There are few better places to look for a formal approach to preference change than Sven Ove Hansson’s work (see Hansson 1995, Grüne-Yanoff & Hansson 2009). In his texts one will find roughly the following types of preference change:
- The preferences can be acquired or lost
- More alternatives can be added to those considered or alternatives can be removed
The case were indifference is replaced by preferring one or the other alternative, or vice versa, and the case where preference change direction can be considered combination of acquiring and losing a preference.
Hansson (1995) calls the four types of preference change listed Revision, Contraction, Addition, and Subtraction. But are these all the options?
Under one assumption, we might introduce at least two more types of preference change. This assumption is that motivational states can have indeterminate content. By “indeterminate conten”t, I mean content that is not adequately expressed with any proposition. If one accepts that existence of such content, then one might expect that a preference could get more or less determinate.
Hansson did not simply overlook this option. It plainly does not make sense in typical accounts of decision theory. It certainly does not make sense according to Jeffrey’s (1990) approach, where a preference is roughly a relation to a proposition. After all, propositions are the standard of determinacy as I have introduced it here. As far as I can see, the indetermincay does not fit into the formal framework of Hansson 1995 either.
The new types of preference change would require more than preferences just becoming more specific. I am not discussing the case where I at first prefer to read a book over watching a movie and later develop the more specific preference for reading a history book over watching an adventure movie. That kind of precisification can be described using Hansson’s apparatus because the content is still determinate. What are introduced are new and more specific alternatives, but they are determinate. What the new type of preference change would require is content that cannot be described in any determinate form, not just one that is less specific - and I have not argued for such a kind of indeterminate content. It might also be possible that preferences can be replaced in the same way by other preferences with more or less determinate content, so that Hansson’s types would suffice. However, as long as we have no formal model of such preferences, it is not clear that we can do that. All I am suggesting here is a conditional conclusion: If it existed, we might have to extend our list of types of preference change.
 It might be important to distinguish the formal representation from the psychological account/philosophy of mind. It could turn out that our best psychological account/philosophy of mind ascribes mental states with indeterminate content, while it remains easier to model this indeterminacy using a standard framework the four types of preference change distinguished by Hansson.
Hansson, S. O. (1995). Changes in Preference. Theory and Decision, 38(1), 1–28.
Grüne-Yanoff, T., & Hansson, S. O. (Eds.). (2009). Preference Change: Approaches from Philosophy, Economics and Psychology. Dordrecht; London: Springer.
Jeffrey, R. C. (1990). The Logic of Decision (2nd edition). Chichago: University Of Chicago Press.