The Problem Hidden Inside an Ordinary Word
Change is one of those words that seems clear until precision becomes necessary.
We say that someone has changed because they behave differently, believe something new, recover from illness, become less anxious, acquire a skill, abandon an ambition or begin to see themselves in another way. We use the same word for a child's development, a patient's recovery, a shift in political conviction, the effects of neurological injury and the slow alteration of a relationship. In ordinary conversation this flexibility causes little trouble. Context usually tells us what kind of difference matters.
For a science of human change, that looseness becomes consequential.
The problem is not that everyday language is defective. Broad concepts are useful precisely because they allow us to group related experiences without stopping to specify their structure each time. Memory can refer to recognising a face, recalling a conversation or retaining a motor skill. Stress can refer to an experience, a physiological response or the demands placed upon a person. The fact that a word covers several related phenomena does not make it meaningless.
But it does mean that the word itself cannot tell us which phenomenon is being explained.
A person who stops avoiding public speaking has changed behaviourally. Whether their fear has diminished is another question. Someone may revise a deeply held belief without encountering a situation in which the new belief alters what they do. A person can regain motor capacity after injury while continuing to experience themselves as fundamentally unchanged; another may describe a profound shift in identity while their daily routines remain recognisably similar.
All of these can reasonably be called change. They are not therefore the same change.
That distinction is easy to lose because ordinary language tends to compress changes in particular aspects of a person into global statements about the person as a whole. She has changed. He is a different person now. Such descriptions may be meaningful, sometimes deeply so, but they gather many possible claims into one expression. Behaviour may differ. Values may have shifted. Emotional responses may be different. A capacity may have been acquired or lost. Relationships may have reorganised. A person's interpretation of their own history may have altered.
Scientific inquiry has to unpack that compression.
Strictly speaking, the object that changes is rarely "the whole person" in any simple sense. What can be identified are differences in particular properties, capacities, patterns or relations involving that person. A person's behaviour can change. So can their expectations, physiological regulation, skill, emotional responses or recurrent way of interacting with others. Several such differences may occur together and justify the broader judgement that the person has changed. But the broader judgement should not be allowed to conceal the more specific claims on which it depends.
This matters because different sciences often focus on different parts of that picture. Psychology may investigate cognition, emotion, personality or behaviour. Neuroscience may examine alterations in neural organisation or function. Developmental science studies changes across the lifespan. Sociology and anthropology may focus on roles, relationships, institutions or cultural conditions within which individual lives unfold. Medicine may be concerned with symptoms, capacities, pathology or recovery.
These disciplines may all be studying genuine changes involving the same human being while asking different questions about them.
Confusion begins when findings from one kind of change are allowed to stand in for another. A behavioural difference does not by itself establish a corresponding change in belief or emotion. A neural difference does not, without further argument, tell us what has changed psychologically. A person's report that they feel transformed is important evidence about experience, but it does not automatically specify what else has changed or how.
The point is not that these dimensions are independent. They plainly are not. Human beings are integrated organisms embedded in social environments. Changes in one aspect of a person's life may accompany, constrain or contribute to changes elsewhere. The conceptual problem is that connection does not erase distinction.
Before asking how people change, then, a more basic question has to be made explicit: what, exactly, is the claim when we say that change has occurred?
At minimum, some specified aspect of the person must differ across time. But even this needs care. Human organisms are changing continuously. Neural activity fluctuates, hormones vary, attention shifts, cells turn over, posture changes. If every detectable difference counted equally as the kind of human change we are trying to explain, the concept would become too broad to do useful work.
The relevant difference has to concern some property, capacity, pattern or relation under investigation. That does not yet tell us whether the change is important, durable, adaptive or transformative. It simply prevents the word change from becoming a substitute for specification.
The ordinary question—Has this person changed?—therefore contains several more precise ones.
What changed? Compared with when? In what respect?
Those questions are where a coherent science of human change has to begin.
Why Science Must Specify Its Object Before Explaining It
Scientific inquiry does not ordinarily proceed by fixing a perfect definition and only then beginning observation. Concepts and evidence develop together. Researchers revise categories when measurements fail, sharpen definitions when theories expose ambiguities and sometimes discover that phenomena once treated as one thing are better understood as several.
What scientific explanation does require, however, is enough clarity about its object for claims to accumulate rather than pass one another unnoticed.
If two research programmes use the same term while measuring different phenomena, apparent disagreement may be difficult to interpret. They may genuinely conflict. They may be describing different dimensions of the same system. Or they may simply be attaching one name to different objects. Until that ambiguity is resolved, more data do not necessarily produce more understanding.
This is especially important in psychology because many of the properties psychologists study are not objects that can be observed directly.
Researchers can record a person answering a question, pausing before making a choice, leaving a room, remembering a word or returning repeatedly to the same activity. These events are observable. What they count as is already partly a matter of classification: a movement may be coded as avoidance, a response as recall, a repeated action as persistence. Claims about anxiety, memory, motivation, self-confidence or belief require another inferential step. The construct is not identical to any one observation used to investigate it.
This problem sits at the centre of psychological measurement.
Cronbach and Meehl's classic work on construct validity gave influential form to a problem that remains current: how can a science justify claims about attributes that are not directly observable? Their answer was not that an invisible psychological construct becomes legitimate once it is given a test. Its scientific meaning depends upon the pattern of relationships connecting the construct to observations, other constructs and theoretically relevant outcomes.
Later work in validity theory made this picture more explicit. Contemporary psychometrics does not treat validity as a permanent stamp attached to a questionnaire or test. Validity concerns the strength of the evidence and reasoning supporting particular interpretations of measurements for particular purposes. No single score reveals a psychological property in isolation.
That distinction matters immediately for claims about change.
Suppose a person's score on a self-confidence questionnaire rises over six months. Something measurable has changed: their responses. Whether self-confidence itself has changed is a further inference. The difference might reflect genuine psychological change, but it could also be influenced by measurement error, transient mood, changes in interpretation, altered circumstances or the limitations of the instrument. Additional evidence can make one interpretation more credible than another, but the measurement and the construct should not be treated as the same thing.
Behaviour involves a similar distinction. The event may be directly observable while its psychological significance remains uncertain. If someone leaves a crowded room, we can observe the departure. We cannot infer anxiety from the movement alone. They may be overwhelmed, bored, late for another appointment or looking for fresh air.
This distinction between observation, classification and inference is easy to blur because in everyday life we perform all three almost instantaneously. We see a friend withdraw from a conversation and experience the action immediately as hurt, anger or disinterest. Scientific inquiry has to separate what was observed from what was inferred, not because inference is illegitimate, but because the inference is part of the claim that requires support.
For much of twentieth-century psychology, behaviourism offered one response to this difficulty: restrict scientific analysis as far as possible to observable relations between behaviour and environment. That discipline was not intellectually trivial. It imposed a valuable demand that explanations of human behaviour remain answerable to public evidence rather than unconstrained speculation about hidden mental causes.
Yet observable behaviour cannot by itself exhaust the subject matter of human change.
People may acquire knowledge they have not yet had occasion to use. Emotional experience can alter while outward behaviour remains constrained by circumstance. Two people can perform the same action while differing substantially in intention, expectation or subjective experience. A person may learn to tolerate a feared situation while continuing to feel fear, producing behavioural similarity to someone whose fear has largely disappeared.
These are not arguments for privileging an invisible "inner self" over observable behaviour. They show why a single class of observation cannot settle every question we might ask about a changing person.
The same caution applies to biological measurement. Neural activity, endocrine responses and other physiological measures provide evidence about biological processes. They can be indispensable. But a change described biologically does not automatically establish a particular psychological change, just as a psychological description does not specify the biological processes that make it possible.
Different kinds of claims therefore demand different forms of specification.
If the claim concerns behaviour, the relevant behaviour has to be identified. If it concerns belief, emotion or identity, the construct and the evidence used to infer it need to be made clear. If it concerns neural organisation, the biological property under investigation must be specified. These forms of description may eventually be connected, but their connection cannot be assumed merely because they concern the same person.
This is the methodological reason the word change cannot remain undefined at the centre of the Library.
Before explanation, there has to be an object precise enough to explain.
That does not mean discovering one definition that every discipline must adopt. The more plausible starting point is the opposite: human change encompasses several distinguishable kinds of difference, and scientific clarity depends upon identifying which of them a particular claim concerns.
The next task, then, is not to search for the one phenomenon hidden beneath every use of the word. It is to ask what kinds of differences can legitimately count as changes in a person without pretending that they are interchangeable.
Human Change Is a Family of Phenomena
Once change is treated as something that must be specified rather than assumed, the apparent simplicity of the concept begins to dissolve.
A person can change in what they do, what they know, what they expect, what they feel, what they can tolerate, how their body functions, how they relate to others or how they understand themselves. These differences can unfold together, but they need not. Nor do they necessarily share the same causes, timescales or evidential signatures.
This is why human change is better understood as a family of related phenomena than as one uniform event.
The claim is modest but consequential. It does not propose a grand taxonomy, and it does not imply that the different forms of change are independent. It simply recognises that the category human change contains several distinguishable kinds of difference involving the same person.
Consider someone who previously avoided speaking in front of groups and now gives presentations regularly. The behavioural change is clear enough: the person does something they did not reliably do before. But several psychologically distinct possibilities remain open.
Their fear may have diminished. Their fear may be unchanged while their willingness to act has increased. They may have become more skilled and therefore less uncertain about what to do. Their role at work may now require public speaking, making avoidance less available. The environment may have changed in ways that make the behaviour easier. Several of these may have occurred together.
At the level of observable action, the cases can look similar. As accounts of what changed in the person, they are not equivalent.
The reverse also occurs. A person can change in ways that are not immediately visible in behaviour. They may abandon a conviction without yet encountering a situation in which the new belief affects action. Their emotional response to a past event may alter while daily routines remain almost identical. A previously central ambition may cease to matter, even though external commitments keep behaviour stable for some time.
Such examples reveal why the person changed is too coarse to function as a complete scientific description.
What science can ordinarily identify are changes in particular properties, capacities, patterns or relations: a change in performance, a change in expectation, a change in reported affect, a change in physiological response, a change in recurring interaction with others. The broader statement that the person has changed may be a reasonable summary when several of these differences accumulate, but it is a summary of more specific claims rather than a separate phenomenon that can be measured directly.
This distinction also helps separate change within a person from change involving a person.
Suppose someone becomes markedly more talkative after joining a group in which they feel safer. Their behaviour has changed. Whether the relevant change should be attributed primarily to an enduring property of the individual, to the social environment, or to the relation between person and context is not settled by observing the difference alone.
Likewise, a person may behave differently because their circumstances have changed while many of their dispositions remain relatively stable. Someone begins cycling to work after a protected cycle route opens. A child speaks more freely in a classroom after moving from a punitive teacher to a supportive one. An employee becomes more participatory after organisational incentives change.
These are genuine changes in behaviour involving a person. They do not automatically establish a corresponding transformation in the person's enduring characteristics.
This matters because human behaviour is rarely generated by the person in isolation. It emerges through transactions among capacities, expectations, histories, immediate situations and broader environments. A difference in action can therefore be located at more than one place in the system that produced it.
The purpose here is not to decide where every change ultimately resides. In many cases, that would require a causal explanation we do not yet have. The more limited task is to avoid attributing a change globally to the person when the evidence establishes only that some aspect of the person–environment system differs from before.
The same discipline is needed for internal change.
Psychological language often encourages us to treat constructs such as confidence, resilience or identity as if each were a single property that rises or falls uniformly. In practice, broad psychological concepts can contain several distinguishable components. Someone may report feeling more confident in one domain while remaining uncertain in another. They may describe a changed identity while retaining many established habits. Emotional responses may shift in one relationship and remain largely unchanged elsewhere.
This contextual dependence does not make the change unreal. It tells us that its scope matters.
A scientifically useful claim therefore needs more structure than something is different now. It must identify the dimension in which the difference is alleged to exist.
That may involve a capacity a person possesses; a pattern they repeatedly enact; a belief or expectation they hold; a physiological response; a subjective experience; or a relation between the person and the surrounding environment.
These categories are not meant as a final classification of human change. Their role here is to expose the basic plurality hidden by the ordinary verb to change.
They also prevent a common mistake in explanation: treating evidence in one dimension as if it automatically settled what happened in another.
A person who repeatedly enters situations they once avoided has changed behaviourally. Whether they have also changed emotionally is an empirical question. Someone whose reported beliefs alter has changed in what they report and may have undergone a cognitive change; whether their behaviour will follow cannot be assumed. A person whose physiological response changes under repeated exposure has changed biologically in some respect; what that means for conscious experience requires further evidence.
This is the stronger sense in which human change is plural. It is not merely that different disciplines use different vocabulary. The underlying object itself can differ according to which property, pattern or relation is changing.
The most useful foundational question is therefore not:
Has this person changed?
It is:
What about this person has changed, relative to what, and in what respect?
That question does not yet tell us whether the change is lasting, important or causally significant. It does something more basic. It prevents a global judgement about the person from outrunning the specific difference that has actually been established.
And once the question is framed this way, another difficulty becomes unavoidable.
Some of the properties we care most about are not directly visible. They have to be inferred from evidence.
Seeing Change and Inferring Change
Scientific claims about human change differ not only in what they describe, but in how directly the relevant phenomenon can be observed.
Sometimes the object is comparatively accessible. A researcher can record whether a person completes a task, enters a room, presses a button, speaks during a meeting or returns repeatedly to an activity. These events can be documented.
Even here, however, observation is not the whole process.
The event must usually be classified before it becomes scientifically meaningful. Leaving a room is an observable movement. Calling it avoidance is already an interpretation based on criteria, context and theory. Speaking less frequently can be counted, but deciding that the reduction reflects withdrawal, inhibition or disinterest requires a further claim.
This distinction between observation, classification and inference is especially important in psychology because many of its central objects appear only at the third level.
Anxiety cannot be observed in the same sense as a person leaving a room. Neither can confidence, belief, motivation, intention or identity. Researchers study such constructs through indicators: reports, behaviour, performance patterns, physiological measures, informant judgements and other observations that are expected to bear some relationship to the underlying property.
The construct is not identical to any one of those indicators.
This was the problem at the heart of Cronbach and Meehl's influential account of construct validity. If psychology proposes an attribute that cannot be observed directly, its scientific usefulness depends upon the web of expected relationships connecting that construct to measurements, other constructs and observable outcomes. Giving an attribute a name, or designing a questionnaire intended to measure it, does not by itself establish that the underlying concept has been captured.
Subsequent validity theory pushed this principle further. Contemporary psychometrics generally treats validity not as a permanent feature possessed by a test, but as a judgement about whether the available evidence supports a particular interpretation and use of measurement results. The same instrument may provide stronger evidence for some conclusions than for others.
For the study of change, the implication is straightforward.
A difference in measurement is not automatically identical to a difference in the psychological property the measurement is intended to represent.
Suppose someone's score on a measure of social anxiety falls substantially between two assessments. The score has changed. Whether social anxiety itself has changed depends on what the measure captures, how reliable the difference is, whether temporary conditions could explain it and how the result fits with other relevant evidence.
The purpose of making this distinction here is not to decide what evidence would be sufficient. That belongs to a separate question. It is to establish that different kinds of human change stand at different distances from direct observation.
Behaviour sometimes allows relatively direct description of what occurred. Psychological constructs generally require inference. Physiological measures provide observations about biological processes from which broader conclusions may or may not follow.
This becomes especially clear when different indicators diverge.
Imagine two people who have both stopped avoiding a particular situation. One reports that the fear which previously drove the avoidance has largely disappeared. The other reports feeling the same fear as before but entering the situation anyway.
If the behavioural measure is the only one considered, both people appear to have changed in roughly the same way. Once subjective experience is included, the cases separate.
Neither account makes the other false.
The behavioural change is real in both cases. The experiential change appears different.
This is not a minor methodological detail. It changes what claim can legitimately be made about each person.
The same issue arises when subjective reports change without obvious behavioural movement. A person may report a major shift in values while circumstances continue to constrain action. Someone may experience less resentment towards another person but behave similarly because the structure of the relationship has not changed. A belief may be revised long before a relevant opportunity arises to test whether behaviour has followed.
Such cases make it difficult to define human change entirely in terms of overt behaviour.
Behaviourism historically represented the strongest attempt to impose that restriction. Its motivation was serious: a science should be cautious about explaining behaviour through unobservable entities that can be invoked without independent constraint. The insistence on observable relations between behaviour and environment helped establish standards of empirical discipline that continue to matter.
But the methodological value of behavioural observation does not require treating behaviour as the only legitimate object of psychological science.
A science can study inferred phenomena provided the inferential chain is explicit and evidence constrains it. Much of mature psychology already operates this way. The difficulty is not that constructs such as memory or anxiety are invisible; it is that conclusions about them must remain proportionate to the evidence available.
This also guards against the opposite error: treating subjective experience as uniquely authoritative because it feels closer to the person.
First-person reports provide access to aspects of experience that external observation cannot supply, but they are still measurements of a particular kind. Memory is fallible. Self-description is shaped by language and context. People can misunderstand their own motives or interpret the same internal experience differently at different times. None of this makes self-report worthless. It makes its epistemic status specific.
Biological measurements deserve the same discipline. A change in neural activation, hormone concentration or autonomic response is an observation at a biological level. Connecting that observation to a psychological construct requires additional theoretical and empirical work. Biological proximity should not be mistaken for explanatory completeness.
The result is a more demanding but more coherent picture.
Some human changes can be described relatively directly. Others are known primarily through inference. Many require several kinds of evidence before the underlying claim becomes persuasive. And because the route from observation to interpretation differs depending on what is alleged to have changed, the phrase evidence of change cannot mean precisely the same thing in every case.
That question—what should count as sufficient evidence that a particular form of change has occurred—belongs elsewhere in the Library.
For present purposes, the conceptual boundary is enough.
To say that a person has changed is to make a claim about some specified aspect of that person across time. Sometimes the relevant difference can be observed relatively directly. Sometimes it must be inferred from a pattern of indicators. In either case, the observation should not be allowed to carry a broader conclusion than it can support.
This gives the definition of human change another necessary constraint.
It is not enough to say what changed.
We must also know at what level the change is being described.
A behavioural pattern, a psychological construct, a physiological process and a social relation can all change within the same human life. Understanding how those descriptions fit together requires one final distinction: the levels of organisation at which a changing person can be understood.
Multiple Levels of Organisation
A human being can be described at several levels at once.
We can describe what a person does. We can describe what they report thinking or feeling. We can examine physiological regulation, neural activity, learned capacities, recurring patterns of interaction or the social conditions that shape what actions are possible. None of these descriptions automatically cancels the others. They answer different questions about an interconnected system.
This is where discussions of change often become confused.
Suppose someone learns to read. At one level, the change can be described behaviourally: accuracy improves, comprehension deepens, reading becomes faster. At another, cognitive capacities involved in language and symbolic processing have altered. Neuroscience may identify changes in the recruitment or organisation of neural systems associated with reading. Education research may focus on instruction. Developmental science may ask how the capacity emerged over time. Sociology may examine how access to literacy depends on institutions, class, language or culture.
Each account concerns the same human being.
They are not interchangeable accounts.
Knowing that reading performance improved does not tell us everything about how the underlying capacity changed. Identifying neural differences does not tell us what the person understood, what instruction produced the change or what reading now enables them to do. Describing the educational environment does not specify the biological processes through which learning was instantiated.
The point is not that these levels are independent. They are linked because they belong to the same organised system. The point is that explanatory relevance depends on the question being asked.
This distinction has appeared repeatedly across the sciences of complex systems. David Marr famously argued that cognitive systems can be understood through different kinds of explanation: what problem a system solves, how it represents and processes information, and how those processes are physically implemented. Herbert Simon emphasised that complex systems are often organised hierarchically, with meaningful structure appearing at more than one scale. Philip Anderson, writing from physics, argued that knowledge of lower-level laws does not make higher-level organisation scientifically redundant.
These frameworks are not identical, nor do they amount to a single theory of human change. Their relevance is narrower. They illustrate why a description at one level cannot simply be assumed to replace every description at another.
That matters because human sciences often carry an implicit hierarchy of explanation.
Biological measurements can appear especially authoritative because they seem closer to physical mechanism. A brain scan or hormonal measure may be treated as revealing what is really happening beneath behaviour or experience. But greater physical detail is not automatically greater relevance.
If the question is whether someone now enters situations they once avoided, behaviour may be the most direct object of study. If the question is whether their fear has changed, subjective and behavioural measures may both matter. If the question concerns neural adaptation accompanying repeated experience, neural measures become central. The appropriate level depends on what is being explained.
The same caution applies in the other direction. Psychological language can appear more complete because it maps readily onto lived experience. Saying that someone became more confident, changed their identity or developed greater self-control may capture something important. But such descriptions can also conceal unresolved questions about behaviour, context, physiology and measurement.
A higher-level description is not automatically vague, just as a lower-level one is not automatically fundamental.
The important distinction is between levels of description and levels of reality. Science may describe the same person in behavioural, cognitive, biological, interpersonal or cultural terms without committing itself to the claim that only one of those levels is ultimately real. In many cases, the levels constrain one another. A change in context may alter behaviour. Repeated behaviour may alter skill. Bodily states may change what actions are possible. Social relationships may affect what a person attends to, expects or fears.
What follows is not a ladder but a network of dependencies.
This also complicates where change should be located.
Imagine that a person who was previously silent in meetings begins contributing regularly after joining a team in which disagreement is welcomed rather than punished. The behavioural difference is straightforward. But where, exactly, is the change?
Perhaps the person has become more confident. Perhaps their expectations about social consequences have changed. Perhaps nothing durable in the person has changed at all and the new environment simply makes an existing capacity easier to express. Perhaps several of these are true.
The observation alone cannot settle the issue.
This is why stable behaviour does not necessarily imply an unchanged person, and changed behaviour does not necessarily imply a stable change within the person.
A person may maintain the same outward performance under increasingly demanding circumstances because skill, regulation or support has changed. Another may behave differently because the environment has changed while many personal dispositions remain much the same. What appears as stability or change at one level can coexist with something different at another.
That does not make the concept hopelessly relative. It makes specification essential.
A claim about human change becomes more informative when it identifies at least three things:
what property, pattern, capacity or relation differs;
the level at which that difference is being described;
and the comparison across time that makes the difference visible.
These are not yet criteria for deciding whether change is genuine or lasting. Nor do they explain how the change occurred. They simply keep separate questions from collapsing into one another.
This distinction is especially important when the phrase the person changed is used as if it referred to a single global event.
In most scientific contexts, what can actually be established are more specific changes: in behaviour, performance, reported experience, physiological response, social interaction or some other defined property. A broad judgement about the person may be justified by the accumulation of several such changes, but the global description remains shorthand.
That shorthand is useful in ordinary life because people are experienced as wholes.
Science has to be more patient.
It has to ask which part of the whole changed, how that difference is being described and what sort of claim follows from it.
Only then can the larger question of human transformation become tractable.
Defining the Object Before Explaining the Process
We can now return to the question that seemed simple at the beginning.
What does it mean for a person to change?
A useful starting definition is this:
Human change is a difference across time in some specified property, capacity, pattern or relation involving a person. Different forms of change can occur at different levels of organisation and need not coincide.
This definition is intentionally minimal.
It does not tell us whether the change is good or bad. It does not tell us whether it was chosen, learned, imposed, adaptive, pathological or accidental. It does not tell us whether it will persist. It does not tell us what mechanism produced it.
It tells us what kind of claim we are making when we say that change has occurred.
That clarification matters because the ordinary phrase a person changed can conceal several distinct possibilities.
Perhaps behaviour changed but subjective experience did not. Perhaps beliefs changed while behaviour remained stable. Perhaps physiological responses changed without obvious conscious awareness. Perhaps the person's interaction with a particular environment changed while their behaviour elsewhere remained the same. Perhaps several dimensions shifted together strongly enough that the global description feels appropriate.
The phrase is not wrong.
It is simply less precise than the phenomena it gathers together.
This is also why no single indicator can function as a universal measure of human change. Behaviour may be the relevant object in one question, subjective experience in another, biological function in another. Some changes are comparatively observable; others depend heavily on inference. A difference established at one level may provide evidence relevant to another, but it does not automatically establish it.
The result is a definition broad enough to accommodate the plurality of human change without treating every difference as equivalent.
That distinction matters because difference itself is everywhere. Human organisms are never literally static. Bodies fluctuate, attention shifts, neural activity changes, environments evolve. A useful science of human change cannot simply ask whether something differs from one moment to the next. It must specify which difference matters for the question under investigation.
This is where the next problems begin.
How persistent must a difference be before it deserves to be described as more than momentary variation? When does a temporary state become a recurring pattern? What distinguishes adaptation from transformation? What should count as evidence that a change is genuine rather than measurement noise or circumstance? If a new behaviour appears and later disappears, did change occur and reverse, or was the original change never established in the first place?
Those are not semantic puzzles.
They are different scientific questions.
They cannot be answered merely by making the definition of change broader. They require distinctions concerning time, stability, context, measurement and causal process. The purpose of this first essay is to establish enough conceptual order for those questions to be asked without confusion.
The central point is therefore not that people change in many ways. That much is obvious.
The more important point is that claims about change are only as precise as the dimension they specify.
To say that behaviour changed is one claim. To say that emotion changed is another. To say that neural organisation changed is another still. These claims may eventually form part of the same explanation, but they should not be treated as interchangeable simply because they concern the same person.
This matters most when we turn from description to explanation.
If we do not know what changed, we cannot know what mechanism needs explaining. If we do not distinguish behaviour from experience, we may mistake one for evidence of the other. If we collapse levels of organisation, we may confuse correlation with explanation or treat one form of description as if it rendered the others unnecessary. If we treat every temporal difference as equivalent, we lose the ability to distinguish momentary fluctuation from more consequential forms of change.
A coherent account of lasting human change therefore has to begin with disciplined modesty.
Not How do people change?
First:
What changed? In what respect? At what level? Relative to when?
Only after those questions are clear can the Library move on to the harder ones: what kinds of change should be distinguished, how stability and recurrence should be understood, what evidence justifies claims of change, and how any of these differences become durable.
The person remains one person.
The changes that can occur within, through and around that person are many.
Any explanation of lasting human change has to begin by keeping those two truths in view.