Crossroads The ACM Magazine for Students

Sign In

Association for Computing Machinery

Magazine: Features
Socializing the web

Socializing the web

By ,

Full text also available in the ACM Digital Library as PDF | HTML | Digital Edition

Tags: Collaborative and social computing, History of computing

back to top 

The history of networked computing is also a story of people struggling to reconfigure technologies designed and built for "practical" purposes into something usable for socializing. The precursor to the internet, the U. S. Defense Department's ARPANET, was built in the late 1960s to enable computer resource sharing; within a few years, interpersonal messaging, an initially unauthorized use, had become the primary source of network traffic. Over the next two decades a thriving ecosystem of online social spaces was created, including bulletin boards, chat services, Multi-Users Dungeons, the massive Usenet newsgroup system, and email. The simplicity of these ASCII text-based programs belied the radical newness of the massive globe-spanning conversations, pseudonymous communities, and other social situations they enabled.

When the web was launched in the early 1990s, the internet was visually austere yet highly interactive. A world in which information was ephemeral and anonymity was rare. Access to the internet had initially been via one's university or job, making users easily traceable and accountable. Though this was changing by the late 1980s with the arrival of publicly available commercial access, including some that provided effectively disposable accounts, most email and newsgroup users were still readily identifiable. The web, with its relatively permanent pages of text and images, was its complement. It was designed for publishing information, not for interaction. The pages were static and visitors to a site were nearly anonymous, identified only by their IP address, which might be shared among hundreds of people.

Yet people quickly adapted the web for social uses. They created home pages, writing mini-autobiographies enumerating, along with their research papers (the early users were primarily academic), lists of their favorite bands, restaurants, or logic puzzles. They posted photographs of themselves, their travels, and their pets. Most importantly for the web's development, they posted lists of useful and intriguing sites they had found. For initially there was no Google, no oracular search engine that would bring up hundreds of answers to your most esoteric questions. Instead, you sought information by following trails of links, a journey of recommendations from homepage to homepage. The web felt social—you could form vivid impressions from HTML self-portraits—but also solitary. Hundreds of others might be visiting a page, enthused or enraged by the opinions expressed there, but there was no sign of that, or any, presence.

At the time, I was a doctoral student at the MIT Media Lab. As a design researcher I was building experimental computer-mediated social spaces: a chat program showing who was paying attention to whom [1], a visualization of the people in a community based on their roles and interests [2], and others. When the web appeared, I was intrigued by seeing how people chose to present themselves, but the lack of interaction seemed a step backward from the visually plain but lively ASCII internet. Computers and networks were becoming fast and powerful enough to enable more creative and expressive media, but the web was like digitally distributed paper. Over the next few years, I built, and collaborated with others in building, a series of projects exploring ways to make the web social. Some things we struggled to achieve, such as showing who else is visiting a website, have become easy to implement with changes to the web's protocols. Others, such as representing people in a legible and evocative way, are still open areas for research and experimentation.

back to top  The Electric Postcard

The Electric Postcard [3] was the first project I created for the web, and while it would later turn out to be unexpectedly and wildly popular, it started as just a casual exercise, and a bit of procrastination from working on my thesis. My officemate was so zealously enthusiastic about the programming language Perl that he was writing a book about it [4], and quite insistent that I try it. So, as a learning exercise, I assigned myself a project: Implement "postcards" that people could send to each other online. I finished the site in early December 1994 (though I never came to like programming in Perl), and sent a few cards to friends. Over the next few weeks, news of the site slowly spread: 18 cards were sent the first day; a month later the daily number hovered in the hundreds. Then Valentine's Day came and the number shot up to more than 1,000; by Christmas time a year later, nearly 20,000 cards were sent daily. It was written up in Wired and other tech-watching publications. It eventually accounted for a sizable amount of the Media Lab's network traffic, enough so that my desktop computer—on which it ran and on which I was working on my dissertation when I was not dealing with the deluge of emails about lost cards, anonymous cards, and the ire of card recipients who had gone to the physical post office demanding their postcards—had to be moved to its own dedicated internet connection.

One reason for its popularity was that postcards, physical or virtual, provide a pretext for being in touch without a need to have anything notable to say. Written letters and email typically require some substantive content, especially if the recipient is outside of one's immediate circle. But much of our face-to-face communication consists of small talk—greetings, pleasantries, and ritualized exchanges—that functions as a verbal form of grooming, helping us assess and reaffirm personal bonds [5, 6, 7]. The Electric Postcard supported this sort of social communication online: Cards could be sent with only minimal text without seeming terse or inarticulate.

Postcards are a form of lightweight gifts, signaling not only "I am thinking of you" but also "here is what I think you would like." They are a demonstration of how well I know you. The wide choice of images the Electric Postcard provided—paintings from the Louvre, distant planets from NASA, protest posters from Sarajevo, hearts and cherubs for Valentine's Day—contributed to this social function by encouraging the sender to seek one that would be especially apt. Furthermore, sending a link to a funny image or video is a perpetual online gifting ritual [8, 9]. The postcard site was dryly humorous, presided over by a bored postal clerk whose cranky remarks were inspired by exchanges at my former post office, where I had a PO Box, necessitated by the surly mail carriers (this was the era when the phrase "going postal" was coined) who at times simply dumped our building's mail in the trash rather than deliver it.

Being a human online today is a much more fraught experience than it was in the web's early days, and much more so if one is a woman.

People are social beings—even online. If they find something interesting, they want to show others; even if they have nothing to say, they still want to be in touch. Today, it is no longer necessary to argue this point. For many people, the web is "social media" i.e. Facebook, Twitter, Instagram, Reddit, etc. But in the 1980s and 1990s, this was far from universally accepted. In particular, social interaction was often dismissed as a waste of time and bandwidth by executives who controlled the development of and access to online social spaces, but did not themselves participate in them. Examples abound in the history of computing. In the early days of computer networks the ARPANET administration threatened to cut off access to nodes hosting frivolous conversations [10]. A decade later the online service provider Prodigy, worried that socializing would distract people from shopping, not only charged for messaging, but also made it deliberately difficult to use. (I worked at Prodigy from 1986 to 1990 in their "advanced technology group" and recall frustrating discussions with senior executives over the value chat and email provide.)

Undervaluing the social is not limited to computer interaction; it is a pervasive issue in many areas. We prize language as the ability distinguishing us from other animals [11], yet denigrate gossip, the social use that is arguably language's evolutionary origin [5, 12]. We prize productivity and efficiency over, and often in opposition to, personal interaction. Urban sociologist William H. Whyte studied how people use and move around city streets and parks, and documented how they shunned the vast plazas urban planners promoted, and instead gravitated to more organically complex and crowded settings. He concluded, "What attracts people most, in sum, is other people. If I labor the point, it is because many urban spaces are designed as though the opposite were true and as though what people liked best are the places they stay away from" [13]. When planners and builders, whether of physical plazas or computer protocols, disregard social elements, the result is barren infrastructure that provides few amenities or opportunities for interaction among its users. The Electric Postcard's popularity is what made it significant. It vividly demonstrated, at a time when this was considered radical or perhaps just frivolous, that people are interested in other people, even online.

back to top  Data Portraits

Today, few would claim that social interaction is an unimportant component of networked computing—you are much more likely to hear complaints that the online world is excessively, detrimentally social. While social media has enabled many positive experiences—from staying in touch with friends over decades and vast distances, to playing games and arguing with complete strangers—it has also proved to have a dark side. Much of which stems from the interrelated problems of deindividualization (people online are not perceived as fellow humans but as abstract ciphers, and thus outside the norms of politeness and empathy that keep face-to-face conversations from routinely devolving into exchanges of insults) and anonymity (which allows people to act badly with little fear of reprisals, reputational harm, or other punishment). Yet I would argue the problem is not that the web is too social, but that it is still not social enough. The interfaces we have today still provide only sparse cues about people and their activities. Without this information, we have a hard time knowing who to trust and believe, and we lack the context to fully understand the meaning of what others say.

People are social beings even online. If they find something interesting, they want to show others; even if they have nothing to say, they still want to be in touch.

Online, we typically see others through the words and images they post, but not the person themselves. For the cyber-utopians of the early internet, this absence of the embodied self was a key element of the networked world's promise. We would form our impressions of each other through what we say, avoiding the instant sorting into age, gender, and race categories we inevitable do upon seeing another's face. Reality, however, turned out to be more complicated: Words, too, reveal social cues, and thus our categorizations persist online [14]. That said, impressions formed through words develop slowly, and the text-based strangers we encounter online often remain opaquely enigmatic. Data portraits—visualizations of people based on their words and actions—potentially provide the best of both worlds: evoking the subject's concerns, beliefs, and personality while leaving their race, gender, and age ambiguous if that is preferred.

Two of the other web projects I worked on in the mid-1990s were early experiments in "data portraiture" [10, 15]. One was an all-online juried competition and exhibit called "Portraits in Cyberspace" that sought works exploring novel stylistic and interactive elements for computer-based portraiture. The other was part of a web-based "Day in the Life of Cyberspace" festival, in which people worldwide participated in various quizzes, games, and questionnaires, many of which asked you to reveal something about yourself, from basic statistic to esoteric interests. Using these answers, the site built a simple portrait of you; it was a prototype for a portrait that could function as a more-meaningful avatar, to be generated as needed by individual sites.

The topic of portraiture, and the problem of how to create meaningful representations of people for the digital world, remains a deep interest of mine. Both the choice of data and the design of the representation pose challenging questions. When I became a professor, my students and I did a series of data portraiture explorations, depicting people through their contributions to online discussion, their email exchanges, their Twitter community, etc. Such portraits, I believe, would benefit many online discussions, for they help the participants perceive each other as individuals and understand each other's writings in the context of their interaction history. The ability to see at a glance the topics the other writes about, the extent of their participation history, and the rhythm of their exchanges with others would be especially helpful on sites where contributors are often strangers to each other, making it easy, for example, to quickly distinguish a frequent poster with a history of constructive interactions from a one with a controversial history of provocations.

Today, the popularity of online social experiences—social networks, recommendations, comments, and other forms of virtual social interaction—mask how primitive the interfaces for interaction still are. In the early days of the web, the technology itself was a major constraint on design. Today, however, it has a far more flexible infrastructure—and many limits are due simply to the weight of habit and success: Why change what is known to work? But "work" in this case means what is profitable, not what is most usable or conducive to harmony. Today, too, interfaces that encourage cooperation are needed more than ever. By current expectations, the early days of the web were halcyon. For example, the "Portraits in Cyberspace" exhibit allowed visitors to comment on each artwork and for several years needed no moderation, no verified sign-ins, etc. There were no mysterious off-shore or in-basement factories churning out disruptive discussion-breaking bots. Today, such an open site would be overrun with spam almost instantly. Better design would help people easily identify genuine, engaged participants and encourage cooperative behavior.

back to top  Being a Woman on the Web

Given that the theme of this special issue is "being a woman on the web" you might be wondering what in this account is specific to being a woman? In terms of creating sites on the early web, very little—and that is the point. It is tremendously important that women's contributions to the creation of various technologies be recognized and celebrated. But at the same time they should, ideally, become unremarkable, and the designers and builders of technology no longer be presumed male.

As far as being a woman present on the web in its early days, I was fortunate not to encounter overt sexism or hostility—due to a combination of luck, a somewhat ambiguously gendered web presence, and the generally gentler nature of online interaction in the 1990s. (While I was clearly identified on the postcard site as its creator, the cranky male cartoon clerk was the more vivid, if titular, identity. I also kept an online diary, but in the guise of my car's daily musings [16].) That cultural change is important to note. I had several sites that featured open and anonymous commenting, and in the first few years of the web there were very few hostile or inappropriate entries. Today, such forums would be immediately overwhelmed with spam. Being a human online today is a much more fraught experience than it was in the web's early days, and much more so if one is a woman.

This deterioration of the online social fabric can, I believe, be mended. But doing so requires making the social side of the technology a central function, not an afterthought. It requires bringing informed and intentional design to the way people are identified, the control they have over these identities, and the technologies they use to input their thoughts and emotions. Here, the wide gamut of women's experiences, from empowered to vulnerable, are crucial to designing the next iteration of the virtual world.

back to top  References

[1] Donath, J. The illustrated conversation. Multimedia Tools and Applications 1 (1995), 79–88.

[2] Donath, J. Visual who: Animating the affinities and activities of an electronic community. In Third ACM International Conference on Multimedia. ACM Press, San Francisco, 1995, 99–107.

[3] Donath, J. The Electric Postcard. Internet Archive. (Retrieved February 8, 2021); https://web.archive.org/web/20060428042845/http://postcards.www.media.mit.edu/Postcards/

[4] Orwant, J. Games, Diversions & Perl Culture: Best of the Perl Journal. O'Reilly Media Inc., 2003.

[5] Dunbar, R.I.M. Grooming, Gossip, and the Evolution of Language. Harvard University Press, Cambridge, 1996.

[6] Laver, J. Communicative functions of phatic communion. In Organization of Behavior in Face-to-Face Interaction, Adam Kendon et al. (Eds.). De Gruyter Mouton, New York, 2011, 215–238

[7] Miller, V. New media, networking and phatic culture. Convergence 14, 4 (2008), 387–400.

[8] Highfield, T. Tweeted joke lifespans and appropriated punch lines: Practices around topical humor on social media. International Journal of Communication 9 (2015), 22.

[9] Taecharungroj, V. and Nueangjamnong, P. Humour 2.0: Styles and types of humour and virality of memes on Facebook. Journal of Creative Communications 10, 3 (2015), 288–302.

[10] Donath, J. The Social Machine: Designs for Living Online. MIT Press, Cambridge, 2014.

[11] Hauser, M.D. The Evolution of Communication. MIT Press, Cambridge, 1996.

[12] Emler, N. Gossip, reputation, and social adaptation. In Good Gossip. University Press of Kansas, Lawrence, 1994, 117–138.

[13] Whyte, W.H., City: Rediscovering the Center. Doubleday, New York, 1988.

[14] Herring, S. Gender differences in computer-mediated communication: Findings and implications. Computer Professionals for Social Responsibility Journal 18, 1 (2000).

[15] Donath, J., et al. Data portraits. Leonardo 43, 4 (2010), 375–383.

[16] Donath, J. 1964 Ford Falcon. In Evocative Objects. Things We Think With. MIT Press, Cambridge, 2007, 153–161.

back to top  Author

Judith Donath is a writer, designer, and artist whose work is about the co-evolution of technology and society. She is the author of The Social Machine: Designs for Living Online and numerous articles about identity, AI, social media, and the future. Formerly the director of the MIT Media Lab's Sociable Media Group, she synthesizes knowledge from urban design, evolutionary biology, and cognitive science to design innovative interfaces for online communities and virtual identities; her work and that of the Sociable Media Group has been shown in museums and galleries worldwide. Currently, she is an advisor at Harvard's Berkman Klein Center and is writing a book about technology, trust, and deception. She received her doctoral and master's degrees in media arts and sciences from MIT and her bachelor's degree in history from Yale University.

back to top 

© 2021 Copyright held by Owner(s)/Author(s). Publication rights licensed to ACM. 1528-4972/21/03 $15.00

The Digital Library is published by the Association for Computing Machinery. Copyright © 2021 ACM, Inc.


There are no comments at this time.


To comment you must create or log in with your ACM account.