Join our Mailing List.
Published: March 1999.
On-line Exhibit Design: The Socio-Technological Impact of Building a Museum over the World Wide WebPaul Marty, Spurlock Museum, University of Illinois, USA
IntroductionThe museum offers a unique environment from which to study the way knowledge is accumulated, analyzed, and distributed by information professionals. Researchers world-wide have recognized the museum environment to be an emerging important field of study for information science (Rayward, 1998; Thomas & Mintz, 1998; Jones-Garmil, 1997; MacDonald, 1992). Today, as museums work to develop digital collections of artifact images and data, new needs and challenges arise (Gladney, 1998; Lucas, 1998; Moen, 1998; Samuels, 1998). As these opportunities for growth are met and evaluated, museum professionals and information scientists alike are afforded the opportunity to explore ways in which advanced information technology can affect collaborative activities within the museum.
In 1996, we inaugurated an ethnographically-informed study of the Spurlock Museum at the University of Illinois in which we are examining how the Spurlock's professionals and patrons have been affected by the introduction of advanced information technology into the museum setting. The research agenda of this study has been the design and development of information systems that support the collaborative production and use of the specialized knowledge contained within the museum environment. One of the more interesting aspects of this study has surrounded the development of advanced information technology to support asynchronous collaboration between curators and exhibit designers. This paper presents the findings of this individual aspect of the overall comprehensive study of museum informatics currently under way at the Spurlock Museum.
The Spurlock Museum: a holistic approach to museum informaticsThe Spurlock Museum at the University of Illinois in Urbana-Champaign offers an ideal environment for research into museum informatics. The Spurlock is a museum of world history and culture; it possesses a collection of approximately 40,000 cultural and ethnographic artifacts from diverse cultures and varied historical time periods, ranging from ancient Sumeria to modern America. Staff members at the museum include ten permanent employees, several dozen student assistants, and a handful of affiliated academic curators. Thus, in its size, the Spurlock represents a comprehensive, yet still manageable and readily understood research environment. Moreover, as a university museum, the Spurlock is situated within an extremely supportive academic environment that encourages innovative research into information science, such as this study of technology-enhanced collaboration in a museum environment.
In 1996, the museum initiated a five-year program designed to prepare the museum, its staff, and its collections for a move across campus to a brand-new facility scheduled to open in the year 2001. As part of this program, the museum implemented a comprehensive reorganization of its use of advanced information technology in the fall of 1996. Today, the technologies and information systems in use at the Spurlock include the following features:
Information science and exhibit design: the need to communicateOver the past two and a half years, we have been able to observe the effect of these new technologies upon the social infrastructure of the museum professionals at the Spurlock. It is interesting to note that what initially prompted such far-reaching technological innovations was in fact the need to provide a more effective means of communication between the staff members at the Spurlock Museum and the exhibit designers planning the new museum. In 1996, it became clear that the task of designing a new facility had placed unfamiliar information resource management demands on the museum staff. The exhibit designers needed access to accurate, comprehensive, and up-to-date data pertaining to artifacts in the museum's collection. However, no reliable means of communication existed that would allow Spurlock staff members to communicate pertinent artifact information-detailed artifact specifications as well as information about where in the new museum artifacts would be displayed-to the designers on a regular basis.
This need initiated a process that would shape the role of the Spurlock's dedicated information systems vis-ˆ-vis the social infrastructure of the museum itself, a classic example of social informatics (Kling, 1998). In this process, information systems were developed that would support the inherent needs of the museum's social environment. As the process proceeded, the museum's social structure was simultaneously redefined in light of the museum's new technological dimension. In this instance, the need to exchange regular and reliable information between museum curators and exhibit designers prompted the technological innovations necessary to make this communication possible. In turn, the implementation of these new technologies altered the social infrastructure of the museum by creating an environment that encouraged collaborative work and required a heightened dedication to the tasks of data entry and dissemination of artifact data.
In 1997, once the new technologies were in place, the museum's registrar began a comprehensive program to re-inventory every artifact in the museum's collection and enter this data into the museum's newly developed collections management system. Dozens of undergraduate assistants were hired to make this task move forward as quickly as possible, while sacrificing neither accuracy nor detail. The resulting artifact data (combined with data from supporting systems) was then dynamically linked to the museum's web site to make the relevant portions of the museum's entire collections management system accessible to the exhibit designers over the Internet in an accurate, instantaneous, and up-to-date manner.
By dynamically linking artifact data contained within the museum's databases to exhibit information contained on the museum's web site, the Spurlock gained the flexibility necessary to portray page after page of detailed artifact information on-line while dynamically organizing this data in a fashion that reproduced the current plans of the exhibit designers. Through this technology, the designers could virtually examine any artifact to be displayed in the new museum, look at its digital photograph, read a textual description, or peruse its particular details and specifications. Moreover, they could examine this data either by directly searching the museum's databases for any particular artifact or by accessing sets of artifacts automatically grouped by the information system based on where the artifacts were to be displayed in the new facility.
This new system, therefore, presented the ideal environment in which to study the social impact of information technology on the collaborative activities of the curators and exhibit designers planning the new Spurlock facility. These individuals now had access to an on-line technological forum that supported many aspects of their collaborative activities. As a result, the system had two effects on the museum's social environment: 1) it encouraged extensive collaboration and data exchange between curators and designers, and 2) it required the creation of a virtual representation of the new museum that could accurately maintain data on the current status of the proposed galleries from the point of view of both curators and designers. We will now examine these two effects in more detail.
On-line collaboration between curators and designersDuring the course of this study, we have observed many different types of collaboration between curators, exhibit designers, and other affiliated scholars engaged in the task of selecting artifacts for display in the new Spurlock facility. These collaborative activities have been encouraged by the very technology that has made the collaboration possible. When working with the Spurlock's information systems, each curator is required to carefully mark each record in the main artifact database to indicate whether the artifact represented by that record is to be displayed in the new museum. It is vital that this data be accurately maintained so that the designers, scholars, and other curators involved in planning the museum's galleries can remain informed as to the exhibit status of each artifact. This information is then made available over the Internet by dynamically linking the museum's collection management database systems with on-line information resources for the exhibit designers and other experts participating in the design process.
When these exhibit resources are accessed over the Internet, requested artifacts are presented in a concise format, listing the artifact's name, accession number, measurements, and visual description (see Figure 1). Clicking on the name of the artifact will retrieve a detailed artifact record, including a larger image, nomenclature specifications, physical analysis, geographical data, scholarly research notes, and so on.
In this way, the exhibit designers can remain abreast of the curators' plans for the new museum merely by accessing the available on-line information.
Example. The exhibit designers are working on designing a particular section of the new Spurlock facility; they need to know which artifacts the curators have selected for that particular exhibit. Accessing the exhibit listings over the Internet, they enter the code for that section of the museum and immediately receive an accurate and up-to-date list of all artifacts scheduled for display in that exhibit. By regularly checking this site, the designers can follow the evolving thoughts of the curators through time and therefore prepare their own suggestions for the curators asynchronously.The accuracy and immediate availability of information pertaining to artifact selections helps reduce inconsistencies in data files and ensures that no one artifact is claimed by two different curators for two different locations in the new museum.
Example. A curator for the European gallery is selecting artifacts for display. When he retrieves the record for one particular artifact, he discovers that it has already been marked for display elsewhere in the new facility. A click of the mouse tells him who made the selection, when, and where this individual believes the artifact should be located. The curator can now discuss this selection with the other individual and the two of them can decide the appropriate location for this artifact.The ease of use which this system affords the exhibit designers also offers them a simple means of receiving answers to straightforward questions about artifact specifications, without the need to contact the museum staff by phone or fax or to travel to the museum itself.
Example. While designing cases for the Ancient Mediterranean gallery, the exhibit designers realize that they do not know the size of a particular artifact: a large bronze money chest from Italy. Since this artifact is going into a display case, it is imperative that the exhibit designers know the exact dimensions of the piece. Instead of having to spend time locating a museum staff member who would know this information, the designers can access the record for this artifact directly over the Internet and download its dimensions with their Internet browser.The museum maintains relationships with many external experts, such as anthropologists, historians, and classicists. Some of these external experts, usually university professors affiliated with the museum, perform the role of "zero-time curators" for the museum and have been valuable consultants for the museum's full-time curators in the past. Now, with the ability to collaborate on-line in artifact selection and exhibit design, these experts can assist museum staff members without leaving their own departments.
Example. A professor in the anthropology department is asked to review a list of Indonesian artifacts being considered for a particular exhibit in the Asian gallery. A museum curator sends this professor an email containing a pre-scripted URL which links automatically to the museum's database systems. With one click on this embedded URL, the professor accesses the entire list of artifacts under consideration and can review them at her leisure. She then emails her recommendations back to the museum staff.The system also makes it simple for multiple individuals working on the same exhibit or gallery to view groupings of artifacts on-line independently, thereby collaborating asynchronously, confident that the data they are viewing is accurate and timely.
Example. As two different curators work on selecting artifacts for the African gallery, they can each follow the actions of the other by accessing the on-line exhibit guide. Even though they may have been marking artifacts asynchronously and in separate locations, the web site will merge together their selections automatically and dynamically so that each can see a single list of selected artifacts.As instances of these types of collaboration grew more frequent, it became necessary to develop a system to maintain groupings of artifacts in a logical manner. The obvious choice was to group the artifact selections around the organization schemes for the new galleries, which were being developed at the same time as artifact selections were being made. This led to the creation of the Virtual Spurlock Museum.
The Virtual Spurlock Museum: a digital representation of a future facilityThe task of deciding how artifacts chosen for display will be organized and where in the new museum they will be located is just as complex a task as deciding which artifacts will be displayed in the first place. It is therefore very important that every individual working on the design of any new museum have access to up-to-date information about artifact arrangements as well as artifact specifications.
As the curatorial staff members plan each gallery in the new museum, they specify electronically not only which artifacts will be displayed but where they will be located in the new building. Then, when the exhibit designers access the on-line version of the new Spurlock Museum, the system links to the museum's internal databases and automatically arranges the digital artifacts on-line according to the available gallery schematics. In this fashion, accurate and timely information about where the various artifacts will be exhibited in the new facility is dynamically displayed to architects, exhibit designers, and interested members of the general public. Thus, years before the Spurlock actually opens to the public, interactive technologies allow electronic visitors to browse the plans for the new museum and virtually visualize artifact locations just as they will be in the new building.
Currently, given the fluid nature of the design process and the rapidity with which the proposed gallery designs change, we have implemented a primarily text-based virtual museum system. This system conforms to the following navigational format: upon entry to the museum, the visitor receives a textual list of the museum's galleries, along with brief descriptions. The visitor then selects a gallery, and is given a new screen of data, which includes a textual description of the gallery's theme and general orientation as well as a list of exhibits contained within the gallery. The visitor then chooses an exhibit and is presented with a textual description of the exhibit's message and sees a list of artifacts to be displayed in that exhibit. This list follows the concise artifact specification format illustrated above (see Figure 1); clicking on any particular artifact will retrieve a more detailed artifact record. Depending on the navigational options selected by the curators, the visitor will be able to proceed from exhibit to exhibit, learning what topics will be covered in each gallery and what artifacts will be displayed in each exhibit.
This logical choice for an organizational scheme enables the designers to view entire galleries and exhibits, display by display and case by case. Each set of artifacts is grouped together as they will be in the new facility, making it easy for the designers to visualize the settings and plan the casework. Moreover, the virtual museum has the added benefit of providing the general public with a way of following the planning for the new museum, keeping the local community aware of the Spurlock's progress.
Building a virtual museum: linking museum databases to the world wide webWe turn now to a brief examination of the technology behind the virtual museum and how the databases were dynamically linked to the museum's web site. First and foremost, we must dispel the misconception that implementing collaborative technologies of the sort described above is difficult to initiate, expensive to create, and hard to maintain. The Spurlock Museum, like most small, underfunded university museums, has very little money and few technological resources. Moreover, when we began this project in 1996, we had to proceed under certain conditions which had the potential to limit our technological possibilities: the museum had already invested in certain software and hardware systems, which we had no choice but to use; the resulting system had to operate on a network of multiple Macintosh PowerPC computers and be developed in FileMaker Pro 3.0.
That we were able to set up a system this powerful and this quickly is a testament to the growing ease of linking virtually any database system running on virtually any platform to the Internet with a modicum of effort. All that is required is persistence to locate the appropriate tools, some experience in database design, and adequate time to develop the system. At the Spurlock Museum, we utilize the FileMaker Pro database interface program called Lasso, produced by Blue World Communications (www.blueworld.com). Lasso is a robust yet simple to use CGI that provides a set of extensions to HTML called LDML, the Lasso Dynamic Markup Language. By inserting the appropriate LDML commands into our HTML source code, we are able to access any given record in any given database and organize the results in a fashion appropriate to any given user, exhibit designer, curator, or scholar.
In order to explain in as simple a manner as possible how this is done, we will first examine the relevant fields in the two main databases that control the virtual museum: Artifacts, which contains one record per artifact in the museum's collection, and Nodes, which contains one record per area of the new Spurlock facility.
ArtifactsArtifacts, the primary collections management database for the Spurlock Museum, contains over 100 fields of data for each artifact record. However, only a handful of these fields are used by the Virtual Spurlock Museum (see Table 1). Fields such as the artifact's accession number, name, measurements, visual description, and graphic image are used to identify and display the artifact in the virtual museum.
Two primary fields, Status and Node, are used to indicate whether each artifact will be displayed in the new facility, and if so, where. Status must be set to any one of a list of possible options, of which the only two relevant settings are "Display" or "Storage." Naturally, for an artifact scheduled for display in the new museum, the status field will be set to "Display." Node contains a code number that identifies a particular section of the new museum; each area of the new facility has been assigned a unique identification number (see below) which is used for sorting the artifacts scheduled for display. When the curator specifies where the artifact is to be displayed, it is this field (the Node field) that is set.
A Remarks field allows the curators to indicate the reasons a particular node was chosen for any one artifact and therefore represents a purely internal method of managing collaboration between multiple curators working on the same set of artifacts. Finally, the fields WorkingSet1 and WorkingSet2 function the same way as field Node but allow for the temporary assignment of an artifact to a given area of the museum without having to permanently alter its primary node setting.
Table 1, Relevant fields in database Artifacts
NodesNodes, the database which governs the layout of the virtual museum, contains approximately a dozen fields per record, of which only eight are relevant to this discussion (see Table 2). This database represents the current plans of the curators and exhibit designers; each record in this database represents a given section of the new museum. By accessing any particular node record, we can obtain data on how this node was reached, what other nodes this node can access, what exhibit topic this node represents, and what artifacts are on display in this node.
The Nodes database is in a state of continual development; as the curators' plans for the galleries develop, nodes are merged, dissolved, or restructured as exhibits are created, shifted, or removed. As the Nodes database changes, so does the Node field contained within the Artifacts database as curators virtually move artifacts to represent their new location in the gallery plans. Clearly, only a dynamic database system could accurately represent on-line the ever-changing nature of the gallery design process.
In this database, two primary fields, NodeNumber and NodeName, identify each node record and its place in the new Spurlock Museum facility. PriorNode and NextNode provide default navigational points for going forwards or backwards in the virtual museum (it may help to imagine a binary tree, with each leaf representing a node). SubNodesExist and ArtifactsExist are indicators that tell the virtual museum engine to look for and compile a list of destination nodes and artifact records respectively. Finally, two fields, DisplayHTML and TextualDescription, provide the text and other elements that will describe this node-its topic, theme, and place in the museum-to the virtual museum visitor. Note that the first of these fields is HTML-enabled, allowing the insertion of links, tables, graphics, or multimedia elements into the node display without the need to edit the primary source code for the virtual museum.
Table 2, Relevant fields in database Nodes
We will now briefly examine how the LDML extensions to HTML allow the Virtual Spurlock Museum engine to access the relevant databases and thereby dynamically create each page of the virtual museum. Here, it is important to understand that we use the Lasso CGI as a preprocessor for the HTML code that runs our virtual museum; this means that for any given location in the virtual museum, we have already executed a search of the Nodes database and retrieved all the data for the Node which the visitor is examining at that time. The very first search, which is executed when the visitor elects to enter the virtual museum, is hardwired into a web page on the museum's server as a pre-scripted URL. When clicked, this pre-defined search sends a command to the Nodes database and retrieves the very first node in the node tree, which represents the entryway to the Spurlock Museum. From this point on, all node accesses are dynamic and determined by the visitor to the virtual museum.
Once the virtual museum engine knows which node the visitor is currently browsing, it is a simple matter to display the contents of fields owned by that node record. The LDML extension follows the format of [field: name_of_field] as shown in Code Fragment 1, which displays the description of the current node as contained in two fields. Note that the tag "raw" accounts for the possible presence of HTML codes in the DisplayHTML field.
Code Fragment 1, Display descriptions of node
Code Fragment 2, Create link to default next node
By introducing a second search, within the confines of the first search, it is possible to have access to variables from new records while retaining information about the identification of the original record. This is accomplished through the use of an inline search within a given HTML file. In Code Fragment 3, we access and display a list of all subsequent nodes to which the visitor can travel from any given current node. The code first checks to be sure that subnodes exist for the current node. If so, it executes an inline search on the nodes database, looking for all records where the PriorNode (or Parent Node) field matches the current node's NodeNumber; this search returns a list of all nodes for which our original node is the parent. If the inline_result returns no error, the code then generates a list item for each record found, presenting the name of each new node as a link to a command. This command uses the recid function to pass the identification of the selected node to "node.html" (the primary controlling HTML file) which will dynamically construct a new page for that node. Thus, by clicking on the node's name, the visitor can travel from node to node in the virtual museum.
Code Fragment 3, Display list of subsequent next node options
Finally, in Code Fragment 4, we use the inline command to execute a search on a different database (Artifacts) while still retaining information about our current node location. After checking to be sure that artifacts are available for this node, the code performs an inline search of the Artifacts database to find all artifact records where the field Node matches the current node's NodeNumber. If the inline search is performed without error, the code then displays an entry for each artifact record located. This entry, with a thumbnail image on one side, and the artifact's name, accession number, measurements, and visual description on the other, is in fact equivalent to the concise artifact listing format as seen in Figure 1 above. The artifact's name is linked to a command to pass the record's identification code (recid) through to a new HTML file ("details.html") which will present the detailed artifact record (containing many more fields) to the virtual visitor.
Code Fragment 4, Display list of artifacts in node
Conclusions and future directionsWe are currently working on refining the Virtual Spurlock Museum in terms of both its capabilities and its interface. Specifically, we are extending the capabilities of the Nodes database to accommodate the introduction of Tours, scripted narratives or trails through the museum's nodes. While traveling a tour, visitors will be guided by an electronic tour guide (which we are calling a CyberDocent) which will present a tour-specific explanation of each node visited. Significantly, these tours will allow the Education department within the Spurlock Museum to develop, test, and refine educational tours of the new facility two full years before the museum opens to the public.
The interface is also being re-designed to incorporate new graphic elements we are beginning to receive from our exhibit designers. When we began this project two years ago, we had access to no materials beyond basic gallery outlines and simple text, and even that was in a state of continual change. Now, as gallery plans are finalized, we have access to detailed artist's renderings depicting the exhibits in each gallery. It is our intention to incorporate these graphics into the virtual museum as a mapped graphic interface, allowing the visitor to navigate the galleries merely by clicking on an interesting area of the illustrations.
When discussing the Virtual Spurlock Museum in such detail, it is perhaps easy to forget that this virtual museum is modeling a facility that is still under construction, that is not expected to open to the public before 2001. It is, after all, one thing to build a virtual museum when the physical building already exists; it is quite another to create a virtual museum as part of the design process while constructing a brand-new museum facility. It requires a totally new set of skills as well as the understanding that access to information will be limited and continually changing in an often exasperating manner. And yet we highly recommend incorporating dynamic access to databases to anyone designing virtual exhibits for a new museum building.
This is, of course, not the first paper to recommend integrating collections management systems with on-line exhibits in a dynamic fashion; at the very first annual Museums and the Web conference, Howard Besser (1997) recommended the very same. But we may be the first to recommend doing it 1) for a museum that does not yet physically exist, and 2) as part of a holistic approach toward museum informatics. The Virtual Spurlock Museum is merely one part of an integrated system designed to support the museum professionals at the Spurlock in each of their regular activities, only one of which is exhibit design. Throughout the course of this paper, we have examined only one aspect of the Spurlock's overall program in museum informatics. It is our hope that by examining the social and technological systems of museum informatics and the way in which they affect communication between museum professionals, exhibit designers, and other affiliated scholars, we can prove the value of museum informatics as a tool for advancing and supporting collaborative activities for all museum functions.
Besser, H. (1997). Integrating collections management information into online exhibits: the web as facilitator for linking two separate processes. In D. Bearman & J. Trant (Eds.) Museums and the Web, Selected paper from Museums and the Web 97 (pp. 201-205). Pittsburgh: Archives & Museum Informatics.
Gladney, H. (1998). Digital access to antiquities. Communications of the ACM, 41 (4), 49-57.
Hooper-Greenhill, E. (1995). Museum, media, message. London: Routledge.
Hooper-Greenhill, E. (1992). Museums and the shaping of knowledge. London: Routledge.
Hughes, J.; King, V.; Rodden, T.; Andersen, H. (1994). Moving out from the control room: ethnography in system design. In Proceedings of CSCW â94. (pp. 429-439). New York: ACM Press.
Jones-Garmil, K. (Ed.) (1997). The wired museum: emerging technologies and changing paradigms. Washington DC: American Association of Museums.
Kling, R; Rosenbaum, H; Hart, C. (1998). Social informatics in information science: an introduction. Journal of the American Society for Information Science, 49 (12), 1047-1052.
Lucas, R. (1998). Digital imaging: how libraries, museums, and other image banks are managing a digital world. Syllabus. 11 (9): 39-46.
MacDonald, G. F. (1992). Change and challenge: museums in the information society. In Museums and Communities. Washington: Smithsonian Institution Press, 158-181.
MacDonald, G. F. (1991). The museum as information utility. Museum Management and Curatorship, 10, 305-311.
Moen, W. (1998). Accessing distributed cultural heritage information. Communications of the ACM, 41 (4), 45-48.
Rayward, W. B. (1998). Electronic information and the functional integration of libraries, museums and archives. In E. Higgs (Ed.), History and Electronic Artefacts. (pp. 207-224). Oxford: Oxford University Press.
Samuels, G. (1998). The museum digital licensing collective: serving all museums. Museum News, 77 (3), 41-65.
Sommerville, I; Rodden, T; Sawyer, P; Bentley, R; Twidale, M. (1992). Integrating ethnography into the requirements engineering process. In IEEE International Symposium on Requirements Engineering. (pp. 165-173). Los Alamitos, CA: IEEE Computer Society Press.
Thomas, S. & Mintz, A. (Eds.) (1998). The virtual and the real: media in the museum. Washington: The American Association of Museums.
Twidale, M. B.; Randall, D.; & Bentley, R. (1994). Situated evaluation for cooperative systems. Transcending Boundaries: Proceedings of CSCW â94. Chapel Hill: ACM Press.