Families of Emotion Theory

Emotion science can look more cleanly divided into camps than it really is.

A reader may encounter basic-emotion theory, bodily theories, appraisal theory, dimensional models, psychological construction, predictive processing and active inference as though these were separate contenders standing in a line, each offering a different answer to the same question.

That is useful for orientation, but it hides something important.

As established in What Must a Theory of Emotion Explain?, there is not just one explanatory problem called emotion. A theory may be trying to explain why recurring emotional forms exist, how bodily changes contribute to experience, why the same situation produces different emotions, how affective states are organised, how variable episodes acquire emotional meaning, or how uncertain bodily and environmental information is interpreted.

Different theories often place different parts of this landscape near the explanatory centre.

They can also share mechanisms, operate at different levels, or combine commitments that are usually associated with more than one tradition.

A single theory may therefore appear in more than one part of the map.

The most useful way to understand the field is consequently not as a set of mutually exclusive camps but as a set of overlapping families of explanatory approaches.

These families are classifications, not natural divisions inside emotion itself. As Scientific Models Are Tools establishes, a map can help organise a complex target without reproducing its structure literally.

The purpose of this map is therefore not to decide which family is correct.

It is to clarify what each family characteristically emphasizes, what kind of problem it is trying to solve, and where apparent theoretical differences become genuine disagreements.

Evolutionary and Basic-Emotion Approaches

One influential family places evolved function and recurring emotional differentiation near the centre of explanation.

Many approaches in this tradition begin from the observation that humans and other organisms repeatedly confront problems involving danger, loss, conflict, contamination, affiliation, caregiving and opportunity. Emotional organisation may, in some formulations, reflect adaptations or evolved systems that help coordinate responses to recurrent classes of situations.

Illustrative functional interpretations might connect fear with responding to danger, disgust with contamination avoidance, or anger with certain forms of obstruction or conflict.

But these examples should not be mistaken for fixed definitions of the emotions themselves.

The family is internally diverse.

Different theories disagree about what makes an emotion "basic," which emotions qualify, how evolution should constrain category structure, how much learning modifies emotional responding, and how specific any underlying neural or bodily organisation must be.

Some formulations emphasize characteristic expressions or action tendencies.

Others place greater weight on neural systems or cross-species continuity.

Still others are more fundamentally concerned with evolutionary function than with finding one unique biological signature for every emotion category.

This matters because evolutionary is broader than any one version of basic-emotion theory.

The two overlap substantially, but an evolutionary account need not commit itself to one rigid set of discrete emotional modules.

Likewise, a basic-emotion account need not claim that every instance of fear or anger must display an invariant face, physiological pattern and neural circuit.

The family is better understood through a recurring concern:

Why do differentiated and recurrent emotional forms exist, and how might they organise adaptive action?

That question connects directly to D7.1's problems of differentiation, regularity and action.

It also overlaps with other families.

An evolved emotional system could still depend on rapid evaluations of what a situation means for the organism. Such a theory can therefore contain appraisal-like processes without ceasing to be evolutionary.

Similarly, an evolved functional account might be implemented through bodily, neural or even predictive mechanisms.

The family label identifies where explanatory emphasis falls.

It does not determine every mechanism the theory is allowed to contain.

Bodily and Physiological Approaches

Another recurring family assigns substantial explanatory importance to bodily state and the sensing of bodily change.

Historically, influential bodily theories proposed that emotional feeling is closely related to perceiving physiological changes produced during an encounter with the world.

That historical tradition is sometimes compressed into a simple sequence:

event → bodily change → emotional experience

But contemporary bodily approaches are considerably more varied.

Researchers and theorists may focus on autonomic regulation, visceral signals, breathing, cardiovascular activity, bodily preparation for action, neural representation of internal states, or the processing of interoceptive information.

As explored in Interoception, the nervous system continually receives information about conditions inside the body.

This makes questions about bodily sensing central to several emotion theories:

How does the changing body contribute to the character of emotional experience?

How does the nervous system represent those changes?

When does bodily information help distinguish one emotional episode from another?

Different bodily theories answer those questions differently.

Some formulations treat physiological change as causally important.

Others treat bodily feedback as partly constitutive of experience.

Still others treat bodily signals as information that must be integrated with context, memory or expectation.

Bodily processes can also play regulatory roles without being sufficient to determine the emotional episode by themselves.

So the bodily family cannot be reduced to one linear body-first theory.

Nor can modern interoceptive approaches be treated as synonymous with the entire bodily tradition.

This is also where Mind, Brain and Body Form One System becomes essential.

A bodily theory is not a theory of a separate "body system" competing with a separate "mind system." Bodily, neural, cognitive and experiential processes occur within the same interacting organism.

Theoretical families differ in which processes they place near the explanatory centre.

They do not describe separate human beings.

This family also overlaps naturally with others.

Appraisal processes may influence bodily organisation.

Constructionist theories may treat bodily affect as one ingredient in emotional episodes.

Predictive or inferential models may interpret interoceptive signals as evidence about hidden bodily states.

Those overlaps are not exceptions.

They are part of the theoretical landscape.

Appraisal Approaches

Appraisal approaches place evaluation of the person–situation relationship near the centre of emotion.

Their characteristic question is not simply:

What happened?

It is closer to:

What does what happened mean for this organism, given its goals, concerns, possibilities and circumstances?

This makes appraisal theories especially well suited to a problem established in D7.1:

why can similar situations produce different emotional outcomes?

Criticism in public may become anger for one person, shame for another, relief for another and little of emotional consequence for someone else.

Appraisal approaches attempt to explain such differentiation by examining how situations are evaluated in relation to factors such as relevance, goals, agency, coping possibilities, novelty or significance.

Those are common examples across appraisal traditions, not a universal checklist shared identically by every theory.

An important misconception must also be removed.

Appraisal does not mean conscious verbal thought.

Many major appraisal models allow evaluative processing to be rapid, automatic, recurrent, multi-level and partly outside conscious awareness.

An organism need not first think:

"I have analysed this situation and concluded that I should now become afraid."

Appraisal refers more broadly to processes through which circumstances are evaluated in relation to what matters for the organism.

Appraisal theories themselves differ substantially.

They disagree about which evaluations are important; how appraisal unfolds through time; whether different appraisal processes occur sequentially or recursively; how appraisal relates to bodily changes; how tightly appraisal patterns map onto familiar emotion categories; and how much of the process reaches conscious awareness.

No single appraisal model therefore represents the whole family.

The family also overlaps with evolutionary approaches. Rapid evaluative processes could themselves be biologically prepared.

It overlaps with bodily approaches because evaluations and physiological responses can influence one another within an unfolding episode.

And some predictive or inferential approaches reinterpret parts of the evaluation problem using probabilistic or generative architectures.

The family distinction remains useful because appraisal theories characteristically place evaluation of significance near the explanatory centre.

That emphasis does not make appraisal the only process involved.

Dimensional Approaches

Dimensional approaches begin from a different problem.

Instead of asking first which discrete category an emotional state belongs to, they ask whether affective states can be organised along shared dimensions.

Two especially influential dimensions are valence — roughly, pleasant versus unpleasant — and arousal — lower versus higher activation.

Other dimensional models propose additional or differently defined dimensions, including motivational direction in some formulations.

The central insight is that emotional states can resemble one another along some dimensions while differing along others.

Fear and anger may both involve high activation while differing in experience, interpretation and action.

Calm satisfaction and excitement may both be pleasant while occupying very different positions in arousal.

Dimensional approaches therefore speak directly to D7.1's problem of overlap between apparently different emotional categories.

They can provide a compact description of how affective states relate.

But an essential distinction follows from Correlation, Prediction, Causation and Mechanism:

A map of affective structure does not automatically explain how an emotional episode was generated.

A dimensional model may tell us that two states are similar in valence and different in arousal.

That does not, by itself, tell us what caused either state; how bodily responses arose; what differentiated their meaning; what neural mechanism implemented them; or why one particular episode developed when it did.

Some dimensional theories are primarily structural or descriptive.

Others embed affective dimensions within broader explanatory frameworks.

The family also contains disagreement about which dimensions are primary, how many are needed, how emotion categories relate to them, and what aspect of affect the dimensions are intended to describe.

So dimensional approaches should not be reduced to one circumplex model.

Nor should they be described as theories that simply "reduce emotion to valence and arousal."

Their recurring emphasis is more modest and more useful:

Can shared dimensions reveal important structure across affective states that categorical labels alone obscure?

That structural emphasis can also become part of other families.

A constructionist theory, for example, may use dimensional affect as one ingredient in a broader account of differentiated emotion.

Psychological Construction Approaches

Psychological construction approaches place another problem near the centre:

How can differentiated emotional episodes emerge from combinations of more general psychological processes?

This becomes particularly important once D7.1's within-category variation is taken seriously.

If two episodes both called anger can differ in bodily response, context, action and experience, must there still be one dedicated "anger mechanism" producing every instance?

Constructionist approaches characteristically resist assuming that every conventional emotion category requires its own dedicated mechanism.

Instead, many formulations explain emotional episodes through combinations of processes that are not unique to any one emotion.

Depending on the theory, these may include affective states; bodily information; attention; memory; learned category knowledge; conceptual processes; context; language; prediction; or other more general psychological operations.

But these ingredients are not universal requirements across the entire family.

Some constructionist theories place strong weight on core affect.

Some emphasize conceptualization.

Some give language and learned category knowledge a larger role.

Others emphasize bodily regulation, interoception or predictive mechanisms.

The family is therefore internally heterogeneous.

The word constructed is also easy to misunderstand.

It does not mean that emotions are imaginary.

It does not imply that people consciously invent them.

It does not mean that emotional episodes can take any arbitrary form.

And psychological construction is not the claim that society simply fabricates emotions.

Constructionist theories still operate under biological, bodily, neural, developmental, learned and environmental constraints.

Their disagreement with other approaches concerns how those constraints are organised into emotional episodes and categories.

This family is especially relevant to D7.1's questions about within-category variation; context sensitivity; category formation; and the relation between general processes and differentiated experience.

Its boundaries are also particularly porous.

Dimensional affect can form part of a constructionist account.

Interoceptive information can be one ingredient.

A constructionist theory can also adopt predictive or active-inference mechanisms.

That does not make psychological construction synonymous with predictive processing.

It means one theory may occupy more than one region of the map.

Predictive and Inferential Approaches

"Predictive and inferential" is the loosest family label in this taxonomy.

It is useful as a convenience grouping for approaches that place prediction or inference about hidden bodily or environmental states near the explanatory centre.

Their formal and mechanistic commitments can differ substantially.

Some approaches focus on inference about the causes of bodily signals.

Others explicitly use predictive-processing architectures.

Some invoke Bayesian ideas more broadly.

Some extend related frameworks into action and regulation through active inference.

The family should therefore not be treated as one theory with one architecture.

A recurring question across several formulations is:

How does an organism interpret signals whose meaning is not uniquely given by the input itself?

Bodily sensations, environmental cues and contextual information may all be compatible with more than one interpretation.

Many inferential approaches propose that prior information, current evidence and uncertainty shape how those signals are interpreted.

Some models also use quantities such as precision or prediction error.

Those concepts are formulation-specific rather than universal requirements of every inferential emotion account.

Emotion-specific interoceptive inference provides a particularly clear overlap with bodily approaches. Internal bodily signals are treated not as fully self-interpreting messages but as information from which underlying bodily causes must in some sense be inferred.

Predictive processing is one influential framework within this broader landscape.

As established in Predictive Processing, predictive processing itself refers to a family of related approaches rather than one uniformly specified model.

And Predictive Processing: Scope and Limits adds an important safeguard: computational claims, broader theoretical frameworks and proposed neural implementations must not be collapsed into one evidential object.

The same distinction matters here.

Predictive/inferential emotion approaches ≠ predictive processing as a whole.

And: predictive processing ≠ active inference.

Active-inference formulations may add commitments concerning action, regulation or policy selection within related formal architectures, but those commitments are not automatically part of every predictive account.

Nor are predictive approaches simply psychological construction under another name.

Some constructionist theories adopt predictive mechanisms.

Other predictive theories need not make the same claims about emotion categories, conceptualization or construction.

The overlap is genuine.

The equivalence is not.

This family is especially relevant to D7.1's problems involving ambiguous bodily information, context-sensitive interpretation and inference under incomplete information.

But its place at the end of this taxonomy should not be mistaken for a claim of theoretical progress.

Newer terminology is not evidence of greater truth, completeness or scientific status.

Predictive and inferential approaches simply foreground a different set of explanatory problems.

The Families Overlap More Than the Labels Suggest

Once the families are viewed together, the boundaries become visibly porous.

That is not a flaw in the map.

It follows from what the families are classifying.

They identify explanatory emphases, not exclusive identities.

An evolutionary theory can contain rapid evaluative processes.

An appraisal theory can include substantial bodily organisation.

A dimensional model can supply affective structure to a psychological construction account.

Interoceptive theories can use predictive inference.

Constructionist accounts can adopt predictive or active-inference mechanisms.

A single theory can therefore occupy more than one region of the taxonomy.

A hybrid theory is not a classification failure.

Nor, however, does hybridity imply that every theoretical commitment can be reconciled.

Theories may genuinely disagree about whether dedicated emotion-specific mechanisms are necessary; general psychological processes are sufficient for differentiation; particular emotion categories have privileged biological status; concepts play a constitutive role; bodily information has causal priority; certain appraisal processes are necessary; categories or dimensions provide the more fundamental organisation; or one computational architecture correctly describes the generation of emotional episodes.

These are real candidate disagreements.

This piece does not decide them.

The point is to locate them accurately.

This is where Human Behaviour Has Multiple Causes also matters.

Human outcomes can depend on multiple interacting causal processes. Discovering one relevant contributor does not automatically make all others irrelevant.

But multiple causation does not imply that every proposed theory is partly correct.

Two theories can identify different processes and still make incompatible claims about how those processes are organised.

Likewise, Mind, Brain and Body Form One System means that bodily, cognitive and neural explanations need not be treated as explanations of separate systems.

A theory may emphasize one level while remaining compatible with accounts at another.

Or it may make claims that genuinely conflict across levels.

The taxonomy helps us ask which is happening.

A Map Is Not a Verdict

The six families can now be summarized by their recurring emphases.

Evolutionary and basic-emotion approaches often foreground evolved function, recurring differentiation and emotional organisation.

Bodily and physiological approaches characteristically give bodily change and bodily sensing substantial explanatory importance.

Appraisal approaches foreground evaluations of situations in relation to goals, concerns and circumstances.

Dimensional approaches often organise affective states according to shared dimensions and relationships.

Psychological construction approaches characteristically explain emotional episodes through combinations of more general processes.

Predictive and inferential approaches place prediction or inference about hidden bodily or environmental states near the explanatory centre.

Those summaries compress substantial variation.

They are useful because they make comparison easier, not because every theory fits cleanly into one slot.

This is also where differences in kind of explanation become important.

A theory may provide a functional explanation of why an emotional system exists.

Another may offer a structural description of how affective states relate.

Another may propose a cognitive or computational process.

Another may identify causal mechanisms.

Another may propose a neural implementation.

These are not automatically rival achievements.

A functional account of why an emotional system might have evolved does not by itself compete with a computational account of how information is processed.

A dimensional description of affective structure is not automatically an alternative to a neural implementation.

A cognitive appraisal account and a bodily mechanism may be complementary—or they may conflict, depending on what causal claims each actually makes.

Different language can therefore conceal overlap.

The same language can conceal disagreement.

Before two theories are treated as direct competitors, the useful questions are more precise:

What phenomenon is each theory trying to explain?

What does each place near the explanatory centre?

At what level is the claim being made?

Is the theory describing structure, proposing a cause, specifying a mechanism or suggesting an implementation?

Which commitments does it share with other families?

And where do the claims actually become incompatible?

That is what this taxonomy is for.

It does not answer: which emotion theory is correct?

It makes that question more scientifically precise.

Once theories have been aligned by explanatory target, level and claim, the next problem becomes possible to address:

Where does emotion science genuinely agree, where does it genuinely disagree, and which apparent disputes arise because theories are using different terms or explaining different parts of the phenomenon?

That is the task of D7.3 — Where Emotion Science Agrees and Disagrees.