Inhabiting the virtual city
Judith S. Donath

4 Sociable information spaces
Physical spaces have important social functions. They demarcate public and private, formal and casual, natural and man-made. ``A vivid and integrated physical setting, capable of producing a sharp image, plays a social role as well. It can furnish the raw material for the symbols and collective memories of group communication.'' ( Lynch 1960) My argument in this section is that a vivid and integrated virtual setting - one made of information rather than rocks or benches or bricks - can play a similar role.

Published media - books, television shows, newspapers, etc. - can be thought of as neighborhoods, inhabited by the publication's readers and viewers. Like real-world neighborhoods, virtual neighborhoods are both a community's home and an important source of its cultural vocabulary.

Neighborhoods are the spatial manifestation of community, ``a place where people live and associate on the basis of certain shared myths, or assumptions about the world'' (Greenbie 1981). The audience for a publication, like the inhabitants of a neighborhood, is often a homogenous group, with shared myths and assumptions about the world. As Fiske put it: ``Choosing texts is choosing social allegiances'' (Fiske 1989a). An action movie, a scholarly journal, a fundamentalist website - all are ``texts'' that bring together a group of people united by common interests and concerns; in the world of information, publications are neighborhood enclaves, places where a community gathers.

Yet as many urban theorists have noted, a neighborhood is not merely a container for the community. As Lynch pointed out, its architecture, its history, and the things within it - stores, parks, signs, etc. - provide the community with the settings and symbols for their collective memory. The virtual neighborhood is even more strongly the source of a community's cultural vocabulary: from song lyrics to political rhetoric, our cultural references are drawn from common media experience. Even our understanding of facial expressions may be drawn from this source. Ekman postulated that ``John Wayne's look of anger on the TV set - not man's evolutionary history - may be responsible for people's ability to recognize the same anger expression across cultures'' (Ekman 1973).

Today, the neighborhood as a source of cultural material has been surpassed by media. In No Sense of Place Meyrowitz wrote:

Electronic media have had a tremendous impact on group identity by undermining the relationship between physical location and information access... Electronic media begin to override group identities based on ``co-presence,'' and they create many new forms of access and ``association'' that have little to do with physical location. ( Meyrowitz 1985)
But the neighborhood still provides a powerful metaphor. And as such, it points out an important thing absent from traditional media - the presence of people. Traditional media is not in any real sense ``inhabited''. The viewers of a television program or the readers of a magazine share a common experience in collective isolation.

Publishing media (TV, books, etc.) which are rich in cultural material, have traditionally been separate from communication media (telephones, email, newsgroups, etc.), which is where the people are. Existing technologies enforce this segregation: the newspaper is inert; the TV set only receives. Today these technological constraints are dissolving: the computers that house the text-based virtual-communities are now also capable of playing movies. Yet the point is not simply to unify the machinery. Playing a movie in one window while writing a letter in another does not constitute the integration of publishing and communication. Rather, the point is to create environments that combine a rich information landscape with the ability to communicate with others - information spaces that provide a context for community.

The dissolution of the boundaries between publication and communication is most noticeable on the Web. For example, discussion forums accompany a variety of Web-published materials: City of Bits ( Mitchell 1995), William Mitchell's book about virtual society, is available on-line, with section by section reader commentary; Hotwired ( HotWired 1996), the electronic version of Wired magazine, includes a wide range of ever-changing reader forums. These discussions occur within an information context provided not only by the host publication, but by the Web as a whole. Each statement can link to other information sources and the participants themselves may be represented by links to their home page self-portrait.

The Web's emergence as a sociable information space - and the limits its design imposes - is the focus of this section. I will discuss three projects here, each of which incorporates a different type of communication with the Web.

4.1 Portraits in Cyberspace

Portraits in Cyberspace main entrance

Portraits in Cyberspace is an on-line exhibit about identity in the virtual world. The call for entries said:

We are seeking images that address: Who is on-line? Who inhabits the edges and margins of the on-line world? What constitutes identity in cyberspace -- and how can it be portrayed? How are essential human experiences -- such as family, religion, community, sex, ethnicity, childhood, personality -- being transformed in the digital era? We are interested both in documentary portraits of people in the virtual world and in experimental depictions of on-line society.
All aspects of the exhibit are virtual: the competition was publicized via email and announcements in art-related sites; the jury deliberations were held on-line; and the exhibit itself is a Web site, where viewers can see the works and participate in discussions about them.

4.1.1 The jury site: the web as a tool for collaboration

Judging the Portraits in Cyberspace entries on-line made it possible to convene a group of experts in the field of new media art without regard to geographical location: there were jurors in Boston, Finland, the Netherlands, Germany and Japan, plus one judging while en route from Croatia to Israel [11]. The virtual jury had neither the scheduling difficulties or nor the expense of a face-to-face meetings. And, since the works were designed for on-line viewing, the judges were seeing the pieces in their intended viewing conditions.

The members of an art jury work together to come to a consensus about the submitted works. Though they may have strong and conflicting views, their task is to decide, by compromise or persuasion, which works are to be accepted to the show or awarded a prize. The on-line jury site thus needed to be a place where the judges could rate the works, exchange opinions about them, and reevaluate their judgements as the process continued. Over 300 works were submitted for consideration. The number that would be accepted into the show was not fixed: with no constraints of limited wall space, all entries judged to be of high quality could be accepted. In addition, up to 6 prizes were available to award to outstanding works. The judging process is a sorting problem: first the unacceptable pieces are rejected, then the uninspiring ones, and eventually the top pieces are culled from the field of accepted works.

The jury site opened with a list of all submitted works. The works were referred to by the artist's name and an entry number; next to each name was list of the jurors who had already rated this work and the work's current cumulative rating, e.g.

Morrison.121 Sakane Kusahara Shaw Huhtamo Mitchell Naimark (3)
Dunn.249 Sakane Kusahara Mitchell (2.3)
The list started in alphabetical order and was grouped by rank as the jurists rated the pieces. Upon entering the site a juror could thus quickly see which pieces still needed to be reviewed, and approximately where a piece was ranked in the group's opinion. From this list, the jurors could access the individual pieces.

Each artwork had its own page, which included the artwork itself, plus the jurors' ratings and comments about it. For this show, an artwork had to include an image and a short statement by the artist. A link to an external site was also permitted; if included, that site was considered an integral part of the artwork.

The jury ranked the works on a scale from 1 (Does not belong in show) to 5 (Should win a prize). They could comment on the work in a private forum seen only the other jurors (and the curator); they could also leave public comments about the work, quotes to be available for the catalog or articles. The jurors were free to amend their ratings throughout the judging period.

Aside from some technical difficulties, this experiment in virtual judging worked well. During a ten day period, the jurors viewed and rated a large number of pieces. Their comments reveal a range of approaches to judging the aesthetics of this new medium, but also a great deal of agreement about what constituted a creative and expressive ``portrait in cyberspace''. The most highly rated works explored ways of using the Web as a new medium, and were unanimously praised by the jury.

The intra-jury communication - the commentaries and the ability to see each other's ratings - was certainly very important. Jurors pointed out features they felt others had missed, from relevant notes about an artist's background to suspicions of irregularities in the attributions [12]. They also argued among themselves: a work that one juror praised as a ``beautifully crafted image'' was described by another as the ``kind of kindergarten metaphysics [that] gives the net a bad name.'' (The former view prevailed and the piece was awarded one of the prizes). Also helpful was the list of entries accompanied by the names of the jurors who had completed their review. Here, as the list filled with jurors names and the best entries emerged at the top, the jurors could see the progress of the ongoing collaborative task.

The most serious problems with the judging process were slow network connections and difficult access. Several of the jurors were travelling during the jury period and had only intermittent or very slow connections. Downloading the images took considerable time, which limited the amount of time jurors could spend on commentary. Also, some of the external sites were themselves slow or not always accessible: one of the prize-winning pieces came close to rejection since its site was down when several of the jurors tried to see it. A jury convening simultaneously on four continents to judge a show of cyberspace portraits was conceptually engaging, but the state of connectivity in the fall of 1995 made it come close to falling short of fully practicable.

4.1.2 The exhibit

Portraits in Cyberspace opened on October 10, 1995. It is superficially similar to the jury site: there is a list of the included works, from which one can access the individual pieces. There is also a ``Gallery of Prize Winners'' entrance, with short introductions to the top works. There are about 80 pieces in the show, and they range from naive art to pretentious undergraduate pieces to some serious explorations of the web as a portrait site. Comments about the works are invited and most pieces have by now accumulated a lengthy and often lively series of remarks.

The exhibit site is differs from the jury site in several significant ways. The jury site was a tool for accomplishing a specific task; the exhibit site is a cultural space for seeing the art works and, if one chooses, the reactions they have provoked in one's fellow viewers. Most significantly, the jury site and the exhibit have very different audiences.

Access to the jury site was limited to a select group of professionals, most o whom knew each other, by reputation if not personally. The exhibit is open to anyone. Although comments must be signed, there is no verification of the names and many of them are clearly pseudonymous [13].

Several paths lead to the Portraits exhibit. In the topology of the Web, these incoming links are the site's surrounding neighborhoods. It is featured on the A Day in the Life of Cyberspace site and listed in various arts related links. The Electric Postcard includes several of the pieces and a number of the artists in the show link to it from their own home pages, as do other people who particularly like the site. People arrive at the Portraits exhibit by following links from one of these neighboring sites. The nature of those sites thus determines the exhibit's audience, which, in this case, is rather eclectic. From the arts links come serious art viewers, other artists, and budding cyber-media critics. From the artists' home pages come their friends. And from The Electric Postcard and the Day in the Life site comes the full range of Web explorers.

Character Site, by Stephanie Cunningham. Work displayed in the Portraits exhibit.
On the left is the piece itself: an image, the artist's statement, and a link to the site.
On the right is the top segment of the commentary page for this piece. (The scroll bar on the right shows how much further the discussion continued).
The strength of this piece (the most highly rated in the judging) is the site rather than the displayed image. The five faces were created by morphing: they are thus recognizable but with no specific identity. On the site, people were asked to write what the imagined the character of each woman to be. It is a fascinating record of what we read into the face.
The diversity of the audience is apparent in the commentaries on the artworks. There are sophisticated critics, conversant with irony and postmodernism, who think art should be thought-provoking and there are visitors who think art should be pretty pictures. The commentary often takes the form of debate, with subsequent visitors arguing the merits of both the work and of the preceding comments. In the Portraits gallery, there are debates about abstraction, about whether the artist should paint a universal or personal message, about whether art should have any message at all, and whether it is art if it was easy to make. There are pieces that have sparked arguments about race, feminism, and pornography.

As new art forms and technologies emerge, they engender new controversies about the role of the artist, the necessity of representation, etc. Abstraction in art has been controversial since painters began turning away from realistic representation; photography as an art form took decades to achieve academic respectability (Grundberg 87). In the commentaries of this exhibit, one can see a new set of debates emerging, centered on the issues of interactive and collective art.

The call for participation for this exhibit invited pieces that used the net as an artistic medium, not merely as a distribution mechanism. A few of the pieces, and arguably the strongest ones, do so and these were generally the pieces that the jury (and curator) found to be the most original and thought-provoking. In the exhibit, they have proved to be the most controversial. For instance, a piece called Curator: Not Quite Nirvana (Scott Matthews) is an installation of images from a kiosk the artist created. The kiosk, which was installed in several Soho galleries, allowed visitors to draw over digitized versions of art on display. Curator was one of the prize-winning pieces. Juror Itsuo Sakane wrote in his critique:

The idea of ``Interactive Kiosk'' for everybody is excellent. Especially the image selected for this contest was appealing with its rather primitive but mystical folk art taste.
The audience on the net was of mixed opinions. The originality of the piece was questioned:

Hey Scott, how can you call this YOUR work when it was actually created by numerous people in a real-world setting? It seems to me you let someone else do the work and you take the credit.

- NOMAD4

As was the quality of the resulting images:

The process is more intriguing than the result. Much the same as the hitting of a golfball on the moon. It is oddly disappointing that for all this technology we have acheived communal doodles. The minds reflected in the comments are more telling than the work itself. In that vein, I suppose the work was necessary as a catalyst. It seems moreso that an intricacy of a Star Trek episode would gaurantee an equal deluge. The exchange of ideas in this medium is often counteracted by the dumbing effect anonymity has on the worth of contributions.

- Quinton Donleavy

And some viewers championed the idea of interactive art:

Forget ``dumbing effect'' I think that for some visitors to this exhibit, being able to interact with the art may have made it more accessible! I believe that creative impulses are deadened by the influences of menial work and subsequent television watching. An exhibit such as this one allows otherwise artless persons to participate (albeit in a limited fashion) in the creative process.

- tlarson@teleport.com

Opinions about the other interactive pieces were equally wide-ranging.

In a real-world gallery, one may overhear a conversation or two, but in general, one sees only the opinions of art world professionals - the curators and critics, steeped in the history and language of art, who write wall plaques and reviews. How other visitors react to a piece remains, for the most part, a mystery. Here in the on-line gallery, however, the viewer can read - and respond to - the comments and reactions of a wide range of fellow viewers [14]. As many social observers have pointed out, debates about art are often at heart debates about class and social values ( Bourdieu 1984; Halle 1992; Fiske 1989a). The discussions. while grounded in their critique of a specific piece, reflect the participants' underlying beliefs about the nature of beauty, the role of the artist, the significance of technology - topics that in the real world are seldom debated among people of widely varying backgrounds [15].

4.2 The Electric Postcard

The Electric Postcard Pick-up Window
The Electric Postcard is a simple concept. On the postcard site there are a number of images (famous paintings, urban photographs, drawings of insects, etc.). You choose an image, write a message, fill in the recipients email address, and send it off. The recipient is automatically notified, by email, that are card awaits. Upon going to the site and giving the claim number to the postmaster, the recipient is presented with a page showing the chosen picture and the message.


 
Date: Thu, 13 Apr 1995 14:20:01 -0400
From: cardmaster@postcards.www.media.mit.edu
To: 
Reply-To: 
Subject: Greetings from Cyberspace

There is postcard waiting for you in the Post(card) Office.
You may claim it at the Pickup Window, which is located at
http://postcards.www.media.mit.edu/Postcards/
Your claim number is:	.170227
Please have this number available when you claim your postcard.

		Thank you,
	 	The Postmaster

Messages left unclaimed after 2 weeks may be discarded.
The Electric Postcard recipient notification. The Web address was left deliberately unexplained: while a notice that acknowledges that WWW address are still rather esoteric and explains the downloading process may be more "user-friendly" it is not as striking as the unadorned one, with its implied belief that an electronic postcard is an everyday object (and its terse parody of a real world delivery notice).

What is interesting about this project is its unexpected success. It went on-line at the end of December 1994; a year later, about three-quarters of a million cards had been sent; half a year later, late spring in 1996, it was close to 1.7 million. It started slowly: 10-20 cards a day in the first weeks, 1000-2000 a day over the first summer, and then it gained momentum rapidly. During the 1995-96 Christmas season, there were days when over 19,000 cards were sent (and there would have been more had the server been able to keep up).

It is much more time consuming than sending email - you have to go to the site, the pictures take time to download. Why do people like it so?

When I built the Postcard server, I thought that the fact that the message contents can be hypertext would be a big appeal. Images and sounds can be interspersed with the text and one can send, along with one's words, links to anything within the Web's vast store of information and arcana. Yet it turns out that relatively few people use this capacity, perhaps 1 in 15-20 cards has some HTML embedded in the message. The appeal appears to be more social.

It is often used as a lure to get a reluctant friend to explore on-line:

Hi!
Just wanted to say I think this is one of the best sites on the net... I have introduced many of my frineds to the internet by sending them postcards.. it becomes an adventure for them to browse through and find a way to retrieve their postcard! :-)
It is an amusing twist on ordinary email:

I really like The Postcard Store. What a great idea! Email can get dull after a while, this is a great way to spice it up.
And it is a way to show friends and colleagues that one has found something new and different on the Web:

Hi there!
Your page is a hit! I have sent several cards out and everyone loves them!
The most significant function of the postcard, and the reason, I believe, for the great popularity of The Electric Postcard, is that they allow people to keep in touch without having to actually say anything. A notable thing about postcards is how trite the messages often are: ``The weather is great. Wish you were here.'' A letter like that would be ludicrous, even rude. Yet the main point of a postcard is its subtext: I'm thinking of you, just checking in, making the rounds remotely. The picture on the card takes the place of the message. It lets the sender express a bit of his or her taste (for humor or for the macabre or, most popularly, for Impressionist prints), like sending a little gift.

What a fun thing on a cold and dark December eve to get a nice warm postcard. Thanks for a cool service :)
The Electric Postcard is an interim technology. A real integration of email and the Web is not far off - the intermediary step of having a claim number and a pickup window are only temporary measures. Yet the ``Postcard'' metaphor is a good one to retain, for it captures the sense that these electronic messages are from some place. Postcards, both real and electronic, indicate where you have been: they are a means of letting your friends know that you are in Paris - or that you have found a cool site on the Web.

The Electric Postcard lets the user send a piece of the Web as a personal statement. The postcard itself, in addition to the picture, includes links to the image source, such as the artist's home page or the Web exhibit that houses the ``original''. These links can be an integral part of the card's appeal: the Artamnesia cards bring the viewer to a virtual club; the Alice in Wonderland cards feature the Tenniel drawings and have link to the chapter from which each illustration was drawn.

Postcard sites are becoming increasingly common on-line. There are card sites in China, Syracuse, and the Boston Computer Museum; there are card sites that feature the Car-Talk mechanics and characters from Griffin and Sabine. There are sites that have transformed the metaphor and that let you send virtual flowers, Indian candies or imaginary presents. ``What's the point of visiting a really nice web site if you can't make all your friends insanely jealous of you?'' asks Royal Caribbean Cruise Lines in their version of the postcard site.

Everywhere in the minutiae of our material culture, we encounter reminders of the availability of authentic experiences at others times and in other places.... Picture postcards circulate throughout the world tying tourists together in networks and linking the tourist to the attraction and to his friends at home. (MacCannel 1976)
Here, the ``authentic experience'' is a virtual location; the postcards make the information space into a source for personal expression.

4.3 WebTalk

The third project I am discussing in this chapter is a bit different than the first two. Both Portraits in Cyberspace and The Electric Postcard are active sites, accessed daily by people from around the world. They are design experiments that show how particular interfaces and metaphors resonate in real use. WebTalk, on the other hand, is a technological experiment, a proof of concept implementation for extending communication and awareness on the Web.

The Web was designed for publishing, not communication. At the time that WebTalk was built (1994-95) neither the httpd protocols nor the available browsers supported live, synchronous communication nor was there any mechanism for awareness of the presence of others on a page.

Though unsupported, the demand for Web-based communication was - and is - strong, as evinced by the numerous ways people have devised for communicating with each other within the Web's context. There are ``graffiti wall'' and ``guest books'' where visitors can post comments (e.g. Portraits in Cyberspace) and Web pages with links that jump to Usenet newsgroups or that bring up telnet windows to MUD sessions. There are collaborative annotation systems and cooperative story building mechanisms; there are voting systems where you can rate Star Trek episodes or recent movies. Some of these function quite well for the communicative task they address; others are clearly interim solutions, waiting for a more elegant and integrated solution.

In particular, two social functions remain unaddressed by the systems described above. First, exploring the Web is remains a solitary pursuit: one is unaware of the presence of the many fellow visitors. Second, one cannot communicate directly to another person on-line within the rich hypertext space of the Web.

The WebTalk project (Donath and Robertson 1994) was a modified Web server and browser that addressed these problems. Most notably, it showed who else was on a page and it allowed the user to strike up a conversation or to join an ongoing discussions. Unlike the graffiti boards and other Web-messaging techniques WebTalk was real-time - one came to be engaged in dialog, rather than to write a message and leave.

The WebTalk "Who's On-line" window.
Shown here both in text and in a graphical version. (Full scale versions available by clicking on above images).
WebTalk provided a sense of the presence of others. A ``Who's On-line'' window showed who else was on the page. One could quickly scan for a particular person, such as the owner of the page or an acquaintance often found at this spot. The size of the list provided a sense of presence and of the activity level of the page: are there only a couple of people here or is this a major gathering spot, a favorite Web meeting point?

Designing this feature required defining presence, a somewhat elusive concept in the Web environment. WebTalk was based on the concept of pages as locations: you appeared in the window of the page you were on and you could only talk to people on that page. Wandering, however, is an essential activity on the Web - following links, jumping from page to page. It would be confining to have to literally stay on the same page when conversing with others. The concept of virtual locations was developed to allow users to put down an anchor on a particular page - where they would appear to remain - and still wander about the Web with their main browser window. Virtual locations mandated and enabled a singular presence: you could appear to be in only one place at a time.

The main feature of WebTalk was its integration of real-time discussion with the resources of the Web. One could easily link one's statements to references, counterarguments and examples - and have an entire library of eloquent pictorial or auditory interjections.

The WebTalk conferencing screen.
Judging from the writing styles that have evolved in ASCII-only discussion spaces, the demand for such meta-linguistic means of expression is strong. Extensive punctuation, smiley faces, and emphatic asterisks and capitalization are very common in these on-line conversations. They are quite easy to use - as easy as typing any ordinary word. Yet these inflections are limited by the ASCII character set, which is neither extensive nor well-suited for this purpose.

Having access to the full resources of the Web greatly extends the expressive range. Yet HTML, the code of the Web, is awkward and indirect: one types a long address string in a source document in order to bring up a picture or link in the viewer's browser. WebTalk, with its modified browser, was able to add fluency to hypertext conversation by providing point-and-click tools: one needed to simply click on an image or link in order to add it to one's message.

We conceived of the WebTalk conferencing mechanism as a sort of prefabricated gathering space: the basic tools were provided for the host, who then set a few parameters and opened up for business. Typically, a Web page designer is a publisher, putting text, pictures, and sounds on-line. WebTalk added conferences to that list, making the page designer (or owner) into a host. This new role raises several issues about control, for establishing norms of behavior becomes an important task once the visitors to one's space are no longer passive observers but active participants.

The parameters were concerned primarily with access: could anyone access the site or was it membership-based? were the conversations moderated? were all participants visible? was there a mechanism for bouncing unruly participants? A moderated, members-only discussion is more formal (and time-constrained) than an open chat space; the former may be appropriate for a scheduled debate on a specified topic, the latter may work well as a meeting place on a popular site. By adjusting the parameters of the conferencing environment, the host would be able to modify the general atmosphere of the discussion [16].

The WebTalk project was discontinued in the spring of 1995. The server and browser were demo-able, but not yet ready for release to the general public. Browser technologies were changing rapidly, and the most advanced and widely used ones were no longer making their source code available. Sophisticated application languages, such as Java, were not yet available but were clearly the shape of things to come. WebTalk remained useful as a demonstration of a sociable information space, but its incarnation as an integrated browser and conferencing system, though providing it with features that stand-alone systems can not offer (such as the point and click links), also made it impractical to introduce at large.

Today, there are a number of sites that offer on-line chat from the Web. Most are simpler than WebTalk and few of them allow HTML to be used in the messages at all, let alone providing an interface for it. Some are simply text-based chat on web page, while others are full graphical environments.

One feature from WebTalk that is not part of any mainstream Web/chat programs - and it is the one that would be the most interesting to see develop and evolve in real use - is the easy to use, point-and-click html editor. As I mentioned in the discussion about The Electric Postcard, I had thought that the ability to add links and other hypertext elements would be a popular feature. But writing HTML is tedious; relatively few people are familiar with it, and even those who are find typing out long site addresses a graceless interface. In a live chat environment, with its emphasis on speedy repartee, the interminable site strings are even less appealing. So, while home pages and other carefully prepared documents are often replete with links, it still has not become an integral part of interpersonal correspondence.

4.4 Remarks
The projects discussed in this chapter each incorporate social interactions with the information landscape. In Portraits in Cyberspace, information (i.e. the exhibit) functions as the setting for the discussions, each of which is situated in the context of a particular work of art. Here, the information setting provides a focus and catalyst for the discussion. The WebTalk conferences were also designed to be situated in a relevant information setting, with a focus on making the information space inhabited, bringing live interactions and representations of presence to the on-line environment. Information is used somewhat differently in The Electric Postcard. Here, the information (in the form of images) is made into personal object (postcards) by means of the metaphor of the postcard and through the individual composition and addressing of a message. The Postcard site is also a setting - like a virtual tourist spot, it is a place from which to send a souvenir.

An information space, though it is not tangible like wood and bricks, is still a ``place'' in the sense that is a particular locale: specific, unique, and recognizable. In the real world, we associate specific places with particular types of discourse: conference rooms, writer's bars, kitchen tables all lend themselves to different tones and topics of discussion (Oldenburg 1991). The physical space shapes the conversation both by attracting a particular group of participants and by creating a particular atmosphere, whether of formality or casualness, intimacy or openness, etc.

An information space similarly attracts a particular audience, one that is likely to share similar interests. Gathering an audience of people with shared interests is one of the Web's strengths: many sites are narrowly topical; others, though broader, are designed to attract a specific group. And the Web's topology of links, as discussed in the analysis of the Portraits audience, makes an interesting flow of visitors likely. Lynch's assertion that the well-designed space should ``furnish the raw material fro the symbols and collective memories of group communication'' is also well served by the information space. Here, the materials - words, images, sounds, etc. - are themselves symbolic constructs, transformable from gathering space to communication element. This is encouraged by the Web's hypermedia structure, in which references to sites are easily integrated into any message. With the , for example, while hypertext was not often used in people's letters, the site itself is both an information space (a known and distinctive site) and an element of communication (pieces of which are ``sent'' as messages).

One way in which the virtual world lags far behind the real is in the creation of a distinctive atmosphere particular to a space. Today, with ``information spaces'' primarily represented by the Web, the atmosphere - the tonal range - of these spaces is quite limited. Pictorial backgrounds have replaced the universal gray of early browsers and methods for adding sounds and animations are starting to appear, yet the options for interaction remain limited and our sense of the presence of others remains dim. Although the Portraits discussions are located in a virtual art gallery, in format they are only slightly different than say, a Web-based discussion about health or a home-page guestbook. Two such information spaces may be quite different in content, but their form and overall visual quality will be relatively similar. In the next two chapters, I discuss projects with highly visual and interactive interfaces. The eventual goal is to unite these two design directions: to make sociable information spaces that are as distinctive in their appearance and interaction as they are in their content.

previous · · · contents · · · next