East west unity

It is rare in the natural sciences to find agreement between Soviet and western approaches. Indeed, the approaches adopted and the subsequent theories developed by the two schools frequently appear not to speak to each other. To some, this counts against the objectivity of natural science. To others, the differences in perspective beg to be unified.

William Frawley is one of the others. In Vygotsky and Cognitive Science, Frawley reminds us of the continuing relevance of early Soviet psychology to the understanding of cognitive development and attempts to unite this with modern-day (western) computationalism.

At first sight Frawley's programme appears to be headed for failure: the contribution of the Soviet school, and Vygotsky in particular, to cognitive development centres on the view of development as the internalisation of social processes, but computationalism typically pays little if any attention to the position of the individual in society, or to the consequences that might ensue from that position for cognitive development.

Indeed, computational approaches to development have tended either to side with nature in the nature-nurture debate, or, if siding with nurture, to ignore social processes. Most readers will therefore approach this book with extreme scepticism tempered by intense curiosity. Frawley begins his quest for unification with a thorough, and frequently masterful survey of the background terrain. The breadth of coverage is impressive, although the material is difficult at times.

However, readers who survive the foundational chapters should find the second half of the book more straightforward. It is here that Frawley lays out the connections between Vygotsky and computationalism in his presentation of "socio-computationalism". Though more a position than a theory, Frawley does explore two principal theses: that consciousness and meta-consciousness are distinguishable and dissociable; and that a private mental control language (a language for thought) develops from public, social speech.

The first of these theses is, to my mind, relatively uncontroversial. The distinction between consciousness and meta-consciousness is essentially the distinction between being aware and being aware that you are aware. Meta-conscious awareness is a central plank in Vygostkian theory. It has also received some attention in western cognitive psychology, particularly now that the study of consciousness has become respectable. The arguments of both schools are marshalled in support of Frawley's position.

The true purpose of the book, however, lies in the second thesis. Jerry Fodor has long argued for a language of thought, a language in which thoughts are encoded and manipulated. Frawley accepts this (on the grounds of no reasonable alternative), but argues that thinking also requires a language for thought, a control language that sequences thought and action. This language is supported by the distinction between consciousness and meta-consciousness, and held to develop through the internalisation of public, social speech.

Personally, I remain in two minds. Frawley's arguments concerning consciousness and meta-consciousness are compelling, but his analysis of the development of a language for thought leaves me unconvinced. The root of my discomfort lies in Frawley's unquestioning acceptance of the computational metaphor that historically has pervaded western cognitive science. The language for thought looks very much like a kind of control language deployed in modern computers. Indeed, Frawley explicitly considers the properties of such languages and then correlates between them and his language for thought. To many, this will be overstepping the mark.

Modern cognitive science, as it moves towards distributed control that eschews sub-routines and co-routines, will not accept Frawley's socio-computationalism without considerable debate. Nevertheless, Frawley has presented a coherent position that modern cognitive science cannot ignore.

Richard Cooper is lecturer in psychology, Birkbeck College, London.