Accounting for Creativity Within a Psychologically Realistic Cognitive Architecture - Computational Creativity Research: Towards Creative Machines - page 146

Cognitive architectures, in general, are becoming increasingly important in
cognitive science, in psychology, and in artificial intelligence (AI) [ 17 ]. Among these
cognitive architectures that have been proposed, the CLARION cognitive architec-
ture [ 26 , 27 , 31 ] tries to provide amore unified explanation of a wide variety of
psychological phenomena. It tries to do so using mostly five basic principles: (1)
The co-existence of, and the difference between, explicit and implicit psychological
processes; (2) The simultaneous involvement of implicit and explicit processes (in
most tasks); (3) The “redundant” representation of explicit and implicit knowledge;
(4) The integration of the results of explicit and implicit processing; and (5) The iter-
ative (and possibly bidirectional) processing. This cognitive architecture has already
been used to account for many psychological phenomena (such as implicit learning,
bottom-up learning, cognition-motivation interaction, creativity, and so on) and to
simulate a great deal of relevant human behavioral data (e.g., with respect to low-
level skill learning, high-level cognitive skill acquisition, and reasoning; see e.g.,
[ 26 , 32 , 33 ]).
In relation to problem solving, some existing psychological theories of problem
solving and reasoning have highlighted a role for implicit cognitive processes. For
instance, implicit processes are often thought to generate hypotheses that are later
explicitly tested [ 10 , 13 , 15 ]. Also, similarity has been shown to affect reasoning
through processes that are mostly implicit [ 25 ]. Yet, most theories of problem solving
have focused on explicit processes that gradually bring the problem solver closer to
the solution in an explicit, deliberate way. However, when an ill-defined or complex
problem has to be solved (e.g., when the initial state can lead to many different
interpretations, or when the solution paths are highly complex), the solution is often
found by sudden 'insight' [ 4 , 22 ], and theories of regular problem solving are, for
the most part, unable to account for this apparent absence of deliberate processes.
Hence creative problem solving needs to be examined specifically.
Research on such creative problem solving has tried to tackle more complex, more
ambiguous problems. However, psychological theories of creative problem solving
tend to be fragmentary and usually concentrate only on a subset of phenomena, such
as focusing only on incubation (i.e., a period away from deliberate work on the
problem; see [ 24 ]) or insight (i.e., the sudden appearance of a solution; see [ 22 ]).
The lack of detailed process models (e.g., detailed process oriented computational
models) has resulted in their limited impact on the field of problem solving and
creativity [ 8 ].
In this chapter, we explore an integrative theory of creative problem solving that
is based on a psychologically realistic cognitive architecture, that is, the CLAR-
ION cognitive architecture. The integrative theory and the cognitive architecture
on which it is based will hopefully transcend the shortcomings of many existing
models/theories, and address many relevant aspects of creative problem solving,
from incubation to insight, and from motivation to personality.
The remainder of this chapter is organized as follow. First, we discuss the rele-
vance of psychologically realistic cognitive architectures to AI, cognitive science,
and psychology. Second, the CLARION cognitive architecture is sketched. Third, the
Explicit-Implicit Interaction (EII) theory of creative problem solving, derived from