Tom Erickson
  home •• pubs essays  HCI Remixed HICSS PC patterns  Apple Human Interface Alumni page

The World Wide Web as Social Hypertext

Thomas Erickson

Advanced Technology Group

Apple Computer, Inc.

(now at)

A version of this essay appeared in the Viewpoints column of Communications of the ACM, January, 1996.

Something curious is happening on the World Wide Web. It is undergoing a slow transformation from an abstract, chaotic, information web into what I call a social hypertext.

Initially, I didn't pay much attention to the Web. After all, it was just a new take on distributed information server systems such as WAIS[8] and Gopher[1]. True, it was easier to use than WAIS, and the ability of Web browsers to display formatted text and graphics with embedded links made it more attractive and engaging than either WAIS or Gopher. But there was nothing really new; it was an incremental advance, a new combination of well known functionality. So I mentally categorized the Web as just the latest fashion to sweep the internet.

In this I was quite wrong, although the phrasing of my dismissal in terms of fashion contained a deep truth.

This isn't to say I ignored the Web. As a user experience specialist in Apple's Advanced Technology Group one of my jobs is to stay abreast of new things. So, I occasionally browsed it to see what was happening. And early in 1995 I had a conversion experience.

The cause of my change of heart was the widespread appearance of personal pages. Personal pages are something like informal resumes, except that in addition to professional material they often contain personal information. Hobbies, research interests, pets, professional publications, children, politics, friends, colleagues, all are grist for the personal page. I believe that this seemingly frivolous blending of the professional and the personal is the key to why the web is becoming a fundamentally different thing from the systems of information servers that preceded it. Personal pages and the world wide web are not being used to 'publish information;' they are being used to construct identity--useful information is just a side effect. A personal page is a carefully constructed portrayal of a person.

It is this that leads me to characterize the World Wide Web as a social hypertext. The nodes -- at least some of them -- are becoming representations of people. And this, in turn, enables another critical feature to emerge: links from a personal page often point to socially salient pages. A common feature of the personal page is a list of pointers to "interesting people and places." What and who counts as interesting? That depends on the person, and hence also tells us more about the person. Thus, the links, as well as the page itself, participate in the personal portrayal; in a sense, they embody a sort of social logic, providing us with a view of that person's network of friends, colleagues, and concerns.

The transformation of the Web into a social hypertext has a number of interesting ramifications. Perhaps the most immediate and practical is that social hypertexts allow a fundamental shift in the way people search for information. Rather than composing queries for search engines or going to likely places to browse, something that many ordinary users find foreign and daunting, people can instead pose the question: Who would know? Or who would know someone who would know? Navigating from one personal page to another we suddenly have a new sort of search strategy, although this sort of social navigation [3] is new only in the context of computer networks--it is, of course, an old and familiar way of finding things out in the real world. We are social beings, and social hypertext provides the opening for us to use our immense store of social knowledge to make inferences about where to find information on the net.

Now it might be argued that this is nothing new, either. After all, can't we achieve the same end just by picking up the phone or sending email? Well, no, not quite. An important difference is that on the Web I can find out about what people are doing and writing without becoming obligated to them. That is, any time I ask someone to do something in my behalf, I accrue a social debt to them. So, if I contact someone I don't know and request a paper, there is now likely to be an expectation that I will read the paper, and perhaps comment on it. The difficulty is that I may not have time to comment on it, or I may glance at it and find it uninteresting, and so I am left in an awkward position. The same issues arise, mutatis mutandis , in sharing one's own work. In short, the ability to find out what someone else is doing, without mutual knowledge of what's happening, is a boon to both parties. This non-mutuality of knowledge is one of the characteristics that makes social hypertext different from more direct forms of communication.

While the non-mutuality and socially-driven search strategies are important practical outcomes of the transformation of the web into social hypertext, I believe that in the long run it is the use of web pages for personal portrayal that will have the most profound effects. Note that while personal portrayal is a new type of usage of the web, it is very ordinary behavior in the real world. People go to considerable effort and expense to manage their appearance. Portrayal management ranges from the ways in which people act in public [4, 5], to the clothes they wear [7] and the goods they possess and consume [2]. The World Wide Web is one of the first venues where individuals can construct portrayals of themselves using information rather than consumer goods as their palette. For the first time, individuals can project huge amounts of detailed information about themselves to a mass audience -- they are no longer limited to assembling montages of consumer items that at best suggest where they fall on political, social, and cultural spectra. If personal pages are indeed a new manifestation of the human impulse to manage personal representations, then we need only look at markets like cosmetics and clothing to get an idea of the force behind this impulse.

One reflection of the social importance of personal portrayal can be found in the kinds of debate it occasions. Creating personal pages is not just an exercise in vanity publishing; as with the real world analogs of personal appearance, consumption, and comportment, people are quite aware of potentially negative consequences of their personal portrayals. Consider the following bit of email describing discussion about personal pages among a community of graduate students and junior faculty: "There's a real dilemma for these people about how much to 'personalize' their home pages. They're concerned that presenting self means offending others or worse yet projecting bad self image to recruiters or peers " This is the negative side of the non-mutuality of social hypertext: it is more difficult to assess how others are responding to one's portrayal. The email continues, describing responses to these concerns: " recently (at least among the people I spoke with) a lot of folk have begun de-personalizing their pages. I've thought this through and put a disclaimer at the top of my page as I don't really want to loose the intimacy of putting things on my page that I think are fun... although I am still debating removing a fairly strong political statement " [6.].

The issue of the intimacy of personal portrayals arises in the corporate context as well, where the diversity and intimacy of personal portrayals creates tension with attempts to project a coherent, business-like image of an organization. At Apple there was considerable internal debate preceding the decision to allow members of the Advanced Technology Group to make their personal pages available on the public web. An initial veto by upper management met with vigorous protest from the researchers. Management concerns ranged from accidental revelation of confidential information to the question of how personal personal pages should be. The concern, of course, was primarily with the latter question--there are well established policies and procedures governing publication of research and other professional information. But there were no policies about what to do if employees wish to publish poetry, samples of music from their punk garage band, or political satire. Nor was management eager to enter this new frontier of corporate policy.

For Apple, the resolution of the debate was relatively easy. In part this was because the Apple Library -- a group respected by both managers and researchers -- stepped in to play a mediating role and to guide the development of web policy. And in part, the actual tension between corporate image and personal portrayal was not that great. Apple in general, and the Advanced Technology Group in particular, values its reputation as an innovative, cutting edge organization, and thus, its portrayal as a diverse community of creative individuals is not at great odds with its image as a leading edge technology developer. That is not to say that concerns about what constitutes the bounds of appropriate behavior have vanished. Rather, the strategy is to encourage a community based approach to defining what is acceptable. At present, new personal pages are posted on a private server so that members of the Advanced Technology group can review proposed pages and discuss any questionable issues. It is important to recognize that this method is deliberately restricted to the hundred or so members of the Advanced Technology group: if there is a problem, the method of resolution should be by personal discussion between two people who know one another (the alternative, discussing the questionable details of an individual's personal page on a mailing list open to hundreds or thousands of strangers, seems like quite a bad idea). Thus, people from other areas of Apple who wish to publish their personal pages need to work out methods and appropriate norms with their management and their peers. It is no coincidence that it is this same method -- albeit in a more subtle form -- that governs other aspects of personal portrayal such as dress and hairstyle. The creation and evolution of social norms is a community-based process, not something that is centrally determined; so far it appears to be working well for social dimension of personal pages.

Clearly, the desire to construct personal portrayals is a deep seated human trait. I believe that this is one of the key drivers behind the expansion of the world wide web. The mix of the professional and the personal is fraught with potential. On the positive side it enables new search strategies based on our social knowledge, it lowers the social cost of accessing and sharing information, and it makes the web a more interesting and engaging place. On the negative side, it creates a host of new opportunities for social gaffes, and defines a new realm in which tensions between organizations and individuals may be manifested. It will be interesting to see how this plays out, and whether the tensions can be resolved without losing the provocative blend of personal and professional that can engender social activity. Although the web may be just the latest fashion to sweep the Internet, if it turns out to be a medium that allows the construction, negotiation, and propagation of the styles of appearance that we refer to as fashion, then its impact may be profound indeed.


Thanks to Jonathan Grudin for comments on a previous version of this essay, and for more general discussions on the social impacts of technology, and to Janet Vratny of the Apple Library for discussions about the development of Apple's Web policy. And kudos to the Apple Library, whose leadership is in no small part responsible for the painless nature of Apple's transition to the public web.


[1] Alberti, B., Anklesaria, F., Lindner, P., McCahill, M., Torrey, D. The Internet Gopher Protocol: A distributed document search and retrieval protocol. University of Minnsota. Anonymous FTP from <URL:>

[2] Belk, R. W. Possessions and the Extended Self. The Journal of Consumer Research , vol. 15, pp. 139-168, 1988.

[3] Dourish, P. & Chalmers, M. Running Out of Space: Models of Information Navigation. Proceedings of the Conference on Human Computer Interaction. (Glasgow, August 1994).

[4] Goffman, E. Behavior in Public Places: Notes on the Social Organization of Gatherings . New York: MacMilliian, 1963.

[5] Goffman, E. The Presentation of Self in Everyday LIfe . New York: Doubleday, 1959.

[6] Grinter, R. Personal Communication (Email dialog between R. Grinter and J. Grudin, 8-20-95).

[7] Hebdige, D. Subculture: The Meaning of Style . London: Routledge, 1988.

[8]. Kahle, B., Morris, H., Goldman, J., Erickson, T. D. & Curran, J. "Interfaces for Distributed Systems of Information Servers." Proceedings of the American Society for Information Science Conference , (May 1992).


Tom Erickson

home ••• pubs essays  HCI Remixed HICSS PC patterns Apple Human Interface Alumni page •