Skip to main content

Museums and the Web

An annual conference exploring the social, cultural, design, technological, economic, and organizational issues of culture, science and heritage on-line.

Tag! You're It! What Value Do Folksonomies Bring To The Online Museum Collection?

Susan Cairns, The University of Newcastle, Australia

Abstract

As growing numbers of museums invite users to digitally tag online collection objects with key words, it is timely to examine the value that this practice can bring to the sector. These tags in their aggregate are known as ëfolksonomiesí and are broad user-derived classifications that are, at their heart, a form of communication (Huang & Chuang, 2009). As a collaborative process with no hierarchical structure (Cameron & Mengler, 2009), folksonomies stand in opposition to traditional museum taxonomies, and present museums with new opportunities for gaining insight into, and interacting with, audiences. Further, through recording evolving language and tag behaviour, folksonomies have the potential to become autonomously important to the museum in its study of people and written historical documentation. This conceptual paper will present an exploratory examination of the value that folksonomies provide to the museum and gallery sector in the digital age. It will demonstrate that incorporating folksonomies with traditional taxonomic museum classification systems provides the museum with an important new way through which to maintain and grow its relevance in the digital age.

Keywords: folksonomies, museum documentation, social media, tagging, taxonomy

1. Introduction

Recent analysis by the Internet monitoring firm Netcraft places the number of autonomous Internet sites available online at over 210 million (Netcraft, 2010). According to predictions by Guo-Qing Zhang et al. (2008), this number is set to grow exponentially, doubling approximately every 5.32 years. With this growth, the Internet is evolving from being a "collection of hyperlinked documents to a hyperlinked Web of Data," (Ding et al., 2009) and concurrently, is moving towards what has been coined 'Web2.0' (also known as the 'Social Web' (Gruber, 2008)). Within this philosophical evolution of the Internet, the Web serves as a platform; harnesses collective intelligence; features database management as a core competency, and encourages and enables rich user experiences (O'Reilly, 2005). Ultimately, the Social Web is a hub of interaction, where the aggregate contributions of many individual users create value (Gruber, 2008). This collaborative approach to information stands in opposition to the formal top-down classifications of objects by museums, and opens the door to new possibilities for the sector in understanding the semantic and informational requirements of their users.

The Internet's evolution into a hyperlinked web of data has accelerated during the past seven years, as the digital world matures into Web2.0. The Web2.0 concept first emerged in a discussion of changing Web design and aesthetics by Darcy DiNucci (1999) in Print. She wrote that:

the first glimmerings of Web2.0 are now beginning to appear, and we can start to see just how that embryo might develop... The Web will be understood, not as screenfuls of text and graphics but as a transport mechanism, the ether through which interactivity happens.

Although it was not until 2003 that the term found its place in mainstream discussion of the changing direction of the Internet, even in this first use of the terminology, the notion that the Web would become a hub for interaction was central to the concept of Web2.0. It is these socially focused principles of Web2.0 that are redefining the interaction between a growing number of museums, their digital collections, and their publics.

Digital object tagging is proving to be a one of the most popular ways of communicating and curating data online (Digital Curation Centre, 2007), providing an informal classification system for the growing mass of data available on the World Wide Web. These tags were coined 'folksonomies' by Thomas Vander Wal in 2004 as a way to set apart and identify this type of bottom-up classification system (Vander Wal, 2007). Social in nature (Huang & Chuang, 2009), folksonomies sit in complement to traditional taxonomic classification schemes, and were first used by social networking sites such as Flickr and YouTube. Folksonomies provide all users of a Web resource with the opportunity to contribute to, or alter, the digital tags linked to the associated online Web object. These tags then become metadata, or data about data.

This paper will present an exploratory examination of the value that folksonomies provide to the museum and gallery sector, with specific focus on the use of language in the construction of meaning in the museum sector. The growing incorporation of folksonomies and other Web2.0 technologies into museum collections practice is a very current and important issue for the sector to address and understand. Changes in the technological landscape frequently occur more quickly than the associated theoretical analysis. Therefore the investigation that this dissertation presents into the value that folksonomies provide the online museum collection gives important and timely context for this practice. This research offers an original perspective on these issues.

This paper will not speak to the technical aspects of incorporating folksonomies with online museum collections, nor will it address the wider implications of online collections practice, such as the erasure of temporal distance that occurs when uploading images of museum objects to the Internet. These issues are beyond the scope of this current investigation.

2. The social value of folksonomies

Taxonomy is the science of the classification and identification of living things and objects (Sokal, 1974). It arises out of the human need to perceive similarities and differences in stimuli, to create groups or sets based on these relationships, and to articulate such differences linguistically (Raven, Berlin & Breedlove, 1971). Robert R. Sokal (1974) says that:

the paramount purpose of a classification is to describe the structure and relationship of the constituent objects to each other and to similar objects, and to simplify these relationships in such a way that general statements can be made about classes of objects.

A classification system is thus the system of rules or regulations by which objects are grouped together into defined relationships, and at their most effective, allow for the easy manipulation and retrieval of information (Sokal, 1974).

Traditional taxonomic structures take the form of a tree-like system, with hierarchical lines or branches of information spreading out, from trunk to branches, to twigs and leaves (Weinberger, 2005). This structure, which was first enunciated by Carolus Linneaus in his 1785 publication Systema Naturae, evolved out of the European Enlightenment's desire for rationalisation and systematic linguistic reform (Spary, 2003). Linneaus' top-down hierarchical system sought to universalise and make sense of the natural world, by providing guidelines through which all known and yet to be discovered animals, minerals and plants could be grouped. The binary structure of this system, in which the properties of an individual item are inherited from each of its parent categories (Weinberger, 2005), has remained the dominant structure for classification systems since that time.

An alternate way of presenting these tree-like classification systems was created in the early 1930s by Shiyali Ranganathan, who proposed a faceted classification system as a more dynamic way of creating and presenting taxonomies (Weinberger, 2005). This system, while still hierarchical in nature, applies pre-defined parameters to the sorting of objects without deciding in advance how branches are arranged. Using the metaphor of a Mechano set, Ranganathan explained that his classification system could be constructed in numerous ways using standardised pieces and connecting symbols (Beghtol, 2008). This is an example of an analytic-synthetic method of classification, which requires document analysis to determine into which particular facets of classification an object or article fits (Beghtol, 2008). In a museum collection this could be manifest by allowing for a collection search by a creator facet, a materials facet or a date facet. These faceted systems of classifications are currently utilised online by museum collections websites such as Sydney's Powerhouse Museum online collection (http://www.powerhousemuseum.com/collection/database/browsecategories.php), where a user searching for 'lace collars' can locate them as a second level search under at least three distinct top-level searches (Lace, Clothing and Dress, or Textiles). Faceted classification systems are a dynamic hierarchical system that provides museum collection users with a variety of paths to reach a desired museum object.

In opposition to the traditional tree structure of taxonomies, David Weinberger (2005) proposes that folksonomies can be interpreted as being a pile of leaves, composed of non-hierarchical clusters of information. Unlike ordered formal systems of classification, folksonomies are messy, made of overlapping and ambiguous information. Further, folksonomies are not designed in advance, nor owned and controlled centrally. Instead, they are organic bottom-up collaborative systems of categorisation and communication, bearing resemblance to Gilles Deleuze and Felix Guattari's concept of a rhizome. As Deleuze and Guattari (1987) spell out:

unlike trees or their roots, the rhizome connects any point to any other point, and its traits are not necessarily linked to traits of the same nature; it brings into play very different regimes of signs, and even non-signs.

While logical and linear information retrieval is a key characteristic of formal taxonomies, whose top-down classifications inform the properties of those below, folksonomies have evolved alongside Internet search platforms such as Google, and therefore they do not rely on linearity and hierarchy for information retrieval. Folksonomies are inherently social, and much of the value that museums can gain from their integration with online collection databases is through increased interaction with, and understanding of, audience.

According to classification theory, taxonomies can be either monothetic, in which uniform properties are used to define differences between sets or classes along strict lines, or polythetic, meaning that the objects grouped together into a classification set share a large proportion of their properties, but are not necessarily similar in any one property (Sokal, 1974). Major museum classification schemes such as the Social History and Industrial Classification System utilise a polythetic and faceted approach in structuring their taxonomies. The primary motivation for the Social History and Industrial Classification system is to make relationships between museum objects and groups of objects based on their social contexts and "according to the sphere of human activity with which they are primarily associated" (http://www.holm.demon.co.uk/shic/shicint.htm). The polythetic nature of this museum classification process, and others like it, means that the choices made by curators in classification can have significant implications for the contextual reading of the object, and impact upon the ease with which collection objects can be located.

Within the museum, the curator imparts meaning onto museum objects from the moment of accessioning, through classification and categorisation, and by creating exhibition and collection storylines (Dallas, 2007). The decisions made by the curator in acquisition, classification and display are political and ideological and key in the construction of meaning and context. As Terry Cook (2001) writes, facts:

cannot be separated from their ongoing and past interpretation, not author from subject or audience, nor author from authoring, nor author from context. Nothing is neutral. Nothing is objective. Everything is shaped, presented, re-presented, symbolized, signified, signed, constructed.

In the process of collecting, museum objects are selected and brought together from diverse backgrounds and placed into new frames of meaning. Taxonomically different objects are grouped together in new ways and ordered to suit the collecting institution's own collection strategies (Alberti, 2005). Museum objects are polysemic, open to a number of different - and possibly contradictory (Hooper-Greenhill, 2000) - interpretations, and many objects can be placed into multiple and intersecting classifications (Alberti, 2005). The classification of museum objects within a collection is made based on institutional decisions regarding significance, and the meanings of these objects when collected come "already calculated, not because of norms embedded in the language but because language is always perceived, from the very first, within a structure of norms" (Fish, 1980). The collecting institution places objects into a classification system based on a pre-determined and structured vocabulary.

While museum objects are collected with a particular institutional reading in mind, visitors to museums and museum websites bring with them their own interpretations. Visitors, and website users, are "autonomous agents with their own agendas" (Alberti, 2005). They bring their own experiences and understandings to the interpretation of museum objects, reading the objects within the context of their own pre-stored experiences. According to Hermeneutic Theory, understanding is found neither wholly in the object, nor in the viewer, but in the dialogue between the two (Hooper-Greenhill, 2000). In the process of creating folksonomies, users are able to tag museum objects with words that reflect their own experiences and understandings of the object. While traditional museum classifications are binary, with objects that either do or do not fit into a category, these folk-created classifications reflect the polysemic nature of objects, and the multitude of meanings that they may carry.

3. Folksonomies and museum documentation

One of the largest research groups currently studying the incorporation of folksonomies into online museum collections is the steve.museum, based at the Metropolitan Museum of Art and the Indianapolis Museum of Art in the USA, and including the Cleveland Museum of Art, the Denver Art Museum, the Guggenheim Museum, the Minneapolis Institute of Arts, the Rubin Museum of Art in Manhattan and the San Francisco Museum of Modern Art. The fully Internet-based steve.museum was formed out of a desire to understand the "semantic gap that separated museums' formal descriptions of works - usually created by art historians or other specialists - and the vernacular language used by the general public for searching" (Steve: the Museum Social Tagging Project, http://steve.museum/index.php?option=com_content&task=blogsection&id=1&Itemid=2). Its principal researcher Jennifer Trant is leading much of the discussion about the use of folksonomies in museum classification today. Trant's early studies through the steve.museum sought to "validate the proposition that social tagging could add value to existing museum documentation" (Trant, 2006). As a result of these studies, Trant demonstrated that museums could utilise folksonomies to open collections up to unexpected and more personal meanings, and that the content elements found in folksonomies were missing from formal museum documentation. She further discovered that non-specialists could provide new access points to works of art online, and augment the formal descriptions provided by museum professionals. These new meanings reflect the dialogic relationships that website users have with online museum objects, and provide the museum with valuable insight into the multiple readings that users have of their objects.

By actively participating in the act of tagging digital museum objects, users are able to connect with works of art and other objects in a more personal and direct way (Trant, 2006). Museum objects are the catalyst for a series of relationships, being person-to-person, object-to-object and object-to-person (Alberti, 2005). The museum contains and fosters these relationships by bringing objects, and audiences, together in interaction. In the online museum, this interaction takes places in the digital collection space. Tagging practices can vocalise these dialogues between audience, object and museum, and invite active involvement in the museum process. Through this type of reciprocal interpretation, museum objects are opened up to new types of interactions (Bearman & Trant, 2005). Folksonomies provide museums with a new platform for dynamic audience participation, and create an environment for active engagement.

In a 2005 experiment of digital tag language run by the Metropolitan Museum of Art, it was found that 80% of tags applied to museum objects utilised different terms from those found in official museum classifications (O'Connell, 2007). Where formal classification systems are used to identify specific relationships between items, the associations made in folksonomies are more fluid. The approach to this type of classification is more natural, intuitive and less reliant on understanding specific predefined categories (Ding et al., 2009). While traditional museum documentation is reliant on an understanding of predetermined language and categorisation approaches, tags applied by non-experts can accurately reflect a user's immediate responses to an object. By inviting museum website users to add their own tags to online museum objects in complement to formal taxonomies, new linguistic entry points can be made. These intuitive and responsive tags can provide new access to users who are untrained in traditional museum terminology.

In creating tags, users include information that they personally think is significant about a museum object, and also that seems important for the retrieval process (Ménard, 2007). The language used tends to be flexible, descriptive and responsive, with neologisms appearing quickly in reflection of real time changes in language trends. Folksonomies are imprecise but responsive, while formal taxonomies are slow to change and symptomatic of the systems that created them. The imprecision of folksonomies is often held up as one of the great inadequacies of such classification systems, in that they are inexact and informal (Ding et al., 2009). However, this reflexive and reactive language reflects the real world nature of communication and knowledge sharing, and the need for interactivity to create understanding (Saab, 2010). By analysing and understanding the semantic gap between the formal classifications used in museum taxonomies and the responsive and informal language used by digital taggers, those working in the museum sector can gain insight into possible barriers to understanding in their audiences. Through this, museum educators and curators can cater more specifically to the language needs of users and audiences when creating museum documentation.

Further, through archiving folksonomies as specific to a particular period of time, the sector can track changing language trends and public reactions to museum objects and groups of objects, and record how these evolve over time. Sharon McDonald, cited in J.M.M. Alberti, calls this the "repertoire of prevalent interpretations" (Macdonald, 2002 in Alberti, 2005).. By looking at digital tag terms applied to objects and utilised in searches, museums can gain understanding into the current concerns of their audiences and the surrounding culture. The reflexive nature of folksonomies gives insight into language, into the trends of terminology and the transient and consistent aspects of a culture.

The opportunity to track historical public contributions to museum collection documentation provides museums with a possible new avenue for historical document analysis. As Diderot wrote:

The language of a people gives us its vocabulary, and its vocabulary is a sufficiently faithful and authoritative record of all the knowledge of that people; simply by comparing the different states of a nation's vocabulary at different times one could form an idea of its progress.

(Diderot, in Foucault, 2002)

Through recording evolving language and tag behaviour, folksonomies have the potential to become autonomously important to the museum in its study of people and historical documentation.

Because no specific skills are required to apply tags to museum objects, folksonomies are inclusive by their very nature. As such, they can provide members and groups who are traditionally left out of the descriptive process with a voice, and an opportunity to reflect their own experiences of an object (Jorgensen, 2004). People from different cultures, nationalities and ethnicities, and from a variety of spheres of influence, can all contribute to a museum's collection of tags (Saab, 2010), allowing for a minority discourse that may be otherwise removed from the museum experience. Through folksonomies, museums may be able to give a broader, more diverse and socially inclusive audience a voice. This is particularly valuable in light of the perceived differences between institutional and personal memories of events, objects and histories. Folksonomies provide the museum sector with new opportunities for cohesive and comprehensive storytelling, which directly records the voice of minority discourse.

However, although multiple users with different knowledge and experiences are empowered to tag online museum objects, folksonomies are only created by the users of a Web resource. These users are likely to be limited to interested parties and those with access. David J. Saab (2010), when discussing the ontology of tags, argues that the:

collective tags of a folksonomy will certainly reflect the dominant cultural schemas of a broad population, but the assumption that collective tags represent a shared conceptualization, interferes with discerning minority cultures, whose schemas may overlap with but are not necessarily entirely consistent with those of the dominant cultural group.

While folksonomies are theoretically open to contribution from all users who have access to the Internet and to the museum collection website, museum users who have traditionally been left out of the descriptive process may need to be encouraged to contribute in this new platform. Despite this, the open and inclusive nature of folksonomies holds promise for creating new ways through which museums can interact with the traditionally unempowered.

It is this same notion of inclusion, through opening the online collection to untrained and unskilled users, that runs counter to the hitherto required levels of expertise and training of museum professionals through which the sector has maintained its aura of authority. Fiona Cameron and Sarah Mengler (2009) suggest that the use of folksonomies and other digital interactions can provide museums with the means to grow their complexity practices, through which the fluidity of cultural interactions produce new levels of interactivity and connection. This is at odds with the fixed and abstracted aspects of collections classification practice, but bridges the gap between these systematic approaches to object classification, and the fluid and responsive interactions made available by inviting public response to digital museum objects. They further propose an acceptance of the contradictions inherent in the differences between the two types of classification systems, to:

hybridize museum order, making the boundaries of classification and description, contextualization and significance permeable; and to acknowledge that there will be a degree of structural instability in the system of collections information and significance previously denied. (Cameron & Mengler 2009)

Corinne Jorgensen (2004) has envisaged the new levels of complexity in recontextualising and classifying online museum collections provided by folksonomies as a creative act. Jorgensen sees possibilities within the sector for the creation of new knowledge through the manipulation of museum classification systems. Through folksonomies, the museum collection becomes the locus of a new arena of participation in the sector, beyond the tightly controlled curatorship of the traditional museum. Digital object tagging may open the collection up to previously undiscovered relationships, and become part of the creative process through encouraging interaction with visitors to the online museum.

4. Who is creating folksonomies?

In order for the value of folksonomies to the sector to be maximised, it is important to understand who is creating folksonomies, and what their tag behaviour is. Christian Korner, Dominic Benz et al. (2010) describe two distinct types of users, each with different tagging patterns. They describe both 'categorizers', who apply a small set of tags to objects as a replacement for hierarchical classification schemes, and 'describers', who use more descriptive and freely associated language. In research investigating the pragmatics of tagging on tag semantics, it was discovered that the emergent semantics of tags in folksonomies were influenced by the tagging practices of the individual, and that a small number of users contribute the majority of terms to the 'collaborative verbosity'. Further, they found that it is possible for a single tag term to be applied by both a categorizer and a describer, each utilising the term for a different purpose (as part of a pre-determined category, or as a descriptive annotation). Finally, it was discovered that:

in general, the collaborative verbosity of describers provides a better basis for harvesting meaningful tag semantics. However, this observation comes with a limitation: The most verbose taggers (in our case mostly spammers) negatively influenced semantic accuracy. (Korner, Benz et al., 2010)

The presence of spam, or electronic bulk messages, is a major problem confronting websites that utilise Web2.0 platforms. In global folksonomies, spammers generate around 40% of all tags, with the primary intention of manipulating search engines towards their own websites (Korner, Benz et al., 2010). This leads to a significant increase in semantic noise, and undermines some of the value of folksonomies for the museum sector. The need for policing and evaluating tags for usefulness runs counter to the democratic ideal of collective intelligence in Web2.0. While manually inspecting and removing tags that are obviously spam will increase the accuracy and relevance of folksonomies on online museum collections, this process requires the time, effort and expertise of a member working within the museum.

The very act of putting a museum collection online has further implications for museum workloads, from the increased time required for digitising and uploading images and metadata through to dealing with enquiries arising from the collection's increased visibility. As Sebastian Chan, Head of Digital, Social & Emerging Technologies at Sydney's Powerhouse Museum, highlights, the digital exposure of museum objects online has led to an increased number of enquiries to the museum regarding the collection (Chan, 2007). These enquiries are increasingly detailed, and require additional research by curatorial staff. Some of these searches have yielded new information about collection items, including contextual information or corrections to current records. However, the time required to look into enquiries of this nature can impact negatively upon museum staff members whose focus may need to be on other things, such as exhibitions.

An antidote to this dilution of museum staff's time may be provided through folksonomies themselves, and the associated formation of interpretive communities or 'communities of passion'. Interpretive communities form around a shared interest and create normative and shared languages and perspectives for creating meaning in the writing of texts (Fish, 1980). An example of this comes from the Powerhouse Museum, where the most viewed museum object online in 2006 (within the first six months of the Museum's collection being made available online via the Powerhouse Museum website) was a dress worn by Australian entertainer Delta Goodrem (Chan, 2007). This object had never been displayed in the museum, but through Google searches and links displayed on Goodrem fan sites, became highly visible online. The tags applied to this object have evolved through the item's digital lifecycle, with certain tags, such as those that included the names of other celebrities, edited over time by users. This internal policing of tags occurs as a community - in this case likely to be composed of fans of Goodrem or those interested in fashion - finds a common or shared language that is indicative of the community perspective. The effectiveness and value of museum folksonomies can be increased as their creators form and become involved in communities of passion.

Over time, interpretive communities develop a mutual language and body of common knowledge through which to write about objects, as seen by the evolving tags applied to the Delta Goodrem dress. According to Stanley Fish (1980), meanings are a:

function of precisely the public and constituting norms (of language and understanding). These norms are not embedded in the language (where they may be read out by anyone with sufficiently clear, that is unbiased, eyes) but inhere in an institutional structure within which one hears utterances as already organised with reference to certain assumed purposes and goals.

Museum taxonomies are an example of fully developed discourse created by an interpretive community composed of museum experts and those working within the field. Folksonomies, however, provide opportunities for new interpretive communities with new language requirements to form. Rick Borovoy, Fred Martin and Mitchell Resnick (1999) write that:

edges of interpretive communities occur both when a newly forming community struggles to establish some common ground, and when a newcomer attempts to acquire enough of the mutual knowledge of an established community to being to make appropriate sense of its discourse.

As communities develop around online museum objects and collections, we can expect to see a growing internal and mutually agreed-upon language, improving the usefulness of online museum collection folksonomies over time.

In the process of defining the borders of accepted and acceptable tag terms, it is likely that conflict will arise between some members of new and growing interpretive communities (Landbeck, 2007). This conflict may occur in a number of different ways, including during the process of establishing the boundaries of accepted tag language, or in establishing and publishing facts about a controversial object (Landbeck, 2007). Museums may need to create and apply conflict resolution techniques. These conflict resolution techniques may require different approaches in different institutions, depending on whether the integration of folksonomies into the online collection is viewed as important for their social indexing or community building properties. However, strategies should include defining who will oversee the resolution of conflicts, and under what circumstances, and will ideally provide a clear guide for museum experts in dealing with circumstances of conflict within the digital community.

One of the likely consequences of the development of communities built around the online museum collection will be the growing stabilisation of the mutually agreed upon language utilised in tags. Although one of the main arguments for incorporating folksonomies into websites is that they are responsive and flexible, folksonomies can grow into a stable resource given enough time and contributions (Mikroyannidis, 2007). Stability in this case means "the tagging eventually settles to a group of tags that describe the resource well and where new users mostly reinforce already present tags in the same frequency as in the stable distribution" (Halpin, Robu & Shepherd, 2007). This has led to observations by Harry Halpin, Valentin Robu and Hana Shepherd (2007) that folksonomies have a number of features that are indicative of their nature as a complex system, including the lack of central coordination, non-linear dynamics and large number of users. These complex systems produce a type over time distribution called a power law. Power laws produced in this fashion "can often be scale-free, such that regardless of how large the system grows, the shape of the distribution remains the same, and thus stable" (Halpin, Robu & Shepherd, 2007). At this time, folksonomies on many museum websites are still new enough that a single user can skew results, in the event that only a small number of contributors have added to the tags for an online museum object (Hotho et al., 2006). However, the growing stability of folksonomies as a resource with increased contributions over time indicates that the sector may be able to rely more heavily on folksonomies to provide meaningful search terms for improved collection navigation and information retrieval.

As tag terms applied to online museum objects become more stable, it is likely that museum users will be able to increase their trust in, and reliance on, folksonomies as a source of meaningful information. The current research being conducted through the steve.museum, in partnership with the University of Maryland and the Indianapolis Museum of Art is focused on the relationship between folksonomies and trust, and is entitled "T3: Text, Terms and Trust". T3 "is a research effort that combines text mining, social tagging, and trust inferencing techniques to enrich metadata and personalize retrieval" (http://steve.museum/index.php?option=com_content&task=blogsection&id=1&Itemid=2). The relationship between tagging and trust is a key area of interest in the growing use of folksonomies by the museum sector. Recent findings by the Institute for Museum and Library Studies in Washington DC indicate that "libraries and museums are trusted far more than other sources of information including government, commercial and private individual websites" (Griffiths & King, 2008). Information found elsewhere on the Web was rated as being less trustworthy than that found on museum websites. As such, maintaining this trust is a clear priority for the museum, whose central aims include disseminating information about collections and the meanings of the objects held within them.

The cumulative effect of the stabilisation of folksonomies along the lines of power laws is in some ways inherently problematic. As the number of contributors to a folksonomy rises, building a consistent and accepted language of tags, there will be a growing homogenisation of tags, and the opportunity for minority discourse to provide significant input may diminish. David J. Saab suggests this effect is compounded when "we begin aggregating tags into tagclouds and broad folksonomies associated with particular perspectives - cultural identities and schemas" (Saab, 2010). While folksonomies can open the museum collection up to new interpretations that honour the polysemic nature of objects, there remains a risk that acceding to majority rule, as is the case with increasingly stable terminology, will cancel out the potential benefits of the inclusive nature of folksonomies to those traditionally removed from the descriptive process. The collective action component of tags, as users try to reach a general consensus of meaning and accepted terminology (Spiteri, 2006), is in some ways contradictory to the open and inclusive platform that folksonomies are built upon at their base level. As Saab iterates, "when making meaning of a particular phenomenon, individuals will rely upon the cognitive and cultural schemas that are integral parts of their salient, contextualized identities" (Saab, 2010). This complicates, although does not necessarily reduce, the value that folksonomies can bring to the online museum user.

5. Folksonomies, language and collected intelligence

Like all unregulated vocabularies, folksonomies have a number of ambiguities inherent to their nature, as different users apply tags to museum objects in different ways (Spiteri, 2006). Terms applied to online museum objects have no synonym control, and so one item can be described by multiple words with the same meaning. Tags can take both singular and plural forms, they can be polysemous (such as the term 'apple', which can be applied both to the fruit and the computer manufacturer), they can be abbreviated, or they can simply be applied incorrectly. Further, the tags applied to online museum objects might be too specialist for some users, and too general for others. These word variations may be meaningful to the individual who applied them to the object, but can further complicate the search process for the searcher, who must put in more effort and be more creative in thinking of all possible variations under which an item they are seeking could be listed (Ding et al., 2009).

The ambiguity in tagging language is caused in part because a single tag when used in isolation is not very semantic (Saab, 2010). As Saab discusses:

A word isolated from the entity it was intended to describe and from the person who created it can mean or refer to many things, and many people may interpret the same tag differently based on their personal histories. In order to make sense of a semantic tag, it is important to understand the perspective from which it is offered. (Saab, 2010)

People create tags based on their own unique experiences, to make sense of a museum object. Tags therefore make the most sense when they are read together with the object that they are being applied to, rather than in isolation (Saab, 2010). Although tags are useful as search terms to encourage external and internal discovery of online museum objects, museums can derive the most value from them when they are located in digital context with the item to which they are applied.

The linguistic limitations of folksonomies are symptomatic of the broader limitations of the Social Web, in which "current prevailing keyword-based Web information search is neither expressive nor accurate enough to support semantics-rich personalized queries" (Sun, Zhuge & Li, 2009). However, groups such as the World Wide Web Consortium (W3C) believe that the Internet is now heading towards what is known as the 'Semantic Web' (W3C, 1994-2010). In the Semantic Web, it is anticipated that human contributions to knowledge will be better integrated with online systems, enabling "computation and inference over the collected information, leading to answers, discoveries, or other results that are not found in the human contributions" (Gruber, 2008). These changes, should they come to fruition, hold promise for incorporating folksonomies into the construction of knowledge. The W3C website explains that the Semantic Web will provide "a common framework that allows data to be shared and reused across application, enterprise, and community boundaries" (W3C, 1994-2010).

Tom Gruber, a researcher into collective knowledge systems, believes that ultimately the Social Web will meet the Semantic Web. He explains:

The Social Web is an ecosystem of participation, where value is created by the aggregation of many individual user contributions. The Semantic Web is an ecosystem of data, where value is created by the integration of structured data from many sources. (Gruber, 2008 )

Gruber further proposes that the possibilities offered by folksonomies and the Social Web offer unmatched potential for knowledge sharing and development than at any previous time in human existence. As he writes:

With the rise of the Social Web, we now have millions of humans offering their knowledge online, which means that the information is stored, searchable, and easily shared. The challenge for the next generation of the Social and Semantic Webs is to find the right match between what is put online and methods for doing useful reasoning with the data. True collective intelligence can emerge if the data collected from all those people is aggregated and recombined to create new knowledge and new ways of learning that individual humans cannot do by themselves.

For Gruber, the current state of Web2.0 is one of collected and not collective intelligence, whereby the value of folksonomies and other user-derived contributions to the Social Web is in their collection and aggregation. As such, the growing involvement of the public in the addition of folksonomies to online museum objects is valuable for increasing our collected intelligence in relation to museum collections.

6. Conclusion

Although folksonomies are useful to the museum sector in their current form and function, it will be their integration into future iterations of the Internet that fully cements their use in online museum collections. The digital tags that are applied to objects in online collections already provide the sector with important insight into user language, and evolving language trends. Digital tags also help the museum sector by growing metadata attached to online museum objects, data which can be used to build both internal and external access to the collection. However, the long-term potential for folksonomies to become more than simple devices for the categorisation, search and retrieval of data will be reliant on the sector's ability to synergise their use with traditional museum classification systems. Ultimately, this could lead to new ways through which to disseminate information about the museum collection.

7. References

Alberti, S. J.M.M. (2005). "Objects and the Museum." Isis 96, 559-571.

Bearman, D. & J. Trant (2005). ''Social terminology enhancement through vernacular engagement: Exploring collaborative annotation to encourage interaction with museum collections." D-Lib Magazine, 11. Available at: http://www.dlib.org/dlib/september05/bearman/09bearman.html Accessed online 16 July 2010.

Beghtol, C. (2008). "From the Universe of Knowledge to the Universe of Concepts: The Structural Revolution in Classification for Information Retrieval." Axiomathes, 18, 131-144.

Borovoy, R., F. Martin, and M. Resnick. (1999). "Revealing Common Ground: Augmentation on the Edges of Interpretive Communities." Paper presented at the Conference on Computer Supported Collaborative Learning, Stanford University, Palo Alto, California.

Cameron, F. & S. Mengler (2009). "Complexity, Transdisciplinarity and Museum Collections Documentation: Emergent Metaphors for a Complex World." Journal of Material Culture, 14, 189-218.

Chan, S. (2007). "Tagging and Searching - Serendipity and museum collection databases." In J. Trant and D. Bearman (eds.). Museums and the Web 2007: Proceeding, Toronto: Archives & Museum Informatics, published March 1, 2007 http://www.archimuse.com/mw2007/papers/chan/chan.html Accessed online June 22, 2010.

Cook, T. (2001). "Archival Science and Postmodernism: New Formulations for Old Concepts." Archival Science 1, 3-24.

Dallas, C. (2007). "An Agency-Oriented Approach to Digital Curation Theory and Practice." Paper presented at the International Cultural Heritage Informatics Meeting (ICHIM07), Toronto.Available http://www.archimuse.com/ichim07/papers/dallas/dallas.html

Deleuze, G. & F. Guattari (1987, sixth reprint 1996). A Thousand Plateaus: Capitalism & Schizophrenia. University of Minnesota Press, Minneapolis and London.

Diderot, article on 'Encyclopédie' in the Encyclopédie, t. V, 637, cited in Foucault, M. (2002) The Order of Things. London and New York: Routledge Classics.

Digital Curation Centre (2007) "About the DCC: What is Digital Curation?" Published: 31 May 2005, Modified: 26 April 2007, http://dcc.ac.uk/about/what/ Accessed online December 27 2009.

Ding, Y., E.K. Jacob, Z. Zhang, S. Foo, E. Yan, N.L.George & L. Guo (2009). "Perspectives on Social Tagging." Journal of the American Society for Information Science and Technology, 60, 2388-2401.

DiNucci, D. (1999). "Fragmented Future." Print 53, 32, 221-222.

Fish, S. (1980). Is There a Text in This Class? The Authority of Interpretive Communities. Harvard University Press, Cambridge, Mass. & London.

Foucault, M. (2002) The Order of Things. London and New York: Routledge Classics.

Griffiths, J. & D.W. King (2008). "Interconnections: The IMLS National Study on the Use of Libraries, Museums and the Internet - Conclusions." Institute for Museum and Library Studies, 1-27.

Gruber, T. (2008). "Collective knowledge systems: Where the Social Web meets the Semantic Web." The Journal of Web Semantics: Science, Services and Agents on the World Wide Web 6, 4-13.

Halpin, H., V. Robu, & H. Shepherd (2007). "The Complex Dynamics of Collaborative Tagging." Paper presented at the 16th Annual International World Wide Web Conference, May 8-12 2007, Banff, Alberta, Canada.

Hooper-Greenhill, E. (2000). Museums and the Interpretation of Visual Culture. London; New York: Routledge.

Hotho, A., R. Jäschke, C. Schmitz, & G. Stumme (2006). "Information Retrieval in Folksonomies: Search and Ranking." In Y. Sure and J. Domingue (Eds.): ESWC 2006, Lecture Notes in Computer Science 4011, 411-426.

Huang, A.W. & T. Chuang (2009). "Social Tagging, Online Communication, and Peircean Semiotics: A Conceptual Framework." Journal of Information Science 35, 340-357.

Jäschke, R., L. Marinho, A. Hotho, L. Schmidt-Thieme & G. Stumme (2007). "Tag recommendations in Folksonomies." Lecture Notes in Computer Science, 4702, 506-514.

Jorgensen, C. (2004). "Unlocking the Museum: A Manifesto." Journal of the American Society for Information and Technology 55, 462-464.

Korner, C., D. Benz, A. Hotho, M. Strohmaier & G. Stumme (2010). "Stop Thinking, Start Tagging: Tag Semantics Emerge from Collaborative Verbosity." Proceedings from the 19th International Conference on the World Wide Web, 521-530.

Landbeck, C. (2007). "Trouble in Paradise: Conflict Management and Resolution in Social Classification Environments." Bulletin of the American Society for Information Science and Technology, 16-20.

Macdonald, S. (2002). Behind the Scenes at the Science Museum. Oxford/New York: Berg, cited S.J.M.M Alberti (2005). "Objects and the Museum," Isis 96: 559-571.

Ménard, E. (2007). "Image Indexing: How Can I Find A Nice Pair of Italian Shoes?" Bulletin of the American Society for Information Science and Technology, October/November, 21-25.

Mikroyannidis, A. (2007). "Toward a Social Semantic Web." Computer 40, 113-115.

Museums Australia Incorporated Constitution and Rules, with amendments adopted 20 May 2008 http://www.museumsaustralia.org.au/site/pageconst.php Accessed online 2 August 2010.

Netcraft. "Web server survey." http://news.netcraft.com/archives/category/web-server-survey/ Accessed online 14 September 2010.

O'Connell, P.L. (2007). "One picture, 1,000 tags." The New York Times, March 28 2007. http://www.nytimes.com/2007/03/28/arts/artsspecial/28social. Accessed online 30 March 2010.

O'Reilly, T. (2005). "What is Web2.0: Design Patterns and Business Models for the Next Generation of Software." What is Web2.0 - O'Reilly Media. http://oreilly.com/web2/archive/what-is-web-20.html Accessed online 10 July 2010.

Powerhouse Museum Collection Search 2.5, http://www.powerhousemuseum.com/collection/database/browsecategories.php Accessed online 25 July 2010.

Raven, P.H., B. Berlin & D.E. Breedlove (1971). "The Origins of Taxonomy." Science, 174, 1210-1214.

Saab, D.J. (2010). "The Ontology of Tags." Paper presented at iConference 2010, University of Illinois at Urbana-Champaign, February 3-6 2010.

Social History and Industrial Classification. "SHIC Introduction." http://www.holm.demon.co.uk/shic/shicint.htm Accessed online 25 July 2010.

Sokal, R.R. (1974). "Classification: Purposes, Principles, Progress, Prospects." Science, 185, 1115-1123.

Spary E. C. "Linnaeus, Carolus" Encyclopedia of the Enlightenment. A.C. Kors (ed.) (2003) Oxford University Press http://0-www.oxfordreference.com.library.newcastle.edu.au/views/ENTRY.html?subview=Main&entry=t173.e404. Accessed online 22 July 2010 .

Spiteri, L.F. (2006). "The Use of Folksonomies in Public Library Catalogues." The Serials Librarian, 51, 75-89.

Steve: the Museum Social Tagging Project, http://steve.museum/index.php?option=com_content&task=blogsection&id=1&Itemid=2 Accessed online 8 June 2010.

Sun, X., H. Zhuge & Q. Li. (2009). "A framework for the massive knowledge Web." Concurrency and Computation: Practice and Experience 21: 705-723.

Trant, J. with the participants in the steve.museum project .(2006). "Exploring the potential for social tagging and folksonomy in art museums: Proof of concept." New Review of Hypermedia and Multimedia 12, 83-105.

Vander Wal, T. (2007). Folksonomy. Article posted 2 July 2007. http://vanderwal.net/folksonomy.html Accessed online 4 September 2010.

W3C, W3C Semantic Web Activity (1994-2010) W3C. http://www.w3.org/2001/sw/ Accessed online 3 August 2010.

Weinberger, D. (2005). "Taxonomies to Tags: From Trees to Piles of Leaves." Release1.0, 23, 1-32.

Zhang, G., G. Zhang, Q. Yang, S. Cheng & T. Zhou (2008). "Evolution of the Internet and its cores." New Journal of Physics 10, 123027.

Cite as:

Cairns, S., Tag! You're It! What Value Do Folksonomies Bring To The Online Museum Collection?. In J. Trant and D. Bearman (eds). Museums and the Web 2011: Proceedings. Toronto: Archives & Museum Informatics. Published March 31, 2011. Consulted http://conference.archimuse.com/mw2011/papers/tag_youre_it_what_value_do_folksonomies_bring