IEEE Xplore is transitioning to HTTPS on 9 April 2018. Customer access via EZproxy will require version 6 or higher with TLS 1.1 or 1.2 enabled.
Review our EZproxy Upgrade Checklist to ensure uninterrupted access.
By Elizabeth Cha; Yunkyung Kim; Terrence Fong; Maja J. Mataric
Now Foundations and Trends Books
The primary expression of biological intelligence consists in the production of an ongoing sequence of actions that flexibly respond to a constant stream of incoming environmental stimuli. Recent research into artificial autonomous agents (animats) suggests that a great deal of such adaptive behaviour may be grounded not in the systematic activity of internal represeiltations, but in complex interactions involving neural, bodily, and environmental factors. In this paper we investigate the threat to representational explanatory strategies posed by this form of causal complexity in behaviour-generating systems. First we try to identify the true nature and extent of the threat, by considering some telling examples of animat research. Then we pursue an extended analogy between the role of neural states in the causal determination of adaptive behaviour, and the role of genes in the causal determination of phenotypic traits. On the basis of this analogy, we develop a concept of genie representation that, we suggest, might offer a way forward. A crucial part of the genic strategy is to appeal to the normal ecological context of the inner states in question, in order to balance (a) the fact that adaptive behaviour is often the result of complex multi-factor causation, against (b) the idea that the notion of internal representation plays a special role in explaining such cases.