Sense of should

Introduced in Theriault, Young & Barrett (2021). A felt obligation to conform to others’ expectations — the subjective experience of social pressure — held to be biologically motivated and separable from two neighbouring motives. See jordan-theriault, lisa-feldman-barrett.

The three-way distinction (Figure 1)

The paper’s opening move, borrowed from Adam Smith’s two motives, is to carve normative influence in three:

motive	do you copy others?	motivated by others’ expectations?	feel obligated?
reputation-seeking	not necessarily	yes	no
sense of should (social obligation)	not necessarily	yes	yes
moral obligation	no	no	yes

Reputation-seeking wants the appearance of compliance (praise, avoided sanction). Moral obligation is driven by an internalized value and can oppose others’ expectations (“tell the truth even if it is painful”). The sense of should is the middle case: you feel you should behave as others expect, without aiming at reward and without an internal value doing the work. It is far more common than moral obligation — it covers suppressing a cough in a quiet hall, waiting to use the restroom during a lecture, passing the salt when asked.

Why it exists: predictability as a resource

The sense of should is derived, not posited. Because prediction error is metabolically costly (metabolic-cost-of-prediction-error) and the brain avoids such costs prospectively (allostasis), an unpredictable environment is a cost the organism will act to reduce. Other people are part of the environment, and conforming to their expectations makes them less likely to change their behaviour in ways you cannot predict. So conformity buys a predictable — and cheaper — social world. The felt version of that motivation is the sense of should. It is “individually adaptive” in an ultimate, metabolic sense, even though it is experienced as the opposite of self-interest.

The paper coins two terms for the influence this produces:

• Stabilizing influence — control exercised over others by conforming to them. It does not make others do anything specific; it makes them less likely to react unpredictably.
• Directing influence — causing a specific behaviour in others (coercion, teaching, or communication). The two combine: “the most straightforward way to control someone else’s behaviour may be just to make them aware of what you want,” because a listener with a sense of should will then conform of their own accord (the warden’s observed prisoner; Mameli’s value-transmitting father).

What it feels like: an interoceptive punishment, self-inflicted

The subjective side is pure core-affect/constructed-emotion machinery. Violating others’ expectations raises your prediction error, which generates arousal (an ANS response), which — interpreted in context — becomes aversive. Crucially the aversive outcome is not administered by anyone: no second party pays a cost to punish you; others simply react to being surprised, and you make aversive meaning of your own interoceptive sensations. Theriault et al.’s phrase: “a punishment that the brain literally inflicts on itself.” Kohlberg’s two adolescents feeling the same stomach anxiety and reading it as “being chicken” versus “the warning of my conscience” is the categorization step made concrete.

Because the brain regulates allostatically, the mature form is anticipatory: a sense of should is experienced as an anticipatory aversion to violating others’ expectations, felt before the violation. This gives it a clinical seam — social anxiety may be an over-estimation of how disruptive one’s behaviour will be, or an over-weighting of the aversive interoceptive experience (the paper cites Khalsa et al. 2018). See anxiety-sensitivity.

Relation to guilt

Not synonymous with guilt. Under constructed emotion, “guilt” names a category of heterogeneous instances that vary by situation; what unites them is the sense of should — the motivation to conform to others’ expectations. Guilt is therefore one instance under the sense of should, which is the more general motive and shows up in non-guilt instances too (a servile lackey, a parent caving to a child). The paper shows its formal model is equivalent to the guilt-aversion model of Chang et al. (2011): both make the agent minimize the discrepancy between behaviour and inferred expectation, reached “from drastically different foundations.”

What to hold loosely

The construct’s reach is very wide — the parent paper uses it to unify conformity, altruism, status-quo bias, communication, culture, and belief maintenance — with a flexible core and no clearly stated disconfirmer, the same structural worry the wiki records for theory-of-constructed-emotion and degeneracy. And it inherits the load-bearing assumption of metabolic-cost-of-prediction-error: that prediction error is costly enough to motivate, a link the source states by proportionality rather than measurement.