Page 342 - Deep Learning

The Formation of Belief 325

The issue is muddied by a confusion between basic cognitive processes
and acquired cognitive strategies. scientists are professional theorists. They
gather data, solve research problems and evaluate theory-data relationships as
part of the work they are trained and paid to do. scientific theory formation is
deliberate and supported by cognitive technologies like special-purpose nota-
tions and mathematical software; it is also public and guided by codified dis-
ciplinary norms and standards. The associated practices are acquired through
a lengthy, multi-year training process called “graduate school.” Consequently,
scientists are aware of how they form theories – there is no end to the stream
of books about research methodology – and the tasks of interpreting data and
evaluating hypotheses are at the focus of their conscious attention. in con-
trast, most laypeople could not give a clear account of how they formed any
one belief, and even less how they form beliefs in general. Beliefs are formed
implicitly, on the run and as a side effect of activities that are undertaken in
pursuit of some more pragmatic purpose than to formulate true descriptions
of the world.
Given these differences, it is not obvious that the professional practices
of scientists hold any clues to the cognitive processes of belief formation and
belief revision. Consider an analogy to the voting practices adopted in most
Western nations for deciding social and political issues. voting is clearly an
invented practice and we would not expect it to provide clues to the nature
of the basic cognitive processes by which people decide what stand to take on
social and political issues. nobody would argue that because political deci-
sions are made by many individuals casting votes, the individual human mind
is therefore likely to consist of a population of, say, “evaluation modules,” each
of which casts a subconscious “vote” for one political stance over another. The
distinction between a description of societal voting practices and a psycho-
logical theory of individual decision making is obvious. The voting theory of
individual decision making could conceivably turn out to be true, but its prob-
ability of being true is not increased by its structural similarity to the social
practice of voting. But if so, why should we believe that scientific practices
provide clues to the psychology of belief formation and belief revision? if the
fact that political decisions are made by voting is not sufficient grounds for
inferring that voters’ minds consist of populations of voting modules, then the
fact that scientists evaluate theories by comparing them to evidence is not suf-
ficient to conclude that scientists’ minds contain processes for relating beliefs
to evidence.
The distinction between the basic cognitive processes of individuals and
the acquired methods and practices of scientists is easily blurred because