Systems, Interactions and Macro-theory

Philip J Barnard

MRC Cognition and Brain Sciences Unit,
15 Chaucer Rd,
Cambridge CB2 2EF
phil.barnard@mrc-apu.cam.ac.uk

The ecosystems populated by information technologies have diversified to an extraordinary extent - from line editor to embodied conversational characters and from radar to virtual battlefields. The diversity of applications has been mirrored in the ways in which we have come to think about interactions with technologies. With each advance, new empirical techniques are brought centre stage,more often than not coupled with changes in the theoretical models of, and perspectives on, interaction. HCI is not simply "interdisciplinary." It is now hard to think how we might put any kind of effective boundaries on what disciplines and topics aren'tof potential relevance to the study of HCI. In a scientific context where many things are of relevance, a key issue is the management of analytic complexity. It will be argued that there is a pressing need to go beyond the development and validation of "local" theory and to put in place macro-theories than structure and bind local theories or concepts into a coherent whole.

This presentation will examine the requirements that a body of macrotheory should meet. In doing so it will draw upon earlier theoretical work dealing with interactions within the human mental mechanism (Barnard & May, 1993) and upon concepts developed within HCI specifically to capture properties of interactions - syndetic modelling (Duke, Barnard, Duce & May, 1995),and Interaction Framework (Blandford, Harrison & Barnard, 1995).
Whether dealing with individual human cognition, a human interacting with a computer, a small group collaborating on a task with the support of a range of technologies, or even an entire organisation, each level can be considered as relying on a "system of interactors".

Any given system has at its heart a set of basic units that "behave". The basic units at any given level can be decomposed into their constituent parts and, the basic units themselves form part of a superordinate interactor. The behaviour of a system of interactors can then be thought of as a trajectory of interaction, constrained by the configuration of interactors, their capabilities, the requirements that must be met to use that capability, and the regime through which the interactors are co-ordinated and controlled. When viewed this way relatively simple theoretical abstractions of "interactions" can applied applied across levels of analysis and macro-theories developed to bind content at one level of analysis to that at another.