Psychological Anticipation

Abstract

Psychological anticipation is a type of mental process: a future-directed state in which the system represents and prepares for an event before it occurs, so that perception, feeling, and action are shaped by what is predicted rather than by what has arrived. This article distinguishes the preparatory face of anticipation, visible in slow anticipatory brain potentials, from its motivational face, the anticipatory affect and reward prediction that precede an outcome, and its constructive face, the episodic simulation of future events. It sets out how expectation is thought to shape perception under the predictive-processing account, why forecasts of one's own future feelings are systematically biased, and how the same machinery, turned toward uncertain threat, becomes anxiety. Three demonstrations let the reader run reward-prediction-error learning, combine a prior expectation with sensory evidence, and trace the impact bias in affective forecasting.

Keywords: anticipation, expectation, prediction, prospection, affective forecasting

Much of what the mind does is not a reaction to the present but a preparation for the future. Before a sprinter's gun fires, the muscles are already poised; before a familiar phrase ends, the next word is already half-formed; before a meal arrives, its pleasure is already being tasted in imagination. Anticipation names this pervasive forward tilt of cognition — the readiness for an expected event that begins to organise perception and behaviour in advance of the event itself. The economist George Loewenstein observed that anticipation is not merely a means to future consumption but a source of present value in its own right, since a person will pay to savour a good outcome ahead of time and pay again to have a dreaded one over with sooner (Loewenstein, 1987). This article treats anticipation as a mental process with distinct but connected faces: a preparatory face that tunes the sensorimotor system before a stimulus, a motivational face that assigns value to outcomes not yet delivered, and a constructive face that simulates the future in enough detail to guide choice — together with the systematic errors that each face is heir to.

What Psychological Anticipation Is

Psychological anticipation is the internal representation of a future event together with the preparatory adjustments that representation sets in motion. It is more than a passive expectation that something will happen: to anticipate is to be readied, so that the anticipated event, when it comes, meets a system already leaning toward it. That leaning can be perceptual, as when an expectation biases what is seen; motoric, as when a response is pre-programmed before the go signal; or affective, as when the prospect of a reward or a punishment is already felt. What unifies these is temporal direction. Where memory faces the past and perception registers the present, anticipation faces the future, and it does so not as idle prediction but as a commitment of present resources to a state that has not yet arrived.

It helps to separate two components that ordinary language runs together. Expectation is the cognitive component — a representation, more or less precise, of what is likely to occur and when. Anticipatory affect is the motivational-emotional component — the pleasure or dread that the represented outcome produces in advance (Knutson & Greer, 2008). The two can dissociate: a person may expect an event with high confidence yet feel little about it, or feel intense dread about an outcome whose probability is low. Much of the psychology of anticipation is the study of how these components are computed, how they drive behaviour, and how they mislead. Table 1 sets out the principal faces of anticipation that the rest of the article develops, each with the kind of event it prepares for and the measure that reveals it.

The Anticipatory Brain

The most direct evidence that anticipation is a real process, and not a way of describing a fast reaction after the fact, is that the brain can be caught preparing before anything happens. In 1964 Grey Walter and colleagues recorded the electroencephalogram while people waited in a simple task: a warning stimulus was followed, after a fixed interval, by an imperative stimulus that called for a response. During the interval between the two, a slow negative shift developed at the scalp and grew until the imperative stimulus arrived — an electrical sign of the association between the warning and the required action, which they named the contingent negative variation (Walter et al., 1964). The potential is contingent because it appears only when the first stimulus reliably predicts the second; break the contingency and it fades. It is a physiological marker of expectancy: the interval is not empty waiting but active preparation, measurable as it unfolds.

Later work resolved this anticipatory activity into more than one component. Cornelis Brunia and colleagues distinguished a motor-preparation component, tied to readying the specific response, from a non-motor component that builds in anticipation of information itself — the stimulus-preceding negativity, which grows before a stimulus that carries feedback or instruction even when no immediate response is required (Brunia & van Boxtel, 2001). The stimulus-preceding negativity shows that anticipation prepares perception and evaluation, not only movement: the system tunes itself toward an informative event as well as toward an action. Figure 1 places these anticipatory signals on the timeline of a warned reaction, between the cue that permits prediction and the target that fulfils it.

Prediction and Reward

If anticipation prepares the system for an outcome, the system needs some way to learn what to anticipate, and the clearest account of that learning came from the study of reward. Wolfram Schultz and colleagues recorded from midbrain dopamine neurons while monkeys learned that a cue predicted juice. Early in learning the neurons fired when the juice arrived; after the cue-reward relation was learned, they fired to the cue and fell silent at the now-fully-predicted reward — and if the predicted reward was omitted, they dipped below baseline at exactly the time it was due (Schultz, Dayan, & Montague, 1997). The signal, in other words, did not track reward as such but the difference between anticipated and actual reward: a reward prediction error. This is the neural currency of anticipation. A positive error says the world is better than expected and revises the anticipation upward; a zero error says the outcome was exactly foreseen; a negative error signals disappointment. Over repetitions the anticipation converges on the world, and the error shrinks toward zero. The computational model that formalises this learning is temporal-difference learning, a reinforcement-learning scheme in which each prediction of future reward is corrected by the error between successive estimates; the dopamine response behaves as precisely the temporal-difference reward-prediction error the model prescribes, which is why the same error term appears in both the theory and the neuron (Niv, 2009).

In humans, anticipation of reward has a characteristic signature of its own. Brian Knutson and colleagues, using functional magnetic resonance imaging while people awaited monetary outcomes, found that the anticipation of an increasing reward selectively activated the nucleus accumbens, a central node of the brain's valuation circuitry, before any reward was delivered (Knutson, Adams, Fong, & Hommer, 2001). Anticipatory affect of this kind is not an idle preview; it feeds forward into choice, so that the intensity of the pre-outcome signal helps determine which option a person selects (Knutson & Greer, 2008). Anticipation can also be aversive in a way that pure expected value cannot explain. Gregory Berns and colleagues showed that many people, offered the choice of a stronger electric shock sooner or a weaker one later, chose the stronger-sooner option: they were paying, in pain, simply to stop having to wait — a neural substrate of dread in which the anticipation of a bad outcome is itself costly (Berns et al., 2006). Loewenstein had drawn out this logic decades earlier: because waiting for a good thing can be pleasurable and waiting for a bad thing painful, anticipation enters the valuation of an outcome directly, sometimes reversing the discounting that would otherwise make the sooner option preferred (Loewenstein, 1987). The first demonstration lets the reader drive a reward-prediction-error learner and watch anticipation converge on the reward.

Expectation Shapes Perception

The reward findings suggest a system that constantly compares what it anticipated with what it got. An influential framework generalises exactly this idea to perception and beyond. On the predictive-processing account, the brain is not a passive receiver of sensation but an active anticipation engine that continually generates top-down predictions of its sensory input and uses the incoming signal mainly to correct them: what propagates forward is not the raw input but the prediction error, the part of the input that was not already anticipated (Clark, 2013). Perception, on this view, is controlled hallucination kept in check by sensory evidence — the percept is the brain's best prediction, updated by whatever the senses did not foresee. Clark's account is an exposition of Karl Friston's free-energy principle, the framework's most general formulation, which casts perception, action, and learning alike as the minimisation of a single quantity — the long-run prediction error, or free energy, between the brain's internal generative model and its sensory input (Friston, 2010). The proposal is bold, and its empirical standing is genuinely contested; a careful appraisal of the neurophysiological evidence finds robust support for some claims (that expectations modulate sensory responses) and much weaker support for the framework's stronger architectural commitments (Walsh et al., 2020).

The better-supported half of that appraisal is substantial. Floris de Lange and colleagues, reviewing how prior expectations shape perception, showed that an expectation can bias what is seen, sharpen the neural representation of an anticipated stimulus, and speed its recognition — expectation acts before and during perception, not only as a later guess (de Lange, Heilbron, & Kok, 2018). The mechanism is naturally described in the language of probability: a prior expectation and a noisy sensory signal are combined, and the resulting percept is pulled toward whichever is more reliable. A large meta-analysis of human neuroimaging confirms that prediction-error signals are not confined to reward but appear across perceptual, cognitive, and motor domains, consistent with error-correction as a general principle rather than a quirk of the dopamine system (Corlett, Mollick, & Kober, 2022). The connection to the older reward literature is exact: the reward prediction error of the previous section is the motivational special case of a computation the whole cortex may perform. The second demonstration lets the reader combine a prior expectation with sensory evidence and watch the percept shift.

Prospection and Episodic Future Thinking

Anticipation reaches its most distinctively human form when the future being prepared for is not the next second but a specific, personal event days or years away. Daniel Schacter and colleagues proposed that the same constructive network the brain uses to remember the past is turned to imagining the future: recalling an episode and simulating a future one recruit a common core of regions, and the memory system is, on this view, fundamentally a prospective one, built as much to anticipate as to retrieve (Schacter, Addis, & Buckner, 2007). Episodic future thinking — the detailed mental pre-experiencing of a particular event — inherits both the power and the fallibility of memory: it is flexible and generative, and, because it is reconstructed rather than replayed, it is prone to distortion (Schacter, Benoit, & Szpunar, 2017). This capacity for mental time travel, Thomas Suddendorf and Michael Corballis argued, may be a defining and possibly uniquely human achievement, and its emergence in evolution and in child development is closely tied to the corresponding ability to travel mentally into the past (Suddendorf & Corballis, 2007).

Prospection matters because it changes what a person will do now. When people can vividly pre-experience a future reward, the pull of immediate gratification weakens: episodic foresight lets a simulated future outcome compete with a present one, reducing the steep discounting of delayed rewards (Bulley, Henry, & Suddendorf, 2016). Martin Seligman and colleagues pressed the point into a general thesis, arguing that much of mental life is better understood as drawn by the anticipated future than as driven by the accumulated past — that guidance by prospection is the more accurate frame for cognition and action than a purely backward-looking, association-driven one (Seligman, Railton, Baumeister, & Sripada, 2013). Anticipation, on this view, is not one process among many but a central organising principle of the mind.

Forecasting Feelings and Anticipating Threat

If prospection guides present choice, its accuracy matters, and here anticipation reveals a systematic flaw. When people forecast how they will feel about a future event — winning a prize, ending a relationship, receiving a diagnosis — they reliably overestimate both how intense and how long-lasting the emotional reaction will be. Timothy Wilson and Daniel Gilbert named this the impact bias and traced much of it to a failure of anticipation to model adaptation: forecasters imagine the event in isolation and neglect the mind's capacity to make sense of, and recover from, even large changes (Wilson & Gilbert, 2005). That mechanism has a name. In the founding study, Gilbert, Elizabeth Pinel, Wilson, and colleagues showed that people harbour a psychological immune system — a set of largely nonconscious processes of rationalisation and sense-making that quietly blunts the sting of adverse events — and that forecasters routinely fail to anticipate it. This immune neglect is a principal source of the durability bias, the specific tendency to overestimate how long a negative reaction will last: because the immune system repairs mood faster than expected, and, paradoxically, works better on major blows than on minor irritations, people over-predict the persistence of their worst-imagined feelings most of all (Gilbert, Pinel, Wilson, Blumberg, & Wheatley, 1998). Prospection here is a construction that leaves things out, and the omissions are lawful rather than random. Gilbert and Wilson set this within a broader account of prospection as pre-experience: because imagining an event elicits a preview of the feelings it would cause, and because those previews are truncated and decontextualised, the anticipated emotion is a biased sample of the real one (Gilbert & Wilson, 2007). Anticipation of one's own affect is thus useful but calibrated poorly, and knowing its characteristic errors is part of using it well. The third demonstration lets the reader compare a forecast with the experience it predicts.

The same forward machinery, directed at threat rather than reward, becomes a principal engine of anxiety. Dan Grupe and Jack Nitschke argued that anxiety is best understood as maladaptive anticipation of an uncertain future: heightened estimates of the likelihood and cost of potential harm, an intolerance of uncertainty, and hypervigilant preparation for threats that may never arrive (Grupe & Nitschke, 2013). Anticipation here is not broken but miscalibrated — the preparatory system, valuable when threats are real and predictable, imposes a heavy cost when it treats the merely possible as imminent. Anticipation also shapes ordinary choice through the body. Antonio Damasio's somatic-marker hypothesis holds that anticipated outcomes are tagged with bodily-emotional signals — somatic markers — that are re-evoked when a similar choice arises and bias decision before, and faster than, explicit reasoning can (Damasio, 1996). Across these cases the lesson is constant: anticipation is a single forward-looking capacity whose value and whose pathology are two sides of the same computation.

Worked Example

Reward-prediction-error learning can be worked by hand, and the first demonstration reproduces the calculation. A learner holds a value estimate $V$ of how much reward a cue predicts, and updates it after each trial by the rule $V_{\text{new}} = V + \alpha\,\delta$, where the prediction error is $\delta = r - V$, $r$ is the reward actually received, and $\alpha$ is a learning rate between 0 and 1. Suppose the cue always yields a reward of $r = 1$, the learner starts fully naive with $V_0 = 0$, and the learning rate is $\alpha = 0.5$.

On the first trial the learner anticipates nothing, so the prediction error is the full surprise of the reward: $\delta_1 = 1 - 0 = 1$, and the value becomes $V_1 = 0 + 0.5 \times 1 = 0.5$. On the second trial the reward is now half-anticipated: $\delta_2 = 1 - 0.5 = 0.5$, and $V_2 = 0.5 + 0.5 \times 0.5 = 0.75$. On the third, $\delta_3 = 1 - 0.75 = 0.25$ and $V_3 = 0.875$; on the fourth, $\delta_4 = 0.125$ and $V_4 = 0.9375$. Two things happen together, and they are the whole point. The anticipated value climbs toward the reward — 0, 0.5, 0.75, 0.875, 0.9375 — while the prediction error shrinks toward zero — 1, 0.5, 0.25, 0.125. A fully learned cue produces almost no error at reward, because the reward was fully anticipated; the surprise has migrated from the outcome to the cue that predicts it. This is exactly the shift Schultz and colleagues recorded in dopamine neurons, and it shows in miniature why a prediction-error signal is the natural teacher of anticipation: it is loud precisely when anticipation is wrong and silent once anticipation is right (Schultz, Dayan, & Montague, 1997).

Discussion

The study of anticipation is unusual in cognitive psychology for how far its several literatures converge on one idea. The slow negativity that builds before a target, the dopamine signal that shifts from reward to cue, the expectation that sharpens a percept, the simulation that pre-experiences a holiday, and the dread that makes a person pay to end the waiting are not five unrelated phenomena but five expressions of a system organised around prediction (Walter et al., 1964; Schultz, Dayan, & Montague, 1997; Clark, 2013). The predictive-processing framework offers the most ambitious unification, casting perception, action, and valuation alike as the minimisation of the error between what was anticipated and what occurred — though its stronger claims remain to be secured, and a sober reading keeps the well-evidenced modulation of perception by expectation separate from the more speculative architectural program (de Lange, Heilbron, & Kok, 2018; Walsh et al., 2020).

Two themes deserve emphasis. The first is that anticipation is constructive, and construction has costs: because a simulated future is assembled rather than perceived, it can be assembled wrongly, and the errors are systematic — the impact bias in forecasting feeling, the miscalibrated threat estimates of anxiety, the somatic markers that can bias as easily as they guide (Wilson & Gilbert, 2005; Grupe & Nitschke, 2013). The second is that this forward orientation may be less a special faculty than the default mode of the mind. If cognition is drawn by the anticipated future as much as driven by the recorded past, then memory, reward processing, and decision making are all, in part, instruments of anticipation (Seligman, Railton, Baumeister, & Sripada, 2013; Schacter, Addis, & Buckner, 2007). What the anticipatory tradition establishes is that the mind spends much of its effort not on the present but on a future it is perpetually building, betting on, and bracing for.

Common Misconceptions

Anticipation is just the expectation that something will happen.

Expectation is only the cognitive half. Anticipation also carries an affective, preparatory component — the pleasure or dread felt in advance and the sensorimotor readying it produces — that can be dissociated from, and can even run counter to, the bare probability estimate. Anticipatory affect recruits reward circuitry before any outcome arrives and feeds forward into choice (Knutson & Greer, 2008).

People are good at predicting how future events will make them feel.

Forecasts of one's own future emotion are systematically biased. People overestimate how intense and how long-lasting their reactions will be — the impact bias — largely because they imagine the event in isolation and neglect their own capacity to adapt to it (Wilson & Gilbert, 2005).

Dopamine signals pleasure or reward received.

Midbrain dopamine neurons track a reward prediction error — the difference between anticipated and actual reward — not reward itself. Once a cue fully predicts a reward, the neurons stop responding to the reward and respond instead to the cue, and they dip below baseline when a predicted reward is withheld (Schultz, Dayan, & Montague, 1997).

Imagining the future is simply replaying stored memories.

Episodic future thinking is constructive, not reproductive. The brain recombines elements drawn from memory into a novel simulation of an event that has not occurred, using a network that overlaps with, but is not identical to, simple recollection — which is why imagined futures are flexible and also prone to distortion (Schacter, Benoit, & Szpunar, 2017).

Glossary

Affective forecasting.

Predicting one's own future emotional reactions to events; reliably subject to the impact bias, an overestimation of their intensity and duration.

Anticipation.

A future-directed mental process in which an expected event is represented and prepared for in advance, shaping perception, affect, and action before the event occurs.

Anticipatory affect.

The emotional component of anticipation — the pleasure or dread produced by a represented future outcome before it is delivered — which recruits reward circuitry and biases choice.

Contingent negative variation.

A slow negative brain potential that develops during the interval between a warning stimulus and an imperative one, indexing expectancy and motor preparation for the required response.

Dread.

The aversive experience of anticipating a negative outcome, costly enough that people will accept a worse outcome sooner simply to end the waiting.

Episodic future thinking.

The detailed mental pre-experiencing of a specific personal future event, constructed by the same network that supports episodic memory.

Expectation.

The cognitive component of anticipation: a representation, more or less precise, of what is likely to occur and when, independent of the feeling it may arouse.

Immune neglect.

The failure to anticipate the operation of one's own psychological immune system when forecasting future feelings; a principal source of the durability bias, the overestimation of how long a negative reaction will last.

Impact bias.

The tendency to overestimate the intensity and duration of future emotional reactions, driven largely by neglect of one's own capacity to adapt.

Mental time travel.

The capacity to project the self mentally backward to relive the past or forward to pre-experience the future; closely tied to episodic memory and possibly uniquely human.

Predictive processing.

The framework in which the brain continually generates top-down predictions of its input and updates them from the prediction error, casting perception and action as anticipation corrected by evidence.

Prospection.

The general capacity to represent and be guided by the future, encompassing prediction, planning, simulation, and the anticipation of one's own reactions.

Psychological immune system.

The set of largely nonconscious processes — rationalisation, sense-making, and dissonance reduction — that blunt the impact of adverse events; because forecasters neglect it, they overpredict the intensity and duration of future distress.

Reward prediction error.

The difference between anticipated and actual reward, signalled by midbrain dopamine neurons and used to update the value assigned to predictive cues.

Somatic marker.

A bodily-emotional signal that tags an anticipated outcome and is re-evoked at a similar choice, biasing decision before explicit reasoning completes.

Stimulus-preceding negativity.

A slow anticipatory brain potential that builds before an informative or feedback stimulus even when no immediate response is required, indexing anticipation of information rather than of action.

Temporal difference learning.

A reinforcement-learning scheme in which predictions are updated from the moment-to-moment prediction error, providing the computational account of the dopamine reward-prediction-error signal.

Key Researchers

Andy Clark. Professor of Cognitive Philosophy at the University of Sussex; he is a leading proponent of the predictive-processing framework, in which the brain is fundamentally an anticipation engine that perceives and acts by minimising the error between top-down predictions and incoming signals. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Karl J. Friston. Professor of Imaging Neuroscience at University College London and Scientific Director of the Wellcome Centre for Human Neuroimaging; he formulated the free-energy principle, the most general statement of the predictive brain, which casts perception, action, and learning as the minimisation of the prediction error between an internal generative model and sensory input. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Daniel T. Gilbert. Edgar Pierce Professor of Psychology at Harvard University; with Timothy Wilson he established the study of affective forecasting and prospection, showing that people systematically mispredict the intensity and duration of their future emotional reactions. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Brian Knutson. Professor of Psychology and Neuroscience at Stanford University; he used functional neuroimaging to isolate anticipatory affect, showing that anticipation of monetary reward recruits the nucleus accumbens before an outcome is delivered and that these signals predict subsequent choices. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Daniel L. Schacter. William R. Kenan Jr. Professor of Psychology at Harvard University; he advanced the prospective-brain hypothesis and the study of episodic future thinking, showing that the constructive memory system that recalls the past is also used to simulate and anticipate future events. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Wolfram Schultz. Professor of Neuroscience at the University of Cambridge; he discovered that midbrain dopamine neurons encode a reward prediction error, the signal by which anticipated and actual rewards are compared and by which anticipatory value is learned. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Thomas Suddendorf. Professor of Psychology at the University of Queensland; he developed the theory of mental time travel, arguing that the capacity to anticipate and mentally pre-experience future events is a defining, possibly uniquely human, cognitive achievement. Faculty Page - Google Scholar - Wikidata - ORCID

Timothy D. Wilson. Sherrell J. Aston Professor of Psychology at the University of Virginia; with Daniel Gilbert he developed affective-forecasting research, documenting the impact bias by which people overestimate how long and how intensely anticipated events will affect their well-being. Faculty Page - Google Scholar - Wikipedia - Wikidata - ORCID

Frequently Asked Questions

What is psychological anticipation?
It is a future-directed mental process in which an expected event is represented and prepared for before it occurs, so that perception, affect, and action are shaped in advance by what is predicted. It has cognitive, affective, and preparatory components that can be measured separately (Knutson & Greer, 2008).

How is anticipation different from expectation?
Expectation is the cognitive component: an estimate of what is likely to happen and when. Anticipation adds an affective and preparatory component, the pleasure or dread felt in advance and the readying of the sensorimotor system, which can run counter to the bare probability (Loewenstein, 1987).

Can anticipation be seen in the brain before anything happens?
Yes. A slow negative potential, the contingent negative variation, builds during the interval between a warning stimulus and a predicted target, and a related stimulus-preceding negativity builds before informative input, showing that the waiting interval is filled with active preparation (Walter et al., 1964).

What is a reward prediction error?
It is the difference between anticipated and actual reward. Midbrain dopamine neurons signal this error rather than reward itself, shifting their response from the reward to the cue that predicts it as the association is learned, which is how anticipation is updated (Schultz, Dayan, & Montague, 1997).

Does expectation change what people perceive?
Prior expectations can bias what is seen, sharpen the neural representation of an anticipated stimulus, and speed its recognition. On the predictive-processing account, perception is the brain's best prediction of its input, corrected by the part of the signal that was not anticipated (de Lange, Heilbron, & Kok, 2018).

Why are people bad at predicting their own future feelings?
Forecasters imagine an event in isolation and neglect their capacity to adapt to it, so they overestimate how intense and how long-lasting the emotion will be, an effect called the impact bias. The anticipated feeling is a truncated, decontextualised preview of the real one (Wilson & Gilbert, 2005).

Is imagining the future related to remembering the past?
Closely. The same constructive network supports both, and the memory system appears to be fundamentally prospective, built to simulate and anticipate future events as much as to retrieve past ones. This shared machinery is often called mental time travel (Schacter, Addis, & Buckner, 2007).

How does anticipation relate to anxiety?
Anxiety can be understood as maladaptive anticipation of uncertain future threat, marked by inflated estimates of the likelihood and cost of harm, intolerance of uncertainty, and hypervigilant preparation for dangers that may not arrive (Grupe & Nitschke, 2013).

References