Designing for the Collective

After reading Dav’s post about the Future Commons event, I was inspired to start thinking about what my own Theory of Everything would be. I’ve been following up on a bunch of different threads recently, and I was thinking about ways I might tie them all together.

One issue I have right from the start is that I’m not sure I believe in a Theory of Everything, despite having spent ten years of my life as a particle physicist in pursuit of same. My recent kick about conflicting realities has led me down a different path, one where global rankings are insufficient. It’s all about the context, giving people the tools necessary to adopt the global to their local situation.

Dav mentioned in his post his fondness for the Global Brain idea proposed by Howard Bloom. I was thinking about this today and got caught again in this apparent conflict between global and local. I love the idea of the collective learning machine that Bloom develops, but I want to be able to specify which elements to use. Some people (elements) are going to contribute ideas and preferences that I don’t agree with or don’t apply to me. How do I restrict the inputs to be only people whose opinion I value?

While poking around this idea some more this morning, I realized that one way to view it would be to use the language of Bruno Latour’s book, and call it “Designing for the Collective”. Latour uses the term collective to, as I put it, “indicate everything that is part of our currently described reality”. It is not so much a well-formed object, as a process, one that is “continually encountering new external influences and finding ways to absorb them, such that the collectives are always growing.” What I am really interested in is finding other people whose reality overlaps mine, whose collective I want to participate in. It doesn’t have the scope of the Global Brain, but it has a much better chance of mapping to things I care about.

What would it mean to take a set of overlapping ever-growing collectives as fundamental elements in the social design of technology? For one thing, I would need ways to find out who else might share my values and therefore be prospective members of my collective. I would also have to be able to inspect their collective (what they value) and see which elements of their collective I would like to incorporate into mine. In some sense, I’m asking that the collective’s values be exposed through technology that is accountable in the sense of Garfinkel, where they can be observed and inspected. This is kind of fuzzy, so let’s go to an example that I think illustrates the point nicely.

In a comment on one of my posts, Jofish pointed to a paper that he had co-authored on reflective design (I’ll skip what they mean by reflective design because it’s not relevant to this post, but it’s a good paper – worth reading). One of the case studies they used was of how to apply reflective design in an art museum. The idea was that patrons of an art museum were given a handheld device, from which they could request further information about works of interest. And it stored individual trails through the museum, such that one could look up what other people had seen. More importantly, one could see the trails of other patrons interested in the work that one was viewing, and use those trails to find other works that one might find appealing. It’s not a strict “majority rules” democracy that a global brain might be, only rewarding the popular works. It allows people to find kindred spirits (other collectives) and follow them through the museum. In a sense, it’s exposing the long tail of the global brain.

When I was describing this to my friend Eli, he pointed out a much more well-known example would be Amazon’s use of such technology to let you know that “Customers who bought this book also bought” or “Customers who viewed this book also viewed”. Again, it allows me to find “people like me” and see what they found interesting.

So after mulling it over today, I think that this is my current Theory of Everything: taking the idea of ever-growing collectives seriously as the basis for our cooperatively-constructed, sometimes-overlapping realities. I think that it could be used to come up with design principles that are fundamentally different than what we have now. There is no one “right” way to design something, because different collectives have different needs and different values. Figuring out ways to allow different people access to other collectives’ values in a way that respects privacy is going to be interesting. But I’m attracted to the concept of these little small-scale collectives agglomerating together in support of various things (probably my romantic anarchic tendencies showing).

I also think this idea of designing for the collective could be the basis for a larger scale work pretty easily. I’m inspired by Dourish, who uses the somewhat obscure philosophical ideas of phenomenology to construct new design principles for human computer interaction. I think that it would be just as valid to take Latour’s idea of collectives and use that as the basis for design principles in social software. The posts I’ve done on Latour could serve as the basis for the introductory chapter on his philosophy, and I could work from there. I’m going to kick this around some more, and see if I can figure out practical ways in which I can apply these ideas. Feedback always welcome.

9 thoughts on “Designing for the Collective

  1. I like this. One quibble I have, though, is that I don’t think collectives are ever-growing, but ever-changing. People leave collectives, a collective can shrink, its center of mass can shift as its membership turns over, etc.

    One of the areas where we could use some advancement in social technology is in handling arrivals and departures from collectives gracefully. We have lots of well-polished rituals for the entrance of someone new into a family (marriage, etc.), but what about voluntary departure? There are forms for the retirement party at work, but nothing so well-defined for new hires. Most collectives have no reactions in place for a member whose identity shifts (changing gender, for example)… and so on.

    I think the idea of software supporting the discovery of (long-tail) collectives you harmonize with is a great idea. And it sounds like your suggesting it should be “a design principle”, rather than the purpose of the software, yes? Just another element of usability? That’s really cool. I like it.

  2. yes. exactly.

    i think you need to start writing a book, and/or go to goddamn grad school and start doing this shit you’re good at 95% of the time instead of 5% of the time. ok, so you don’t get the big bucks for a year or two. only a year or two — you want to get a masters, don’t get a phd, it’s not important for you. but you need a chance to really be forced to think through this and do it. you’re good at this stuff and you grok it from the inside.

  3. Beemer, nice catch on the ever-changing rather than ever-growing. Although I think you may be interpreting collective too narrowly. At least in my head, it can refer not only to groups of people, but also to groups of ideas, or groups of other objects. It’s any hierarchy which can change in response to new inputs, by absorption, assimilation or rejection. I don’t know if Latour would buy that, but what the heck.

    Your mention of the new hire process is interesting, because it reminds me of this post on team building, where I make some recommendations on how to integrate new people into an existing team, or how to bring a new team together. It’s all about building trust, which is the same as building other people into your cognitive subroutines, or collective.

    And, yes, a design principle. The point is that it shouldn’t be a goal in itself, it should be part of the interface of all technology. The examples of the museum and of Amazon show the power of it. Now it’s a matter of seeing if there are other ways to leverage these ideas.

  4. Oh, I wasn’t thinking of collectives as just people, but that’s where all the easy examples are.

    (Though I might lean towards thinking that all collective contain people in addition to other things — or maybe all interesting/non-trivial collectives…)

  5. sounds a lot like the field of recommender systems research–case-based reasoning, collaborative filtering, all that amazon-y stuff. they go from the point of view that if you know a little about the target individual, you can assume a bit more based on other similar people, and over time build up a network of taste-overlaps. a lot less philosophical from the CS side of things, but the research is there.

Leave a Reply

Your email address will not be published. Required fields are marked *