A Conversation on Social Interaction Design with Adrian Chan

by Thomas Vander Wal in , , , , , ,

Earlier today Adrian Chan and I had an e-mail exchange that both of us enjoyed and agreed it would be good to place it out for public consumption on our blogs. The ideas and concepts differ in their focus and approach, but are similar in that they are trying to reconstruct a much improved social interaction design understanding than the light understandings that are perceived and built upon in many of the social service on the web today.

What follows is the same content (directly from our e-mail exchange) that Adrian posted earlier today. It is rough form (I added markup for clarity in structure, we didn't use bullet points in our e-mail), but good understanding of what we are thinking. What I am discussing is a small part of what was in my recent workshop from last week.

On Jul 21, 2009, at 8:59 AM, Thomas Vander Wal wrote:

The conversations models & how they map to the difference faces & steps in the communication progression from personal, collective, community/group, and collaborative have interest to me. Each are different design problems with very different interaction & communication needs, hence leading to different conversation models.

  • Personal: Focussed on holding on to objects (including people & relationships) and annotating for refinding and aggregating as needed.
  • Collective: Open sharing/stating around objects (with various possibilities around level of sociality) with some conversation directly with them in comments, but also indirect conversations (friendfeed, microsharing, etc.)
  • Community/Group: Fully aware of others with interests around the object and interacting with the others in a manner that is open to others in the community/group.
  • Collaborative: Goal is getting down to one view and one product. This requires the means to identify and work through conflicting concepts and understanding. Requires working together and identifying, addressing, and working through conflict to come to one resolutions (there can not be more then one personal day policy in an organization).

On Jul 21, 2009, at 12:25 PM, adrian chan wrote:

these are cats used by ross, clay and others that i'm not totally aligned with. primarily because I don't think they reflect anthropological or sociological distinctions in interaction systems or situations. (e.g. paired interactions, triangulated interactions, group membership, inter-group interaction, alliance, family, tribe, community, or now the social media-specific formations which seem to be "invisible audiences," "publics" or "audiences" depending on who you talk to.)

for example i don't think "collective" is a natural social phenomenon but if it occurs is a byproduct or outcome of carefully structured interactions in which personal social dimensions are minimized to reduce the bias of status, rank, hierarchy and other attention-getting behaviors. Which is why Hunch.com has shirky written all over it, or why we all use wikipedia as our reference standard for collective action!

in other words,

  • a structuralist would tell us that these categories don't exist.
  • a sociologist would say that forms of communication and social practices transcend these categories and may be found in the reproduction in any of these categories, so cant be the causal explanation for how these categories of content production are realized.
  • a psychologist would say that user motives are not a reflection of a kind of social arrangement, that for example interpersonal stuff, attractions and flirting, lurking etc can all occur in social groups of different sizes and structure
  • a social media theorist might say that it matters more how people see others, see themselves, and think they see how they are seen by others, and that the constraints on action in and results out are what govern behavior -- but that users wont have "collective" or "collaboration" etc in mind when they're acting -- that user centric view will prevail over an architectural one

i think where shirky has a blindspot is in motives -- he's a good pattern recognizer but patterns can be effects without being causes, or without being the goal or the motive of a certain user's activity.

where shirky sees structure as a way of possibly eliminating social distortions, i still think it's essential to know how the user sees himself in the social field to know where bias may be introduced.

and in today's highly conversational mediaverse, these structures are hard to map to aggregation, disaggregation, and other twitter/status feed phenomena. twitter and its kin are so fluid, so ephemeral and time-based, that it's hard to grasp the causes of social outcomes without using communication theory and interaction dynamics (which i sloppily call "conversation models"). challenge being that one has to capture what interests a user -- could be their own status, could be their reputation, their commitment to a higher goal, their need for attention, etc, all of which come out in conversation but none of which are governed by structural arrangements (like collab, collective, or community)....

in short the question you raise is: does the social order account for user behavior? Is the social order the user's orientation. I don't think it is, but that would be my bone to pick with ross or shirky (some day....)

what do you think? am i making sense?

On Jul 21, 2009, at 9:52 AM, Thomas Vander Wal wrote:

Your approach makes sense and fits wonderfully within social comfort. One of the things I have found working with organizations on the inside is the assumptions from the outside (open web tools) are broken. Adoption of the same patterns outside don't happen inside organizations, as the measures are vastly different (outside pure numbers (100k to millions of users) and inside is percentage of employees/customers). Our assumed understanding for tools and models from web 2.0 don't really work well when dealing with closed populations. What we realize is these tools are less than optimal on the web too. This was my huge problem in writing my book (Understanding Folksonomy) for O'Reilly, I could not explain value that was derived nor could I explain things that were broken.

Conversation models fit nicely in social comfort, which I currently have set within the elements of social software and build order. Unless the prior elements are met, there is no communication/conversation. The realm of social is far more complex and runs on many different planes and models at once. There is no pure model, but a mixture of models and understandings.

The elements of social software and social comfort are important in all of the faces of perception (where personal, collective, community, collaboration, newbie, system owner, and external developer) come into play as task roles. But, seen from the perspective of a cube or other polygon, we can see many sides at once and are participants in the various tasks and faces.

I agree and disagree with "but that users wont have "collective" or "collaboration" etc in mind when they're acting" as I see the mindset of whom am I sharing with (how broadly) and goals (stated or inferred) with the task type, when users are interacting with others on internal social tools. But, it is not the user's perspective that is at the forefront as much as it is having the proper tools with the proper elements to achieve each type of task. Most organizations do not think of the progression of tasks and ensure their tools embrace the needs at the various stages. Often true collaboration elements are missing as well as desperately needed tools for personal tasks.