Reinterpreting Emotional Responses in PTSD as Adaptive Survival Strategies — Epoche C2
What the manual actually says Hypervigilance, exaggerated startle and intrusive re-experiencing sit among the twenty symptoms that the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (2013) distributes across the four symptom clusters of post-traumatic stress disorder, and considered as machinery they are indistinguishable from the machinery that keeps an animal alive under threat. This essay takes up what follows from that — specifically, whether it licenses describing these responses as adaptive survival strategies dislocated from their original context, which is the thesis the shorter version of this essay advanced and which, in a form made precise below, survives. It survives, but not against the opponent originally named. The earlier version said that the prevailing discourse, "deeply entrenched in diagnostic criteria like the DSM-5", frames these responses as inherently pathological and overlooks their adaptive significance. That is fair as a description of a great deal of clinical rhetoric, and unfair to the manual, which is worth reading closely at this point because it concedes more than the charge allows. Criterion A of the PTSD entry requires exposure to actual or threatened death, serious injury or sexual violence: PTSD is one of the very few diagnoses in the manual defined by its cause, and in DSM-5 it was moved out of the anxiety disorders into a new chapter of trauma- and stressor-related disorders precisely to mark this. Criterion G requires clinically significant distress or impairment in functioning. Nowhere does the manual assert that the symptoms are malfunctions of a mechanism; it asserts that, taken together and persisting beyond a month (Criterion F), they impair. Those are different claims, and the difference is the space in which the present argument works. The counts are worth stating because the fourth edition's revision shows the direction of travel. DSM-IV listed seventeen symptoms in three clusters: five intrusion items, seven avoidance-and-numbing items, five hyperarousal items. DSM-5 split the middle cluster in two and added three symptoms — persistent distorted blame of self or others, persistent negative emotional states, and reckless or self-destructive behaviour — giving twenty items in four clusters: five intrusion, two avoidance, seven negative alterations in cognition and mood, six alterations in arousal and reactivity. The additions are all items of altered appraisal and altered conduct , not of heightened arousal. The manual's own drift is away from a picture in which PTSD is an excess of alarm. One element of the earlier version has been removed rather than expanded. It carried three sayings attributed to Akan, Zulu and Bambara tradition. The first — that a tree falling in the forest makes a sound whether or not anyone hears it — is not an African proverb but Berkeley's philosophical example in eighteenth-century dress; the other two could not be traced to any collection. The formula "I am because we are" is John Mbiti's, from African Religions and Philosophy (1969), rather than a Xhosa proverb. Decoration that cannot be sourced is worse than no decoration in an essay whose whole burden is that a claim must carry its grounds, so the sayings are gone. The substantive point they gestured at, that healing may be constituted communally rather than individually, is a real research question and is not argued here, because nothing in what follows establishes it. Why the responses exist, and why false alarms are the sign of good design The first thing the thesis needs is a criterion under which "adaptive" is a testable claim rather than a consolation. Randolph Nesse's smoke-detector principle (2001) supplies one, and it is worth deriving rather than invoking, because the derivation is what shows why the false-alarm rate of a well-built defensive system should be very high. Treat the expression of a defensive response as a decision under uncertainty. Let $c$ be the fitness cost of expressing the response when no danger is in fact present — energy, forgone foraging, forgone mating, the risk incurred by conspicuous flight. Let $h$ be the fitness harm suffered if danger is present and the response is withheld. Let $p$ be the probability of danger given the cue the organism has. Expressing the response costs $c$ whether or not the danger is real; withholding it has expected cost $p\,h$. Expression pays whenever $p\,h > c$, so the optimal threshold on the posterior probability is $$p^{*} = \frac{c}{h}.$$ The exponent-free simplicity of this is the point. Put numbers to it in a common currency: a gazelle's sprint costs on the order of $c = 100$ units, and being caught costs $h = 100{,}000$ units, since it forfeits all future reproduction. Then $p^{*} = 100/100{,}000 = 10^{-3}$: it pays to run whenever the chance of a predator exceeds one in a thousand. At exactly that threshold, $999$ of every $1{,}000$ sprints are responses to nothing at all. The false-alarm rate is not a defect of the system; under this model it is the signature of correct calibration, and it rises as the ratio $c/h$ falls. A system that ran only when it was more likely than not that a predator was present would be catastrophically miscalibrated. Applied to trauma, this establishes something and fails to establish something else, and the two must be kept apart. It establishes that a threshold low enough to produce frequent responses to non-threatening cues is what a well-designed system looks like, so that the sheer frequency of alarm in PTSD is no evidence of malfunction. It does not establish that a response sustained for months or years is adaptive, and it is important to see that the model actually cuts the other way here. The derivation assumed that $c$ is paid per episode and is small. If arousal becomes chronic — sleep disruption, sustained sympathetic activation, exhausted attention — then $c$ is no longer small, the ratio $c/h$ rises, and the same calculati