Source
Barandiaran (2025), LIT-1026, read in NOTE-730: §3 (the dual-order account and its problems), §4.1–4.4 (regions, the normative field, the model, the three kinds of dysfunction), §4.5 (attribution and compensation), §5.1–5.2 (norms without orders, illness as a shrunken viability space), note 23 (nested autonomies).
What was actually shown
The paper argues from definitions, counterexamples and a one-variable model of a metabolite fed by food. It measures nothing.
Against the dual-order account. As quoted from Moreno & Mossio (2015, LIT-536), a trait malfunctions when it fits first-order norms but fails the second-order regulation that shifts the system to another regime. That definition cannot count a stiffening coronary artery at rest, since no regime switch is involved. It cannot count an autonomic nervous system that fails to trigger the stress response, since only the regulated trait can malfunction. And it cannot tell which of the two failed. Barandiaran traces this to two assumptions: that first-order norms are all-or-nothing, and that regulation happens only as switching between regimes. Physiological regulation is mostly continuous modulation of rates.
The replacement. Essential variables span a viability space. Holding a parameter fixed, the space divides into a viable region (the system stays alive), a precarious region (“(still) alive but will inevitably die if no parametric change occurs”) and a terminal region (beyond recovery). The normative field is “the minimum constant parametric change, for each point of the precarious region, necessary to avoid the irreversible disintegration of the system”. A trait functions when “its effects on the viability space correlate positively with the normative field”. The three failures are subfunction (right direction, too slow), malfunction (wrong or no direction) and nonfunction (no operation or no effect where operation is required). Regulatory traits fall under the same standard, and a regime switch becomes an added dimension of the viability space. One can still be “a dualist on organizational orders … but a monist regarding norms”.
Attribution. Identifying the unmet norm is not locating the failing part. The paper sketches a recursive search over each trait’s feasible range, and argues that compensation does not hide a dysfunction, because “compensations have a cost on the viability of the system”.
What this does not say
- It does not compute a field for any real system. The author says doing so “is extraordinarily complex and may not be feasible”.
- It does not derive the essential variables. They are given, as K is given in viability theory (THEORY-134). Which states count as death is an input.
- It is not Aubin’s viability theory. The paper works with Ashby’s essential variables and bifurcation diagrams. Its “viable region” is where the system survives with the parameter held constant, not a viable set. Read in Aubin’s terms, the terminal region is outside the viability kernel of the controlled system, and the precarious region is inside it but not invariant at the current parameter. That correspondence is the record’s.
- It does not say what ought to persist. The norms are for the organism, relative to its persistence. CLAIM-075 stands.
- It does not refute the organizational account of function. It is an organizational account. It disputes only one definition of malfunction, taken from the 2015 book. The 2009 paper (LIT-570) is cited as part of the standard approach; its dysfunction is a produced trait that fails to contribute, and it is not the dual-order definition.
Connections
- QUESTION-087 (answers, in part). The question asks whether a self-reproducing authority structure can be pathological by a standard the organization itself supplies. This account supplies one: the normative field of the whole organization. Since regulatory traits can malfunction, an authority structure, a regulation of other processes, can be judged by whether its operations correlate with that field. Two further results widen the net. Illness can be a shrinking of the viability space rather than a failure to contribute (§5.2), and protective lock-in’s fifth step (TERM-020), reduced capacity to leave the regime, is harm of that kind on the record’s mapping. And compensation shows up as cost. The answer is partial. An authority that perpetuates itself while leaving the organization as viable as before is not dysfunctional on this account, so it does not separate merited trust from entrenched advantage (CLAIM-277) where both keep the organization alive equally well. And note 23 allows that an entrenched authority may become an autonomous level with its own norms in conflict with the organization’s, without saying which prevails.
- QUESTION-086 (no relation declared). The attribution procedure, varying one trait within its feasible range, is an intervention design. It locates where a norm fails. It does not test whether the norm is the system’s own, which is what that question needs.
- CLAIM-079 and THEORY-134. A system in the terminal region is alive with no trajectory left that keeps it so; that is CLAIM-079’s gap between occupying a state and keeping access to viable futures, in the record’s reading.
- Di Paolo (2005, LIT-1017). The precarious region is where Di Paolo’s adaptivity acts. This account turns his graded inner norm into a field that a part can be judged against.
Note of 2026-10-28: what holds it back
Kept Proposed by the sixth status review. No normative field has been computed for a model richer than the paper’s
one-variable toy (several traits, one regulatory, a regime switch), so the
classification test in promote_when has not been run.
CLAIM-368 now cites this reading as corroboration, not as a ground.