ARE STATES OF NATURE PRESCRIPTIVE OR DESCRIPTIVE?

ARE STATES OF NATURE PRESCRIPTIVE OR DESCRIPTIVE?
The second section in this chapter argued that the scientific suitability
of a particular theoretical construct—in that case it was a particular
classification scheme for food—could depend on whether the construct was meant to be applied to the independent variables of a theory
or its dependent variables. This section addresses a similar issue, namely that certain criteria for suitable specification of the states of nature can depend upon whether the states are to be used for positive (that is, descriptive) versus normative (that is, prescriptive) purposes.
30 Machina
Since its inception, expected utility theory has always straddled the boundary between being a descriptive and a prescriptive model of decision
making under uncertainty. Even its original presentation by Bernoulli
(1738) as a “solution” to the St. Petersburg Paradox can be alternatively interpreted as either a description of why people don’t assign an infinite certainty equivalent to the Petersburg Game, or a prescription
for why an individual shouldn’t assign an infinite certainty equivalent to the game. Two centuries later, proponents of objective expected utility theory defended it against the Allais Paradox by shifting
their emphasis from the alleged descriptive power of the theory to its alleged normative power.
The same points can be made about the particular component of subjective expected utility theory that forms the central topic of this chapter—namely the notion of states of nature. It is one thing to assert that the states of nature approach offers a useful normative framework for decision making. It is quite a different thing to assert that, for the most part, this is how individuals actually do go about making decisions
in the absence of probabilistic information. We shall consider each of these two domains in turn—in each case, with the goal of identifying
the proper scientific criteria for states.