Computer Science at Kent Semantic Modulation of Temporal Attention: Distributed Control and Levels of Abstraction in Computational Modelling

1 Introduction There are now many different approaches to the computational modelling of cognition, e.g. and neurophysiologically prescribed connectionist models (O'Reilly & Munakata, 2000). The relative value of different approaches is a hotly debated topic, with each presented as an alternative to the others; that is, that they are in opposition to one another, e.g. However, an alternative perspective is that these reflect different levels of abstraction / explanation of the same system that are complementary, rather than fundamentally opposed. Computer science, which has often been used as a metaphor in the cognitive modelling domain, gives a clear precedent for think in terms of multiple views of a single system. In particular, in computer science, single systems are routinely viewed from different perspectives and at different abstraction levels. An illustration of this is what is now probably the most widely used design method, the Unified Modelling Language (UML) (Booch, Rumbaugh, & Jacobson, 1999). This approach incorporates multiple modelling notations, each targeted at a particular system characteristic. Furthermore, UML is not unique in emphasizing a multiple perspective approach; see, for example, viewpoints (H. It is not that this perspective has been completely lost on cognitive scientists; indeed, Marr famously elaborated a version of this position in his three levels of cognitive description (Marr, 2000). However, despite Marr's observations, concrete modelling endeavours rarely, if ever, consider multiple abstraction levels in the same context and particularly how to relate those levels. 1 Full reference is, "Semantic modulation of temporal attention: Distributed control and levels of abstraction in computational modelling.