It is possible to understand a pattern in considerable depth and still find that the pattern continues. A person may recognise why a particular situation affects them, identify assumptions that shape their interpretation, understand something about the history that contributed to a recurring response, and describe that response with considerable accuracy. Such understanding can be consequential. It may reduce confusion, expose relationships that were previously difficult to see, improve anticipation, and sometimes contribute to substantial behavioural change. Yet the presence of understanding does not, by itself, establish that functioning has changed in the same way.
This is where discussions of insight often become unnecessarily polarised. One position assumes that once something has been understood, meaningful change should follow. Another dismisses intellectual understanding as superficial until it becomes somehow “embodied”. Both positions compress several different questions into one. Insight can matter considerably without necessarily producing every other form of change, and a change in behaviour does not automatically tell us how or why that change occurred. A more precise discussion therefore begins by distinguishing what has changed in understanding from what has changed in functioning.
Research offers good reason to take insight seriously. A systematic review and meta-analysis by Jennissen and colleagues found a moderate association between insight and psychotherapy outcome across a range of treatment approaches. The authors also noted important differences in how insight had been conceptualised and measured, and an association between insight and outcome does not establish that insight alone is responsible for improvement. The finding nevertheless challenges the idea that insight is merely an intellectual side-effect of more important processes.
The more useful question is therefore not whether insight “works”. It is whether insight and integration describe the same kind of change. In this article, insight refers broadly to a change in what has become known, recognised or understood, while integration refers to the coordination of relevant aspects of functioning rather than simply to the possession of that knowledge. The distinction is analytical rather than diagnostic, and it does not assume that integration is a single mechanism, a final state, or something that can currently be measured through one universal criterion.
Insight changes what is understood
Insight is not a single, universally agreed psychological construct. Its meaning varies across theoretical traditions and research contexts. Within psychotherapy research, it can include recognition of relationships between past and present experience, awareness of recurring interpersonal patterns, and greater understanding of relationships between emotional experience, behaviour and psychological difficulties. Related concepts elsewhere in psychology may involve metacognition, self-knowledge, cognitive reappraisal or recognition of behavioural contingencies.
What these different uses share is an epistemic dimension: something has become available to understanding that was previously absent, unclear or organised differently. A person who previously experienced a recurring response as inexplicable may begin to recognise the circumstances under which it appears. Someone who interpreted a difficulty as evidence of an unchanging personal characteristic may discover that it varies markedly by context. A repeated interpersonal difficulty may become recognisable as a pattern rather than a sequence of unrelated events.
These changes can matter because representation affects what can be noticed, anticipated and considered. Better understanding may alter the meaning assigned to an experience, make alternative interpretations possible, improve decision-making, or enable someone to recognise a response earlier than before. In some circumstances, that may be enough to produce significant behavioural change without any lengthy additional process.
However, insight remains evidence about what has become understood. It does not automatically establish whether an alternative response is readily accessible under the conditions in which the familiar response occurs, whether relationships supporting that response have changed, whether new learning transfers across situations, or whether multiple capacities are now functioning together differently. Those are further questions, and treating them as already answered because insight has occurred makes it harder to understand what happens when knowledge and behaviour diverge.
Knowing, intending and doing are related but not interchangeable
Research on the relationship between intention and behaviour provides a useful illustration of this distinction. It is not research on integration itself, and it should not be treated as direct validation of any particular definition of integration. Its value is narrower: it demonstrates that what people consciously intend and what they subsequently do are related without being equivalent.
Webb and Sheeran examined 47 experimental tests in which interventions successfully changed participants’ behavioural intentions. Across those studies, the change in intention was medium to large, while the resulting change in behaviour was smaller. The conclusion is more nuanced than either “intentions determine behaviour” or “intentions do not matter”. Changing intention can influence subsequent behaviour, but a change in conscious intention does not translate automatically or proportionately into enacted change.
This distinction becomes important when thinking about familiar human experiences. A person may know how they would prefer to respond during disagreement, understand why a different response would be beneficial, and sincerely intend to behave differently. Yet when the relevant interaction occurs, the familiar response may still become more readily available. Someone may understand why a repeated behaviour is costly while continuing to enact it in the contexts in which it has become established.
The persistence of the behaviour does not necessarily mean that the person has failed to understand the issue properly. It may instead indicate that understanding is only one of several influences relevant to what occurs. Human behaviour is affected by learning history, contextual cues, competing goals, attention, environmental structure, interpersonal consequences and other processes whose importance varies according to the phenomenon being examined.
This is why “I know this, so why do I still do it?” is a more complex question than it initially appears. The presence of knowledge changes one part of the problem. It does not necessarily settle every question about what becomes possible in the moment.
An understood alternative is not necessarily an accessible alternative
There is an important difference between knowing that an alternative response exists and being able to access that response reliably under the circumstances in which it is required. A person may describe an alternative clearly during reflection but find it much harder to enact during a fast-moving interaction. They may demonstrate a skill effectively when demands are manageable and struggle to access the same skill when several demands compete simultaneously. They may function differently in one relationship, role or environment than in another.
These variations are sometimes interpreted too quickly. A person may be described as resistant, unmotivated or unwilling to apply what they know. In other settings, the discrepancy may be attributed immediately to trauma, the body or the nervous system. Any of these factors could be relevant in a particular case, but the presence of a gap between understanding and behaviour does not identify the explanation by itself.
A more careful approach begins with description. Under which conditions is the alternative available? Under which conditions does the familiar response recur? What changes between those contexts? Is the difference associated with attentional demand, repetition, social consequences, environmental structure, emotional intensity, familiarity or some other factor?
This way of approaching the problem is consistent with the broader argument developed in What Do We Mean by Human Organisation?. Complex human functioning is more clearly understood when description, level of analysis and explanation are not collapsed into one another. The same principle applies here. Variability in access is an observation that requires explanation rather than an explanation in itself.
Habit shows how understanding and repetition can coexist
Habit research provides one well-studied example of why explicit understanding may coexist with repeated behaviour. Wood and Rünger describe habits as responses that become associated with recurring contexts through repetition. Once established, familiar cues can support well-practised responses alongside deliberate goal pursuit rather than simply disappearing when a person forms a different intention.
This makes it possible to understand how someone may recognise a pattern, prefer an alternative and still repeat the familiar behaviour without assuming that the person is irrational. Knowledge may have changed while the contextual relationships supporting the established response remain influential. The person can genuinely understand the pattern while the conditions under which the pattern is enacted remain largely intact.
The habit literature should not be stretched beyond what it can support. Not every persistent emotional, interpersonal or behavioural pattern is a habit, and complex human difficulties should not be reduced indiscriminately to automatic cue–response relationships. The relevance of the research lies in the broader principle it demonstrates: explicit knowledge is one influence on behaviour, but it is not the only influence.
This distinction is especially useful because it avoids replacing one oversimplification with another. There is no need to claim that insight is powerless, and there is equally little reason to assume that awareness should automatically dissolve a well-established response.
New learning also has to generalise
Another distinction concerns the difference between acquiring something under one set of conditions and being able to express it across others. Learning demonstrated in a particular setting does not necessarily transfer perfectly to every environment in which the relevant knowledge or behaviour might later be required.
Raviv, Lupyan and Green reviewed research on variability in learning and generalisation and highlighted an important tension: variability can sometimes make initial acquisition more difficult while supporting broader generalisation later. Their review spans several areas of learning rather than the specific problem of psychological insight, but it reinforces the importance of distinguishing what has been learned from how widely that learning transfers.
This distinction has direct conceptual relevance to insight. A person may arrive at an important understanding while reading, reflecting, writing or speaking in a supportive environment. They may also demonstrate an alternative response under similar conditions. What happens when the circumstances change remains a separate empirical question. Different social pressures, cues, consequences or competing demands may affect which responses become available and how learning is expressed.
That does not mean genuine change should produce identical behaviour everywhere. Human functioning is naturally context-sensitive, and an expectation of complete consistency would be unrealistic. The important distinction is between acquisition and generalisation. Evidence that a person can think or behave differently under one set of circumstances demonstrates something meaningful, but it does not necessarily tell us how robustly that difference will extend to every relevant context.
This is more precise than saying that an insight has not yet “gone deep enough”. Depth is difficult to define and harder to test. Questions about context, accessibility and generalisation can be investigated much more clearly.
Integration is not simply another word for embodiment
The language of embodiment often appears in discussions of this problem. Someone may say that they understand something “in their head” but have not yet “embodied” it, or that they know something intellectually without knowing it at a deeper level.
Such language can describe a recognisable subjective experience. It may be attempting to express the discrepancy between being able to articulate something conceptually and responding differently when a relevant situation actually occurs. The difficulty arises when metaphor is treated as explanation.
Saying that knowledge has not moved “from the head into the body” can imply a division between cognition and bodily functioning that is far less straightforward than the language suggests. Similarly, attributing the discrepancy automatically to “the nervous system” may imply a specific physiological mechanism without identifying or demonstrating what that mechanism is. Physiological processes may be highly relevant to some questions, but behaviour, attention, environmental structure, learning, social contingencies or other influences may be more informative in others.
A more precise use of integration avoids deciding the mechanism in advance. Here, the term refers to the extent to which the aspects of functioning relevant to a particular task or response operate in a coordinated way. Depending on the phenomenon, those aspects might include knowledge, attention, behavioural repertoire, emotional response, environmental conditions, interpersonal expectations or physiological processes. There is no requirement that every analysis include the same components.
This bounded definition also prevents integration from becoming vague wellness language. It does not mean feeling whole, reaching an ideal state, becoming permanently regulated, or transforming an insight into an undefined bodily truth. It asks a narrower question about coordination within the function being examined.
Integration is not simply behaviour change
If integration concerns coordination, it should not be inferred merely because different behaviour appears. Behaviour can change for many reasons, and similar observable outcomes can arise through different processes.
A person may respond differently because the environment has changed, because another person is present, because the consequences associated with a behaviour have altered, or because substantial conscious effort is temporarily supporting an alternative response. External structure may make a behaviour possible under conditions in which it would otherwise be difficult. Repetition may gradually make the new behaviour more familiar. A different action may therefore be meaningful without telling us precisely what has changed in the broader organisation of functioning.
The reverse is also important. Variability does not necessarily mean that integration has failed. A response may be broadly reliable while still changing under fatigue, high demand, unfamiliar circumstances or unusually strong competing pressures. Human functioning remains responsive to context even when learning is well established.
For this reason, integration should not be treated as a binary state in which a person either “has integrated” something or has not. Such language encourages false certainty and can easily turn a complex pattern of functioning into a label. A better question is how different aspects of functioning interact across the conditions that actually matter.
Integration should not be confused with durability or structural change
A further problem arises when integration is treated as though it automatically means that a deeper or permanent transformation has occurred. Coordinated functioning, durability and structural change are related questions, but they should not be collapsed.
A new response may function well while still requiring substantial external support. A person may develop an effective compensatory strategy without altering every process that contributed to the earlier pattern. Environmental change may produce substantial improvement while leaving other aspects of functioning relatively unchanged. Equally, a meaningful shift may become progressively more stable without ever producing identical behaviour across all conditions.
This means that improvement alone does not tell us whether a change is temporary, context-dependent, compensatory, increasingly established or reflective of something more fundamental. Each description requires different evidence. Treating them all as “integration” would make the term so broad that it loses analytical usefulness.
Durability therefore deserves separate attention. A response that becomes more reliable over time, generalises appropriately across relevant contexts and requires less exceptional support provides different evidence from a response that appears only under narrow conditions. These differences are important, but they should be described rather than converted prematurely into claims about a hidden internal state.
What would count as evidence of integration?
The concept becomes more useful when it is treated as a question that requires evidence rather than as a label applied retrospectively. There is currently no universal scientific measure by which an individual can simply be declared “integrated”, and different fields use the term in substantially different ways. Any strong claim that integration has occurred therefore needs to specify what is actually being observed.
For a functional definition, evidence would need to concern coordination rather than the mere presence of individual capacities. Knowing that a person understands an alternative does not establish how that understanding interacts with attention, behaviour or context. Knowing that someone possesses several relevant skills does not tell us whether those skills work together effectively under the conditions in which they are needed.
Reliability across relevant circumstances may also be informative. A response demonstrated once under unusually favourable conditions provides different information from one that becomes increasingly accessible across the range of situations in which it matters. Generalisation is relevant when the desired function requires transfer, while dependence on exceptional effort, intensive monitoring or unusual external support may describe a different form of organisation from functioning that has become less dependent on those conditions.
None of these observations should be turned into informal diagnostic criteria. They are examples of the kinds of evidence that a rigorous discussion of integration might eventually need to consider. Their importance lies in preventing the concept from becoming circular. If every improvement counts as integration and every difficulty is explained as incomplete integration, the concept cannot meaningfully be questioned or tested.
A useful analytical concept has to permit uncertainty and alternative explanations.
Sometimes insight really is sufficient
An important qualification follows from the same reasoning: not every change process requires something beyond insight. Sometimes the primary problem is informational or representational, and better understanding is enough to alter behaviour.
A person may misunderstand what is expected of them. They may not recognise an available option, may be operating from an inaccurate assumption, or may simply not know a more effective strategy. Once the relevant information becomes available, behaviour may change rapidly. An interpersonal misunderstanding may resolve through clarification, a mistaken belief may change when stronger evidence appears, or a person may adopt a new method immediately because the main barrier was not knowing that it existed.
In such cases, insisting that an additional process of “integration” must occur would add complexity without evidence. The psychotherapy literature also cautions against dismissing insight too readily, given the observed association between insight and outcome.
The more defensible position is conditional. When the difficulty is primarily one of knowledge, new understanding may be sufficient. When understanding changes but functioning remains relatively stable, inconsistent or highly context-dependent, further questions become necessary. The purpose of analysis is to determine which situation is present rather than assuming that insight must always be either sufficient or insufficient.
Why insight still matters when behaviour has not changed
Even when insight does not immediately produce a visible behavioural change, its value should not be judged solely by whether an outcome follows. Better understanding can improve the quality of the next question.
A person may discover that a pattern occurs only in particular relationships rather than everywhere. They may recognise that a response is readily available under manageable demands but becomes difficult to access when demands accumulate. They may notice that a behaviour appears in highly familiar contexts but not in others, or that an alternative response is possible when certain environmental supports are present.
These observations make the problem more precise. Instead of remaining with the broad statement “I know better, but I still do it”, it becomes possible to ask what differs between the circumstances in which the alternative response is available and those in which it is not. The enquiry can move towards learning, context, accessibility, reinforcement, competing demands or other processes supported by evidence.
This matters because the gap between knowledge and action is easily moralised. Continued repetition can be interpreted as weakness, lack of commitment or resistance to change simply because the person already understands the issue. Yet none of those conclusions follows automatically. The coexistence of insight and persistence demonstrates only that a change in understanding has not answered every question relevant to functioning.
The trauma-specific question is related but narrower
The difference between understanding and functioning can be particularly visible in trauma-related experiences, but trauma should not become the general explanation for the distinction itself. The earlier article Why Insight Alone Does Not Heal Trauma examines the narrower trauma-specific question through concepts such as conditioned responses, hypervigilance and the persistence of responses after danger has passed.
This article addresses a broader conceptual issue. Research on intention, habit and generalisation shows that knowledge and enacted functioning can diverge across many forms of human behaviour without trauma being the underlying explanation. Keeping those questions separate allows the trauma article to remain specific to its own evidence base while avoiding the implication that every persistent pattern must ultimately be understood through trauma or nervous-system explanations.
The broader issue is therefore not whether insight has become sufficiently “embodied”. It is whether changes in understanding and changes in functioning should be treated as the same phenomenon. The evidence gives little reason to assume that they should.
From understanding differently to functioning differently
The familiar question “I understand this, so why do I still do it?” contains an assumption that often goes unnoticed. It assumes that once the relevant knowledge is present, whatever remains should already have changed. When that does not happen, the discrepancy can seem irrational or personally deficient.
A more precise approach removes that assumption and examines the conditions under which functioning actually occurs. The established response may be strongly associated with a recurring context. The alternative may be understood but not equally accessible under all conditions. New learning may be available in one environment without generalising fully to another. A familiar behaviour may still be efficiently supported by repetition, environmental structure or interpersonal consequences. None of these possibilities should be assumed in advance, but each provides a more useful direction for enquiry than simply concluding that the person has failed to understand deeply enough.
This is where the distinction between insight and integration becomes valuable. Insight can alter what is recognised, represented and understood, and that change may itself be significant. Integration concerns a different question: whether the aspects of functioning relevant to the situation are coordinated in a way that allows that understanding to participate effectively in what actually happens.
Sometimes those changes occur almost together, and insight proves sufficient. In other circumstances, knowledge develops considerably before functioning changes to the same extent. Neither outcome diminishes the importance of insight, and neither requires a vague appeal to a deeper hidden process.
Understanding a pattern and functioning differently within it are related achievements, but they are not identical ones. Recognising that distinction makes it possible to examine human change with greater precision, without reducing it either to conscious understanding or to an unspecified process beyond it.
Selected research
Jennissen, S., Huber, J., Ehrenthal, J. C., Schauenburg, H., & Dinger, U. (2018). Association Between Insight and Outcome of Psychotherapy: Systematic Review and Meta-Analysis. American Journal of Psychiatry, 175(10), 961–969.
Webb, T. L., & Sheeran, P. (2006). Does Changing Behavioral Intentions Engender Behavior Change? A Meta-Analysis of the Experimental Evidence. Psychological Bulletin, 132(2), 249–268.
Wood, W., & Rünger, D. (2016). Psychology of Habit. Annual Review of Psychology, 67, 289–314.
Raviv, L., Lupyan, G., & Green, S. C. (2022). How Variability Shapes Learning and Generalization. Trends in Cognitive Sciences, 26(6), 462–483.
Metamorphosis Wellness 
Leave a comment