What looks like change is a change of circumstance, and it ends when the circumstance does

The argument

A second argument concedes that people are seen to behave differently and denies that this is change in the person. Proponents observe that nearly every impressive transformation arrives with something else attached: a new city, a new partner, a crisis, a period of supervision, a supportive group, or a role that made the old behaviour impossible. While the new conditions hold, the new behaviour holds. When they are withdrawn, and this is the part the hopeful account passes over, the old behaviour returns, often surprising everyone including the person themselves. On this reading the apparent change was never located in the person at all. It was located in the arrangement, and the person was behaving as that arrangement required, exactly as they had previously behaved as the old one required. Proponents regard this as the honest explanation of why so many resolutions, treatments and fresh starts hold for a season and then give way, and of why those closest to someone learn to wait before believing it, having waited before. The argument also accounts for the sincerity, which is otherwise puzzling. The person meant it. They were not pretending during the good period, and they are not pretending when they are dismayed by the relapse. That combination is what a disposition held in check by circumstance looks like from the inside. Proponents conclude that the only durable way to get different behaviour from someone is to keep the circumstances that produce it permanently in place, which is a strategy for managing people rather than evidence that people change.

Premises

Counter-arguments

Opponents answer that this proves too much. By the same reasoning nobody ever learns anything, since every skill is maintained by the conditions in which it is used, and learning to drive would be a property of roads. They add that the premise overstates the evidence: plenty of changes outlast the circumstances that began them, and the relapses the argument leans on are the cases that get noticed, while the person who quietly stopped behaving badly and never resumed gives nobody anything to remark upon.

Rejecting the premises

[Rejecting P1] Almost always rests on the cases that come to notice. Changes that simply hold produce no story and no relapse to explain, so the evidence the premise draws on is selected by the outcome it is used to support. [Rejecting P2] Characteristically overstates it. Relapse is common in the hardest changes and uncommon in ordinary ones, and treating the hardest as representative settles the question in advance. [Rejecting P3] By this standard nothing anyone learns belongs to them, since every capacity is sustained by the conditions in which it is exercised. [Rejecting C] Managed behaviour and changed behaviour are not alternatives. Someone who arranges their life so that they reliably behave differently has changed what they do, which is what the question asks.