Cognitive appraisal
The lineage that, per Seth’s Glossary, dates “back to James (but not Lange)”: emotions depend on cognitive interpretation of physiological changes. Its most famous experimental form is the Schachter–Singer (1962) two-factor theory — emotion as interpreted bodily arousal (physiological change × cognitive appraisal). See stanley-schachter.
Four readings in this cluster
- As the target of a specificity critique — For Friedman, Schachter–Singer’s appraisal model presumes undifferentiated arousal (see autonomic-specificity) and was, physiologically, “exquisitely bad,” yet it eclipsed specificity research during the cognitive revolution.
- As a historical artifact — For Dror, the “cognitive” element was not a discovery about emotion but a device to extend Festinger’s social-comparison theory to emotions; the arousal-labeling framing was retrofitted onto the James–Lange debate. See origins-of-two-factor-theory.
- As a special case to be generalized — For Seth, appraisal theories are the ancestor of interoceptive-inference: “interpretation” of arousal is Bayesian inference on interoceptive causes. Appraisal becomes top-down prediction.
- As the consciousness half of a two-part machine — For LeDoux, the appraisal structure survives, relocated. Feelings occur when consciousness “(1) detects that a survival circuit is active or witnesses the existence of a global organismic state… and (2) appraises and labels this state.” He cites Schachter (1975) directly; the resulting state “could be dimensional in nature (just based on arousal and valence) or could take on specific qualities,” with emotion words adding the specificity. See global-organismic-state.
Automatic vs. reflective appraisal — and a caution
LeDoux also engages the appraisal literature’s own internal distinction: automatic appraisal (significant stimuli eliciting responses without deliberate control) vs. cognitive/reflective appraisal (deliberate, controlled, often conscious determination of stimulus meaning — Arnold, Lazarus, Scherer, Frijda, Smith & Ellsworth).
Survival circuit trigger-evaluation is “clearly an example of automatic appraisal.” But he warns against the easy identification: “one should not be too quick to assume that what psychologists refer to as automatic appraisals in humans is identical to survival circuit processing. The latter probably refers to a narrower set of phenomena than the former.” The categories overlap; they are not the same category. This is a useful check on treating survival-circuits as a neural redescription of appraisal theory — it isn’t one.
A fifth reading: the ordering is wrong (Barrett 2017)
The first source in this wiki to attack appraisal theory rather than absorb, relocate or generalize it. Barrett (2017) splits the tradition in two and keeps only half.
What she rejects — causal appraisal. The objection is structural, not empirical:
Meaning does not trigger action, but results from it. This makes classical appraisal theories highly doubtful, because they assume that a response derives from a stimulus that is evaluated for its meaning.
In a predictive brain, visceromotor and motor predictions run ahead of the input; “for a given event, perception follows (and is dependent on) action, not the other way around.” There is no moment at which an unevaluated stimulus arrives and waits to be appraised — so the stimulus → evaluation → response architecture has no slot to occupy. She names Lazarus (1991), Scherer (2009) and Roseman (2011) as targets, and extends the verdict to “all classical theories of emotion… even those that explain emotion as iterative stimulus → response sequences.”
What she keeps — descriptive appraisal. The concession is real:
Appraisals as descriptions of the world, however, are produced by categorization with concepts.
The taxonomy survives; the mechanism claim does not. Novelty, goal-conduciveness, agency and coping potential may describe perfectly well how a situation is represented — they just don’t make the emotion. On this reading a categorization produces the emotion and the appraisal-describable content in one act, and the appraisal is a readout. This is Clore & Ortony’s (2008) position, which Barrett cites approvingly: she is siding with one wing of appraisal theory against another, not opposing the field.
See are-appraisals-causes-or-descriptions, where this becomes a debate — including the reply available to Scherer, and the structural problem (now partly relieved — see the next section) that the causal appraisal tradition was long present here only as a target, quoted by critics and represented by no primary source.
The tradition itself, first-hand at last (Tong 2015)
Every reading above holds appraisal theory through someone doing something to it — generalizing it, relocating it, historicizing it, refuting it. Tong (2015) is the first source in raw/ that simply works inside the tradition: a mainstream appraisal theorist (in the Smith & Ellsworth / Lazarus / Scherer / Roseman line — see eddie-tong) using the standard recall-and-rate paradigm to show that 13 cognitive appraisals classify 13 positive emotions well above chance (42.4% vs 7.7%), cross-culturally, with 91 of 95 predicted pairwise differences confirmed.
Two things to take from it, and they pull in opposite directions:
The taxonomy is real and rich. Whatever appraisals are — causes or descriptions — the dimensions the tradition named (pleasantness, goal attainment, agency, control, certainty, effort) carry enough information to sort fine-grained positive states apart at high accuracy. Barrett grants appraisals are good descriptions; Tong shows just how discriminating those descriptions are. That is a datum the constructionist reading has to accommodate, not one it can wave away.
But it does not settle causation, and cannot. Tong measures the correspondence between recalled emotions and reported appraisals — which the causal reading and Barrett’s descriptive reading predict equally well. He flags the problem himself (the correspondence “could reflect actual experience or culturalised conceptual knowledge”). So the primary source the debate lacked turns out to strengthen the appraisal tradition’s empirical standing without touching the mechanism dispute. See are-appraisals-causes-or-descriptions.
There is also a quieter finding that cuts against the tradition from inside it: some canonical positive emotions (interest, joy, relief) classify near chance, and contentment and joy cannot be told apart by any appraisal at all. An appraisal theorist reporting that the machinery fails to individuate certain emotions is the same shape of result the constructionists reach (lindquist-2012-brain-basis-of-emotion, theory-of-constructed-emotion) — reached from the opposite side.
The experience-sampling sequel closes the memory loophole and adds a co-occurrence claim. Tong & Jia (2017) rerun the taxonomy with Ecological Momentary Assessment (experience-sampling) — Palm handhelds beeping 395 people through two ordinary days — and recover the same profiles (61/75 predictions) from ratings made in the moment. That answers the one concrete objection to the 2015 result: the appraisal–emotion correspondence is not a memory reconstruction. It does not answer the causal question (EMA is still non-experimental, and the authors say so outright). The new paper’s own contribution is appraisal-overlap: positive emotions co-occur to the degree they share their predicted appraisals, and — after pleasantness is removed — still do, so co-occurrence is not merely a valence effect. The recall study’s contentment≡joy null even softens: understandability now splits joy from contentment and serenity.
A note on the other appraisal tradition: stress and coping (Lazarus)
Everything above is appraisal theory of emotion — Schachter–Singer and its descendants, the arousal-labelling lineage. There is a second, parallel appraisal tradition this page had left implicit: Lazarus’s appraisal theory of stress and coping, now held first-hand through Folkman & Lazarus (1980). It uses the same word for a related but distinct construct.
In the transactional model, appraisal is the evaluation that makes an encounter stressful: primary appraisal (what is at stake — harm/loss, threat, or challenge) and secondary appraisal (what coping resources and options are available). Its downstream object is not which emotion but which coping response — problem-focused or emotion-focused. So the two appraisal traditions share a founder-figure (Lazarus appears in both) and a core idea (evaluation mediates the person-environment relationship), but they answer different questions: one asks what appraisal individuates (emotions), the other what it determines (coping).
The link matters for the debate below. The causal-appraisal camp Barrett attacks is this Lazarus — and the wiki now holds him first-hand, though on coping rather than on the emotion-causation claim, and in an explicitly reciprocal (‘appraisal and coping continuously influence each other’) rather than serial form. See are-appraisals-causes-or-descriptions.
The through-line
All five readings pivot on the same hinge: how much of emotion is bodily signal and how much is interpretation of it. Appraisal theory answers “interpretation dominates (given undifferentiated arousal)”; autonomic-specificity answers “the signal is already differentiated”; interoceptive inference answers “the distinction dissolves — interpretation and signal are two directions of one predictive hierarchy.” LeDoux answers “both, but in series and in different tissue” — a partly-differentiated state (survival-circuit-specific components plus generalized arousal, so not Cannon’s undifferentiated soup) that a separate cortical workspace then labels. That the appraisal skeleton persists in a 2012 comparative-neuroscience paper written to purge feeling words from the field is a reasonable measure of how durable the two-factor structure has been.
Barrett answers something else: the hinge presupposes an ordering that doesn’t exist. There is no unevaluated signal waiting to be interpreted, so the question of how much is signal and how much is interpretation has no well-formed answer. That is a different kind of move from the other four — they distribute the labour between signal and interpretation, she denies the two are sequential stages that labour could be distributed across.
Note how close this looks to Seth’s reading (3) and how different it is. Both dissolve the signal/interpretation boundary into one predictive hierarchy. But Seth’s dissolution is friendly — appraisal theory was right in outline, and “interpretation” just needs renaming to “inference,” so the tradition is vindicated by absorption. Barrett’s is hostile — inference is not appraisal renamed, because appraisal claims to explain how a response is generated and her account denies there was a response-generation step to explain. The same architecture, imported by two theorists, saves the tradition for one and buries it for the other.