A Social Constructionist Framework for Artificial Intelligence

This research aims to build a new conceptual framework for artificial intelligence (AI) that gives priority to social relationships as determining intelligent behaviour. It starts from the premise that intelligence manifests itself only relative to specific social and cultural contexts. This is in contrast to the prevailing view of most AI research, which sees intelligence as an abstract capability of the individual agent based on a capacity for rational thought. The main consequences of the new approach are to reject the idea that the brain is a rational processor of symbolic information, and to reject the idea that thought is a kind of abstract problem-solving having a semantics that can be understood apart from its embodiment. Instead, priority is given to emotional and mimetic responses that serve to engage the whole organism in the life of the communities in which it participates. Intelligence is seen not as the deployment of capabilities for problem-solving, but as the continual and unfinished engagement with the environment. The construction of the identity of the intelligent individual involves the appropriation or taking up of positions within the narratives in which it participates. Thus, the new approach argues that an individual's intelligent behaviour is shaped by the meaning ascribed to experience, by its situation in the social matrix, and by practices of self and of relationship into which its life is recruited. This social constructionist perspective is at variance with the dominant structuralist (behaviour reflects the structure of the mind) and functionalist (behaviour serves a purpose for the system) perspectives.

The new approach challenges the radical reductionist understanding of intelligence (e.g. Paul Churchland). Other AI researchers who have challenged the reductionist view in a similar way include Drew McDermott (questioning the role of logical deduction for reasoning), Terry Winograd (priority to hermeneutics), and Joseph Weizenbaum (social/cultural context for intelligence).

The widespread interest in cyborg discourse and human nature may also offer a new way into these topics, particularly in relation to ethical concerns.

Relevant Publications/Talks
W.F. Clocksin, 1995. Knowledge Representation and Myth. Chapter 12 of Nature's Imagination: The Frontiers of Scientific Vision (J. Cornwell, ed). Oxford University Press.
W.F. Clocksin, 1998. Artificial Intelligence and Human Identity. Chapter 6 of Consciousness and Human Identity (J. Cornwell, ed.) Oxford University Press.
W.F. Clocksin, 2000. A narrative architecture for functioning minds: a social constructionist approach. Proceedings of the AISB `00 Symposium on How to Design a Functioning Mind, 30-37, Birmingham, April.
February 2000. Talk on social construction and AI, Society of Ordained Scientists, Windsor Castle.
February 2003. Talk on AI and the human person, World Council of Churches, El Paso, Texas.
W.F. Clocksin, 2003. AI and the future. Philosophical Transactions of the Royal Society.