Saturday, October 19, 2002

Seamfulness - or, "Where did that joint go?"

I was recently pointed towards the fascinating work of Matthew Chalmers and a concept for ubiquitous computing he is working with. The disappearing or invisible computer has so far been identified with seamlessness - a notion largely equated with physical and material reality, but somewhat poorly adaptable to articulating human interaction with computers. Weiser and Seely-Brown called for a focus on calm computing, or a way of interaction that centres on peripheral engagement. The basic idea is that we may not be explicitly attuned to matters on the periphery of consciousness, but they nonetheless profoundly shape our everyday interactions with the world. As such, computers may be more "effective" if they are able to work in the background.

I believe the main problem with seamlessness - or the relative inability to distinguish computers in the world - is the subsequent difficulty in locating agency and accountability. So I was thrilled to learn that Chalmers proposes the notion of seamfulness, an intelligent and informed attempt to expose users to the points and spaces of interaction with the machine. He is interested in "relating contemporary semiology/philosophy to computational representation... drawing on work in linguistics, architecture, neurophysiology and philosophy, trying to understand the similarities and differences between the different fields that deal with the human use and interpretation of information." Cool. And his Equator project focusses on "going beyond the traditional and naive way of treating digital and physical media as separate 'worlds'. Human activity continually interweaves them and makes them interdependent, and so Equator intends to treat them as two halves of the same world. Equator works on the borderline between the two." Still cool.

So here are my questions: First, are interwoven and interdependent fields understood to be (actual) self-contained parts of the same world? Then, what kind(s) of space might constitute(s) the borderline between fields? And perhaps most importantly, what sort of movement occurs between fields? In other words, what and who are constructed in our representations of these spaces of interaction?

I would suggest that there are no such "things" as closed fields - and I am interested not in how these fields interact with each other, but rather in how these fields and interactions emerge as such. I guess this means that, ultimately, I am concerned with the limitations of representation and how we may therefore account for emergence.

Many of you are familiar with the popular scientific press on networks and emergence (such as books by Steven Johnson or Albert-László Barabási), or if you want something more hardcore academic, Greg Smith's recent paper, Notes on Interaction. Within this larger body of work is the notion that "real time systems cannot be modelled as algorithms for 2 reasons: the TM [Turing Machine] formalism is independent of time and it is required that an algorithm eventually halt whereas a real time system need not. [And accordingly] both concurrent and object oriented systems are not formalisable as algorithms."

Smith continues to explain that "Applied to design, this means that emergence occurs when an agent perceives or explains some property in a working design that it would not have been not capable of computing given it's current knowledge and bounded rationality.... Supervenient properties are those which can be explained in a reductionist manner, and emergent properties are those which cannot." According to Wegner, "interactive systems are grounded in an external reality both more demanding and richer in behavior than the rule-based world of noninteractive algorithms," which Smith continues, "makes explicit the notion of interaction: the agent perturbs its environment and then looks to see its impact before deciding on another action... Interactive agents can make nonenumerable distinctions about their environments. It is this ability to make determinations not predictable before environmental interactions have occurred that enables emergence."

Okay, I know that's a lot to absorb, so let's put it back in social terms. If society is a bounded entity, then the performance of the social or sociality (and of virtual spaces) is unbounded, emergent and unpredictable. So how can we represent - and design for - that sort of space and interaction? Social and cultural theorists, architects, designers and artists have been asking these questions for a long time, and my first reaction is to move towards what is called "non-space," defined through notions of mobility and movement. In this sense, it is not the place that is of interest as much as what moves through the space. Airport lobbies, motels, nomads, vagabonds, ships and cars can all be seen to embody the simultaneous movement of people, objects and ideas. Their very hybridity and instability make them difficult to represent through static models and maps, which at best represent partial truths. This leads me to question how we can even know these movements - and I worry that there is no "external reality" to which we may look for answers, and that the computational notions described earlier rely on our ability to recognise and understand intention or motivation in (linear/real) time and space.

Now here's the killer for those of you with the patience to have followed these half-baked ideas of mine: for the fully-baked version we will all have to wait for me to finish my dissertation. And since that's still a long way off, I only hope you'll hang around for the ride.


Post a Comment

<< Home

CC Copyright 2001-2009 by Anne Galloway. Some rights reserved. Powered by Blogger and hosted by Dreamhost.