Best of the Web
Key Dates

When is the next
Museums and the Web?

A&MI home
Archives & Museum Informatics
158 Lee Avenue
Toronto, Ontario
M4E 2P3 Canada

Search Search

Join our Mailing List.

Archives & Museum Informatics

On Beyond Label Copy: Museum-Library Collaboration In The Development Of A Smart Web Exhibit

Bernadette G. Callery and Robert Thibadeau, Carnegie Mellon University, USA


Visitors to modern museum exhibitions see only a limited number of objects selected from the collections. With only limited label copy available, it can be difficult for the visitor to understand the object in the context of the scientific, artistic or cultural theory presented in the exhibition. What happens when the visitor wants to know more? How can museums offer the original documents of discovery, the raw material of research, to the museum visitor and the World Wide Web user? Carnegie Mellon University Libraries, the School of Computer Science at Carnegie Mellon University and the Carnegie Museum of Natural History are collaborating to develop the prototype of a Smart Web Exhibit (SWE) designed to deliver information online, on target and on time, to a diverse user community. Visitors will view digitized archival material drawn from signature collections at both institutions. Users of the SWE will select material based on their level of interest (basic curiosity, generally informed interest, or detailed technical interest) or the amount of time they want to spend. Both onsite and online users will have the option of viewing a choice of pre-selected online exhibits of this material or searching the online collections by the materials' type, keyword or subject terms. The University Librariesí SWE will provide access to photographs, correspondence, lecture notes and published and unpublished papers of two early innovators in computer science, Herbert Simon and Alan Newell. The Carnegie Museum's SWE will display unpublished field notes, correspondence, photographs, published popular and scientific articles dealing with the Museum's early paleontological discoveries, especially that of Diplodocus carnegii. Challenges and compromises of the initial stages of this collaborative project are discussed.

On Beyond Label Copy

Non nova sed nove: Not new things but in a new way


Visitors to modern museum exhibitions see only a limited number of objects selected from the museum's collections. Since the late nineteenth century, when the concept of the "index" museum was proposed by J. E. Gray of the British Museum (Natural History) and promoted by W.H. Flower of the Hunterian Museum of the Royal College of Surgeons, many public museum collections have been divided into those items selected for exhibition and a much larger research collection that is not usually available to the public (Stearn, 1981). In an index museum, selected typical specimens summarize the whole of a collection in a relatively small space, as it was felt that the museum's large collections organized for scientific study, with their masses of unique, but related objects, would bewilder and tire the public. Museum visitors, unaware of the existence of these separate research collections, may assume that the entire holdings of the Museum are on permanent display. Collections of archival material, including correspondence, photographs and other records dealing with the acquisition and use of the museum's objects, are less known than the research collections themselves, as the visiting public seldom sees any archival material on display. Moreover, unless the formation and development of the museum's collection is the subject of the exhibition, there is little information in most exhibition label copy to indicate when and under what circumstances a particular item was added to the collection or the context from which it was taken. This research explores the idea that the box office appeal of seeing the "real stuff" in museums could be strengthened by access to the collection records and the related documentation that supports the evidence for that reality.

Proponents of the New Museology, as summarized by Stam (1993), emphasize the value of the museumís information base and suggest that the wider availability of information about the acquisition of collections is a way of fulfilling the social responsibility of museums. Those who see the museum visit as an interactive educational encounter such as Falk (1992) and Hooper-Greenhill (1994) emphasize the need to present contextual information about the cultural materials exhibited in museums as this context provides a basis for the visitor to make a connection with the content and concepts of the exhibition. Hooper-Greenhill (1994) speaks of the development of museum education programs, from the early part of the twentieth century when "curators struggled to establish museums as places where important objects were collected and cared for" to the end of the 20th century, when new ways of communication were developed using "discovery galleries, interactive exhibits and new relationships to new audiences." The display of and possible interaction with "real objects" has considerable power when viewed in the social situation of a museum visit. Museum visitors want to see some reflection of themselves in an exhibit. While not all museum visitors will have gone on a field expedition, many can relate to the theme of exploration. With the proposed Smart Web Exhibits (SWE), museum visitors will be able to view both the mounted specimen in the museum gallery and the correspondence and field photographs related to that specimen, sharing in the tedium of fruitless exploration and the excitement of actual discovery.

The development of the Smart Web Exhibits will explore a portion of the larger research question that asks what criteria museum visitors apply in choosing one exhibit over another. Experimental variables that the project hopes to isolate and test through the end user interface will include the level of interest as indicated by the visitor and the amount of time the visitor wants to spend looking at the exhibit.

Responding to the expectation that museums will provide access to information about their collections via their websites, this project plans to provide information through web-based kiosks on-site in the museum's exhibition galleries and on the museum's and the library's websites. The University Librariesí SWE will provide access to photographs, correspondence, lecture notes and published and unpublished papers of two early innovators in computer science, Herbert Simon and Alan Newell. Web access to the congressional papers of U.S. Senator John Heinz is already available on Carnegie Mellon University's site and will serve as a model for the Simon and Newell papers. The Carnegie Museum's SWE will display unpublished field notes, correspondence, photographs, published popular and scientific articles dealing with the Museum's early paleontological discoveries, especially that of Diplodocus carnegii. Evaluation techniques will be designed to explore the differences in the visitorís queries and other uses of the information system depending on whether they are onsite in the galleries or searching the collection information via the website.

While the popularity of behind-the-scenes tours of the museum's research collections suggests that visitors are interested in seeing more of a museumís collections, the specific research questions asked in this project include

  • Do museum visitors want to know more about exhibited objects?
  • How can this additional information be organized and presented for the visitor's use?
  • What type of background and contextual information would visitors select, such as correspondence or photographs, if given a choice?
  • How can archival documents be organized for presentation as an adjunct and online exhibition?
  • How does online use of these adjunct archival materials differ from their onsite use, particularly in frequency of selection?
  • Do online visitors spend more time searching for and viewing this associated archival material than onsite visitors?
  • What amenities, such as a separate seating area or a separate, sound-insulated space within the galleries, would increase use of this material onsite?

With funding from the Institute of Museum and Library Services, Carnegie Mellon University Libraries, the Carnegie Museum of Natural History and the School of Computer Science at Carnegie Mellon are collaborating to develop, document and disseminate prototypes of Smart Web Exhibits. Smart Web Exhibits (SWE) are designed to deliver information online, on target and on time to a diverse user community. The typical museum exhibition is a distillation of information gathered from many sources, including the correspondence of collectors, the records of acquisition, the dry bones of scientific description and subsequent publication - all part of the raw material from which the curatorial narrative of exhibitions are derived. Using the SWE, museum visitors will have direct access to this material. Like the user selecting material from the library catalog, the museum visitor may select material from the SWE to supplement the museum's exhibits. Visitors may either follow the suggested trails of an online exhibition or strike off on new ones of their own choosing by searching the document base on the basis of keyword, an individual's name or some other specific term.

Telling the Story

Vannevar Bush (1945) wrote that "a record, if it is to be useful to science, must be continuously extended, it must be stored and above all it must be consulted." Museum exhibitions are a form of record and represent the choice of curators, educators and exhibition designers to present the curatorial narrative. With the Smart Web Exhibit, we are combining the traditional selection of quality materials with access to the raw materials of scholarship, including unpublished manuscript sources that have been inaccessible and seldom included in exhibitions, except by reference. In the SWE's display, transcriptions of handwritten letters will accompany the page images, creating a self-documenting exhibition label. By including easily accessible original documents in the information base organized by the SWE, we are diversifying the voice of the exhibition. The Smart Web Exhibit goes on beyond the label copy by including the writings of those directly involved in discovering the objects displayed.

The SWE extends the museum exhibition, allowing the visitor to see the full text of the selected article or manuscript letter or the field photograph or video, not just the index terms in the catalog and the bibliographic entry for the retrieved text. The desirability of this immediacy of access is favorably noted by Gurian (1991).

We understand that not everyone wants the same additional information, but the individual who does want more information wants it then and there. Immediate access to information is satisfying to the audience; therefore the task is to provide information in the exhibition in a manner such that the audience knows it is available without being intrusive.

To tell the story of early paleontological discoveries at the Carnegie Museum of Natural History, we are creating a curatorial narrative by using the actual documents which recorded those discoveries. The inclusion of the extensive correspondence between the museum's administrators, scientists and field collectors, the required field expense reports and the photographs capture the museum's side of the story. Clippings from local newspapers and early scientific articles interpreting the discoveries add the popular and scientific dimensions. The complete text and illustrations of the first scientific publication describing the Carnegie's signature dinosaur, Diplodocus carnegii, is already available online and can now be put in context along with the correspondence between the Museum's director and the field collectors. This work was scanned as part of the celebration of the 100th anniversary of its discovery and is available at the Antique Books site ( For background on the Antique Books project, see Thibadeau and Benoit (1997). The online scrapbook approach has been tested on the Museum's website, with the slideshow "This is your life, Diplodocus Carnegii" ( Picture books of field photographs, recording actual field activities of the Museum's recent collecting seasons, are a popular feature of the Museum's simulated dinosaur dig, the Bonehunters's Quarry ( These albums continue the tradition of documenting the discoveries and field preparation of fossils, as well as the perils and pleasures of modern camp life.

In designing a museum exhibition, the earliest, and the most influential decisions involve the choice of audience for whom the exhibition is intended and the point of view or bias used in selecting and connecting the items used. Lavine (1991) questions the ability of an exhibition to maintain a single voice, given the evolution of scholarship within a discipline and the resultant divergence of voices and views. He suggests that different types of exhibitions have different modes of presentation and organization, noting that "exhibitions of social history are driven by interpretive ideas and are usually organized as narratives. As a result, objects are not made the primary focus of exhibits but retain significance as corroborative evidence." Providing the museum visitor with a choice of layers of information "of successively greater depth and detail" is one technique used by museums in providing diversity of information about the objects and ideas presented in their exhibitions. MacDonald (1991) notes that "Diversity is the key to a museum fulfilling its role as an information provider in the service of society." The Smart Web Exhibits may be one technique for providing those alternate voices.

Labels - What Do Visitors Want To Know?

Museum exhibition labels are the most evident types of the passive information sources provided to museum visitors. George Browne Goode (1895), writing on museum administration from the vantage point of assistant secretary of the Smithsonian Institution, recognized the positive value of labels, noting "an efficient educational museum may be described as a collection of instructive labels, each illustrated by a well-selected specimen."

Like many other aspects of museum exhibition technique, fashions on the length of labels have changed. Serrell (1996), whose writings on the nature and extent of the museum exhibition label are widely referred to, is quite firm on the length of the label, noting that "all types of labels should be kept as brief as possible." She does recognize that there are other types of interpretation, which can be used to inform and engage the visitors, ranging from handouts to extensive exhibition catalogs. "Videos, audio tours, computer databases and demonstrations by staff can provide other means and modes of presenting information and interpretation. All of these supplementary forms of interpretation will allow interpretive labels on walls or in cases to remain brief, as they should be." Gurian (1991) also mentions the "branching programs" of interactive videodiscs as a means of bringing additional information to the user without cluttering up the walls of the exhibition.

In selecting material for exhibition, as well as the label information accompanying them, curators and designers are guided by Serrel's warning (1996) "The museumís mandate to present information must be tempered by a responsiveness to what visitors are interested in and capable of receiving and processing."

Organizational Stratagies

The focus of the early phase of the Smart Web Exhibit project will be to develop techniques for matching the visitor to the material available. This may be done by beginning with the documents and identifying individual documents as being of interest to a visitor with a particular level of interest. Another technique would be to begin with the visitor, as in the typical library catalog search, allowing the visitor-as-user to search documents for keywords of interest to them.

One technique of categorizing the visitor's level of interest would be to include this information in the coding of the documents. One possible approach to the conditional interpretation of documents would be to organize them in hierarchical type trees as is normal with collections. Documents would be labeled as described in Thibadeau (1996) and the labels then formatted in XML format and manipulated using advanced XML scripting as described in Thibadeau, Balderas, Snyder, and Nestor (1999). Similar systems have been built for books at the National Academy Press ( and World Resources Institute (, for wide-ranging materials including videos (, and for art, art history, and architecture images at the Academic Image Cooperative (

Another method for matching the material and the visitor would be to apply a technique under development at Carnegie Mellon University Libraries. The Automated Reference Assistant (ARA) provides basic reference assistance to remote users by suggesting subject-specific reference tools based not only on the subject of the user's question, but also on the level of information desired. In the proposed model, resources are categorized as basic and advanced. If a user retrieves too many resources, only the ones identified by the system as basic are presented. The user may then request additional information, at which point the system provides the full list, including both basic and advanced resources.

One technique will follow the library search model, using the organizational strategy developed for the HELIOS project ( In this model, individual documents organized by traditional archival finding guides are scanned, the digitized images are converted to searchable text using Optical Character Recognition and the quality of the resultant page images and corresponding texts verified. Information on the archival organization of the documents is preserved and metadata is attached as part of the scanning and verification process. User will be able to search the material by noun phrases in the text or subject terms and limit the search by document type or date. A list of proposed document types for the Museum's collection is included in the appendix to this paper.

One of the features of the HELIOS project used to organize archival collections is its notion of "bundles." This is their term for identifying groups of different formats of material fastened together. Bundles are a technique which can be used to record the physical relatedness of documents. By this definition, there are "bundles" in the archival material supplied by the Carnegie Museum. In addition to maps and other field sketches inserted in the body of letters and field reports, there are attached photographs and the all-important field expense reports and receipts from workmen. In the manner of television documentaries, the text of the letters could be interrupted, and interpreted, by photographs which illustrate the action described in the letter.

The strategy that follows the exhibition organizational model is one in which the documents are grouped into thematic hierarchies. Access to the content of the archival documents available via the Smart Web Exhibit can be provided based on the userís expressed level of interest similar to the selection made for a traditional on-the-walls exhibition or in a less-structured way similar to that of users browsing a library catalog. In the first mode, the archivist or curator designs a hierarchical structure that categorizes documents into these exhibition schemes. In the second, the users search for documents that match the search keys that they have input.

In the case of the material dealing with the Carnegie paleontological discoveries, one proposed structure dealing with the nature of research in the field of paleontology is outlined below. Individual documents will be slotted into these categories as well as being flagged for the user's level of interest.

Figure 1: Thematic organization of archival material

In all cases, one of the measures of the success of the organizational and indexing strategies used in this project will be the ability to read a collection of letters dealing with a particular field discovery. The challenge will be to present these materials in chronological order, regardless of how the actual documents may have been scattered across different archival series or record groups. Here will be an opportunity for the museum visitor to share in the excitement of discovery.

Identifying The User

Without abandoning the added value of the curatorial narrative, the SWE is designed to adapt to the user's expressed level of interest. Serrell (1996) reports that the Denver Museum of Natural History used discovery, exploration and study as the three levels of information supplied to their visitors. The SWE proposes the following as its initial categories of these levels of interest.

  • simple curiosity, distinguished by an emphasis on graphics
  • general interest, the museum's standard level of presentation
  • intense technical detail, including scientific papers with specialized vocabulary

It may be important to categorize users according to the amount of time they have available to "view" the exhibit. The amount of available time could be matched with the levels of interest by offering a "highlights" tour for each interest level. Each set of highlights would present items determined to be of interest at the appropriate level.

Demographic breakdowns by the age of the visitor does not appear to be adequate - as age isn't necessarily an indicator of level of interest. Level of interest could translate roughly into age if need be, but documents would be identified as fitting into one of the three levels of interest, that is, basic curiosity, informed interest and technical interest. The basic curiosity material could include the "high points" that we would include in one of the system-supplied "tours" that would be in addition to user-selected searches.

Some mechanism would need to be developed to allow users at any level to have the opportunity to see more than is initially presented to them, by either searching the information base directly or changing their stated level of interest.

Challenges And Compromises

Largely due to maintenance considerations, the Museum prefers not to use keyboards as input devices in the majority of its computer-based exhibits, using instead trackballs to point to menu choices. This limited form of user query may not be adequate as a mechanism to select search strategies that take advantage of the textual richness of the archival documents included in the SWE. It may be that two different user query approaches will have to be developed -- one for the onsite user and one for the offsite or Web-based user. Users of the non-keyboard interface may lose subject access to much of the material unless some other sort of navigational tools is developed.

Members of the museum's education and exhibition staff are concerned that all newly developed exhibits appeal at some level to children. This will affect the identification of material and may necessitate the construction of a "just-for-kids" mode, possibly with a separate user interface.

The extent to which specimen-level data will be provided is still undecided, as there is a long tradition of limiting access to particular types of data, such as specific location information. This type of sensitive information may simply be omitted from the public display at all levels.

Other Aspects Of The Propsed Smart Web Exhibit

Not discussed in this paper are the plans for the evaluation of the user interface, the tools used to index and provide access to media other than text or still images or the process for obtaining reproduction rights to material under copyright. Links may be provided to other similar resources outside the individual museum or library.


The Smart Web Exhibit project is still in the specification and development phase. The challenge of providing information to a diverse audience online, on target and on time, will be met in the coming year.

Appendix: Document Types In The Museum's Archival Collection

The proposed document types are based on past user queries. Terms listed under each standard category are examples of documents that might be encountered in the collection that would be included under the standard category.

Annual reports

Departmental reports
Curatorial reports

Biographical notes




Expedition reports

Field reports

Field expenses

Field expense sheets
Receipts from workmen

Field notebooks

Financial records

Shipping invoices


Line art
Quarry maps

Mining claims


Newspaper clippings


Collecting permits


Field photographs
Installation photographs
Preparation photographs

Popular publications

Scientific publications



Bush, V. (1945). As we may think. Atlantic Monthly 176(1), 101-108.

Falk, J.H. and L. Dierking. (1992). The museum experience. Washington: Whalesback books.

Goode, G.B. (1895). The principles of museum administration. New York: Coultas & Volans.

Gurian, E.H. (1991). Noodling around with exhibition opportunities. In I. Karp and S. D. Lavine, (Eds.) Exhibiting cultures: the poetics and politics of museum display (pp. 176-190). Washington, DC: Smithsonian Institution Press.

Hooper-Greenhill, E. (1994). The past, the present and the future: museum education from the 1790s to the 1990s. In E. Hooper-Greenhill, ed. The educational role of the museum (pp. 258-262). London: Routledge.

Lavine, S.D. (1991), Museum practices. In I. Karp and S. D. Lavine, (Eds.) Exhibiting cultures: the poetics and politics of museum display (pp. 151-158). Washington, DC: Smithsonian Institution Press.

MacDonald, G.F. and S. Alsford. (1991). The museum as information utility, Museum management and curatorship 10, 305-311.

Serrell, B. (1996). Exhibit labels: an interpretive approach. Walnut Creek: Alto Mira Press.

Stam, D.C. (1993). The Informed Muse: the implications of ëThe New Museologyí for Museum Practice, Museum Management and Curatorship 12, 267-283.

Stearn, W.T. (1981). The Natural History Museum at South Kensington: a history of the British Museum (Natural History) 1753-1980. London: Heinemann.

Thibadeau, R. & Benoit, E. (1997). Antique books., D-Lib Magazine September 1997, change made at the request of the author, March 4, 1998, consulted February 14, 2000.

Thibadeau, R. (1996) Digital labels for digital libraries., D-Lib Magazine October 1996, consulted February 14, 2000.

Thibadeau, R., Balderas, J., Snyder, A. & Nestor, J. (1999). E-Commerce catalog construction: an experiment with programmable XML for dynamic documents., D-Lib Magazine 5(2) February 1999, definitive version released February 16, 1999 12:30 PM, consulted February 14, 2001.