate this local level of description into the more general ontology for wider usage (See also the MLA Net-Gain initiative (http://www.mla.gov.uk/action/online/netgain_vision.asp)). Another example for the Web is the Dublin Core Metadata Initiative (Duval, Erik et al (2002); See also the Dublin Core (http://dublincore.org/)). Here a very basic set of categories should be added to every web page, using standardised tags, so that a wide community can search these basic and universal categories. But does this really solve the problem of interoperability? If this disparate data is translated and annotated with metadata on every digital resource, will it allow a vast multicultural, multilingual and multipurposive community find what they want, in the way that they want, in terms that they define, and, then, understand what it is? The crucial weakness of standardizing programmes is that they either relegate this distributed difference to secondary or peripheral status, or they assume that the metadescription can encompass 'lower' levels of description. However, if every description obtains validity from a local contextualised ontology, then these descriptions must be accorded the same validity as the meta-description, in which case the meta-description becomes but one description among many. Therefore, the goal must be to abandon the search for a perfect meta-description, and instead to find ways to accommodate and embed multiple descriptions and their ontological contexts (Eco, Umberto (1997)). The fact is that individuals within a given community attach different descriptions to shared phenomena, and they need to continue to describe the world differently. As each of us is a member of different communities, we each describe and classify our DRAFT: Not to be quoted without permission of the Authors (December 2005) 13 world using different concepts at different times and for different purposes. And these different descriptions -- these contrasting and fluid ontologies (Boast, R. (1997); Boast, R. (2000); Srinivasan, R.and Huang, Jeffrey (2005)) -- remain important. Not only are they useful, but they are the ontological keys that unlock the doors to diverse, rich and incommensurable knowledge communities. They are not merely alternative translatable ways of expressing the same piece of knowledge, but more accurately are diverse 'ways of knowing' about the world and are necessary to organise, find and use information. Hence, embracing these multiplicities is fundamental to experiencing and creating faithful representations of knowledges. To summarise our position thus far, we recognise that it is important to explore how resources can be integrated while ensuring that they continue to be distinguished in terms of classifications and distinctions that are meaningful to the diverse knowledge communities from which they arise. The universalist and the pluralist approaches should coexist and be held in dialogical tension, while maintaining spatial and temporal unity -- but without a unified meta-ontology. We further posit that developing information systems that show differences across interpretation and use of objects is an important development in its right. It follows from working with multiple ontologies and without a meta-ontology that the interactions/outcomes cannot be anticipated or subjected to a rule-based process. That is to say the process is emergent in a weak sense. But this argument also holds in a stronger sense -- this new approach would enable new synergies, new interactions, unanticipated connections to be uncovered, exposing emergent knowledge. These are the vital elements for communitybuilding that will make the better access to collections so much more significant DRAFT: Not to be quoted without permission of the Authors (December 2005) 14 (Turnbull, David (2003); Turnbull, David (1991)). Tags, RSS, Weblogs and other responses to diversity. The emergence of new approaches toward describing and classifying knowledge is testament to the importance of re-describing digital resources with a more fluid ontological approach (Bowker, Geoff and Star, Leigh (1999); Srinivasan, R. and Huang, Jeffrey (2005)). As an example, we consider the recent explosion of Tagging, RSS and Weblogs. Over the past eighteen months, a number of sites have arisen -- bookmarking sites such as del.icio.us, Furl and Technorati, or photo sites such as Fliker (http://del.icio.us/; http://www.furl.com/; http://www.technorati.com/; http://www.flickr.com/) -- that allow the user to “tag” a website or a submitted digital photo. The Tag can be expressed in the user's own language and will reflect the annotation from which the digital resource can be re-accessed in the future. It can be in any language and use any idiom. There are no restrictions but that it has to be a word (or words). These Tags can be solely for the use of the individual, although the individual can make them public, and, on del.icio.us, you can now cluster your Tags thus creating a kind of grouped categorization (This Clustering is not the same as a taxonomy as it is not hierarchical. It is, even on the individual level, merely a grouping and clusters can and do overlap and are not at all hierarchical or exclusive). Some of these sites offer a folksonomy, or a listed description and ordering of the top 250, or so, accessed public tags. This approach has generated a great deal of literature and commentary (There is a huge amount of discussion out there, but see particularly DRAFT: Not to be quoted without permission of the Authors (December 2005) 15 Wikipedia's entry on Folksonomies (http://en.wikipedia.org/wiki/Folksonomy), Thomas Vander Wal's Blog (http://www.vanderwal.net/random/category.php?cat=153); Clay Shirky's Blog (http://www.corante.com/many/archives/2004/08/25/folksonomy.php), and Peter Van Dijck (http://poorbuthappy.com/ease/archives/2005/01/15/2419/)). Much of this discussion has reflected a sense of amazement that despite predictions by advocates of traditional meta-standards that such a methodology that should lead to chaos and disorder, this breakdown has not materialised; in fact the folksonomies seem to generate great activity and usage, implying that they have engaged their user publics. Technocrati now boasts over 900,000 Tags (http://del.icio.us/; http://www.technorati.com/; http://www.flickr.com/), yet there is pattern and sense to the Tags pages. Users seem to be able to navigate in a richer manner than with either free text searches or structured indexes. And the success of Tags is an emerging reality, Yahoo is introducing Tags and so is Google (http://myweb2.search.yahoo.com/; http://www.google.com/tags/). CityUlike and Conotea (Connotea: http://www.connotea.org/; CiteULike: http://www.citulike.org/) both allow academics to Tag articles on JSTOR and other academic e-publication sites, and Matthew MacLaurin, a program manager in the Social Computing Group at Microsoft Research, predicts that Tagging and Folksonomies are the next quantum leap on the Internet (Quoted in “Steal this Bookmark!”, (http://www.salon.com/tech/feature/2005/02/08/tagging/print.html)). Will Tagging over-populate the Web with meaningless annotations and even more uncontrollable and un-navigable information than ever before? The logical DRAFT: Not to be quoted without permission of the Authors (December 2005) 16 consequence, critics argue, is that mass cultures will populate the Web precluding any notions of hierarchy or priority. Who really wants to search Flicker for the most popular term which happens to be “photo”, or another popular term “me”? This suggests a lack of discrimination, an absence of terminological control, a vacuum where there should be a central authority; so the argument goes (See Matt at http://www.allpeers.com/blog/?p=12). Our response to these concerns is to acknowledge their significance, and to draw attention to the sociological issues at stake. Identifying the presence and location of authority and regulation in a Tagging system cannot be separated from the social norms of the user communities. We posit that structure in Tagging emerges because belonging to a community entails people having access (albeit differential access) to a body of shared knowledge and understandings, being guided by norms for regulating everyday behaviour, possessing a shared vocabulary, and following tacit procedures for producing agreement. The ways that individuals respond to each other through these norms and expectations, are the sources of emergent shared meanings. It follows from our position that the emergent structures of Tagging will reflect particular features of the communities that produce them. In principle, Tagging allows for individual categorisations to be available to all publics simultaneously. However, the significance of these categorisations remain at the level of either the individual or the web-community generally. While this gives us cause for optimism, the challenge for us remains to learn how smaller communities who may not grant any special status or intrinsic social value to digital resources, create emergent structures in practice. There is no privileged methodology to do this, but we believe DRAFT: Not to be quoted without permission of the Authors (December 2005) 17 that attitudes towards authorship and context are of great importance, and have begun to receive attention through RSS, which stands for “Really Simple Syndication” (or “Rich Site Summary”) (Pilgrim, Mark (2002) and Wikipedia, RSS (file format). http://en.wikipedia.org/wiki/RSS_%28file_format%29). RSS is a recent technology that has been appropriated by the grassroots knowledge communities. Basically it is a simple XML tagging convention that allows publishers, bloggers, news agencies, companies, etc. to send dynamic feeds to their web pages for anyone who wishes to subscribe -- but without imposing a standardised ontological schema. It is an interesting example in that it resolves many of the problematic issues critics of Tagging and Wikis have pointed to regarding context and authorship. It maintains signatures and authenticity yet remains grassroots because it is managed by institutions, individuals, groups and communities. Users may 'subscribe' to RSS feeds thus choosing what they want; and RSS is extendable to include filtering of individual RSS feeds for personal interests. RSS feeds can now add a whole new level of trusted local knowledge to different individuals' and communities' web worlds, which encompass weblogs and tags. Crucially the trust in the knowledge and the decision to accept a feed both reside in the subscribing individuals. These are early days for RSS, and much needs to be developed. At the moment, it is easy to subscribe and filter RSS feeds locally. They are integrated into weblog technology and have begun to be incorporated into Search engines. However, what is still missing is the application of these new approaches to the norms and practices of traditional institutions that hold and maintain knowledge, and interactions that engages communities as co-producers and classifiers, rather than simply as technology DRAFT: Not to be quoted without permission of the Authors (December 2005) 18 consumers. A philosophy that recognises the value of structures that enable the authoring of knowledge resources to be distributed and decentralised has motivated technologies like RSS and has arguably been justified by their success. Take for example, weblogs (hereafter abbreviated by the colloquial term 'blogs'). Blogs are useful not only to allow people to 'get things off their proverbial chest', but also because they are a good way for people who may consider themselves however tentatively to belong to nascent communities, to identify, discuss and share common problems and understandings -- to share and develop knowledge through the development of a community. Our analysis of the need for new digital resources recognises the contribution of Tagging, RSS, and Blogs to decentralizing access to knowledge resources away from centres of calculation to individual users. We should stress however that the terms 'individual' and 'decentralise' are not transparent, they carry with them assumptions about the sociality of their members. Anthropologists warn us that the relations implicit in the concept of the individual varies tremendously from one society to the next. In Melanesia for example, the individual is defined through a complex set of relations with all other living beings, not as autonomous beings acting out their own free will. Strathern in her canonical analysis of technology and social structure, argues convincingly that the ways in which technologies constitute individuals and their kinship relations, require deep analysis, and should in no way be taken for granted. Hence if Tagging, RSS, and Blogs currently function to create more autonomous webbased communities of users, we should not assume that these end users are a guide to the many grassroots communities of origin we hope to engage with. By the same token, DRAFT: Not to be quoted without permission of the Authors (December 2005) 19 we should be careful to keep in mind that the significance of 'decentralization' is equally underdetermined and contingent. What knowledge practices can be decentralised -accessing, authoring, and describing collections -- depends on how knowledge is situated or lived in the recipient communities. One possibility may therefore be to allow users not only to assemble their own resources, Tag them and comment on them, but also to assemble together what we might call contacts, storytellers or guides. We must recognise that tagging and other techniques we have discussed are fundamentally textual and pre-suppose a literacy that drives the annotation process. Keeping in mind that for many indigenous communities, literacy is only one element within traditions that maintain more deeply oral, performative or visual forms of description and communication, we will have to adapt our ontology-description process accordingly. In previous research, we have worked with communities that have adopted telecommunications and computing technologies to maintain such traditions and have worked to develop media that have a resonance with these findings (Srinivasan, R. (in press); Bravo M. T. (2002); Bravo, M. T. and S. Sörlin (2002); Michaels, Eric (1986); Michaels, Eric (1990)). Indeed, the methodology of this research points to possibilities by which all parties in this project can develop capacities toward self-reflection, and therefore re-explore notions of history, education, politics, and so on. Particularly in the case of oral communities that have ritualised the notion of progress, development, and future vision, research has found that information projects can generate important collective dialogues and mobilizations from within the community (Srinivasan, R. (2005); Blood, Rebecca (2000)). Indeed, the ongoing research offers the possibility of creating DRAFT: Not to be quoted without permission of the Authors (December 2005) 20 information research relationships that are self-sustaining and nurturing between the participant institutions and such diverse stakeholder communities. These relationships are fundamental to the mission of all public institutions of knowledge, and this approach may uncover an effective ability to allow these connections to be forged and maintained. In parallel, this approach may also enable institutions to develop presentations -- whether written works, on-line presentations or exhibitions -- that are more deeply reflective of the cultures from which the objects have been presented. And, for countless others, our goal is an engagement with the digital resources through the diverse layers of understandings of these various individuals and communities. Virtual society vs. distributed communities The World Wide Web was created as a medium by which highly specialised scientists could interact and engage with different communities of highly specialised scientists. It was a system that was designed for many small, diverse and distributed research communities, but also for the interaction between these research communities (Moschovitis, Christos J.P. et al (1999)). However, as the Web grew it quickly took on the character of other broadcast media -- television, radio, newspapers, publishing. With the exception of the ever-successful email, the Web has become much like the library or the mall. It is not surprising that in attempts to make sense of this great virtual storehouse, the chosen metaphors are those of the index, the stock system, and the accountant (Strathern, Marilyn (2002)). It is also unsurprising that one of the dominant images of the Web is that of the Virtual DRAFT: Not to be quoted without permission of the Authors (December 2005) 21 Society, and that it is closely tied to ideas about consumption. Societies consume, they are large audiences to which messages, information, resources can be broadcast or distributed. Choice exists, often, in that we decide which of the many channels of information, entertainment or goods we allow into our home. But aside from the transaction, there is only one-way communication -- from broadcaster to recipient. The broadcasters or distributors do not enter into a discussion with the consumer, nor do they know anything beyond the generalities of their audience necessary for successful marketing. In contrast the goals of the first Webbed scientists, and the millions now using social computing systems, are much more small scale, intimate and dialogical. Where the Virtual Society is about messages and audiences, about distributors and consumers, social computing is necessarily about small-scale distributed communities interacting (Shirky, Clay (2002)). Scratch below the surface of the Web and you find a multitude of these diverse communities (Star, Susan Leigh (1995); Brown, John Seely and Paul Duguid (2000)), in fact this diversity has always existed. eScience always used the Web this local dialogical way, and the immense success of email over all other forms of Web-use demonstrates the success of such dialogical uses of the Web. Unlike most other broadcast media -film, television, radio, newspapers, publication -- the technical and economic limitations of production and distribution do not apply to any significant degree on the Web. But the Web remains dominated by the ethos, if not the socio-technics, of the Virtual Society with its strong enlightenment-inspired division of the world into individual voices and mass audiences. Though Tagging, Weblogs and RSS all offer the technology to distribute expertise and DRAFT: Not to be quoted without permission of the Authors (December 2005) 22 accommodate diversity, they continue to operate within this dichotomy of the virtual individual and the virtual society. Though these systems overcome many of the limitations of universal classifications or meta-ontologies, distribution continues to be defined in terms of mass audiences isolated within the virtual society; diversity persists through many voices but little dialogue; ownership is only available on the level of the individual and access on the level of the virtual society generally. Though it is tempting to think that we can somehow do away with the effects of mass media with new technology, the difficulty of reaching millions or even tens of thousands of people one community at a time is as much about human wiring as it is about network wiring. No matter how community minded a media outlet is, needing to reach a large group of people creates asymmetry and disconnection among that group -- turns them into an audience, in other words -- and there is no easy technological fix for that problem. Like the leavening effects of Letters to the Editor, one of the design challenges for social software is in allowing groups to grow past the limitations of a single, densely interconnected community while preserving some possibility of shared purpose or participation, even though most members of that group will never actually interact with one another. (Shirky, Clay (2002)) Clay Shirky's point is apt. On the Web, as with other media, it is relatively easy to define audiences and broadcast messages, or distribute information. It is easy, with a well defined and well trained audience to standardise the ontologies by which information, resources and goods are defined and, hence, accessed. It is also relatively easy to create and maintain relatively small and tight-knit special interest groups and 23 DRAFT: Not to be quoted without permission of the Authors (December 2005) communities. We use the methods of the library and the learned society respectively. However, the phenomenal successes of Google tools, del.ici.ous, Flicker, Blogger, and other social computing sites shows that these dominant methods are no longer universally accepted, if they ever were, and are instead meeting with active resistance. The now enormous Web community is demanding the right and ability to describe, use and populate the Web how they want and in ways that are meaningful to the small communities of which they are a part. But Shirky's challenge remains: how do we move beyond many little broadcasters to conversing communities? A working group has been set up to consider this problem. Emergent Databasing: Emergent Diversity (ED2) considers that to truly make the Web a knowledge resource, it is necessary to recognise and accommodate the negotiated, narrative, emergent and incommensurable nature of knowledge production and use. To this end, grassroots social computing is a step in the right direction, but it, largely, does not recognise certain problems. The greatest of these is the effects which result from the transformation of local, timely, contentious, and authored knowledge statements into information. The removal from information objects of temporality, spatiality, authorship, and contention, does not fix or confirm knowledge rather it removes from them the dynamics which make them knowable. We can look at three effects of this transformation: 1. Knowledge claims are of a time and of a place. The negotiated and narrative nature of knowledge requires that when a statement was made and where is essential to assessing not only its authority, but also its place in a knowledge programme. Information objects, on the other hand, are translations of these authored stories to 24 DRAFT: Not to be quoted without permission of the Authors (December 2005) timeless abstractions. This translation can often arise merely by the act of digitisation. However, information objects are not reducible to timeless abstractions, but take on the effect by the loss of temporality and spatiality. 2. Authority is a major problem for all information objects, but as authority is a matter of authorship, the knowledge of who the author is is usually seen as sufficient. As for all knowledge, however, authority cannot be assured centrally, but must be negotiated within different contexts of use. The loss of authorship, and the communities of affiliation, remove any possibility for understanding or accommodating authority for information objects. 3. Mostly, though, the transformation of knowledge to information creates the effect of fixity. Whereas most knowledge (in science, academia, local specialist communities, etc.) is dynamic, negotiated and constantly contentious, the transformative processes that create information remove these dynamics. Neither can the effects of the translation of knowledge objects to information objects, in particular digitally, be equated with the immutable mobile so prevalent in other media. Bruno Latour has argued that immutable mobiles are able to circulate and travel so successfully because they are standardised and even mathematised mobiles (Latour, Bruno, _(1986); Latour, Bruno (1987)). However, as John Law points out, with both Vicky Singleton and Annemarie Mol (John Law and Vicky Singleton (2003); John Law and Vicky Singleton (2003); John Law and Annemarie Mol (2003)), it is not that these mobiles are immutable, but to the contrary, but that they are mutable that allows them to be mobile and operate as objects around which local knowledge can operate. It is the degree that they can be embedded in different configurations -- different dialogues, 25 DRAFT: Not to be quoted without permission of the Authors (December 2005) different practices -- that allows them to operate as mobiles in knowledge systems (Knorr-Cetina, K. (1999)). But it is also often their ability to change their shape as well. We argue that it is only the degree to which these (un)stable forms of knowledge representation -- in science, publication, or any form of local knowledge representation -- remain associated with their place, time and authorship in an unfolding biography that they can they contribute to diversity. They remain useful because, by knowing their identity as a function of their unfolding biographical history, or their journey, can they be set within the dynamic and expanding negotiations that constantly work to constitute the knowledges of which they are an active part. If a mutable mobile is discredited, overruled, or abandoned it does not lose its validity for it retains its place and time in the local negotiations that are knowledge. If however it loses its associations, it becomes uprooted, displaced, and therefore severed from the people and places that validate its social meaning. Databases in particular, and the Web in general, treat information as though it does not have this requirement -- as though it is immutable and mobile. Broadcast media, by its nature, often treat the source of information as though it were timeless, authorless and placeless and, hence, non-negotiable. Standardising and Meta-ontological approaches have clearly signed up to this non-negotiable, immutable, broadcast understanding of information. However, the point we make here is that even many of the grass-roots approaches have signed up to it too. Only blogs seem to accommodate most of the aspects of successful knowledge mobiles -- authorship, temporality, narrativity, and negotiation -- but they do not provide much of a forum for emergence. While Tagging, RSS and Wikis all recognise that knowledge is an emergent effect of a community of DRAFT: Not to be quoted without permission of the Authors (December 2005) 26 knowledgeable individuals, these approaches treat this emergent effect as a timeless, placeless and universal abstract. The community that is allowed to create this effect is greatly enlarged, but the effect remains the same. Concluding thoughts The ED2 project is underway to demonstrate the potential of creating information systems that engage rather than dismiss or restrict cultural and local meaning. In that regard, the project holds the potential to extend and distribute within the cyberinfrastructure research that encompasses multiple nations and cultures. It presents a possible mechanism for understanding how to reconcile multiple traditions of knowledge around such important topics as health, education, and environment. We therefore hope that from the results of this project a pathway will be articulated from which systems can be 'grown' for general application for data-sharing in the Arts, Humanities, Social Sciences, Sciences, Indigenous Knowledge Communities, and the wider Web. Moreover, we anticipate that the ED2 project will speak to the growing body of community-ICT research, demonstrating a means by which public institutions can acknowledge and catalyze their diverse communities in information projects, and that systems may be created that recognise information flows and processing that are culturally differentiated across communities. We strongly believe that from this project, and ones with similar goals, we can present a mechanism by which public institutions can begin to develop, share, and disseminate collections that can acknowledge the diversity of cultures and environments of its constituencies. DRAFT: Not to be quoted without permission of the Authors (December 2005) 27 References: Blood, Rebecca (2000) Weblogs: A History and Perspective. Rebecca's Pocket. Written: 7 September 2000. Retrieved: 17 February 2005. (http://www.rebeccablood.net/essays/weblog_history.html). Boast, R. (1995) The Virtual Teaching Collection: Multimedia access to museum collections. Proceedings of "Information: the Hidden Resource, The Seventh International Conference of the Museum Documentation Association, Edinburgh. Cambridge, Museum Documentation Association, pp. 323-334. Boast, R. (1997) A small company of actors: a critique of style. Journal of Material Culture, 2(2):173-198. Boast, R. (2000) Speculum, Exemplar, Imago. In A. Lowe & S. Schaffer (eds.), N01SE: Universal Language, Pattern Recognition, Data Synaesthetics. Cambridge, Kettle’s Yard. Bower, Jim and Andrew Roberts (revised by Leonard Will) (2001) Developments in museum and cultural heritage information standards. Paris: The International Committee for Documentation of the International Council of Museums (ICOM-CIDOC), Le Comité international pour la documentation du Conseil international des musées (ICOM-CIDOC). (http://www.willpowerinfo.myby.co.uk/cidoc/stand1.htm). Bowker, Geoff and Star, Leigh (1999) Sorting Things Out: Classification and Practice. Cambridge, MA: MIT Press. Bravo M. T. (2002) The geography of an empire licensed by providence: essay review of Nature's government: science, imperial Britain, and the 'improvement' of the world. Annals of Science 59(4):413-418. Bravo, M. T. and Sörlin, S. (2002) Narrative and practice. in M.T. Bravo & S. Sörlin (eds.) Narrating the Arctic: a cultural history of Nordic scientific practices, Nantucket, Mass: Science History Publications, 3-32. Brown, John Seely and Paul Duguid (2000) The social life of information. Boston : Harvard Business School Press. Castells, M. (1997). The power of Identity. Oxford: Blackwell Chenhall, R.G. (1975) Museum Cataloguing in the Computer Age. Nashville, TN: American Association for State and Local History Crofts, Nick, Martin Doerr, Tony Gill, Stephen Stead, Matthew Stiff (2003) Definition of the CIDOC Conceptual Reference Model. Paris: ICOM/CIDOC CRM Special Interest Group. Doerr, Martin & Nicholas Crofts (1999) Electronic Esperanto: The Role of the Object Oriented CIDOC Reference Model. Washington, DC: Proc. of the ICHIM'99. (http://cidoc.ics.forth.gr/technical_papers.html). DRAFT: Not to be quoted without permission of the Authors (December 2005) 28 Doughty, P. S. (1977) Towards a Documentation for British Museums. Museums Journal, 77(2): 133 Duval, Erik, Wayne Hodgins, Stuart Sutton and Stuart L. Weibel (2002) Metadata Principles and Practicalities. D-Lib Magazine, 8(4). Retrieved: 15 Oct 2005. (http://www.dlib.org/dlib/april02/weibel/04weibel.html) Eco, Umberto (1997) The Search for the Perfect Language. London: Fontana. Garshol, Lars Marius (2004) Metadata? Thesauri? Taxonomies? Topic Maps!: Making Sense of it All. Ontopia, Consulted: 5 May 2005. (http://www.ontopia.net/topicmaps/materials/tm-vsthesauri.html#N645). Geertz, C. (1978) The bazaar economy: Information and search in peasant marketing. American Economic Review 68(2): 28-32. Gulia, Milena and Barry Wellman. (1998) Net Surfers Don't Ride Alone: Virtual Communities As Communities. in Networks in the Global Village, edited by Barry Wellman. Boulder, CO: Westview Press, Pp. 331-67. Kavanaugh, Andrea L. and Scott J. Patterson (2001). The Impact of Community Computer Networks on Social Capital and Community Involvement. American Behavioral Scientist 45(3): 496-509. Knorr-Cetina, K. (1999). Epistemic cultures: how the sciences make knowledge. Cambridge: Harvard University Press. Kretzmann, John and McKnight, John. (1993) Building Communities from the Inside Out: A Path Toward Finding and Mobilizing a Community's Assets. Chicago, IL: ACTA Publications. Latour, Bruno (1987). Science in Action: How to Follow Scientists and Engineers Through Society. Milton Keynes: Open University Press. Latour, Bruno (1986) Visualization and Cognition: Thinking with Eyes and Hands. Knowledge and Society: Studies in the Sociology of Culture Past and Present. 6: 1-40 Law, John and Annemarie Mol (2003) Situating Technoscience: an Inquiry into Spatialities, Centre for Science Studies, Lancaster University. (http://www.comp.lancs.ac.uk/sociology/papers/Law-Mol-SituatingTechnoscience.pdf) Law, John and Vicky Singleton (2003) Object Lessons. Centre for Science Studies, Lancaster University. (http://www.comp.lancs.ac.uk/sociology/papers/Law-Singleton-ObjectLessons.pdf) Law, John and Vicky Singleton (2003) This is Not an Object. Centre for Science Studies, Lancaster University. (http://www.comp.lancs.ac.uk/sociology/papers/Law-Singleton-This-isNot-an-Object.pdf) Marcum , Deanna B. (2002) Automating the Library: The Council on Library Resources. IEEE Annals of the History of Computing, July-Sept.: 2-13. DRAFT: Not to be quoted without permission of the Authors (December 2005) 29 MDA Development Committee (1982) The Future Development of the Museum Documentation Association. Museums Journal, 82(2): 71-76. Michaels, Eric (1986) The Aboriginal invention of television in Central Australia, 1982-1986. Canberra: Australian Institute of Aboriginal Studies. Michaels, Eric (1990) A model of teleported texts (with reference to aboriginal television). Continuum, 3(2), 8-31. Mitra, A. (2004) Voice of the marginalized on the Web. Journal of Communication, 54(3). Moschovitis, Christos J.P., Hilary Poole, Tami Schuyler and Theresa M. Senft (1999) History of the Internet: A Chronology, 1843 to 2001 and Beyond. Santa Barbara, CA: ABC-CLIO. Neufeld, M.L. and M.Cornog (1986) Database History: From Dinosaurs to Compact Discs. J. Am. Soc. for Information Science, 37(4): 183-190. Novak, J. D., & D. B. Gowin. (1984). Learning How to Learn. Cambridge: Cambridge University Press. Pilgrim, Mark (December 18, 2002) What is RSS? Retrieved: 15 Oct 2005. (http://www.xml.com/pub/a/2002/12/18/dive-into-xml.html) Rheingold, Howard (1993). The Virtual Community: Homesteading on the Electronic Frontier. Reading, MA.: Addison & Wesley Co. Shapiro, Michael J., G. Matthew Bonham and Daniel Heradstveit (1988) A Discursive Practices Approach to Collective Decision-Making. International Studies Quarterly, 32(4): 397-419. Shirky, Clay (2002) Communities, Audiences, and Scale. In Clay Shirky, Writings About the Internet: Networks, Economics, and Culture. Written: 6 April 2002. Retrieved: 6 May 2005. (http://shirky.com/writings/community_scale.html) Smith, Marc and Peter Kollock (1999) Communities in Cyberspace: Perspectives on New Forms of Social Organization. London, Routledge Press. Srinivasan, R. (2004) Reconstituting the Urban through community-articulated digital environments. Journal of Urban Technology, 11(2): 93-111. Srinivasan, R. (2005). Weaving Spatial, Digital and Ethnographic Processes in Community-Driven Media Design. Doctoral Dissertation, Graduate School of Design. Harvard University. Srinivasan, R., and Huang, Jeffrey (2005) Fluid Ontologies for Digital Museums. Journal for Digital Libraries, 5(3): 193-204. Srinivasan, R. (in press) Where Community Voice and Information Society Intersect. The Information Society, [under review]. Star, Susan Leigh (1995) The Cultures of Computing. Oxford: Blackwell. Strathern, Marilyn (2002) Abstraction and Decontextualization: An Anthropological Comment. In Steve Woolgar (ed.) Virtual Society: Technology, Cyberbole, Reality. Oxford: Oxford University Press. Turnbull, David (1991) Local Knowledge and 'Absolute Standards': A Reply to Daly. DRAFT: Not to be quoted without permission of the Authors (December 2005) 30 Social Studies of Science, 21(3): 571-573. Turnbull, David (2003) Assemblage and Diversity: Working with Incommensurability: Emergent Knowledge, Narrativity, Performativity, Mobility and Synergy. Paper presented at the AAHPSSS, Melbourne. Wiegand, Wayne A. (1996) Irrepressible reformer: a biography of Melvil Dewey. Chicago: American Library Association. Acknowledgements The authors would like to thank David Turnbull and Geoffrey Bowker in particular for their collaboration and inspiration. We would also like to thank Brian Wynne, Claire Waterhouse and Steve Woolgar for their support and guidance of this research. i ii Yahoo’s classifications are the most well known example of an index of the Web. The Directors of ED2 are: Robin Boast (University of Cambridge, UK), Michael Bravo (University of Cambridge, UK), Geofrey Bowker (University of California Santa Clara, USA), Ramesh Srinivasan (University of California Los Angeles, USA), and David Turnbull (Deakin University, Australia). DRAFT: Not to be quoted without permission of the Authors (December 2005) 31