Page 319 - Deep Learning

302 Conversion

interpretation of a discourse is almost guaranteed to be consistent with those
beliefs. Because new information is interpreted in terms of a person’s cur-
rent concepts and beliefs, such information has little power to change those
concepts and beliefs. There is no obvious way to circumvent this assimilation
paradox. As we cannot see without our eyeballs, so we cannot understand a
discourse without our prior concepts.

Center-Periphery structure

The conflict between prior beliefs and a current discourse is sometimes
explicit, as in an exchange between two persons with different political views,
two academics advocating competing theories or two spouses arguing about
parenting or family finance. every disagreement is potentially an opportunity
for productive change: There is a chance, however slight, that the other person
is right. A detected conflict is likely to trigger other assimilation processes than
unconscious interpretive choices.
Clues to how those processes operate emerged in the work of the French
philosopher and physicist Pierre M. M. Duhem in the late 19th and early 20th
centuries. in his 1914 book The Aim and Structure of Physical Theory, Duhem
pointed out that the derivation of a testable prediction or a consequence from
a hypothesis in physics always assumes the truth of multiple other hypotheses
and theories. An experiment might, for example, require multiple measure-
ments, and the measurements obtained with a particular instrument cannot
be interpreted without a theory for how that instrument works. if the pre-
diction does not fit the data from the experiment, something is wrong, but
it is not clear what: “the physicist can never subject an isolated hypothesis to
experimental test, but only a whole group of hypotheses; when the experiment
is in disagreement with his predictions, what he learns is that at least one of
the hypotheses constituting this group is unacceptable and ought to be mod-
ified; but the experiment does not designate which one should be changed.”
22
A judgment of accuracy has to compare an entire group of hypotheses to an
entire body of experiments.
Duhem’s insight was generalized by the American philosopher Willard van
orman Quine and followers to encompass informal as well as scientific belief
formation. Like other theorists considered in this chapter, Quine regarded
23
“the characteristic occasion for questioning beliefs” as a situation in which ”a
new belief, up for adoption, conflicts somehow with the present body of beliefs
24
as a body.” His basic observation is the same as Duhem’s: “We think loosely
of a hypothesis as implying predictions when, strictly speaking the implying is