226 A CREATIVE COGNITIONIn the standard approaches, however, it has been a confusing area
of research and theory. In his book, How the Mind Works, Steven Pinker
mentions how psychologists feel perplexed and baffl ed about the nature
of knowledge. In a review of the subject, Emmanuel Pothos says that
“overall, there has not been a single dominant proposal for understanding
general knowledge.”^23 In the popu lar textbooks of Michael Eysenck, Mark
Keane, and others, knowledge appears to be treated as an ingredient in a
vast diversity of cognitive pro cesses but not as a subject in its own right.
In some theories, some or most of an individual’s (animal or human)
knowledge will be deemed to be innate. To most this usually means genet-
ically determined, or coded in genes, and therefore in the form of if- then
(c ue- response) r u les. Th e same idea is applied to higher concepts of cogni-
tion, like our concept of space and number, and also in the form of con-
straints on learning that limit what can be learned. Just how codes in linear
strings of DNA can specify such complex structures as object concepts
is not described. Other theories have viewed knowledge as cata logues of
simple associations (as in the concept of dog mentioned above); still others
are based on the rather vague notion of constructed schemas.
I discussed prob lems with these views earlier. For example, it is impos-
sible to fi nd a convincing model of what form such associations and con-
cepts actually take in the mind/brain. Symbols, icons, pictures or other
repre sen ta tions are mentioned, but these are just tentative labels. Accord-
ingly, it is diffi cult to describe individual diff erences in knowledge.
In a dynamical perspective, each domain of knowledge is an emergent
attractor consisting of relational par ameters. Th ese are the experienced co-
variations between variables— the structural par ameters— interdependent
at many diff er ent levels. Th ey furnish predictability in changing conditions.
General knowledge consists of attractor landscapes— coalitions of the
more specifi c attractors for each domain of experience. Th e attractor
landscapes consist of relational par ameters abstracted from each domain
of experience and condensed into neural network confi gurations. Any in-
dividual is knowledgeable in a domain when it is pos si ble to utilize those
confi gurations to generate predictions about current or future situations.
However, as described earlier, local attractors tend to form co ali tions,
giving rise to emergent knowledge not pres ent in the original attrac-
tors. Th ese include hierarchies of superordinate concepts, multimodal
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:54:24 UTChttp://www.ebook3000.com