- August
Posted By : Adrian Chan
Notes on a relational theory of social interaction design

The manuscript that rests peacefully in a folder on this hard drive does not contain silly opening sentences such as this one. It’s a serious, perhaps too serious, work on social interaction design. In it I attempt to lay out a theory of relations: mental relations, psychological, and social relations, as preferred by social media. Relations to which the medium lends itself, and on which it relies to supply “compelling” user experiences.

It’s a notional book, insofar as there’s no research available to “prove” what I claim. A hundred years of self psychology, slightly fewer years of sociology and anthropology, and some fifty years of media theory. But no research. That may come next — I have the research questions outlined. But research costs and takes money, and I am not an academic nor do I have institutional funding.

The first hours of returning to twitter and to blogging were interesting. After a couple months away, I quite literally struggled to compose a tweet. I quite physically battled a sense of alienation and strangeness. Have you taken a couple months off before? If not, then it’s worth a try. (It’s good to quite drinking, quit coffee, start running, and many other things I’ve not yet sustained for two months…) The computer was a computer, not yet my social world. Tweets were written, not yet a simple act of expression.

Coming back from time away provided me with a rare chance to re-experience the novelty, strangeness, and distance that new users must come through — as if undergoing the personal and social transformation engendered by tribal initiation practices and rites of passage.

On the other side of tentative, just beyond hesitation and safely past trepidation, is a zone of comfort and familiarity, friendliness and ease, with both medium and friends who use it. I feel it now as I write this. Only this time I am remind myself that I am after all sitting alone at the kitchen counter with my laptop and a cup of coffee. And that all that I just experienced was as good as an illusion if not also as seductive as a mirage.

It is this individual act of world making that I want to touch on just briefly. Because it’s at the crux of my experience (yours also?), and is the basis of my relational theory of mediated social interaction. The confirmation, mental and physical, of this world making is possible only through distance and separation from habits and routines of media use. I had to become estranged in order to then find my way in again as if as a newbie, able however to recognize and name experiences of re-entry.

Did I say yet that social media design, social interaction design as I call it, is the ultimate user-centric discipline? It is. It is understood and practiced not on a series of design concepts. Those are too neat and tidy to be good analogs of actual social activity. And besides, we know that design concepts serve the purposes of designers, are there to help designers observe, describe, think, create, and ultimately communicate. Besides which, many of them are branded by agencies and little more than the long-winded gobbledy gook and keynote pitchspeak you always suspected it to be.

User behavior and social practices of social tools are not explained, either, by social norms. Social norms are of great service to sociologists, and moreso, anthropologists. But they explain individual behavior on the basis of collective force, cultural, traditional, or other authority (right, wrong, should, shouldn’t). They presume conformity with norms, even when used to describe resistance, subversion, and opposition (for behavior then is still bound by norms). But most of all, they depend on facework and co-presence to account for their force, and media simply raise too many obstacles to normative explanations. Is the self as s/he seems; what role attention seeking; what of anonymity and play; and many more.

But why use norms when you can explain social behavior and outcomes in terms of simple, though mediated, interaction and communication? The trouble is that communication theories, speech pragmatics (what we do when talking), semantics, among other linguistic theories, are also inadequate. Any social designer knows that while icons and graphical elements used for gesture, as gifts, as symbolic tokens (to represent, indicate, reference, or suggest), etc may have meaning, their meaning in a context owes much to their use. How used, for what, by whom. In this way the meaning of a smiley is both fixed and floating. Linguistics, semiotics, and pragmatics give us theories of meaning that describe semantic meanings, gestural indications, signs, and more — but they hail from a world that is written and abstract and representational, not one of social action.

Media theories of writing and recording systems are a big help, and go far to account for distortions and system effects unique to media. Clay Shirky writes convincingly about these kinds of things. But here too, the framework describes effects on a grand scale, and which couldn’t possibly account for the individual user experience in its intention, or in its interpretation. Yes, the internet makes everybody immediately available. Yes, the internet changes the mode of production, rechannels broadcast through social interactions and across actual friend and peer relationships. But do these changes tell us which startup will succeed, and which will fail? If success depends on users?

Systems theories describe epiphenomena well, where system observation and meta data serve to render the invisible visible. In design languages, we have theories of our own, for the elements of design as for the activities of actions. Here, however, our action theories, object theories, and theories of context and use are behind the times. Adequate as they may have been to describing the design of software and systems in which objects, actions, information, and other abstractions required the benefits of consistent and durable conceptualization — these theories have not caught up to a world of mediated social interactions; of views and representations of open, apportioned, or closed publics; of disaggregated distribution and recontextualized communication. Of game or rule-bound interactions, sequenced and serial exchanges, uncoupled posts, and so on. Software design language has not caught up with social uses and is very far behind in its understanding of social outcomes.

We have, in psychology, some great insights into self and alter (the other) that serve many of our needs in explaining the behaviors seen in social media. Relational concepts like introspection, projection, externalization, substitution, and other movements of the reflective and self-reflexive, intentional, and acting subject are critical. We need them to describe how it is that abstracted communication, action, interaction, and social representations become meaningful. How it is that we see and reach through the medium — how it can be that our relations, while having no facework or co-presence, can mean “as if” as much as they would were we in the same room. But again, these notions describe individuals, not social outcomes. They need to be differentiated to accommodate many different kinds of personality, character, social skill, communicative intent, and so on. And this is not easy — particularly because we really only have made those distinctions in the pathological domain. We have no psychological distinctions with which to account for differences created by the amplifying and distorting effects of mediated, non face-to-face communication and interaction. As important as it would be to our conceptual framework, a narcissus-like concept for the reflective public sphere (I see myself as if seen by others in the medium, but those others are in part my invention, as is the social that I imagine they comprise, and so my self as seen by others is really entirely a projection of fantasy, mixed with anticipation and expectation of real and actual behavior) does not exist. The doubling of self in self image and self representation, made possible in social media, needs theorizing.


This is the ultimate user centric discipline because it rests entirely on voluntary participation by users. It depends on the user’s ability to act and behave “as if” mediation were not a factor; or to act with and through mediation, as when playing social games.

What I like about the relational approach that I’ve been working towards is its flexibility. Relations are not real. So I don’t have to go about modifying an object theory. Relations are subjective, and describe differences, qualities, directionality, and movement in the way needed to account for mediated social interaction. Objects encourage us to falsely associate causality with an objective object — as if objects were the same one and for all. User centricity means subjectivity. Subjective relation are attachments — the value attached by somebody in a relation to something/somebody is explained as his or her interest. There is no value — there is an attachment of meaning, an association, which could be a response hoped for, a result expected, an achievement earned. It doesn’t matter. What matters are the social dynamics by which some meanings become more relevant and common than others. Foursquare badges are a stupid idea, to 99.99 % of the population. They’re not stupid to those who earn them. The badge, as an object, is meaningless. The actions required to get them; the activities in which they are relevant; the meaning conferred on their owners: those are what matter. (From Foursquare’s perspective, the business relations to badges, and the creative app-like market for new badges, matter also.) It’s social engagement — the participation of others — that leaves behind the content as communication and action with which relational meanings form. And thus from which social binding as well as social meaning simultaneously emerge as new forms of social action.

I hope to get this manuscript out the door some day. It’s so much more difficult to write a book-length work than it is to bang out a blog post. If you think this sounds interesting, do let me know.



  • Adrian,

    You say,

    “Relations are subjective, and describe differences, qualities, directionality, and movement in the way needed to account for mediated social interaction.”

    I don't see how a relation, presumably between people, can be anything other than intersubjective. Maybe I'm missing something.


  • Sociologically speaking I see relations as intersubjective, yes. I can't do it justice here, but I also like Deleuze's relational concepts, among which are virtual relations. He uses the example of Melanie Klein observing a toddler sucking his thumb while crawling towards his mother. The infant is in two relations — one is an actual movement towards his mother; the other is an internal relation of satisfaction/pleasure associated with thumb sucking.

    This strikes me as interesting. Even w/o recourse to the relational concepts of psychology, there were already many subjective relations described by philosophers of the subject. Even Hume, before Kant, described associative relations that are entirely subjective.

    I need to draw on this in order to avoid reifying relations as “real” differences (causal). It's important that much of the world of meaning that makes a difference to and for a user belong to that user, and be supplied by that user. In acts as well as interpretations of others' acts, as well as reflections on each.

  • It's a very interesting topic, and I'd like of course to know more about it. I've been working for more than four years now with the psychologist and researcher Gianandrea Giacoma to what's now is forming as the Social Interaction Design discipline and it's amazing to see that we weren't wrong about it. 🙂

    Unfortunately our “ebook” is in italian at this time, but we're doing posts, talks and workshops to raise awareness about those topics. Our methodology is called Motivational Design, because it's based upon motivations. You can read a small synthesis here: http://www.scribd.com/doc/29989414/Motivational

    About the post, I think that for now you're still talking about ideas, so I'd really like to know more in detail what's your theory about. From what I read, it reminded me the Theory of Social Representations by Moscovici, that is the idea that there are social objects that are shared inside a specific group and he describes also the process how groups build that kind of social object. Since you didn't talk about him, have a look, that's a very interesting and useful theory.

  • Davide,

    I'm interested to see how you approach SxD! Motivational design sounds like a promising angle. Motives play a key role in my view of this also — though I regard them as grounding user activity primarily, and so from my perspective are not designed but designed for. (This may also be your approach)

    On soc objects, I've posted some on that. Not to reiterate but instead to summarize in brief, my framework is perhaps more a view of mediated subjectivities than social objects. I see relations as extending user interests in a wide range of mediated interactions, including those that involve what historically have been social objects. But to account for the signifying, indexical or representative functions of some types of social objects, I've added relational investments in signs and other meaning-based elements (gestures, gifts, etc). Those, to me, exceed social object definitions because they do actually have meanings attributed by a system of references, as well as having intentional meanings to users, some of which fall outside of group or context-based investments (objects as boundary objects).

  • If your interested in novel relationship theories of social interaction, try checking this out.


  • I’ve just started following you, and will keep this brief for now. I do think this is interesting, and would love to continue to know your thoughts on this as it evolves, particularly as you grow it to book length. I’ve recently seen one of your presentations about Social Media Personality types and am intrigued about the potential for research in this area.

  • Quitting coffee and stop drinking alcohol are just the two steps in making your life free from toxins. This is a good start and may you continue doing so. Kudos to you!

Leave a Reply