Page 144 - Deep Learning

Creative Insight: The Redistribution Theory 127

accessible in the context of the subgoal that was in effect when the information
was learned, but becomes generally accessible when that subgoal fades from
memory. These assumptions yield two predictions: First, a person might per-
sist in following an unsuccessful approach to a problem, because factual infor-
mation learned in the course of exploration might not be accessed when it is
most needed. Second, when a person returns to a problem after a rest period,
he is able to access all the factual information learned in the course of prior
explorations and as a consequence constructs a different search space. These
hypotheses are entirely ad hoc and have no independent support in psycho-
logical research. It is nevertheless possible that some version of the differential
forgetting idea will turn out to be true.

Resolving impasses: The role of externalities
The oldest explanation for how an impasse is broken is embedded in the
ancient story of the philosopher Archimedes’ efforts to measure the volume of
66
an elaborately wrought gold wreath without destroying it. After several fruit-
less attempts, he decided to take a break and lowered himself into the classical
version of a jaccuzzi. Seeing the water level rise, he realized that the volume
of a body, no matter how irregular, is equal to the volume of the water that it
displaces when submerged. The perception of the rising water level, an event
seemingly unrelated to the problem of measuring the volume of irregular sol-
ids, triggered the activation of a previously unheeded but crucial fact.
Although the story is unlikely to be true, it is useful as a striking symbol for
the general principle that fortuitous events in the environment might activate
knowledge elements that are crucial for a problem solution, but which were
nevertheless not activated by the problem solver’s perception of the problem.
Archimedes already knew, in some implicit sense, that a submerged body dis-
places a volume of water equal to its own volume, but this piece of knowledge
was not activated by his initial, geometric approach to the problem.
Like many other ideas in the study of creativity, the principle of fortu-
itous reminding has been rediscovered multiple times since Archemedes. For
example, Michael I. Posner wrote in 1973 that a pause provides an opportunity
for “the chance occurrence of an external event which completes the solu-
tion. The external event might retrieve the incomplete solution and provide
the missing association to complete it, leaving the subject aware only of the
solution and not the event.” Colleen Seifert and co-workers have called this
67
68
opportunistic assimilation or opportunistic reminding. Ilan Yaniv and David
E. Meyer added the twist that the initial attempts to solve the problem partially
activate the required knowledge structures in memory, making the problem