A CREATIVE COGNITION 225
attractors (discrete fi les, sequences, or traces). Again this is a resort to
inappropriate meta phors. It should be clear by now that neither the cog-
nitive system nor the brain is fashioned to handle information in that
form. Rather they evolved to handle abstract statistical structures, the
grammars, from which most likely specifi c instances can be generated
from constantly novel inputs. Memories, as dynamic entities, refl ecting
dynamic experience, are enshrined in such grammars (or attractors).
Th is explains several prominent fi ndings in memory research. Se-
mantically related memories will tend to have related pa ram e ter struc-
tures. Accordingly, a person being asked to recall items in a category
l i ke “a n i ma ls” will tend to do so in groups of related species: a ll the birds,
fi rst, say; then all the fi sh; and so on, rather than by recency of experi-
ence. It also explains why memories become distorted over time, but in
par tic u lar ways. For example, Frederick Bartlett in the 1930s described
studies in which participants were asked to reproduce stories and pic-
tures experienced earlier. Th e participants typically elaborated or distorted
the originals in ways that imposed their personal and social experiences
over the original story line (another example, in fact, of the workings of
ideology).
Th is makes sense in a dynamical perspective, because events tend
to become assimilated into more general pa ram e ter structures—
attractors— with experience. Over time, these also come to include per-
sonal and cultural ideals, as in the developing child. In other words, with
continuing learning in a domain, memory attractors become revised
and reor ga nized in relation to one another. Th is is a well- established
fi nding in research and is refl ected in the nature of knowledge. (In fact,
in recent years, most cognitive models have been concerned with working
memory, which I scrutinize below).
KnowledgeIn the cognitive sense, knowledge is obviously the product of learning
(indeed, the word “cognition” stems from the ancient Greek for “to
know”). For the past few de cades, at least, it has been assumed to reside
in patterns of neural connections. As David Elman and colleagues put it,
“Knowledge ultimately refers to a specifi c pattern of synaptic connections
in the brain.”^22
This content downloaded from 139.184.14.159 on Tue, 17 Oct 2017 13:54:24 UTC