Conference Papers

Museums and the Web: An International Conference
Los Angeles, CA, March 16 - 19, 1997

Terry Hemmings
Dave Randall
Dave Francis
Liz Marr
Manchester Metropolitan University

Colin Divall
National Railway Museum, York and The University of York, U.K.

Gaby Porter
The Museum of Science and Industry in Manchester, U.K.

Situated Knowledge and the Virtual Science and Industry Museum: Problems in the social-technical interface.

This paper was produced by the Information in Museums Project, Department of Sociology, Manchester Metropolitan University. UK. URL:


The Museum is a perspicuous site for analysing the complex interplay between social, organisational, cultural and political factors which have relevance to the design and use of 'virtual' technologies. Specifically, the introduction of virtual technologies in museums runs up against the issue of the situated character of information use. Across a number of disciplines (anthropology, sociology, psychology, cognitive science) there is growing recognition of the 'situatedness' of knowledge and its importance for the design and use of technology. This awareness is fostered by the fact that technological developments are often associated with disappointing gains for users. The effective use of technology relies on the degree to which it can be embedded in or congruent with the 'local' practices of museum users. Drawing upon field research in two museums of science and technology, both of which are in the process of introducing virtual technologies and exploring the possibilities of on-line access, findings are presented which suggest that the success of such developments will depend on the extent to which they are informed by detailed understanding of practice- practices that are essentially socially constituted in the activities of museum visitors and museum professionals.


The New Museology recommends that the study of museums and professional work within museums should adopt a greater degree of self awareness, inquiring not just into methods, but into purpose, context and consequence (Vergo 1989) . When this term was coined in 1989 1 most museologists were blissfully unaware of the potential problems and promises generated by Information and Communication Technologies (ICT's). Museums are no longer, if ever they were, innocent pedants engaged in the collection, conservation, classification and display of objects. On the contrary, museum practitioners now recognise they are one among many components in panoply of cultural amenities. Add together determinants that include the external pressure of political and market forces with an increasing internal reflexivity and the result is an institution no longer professionally or intellectually isolated from a commercial array of cultural and leisure sites 2 and no longer certain of its role or identity in the explosion of contested identities, image and meaning (MacDonald and Silverstone, 1996).


In many cases the security of museums will increasingly rely on their position as resource providers and mediator within various educational infrastructures, including the World Wide Web. However pedagogists within museums have by definition always identified their function as being part local educational landscape. In the UK national science museums have a long didactic tradition, providing educational and informational resources for diverse social constituencies. It is, however, only quite recently that funding bodies 3 have become keenly aware of the putative benefits that new technologies might confer upon museum work. Many museums are now beginning to utilise more recent ICT applications, intra and inter organisationally, to develop database and support curatorial and management networks. 'Designed in', telematics and interactive exhibits are extensively incorporated into galleries in order to enhance the user/visitors' experience. Multimedia applications are also widely used as a novel way of representing illustrated versions of text and artifact , and for the time , images, sound and text are now becoming available and accessible asyncronically through on-line local and global communication networks. Theoretically, these information highways not only extend and increase the available range and types of collections but also the visitor/user demographic.

So far, with the exception of a limited number of European initiatives; similar to the IMIN the International Museums Information Network and RAMA, 4 the process has been piecemeal and ad hoc in character and, while museums are keen to adopt new technologies, not for the first time can two key phenomena be observed. First, there is arguably a tendency towards a 'technology led' approach, whereby the mere existence of new technologies is deemed to constitute a solution to organisational problems, and secondly, the rapid and enthusiastic development of new applications is not always accompanied by rigorous systematic consideration of the use to which they will be put. The need for a strategic thought is now being recognised under the heading ìThe Museum and the Webî. The appeal of such slogans, is that they can be used to mean many things to many people. This slogan does, however, signify a shared sense that ICT's have the potential to fundamentally reshape the very conception of what a museum can be. On-line technologies are emerging which, theoretically at least, have massive implications across the entire range of museum work; ICT's promise to revolutionise the activities involved in collections management and user access and inquiry. Allowing for the associated element of 'hype', which one has come to expect in relation to such claims (Hemmings et al, 1996) and the inevitable 'churn factor', there can be little doubt that elaborating the 'virtual museum' generates a set of issues which museum professionals and academics recognise as critical for the future of the museum as a civic institution. We moot these issues not to question the value of ICT's but to focus attention on a host of enduring tensions that that provide an almost Faustian sub text of problems and promises which Silverstone (1995) succinctly identified as "objects and logics in time and space".

Problems and Promises.

Actual museums occupy physical space; they contain objects and artefacts; the character and circumstances of individual museums and their collections is often unique; they allow visitors to wander through an exhibited collection. The virtual museum provides us with a number of practical and theoretical issues which we list as the problems and potentials:

Set against these hypotheticals, our study covers a range of practical issues, including the work of curators and archivists, designers, marketing personnel, 'educators', management, intra- and inter- museum relations, and so on. Some, but not all, of the work we observed was germane to the use of technology. Where technology was seen as relevant, it was in terms of some very general considerations of 'potential' in areas where our colleagues at the two museums were concerned. We concentrate here on classification work and how it is done by curators and archivists, largely because it relates very much to problems of recording and using database information, because the use of database applications is more advanced in museums than any other kind of application, and because curatorial practices enable us to make classification visible as work and at that cooperative work, rather than as the deployment of pre-existing categories in stable hierarchies. What we try to do below is present an account of the kinds of relevant practice that need to be understood as classification work before we can begin the process of identifying what is 'appropriate' computer support.

Contested Meanings

If the commercial nexus is one dimension in which strategic thinking is emblemised by the concept of the virtual museum, then there are two others of major significance. The first involves a fundamental rethinking of the cultural role of the museum in society. Reiterating our introduction, concerning the fragmentation of meaning and the decentering of social structure in contemporary society, it is clear that cherished views of the value and role of the museum as a social institution have come under intense scrutiny. If the 19th century conception of the museum was to achieve 'by the ordered display of selected artefacts a total representation of human reality and history' (Donato, 1979 quoted by Bennett, 1995: 126), wherein 'as educative institutions, museums function largely as the repositories of the already known.' (ibid:147), contemporary accounts emphasise the multiplicity of narratives and the contestation of knowledge claims. This ongoing intellectual debate, marked by such labels as 'the new museology', is not simply an academic one; it is one which informed museum professionals are very aware of and to which they contribute, along with social scientists and others. Consequently, the issues raised feed back into professional practice in a much more direct way than is usually the case with such discussions in social science, informing and transforming ideas about acquisitions policy, exhibition making, education work and many other aspects of museum life.

The debate centres around the task of formulating a coherent rationale for the museum in place of the traditional one of the museum as a repository of fixed, universal values. Challenges to established discourses, e.g. of progress or enlightenment, mean that the notion of cultural authority which the museum traditionally exemplified is being re-examined in fundamental ways. This process seeks to respond to the displacement of such absolutist discourses by new and competing social narratives. Thus the language of change, even crisis, nowadays is being routinely employed within the professional museum literature. Many professionals would have sympathy with the remark by Sola (1992) that the truth is, we do not know any more what a museum institution is. Similarly, it is only logical to ask the question; do we now or have we ever, provided a perspicuous account of what constitutes a museum visit and how do go about understanding who we are identifying when we talk of 'the visitor' ?

Visitor Behaviour: the social construction of the visitor.

We referred above to the visitor and their visit, we argue, by aggregating these these terms nothing more is constructed than a gloss description for a diversity of socially organised activities. Furthermore, it is not only museum artefacts but also visitors who can be conceived as social constructs and sites of meaning contestation. If it is true, as the discussions referred to above suggest, that museums are taking up the challenge to escape from the tyranny of truth telling (Sledge, 1995), and responding to a recognition that visitors approach their institutions with a wide variety of expectations and questions and, indeed, motivations for asking those questions (Mackenzie, 1996), a key issue is what factors will shape the part to be played by technology in organising this response. Visitor studies are not new in museum work, but these studies for the most part constitute attitudinal surveys of one kind or another. Such studies assume a naive realist conception of the visitor: visitors are objectified and viewed as fixed or stereotypical entities. In contrast, we conceive visitors as sites of contested meaning: who and what a visitor is, or can be, is constructed interactionally as individual and group come into contact with museum environments which are structured in terms of conceptions of the visitor.

Thus, it is not simply that little work has been done to examine the nuances of visitor behaviour in the museum, and particularly distinguishing what takes place when different categories of visitor encounter museum exhibitions. Even less has visitor research examined the ways in which professionalised conceptions of visitors inform the structuring of the museum as a social environment and thus constrain what persons can be as visitors the logical corollary to this way of thinking is the identity of 'the virtual museum user' is constructed from the problematic philosophies. We make no assumptions regarding the visitors/users motives but if the virtual museum is to be more than simply engaging and user friendly then it must provide a 'wide number of subjects areas with an infinite number of routes' (MacKenzie, 1996) that not only 'enable people to think for purposes they have defined for themselves' (Sledge, 1995) but also provide guidance . Assistance in the form of specialist occupational work enables skilled practitioners to design exhibits and text in ways that are specifically pitched qua structured to the users intellectual requirements. Visible hierarchies is not necessarily 'the dead hand of curatorial diktat' (MacKenzie, ibid).

The Diversity of User Constituencies: we're all users now?

Furthermore, the commonsense dichotomy of professionals v. public needs to be deconstructed. Our early field research was focused on the work of curators, lay workers and enthusiasts, and on the organised educational visit. Our interest has been how these different users work with artefacts and documents, according to the purposes to hand and that knowledge work is situated action. These interests bear closely on the virtual provision of text and artefact because they suggest that the claims surrounding the flexibility of, for instance, Hypertext media (Landow, 1994) will, if they are to be realised, demand the same attention to practice as has been given to users in other social and technological contexts (MacKenzie, 1996). The logic of this philosophy is that it is the simply our commonsense abilities that allow us to recognise typical form of activities that allow us to make commonsense inferences about the identity of people wandering round the museum and therefore it is necessary to distinguish and clearly explicate these forms of situated interactional practices, that constitute the setting of both real and virtual museum activity. Understood this way, the dichotomy dissolves with the effect that we all become users.

Curatorial Work

Curatorial work 5 is of special significance in relation to computer technology since collections management relates very much to issues of recording and using database information. It is not surprising, therefore, that the use of database applications is more advanced in museums than any other kind of application. It is also notable that this work is co-operative, in the sense in which that term is understood in the field of CSCW. The museum `collection' constitutes the central reality and rationale of curatorial work. In their daily activities, curators work with `the collection' in myriad, socially shared and socially legitimated ways. The work done includes such activities as accessioning, whereby artefacts and texts come to be considered as candidates for inclusion in collections, and classified as being of a 'kind' within the collection; work involved in identifying and documenting a collection items, including the hidden collection of artefacts and texts not on display; and the work of display/exhibition making.

Again, this work is taking place in a changing organisational context. Our studies thus far have revealed a tension between on the one hand an intellectual climate in contemporary museum life which problematises 'meaning' and on the other a 'rationalising' tendency which seeks to stabilise or fix it for organisational purposes. As we have outlined, the changing status of the museum, and accounts of its 'proper' purpose, along with the pull of the commercial nexus, have substantially affected the professional curators sense of the curatorial function of the museum. (Kopytoff, 1986) In other words, curators perceive a requirement to be responsive to competing views concerning the meaning and organisation of knowledge. The characterisation of curatorial work as interpretation involves acknowledgement of this requirement. Such concerns open up a difficult analytic terrain to do with how one might organise and classify artefacts and texts in more flexible ways.


It is said that the single most obvious and determining feature of the museum is the presence of objects and artefacts (Silverstone,1995). There are, however, similar types of objects in the Mall, but commonsense tells us there is a difference. Goods are provided with a particular type of relevance, they belong in the a class of objects that are readily saleable and which we commonly recognise as the kinds of things sold in kinds of stores. Recognising a class object it seems then, is on the face of it unproblematic - kinds of objects, in kinds places at kinds of times, doing kinds of activities . Recognising an objects or artefacts claim to 'particular' status in 'particular' place- in our case the museum, rests upon it embodying characteristics which are often deemed to include one, some, or all of the following symbolic properties: uniqueness, significance, representativeness. Deeming or judging which objects or artefacts embody these relevant properties is a socially sanctioned act that can be carried out only by only those publicly or officially authorised as competent and qualified. The question is of course, how is this common understanding accomplished? We would suggest that our commonsense knowledge of the world relies upon our understanding the intrinsic and imposed 'domains of relevance' (Schutz,1962). Post-modernists would argue that these domains are becoming increasingly homogenised. Doug Makenzie, in a portentious paper presented at EVA '96 warned that web page design and the attraction of hypertext systems together with the more conservative curatorial procedures could reinforce "dogmatic " he recommends that virtual and physical museums would do well to remember Arthur Danto's (1992) remark 'that Philip IV did not have his portrait painted by Velazquez to have it hung with the Spanish School in the Musee Napoleon. He quotes..."only when the form vanish in which the King could multiply himself....did these paintings reduce to works of art in which aesthetics could murmur about diagonals". Danto continues with the observation that it is "as if the museum came into existence to celebrate the secondary value of such objects once their primary value was gone" What we find interesting here is that Danto recognises that this transformation - recasting objects from one class of thing to another, is a purely social phenomena.

Technology in Context

While academic attention has begun to be focused upon the issues identified above, thus far such attention has been overwhelmingly from within critical sociologies. Our study is predicated upon the assumption that the possibilities of the virtual museum will not be successfully realised without a detailed and sophisticated understanding of these activities as practical and socially organised activities. It is also assumed that adequate understanding of the requirements of the various stakeholders in the social world of the museum must begin from an appreciation of the situated character of such requirements. Sharing the phenomenologists idea of the integrity of the situation we express it in terms of what Schutz, (1962) called the "paramount reality of the Lebenswelt", simply put, the concept of understanding the meaning of a situation or event is made recognisable contextually as social setting is constituted in and through the form social interaction.

Across a number of social scientific disciplines (e.g. Anthropology, Psychology, Sociology, Cognitive Science), there is growing recognition of the situatedness of knowledge and of the importance this has for the design and use of technology (Suchman, 1987; Hughes et al, 1991). This awareness is fostered by a large and growing literature which documents the fact that technological developments are all too often associated with disappointing gains for users (Jirotka and Goguen, 1994). Consensus is emerging that the effective deployment of technology relies upon the degree to which it can be thoroughly embedded in its contexts of use, including the ordinary, practical orientations and competencies of those who are to work with it. The problem of understanding what computer systems are for in socially situated terms, rather than simply in terms of an a-social, technical perspective, is an enduring one, and must be addressed if such systems are to be woven into existing social practices and thus provide for the effective creation of new ways of working. Virtual technologies, whether or not they represent radical innovation, and regardless of their potential impact at a societal level, will be usable or otherwise according to criteria of relevance determined by situated use. Issues of relevance are at the heart of multi-user systems, and translate into specific concerns which have to do with the organisation and use of information. It is apparent that in a context such as the museum, factors such as contested meaning on the part of the professional community, differing organisational roles, including those of educators, curators and managers, and the particular needs of different categories of visitor mean that information retrieval is very much a matter of relevance. Usable technologies should enable disparate users to interrogate information according to their needs and concerns, and thus should furnish the user with relevant search strategies for retrieving information according to what they require from it. In turn, information must be organised in such a way that varied strategies can be supported. In part, these strategies are informed by the classification devices used by information users. Artefacts and text considered as virtual information resources do not exist in an undifferentiated mass, to be searched and sorted more or less randomly, but as resources which are to be interrogated according to professional and lay categories- categories which will vary according to purpose.

Orientation and Methodology

Classification work, which can be defined as the ordering of things according to some scheme of categories, is central both to lay and professional use of museum information. Of course, at least since Durkheim, classification has been understood as a social phenomenon as much as a cognitive one, suggesting that the ways in which individuals carve up reality cannot be adequately understood without reference to their social membership. Indeed, his remark that the ideas which correspond to the diverse elements of language are....collective representations (1976, p.434) is foundational to much work in the sociology of knowledge. However, the sociological analysis of classification is still an open issue. Ethnomethodological approaches, for instance, have frequently criticised the division it assumes between the social and the cognitive, and have argued that semiotics, Levi-Straussian structuralism and the ethnosemantic tradition (e.g. Frake, 1980), transpose culture from the social realm to the inside of the members head. Furthermore, the methodology by which this cognitive order is extracted could be said to leave out the interpreters work (Eglin, 1976). The point here is that treating classification as work done is a methodological choice which we believe provides a useful approach for understanding how information in the virtual society will in practice be used.

Ethnomethodological studies emphasise the practical and contingent character of classification under real world conditions (Garfinkel, 1967). Our own stance on these methodological choices is ethnomethodologically informed and is intended to develop the now substantial literature on ethnography in the context of systems design, mainly to be found in the interdisciplinary fields of Human Computer Interaction (HCI) and Computer Supported Co-operative Work (CSCW). Such an approach has a number of advantages in this context. Firstly, emphasis on the situated nature of work interactions gives an analytic purchase on the different ways in which actors can encounter the museum exhibit, locating their understandings of display in their local and contextual purposes. Secondly, such studies have a strong and distinctive evaluative element, relocating evaluation strategies in the real world of work and organisational context. Thirdly, they can link with user participative methods, the ethnographer acting as a bridge between the communities involved and assisting in the articulation and representation of viewpoints (Twidale et al, 1994). Nevertheless, a substantial part of this literature is concerned with circumscribed locations, and the study of information use in the museum context.

The Ethnography

Our analysis is based on an ongoing two year ethnographic study of the work of museum staff and visitors in two locations in the North of England; the Museum of Science and Industry in Manchester (MSI) and the National Railway Museum in York (NRM), concerned in part with what might be called 'scoping' the problem of information use at the museums in question. That is, we have been mainly concerned with an attempt to understand what kind of work is undertaken, and by whom, and how this work might prove to be relevant to information needs. Our study continues and we have currently completed about a year of observation. Our reasons for focussing on the museums in question had to do with local accessibility and the existence of contacts who could facilitate entry.

We began to recognise early on that some features of museum work, for us at least, had been previously unconsidered. Firstly, each museum has its own biographical history. By way of example, the origination of the Manchester museum was as a small scale, almost personal enterprise prompted by the interests of an individual in the history of engineering. The museum at that time, therefore was seen as 'his personal project', or, 'run like a fiefdom.' The fundamental function of the museum then was perceived as educational, quite specifically as an 'aid to schooling'. It grew when offered a railway goods yard and all the artefacts it contained. The museum then became situated historically at the location of the world's first railway station, and this informed the formation of a set of embedded assumptions about 'what was important'. The original source of funding was the local council, and this in turn influenced policy- a policy that was essentially parochial and very much to do with what was perceived as 'locally relevant'- especially since the 'Manchester' connection determined things like funding support for acquisition. In a nutshell, the museum was run by enthusiastic amateurs for whom the evolution of policy was a question of deciding what it was that they were enthusiastic about. Policy was, of course, progressive, and management and curatorship increasingly professional. Nevertheless, this brief history indicates the way in which we want to understand policy concerning relevant acquisitions and display decisions- as being arrived at in the 'doing' of decisions concerning 'what do we want' and 'what don't we want?'.

Folklore would have it that we might assume that curatorial functions were akin to detective work; happening on clues to the whereabouts of rare and valuable items, and tracking them down, the reality here was startlingly different. The most surprising aspect, for us, was the fact that both museums are confronted with an almost endless supply of artefacts and materials. Their problem is less to do with how to find exhibits, but with how to decide what to reject and what to keep, and the latter greatly exceeds the former. That is, the museums are engaged in sorting work. This is manifested above all in 'accessioning' and 'de-accessioning', or deciding what should constitute the 'collection'. That is, to an extent, an 'embarrassment de richesse' affects acquisition policy, and available space has a significant impact on what can be kept and what cannot.

The local production of order

Understanding museum work, then, means seeing it as having a historical character, as the sedimentation of things that have happened, chronicles of events and significant developments, and of the roles and consequences of artefacts. It is constituted in short as the 'ordering' of history. In the sociology of museums this has been much remarked upon, as outlined above, generally in terms of the way the order of things, to use Foucault's formulation is an ideological or discursive construct (see Bennett, ibid, Macdonald and Fyfe, 1996) and how these are spaces which have been opened up in the 'post-modern' era for contestation, for instance by feminist scholars (see Porter, 1996). Competing views of history, then, inform the work of curators to a greater degree than ever before, along with the recognition that their interpretive work is at best 'partial' and incomplete. Regardless of this rhetoric, however, our interest here is the way in which the coherence of classificatory schema is arrived at and maintained, for in the work one of the most striking features is the way in which agreements at the level of classification need to be reached before there can be any disagreements about significance. Even so, classification is above all the work of making history coherent, of arriving at definitions that are suggestive of the ways thing relate. Relevancy decisions are made precisely with an eye to the way things 'fit'- and using those definitions as 'working hypotheses' or resources for further coherency work. That is, classification is used to produce order, and is produced in turn by it. The point here is that these categorisation devices are the rationale by which curators arrive at mutual understandings of what it is that they are doing. It is, in short, accounting work. Such a description may make it appear that inscribing terms and usages in databases will be no more complex than for any other set of terms, and so it would be if the coherence we refer to led to a single usable hierarchy of terms. That is, if usage could be made 'planful' such that all new objects and texts could be made to fit an existing thesaurus unproblematically. The problem is precisely that it cannot easily been done .

Our interest in the way in which objects are specified as being 'of a certain kind' relies on uncovering the knowledge and experience which informs the practical, day-to-day, work of curators and others. For example, one way in which this work is made visible is when considering acquisitions and their cataloguing. Unsurprisingly, museums need to do this. Institutions of this kind after all would find it difficult to operate without recorded knowledge of the artefacts they own or control. Nevertheless, it is more than simply making a list. The inscriptions which are generated, be it in the catalogues, the cards accompanying displays, or the fields in databases they use, indicate not just the name of the object, but its status. That is, the act of examining and inscribing the object's status is an attribution of 'value'. They are examples of what Berger and Luckmann term 'typifications' (Berger and Luckmann, 1961), but typified according to relevancy criteria determined by expert judgement. Deciding 'what kind of object it is', then, is a matter of providing a brief description which incorporates judgements concerning issues such as 'provenance', 'uniqueness', and 'relevance to the collection'.

In both the museums we are examining, objects arrive and leave in a more or less continuous stream as candidates for inclusion. To reiterate the point, artefacts come to the museum without an 'authority'. They are merely candidates for inclusion, and their inclusion confers an authority upon them. In other words, an authenticity is granted them by virtue of inclusion, and they stand as 'authentic' to the visitor as a result. Procedures for 'theming' exhibits within the museum, for deciding what to display and where, for cataloguing the collection and so on, are reflexively linked to classification schema. Classification, however, is an increasingly problematic area. In this respect, the developing interest in the relationship between image and text (Burgin, 1986a, 1986b) is relevant. Whilst not wishing to engage with the explicitly political rendering of meaning that Burgin and others are concerned with, it is clear that serious issues concerning the classification and 'meaning' of artefacts are raised. Providing 'coherent' and 'relevant' themes is a matter of organising artefacts and text, not merely representing them.

The limitations of physical spaces mean that, often, only highly constrained choices are available. An initial example concerns a large Goss printing machine currently in the possession of the Manchester museum. We were present at discussions concerning whether this machine should be de-accessioned, which in effect means finding another institution that might be interested in it, or otherwise scrapping it. What is interesting here is not the machinery, but the rationale that was brought to the discussion, which included a discussion of 'provenance', a term covering the history of the machine, the site it came front, the local relevance of both the site and the machine itself, the importance of the printing industry to the local area, where it was made and so on. Points at issue included that the machine itself was not made in Manchester, but was a major technological development in the development of the printing trade in Manchester, that chronicling the history of printing necessitated the ownership of key machines, but that printing was 'no longer that important' locally. The point here is not that these factors are the most important in determining the fact of an artefact, but that they were the most important governing the fate of this artefact. That is, an account is provided which has to do with changing assumptions about its value. This whole discussion takes place against another background, that of available space. The fact of the matter is that printing machines are very substantial, and retaining it weighs against the acquisition of other things. In a nutshell, objects have careers. Hence: database management in museums will involve careful control over the recording of both what enters and what leaves the museum, and the collection changes in a more or less continuous way.

Needless to say, not all artefacts are large objects. Many are text based, and as it were, come bundled. The result is that they have to be searched, sorted and identified in order to establish what their value, if any, is. By examining one such example of artefacts arriving at MSI, which consisted of two archives from large electronics companies based locally, we can identify some of these themes. In the first instance, a large collection of text based materials was accepted from a company which had been based in Oldham, a town near to, but not part of, Manchester. Subsequently, a bigger archive, which included photographs and detailed historical records, was offered from GEC, a company based in Trafford Park, a region of Manchester.

Of course, such materials do not arrive prepared and ordered for museum use. The curators, in their own words, must 'do a job on it'. Watching them examine the material, it is clear that the physical nature of the artefacts in question provides a variety of accessing routines. Some of the material is flat, some of it consists of photographs of buildings, machinery and people, some of it is large format paper, and thus can be unfolded, as with blueprints, etc. Some of it is sequentially ordered to reflect a development. The practised eye of the curator can see how this material could in principle be organised- its potential as a display item or as a 'fileable' resource. As the material is accessioned, knowledge is applied to it-

'oh, this would do for ... look, photos of Trafford Park ... you can see the development ...'

'GEC is a big international company, but its got a local history. I mean, GEC is Trafford Park, and Trafford Park was once the world's biggest industrial centre ...'

'the [other] archive is just not so interesting ...'

As they are unpacked, the materials are 'seen' as what they might be. Drawing on the kinds of relevancy we mention above, the archivist is able to tease out the social, political and economic relevance of the material, and thus how it might be ordered for the visitor or the professional interrogator of such material, in and through the sorting that is going on. The 'interpretation work' that is taking place is 'doing' work. It is the physical work of separating material into piles, noting that some parts of it have a particular relevance, some could be incorporated into other displays, some have 'exhibition' potential, and so on. Notably, the criteria being applied centre on the local relevance of the material, as in 'GEC is a big international company, but it's got a local history', and on the significance of the material, as with, 'oh look, photos ... you can see the development ...' and 'Trafford Park was once the world's biggest industrial centre ...'. In other words, the sorting and classifying of the material is done with an eye to the story that can be told, and a story which is in keeping with known-in-common criteria. Curators evidently 'know what they mean' when they use terms like, 'is an example of', 'is typical of', 'is a rare example of', 'is of the school of', is 'the first' or 'last example of' or is 'a common type of' and such terminologies are reflexively deployed against, in this instance, background assumptions concerning the museum's interest in 'local' matters and in 'industry', as well as 'what we can do with it'. All of these usages, despite their sometimes contradictory status, might at some time be used as justifications for retaining or acquiring a particular piece. The objects present themselves as candidates for inclusion and are sorted on the basis of what 'is interesting' about them. What is interesting is determined not only by assumptions about the object itself, but by the determination of the curator's practised eye deciding on its 'place' in the collection, a place which is determined by the status of all the other objects in the collection and the curator's knowledge of their relationships.

For MSI in particular, the judgements arrived at are dependent on criteria such as whether it has local relevance, whether it has significance for the subject area of the museum, whether it is unique, and the object's provenance. These matters, it turns out, are important not only for accessioning but also for de-accessioning. Museums of this kind have any endlessly changing set of potential exhibits and frequently wish to move objects elsewhere, or simply dump them. The point here is that the Accessions catalogue constitutes a brief record of the rationale for accepting objects or disposing of them. Not least, it seemed the entering of descriptions in the catalogue gave a status to the object, a status that is arrived at by discussion and subsequent agreement among participants. In other words, the work of cataloguing is the work of defining objects as being of a 'of value' for specified reasons. What is being recorded, so to speak, are curatorial agreements concerning the 'value' of potential exhibits. A more complete picture of the things that decide 'object value' in this work, we think, would tell us much about appropriate categories for database organisation.

On the back of this example, we can make a number of points about representing the status of objects, including that it is not a matter of imposing a rigid classification scheme, but of using existing classifications for deciding 'what you can do with it.' and it is in this way the we can begin to see knowledge work as situated action. In that classifications deployed are not simply descriptions of the objects themselves, but emanate from 'seeing' the material as it might be as part of a gallery, of an existing or potential exhibition or as a resource for scholars. That is, classifications incorporate assumptions about the 'value' of objects. Classification schema are contingent upon relevancy criteria which are not self-evident, but have to do with evolved policy concerning the museum's 'interests'. They include, in addition to those aspects we mention above, their use when applying for budgets, determining levels of staffing, space to be allocated, etc. Classification is not fixed but is contingent on the practised reading of acquisitions, and is remade reflexively as new acquisitions take their place in the collection. Classification work provides for matching references looking and finding and found, in order to name. Naming then inaugurates the object in preparation for description and measurement and measuring allows for further comparison- treating each as resembling each other- infinitely extendible, recording ordering in categories provides a method of formally documenting inscriptions.

As any practitioner will know, museum work is complex. Classifying work does not go on in any single place, or stage of an object's career, and is not solely the responsibility of the curator. Museums employ staff with various responsibilities, which typically include, for instance, professional designers employed for exhibition purposes, educational work of an organised kind, as well as marketing, and administration functions. In addition, professional staff frequently rely on networks of enthusiastic amateurs for aspects of their work, as we shall see. On many occasions, these roles become rather fluid, and it is difficult in practice to entirely distinguish between what designers and educators do, and what curators do. There is no space here to detail all aspects of the work, so we content ourselves with referring briefly to two others. We do so because in different ways they demonstrate the contingency of classification, according to the purposes at hand. Analysis of artefacts and their allocation to suitable categories takes place then when exhibitions are put together and when responding to enquiries. The 'in house' work of preparing for more permanent exhibition is considerable. Here, the work is that of selecting relevant exhibits, structuring them in an orderly and coherent way according to the 'story' they are trying to tell, and providing a narrative text. Again, the work of producing the story is invisible to the outsider and is produced out of discussion and agreement concerning the status of objects which an exhibition might contain, and how they relate to each other. However, these relations are contingent on the exhibition in question. It is, one might say, deemed 'like this' for 'this purpose' although it might quite well be described in other ways for other purposes. Further, it is not as if the story to be told comes fully formed. The production of the exhibition is as much the production of a story as it is the production of an organised set of artefacts, and they inform each other. The objects, the accompanying text, and the positioning of the object vis a vis others together tell the desired story.

Electronic catalogues of objects, then, may require organisation with exactly the same attention to what stories are to be told. Classification schema sometimes imply a concern for narrative and in our view, inadequate attention has hitherto been paid to how various stories can be woven together, and more importantly what we wish them to be, when electronic media are being used.

Other relevant classification work can be seen in response to enquiries, typically by telephone or in person. Indeed, in the museums in question, this fact is recognised and institutionalised through rota systems for dealing with them. Although many of these enquiries were of a routine nature, some were not. There is, for instance, the occasional request from the media to provide information of a more in-depth kind than one would expect from a typical member of the public. Our point here is that answering enquiries can be far more than merely giving an answer to a self- evident question. In such situations, responding to an enquiry can be a process of identifying exactly what it is the enquiry is about, and can involve identifying, finding, and preparing, the 'local expert' to answer the enquiry in question. The social distribution of knowledge and expertise is a common feature of organisational life, and we have seen elsewhere (for instance in studies we have been involved in the financial sector) that asking and answering questions, whether it be from customers or colleagues, is a frequent and surprisingly complex activity. The point here is that 'answering enquiries' sometimes involves making sense of 'what kind of enquiry it is', a task human beings are well suited to. 'Understanding the question' and 'Finding the person who knows the answer' relies on 'what the respondent knows' as part of his or her professional expertise. Embedding this expertise into systems intended to answer the questions visitors or staff may have is anything but a simple exercise. It relates directly to the problem of database interrogation and certainly involves more than the 'Frequently Asked Questions (FAQ)' framework commonly found on the Internet. The point here is that classification schema embedded in computer databases do not necessarily pay due regard to the fact that relevant classification is contingent upon the task in hand. Museum databases should be flexible enough to be retrievable in such a way that other roles are supported, not merely that of the curator.

Our argument, then, is that classification work involves the production of a 'sense of order' out of the myriad artefacts and texts available. Despite the recognition that historical interpretation is a contingent matter, there is not, and has not been to our knowledge, any serious suggestion that museums could get by without managing objects into 'orderly' collections. The 'sense of order', in other words must precede any subsequent disagreements about its partiality, and we need to understand more fully how it is arrived at, if we are to successfully evaluate the possibility of developing standardised usages. The following example which came to our attention took place over a period of time at the NRM, and involved the senior curator in the engineering department of the museum and a network of enthusiasts as they evolved a suitable classification scheme for a particular kind of railway truck. The background to this is that for the visitor to the NRM, steam trains are redolent of many things. They evoke, for many, a 'romantic' past. Steam engines, it seems are 'of interest', and the museum can trade on a number of assumptions about the public's enthusiasm for their history, including the apparent attraction of 'steam', science and engineering issues, political and cultural background, etc. Not least, they have a large and impressive physical presence. Notably, however, this enthusiasm for the engine on the part of the public is not accompanied by any enthusiasm for the rolling stock. 'Wagons', as aficionados term them, have other characteristics, which seem to make them less interesting to the public, including that they are exceedingly common, they are not easily distinguished one from another by the unpracticed eye, they are not 'nice to look at', and so on. Nevertheless, for the railway expert there is a considerable interest. A number of historical developments meant that certain wagons were manufactured to do certain jobs in certain locations. For instance, wagons specifically designed to transport clay from Cornwall to centres of paper finishing in the North of England, were built as the rail network developed to allow large scale transportation. This, it seems, involved a number of problems, including the fact that the existence of private regional companies meant different sized gauges had to be contended with. Such matters, however obscure to the rest of us, are interesting to the curator. In terms of classification, an operational scheme existed for indexing these wagons according to what is known as a TOPS number. Such a scheme, however, is of no value to the curator since the index is operationally defined, meaning that it distinguishes only by criteria such as axle weight and payload. Operationally, TOPS allowed one to determine whether one set of wagons could be connected to another, and what engine could tow it. Marshalling was dependent on this scheme. The TOPS number is of no value for curatorial purposes, precisely because the historical significance of the object is not equivalent to its operational significance. The object has changed. At the NRM, the group in question had determined that the historical part played by the wagon in the development of the railway had been somewhat under-researched, had if you will something of a 'Cinderella' status, and set out to catalogue and classify all existing wagons of a particular kind. The classification issues here turned out to be markedly different from those implicit in their operational value, and were evolved as part and parcel of the work of recording their existence. In brief, it was decided that the network of enthusiasts would 'spot' these wagons in much the same way that trains are 'spotted', but that in so doing they would record certain relevant details, most notably where the wagon was to be located, the condition it was in, and so on. Each agreed that when embarking on this exercise they would photograph the artefact in question, and record the relevant details in notebooks. This candidate classification scheme had to do with 'provenance' in museum usage, carried and organised in and through a small group's discussions of what needs to be recorded so that the recording could be done at all. It was evolved through the 'taken for granted' background knowledge about typical classification devices that enthusiasts and curators share. The point is that, as with all classifications, the terms used need to be consistent if they are to be recorded for comparison work, so categories such as 'rotting', 'wheels on or off', 'coal', 'clay', 'manufacturer', etc. were agreed on. That is, and in the first instance, some 'broad brush' classification enabled the work of 'collecting' data to take place. We might refer to these, in Berger and Luckmann's terms, as typifications, or as candidate classifications, in that they will 'do' for the purposes in hand. They are socially organised agreements.

As information began to amass, the problem became one of organising it into an orderly record. The brief details were, therefore, brought back and put into an A4 ring binder. Entries were placed in the binder chronologically in order of. Using this binder all the time, and getting used to 'where things are' in the binder allowed the group to do comparison work, such that using the record as a pre-classified resource allows members of the group to compare discoveries with existing finds, and thereby to generate their own knowledge of the 'state of play'. As their endeavours grew, so it became evident that a more effective recording system was required, largely because the work itself had produced a greater sense of its importance. That is, in the words of the senior curator, "its important work ... we can't keep all these wagons, and many of them are deteriorating ... its the only record we have of a part of Britain's industrial and transport history." It was thus decided that the information be transferred to an electronic database, purchased 'off the shelf' for its customisability (called ACCESS). A set of criteria were worked out, which were: TOPS; geographical location; the owner; date of manufacturer and the who manufacturer was; a photograph. On top of this, a scale of 1 to 5 is used to record condition (status) and the criteria became fields inscribed in the ACCESS system. The scale, along with the other fields, was agreed beforehand on the grounds that it was a 'good' method. 'Good' here, consistent with Garfinkel's findings (ibid), meant simply that members arrived at agreements with a degree of consistency. A text box in the system also allowed the group to record biography, special details, whether the wagon is rare or typical, is worth buying or not etc. The data, both in the ring binder and in the system, was then available for interrogation, and thus to interpretation. Thus the curator was able to show us an example and say, 'this wagon is a fine example of a particular type of clay wagon ... it was used in 1947, and played a particularly important part in the development of the paper finishing trade.' One feature of this ongoing work was that it implicated two distinct kinds of classification work. The first is what Coulter (1991) has referred to as taxonomic work, which involves laying down the terms by which artefacts can be organised in the first place. The second, however, lies in the special qualities of individual items or their 'uniqueness'.

Taxonomic work and 'uniqueness' work are distinguishable, and both need to be supported in museum databases. For the most part, this system of coding and incorporating information into a small database worked well. Problems arose, however, as the network of people involved began to expand and we were able to observe first hand some of the consequences when being shown the system. Here, the reliability of members' terms came into question, as the terminology used when new 'spots' were input into the database became increasingly variable. Particular confusions arose over the status of one variety of motorised wagon, termed a 'Mogo'. It seems there are a number of different kinds of Mogo, identified by weight, for instance the 8 and 12 ton versions. As the curator showed us how he used the system, he said,

"I'll bring up Mogo, cos I know there are two of them. I can use search criteria like these ... oh, there are five. Hang on, I'll try again. There you are, those two ... I can't see why I got five the first time. It's to do with the order you search in, but I didn't know there were five of them."

Of course, problems of search ordering are well known, but that is not the point here. What had happened was that different people had made assumptions about the appropriate terminology and reached different conclusions. Thus, 'spots' had been entered variously as '12 ton Mogo' or 'Mogo 12 ton'. Some entries of 'Mogo' had been capitalised, others entered in lower case. Members had taken for granted their knowledge of Mogos and assumed that the database was configured in that respect, whereas it had been configured according to the senior curator's categories alone. There followed a lengthy and animated discussion which we will call a 'what's in a name?' discussion and which we need not detail. Arguing about a name, however, was indicative of the need to identify fields in such a way that they both incorporate and exclude in ways that are in common with members' categories. This lays out a potential problem for the enlargement of use, in that standardisation, however desirable, may require us to be sensitive to the changing user population, and variations in the common-sense categories they deploy. General conclusions we might reach from this example include:

Classification schema evolve in the course of the job of work in hand. The 'value' of artefacts can become decided through the deployment of recorded data, but equally determinations of value produce new versions of the classification scheme.

The effectiveness of classification schema depends on the 'authority' of those involved in producing it. Users have considerable difficulty in inputting and interrogating information where any number of potential descriptions could be appropriate, and where the user population evolves.

The database might support not only the collection but a general history of relevant objects.


'Evaluative' ethnographies (Hughes et al, 1994) have in the main been used to demonstrate specific ways in which technologies have failed to 'fit with' work practice, and on occasion to reveal how 'work arounds' can obscure problems with the technology. More generally, it has been suggested that their task is to deal with 'uncertainty' (see Twidale et al, 1994; Randall et al, 1995) Our task is not to be overly critical about existing systems, for in Europe at least the history of database applications for museum work is a history for the most part of adapting existing software to support some fairly standard functions. Indeed, large parts of museum work currently involve no use of computer technology at all. That it should be so, as we have suggested, is arguably because the naivety of museum staff has made the elicitation of 'requirements' a difficult task, and therefore the prospective success of computer systems in museum work deeply uncertain. Our purposes here lie in the evaluation of database technologies in a rather different sense. We are not concerned with the 'failure' of particular systems, for our impression is that systems in use are hardly of the most advanced kind. The evaluation, if that is what it in fact is, concerns a set of organisational and work- related issues which provide some initial scope for understanding the notion of a 'sense of order', and how deconstructing the term might help us reduce the level of uncertainty contained in database use. That sense of order, we suggest, is arrived at in a series of contingent ways which are reliant on work done cooperatively, sometimes in small and fairly coherent groups, sometimes across the boundaries of responsibility, and sometimes by virtue of the way in which general assumptions about purpose are organised.

In a climate where museums are under an ever increasing commercial pressure, and where once secure roles are becoming challenged, the task is more to do with applying our study of work practices to the general problem of ordering and classifying the myriad objects which museums keep in their collections, objects which can include artefacts, texts, images, whole archives, and so on. What we do want to do is suggest some of the subtlety of classification work in such a way that the next generation of systems might support work as it is practically accomplished, work that for the most part has remained 'seen but unnoticed', in Garfinkel's terms (1984) to the procurers of existing systems. Of course, Garfinkel raised the issue of classifying in terms of 'common sense choices' in his analysis of the documentary method (ibid), and we want to see classification in the museum context in much the same way. Classification, when considered as a job of work, can be seen as orienting to, and contingent upon, a number of factors which have to do with the way objects and texts come to attention, by whom, what the local relevancies of the museum are, the research interests of curators, and the prospect and desirability of display. All of these factors may impact on what classification is deemed appropriate, and lead to the conclusion that artefacts may in principle be classified in any number of ways. Indeed, as we have seen, they frequently are. The contingent status of artefacts, the invisibility of rationale, the 'work done', have direct corollaries in design terms. We have suggested a range of issues that seem germane to the general problem, including the consideration of objects and collections as 'careers' and thus change, how classifications emanate from 'seeing' artefacts in terms of their value in the collection, how they are impacted upon by a series of local relevances, how it is dependent on the practised reading of acquisitions, relate to narrative possibilities, depend on the 'job in hand', and on the 'authority' of the user. The comprehensive electronic organisation of data about museum artefacts would be greatly enhanced by the prospect of parallel classification schema, search pathways, and sophisticated database interrogation techniques which would allow the contingent uses we have outlined above. Those with a more cynical view of sociological input into system design will not be surprised to hear that we proffer no solutions. The point here is to raise initial questions concerning how working with database systems might allow, rather than obstruct, contingent classification. The task here, as we construed it, was simply to do the ethnographic work, evaluate the work of museum staff both in their use of technology and without it, and understand how in practice they 'managed' their collections.

In this paper, we have sought to raise some initial issues concerning the use to which potential electronic resources for collection management may be put, based on the observations we have undertaken. We hope, as the work unfolds, to begin to specify how they may be dealt with through a more detailed and thorough description of classification procedures, visitor behaviour, and educational work. We have not raised at all the problem of relating professional use to visitor needs, and believe that if any of these possible benefits are to be realised, a substantially more nuanced view of the work of museum staff and visitors is needed. After all, the objects and their inscriptions are not merely examples of the past but are instructions as to 'how to tell about the past'. In a sense, a special, respected, experienced, entrusted authority, resides in the museum, whereby a tacit contract exists between the museum and the audience. The issue of providing appropriate 'pathways' through complex and changing information resources, we suggest, is central to this. If such technology were to incorporate visitor uses as well, then the relationship between professional contingencies and the degree to which they match or otherwise the assumptions and practices of visitors will required.


Leonard Will initiated one of the UK's first museum computer databases at the Science Museum in 1982.
At the Euro'96 Football the Microsoft web page was registering around 9 million hits per week. An alternative web site provided by Manchester City Council was still registering around 9k hits per week two months after the competition.
See, A Common Wealth: Museums and Learning in the UK. Dept. of Heritage 1997.
Remote Access to Museum Archives, Race 2041 aims to make possible consultation and dialogue amongst museum databases possible. These data bases exist at institutions but have different internal structures ie. ORACLE, INGRES, BASIC+, TINMAM, DIABOLA etc.. and different vendors such as IBM, HP and Sun.
By curatorial work we mean the work of those whose activities include planing, designing, constructing, collecting and organising museum collections and exhibits.
The principle of contextual determination of meaning has been dealt with extensively (see Douglas,1967 especially 242).


Bennett, T., (1995), The Birth of the Museum, London, Routledge.

Berger, P. and Luckmann,T. (1961), The Social Construction of Reality, NY. Doubleday.

Burgin, V., (1986a), Between, Oxford, Blackwell in association with the ICA.

Burgin, V., (1986b), The End of Art Theory: Criticism and Postmodernity, Basingstoke, Macmillan.

Button, (1990), Going up a Blind Alley, in P. Luff, N. Gilbert and D. Frohlich (eds.), Computers and Conversation, 67-90, San Diego, Calif., Academic Press

Chen, H., Ng, T., Martinez, J. and Schatz, B., (1997), A Concept Space Approach to Addressing the Vocabulary Problem in Scientific Information Retrieval: An Experiment on the Worm Community System, Journal of the American Society for Information Science, 48 (1):17-31.

Coulter,J.(1991) The Praxiology of Perception: Visual orientations and practical action. Inquiry 33.

Danto, A.C. (1992), The Museum of Museums in Beyond the Brillo Box, Farrar et al, NY

Donato, E., 1979, The museum's furnace: notes towards a contextual reading of Bouvard and Pechuchet', in J. Harrari (ed.) Textual Strategies: Perspectives in Post- Structuralist Criticism, Ithaca and London, Cornell University Press.

Douglas,J.D. ed (1971), Understanding Everyday Life: Toward the reconstruction of sociological knowledge, Routledge and Kegan Paul.

Divall, C., (1994), Theories and Things: Using the History of Technology, in L. Fitzgerald and G. Porter, Museums, Collection and Interpreting Domestic Artefacts, Manchester, Science and Industry Curators Group

Durkheim, E., (1976), The Elementary Forms of the Religious Life, London, Allen and Unwin

Eglin, P., (1976), Leaving Out the Interpreters Work; A methodological critique of ethnosemantics based on ethnomethodology, Semiotica, 17 (4) pp339-369

Frake, C., (1980), Language and Cultural Description, Stanford, Calif., Stanford University Press

Francis, D., (1995), Negotiation, Decision-making and Formalism, in J. Wagner and K. Ehlich (eds.), The Discourse of Business Negotiation, Berlin, Mouton- De Gruyter

Fraser, N. and Woofitt, R. C. (1990), Orienting to Rules, in N. Gilbert (ed.), Proceedings of the workshop on Complex Systems, Ethnomethodology and Interaction Analysis, American Association for Artificial Intelligence (pp69-80), Boston

Garfinkel, H., (1967), Studies in Ethnomethodology, Englewood Cliffs, N.J., Prentice-Hall

Hemmings, T., Randall, D., Marr, L., Francis, D., Porter, G. and Divall, C., (1996), Hype and Reality in the Virtual Museum: Some Initial Problems, Proceedings of the EVA Conference, The National Gallery, London

Hughes, J, Randall, D. And Shapiro, D., CSCW: Discipline or Paradigm? A Sociological Approach, (1991), in L. Bannon and M. Robinson, (eds.), Proceedings from Second European Conference on CSCW, Kluwer, 1991.

Jirotka, M. and Goguen, J., (1994), Requirements Engineering: Social and Technical Issues, London, Academic Press.

Kopytoff, I, (1986), The Cultural Biography of Things: Commoditisation as Process, in A. Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective, Cambridge, Cambridge University Press

Landow, G.P. ed., (1994), Hyper/text/Theory, London, The Johns Hopkins University Press.

Macdonald, S. and Fyfe, G. eds., (1996), Theorising Museums, Oxford, Blackwell/ The Sociological Review

Macdonald, S. and Silverstone, R., (1990), Rewriting the Museums Fiction Taxonomies, Stories and Readers, CRICT Discussion paper No. 7, Brunel University

Mackenzie, D., 1996, Beyond Hypertext: Adaptive Interfaces for Virtual Museums, Proceedings of the EVA Conference, The National Gallery, London.

Porter, G., 'Seeing through Solidity: A feminist perspective on museums', in S. Macdonald and G.Fyffe eds.(1996), Theorising Museums, Oxford, Blackwell/The Sociological Review

National Heritage report, January 1997, A Common Wealth: Museums and Learning in the UK

Sledge, J., 1995, Points of View, in D. Bearman (ed.), Multi-Media Computing and Museums, Pittsburg, PA., AIM

Sola , T., 1992, Museum Professionals- the endangered species, P.Boylan (ed.), Museums 2000, London, Routledge

Star, S. Leigh, and Griesemer, J.R., (1989), Institutional Ecology, Translations and Boundary Objects: Amateurs and Professionals in Berkelys Museum of Vertebrate Zoology, 1907-39, Social Studies of Science, Vol. 19, 387-420

Suchman, L. 1987, Plans and Situated Actions, Cambridge, Cambridge University Press.

Twidale, M., Randall, D. and Bentley, R., 1994, 'Situated Evaluation for Cooperative Systems', Transcending Boundaries: Proceedings of CSCW '94, Chapel Hill, North Carolina, ACM Press.

Vergo, P., 1989, The New Museology, London, Reaktion Books.

Archives & Museum

Informatics Museums and the Web Register Exhibit Sponsor Program

Last modified: April 15, 1997
This file can be found below

Send questions and comments to