Body in a Coffin
March 03, 2003
A series of events in the last seventy two hours or so has done some damage to the words I've been using to describe the philosophy that supports my attitudes towards learning, communication, collaboration, tool use, knowledge transfer and a bunch of other stuff. I'm doing some ThinkOutLoud with this entry, so excuse my mess. (0000A3)
Last summer, to gain some credit for my now completed degree, I worked on a couple of independent studies. In one, my research partner and I workshopped a paper in support of formal knowledge access structures for reference, but opposed to formal knowledge representation for communication. We ran up against some opposition from some of the participants who felt we were being scruffy and illogical. (0000A4)
Being scruffy is one of my identifying characteristics and I hope I stay that way. (0000A5)
The other independent study was a readings group called Augury. We created the group in an attempt to discover connections between the philosophy of Doug Engelbart, as described in Thierry Bardini's ''Bootstrapping'' book, and embodied cognition as described in Andy Clark's ''Being There''. (0000A6)
In a fit of zealotry I went off and found what I was looking for: pivot points of connection between the two sets of ideas. It was an extremely productive time for me. I solidified many ideas that had been percolating for a few months, especially ideas associated with knowledge transfer and tool craft. (0000A7)
It was good and it was fun but unfortunately it was built on a stacked deck of cards. (0000A8)
Poupou's in a cogsci class in which they talk about brains. She has some issues with the class and the participants. A long time complaint with cognitive scientists has been that they tend to think of the brain as an entity in a jar that, with perfect information, can be recreated. I've never liked this mechanistic view of things. Embodied cognition, for me, represented a way out of that morass: brains with bodies living in the world, participants in very complex and infinite network. Poupou's comments about her class and this diatribe about embodied cognition (found by my handy little GoogleTracker whoosit) show that's not really the case. I had taken embodied cognition to mean cognition of an individual in their sensory network situated in a world environment; interactive cognition; informed cognition; world conscious cognition. Turns out, for a lot of those folks it just means having some limbs that bump into the things. (0000A9)
So we go from brains in jars to bodies in coffins; a slight movement of the membrane outward. Wee ha! Go team! Great, cheers, thanks a lot; I still think I'm right: my interpretation of embodied cognition, maybe call it environmental cognition (although it looks like maybe that's been taken over by people who think of the environment as "out there" while the brain is "in there"), and the roots it finds in folks like Heidegger and Bateson, is still strongly supportive of Engelbart's philosophy and human-centered improvement and agumentation. (0000AA)
How does this fit in with formal knowledge representations? The gang over at the Blue Oxen Collaboration Collaboratory are having a chat about structured-dialog systems. Many of the participants in the group are big fans of outliners and IBIS. They're neaties, not scruffies. There is some historical precedent that tries to put Doug Engelbart in with the neaties and Ted Nelson with the scruffies (I think this is not right, but people seem to enjoy the fight). Some of the collaboratory members appear to feel formal knowledge representations are a personal aid in communication. I heartily disagree with this principle. Formal knowledge representations are a secondary or tertiary step: communication first (scruffy communication) followed by digestion and summarization. One of the participants touches on this. A process of facilitation is required, either by oneself or someone else, to get to the "formal" state. (0000AB)
Once in the formal state, the information (or knowledge, if you insist) is available for reference and is valuable as such. It is not, however, valuable as communication. Communication is what people do with one another, situated in their worlds. If one insists in the use of formal representations for communication and in the belief of an achievable truly shared ontology one is there in the coffin, not too far from the jar, being a brain, without a world. When the world's not really there, what happens in the brain can sometimes be far more important and enjoyable than what is going on in the world and action outside the world may slip away in favor of a process of regular and eventually useless brain lubrication. (0000AC)
Comments
Not enough time to comment on everything you touch on, but note a coupla things: (1) John Lilly's (and apparently others') experiments with float tanks and sensory deprivation provide significant support for your notion that cognition has a lot to do with environment. Certainly, the self does not appear to be "self-contained" as it were. Or perhaps more accuarately, the "self" contains less than we expect. Obviously, getting your definitions straight is important in discussing these topics. I recommend asking Ken and Brian Bothwell for references on such matters. We discussed it just the other night but I don't remember many details. (2) I think you're taking too black/white a view of "this works"/"this doesn't" when it comes to facilitation, collaboration, etc. Clearly a continuum of things work in different situations and with different people. Remember the situations in which various scruffy approaches alienated and scared people? To ignore this is to ignore a major human criterion in evaluating techniques. (And, beyond, in coming to any higher understanding or theory of collaboration.)
But scruffy is more fun.
1)Getting the definitions straight is part of what I'm railing against. I think in "reality" you can't really get your definitions straight, but you might be able to get them near to one another.
But you're right, you have to lay enormous groundwork to get to the same page to even really begin to have the conversation. It's part of why people like Jack are into the Glass Bead Game because it has strict rules about ontological purity and part of why I often feel like I'm speaking a different language in the collaboratory.
I yearn to get in a sensory deprivation tank.
2) The black/white view is mostly just for exercising the thinking (out loud). What "works" is context dependent for any situation and different for each participant. One person can walk away thinking success while another thinks failure and another thinks "eh".
2a) I'm not convinced that a theory of collaboration is available. In fact I'm not sure that collaboration as an independent entity exists. People collaborate to do things. The things and people part in that statement are what matters; they are the powerful part.To talk about collaboration without a goal feels like masaturbation: pleasant and good for you but not particularly productive outside oneself. I don't mean to claim that sex is for reproduction alone, but that it does do a lot for communication.
SB pointed out that I may be causing some confusion in my use of the term representation. I don't mean the cogsci version of representation wherein the brain has an internal model or prototype of "out there". I mean representation in the sense that it is used in library science: ways of representing information in the world. Binary 0101 is a representation of the number sometimes known as five; the RSS of this blog is a (shortened) representation of the content I create.
My argument is that formal, structured representations are good for making reference but not good for immediate conversation between people--things like IBIS gain effectiveness in the later stages of collaboration, not in the immediate moment of communication. Later may come very soon after now or much later.