“Social Software” as Institutions in On-Line Life?

Debate once flourished over whether ‘cyberspace’ – and more particularly social life on-line – was embedded within, an extension of, or apart from surrounding social, political, economic, cultural organization. Part of the argument was girdled by technological determinism, a strong form of ‘medium theory’, so called in a bow to the media studies guru Marshall McLuhan. Behind his quip that the medium is the message was a serious theory of communication built on the observation that systems of communication (a) tend to be laid on top of previous ones and (b) to differ in what they can communicate (and thus socially organize). Hence, focus on the medium, not the message, to understand the forces in and impacts of communication. The basic idea came from a Canadian geographer of the fur-trade who observed that each new system of communication – roads, railroads, telegraph, telephone, etc. – was laid over those before. An example might be Jordan, through which from south (the Gulf of Aqaba) to north (past Galilee Lake Tiberius toward Damascus) runs the so-called King David Highway, which is actually a Roman road, parts of which date to the neolithic. Just east of it is a modern north-south highway paralleling the Ottoman-period rail route between Damascus and Mecca; alongside are the telephone and telegraph lines, and buried somewhere nearby are the military’s fibre-optic cables. McLuhan’s very sensible view that each of these would ‘communicate’ differently tends to be discredited by its strong form, technological determinism, which overlooks how technologies are socially ‘constructed’ or grown. When examined with their history or social surround restored, technological effects seem almost accidental, at least hardly ever as intended, until experience consolidates and discards surrounding data – back stories – as false starts.

The lesson learned in media studies was social constructionism to get past the simplified social physics of ‘impacts’ to what are now called ‘affordances’. Taking technologies seriously as cultural and not natural puts agency back in users that automation would vest in machines or, in media studies, advertising would vest in the message. The resocialised IT ‘affords’ or facilitates the expression of social acts and relations on-line, in what used to be but is no longer seriously called ‘cyberspace’. This milder form of medium theory has the salutary effect of focusing attention on what people do instead of projecting theoretical capabilities of technologies taken abstractly or in isolation. So, Clay Shirky proposed a weak version of the Media Studies hope that structure can at least guide practice ( Social Software and the Politics of Groups, 2003) as facilitating ‘latencies’ in actors that has become Web 2.0 gospel. That is, social software enables actors to realise desires and capabilities (‘dispositions’ in the practice theory of Pierre Bourdieu, Outline of a Theory of Practice, 1977) to make connections on-line that circumstances preclude off-line. To elicit that, he goes on to distinguished four types or levels of sociality, which he seems to define as intention + commitment (but compare sociological definition of sociality by Wellman et al (2002)) that social software ‘afford’ expression on-line: sharing, conversing, collaborating, collective action.

Sharing, conversing, collaboration and collective action can be more or less associated with particular ’social software’ designed or used, sometimes ‘re-purposed’, for those kinds of interactions. Earlier, electronic bulletin board systems facilitated sharing, now it’s ‘bookmarking’ programs like Digg or del.icio.us, blogs, Flikr and Youtube across a spectrum of media (words, images, sound, video) that either bring things online to whare or find things online to share. For conversing, it was listservs earlier and now its social networking sites for an extended multimedia exchange of information beyond mere sharing. Collaboration, as working together, ups the ante; Shirky’s contemporary example is Wikis, which have a pre-history in groupware. For collective action, his examples of enabling software are fewer: Meetup is his best, and the rest of the discussion goes soft.

What’s going on here is not just matching tools to functions or, from a the point of view of designers (whom Shirky teaches), designing a tool for a function that will take that function (sharing, conversing, collaborating, organizing) on-line. There is an underlying theoretical commitment to a view of on-line life as realised latency. Social software is in this view almost neutral, and ideally transparent. The problem is a deficient view of what’s social in it. It’s a tool-user’s view, which strips down the social to functions.

One needn’t be a technological determinist, or even have recourse to medium theory, to see that tools are not just functional; and you don’t need critical theory to see them as socially constructued. Tools are also institutional. Julian Orr famously demonstrated in his study of Xerox machine repairmen (Talking about Machines, 1996), a machine is not just a physical object, It includes operators and knowledge about operation. Tack back to Marshall McLuhan, who was a student of the Canadian geographer Harold Innes who translated his observations from transportation studies to the humanities. The telegraph is not just keys and wires, or even Morse Code, but also operators and their informal knowledge, which included recognition of personal styles or ’signatures’. It is also the delivery boy culture and customs that arose around sending telegrams (almost like telephone calls later). A technology includes all this; so does social software.

Shirky coined that term and gave an expansive definition (‘anything’ used on-line to conduct social interaction/relations), which we can use similarly to list what, besides code, it involves. Customs arise, even cultures, around specific social software – MySpace or Facebook, for instance. The sociologist danah boyd has observed that the user-communities for social networking software, Facebook and MySpace, tend to diverge class-wise, which she relates to their social histories, on the one hand, and to their ‘network effects’ or externalities, on the other hand. These are patterns of behavior arising apart from and not dependent on reflexive apprehension (the strict sense of what critical cultural studies focus on); and she’s careful not to attribute these patterns to the software but to their social origins and user communities, which are well beyond mere intention or commitment.

One could go further. The anthropologist and ‘digital ethnographer’ Michael Wesch observes how YouTube technically facilitates and so socially encourages ‘remix’, a particular practice that can solidify some non-individual features of relations (bonding) and extend others (bridging). To put it differently, social software, taken in the extensive sense of technologies including not only machines but their operators, practices and practical knowledge, isn’t just’ affordance’ of something ‘latent’ . Social software encode rules, procedures, practices for doing – and, beyond that, for participating in – something on-line.

So it seems more straightforward to think of social software as social institutions for arranging life on-line (as channeling social relations rather than as channeling human nature, as Shirky does). And it doesn’t merely convey social relations on line, as the ‘migration’ thesis that I and others have used would hold; not is it ‘alternative’ to non-IT-enabled interaction and expression as frequently cast in political analyses. First, it couldn’t be mere conveyance since no social software conveys more than a narrow range of behavior and none conveys the full density of social life. They don’t simulate so much as sample. Second, as rules and procedures it does not proceed from nature, but from culture. (Nothing really radical here. These are actually fundamental discoveries on which the modern social sciences were founded – that is, to deal with data that do not depend on and cannot be reduced to properties of individuals or facts about individuals.)

Such a perspective, for one thing, gets past the lingering uncertainty in trying to match particular bits of software with particular types of social behavior, which is that the matching is never one-to-one. Listservs may be used for sharing, conversing, even collaborating; while wikis may be mostly about collaborating, they record conversations and manifest sharing. LIkewise, blogs could be sharing (as publications) or conversations, and so can social networking sites. These relationships are indeterminate. They’re not so indeterminate, howeveer, when social software is thought of as institutions.

First, SS corresponds to the fundamental feature of social institutions, what Durkehim called ‘exteriority’ and ‘constraint’ (to individuals), no matter how much simplified net-praxis theory (you are what you do) would have it otherwise. This doesn’t mean they’re determinative of individual behavior, because their second feature is what Goffman called ‘framing’ (Frame Analysis, 1974). Social institutions, met in the strong form as rules, in weaker ones as expectations, and somewhere in between as procedures, are experienced as frames of reference, subjectively. Such experience doesn’t depend on personal commitment to them. Actors can be alienated from a frame of reference as ontology but still register it, even accurately, as epistemology. You may be alienated from while still thoroughly comprehending corporate life. Indeed, this can be a powerful experience of ‘newbies’, whose socialisation into an institutional way of doing things is partial and on-going; it has not, in the hard form of ‘total institutions’ (Goffman, Asylums, 1961), obliterated memory of before or of any place to stand outside.

Thanks to Wesch, we have fairly clear start on understanding how this happens with YouTube, or “on” YouTube as a platform for constructing and enacting social relations (identities, interactions), or to boyd for social networking sites such as Facebook and MySpace that could be compared to earlier ones like Friendster or more specific ones like LinkedIn. Backing away from the limited sociology needed for media studies (audiences, preference profiles, essentially the sociology of crowds, no matter how ’smart’ the mobs), much as media studies backed away from tech studies (machines as automation) that also limited sociology (essentialy to the structural-functionalist moot that structure determines practice that engineers learned from efficiency studies), what does social software look like as social institutions.

First, as tools, they’re designed to perform a function.

Second, their functions are far less calculation than connection – connecting persons to persons, persons to ideas, ideas to ideas.

Third, they’re not just in but about what’s on-line. This is not a remnant of the notion of cyberspace as a sort of virtual place, mundanely recognised to be a vast realm of information, but a measure of reflexivity.

Fourth, social software offers/promises/is designed to introduce localised measures of organization in this vast sea. Designers dream of their application becoming a ‘platform’ much as the browser is imagined to have, then the portal, and now as some claim social networking sites become for youth.

Even the seemingly simplest and least ’social’ acts, searching for information on Google registers its now-institutional character. It turns that program, as a ‘platform’, into order of the space, which Google is ordering not as a flat data field but by adding the algorithms of users’ searches to its own indexing. Facebook and MySpace, Flikr and Youtube up the ante in trying to match that success, as does socialbookmarking and, its mirror image, wikis, both of which bring bits of information together by structuring relations to do that (membership criteria, rules of behavior, rewards and sanction systems). Likewise, blogs form a continuum between between private musings written down and publications composed for an audience, which is reflexively registered as a tension (and in debates about) self-revelation vs. self-promotion. Quite apart from how actors are so motivated (by human nature, by reference groups), they are able to register and join in these practices. So, what gets in the way of seeing this?

Shirky’s is a sociology for media studies that captures levels of commitment plus density of interaction with an eye to how these can be scaled up through social software, which breaks barriers to connecting with others (to a lesser extent with other ideas). Basically, an Internet-updated version of the Global Village. In a sense, a sociological sense, this is going backwards, for the great discovery of modern social science was how to think coherently about the order not at the level of villages (like-mindedness, multifunctional institutions and few of those) but of complex industrial societies. It’s key was the discovery, or conceptualisation, of institutions; in modern, industrial, complex, mass society you know the institution instead of its individual personnel, bakers and not the baker – or as Simmel (1907) put it, The Stranger as a type, not individual strangers, who if known as individuals wouldn’t be strangers any more. No one’s been nostalgic about this sort of society. Witness Robert Putnam who feels empty when the last vestiges of village-like community seem to recede and what’s left of Metropolitan Life is nothing; it doesn’t have its basis in personal commitment + interaction. It’s diagnostic figures have run from Simmel’s ’stranger’ to Whyte’s ‘organization man’.

To a considerable extent, the Internet was greeted as a way out of this sort of world, while preserving its advantages, not least because that was the way engineers who developed it cast it socially, as renewal of agency. Tthough many have hoped it could, from Howard Rheingold’s Virtual Community (1993) and Lawrence Grossman’s Electronic Republic (1994) to Negroponte’s Being Digtal (1995), it would not seem likely that the Internet is, or could be, taking us back to a pre-industrial sociality.

Cultural critics writing from the other end of the spectrum have a clearer picture of the Internet as a system and where that is taking us, perhaps too clear a picture, and it is not that direction. It is into a world of out-sourcing, free-lancing, work given away while income is made from derivatives and modeled on selling property from selling the business to collecting royalties. Flexible actors for a political economy of flexible accumulation replace the industrial period’s political economy of capital accumulation. Fine as far as it goes, this tends to dichotomise micro-level and macro-level analyses and to miss what’s the sociological middle ground where parameters of order not so much built into social software either by Free Agents or by The System but that social software builds for users.

If we go back to specific social software, we can see how this works. In one of the better pieces of inside reporting on Silicon Valley, Sarah Lacy’s Once You’re Lucky, Twice You’re Good (2008) casts as a tale of Web 2.0 entrepreneurs’ having learned lessons from the 1990s dot.com boom-and-bust to create platforms. To an anthropological eye, hers is an ethnography not just of re-negotiating the developer identity but of developer practice into an extended process of trying to write some social behavior into code, continuously modified or ‘tweaked’ into ever denser (additional) social practices. The goal, as earlier, was to attract users by encoding their desired but especially already practiced uses; the innovation was imagined to be getting out of the way of users’ interacting with each other at a distance with users and using releases to test whether or not they had added desired practices. Engineering values (such as elegant code) as well as practices (testing, quality control) went by the board in favor of social values (connecting to stuff on the Web, to people on – not just through – the Web) and practices (gathering, identity). Developers a sensibility for the Web as a ‘platform’ socially for reference groups, whose structuring principles they would instantiate in code but that, objectively speaking, code provide as the structure of life on-line in the form of routines, scripts, practices for which users would turn to their programs. This is not just agency enhanced or community virtualised but something in between, or that combines both, a sensibility for reference groups as more than categories.

The significance of providing structure for users (aka, ‘platforms’) comes with recognizing practices that can be captured as ‘meta-data’, which would be data about the social fields themselves. Why? These are unconscious patterns that, as categories, are features of structure. Users described in these studies are not ‘using’ the Internet to get to something else (the functional account), nor exploring another continent (the reification account); they not connecting through the Web or with the Web as an object, so much as they are connecting to its properties in the way that people conform to institutions. Directed by institutions, enclosed within them such that whatever else their actions may ‘mean’ in some other terms, they have specific meanings (intentions, interpretations) within the system of actions in which they participate.

When the Internet was young, listservs and newsgroups provided this sort of structure, but not enough to prevent ‘flaming’ or to keep users on topic. They had to do that independently of the software, or not do it at all. Particularly to particularly to early users who were also the Internet’s developers, it registered as a transparent medium; and there was a sense of beginning with a purer or stripped down social environment. It is the addition of sanctions (membership criteria, moderators, content filters), however, that doesn’t just co-opt or ‘appropriate’ the space but, instead, complete its sociality. Contemporary ’social software’ starts off being more social by operating in an environment that is already more social, such as represented in the notion of an application, like the browser, becoming or being transformed by other applications into a platform. This seems a lot less a matter of releasing latent potentials than of structuring behavior on-line with an eye to what’s on-line as well. This is the recursive, self-referential and, experientially, self-evident property of institutions, which are not reducible to aggregation of individual behaviors.

The issue whether people bring their culture(s) on line or fine one there is moot, as is the conflict viewing software as ‘affording’ on-line something already available off-line, on the one hand, or structuring practices directly, on the other hand. So, I think the more productive way to view ’social software’ is neither as determinant of action nor conduit of action – the ‘hard’ and ’soft’ constructions of ‘channel’ – but as institutions in on-line social life, that institutions typically are multi-functional, that they involve both overtly recognized and covertly realised patterns of behavior, that some of those may be reflexively registered in the form of myth and ritual that re-present them in ‘other’ words (concepts) and deeds (actions).

Explore posts in the same categories: online social life, social software

Tags: Erving Goffman, jordan, McLuhan, Simmel, social software, socialnets

You can comment below, or link to this permanent URL from your own site.