Personality Pharmacology

Part IV · Psychopathology on the Same Landscape Draft

Chapter 24

From Temperament to Disorder

Tails, Thresholds, and Hysteresis

A continuum does not have to feel continuous.

Water warms gradually and boils abruptly. Financial leverage can rise slowly before a system fails. A microphone can approach a speaker without incident until feedback suddenly dominates the room. Psychological systems can behave the same way.

This is the missing piece in crude dimensional accounts of mental illness.

Suppose threat sensitivity varies continuously. As it rises, a person may become cautious, then avoidant, then chronically anxious. At some point autonomic arousal, catastrophic interpretation, attention to bodily sensations, and fear of fear can lock into a positive feedback loop. The result is a panic attack, an event that feels qualitatively different from ordinary caution even if the parameters that produced it changed gradually.

The threshold is real even when the substrate is continuous.

The same logic can apply to mania. Sleep gradually shortens. Reward salience rises. Confidence increases. Activity produces more stimulation, which further reduces sleep. Social feedback may initially reward energy and productivity. At some point the state becomes self-sustaining and judgment deteriorates. Restoring one variable may no longer be enough to reverse the whole system.

That persistence is hysteresis.

Interactive figure: Tipping in, tipping back.

Figure 24.1Tipping in, tipping back. A continuous push can produce an abrupt change of state, and the way back requires far less pressure than the way in. Illustrative model, not measured data.

In a system with hysteresis, the path into a state differs from the path out. A drug may trigger psychosis, but removing the drug does not guarantee immediate recovery. Sleep loss may initiate mania, but one good night’s sleep does not necessarily end it. Panic can create avoidance learning that persists after autonomic arousal returns to normal.

The brain remembers its own states.

Learning is one mechanism. Plasticity is another. Receptor regulation, sensitization, circadian disruption, social consequences, self-concept, and environmental changes can all stabilize a condition after the initial trigger has weakened.

This gives personality pharmacology a more realistic model of disorder than “trait plus severity”.

Let a disturbance vector D represent pressures on the system: stress, drug exposure, inflammation, endocrine change, trauma-related learning, sleep loss, neurodegeneration, or other factors. Let W map these disturbances into psychological dimensions T:

ΔT = W D + ε

Different disturbances can produce similar psychological outputs. This is equifinality. Sleep deprivation, stimulants, corticosteroids, or bipolar vulnerability can all contribute to activated states through different pathways.

The same disturbance can also produce different outcomes in different people. That is multifinality. Cannabis exposure may relax one person, provoke panic in another, and precipitate psychosis in a vulnerable third.

Equifinalitysleep deprivationstimulantscorticosteroidsbipolarvulnerabilityAn activated statesimilar output, differentpathsMultifinalityCannabis exposurethe same disturbancerelaxation in onepersonpanic in anotherpsychosis in avulnerable third
Figure 24.2Many to one, one to many. Different disturbances can converge on the same state, and the same disturbance can lead different people to different places. Baseline changes the map.

Baseline changes the map.

Interaction makes the problem even less linear. A moderately elevated reward system may be harmless with strong inhibitory control. Moderate sleep loss may be tolerated when stress is low. Moderate suspiciousness may be contained when social support is strong. Combine several moderate deviations and the system can cross a threshold that none would produce alone.

moderate reward drive alonemoderate sleep loss aloneboth togetherunstable regiondeviation in one system →deviation in another →
Figure 24.3Configuration, not single factors. Moderate deviations that are each harmless can combine to cross a threshold none would cross alone. Illustrative model, not measured data.

This is why single risk factors often predict poorly. The relevant variable is configuration.

The model also clarifies why diagnosis can become necessary even in a dimensional theory. Once a state crosses a threshold and creates a new attractor, treatment may need to be qualitatively different. Mild insomnia and manic insomnia are not merely the same symptom at different intensities because the latter sits inside a self-reinforcing syndrome. Ordinary skepticism and delusional conviction differ not only in degree but in how beliefs update under counterevidence.

Emergent structure can justify categorical clinical language without requiring categorical causes.

This perspective has practical consequences for intervention timing. If thresholds and hysteresis matter, preventing a transition can be easier than reversing it. Protecting sleep in someone with bipolar vulnerability may have disproportionate value. Interrupting escalating stimulant use before sensitization and sleep collapse may prevent a much larger shift. Treating early panic may reduce the formation of an avoidance network.

A small perturbation at the right time can matter more than a large one later.

It also complicates the idea of “returning to baseline”. After a severe state, the old baseline may no longer exist in exactly the same form. The person has learned something, lost something, changed relationships, altered medication exposure, and perhaps undergone biological adaptation. Recovery is often the construction of a new stable region rather than a simple rewind.

This matters for deliberate personality engineering as well. Desired states can leave beneficial hysteresis. Temporary social confidence may produce durable learning. A psychedelic experience may alter values or self-narrative long after the drug is gone. Repeated disciplined focus can build habits.

The same principle that makes pathology persist can make change persist.

The ethical question is whether the lasting trace was chosen, predictable, and still endorsed.

Thresholds therefore connect nearly every part of the theory. They explain why continuous traits can coexist with categorical clinical events, why baseline vulnerability matters, why intervention timing matters, and why temporary state manipulation can have enduring consequences.

The next chapter takes one particularly important threshold seriously: the point at which a mind’s model of reality loses sufficient calibration that perception, belief, mood, or salience becomes dangerously detached from the world it is trying to navigate.