Online Exhibitions: A Philosophy of Design and Technological Implementation

Marc Tinkler, Plumb Design Inc.
Michael Freedman, Plumb Design Inc.

The function of a virtual representation of a museum should mirror that of a physical museum. It should be a tool for sharing a museum's valuable assets with the rest of the world. By assets, I mean more than merely the artifacts that a museum possesses. A true museum is more than a building that houses a collection of objects, it is an institution that establishes and fosters a community. One of a museum's greatest assets is its ability to control what the audience sees, to set an agenda, to establish dialog, to provoke thought and discussion. An effective museum web site should do the same. It should be an extension of the institution, not merely an online information desk that provides helpful information. Of course, any web site should provide schedules, programs, contacts, and other practical information, but an effective web site can also serve the community in many other ways.

First and most basically, a museum web site should act as resource that supports the exhibitions currently on display in a museum. It should generate interest, encourage people in the community to visit the museum, and give them an idea of what they may see there. It should act as a resource for people who have recently visited the museum who want to acquire more knowledge about what they have just seen. Second, a museum web site should be a research tool, a catalog of the museum's holdings and an encyclopedia of information about and related to the collection of the museum. It can point to outside references, house research papers and allow the researcher to drill down into all of the museum's accumulated knowledge. Third, it should act as a community center. An effective web site can become a catalyst for moderated debate and discussion by allowing its visitors to participate in online forums, and by hosting guest lecturers that stimulate critical discussion. A visit to a museum is not a solitary experience, nor does a visit to the museum's site have to be. Fourth, and perhaps most importantly, a museum's web site should become an extension of the museum itself, and provide a virtual space for online exhibitions. It can use the Internet as a medium in and of itself, providing thought provoking exhibitions designed explicitly for the online experience. These exhibitions should exploit the qualities of the digital medium, and use them as opportunities to explore areas that would otherwise be impossible in a physical museum environment.

It is important to note that a web site hosted by a Museum is not in and of itself, an online exhibition. Like any exhibition, to be successful, an online exhibition should do more than put collections online; it should reveal the underlying relationships that transform a random collection of objects into a meaningful exhibition. It must be curated to create an educational and thought-provoking experience. This paper seeks to explore what makes an effective online exhibition. It will look at some of the tools available to online curators, and will examine the opportunities and challenges each of them present from the perspective of creating a compelling experience. Finally, the paper will examine a case study, and look at "Revealing Things", a recent online exhibit hosted by the Smithsonian Institution, to show the practical application of the ideas discussed.

There are two fundamental ways in which an online exhibit differs from a physical exhibit. Firstly, a computer will never be able to replicate, or even closely simulate, the tactile experience a visitor has in a museum when interfacing with an object. Secondly, visiting a web site is a more solitary experience than visiting a museum. Visiting even the most quiet, subdued museum is still a communal experience. Viewing an online exhibition will always be a singular dialog between the user and the computer. It is because of these differences that I argue that an online exhibition should not just be a virtual representation of a physical exhibit hall. An effective online exhibition exploits the hyper-real qualities of digital media and uses them to create a more compelling experience, rather than trying to mimic the structures that have evolved for use in the physical world. An effective exhibition views the limitations of the digital domain as opportunities rather than constraints.

Although there are many differences between physical and online exhibitions, there is still a lot that can be learned from physical exhibition design. Like a physical exhibit, an effective online exhibition is a choreography of the viewer's experience. The designer of a digital exhibit must think about rhythm and movement in many of the same ways that a physical exhibit designer organizes the rhythm and placement of objects within a museum. But an online exhibition can be alive and can actively encourage exploration; it can respond differently to different viewers. If a visitor does not interact with the material, the exhibition may act autonomously to bring an array of images and sounds to the viewer's attention. If the visitor explores actively, the exhibition may take a more passive role.

Like a physical exhibit, an effective exhibition does not rely on one method to present information. An online exhibit should take advantage of a variety of different technologies to provide a myriad of ways for the user to interact with the material. Each section of an exhibition can have its own pace, establishing a varied rhythm for the viewer that helps to hold their interest. Some pieces of an exhibit should be passive, allowing the user to explore an object or read related information. Others parts should take an active role, and rely on narrative to actively tell a story, or require the user to participate in a focused way.

Although variety is an important quality that contributes to effective exhibit design, it is always possible to go overboard. It is important to create an environment within an exhibition that is as seamless as possible. Establishing an overall theme and creating a mood makes it easier for the viewer to connect the information presented in different of parts of the exhibit and to make inferences about the meaning of the overall exhibition. If there is too much disparity between the look and feel of the different areas of an exhibit, it is likely that the viewer will get disoriented and will view the experience as a collection of random objects and not a meaningful experience. A seamless experience encourages exploration at a number of levels: visually, aurally, intellectually, and emotionally.

Tools

Old-fashioned HTML
HTML is still one of the easiest and most effective ways to communicate information. Unfortunately, because of bandwidth constraints, it is difficult to give the viewer of an online exhibition a tactile sense of objects in a collection. This fact makes "tombstones" and related information a much more integral part of any online exhibition. Stories, anecdotes, and other interesting information about objects can help to make an online exhibition interesting. But reading on a screen is hard, so it is important to break pages up into smaller segments. In general, scrollbars should be avoided, as they tend to break the seamlessness of an experience.

JavaScript
For better or for worse, JavaScript is the glue that holds effective, innovative online experiences together. JavaScript allows you to make components on a page work together that would normally exist independently of each other. It is far and away the most important tool in an exhibit designer's arsenal, because it allows you to combine other tools in ways that their designers could not have imagined.

Embedded Audio, MIDI & Beatnik
One advantage of the Web being a solitary experience is that you don't have to give your visitors headphones in order to avoid annoying other people. When used appropriately, audio can greatly enhance an online exhibit. Background audio such as sample loops, MIDI, or Beatnik files allows the designer to create a mood that can permeate an entire site. Real Audio is perfect for narration, and WAV files are best for interaction triggered sound. Not all sound has to be triggered by the user pressing a "play" button. Sound could start and stop on its own accord, or in response to a seemingly unrelated event. In general, try to integrate the sound into the body of the exhibit. Avoid using the pre-packaged "radio controls" that the browser or plug-in offers. These days, most sound features can be accessed through JavaScript, which gives you much greater control, and lets the user concentrate on the material, not the technology.

QuickTime VR and PhotoVista
QuickTime VR and PhotoVista can perform two functions. First, they allow the viewer to rotate an object around an axis, and view it in three dimensions. Second, they can place the viewer in the center of a virtual sphere, allowing the user to pan in any direction, creating an interesting pseudo-virtual-reality experience. These tools are good for allowing the user to explore a small artifact like a piece of jewelry, or experience an immense space like the inside of the Pantheon. They are both fairly bandwidth intensive and should be used sparingly. Both formats allow the designer to place hotspots that either link to other pages, or transport you to a different view. One interesting idea is to use JavaScript to make the hotspots trigger external events, or to use external events such as audio narration to trigger a QuickTime VR or PhotoVista display to rotate to a particular position.

Internet Imaging Protocol
Internet Imaging Protocol (IIP) is not getting a lot of press, but it holds a lot of promise for the Museum community. It allows the user to download a high-resolution image in small pieces rather than all at once. This means that a user can get an overall view of an artifact and then zoom to near-infinite resolution, downloading only the information that they need. This technology, when used appropriately, could, for instance, allow visitors to zoom into a Van Gogh painting on the Web and study the exquisite detail of his brushwork. Unfortunately, IIP is not quite ready for primetime. A plug-in is required, generally the interfaces are quite clunky, and there are no API's that allow you to manipulate the images through JavaScript. If combined with QuickTime VR or FlashPix, IIP could turn out to be quite promising.

Dynamic HTML
Dynamic HTML fully realizes the potential that JavaScript and HTML introduced. It allows the designer to manipulate all aspects of an HTML document in real time. The possibilities that Dynamic HTML has introduced are limitless. The designer can create layers that appear, move, and disappear based on user interaction or precise timing. Because the pages are still HTML, they can be created "on the fly" using a database or a server-side script. This means that the experience can be dynamically altered for each individual user. An item in the exhibition may not only look differently, for example, if a user has seen this item before, but it can behave differently.

Shockwave Flash
Shockwave Flash is unique in that it allows the designer to create low-bandwidth, "high energy" pieces that download very fast, but have an almost theatrical quality. Flash is great for narrative pieces, because it allows you to create very long sequences that can be fast to download while having an enormous experiential value. You can integrate streaming audio with Flash and, using JavaScript, you can make flash trigger any event that may happen elsewhere on the page.

Shockwave for Director
In the same family as Flash, is Shockwave for Director, which is best at creating stand-alone games, puzzles and other educational pieces. Shockwave files are larger than Flash files, but you can do almost anything you can imagine using Director. One downside to Shockwave and Flash is that they are static files. Changing the behavior or content within a piece is difficult, and requires "re-building" them. Pieces that are dynamically generated or altered by multiple users are possible to create, but other technologies like Java may be more appropriate for complex network oriented tasks.

Java
If you can dream it, you can do it in Java. On the positive side, since Java is a full-fledged programming language, theoretically, anything is possible. Most browsers support it, so your users don't have to download any plug-ins. On the negative side, since Java is a full-fledged programming language, in order to develop anything custom-made, you need to have access to experienced programmers to really make it worthwhile. Java also suffers from being a relatively young technology. It is not always consistent, reliable, or as cross-platform as advertised, but if you are willing to spend the time, Java can really pay off.

This is just a small sampling of the technologies that are available to the on-line exhibit designer today. There are countless other tools out there, and new technologies are coming to market every day. What is important to remember, is that every tool, no matter how new and cutting-edge, should enrich the viewer's experience. Again, the goal of an effective exhibit is to create an experience whose meaning and impact is independent of the technology used to implement it.

Case Study: Revealing Things

The goal of the exhibition was to convey to the visitor that everyday objects are more than the sum of their function and use; everyday objects are important signifiers of cultural and personal meaning. Ideally, after experiencing "Revealing Things", visitors will begin to interpret the messages and meanings that everyday objects communicate, and ultimately will gain an increased awareness of the objects that surround them.

Saussure said that a sign was two-parted: the signifier and the signified. The signifier and the signified have an arbitrary, culturally specific relationship. A non-arbitrary relationship between the signifier and the signified is symbolic. Objects are signs or signifiers of meaning when viewed in contrast to other objects. He said, "Nothing has meaning in itself and relationships are all important."

If we accept Saussure's hypothesis, a full understanding of any artifact can only be accomplished by absorbing and comprehending the culture in which that object was made and used. Thus, meaning is attached to an artifact by understanding it as a node in a larger context.

In a physical exhibit, the curator would have to deconstruct the meaning of every object in order to classify it and determine its position in the exhibition. In a physical exhibit, for example, a pair of jeans worn to a rock-and-roll concert either can be placed with other articles of clothing, or with other objects associated with the music of the time. Our goal was to take advantage of the possibilities of the digital medium in order to present the multi-dimensional nature of the relationships between objects, history, and personal meaning. Our solution was not a deconstruction in response to the complexity of the information, but rather a reconstruction of the complex network of meaning that surrounds every object in the collection.

The exhibit uses a Java-based technology called Thinkmap to create a dynamic interface that demonstrates the underlying connections among the objects. Thinkmap animates data, displaying the underlying connections between discrete items. The relationships between items in a Thinkmap database are not absolute and over-determined, but dynamic and subjectively conditioned by the viewer. Information-publishers who use Thinkmap can create a series of flexible rules that define object behavior. These rules ultimately determine the display's look-and-feel. The Thinkmap interface we created enables visitors to choose their path through the exhibition. A visitor examining a lantern made by a Japanese-American in the Manzanar Relocation Center during World War II, for example, could pick a number of different methods to view other objects. She could study objects from the same period (the table setting from a dinner given during Pearl Harbor), "decorative objects" (a tiffany vase), or objects whose stories have similar themes (a cookie mold made by Arab-American immigrants). This flexibility allows visitors to follow their interests and enables curators to display objects in a variety of contexts. In a museum, objects can be in only one place at one time. In Revealing Things, objects are dynamically positioned depending on the preference of the user. The curator becomes a guide, building the underlying structure. The visitor designs the exhibition. Just as a person can walk through a museum, choosing objects that interest her, a user of Revealing Things can pick a few objects and delve deeper into those that she finds particularly attractive.

Each object in the collection is presented in a different way. Most objects are presented using a single page of HTML that starts off with a story or anecdote. Audio is used to reenforce important parts of the narrative and helps to draw users into the second-level pages that delve deeper into the object's meaning. Other objects are presented using Flash movies that tell a story or provide historical background. In the final exhibition, there will be QuickTime VR or PhotoVista views of artifacts, as well as Shockwave games and puzzles to encourage user interaction. The ultimate goal is to present a variety of experiences, to establish a rhythm of passive and active periods for the viewer.

The result not only serves to link disparate objects in novel ways, but also entices people into the exhibit to explore. The interface, written in Java, is always moving, and encourages people to click and explore. What could have been a static exhibition becomes an expandable, lively, interactive, learning platform, where visitors are engaged in seeing and contributing to their own learning experience.

Conclusion