Definition
Outcome specificity is the principle that a claim of human change must identify the particular outcome in which a difference occurred, because evidence that one outcome changed does not by itself establish change in another.
One Word, Different Outcomes
Once we ask what changed?, the answer can split in several directions.
Imagine someone who used to remain silent in meetings and now contributes regularly. The visible difference is behavioural: they speak more often. But that behaviour could sit within very different patterns of change. Their anxiety may have eased. Their skill in expressing disagreement may have improved. Their expectations about colleagues may have shifted. They may have begun to see themselves differently. Or the organisation may simply have become safer, making the same person more willing to speak.
The outward difference is similar. The outcome underneath it may not be.
This is the problem of outcome specificity: a claim about change should identify the particular dimension in which the difference occurred, because evidence for one outcome cannot simply be transferred to another.
The Library distinguishes eight principal outcome classes commonly gathered under the word change. They are not eight mutually exclusive natural kinds, nor an exhaustive scientific taxonomy. They are a canonical way of separating different claims that established research traditions already study under different names.
The categories can overlap. Skill acquisition may alter behaviour. Belief revision may affect emotional responding. Structural change may reshape relationships. Identity change may influence action. Their interaction is precisely why distinction matters.
Nor are these categories simply levels of explanation. A skill can be studied behaviourally, cognitively or neurally. An identity change can be investigated through self-report, behaviour, social recognition or developmental history. The category identifies what outcome is alleged to have changed; the level of explanation identifies how that change is being described or investigated.
With that boundary in place, the principal outcomes become easier to distinguish.
Eight Outcomes Commonly Called Change
State change
A state change is a difference in a relatively current psychological, emotional, physiological or experiential condition.
Mood shifts. Attention narrows or expands. Arousal rises and falls. Someone feels more anxious before an examination, calmer after leaving a threatening setting or more alert after sleeping.
These differences matter even when they are temporary.
What they establish is a changed condition at a particular time. Whether that difference reflects a more enduring trait or recurring pattern is a separate question.
Behavioural change
Behavioural change concerns a difference in action or a pattern of action.
Someone begins exercising regularly, stops checking their phone during meetings, speaks more often during conflict or repeatedly enters situations they once avoided.
The claim is specific: what the person does has changed.
Its explanation remains open. The altered behaviour may reflect skill, belief, emotion, incentives, social pressure, opportunity or some combination of them. Behavioural change identifies the changed pattern of action; it does not, by itself, locate the source of that change elsewhere in the person.
Skill acquisition
Skill acquisition concerns a change in acquired capability.
This is why it cannot be identified with performance on any single occasion.
A pianist may possess a well-developed skill and still perform badly when exhausted. An athlete can execute a familiar movement poorly under pressure. Conversely, favourable circumstances can produce good performance without demonstrating that a capability has been reliably acquired.
Research on learning and performance has repeatedly shown why this distinction matters: conditions that improve immediate performance do not always produce the strongest later learning, while learning can occur under conditions in which practice looks relatively unimpressive.
Skill acquisition therefore asks a different question from behavioural observation. One concerns acquired capability; the other concerns what happened on a particular occasion or across a pattern of occasions.
Belief revision
Belief revision refers to a change in what a person takes to be true, probable or credible.
Someone revises an expectation, abandons a previous conviction or updates their judgement in light of new evidence.
Such a change may eventually influence behaviour, but the relationship is not one-to-one. Adjacent literatures on attitudes and intentions make the broader point clearly: changes in explicit cognition can predict behaviour without mapping perfectly onto it. Context, incentives, habit, social constraints and competing goals can all alter what follows from what a person thinks.
Belief revision is therefore a legitimate outcome in its own right. It concerns altered cognition, not guaranteed behavioural consequence.
Emotional learning
Emotional learning requires a slightly different treatment because the phrase often names a process rather than an outcome.
Learning can alter how situations are emotionally represented and responded to. Research on fear acquisition, extinction and retrieval offers a particularly clear example: explicit expectations, physiological responding, subjective fear and behaviour can follow partially different trajectories.
For this taxonomy, the relevant outcome is not the learning process itself but the changed emotional response, affective expectation or learned association produced through it.
This distinction matters because knowing and feeling are not always synchronised. A person can explicitly know that a situation is safe while continuing to react fearfully. Emotional responding can also shift without a correspondingly elaborate change in explicit belief.
Laboratory fear paradigms should not be treated as universal models of emotional life. They do, however, demonstrate the conceptual point that emotional change cannot simply be collapsed into belief revision or overt behaviour.
Identity change
Identity change concerns some alteration in self-conception, self-representation or socially situated identification.
A person may cease to organise their life around a previous role, begin to understand themselves as belonging to a group they had not previously identified with, or reinterpret their past and future through a different account of who they are.
Identity is not one hidden substance waiting to be measured before and after transformation. Different traditions emphasise narrative identity, social identity, role identity, self-concept and related constructs. The boundaries are therefore theoretically contested.
That does not make identity change meaningless. It makes specificity essential.
A transient feeling about oneself is not automatically evidence of identity change. Nor does a changed self-description establish that behaviour, emotion or relationships changed in parallel. The claim concerns some aspect of self-representation or identification; what follows from it remains an empirical question.
Relational change
Some outcomes belong most naturally to recurring interaction rather than to one individual considered alone.
Relationships develop patterns: who initiates contact, how conflict escalates, how trust is expressed, how decisions are negotiated, whether vulnerability is reciprocated. These patterns can change even when neither participant can be adequately described as undergoing the same internal change.
Dyadic research provides the clearest examples, but the principle extends to broader interpersonal systems as well.
Relational change therefore refers to a difference in the organisation or quality of interaction. The altered property may be a pattern produced between people rather than a trait possessed by either person in isolation.
Structural change
Structural change is the boundary case in this taxonomy.
Unlike the preceding categories, the primary thing that changes may lie outside the individual: an institution, rule, organisational practice, material environment, incentive structure or social norm.
A workplace changes how disagreement is handled. A school alters its disciplinary system. A city builds safer transport infrastructure. A law expands or restricts available choices.
These are not internal personal changes.
Their inclusion matters because behavioural and psychological outcomes are often interpreted as though they must have originated in the person. Sometimes the surrounding conditions changed instead.
If cycling increases after safe infrastructure is built, the behavioural difference is real. It does not follow that residents underwent a corresponding identity or value change. If employees contribute more after punitive norms disappear, the organisation may have changed more than the employees did.
Structural change is therefore included to preserve a boundary: change in the conditions organising human action should not be mistaken for equivalent change within the individual.
The Outcomes Can Separate
The taxonomy becomes useful when several outcomes fail to move together.
Someone may repeatedly enter a feared situation while continuing to feel considerable fear. Behaviour changed; emotional responding may have changed less.
Another person may revise a belief while behaving as before because habit, constraint or circumstance remains unchanged.
Skill can improve even when immediate performance deteriorates under fatigue or pressure.
A structural reform can alter behaviour without producing a corresponding identity change.
These outcomes can also co-occur and influence one another. A changed belief may contribute to new behaviour. Repeated behaviour may alter skill. Structural conditions can reshape relationships. Identity may affect the situations a person enters and the interpretations they bring with them.
Distinction therefore does not imply independence.
The relevant principle is narrower:
when several outcomes appear together, their co-occurrence must be demonstrated rather than assumed.
A visible behavioural difference should not silently expand into a claim that belief, emotion, skill and identity changed with it. Equally, an internal shift should not be treated as equivalent to outward behavioural transformation.
The categories distinguish the property about which the claim is being made, not separate worlds inside the person.
No Hierarchy of Depth
Once different outcomes are named, it is tempting to arrange them from superficial to deep.
Behaviour is sometimes treated as merely external. Belief seems more internal. Emotion may be described as deeper still. Identity is then cast as the most fundamental form of personal change.
The taxonomy does not support that hierarchy.
These categories answer different questions.
A behavioural change can be enduring and consequential. A revised belief may have little effect beyond the moment in which it is expressed. Identity change may reorganise important parts of a life, or it may remain narrow and context-dependent. Structural change can alter what millions of people are able to do without requiring each of them to undergo an internal transformation.
Nor does a biological description automatically outrank the others. Neural change concerns a level of explanation, not a universal measure of depth. A skill, belief or emotional response may have neural correlates without the neural description replacing the psychological or behavioural claim.
Depth, importance and durability are therefore separate dimensions from outcome type.
D1.2 asks only: what kind of outcome changed?
It does not rank the answer.
Why Outcome Specificity Matters
Once the outcome is specified, the next scientific questions become clearer.
A changed state raises questions about recurrence and stability. A behavioural difference raises questions about process, context and persistence. A claimed identity change raises questions about evidence and operational definition. A structural change raises questions about how altered conditions interact with individual outcomes.
Those questions differ because the objects differ.
Evidence for belief revision cannot, by itself, establish behavioural change. Improved performance on one occasion cannot automatically establish skill acquisition. Changed organisational conditions cannot simply be redescribed as internal personal transformation.
The word change becomes scientifically useful only when the claim attached to it is made specific enough to evaluate.
That is the purpose of this taxonomy.
It does not attempt to capture every construct that can change in human life. Nor does it imply that the eight outcome classes are isolated from one another. It establishes a discipline of description.
When someone says that a person changed, the statement may eventually prove to involve several dimensions at once.
But the first question remains the same: what changed?
Before asking how it happened, whether it lasted or what produced it, the outcome has to be specified.
The research basis for the most consequential distinctions is strong: learning and momentary performance are demonstrably separable; recent reviews show that changes in attitudes and intentions translate imperfectly into behavioural change; extinction research distinguishes emotional learning from the later expression and retrieval of learned responses; identity research treats identity development as contextual and dynamic; relationship research supports dyadic and system-level analysis; and contemporary norm and ecological research supports treating social conditions as objects of change in their own right.