As Kasparov (2003) pointed out, Deep Blue not only was unaware of the fact
it was playing a world champion, but had no self-awareness that it was win-
ning or losing. Such lack of self-awareness and consequent emotional reac-
tions would be potentially devastating for human players, because this cru-
cial piece of information would motivate adaptive strategic adjustment (e.g.,
to fight back).^3
The failure to consider subjective experiences also creates blind spots such
as how a thinker’s values, attitudes, dispositions, self-understandings, and
beliefs guide his or her thinking. Because cognitivism focuses on the formal
or syntactic aspect of symbol manipulation (Smith, 2001), and neglects men-
tal or semantic contents of one’s directed consciousness or intentionality
(Searle, 2001), what gets obscured is the entire issue of how the culture, with
its rich historical legacy, enables our thinking through language and other
conceptual tools working seamlessly but potently in an intersubjective world,
without which most of what we call intellectual development is simply out of
the question (Gardner, 1985; see also D’Andrade, 1981, 1995, for a discus-
sion of differences between computer programs and cultural programs of
cognition). The very Kasparov phenomenon (or the phenomenon of Deep
Junior, for that matter) cannot be understood without the proper context of
cultural values, incentives, tools, and resources (including a body of the codi-
fied chess knowledge, coaching, tournaments) supporting the development of
chess expertise.
The Trend Toward Integration
What we have witnessed since about 1990 is, to paraphrase Bruner’s (1994)
comments, a “renewed respect for a rather classical form of functionalism”
(p. 277) that tries to situate perception and cognition in a broader functional
context of human adaptation. Such a change logically calls for a more inte-
grated understanding of intellectual functioning and development. As Newell
(1988) pointed out, cognitive psychology started with elementary cognitive
processes, and only gradually shifted its focus to higher levels of purposive
behavior. Such a shift necessarily brings the whole person and functional
- BEYOND COGNITIVISM 7
3 3 There is a debate as to whether computational models are capable of derived intentionality,
albeit the fact that it cannot produce real conscious experiences (e.g., Dennett, 1991; Searle,
1990). G. Matthews (personal communication, May 12, 2003) pointed out that consciousness
and intentionality are beyond the computational metaphor, but many of the functional attri-
butes of conscious states may not be. Our focus is how the human mind works. Whether compu-
tational models can simulate functional properties of mental states and acquire derived
intentionality is another question. To the extent Deep Junior does not have a functional property
resembling human emotional reactions to an imminent loss or win, we can say the system is not
embodied.