nucleation

THEORY-901: As Mancinelli and Bach's systematic review of avoidance learning in healthy humans reports it, rating both A, an avoidance response left available during extinction persists, in some participants for hundreds of unpunished trials, and keeps expectancy of the aversive outcome higher than in people denied it (nine studies); and less effective or costlier responses extinguish faster (six studies, one exception)

3 min read

Properties7
StatusActive
Version2
Tagslearning-and-conditioning, psychopathology-and-treatment
SourceLIT-1477 — A systematic review of human avoidance learning: Cognition, computation, and methods
Bears onQUESTION-327 — Does available avoidance leave threat expectancy uncorrected in proportion to how few unavoided disconfirming trials it allows, in people with anxiety disorders as in healthy controls, and does performing and then losing an unsignalled avoidance response raise threat expectancy for a stimulus never paired with harm, scaling with how much the person avoided?
SupportsCLAIM-446 — Some persistent anxiety may be governance-maintained rather than belief-maintained: anticipated threat grants defensive processes authority over action, attention and evidence acquisition; under the avoidance that authority selects, the hypotheses that the situation is dangerous and that it is safe predict the same observation, so repeated safety need not correct the threat estimate; relief reinforces the defense without being evidence of danger; and the regime can outlast a corrected belief, its pathology, where present, being defensive authority insensitive to corrective evidence at avoidable cost
Cited as “extends” byTHEORY-1189 — In laboratory avoidance learning in healthy humans, an avoidance response that remains available during extinction persists, in some participants for hundreds of unpunished trials, and keeps expectancy of the aversive outcome higher than in people denied it; the more effective and the cheaper the response, the slower it extinguishes

Graph View

Source

Mancinelli & Bach, LIT-1477, read in full in NOTE-1134 from the PsyArXiv preprint, version 1 (11 June 2025): pp. 17–20 and 28, and the table of phenomena (Fig. 4, p. 53). The version of record is closed.

What was actually shown

The review pools no data; it classifies findings.

  • In two older studies with the response available during extinction, “42% (Malloy & Levis, 1988) and 47% (Williams & Levis, 1991) of participants did not extinguish a unary and costless AR by the 500th would-be-unpunished trial” (p. 18).
  • Two studies comparing groups with and without the option to avoid during extinction found that those allowed to avoid “continue to do so and retain higher US expectancies in a follow-up test where avoidance is precluded” (p. 18).
  • Less effective or costlier responses extinguish faster (Xia et al. 2017; Zuj et al. 2020; Rattel et al. 2017, 2020; Dymond et al. 2023, 2025), with one exception (Meulders et al. 2016) (p. 19).
  • The authors’ explanation: “more effective ARs are harder to extinguish when the response is not prevented, putatively because effective ARs provide fewer opportunities to detect changes in contingency (i.e., the absence of punishment when the AR is not performed), causing slower learning” (p. 28).

What could have come out otherwise: people could have stopped responding once the shocks stopped, or their threat expectancy could have fallen as fast with avoidance available as without.

What this does not say

  • Anything about anxiety disorders. Only healthy participants were included, mostly students, with shock or loud noise.
  • That the mechanism is unidentifiable evidence. That is the authors’ “putatively”. The review offers no test that varies the number of unavoided trials with effectiveness held fixed.
  • That avoidance strengthens threat belief. Here it preserves a belief that would otherwise extinguish. Raising belief in a stimulus never paired with harm is a different phenomenon, THEORY-920.
  • That the effect is the same with real-world avoidance. The responses are mostly signalled, instructed button presses, which the authors rate as their main risk of bias.

Connections

  • CLAIM-446. The laboratory form of the claim’s “avoidance destroys identifiability”: under available avoidance, safe trials do not correct threat expectancy. It supports that step, in healthy people.
  • THEORY-910. Brown et al. propose the same loop with uncertainty, not threat expectancy, as what avoidance preserves. The review does not separate the two.
  • THEORY-285. A model-level instance of the same self-sealing; this is the human laboratory instance.

Backlinks