In the book Breakdown of Will author George Ainslie presents a theory of decision-making where relative preferences for possible alternative outcomes vary over time, unlike the usual “rational” model where relative preferences remain consistent. His model is based on hyperbolic discounting where the value of a future pay-off is discounted more heavily than normal exponential discounting in the period near the pay-off time and less heavily in the long tail well before the pay-off.
Here are some excerpts from the book, organized by the parts and chapters of the book.
Part 1. Breakdowns of Will: The Puzzle of Akrasia
We make irrational choices in the moment that go against our long term interests. Why? Modern psychology avoids talking about willpower. Aristotle talked about akrasia, “weakness of will”. The self is seen to use reason to counter the demands of passion.
The human bent for defeating our own plans has puzzled writers since antiquity. From Plato’s idea that the better part of the self — reason — could be overwhelmed by passion, there evolved the concept of a faculty, will, that lent reason the kind of force that could confront passion and defeat it. The construct of the will and its power became unfashionable in twentieth-century science, but the puzzle of self-defeating behavior — what Aristotle called akrasia — and its sometime control has not been solved. With the help of new experimental findings and conceptual tools from several different disciplines, it will be possible to form a hypothesis about the nature of will that does not violate the conventions of science as we know it.
Ch 2: The Dichotomy at the Root of Decision Science: Do We Make Choices By Desires or By Judgements?
The puzzle of self-defeating behavior has provoked two kinds of explanation, neither of which has been adequate. Cognitive theories have stayed close to introspective experiences of the will and its failure but have shrunk from systematic causal hypotheses, perhaps because they make a person seem too mechanical. Utility-based theories have accounted well for many properties of choice, but seem to predict neither self-defeating behavior nor any faculty to prevent it. Hypotheses to reconcile self-defeating behavior with maximization of utility have cited naiveté, short time horizons, conditioned cravings, and the physiological nature of reward, but all of these explanations have failed on experimental or logical grounds.