Pressure, Hysteresis, and Experience: A Friction-Theoretic Framework for Clinical Intervention. The paper proposes a substrate-computational mechanism specification for psychiatric clinical phenomena and intervention, applying the RACE-architecture and hysteretic-trace mechanism of Friction Theory (FT; Pødenphant Lund 2026b, Paper 1) to the human-clinical content of Behavioural Friction Theory (BFT; Pødenphant Lund 2026a, Paper 0). It is a framework-development article, not a clinical-decision-support tool: the contribution is at the substrate-mechanism level, with empirical adjudication invited at the level of falsifiable predictions on existing cohort data. Framework contributions. (1) The outcome-cluster-versus-route-identity diagnostic-imprecision pattern, which explains why dimensional alternatives (RDoC, HiTOP) do not resolve the clinical heterogeneity that route-distribution under a shared outcome explains. (2) The additive-only intervention asymmetry, which derives a six-row intervention-mechanism taxonomy from hysteresis-asymmetry rather than from tradition-axes. (3) An integrated meta-mechanism unifying the intervention mechanism-axis, the reactance-bypass framing-condition, and pharmacology-as-forward-modelling-modulator under a single signal-budget allocation principle. The framework develops trust as a structurally-defined meta-variable, distribution-shape as a fourth diagnostic dimension, and the prevention-over-cure asymmetry that follows from hysteresis once traces have accumulated. It is developed in depth against treatment-resistant depression as a worked exemplar and illustrated in breadth across addiction, ADHD, autism-spectrum, PTSD/OCD, specific phobias, and aging, to evidence its central claim that one machinery, re-parameterised, produces phenotypes the diagnostic catalogue treats as unrelated. Predictions. Five framework-pivotal falsifiable predictions (whose disconfirmation forces framework-revision), twenty-three supporting predictions, and five negative-prediction R-conditions, several testable on existing public-access cohort data without new collection. A cross-substrate parallel with reinforcement-learning-from-human-feedback language models illustrates the framework's substrate-universal scope as structural isomorphism, not mechanism identity. Author position and invitation. The author is an independent researcher without clinical credentials. The framework is positioned as hypothesis-and-invitation: the paper specifies what should be tested; clinical translation requires domain-expert collaboration. We make no patient-level recommendations and prescribe no doses. Target venue: Frontiers in Psychiatry — Hypothesis Kirby et al. 2017); (§4.5e) self-directed downward escape (non-suicidal self-injury / escape offloading) as a false-exit; (§5.0.2b) expectation as a meta-variable, giving placebo and nocebo a substrate reading as expectation-priors that gate effort and commit-convergence, with a biological bridge (opioid / naloxone-reversible; allostatic load) and falsifiable predictions; (§5.0.6) locus-of-leverage — agency sits at race-initiation and between races, not within a running race, unifying the framing-versus-retrieval delivery rule with delay-discounting; and (§2.5) a habit = pressure-immunity corollary (a consolidated route is won before competition and so sits outside the momentary pressure knob's reach). Builds on the v3 revision.
Tomas Pødenphant Lund (Tue,) studied this question.