Abstract
Intention, which MeSH classifies under motivation, is what a person has in mind to do or bring about. Intentions differ in reach, from plans settled days in advance to the moment-by-moment guidance of a movement. In social and health psychology, stated intentions predict later behaviour moderately well, yet many people who intend to act do not follow through, a shortfall known as the intention-behaviour gap, and if-then plans that tie an action to a cue help to close it. Neuroscience has asked whether brain activity that precedes a voluntary movement shows conscious will arriving too late to matter, a reading that is now disputed. Developmental research shows that infants and toddlers read the goals behind other people's actions.
Keywords: intention, implementation intentions, intention-behaviour gap, readiness potential, sense of agency
A student who resolves on Sunday to revise for an hour every evening, a patient who plans to book a screening appointment, and a driver whose foot is already moving toward the brake all have something in common: each holds a state that points toward a particular action. The student's plan reaches days ahead, the patient's waits on an opportunity, and the driver's is fused with the movement itself. Psychology studies all three under the heading of intention, and the concept links the study of volition to the study of social behaviour, memory and brain function.
Intending is not the same as wanting. A person can want many incompatible things at once, but an intention selects one of them and commits the person to pursuing it. This treatment of intentions as commitments that structure plans and practical reasoning follows the planning account that Bratman developed in his 1987 book (Bratman, 1987). This article first sets out what intention is and how its levels are distinguished, then follows four research traditions: models that predict behaviour from intentions and the gap that opens between the two, if-then planning, the neuroscience of voluntary action and the debate over Libet's timing experiment, and the sense of agency. It closes with how intentions are held in memory and how even young children recognise the intentions of others.
Key Takeaways
- The theory of planned behavior, which treats intention as the nearest predictor of action, accounts for a moderate share of behaviour in meta-analyses, and its fit is flattered when behaviour is self-reported.
- The intention-behaviour gap, the shortfall between what people intend and what they do, is wide; in physical activity most of the mismatch comes from people who intend to act and then do not.
- Experiments that changed intentions produced a smaller change in behaviour (d = 0.36) than in intention itself (d = 0.66), across 47 experimental tests.
- Implementation intentions, if-then plans that tie a planned action to a specific situation, raise goal completion, and combining them with mental contrasting has a small-to-medium effect on goal attainment, with possible publication bias.
- The readiness potential, a slow electrical build-up recorded before a voluntary movement, began before the reported urge to move in Libet's small study of 5 subjects, but whether it reflects a decision at all is contested.
- Intentional binding, the perceived drawing together in time of a voluntary action and its effect, is widely used as a measure of the sense of agency, though the link between the two is not fully explained.
What Intention Is
What a person has in mind to do or bring about.
Scope note for the MeSH descriptor Intention (D033182), U.S. National Library of Medicine.
An intention is a mental state that commits its holder to an action. Where a desire merely registers that an outcome would be welcome, an intention settles what the person is going to try to do and keeps that plan in force until the time to act arrives. The commitment is what makes intentions useful for coordination: other people, and the person's own later self, can rely on a formed intention in a way they cannot rely on a passing wish.
People also apply the concept to one another, judging every day whether an act was done on purpose. That everyday judgement is a research topic in its own right, and Malle and Knobe (1997) studied the folk concept of intentionality directly. Knobe (2003) then posed a sharper puzzle with a vignette about a company chairman who approves a plan while foreseeing a side effect that he does not care about, and asked whether the chairman brings that side effect about intentionally. The case is instructive because it pulls apart two things that usually travel together: aiming at an outcome and knowingly causing it.
In the MeSH vocabulary, intention appears twice. It sits under motivation, as one of the states that energise and direct behaviour, and under mental processes, as one of the operations of the mind. The two placements reflect the two faces of the concept: an intention is both a reason for acting and a representation that has to be formed, stored and retrieved.
Levels of Intention
Intentions are not all of one kind. Pacherie (2008) proposed a three-tiered model of intention as a framework for the phenomenology of action, that is, for how acting is experienced from the inside, and tied its levels to internal-model accounts of how actions are specified and controlled. Pacherie presented the three-tiered model as a conceptual framework, and the paper reports no empirical test of it (Pacherie, 2008). The model has since been revised. Mylopoulos and Pacherie revisited the model, which they call the dynamic hierarchical or DPM model, refining its account of motor representations and of how the levels interact, and restated its core insight that action control results from coordinated activity across distal, proximal and motor intentions (Mylopoulos & Pacherie, 2019). In the vocabulary that grew up around such schemes, a distal intention is a plan formed ahead of time, a proximal intention is the intention to act now, and a motor intention is the fine-grained specification of the movement that carries the action out.
The lowest level has the firmest empirical anchor. Reviewing work on movement, Jeannerod (1994) argued that motor imagery and motor preparation are closely equivalent in function, so that imagining an action can be used to study how it is planned and prepared. He offered this equivalence as a suggestion supported by converging evidence rather than as a demonstrated identity (Jeannerod, 1994).
Figure 1
Three Levels of Intention, From a Plan to a Movement
Intentions and Behaviour
The most widely tested account of how intentions turn into action comes from social psychology. The theory of reasoned action grew out of the programme that Fishbein and Ajzen (1975) presented in a book-length introduction to theory and research on belief, attitude, intention and behaviour. That theory has since been revised: Ajzen (1991) extended it by adding perceived behavioural control, producing the theory of planned behavior, and his later overview discusses how the two theories differ (Ajzen, 2020).
In the theory of planned behavior, intention is the closest predictor of what a person does. Intention in turn depends on three judgements: how much control the person believes they have over the behaviour, whether they think the people who matter to them expect it, and how favourably they evaluate doing it, the last being the person's attitude toward the behaviour. Perceived behavioural control can also predict behaviour alongside intention (Ajzen, 1991). Ajzen's 1991 review judged the theory well supported but listed several issues as unresolved: the expected interaction between intention and perceived control had only weak support, perceived control added little when people's sense of control was unrealistic, and whether the predictors were sufficient remained an open question (Ajzen, 1991).
The theory of planned behavior is contested. Its supporters can point to meta-analyses showing that it predicts a moderate share of what people do. Armitage and Conner (2001) pooled 185 independent studies published up to 1997 and found that the model accounted for 27% of the variance in behaviour and 39% of the variance in intention. They also noted that self-reported behaviour inflated the variance explained, and that the subjective norm, the perceived social pressure to act, was a weak predictor of intentions, partly because of poor measurement. A later meta-analysis restricted to prospective tests, 237 of them drawn from 206 articles, found that prediction of health behaviour depended on the type of behaviour, and that self-reported outcomes and shorter follow-up periods were predicted better than objective outcomes and longer intervals (McEachan et al., 2011). In a preregistered meta-analysis of 39 studies (N = 13,121), perceived behavioural control moderated the path from intention to behaviour but left the paths from attitude and from subjective norm to intention unchanged, and much of the variation between studies remained unexplained (Hagger et al., 2022). Critics have gone further: Sniehotta and colleagues titled their 2014 paper a call to retire the theory (Sniehotta et al., 2014). Ajzen responded directly in a commentary contending that the theory was still alive and well and not yet due for retirement (Ajzen, 2015). Ajzen (2020) later published an overview organised around frequently asked questions about the theory, among them the intention-behaviour gap.
The Intention-Behaviour Gap
Intentions predict behaviour, but imperfectly, and the shortfall has become a research field of its own. Sheeran (2002) combined existing meta-analyses into a meta-analysis of meta-analyses and concluded that a real gap separates intentions from behaviour. His review also argued that the width of the gap depends on moderators, grouped as features of the behaviour, of the intention and of the person, and it considered what habits imply for the reach of the intention concept.
Correlations cannot show that intentions cause behaviour, which is why experimental evidence matters. Webb and Sheeran (2006) gathered 47 experimental tests in which an intervention changed people's intentions and then measured what they did. Intentions shifted by d = 0.66, but behaviour shifted by only d = 0.36, a small-to-medium effect. Changing what people intend does change what they do, then, but by less than the change in intention would suggest.
A useful way to see the gap is to cross intention with action. Orbell and Sheeran (1998) followed 166 women who had never been screened for cervical cancer over a one-year period, checking attendance against medical records. Protection motivation theory predicted who intended to be screened, but no discriminant function separated the inclined actors, who intended to attend and did, from the inclined abstainers, who intended to attend and did not. The theory, they noted, says nothing about the processes that carry an intention into action (Orbell & Sheeran, 1998).
The action control framework turns that four-way split into a measurement tool (Rhodes & de Bruijn, 2013). Applying it to physical activity, Rhodes and de Bruijn (2013) pooled 10 studies with a combined N = 3,899 in a systematic review and meta-analysis. The gap was large, and most of the mismatch between intention and behaviour came from people who intended to be active and then were not; the authors also judged that early intention models had performed weakly for physical activity. A later systematic review of 129 studies found that how wide the gap between intention and physical activity was depended chiefly on the motivational character of the intention and of the person holding it, though most of the studies were judged to be of moderate quality (Rhodes et al., 2022). Rhodes and de Bruijn (2013) used the action control framework to quantify where the gap arises.
Demo 1: how many intenders actually act?
Illustrative only: these are invented proportions for an imaginary group of 100, not data from any study. The four cells follow the way researchers sort people by intention and action.
Some behaviour bypasses fresh intentions altogether. In a meta-analysis paired with a primary study, Ouellette and Wood (1998) found that past behaviour predicts future behaviour by two routes: through intentions, and through habits that run with little deliberation. The strength of the intention matters as well. Reviewing the evidence, Conner and Norman (2022) concluded that strong intentions predict behaviour better and are more stable over time than weak ones. They cautioned that moderators of the intention-behaviour relation have been little studied, that findings on them are mixed, that more tests of stability as a mediator are needed, and that the usual assumption of a linear relation is doubtful (Conner & Norman, 2022). A broader review reaches the same overall verdict: not every intention is carried out (Sheeran & Webb, 2016).
Implementation Intentions
The Rubicon model divides goal pursuit at the point of commitment, separating a motivational state of mind, in which goals are weighed, from a volitional one, in which a chosen goal is carried out. Heckhausen and Gollwitzer (1987) contrasted thought contents and cognitive functioning in these motivational and volitional states. The model is still in use: a 2018 chapter in an edited volume restated how the course of an action divides between motivation and volition (Achtziger & Gollwitzer, 2018). The if-then planning research that followed was reviewed by Bieleke and colleagues after almost 30 years of work (Bieleke et al., 2021).
Within that programme, Gollwitzer (1999) described implementation intentions as if-then plans that tie an expected situation to a response serving the goal. His review argued that such plans improve the attainment of goals by linking a situational cue to a planned response, so that the response can be triggered with some automaticity when the cue appears. The review drew on a programme of studies showing that such plans help people reach their goals, and it examined how they work (Gollwitzer, 1999).
Demo 2: turn a goal into an if-then plan
Menus written for this article. The demo shows the structure of an if-then plan; it makes no claim about how well any particular plan will work.
Early evidence came from one correlational study and two experiments. People who had formed implementation intentions completed their goals at a higher rate, about 3 times as often in one comparison, although the first study was correlational only (Gollwitzer & Brandstätter, 1997). Gollwitzer and Sheeran (2006) later published a meta-analysis of the effects and processes of implementation intentions.
The technique has also been combined with mental contrasting, in which a person sets a desired future against the obstacles in the way before planning how to meet them. A meta-analysis of 24 independent effect sizes from 21 articles, covering 15,907 participants, found that mental contrasting with implementation intentions improved goal attainment with a small-to-medium effect, g = 0.336, 95% CI (0.229, 0.443) (Wang et al., 2021). The authors flagged possible publication bias and few included studies, and only the type of intervention moderated the effect; age, domain and outcome measure did not (Wang et al., 2021). The strategy sits within the wider study of self-regulation and self-control. Bieleke et al.'s review of if-then planning extends the approach to flexible goal striving and physical endurance, while acknowledging many open conceptual questions and calling for work on individual differences (Bieleke et al., 2021).
Intention and the Brain
For much of its history, neuroscience set volition aside as unsuitable for study, and its mechanisms long remained unclear. Haggard (2008) reviewed the evidence that has since accumulated and described linked brain regions behind voluntary action, among them the parietal cortex, the front of the prefrontal cortex and the area just ahead of the supplementary motor area.
Several studies locate specific contributions. In a functional magnetic resonance imaging study of self-paced actions, attending to one's own intention to act enhanced activity in the pre-supplementary motor area, and prefrontal activity, unlike parietal activity, was more tightly linked to activity in that area (Lau et al., 2004). Direct stimulation points to the parietal lobe as well: in patients undergoing brain surgery to remove tumours, electrical stimulation of parietal sites evoked a conscious intention to move with no accompanying muscle activity, although the abstract does not report how many patients took part (Desmurget et al., 2009).
Intentions that have not yet been acted on can also be read from brain activity. Using pattern decoding of functional imaging data while people held a choice between two tasks over a delay, Haynes and colleagues decoded the covert intention from prefrontal cortex, most robustly from its medial part, while posterior prefrontal regions dominated during execution (Haynes et al., 2007). Recordings from single neurons in humans add timing. Across twelve subjects and 1019 neurons, cells in the supplementary motor area and nearby medial frontal cortex predicted a decision to make a self-initiated movement before the subjects reported being aware of it; 256 neurons in the supplementary motor area sufficed to predict single-trial decisions with over 80% accuracy 700 ms before reported awareness (Fried et al., 2011). Because the moment of decision was self-reported, the comparison inherits the uncertainty of that report.
The what, when, whether model organises this literature. Brass and Haggard (2008) proposed that an intentional action involves three components, deciding which action to perform, when to perform it and whether to perform it at all, and argued that treating intention as a single process had blurred an imaging literature that looked contradictory. The model is now partly supported: pooling 15 PET and fMRI studies, a meta-analysis found that its three components could be partly separated in the brain, although the separation also reached regions outside the medial frontal wall, unlike the original model, and in a wider imaging database the same separation came from tasks in which people responded to cues rather than acting on their own intentions (Zapparoli et al., 2017).
Intention also has a representation in the coding of others' acts. Recording from single neurons in the inferior parietal lobule of monkeys, Fogassi and colleagues found cells that coded the same motor act differently depending on the final goal of the action it belonged to (Fogassi et al., 2005). That such coding supports an observer's understanding of intention was an inference from these recordings rather than a direct test.
The Libet Debate
No experiment on intention has drawn more argument than Libet's. Libet and colleagues recorded the readiness potential, a slow negative shift in the electroencephalography signal that builds up before a voluntary movement, while 5 subjects made self-initiated movements across at least 6 sessions (Libet et al., 1983). Subjects watched a spot revolving on a clock face and afterwards reported its position at the moment they first became aware of wanting to move. In series where subjects reported no preplanning, the readiness potential began on average about 350 ms, and at least about 150 ms, before that reported awareness (Libet et al., 1983). The authors read the result as showing that a freely voluntary act is initiated unconsciously. Two limits apply: awareness could only be measured through what subjects recalled of the clock position, and the sample was tiny (Libet et al., 1983).
Demo 3: when did you decide?
A spot will circle the clock. Whenever you feel like it, press the button. Then report where the spot was when you first felt the urge to press.
An illustration of the clock method only. No brain signal is recorded or simulated, and a single trial says nothing about the timing of anyone's decisions.
Later studies refined the measures. Using Libet's method, Haggard and Eimer (1999) found that the lateralised readiness potential, the component that reflects preparation of a specific hand, tracked the timing of awareness, whereas the general readiness potential did not. They described the link as only suggestive: the readiness potential's tendency to start later did not reach significance, and freely chosen and fixed movements did not differ. With pattern decoding, Soon and colleagues predicted the outcome of a free choice from activity in prefrontal and parietal cortex up to 10 s before the subjects reported being aware of deciding, though the interpretation of that delay was tentative (Soon et al., 2008).
The accumulator model of the readiness potential offered a different reading. Combining a computational model with two experiments using electroencephalography, Schurger and colleagues proposed that the build-up reflects spontaneous, subthreshold fluctuations in neural activity that occasionally drift across a decision threshold, modelled as a leaky stochastic accumulator (Schurger et al., 2012). The model assumes only a weak imperative to move and was tested on a task with a delay of several seconds. The accumulator model of the readiness potential is contested but influential: Schurger and colleagues' 2021 review maintains that the potential reflects stochastic fluctuation rather than a decision, while conceding that whether the decision comes early or late remains unresolved (Schurger et al., 2021).
Two experiments probe the account from different sides. Using a brain-computer interface that detected readiness potentials in real time, Schultze-Kraft and colleagues found that people could still veto a self-initiated movement until roughly 200 ms before it began, a point of no return (Schultze-Kraft et al., 2016). The study analysed 10 of 12 recruited subjects and used foot movements only. Maoz and colleagues compared arbitrary choices with deliberate ones, in which subjects decided how to allocate a $1000 donation between organisations they cared about (Maoz et al., 2019). Among 18 subjects, a readiness potential appeared before arbitrary decisions but not before deliberate ones, and the authors read a Bayes factor of 0.332 for the deliberate condition as moderate evidence that the potential was absent, a reading one peer reviewer disputed as inconclusive (Maoz et al., 2019).
What the readiness potential actually is remains open. The 2021 review concludes that it is not established whether the potential exists on single trials, notes that it is weaker than ordinary EEG noise, and observes that imaging-based prediction of choices is only marginally above chance (Schurger et al., 2021). The debate therefore bears on the role of consciousness in action without having settled it.
Sense of Agency
The sense of agency is the experience of being the author of one's own actions and their effects. A widely used implicit measure comes from Haggard and colleagues, who found in psychophysics experiments that the perceived times of a voluntary action and of its sensory effect drew toward each other, an effect called intentional binding (Haggard et al., 2002). When the movement was produced involuntarily by magnetic brain stimulation, the effect reversed. A later review judged the case linking intentional binding to the sense of agency compelling, while noting that the link had not been fully explained and that open questions remained (Moore & Obhi, 2012). Binding may follow causation as well as intention: in a stimulus-anticipation task, people experienced a shorter interval before stimuli they had caused themselves than before equally predictable stimuli they had not caused, which the authors read as suggesting that causality itself produces temporal binding, beyond earlier claims of intentional binding (Buehner & Humphreys, 2009).
The comparator model of agency holds that the feeling of control arises from matching the predicted consequences of a movement against its actual consequences. The comparator model of agency has been revised. Synofzik and colleagues argued, from existing empirical studies, that a comparator alone cannot explain agency, pointing among other things to findings in schizophrenia that it fails to account for, and proposed a two-step account in which a weighting of multiple agency cues produces a low-level feeling of agency, which conceptual processing then turns into a higher-level judgement of agency (Synofzik et al., 2008). Haggard's later review describes the field as advancing, with the brain mechanisms of agency only recently studied and its measures still improving (Haggard, 2017).
The theory of apparent mental causation takes a more radical line. Wegner and Wheatley (1999) argued, in a theory paper supported by one experiment, that the experience of willing an act is an inference: people feel they caused an action when nothing else in view could explain it, when what happens matches what they had in mind, and when that thought arrived just beforehand. The theory of apparent mental causation is contested. Moore's overview sets it beside comparator accounts and a cue-integration view in which several signals are combined, so that inference becomes one cue among others, though he notes that cue integration still needs fuller examination (Moore, 2016). The same overview covers the main measures, including intentional binding and sensory attenuation, and judges applied work on agency to be still in its infancy (Moore, 2016).
Remembering Intentions
Many intentions must wait for the right moment, and remembering to carry them out is called prospective memory. Unlike retrospective memory, which supports the recall of past events, prospective memory concerns actions still to be performed. In two laboratory experiments comparing young and old adults, Einstein and McDaniel (1990) found no age deficit in prospective memory, even though the same older adults showed age differences on retrospective memory tests, and prospective and retrospective performance were unrelated in regression analyses. A later meta-analysis found that older adults performed worse on laboratory prospective-memory tasks, with a smaller shortfall when the task was event-based and relatively automatic, whereas in naturalistic settings the older participants came out ahead (Henry et al., 2004).
The multiprocess framework of prospective memory explains how intentions are retrieved. McDaniel and Einstein (2000) proposed that retrieval can rely either on strategic monitoring for the right moment or on relatively automatic processes in which a cue brings the intention to mind, and that the route used varies with the task, the cue, the ongoing activity and the individual. The multiprocess framework of prospective memory has since been revised: one of its authors and colleagues recast it as a dynamic framework, and in an experiment with 121 undergraduates, monitoring for the intention appeared to start only once the intention had been spontaneously retrieved, rather than coming before that retrieval (Scullin et al., 2013). The subgroups of correct responses were too small for some comparisons, and every participant was an undergraduate (Scullin et al., 2013).
Holding an intention has a price. In 3 experiments using a lexical decision task, Smith (2003) found in the first two that carrying an event-based prospective memory intention slowed performance on the ongoing task, and concluded that retrieving a delayed intention can demand capacity. Smith's work grew into a monitoring-only view, on which sustained preparatory monitoring is needed for an intention to be retrieved, and Scullin and colleagues reported results inconsistent with that view (Scullin et al., 2013). Smith and Bayen split event-based prospective remembering into a preparatory attentional component and a retrospective memory component for the intention, estimated the two as separate parameters of a multinomial model across four experiments, and also evaluated an alternative model (Smith & Bayen, 2004). Intentions also enjoy privileged access in memory. In 4 experiments, Goschke and Kuhl (1993) found that words from a script that participants intended to perform were recognised faster than other words, whereas words from a script they would only observe showed no such advantage over neutral words. This intention superiority effect persisted without rehearsal or imagery, which they took as a sign of persisting activation of intended actions.
Understanding Others' Intentions
Reading intentions in others begins early and underpins theory of mind. Using a visual habituation paradigm with infants aged 5, 6 and 9 months, Woodward (1998) found that infants encoded the goal object of a person's reach rather than the path of the arm, responding more strongly to a change of goal than to a change of path. The pattern was weaker at 5 months, and it did not appear when an inanimate object made the same movement, an early marker of infant behaviour attuned to agents.
By 18 months, children infer goals that an adult fails to achieve. In Meltzoff's (1995) re-enactment procedure, children who watched an adult try and fail to complete an act on an object went on to perform the intended act rather than copy the failed movements, and they did not do so for an inanimate device that went through similar motions. The groups in the first experiment were small, 10 children per group, one test object produced a weaker result, and the study could not decide whether the children represented goals or mental states.
Tomasello and colleagues built on such findings an argument about shared intentionality, the capacity to form joint intentions and joint attention with others, which they proposed distinguishes human cognition and makes cultural learning possible (Tomasello et al., 2005). On their account, great apes grasp intentional action but do not share intentions in this way. The monkey recordings described above (Fogassi et al., 2005) suggest one neural route by which an observer might register the goal of another's act. A critical review listed eight problems with the claim that mirror neurons are the basis of action understanding, observing that no data from monkeys had ever put the hypothesis to a direct test (Hickok, 2009).
Worked Example
The following example is illustrative. Its numbers are invented for this article and come from no study. Suppose 100 imaginary people are asked whether they intend to exercise three times next week, and their behaviour is recorded the following week.
Sixty say yes. Of those 60, 36 exercise as planned and 24 do not. Of the 40 who say no, 4 exercise anyway and 36 do not. Crossing intention with action yields four cells: 36 inclined actors, 24 inclined abstainers, 4 disinclined actors and 36 disinclined abstainers, which together account for all 100 people.
Seen one way, intention looks like a strong predictor. Of the 40 people who exercised, 36 had intended to, and 72 of the 100 people behaved consistently with what they said. Seen from the intenders' side, the picture changes: 24 of the 60 intenders, or 40%, did not act. The mismatch also falls unevenly across the two off-diagonal cells. Of the 28 people whose behaviour contradicted their stated intention, 24 were intenders who failed to act and only 4 were non-intenders who acted anyway, so about 86% of the mismatch sits in the inclined-abstainer cell.
That asymmetry explains why planning interventions target intenders. An if-then plan does not try to create motivation where there is none; it aims to move people from the inclined-abstainer cell into the inclined-actor cell by linking the intended exercise to a specific cue, such as a set time and place. If such a plan converted a third of the 24 inclined abstainers, 8 people would move cells, and the share of intenders failing to act would fall from 40% to about 27% (16 of 60).
Table 1
Illustrative Intentions and Actions for 100 Imaginary People Asked About Exercising Three Times Next Week
| Group | Acted | Did not act | Total |
|---|---|---|---|
| Intended to exercise | 36 (inclined actors) | 24 (inclined abstainers) | 60 |
| Did not intend | 4 (disinclined actors) | 36 (disinclined abstainers) | 40 |
| Total | 40 | 60 | 100 |
Note. This article's own invented numbers, not data. 24 of the 60 intenders (40%) did not act.
Discussion
Some conclusions about intention are secure. Intentions predict behaviour, and models built around them account for a moderate share of what people do, with the estimate inflated when behaviour is self-reported. Experiments confirm that changing intentions changes behaviour, but by a smaller margin than the change in intention, and a substantial share of intenders do not follow through. If-then planning addresses that failure directly and improves goal attainment, with the effect of its combination with mental contrasting estimated as small to medium and possibly inflated by publication bias. In the brain, voluntary action engages identifiable medial frontal, prefrontal and parietal networks, and intentions that have not yet been acted on can be decoded from them.
Other questions are open, and they cut across the sections above. The first is what the readiness potential reflects: a decision taken before awareness, or the accumulated noise against which a later decision is made. The deliberate-choice data, the veto data and the accumulator model each constrain the answer without fixing it. The second is how the sense of agency is computed, given that comparator signals alone are insufficient and the weighting of other cues has not been fully worked out. The third is generalisation. Much of the neuroscience rests on small samples, arbitrary movements and awareness timed by recalled clock positions, while the behavioural literature rests heavily on self-report; whether a lab timing task tells researchers much about consequential decisions, or whether a self-reported intention is the same state that the neuroscience measures, has not been settled. Studies that pair objective behavioural outcomes with deliberate, meaningful choices would bear on all three.
Current Directions
Recent behavioural work has moved from asking whether intentions predict behaviour to asking which intentions do. Reviewing the evidence on intention strength, Conner and Norman (2022) argued that strong, stable intentions predict better than weak ones, while noting that the relevant moderators remain little studied. Planning research has consolidated: Bieleke et al.'s review of almost 30 years of if-then planning identifies individual differences as a priority (Bieleke et al., 2021), and a meta-analysis of mental contrasting with implementation intentions found a small-to-medium benefit for goal attainment, with possible publication bias (Wang et al., 2021). The theory of planned behavior has been restated and defended by its originator in a question-and-answer overview (Ajzen, 2020), and the broader intention-behaviour gap has been reviewed as a problem in its own right (Sheeran & Webb, 2016).
In neuroscience, the readiness potential is being re-examined from several directions. Real-time brain-computer interfaces have mapped how late a movement can still be cancelled (Schultze-Kraft et al., 2016); comparisons of arbitrary and deliberate choices question whether findings from arbitrary movements extend to decisions that matter (Maoz et al., 2019); and theoretical work asks whether the potential reflects a decision at all (Schurger et al., 2021). Agency research is refining its measures and studying its brain mechanisms (Haggard, 2017) while beginning applied work that is still in an early phase (Moore, 2016).
Common Misconceptions
- A strong enough intention reliably produces the action.
- A meta-analysis of meta-analyses found a real gap between intentions and behaviour, whose width varies with features of the behaviour, of the intention and of the person (Sheeran, 2002). The belief persists because intentions do predict behaviour on average, which is easily mistaken for predicting it in every case.
- Changing what people intend changes what they do by the same amount.
- Across 47 experimental tests, interventions shifted intentions by d = 0.66 but behaviour by only d = 0.36 (Webb & Sheeran, 2006). Correlational studies, however strongly intention and behaviour are related in them, cannot show how much behaviour follows a change in intention.
- Libet showed that conscious will plays no role in action.
- A recent review argues that the readiness potential may reflect stochastic fluctuation rather than a decision, and that whether the decision comes early or late is unresolved (Schurger et al., 2021). The misreading treats a build-up of activity before reported awareness as proof that the decision was already made.
- The feeling of having acted comes from a direct readout of the motor command.
- A comparator of predicted and actual movement consequences cannot by itself explain agency; a two-step account derives a low-level feeling of agency from a weighting of several cues and then forms a judgement of agency from it (Synofzik et al., 2008). The belief draws on the comparator idea, which captures part of the process but not all of it.
Glossary
- Action control.
- A scheme that classifies people by crossing whether they intend to act with whether they act, so that the source of an intention-behaviour gap can be located; used as a descriptive tool.
- Distal intention.
- An intention formed in advance of the action it concerns, settling what will be done at some later time.
- Goal intention.
- A commitment to reach a desired end state, without specifying when, where or how the necessary actions will be performed.
- Implementation intention.
- An if-then plan that links a specific situation to a planned response, intended to make the response follow when the situation arises.
- Intention superiority effect.
- Faster recognition of material belonging to an action a person intends to carry out than of comparable material, taken as a sign that intentions stay active in memory.
- Intention-behaviour gap.
- The shortfall between what people intend to do and what they actually do; in physical activity it arises mostly from people who intend to act and then do not.
- Intentional binding.
- The perceived shortening of the interval between a voluntary action and its effect, used as an implicit measure of the sense of agency.
- Motor intention.
- The lowest level of intention, specifying the movement that will carry out an action.
- Perceived behavioural control.
- A person's belief about how far performing a behaviour is within their power; it predicts intention and can predict behaviour directly.
- Prospective memory.
- Remembering to carry out an intended action at the right moment in the future.
- Proximal intention.
- An intention to act now, which initiates and guides an action as it begins.
- Readiness potential.
- A slow negative build-up in scalp-recorded brain activity before a self-initiated movement, whose meaning for the timing of decisions is contested.
- Sense of agency.
- The experience of being the one who causes one's own actions and their effects.
- Shared intentionality.
- The capacity to form intentions and attention jointly with other people, argued to set human cognition apart.
- Subjective norm.
- A person's sense of whether important others expect them to perform a behaviour; a comparatively weak predictor of intentions in meta-analyses.
- Veto.
- The cancellation of a movement that has already begun to be prepared, possible up to a late point of no return.
- W judgment.
- A subject's report, usually read from a clock display after the fact, of when they first felt the urge or intention to move.
Key Researchers
Icek Ajzen
(University of Massachusetts Amherst). Social psychologist who proposed the theory of planned behavior, in which intention is the immediate predictor of action. Faculty page · Wikipedia · Wikidata
Peter M. Gollwitzer
(New York University; University of Konstanz). Psychologist who introduced implementation intentions and, with Heinz Heckhausen, the Rubicon model of action phases. ORCID · Wikipedia · Wikidata
Patrick Haggard
(University College London). Cognitive neuroscientist who described intentional binding and studies human volition and the sense of agency. ORCID · Wikipedia · Wikidata
Benjamin Libet
(University of California, San Francisco, 1916–2007). Neuroscientist whose timing experiments set the readiness potential against people's reported awareness of the urge to move. Wikipedia · Wikidata
Elisabeth Pacherie
(Institut Jean Nicod (CNRS / EHESS / ENS-PSL)). Philosopher of mind who distinguished future-directed, present-directed and motor intentions. ORCID · Wikidata
Aaron Schurger
(Chapman University). Cognitive neuroscientist who proposed that the readiness potential can reflect spontaneous neural fluctuations reaching a threshold. Faculty page
Paschal Sheeran
(University of North Carolina at Chapel Hill). Social and health psychologist whose meta-analyses measured the gap between what people intend and what they do. ORCID · Faculty page · Wikidata
Frequently Asked Questions
What is the difference between an intention and a desire?
A desire marks an outcome as welcome, while an intention commits a person to a course of action. Philosophers and psychologists also distinguish levels of intention, and one conceptual framework separates plans made in advance from intentions to act now and from the motor level that specifies the movement, though that paper offers no empirical test of it (Pacherie, 2008).
How well do intentions predict behaviour?
Moderately. A meta-analysis of 185 independent studies found that the theory of planned behavior accounted for 27% of the variance in behaviour and 39% of the variance in intention, with self-reported behaviour raising the share of behaviour explained (Armitage & Conner, 2001).
What is the intention-behaviour gap?
It is the shortfall between what people intend and what they do. In a meta-analysis of 10 studies of physical activity with N = 3,899, the gap was large and came mostly from people who intended to be active but were not (Rhodes & de Bruijn, 2013).
What is an implementation intention?
It is an if-then plan that names a situation and the response a person will make when it occurs. A review of the research programme argued that such plans improve goal attainment by letting a situational cue trigger the planned response (Gollwitzer, 1999).
Do if-then plans work when combined with mental contrasting?
A meta-analysis of 24 independent effect sizes covering 15,907 participants found a small-to-medium benefit for goal attainment, g = 0.336, while noting possible publication bias (Wang et al., 2021).
Did Libet prove that free will is an illusion?
No. Libet's study of 5 subjects found brain activity beginning before the reported urge to move (Libet et al., 1983), but later work argues that this activity may reflect random fluctuation rather than a decision, and the question of when the decision occurs remains unresolved (Schurger et al., 2021).
Can a person still stop a movement after the brain has begun preparing it?
Yes, up to a point. In a brain-computer interface study that analysed 10 subjects making foot movements, people could cancel a self-initiated movement until roughly 200 ms before it began (Schultze-Kraft et al., 2016).
When do children begin to understand other people's intentions?
Early. In a re-enactment study with small groups of 10, children aged 18 months who saw an adult fail at an act went on to perform the act the adult had intended rather than copy the failed movements (Meltzoff, 1995).
References
Achtziger, A., & Gollwitzer, P. M. (2018). Motivation and volition in the course of action. In J. Heckhausen & H. Heckhausen (Eds.), Motivation and action (3rd ed., pp. 485-527). Springer. https://doi.org/10.1007/978-3-319-65094-4_12
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and Human Decision Processes, 50(2), 179-211. https://doi.org/10.1016/0749-5978(91)90020-T
Ajzen, I. (2015). The theory of planned behaviour is alive and well, and not ready to retire: A commentary on Sniehotta, Presseau, and Araújo-Soares. Health Psychology Review, 9(2), 131-137. https://doi.org/10.1080/17437199.2014.883474
Ajzen, I. (2020). The theory of planned behavior: Frequently asked questions. Human Behavior and Emerging Technologies, 2(4), 314-324. https://doi.org/10.1002/hbe2.195
Armitage, C. J., & Conner, M. (2001). Efficacy of the theory of planned behaviour: A meta-analytic review. British Journal of Social Psychology, 40(4), 471-499. https://doi.org/10.1348/014466601164939
Bieleke, M., Keller, L., & Gollwitzer, P. M. (2021). If-then planning. European Review of Social Psychology, 32(1), 88-122. https://doi.org/10.1080/10463283.2020.1808936
Brass, M., & Haggard, P. (2008). The what, when, whether model of intentional action. The Neuroscientist, 14(4), 319-325. https://doi.org/10.1177/1073858408317417
Bratman, M. E. (1987). Intention, plans, and practical reason. Harvard University Press.
Buehner, M. J., & Humphreys, G. R. (2009). Causal binding of actions to their effects. Psychological Science, 20(10), 1221-1228. https://doi.org/10.1111/j.1467-9280.2009.02435.x
Conner, M., & Norman, P. (2022). Understanding the intention-behavior gap: The role of intention strength. Frontiers in Psychology, 13, 923464. https://doi.org/10.3389/fpsyg.2022.923464
Desmurget, M., Reilly, K. T., Richard, N., Szathmari, A., Mottolese, C., & Sirigu, A. (2009). Movement intention after parietal cortex stimulation in humans. Science, 324(5928), 811-813. https://doi.org/10.1126/science.1169896
Einstein, G. O., & McDaniel, M. A. (1990). Normal aging and prospective memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 16(4), 717-726. https://doi.org/10.1037/0278-7393.16.4.717
Fishbein, M., & Ajzen, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research. Addison-Wesley.
Fogassi, L., Ferrari, P. F., Gesierich, B., Rozzi, S., Chersi, F., & Rizzolatti, G. (2005). Parietal lobe: From action organization to intention understanding. Science, 308(5722), 662-667. https://doi.org/10.1126/science.1106138
Fried, I., Mukamel, R., & Kreiman, G. (2011). Internally generated preactivation of single neurons in human medial frontal cortex predicts volition. Neuron, 69(3), 548-562. https://doi.org/10.1016/j.neuron.2010.11.045
Gollwitzer, P. M. (1999). Implementation intentions: Strong effects of simple plans. American Psychologist, 54(7), 493-503. https://doi.org/10.1037/0003-066x.54.7.493
Gollwitzer, P. M., & Brandstätter, V. (1997). Implementation intentions and effective goal pursuit. Journal of Personality and Social Psychology, 73(1), 186-199. https://doi.org/10.1037/0022-3514.73.1.186
Gollwitzer, P. M., & Sheeran, P. (2006). Implementation intentions and goal achievement: A meta-analysis of effects and processes. Advances in Experimental Social Psychology, 38, 69-119. https://doi.org/10.1016/S0065-2601(06)38002-1
Goschke, T., & Kuhl, J. (1993). Representation of intentions: Persisting activation in memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 19(5), 1211-1226. https://doi.org/10.1037/0278-7393.19.5.1211
Haggard, P. (2008). Human volition: Towards a neuroscience of will. Nature Reviews Neuroscience, 9(12), 934-946. https://doi.org/10.1038/nrn2497
Haggard, P. (2017). Sense of agency in the human brain. Nature Reviews Neuroscience, 18(4), 196-207. https://doi.org/10.1038/nrn.2017.14
Haggard, P., Clark, S., & Kalogeras, J. (2002). Voluntary action and conscious awareness. Nature Neuroscience, 5(4), 382-385. https://doi.org/10.1038/nn827
Haggard, P., & Eimer, M. (1999). On the relation between brain potentials and the awareness of voluntary movements. Experimental Brain Research, 126(1), 128-133. https://doi.org/10.1007/s002210050722
Hagger, M. S., Cheung, M. W.-L., Ajzen, I., & Hamilton, K. (2022). Perceived behavioral control moderating effects in the theory of planned behavior: A meta-analysis. Health Psychology, 41(2), 155-167. https://doi.org/10.1037/hea0001153
Haynes, J.-D., Sakai, K., Rees, G., Gilbert, S., Frith, C., & Passingham, R. E. (2007). Reading hidden intentions in the human brain. Current Biology, 17(4), 323-328. https://doi.org/10.1016/j.cub.2006.11.072
Heckhausen, H., & Gollwitzer, P. M. (1987). Thought contents and cognitive functioning in motivational versus volitional states of mind. Motivation and Emotion, 11(2), 101-120. https://doi.org/10.1007/BF00992338
Henry, J. D., MacLeod, M. S., Phillips, L. H., & Crawford, J. R. (2004). A meta-analytic review of prospective memory and aging. Psychology and Aging, 19(1), 27-39. https://doi.org/10.1037/0882-7974.19.1.27
Hickok, G. (2009). Eight problems for the mirror neuron theory of action understanding in monkeys and humans. Journal of Cognitive Neuroscience, 21(7), 1229-1243. https://doi.org/10.1162/jocn.2009.21189
Jeannerod, M. (1994). The representing brain: Neural correlates of motor intention and imagery. Behavioral and Brain Sciences, 17(2), 187-202. https://doi.org/10.1017/s0140525x00034026
Knobe, J. (2003). Intentional action and side effects in ordinary language. Analysis, 63(3), 190-194. https://doi.org/10.1093/analys/63.3.190
Lau, H. C., Rogers, R. D., Haggard, P., & Passingham, R. E. (2004). Attention to intention. Science, 303(5661), 1208-1210. https://doi.org/10.1126/science.1090973
Libet, B., Gleason, C. A., Wright, E. W., & Pearl, D. K. (1983). Time of conscious intention to act in relation to onset of cerebral activity (readiness-potential): The unconscious initiation of a freely voluntary act. Brain, 106(3), 623-642. https://doi.org/10.1093/brain/106.3.623
Malle, B. F., & Knobe, J. (1997). The folk concept of intentionality. Journal of Experimental Social Psychology, 33(2), 101-121. https://doi.org/10.1006/jesp.1996.1314
Maoz, U., Yaffe, G., Koch, C., & Mudrik, L. (2019). Neural precursors of decisions that matter: An ERP study of deliberate and arbitrary choice. eLife, 8, e39787. https://doi.org/10.7554/eLife.39787
McDaniel, M. A., & Einstein, G. O. (2000). Strategic and automatic processes in prospective memory retrieval: A multiprocess framework. Applied Cognitive Psychology, 14(7), S127-S144. https://doi.org/10.1002/acp.775
McEachan, R. R. C., Conner, M., Taylor, N. J., & Lawton, R. J. (2011). Prospective prediction of health-related behaviours with the theory of planned behaviour: A meta-analysis. Health Psychology Review, 5(2), 97-144. https://doi.org/10.1080/17437199.2010.521684
Meltzoff, A. N. (1995). Understanding the intentions of others: Re-enactment of intended acts by 18-month-old children. Developmental Psychology, 31(5), 838-850. https://doi.org/10.1037/0012-1649.31.5.838
Moore, J. W. (2016). What is the sense of agency and why does it matter? Frontiers in Psychology, 7, 1272. https://doi.org/10.3389/fpsyg.2016.01272
Moore, J. W., & Obhi, S. S. (2012). Intentional binding and the sense of agency: A review. Consciousness and Cognition, 21(1), 546-561. https://doi.org/10.1016/j.concog.2011.12.002
Mylopoulos, M., & Pacherie, E. (2019). Intentions: The dynamic hierarchical model revisited. WIREs Cognitive Science, 10(2), e1481. https://doi.org/10.1002/wcs.1481
Orbell, S., & Sheeran, P. (1998). 'Inclined abstainers': A problem for predicting health-related behaviour. British Journal of Social Psychology, 37(2), 151-165. https://doi.org/10.1111/j.2044-8309.1998.tb01162.x
Ouellette, J. A., & Wood, W. (1998). Habit and intention in everyday life: The multiple processes by which past behavior predicts future behavior. Psychological Bulletin, 124(1), 54-74. https://doi.org/10.1037/0033-2909.124.1.54
Pacherie, E. (2008). The phenomenology of action: A conceptual framework. Cognition, 107(1), 179-217. https://doi.org/10.1016/j.cognition.2007.09.003
Rhodes, R. E., Cox, A., & Sayar, R. (2022). What predicts the physical activity intention–behavior gap? A systematic review. Annals of Behavioral Medicine, 56(1), 1–20. https://doi.org/10.1093/abm/kaab044
Rhodes, R. E., & de Bruijn, G.-J. (2013). How big is the physical activity intention-behaviour gap? A meta-analysis using the action control framework. British Journal of Health Psychology, 18(2), 296-309. https://doi.org/10.1111/bjhp.12032
Schultze-Kraft, M., Birman, D., Rusconi, M., Allefeld, C., Görgen, K., Dähne, S., Blankertz, B., & Haynes, J.-D. (2016). The point of no return in vetoing self-initiated movements. Proceedings of the National Academy of Sciences, 113(4), 1080-1085. https://doi.org/10.1073/pnas.1513569112
Schurger, A., Hu, P. 'B.', Pak, J., & Roskies, A. L. (2021). What is the readiness potential? Trends in Cognitive Sciences, 25(7), 558-570. https://doi.org/10.1016/j.tics.2021.04.001
Schurger, A., Sitt, J. D., & Dehaene, S. (2012). An accumulator model for spontaneous neural activity prior to self-initiated movement. Proceedings of the National Academy of Sciences, 109(42), E2904-E2913. https://doi.org/10.1073/pnas.1210467109
Scullin, M. K., McDaniel, M. A., & Shelton, J. T. (2013). The Dynamic Multiprocess Framework: Evidence from prospective memory with contextual variability. Cognitive Psychology, 67(1-2), 55-71. https://doi.org/10.1016/j.cogpsych.2013.07.001
Sheeran, P. (2002). Intention-behavior relations: A conceptual and empirical review. European Review of Social Psychology, 12(1), 1-36. https://doi.org/10.1080/14792772143000003
Sheeran, P., & Webb, T. L. (2016). The intention–behavior gap. Social and Personality Psychology Compass, 10(9), 503–518. https://doi.org/10.1111/spc3.12265
Smith, R. E. (2003). The cost of remembering to remember in event-based prospective memory: Investigating the capacity demands of delayed intention performance. Journal of Experimental Psychology: Learning, Memory, and Cognition, 29(3), 347-361. https://doi.org/10.1037/0278-7393.29.3.347
Smith, R. E., & Bayen, U. J. (2004). A multinomial model of event-based prospective memory. Journal of Experimental Psychology: Learning, Memory, and Cognition, 30(4), 756-777. https://doi.org/10.1037/0278-7393.30.4.756
Sniehotta, F. F., Presseau, J., & Araújo-Soares, V. (2014). Time to retire the theory of planned behaviour. Health Psychology Review, 8(1), 1-7. https://doi.org/10.1080/17437199.2013.869710
Soon, C. S., Brass, M., Heinze, H.-J., & Haynes, J.-D. (2008). Unconscious determinants of free decisions in the human brain. Nature Neuroscience, 11(5), 543-545. https://doi.org/10.1038/nn.2112
Synofzik, M., Vosgerau, G., & Newen, A. (2008). Beyond the comparator model: A multifactorial two-step account of agency. Consciousness and Cognition, 17(1), 219-239. https://doi.org/10.1016/j.concog.2007.03.010
Tomasello, M., Carpenter, M., Call, J., Behne, T., & Moll, H. (2005). Understanding and sharing intentions: The origins of cultural cognition. Behavioral and Brain Sciences, 28(5), 675-691. https://doi.org/10.1017/S0140525X05000129
Wang, G., Wang, Y., & Gai, X. (2021). A meta-analysis of the effects of mental contrasting with implementation intentions on goal attainment. Frontiers in Psychology, 12, 565202. https://doi.org/10.3389/fpsyg.2021.565202
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. https://doi.org/10.1037/0033-2909.132.2.249
Wegner, D. M., & Wheatley, T. (1999). Apparent mental causation: Sources of the experience of will. American Psychologist, 54(7), 480-492. https://doi.org/10.1037/0003-066x.54.7.480
Woodward, A. L. (1998). Infants selectively encode the goal object of an actor's reach. Cognition, 69(1), 1-34. https://doi.org/10.1016/S0010-0277(98)00058-4
Zapparoli, L., Seghezzi, S., & Paulesu, E. (2017). The what, the when, and the whether of intentional action in the brain: A meta-analytical review. Frontiers in Human Neuroscience, 11, 238. https://doi.org/10.3389/fnhum.2017.00238