Psychophysiology

Abstract

Psychophysiology is the branch of science that infers psychological states and processes from the physiological signals of the intact, behaving organism, chiefly the electrical and mechanical activity of the brain, heart, skin, muscles, and viscera. It is defined by its logic of inference rather than by any one signal: an experimenter manipulates a psychological variable and reads its consequences in a bodily measure recorded noninvasively. This article sets out that inferential logic and its central difficulty, that the mapping between mind and body is rarely one to one; surveys the core measurement domains of electrodermal activity, cardiovascular dynamics and heart rate variability, and the event-related potentials of the electroencephalogram; and develops the doctrine of autonomic space that replaced the single arousal dimension. Three interactive demonstrations let the reader manipulate autonomic control, heart rate variability, and the oddball P300.

Keywords: psychophysiology, autonomic nervous system, heart rate variability, event-related potentials, electrodermal activity

Psychophysiology is the study of the relations between psychological states and the physiological activity of the body, approached from the psychological side: the independent variable is a manipulation of perception, emotion, attention, or cognition, and the dependent variable is a physiological signal recorded from the surface of the intact organism. The discipline is distinguished from physiological psychology, its near neighbor, mainly by direction and method. Physiological psychology manipulates the nervous system, often invasively in animals, and observes the behavioral result; psychophysiology manipulates a psychological state in the awake, unrestrained human and observes the bodily result, typically through noninvasive recording (Cacioppo & Tassinary, 1990). Its instruments are the electroencephalogram, the electrocardiogram, the skin conductance sensor, the electromyogram, and, increasingly, the wearable and the imaging scanner. Its founding problem, and the thread that runs through everything below, is that a physiological signal almost never corresponds to a single psychological cause, so the field's real subject is the disciplined inference from an ambiguous bodily measure to the mental process that produced it.

What Psychophysiology Is

Psychophysiology is best defined by its method of inference rather than by a list of instruments. In the canonical framing, the psychophysiologist treats a physiological signal as a sign of an unobservable psychological event and asks what class of inference the sign supports (Cacioppo & Tassinary, 1990). A signal may be an outcome, reliably produced by a psychological state but also by others; a marker, associated with the state within a defined context; a concomitant, co-occurring without a known functional tie; or, in the strongest case, an invariant, standing in a one-to-one relation with the state across all contexts. Genuine invariants are rare, which is why the discipline is so concerned with specifying the conditions under which a measure means what it is taken to mean. The signals themselves fall into a few families. The central nervous system is read through the electroencephalogram and its time-locked event-related potentials; the autonomic nervous system is read through cardiovascular measures, electrodermal activity, and respiration; the somatic system is read through the electromyogram; and the endocrine and immune systems are sampled through assays of hormones and cytokines. What unites them is the constraint that recording must leave the organism intact and, ideally, unaware of being measured, so that the physiology reflects the psychological manipulation rather than the act of measurement.

Types of Psychophysiology

Psychophysiology is also a formal descriptor in the National Library of Medicine's Medical Subject Headings, which files it beneath Psychological Phenomena in one branch of its classification and beneath the behavioral sciences and physiology in others, at tree position F02.830. Beneath the descriptor MeSH hangs a set of narrower headings, listed in Table 1. Two cautions apply. The list is an indexing classification built to organize the literature, not a theory that carves the mind at its joints, and its members are neither mutually exclusive nor jointly exhaustive: a study of the orienting response touches Arousal, Orientation, and Reaction Time at once. Only the subtypes that are themselves live articles on this site are linked.

Inference From Physiological Signals

The organizing difficulty of psychophysiology is that the relation between a mental event and a bodily signal is many-to-many. One psychological state can produce many physiological effects, and one physiological effect can arise from many psychological states, so a raw signal underdetermines its cause. Cacioppo and Tassinary made this the theoretical heart of the field, classifying the possible relations by their generality and specificity and warning that most useful measures are context-bound markers rather than context-free invariants (Cacioppo & Tassinary, 1990). A rise in skin conductance, for example, follows a startling noise, a difficult mental arithmetic problem, a lie, and a sexually arousing image; the signal alone cannot say which, and only the experimental context licenses the inference. The point is not that physiological measures are weak but that their interpretation is a matter of specifying relations. Two research strategies follow. One holds the psychological state fixed and asks which signals reliably accompany it, the search for the physiological signature of an emotion or a cognitive operation. Kreibig's review of autonomic activity across emotions found that discrete emotions are distinguished by patterns of response across many measures rather than by any single index, precisely the many-to-one structure the framework predicts (Kreibig, 2010). The other strategy holds the signal fixed and asks what it can reveal, the logic behind using a well-characterized measure such as the P300 as a probe of attention or memory across many tasks. The whole enterprise depends on measurement standards rigorous enough that a signal recorded in one laboratory means the same as the same signal recorded in another, which is why the field has invested heavily in published recording and reporting guidelines (Fowles et al., 1981; Boucsein et al., 2012).

The Autonomic Nervous System and Autonomic Space

Much of psychophysiology reads the autonomic nervous system, the division that governs the viscera, the glands, and the vasculature below the level of voluntary control. For most of the twentieth century the autonomic output was pictured as a single continuum of arousal, with the sympathetic and parasympathetic branches acting as reciprocal ends of one seesaw: as one rose the other fell, and a single number captured the organism's activation. Berntson, Cacioppo, and Quigley showed this picture to be too simple and replaced it with the doctrine of autonomic space (Berntson et al., 1991). Because the two branches are anatomically and functionally distinct, their activity is better represented on two independent axes than on one. A target organ such as the heart can therefore be driven by reciprocal control, the classic pattern in which sympathetic activation accompanies parasympathetic withdrawal; by coactivation or coinhibition, in which both branches move together; or by uncoupled control, in which one branch changes while the other holds. The consequence is that the same heart rate can arise from very different autonomic states, and that a peripheral measure cannot be read back to a single dimension of arousal. Porges extended the functional analysis of the parasympathetic branch with the polyvagal perspective, distinguishing an evolutionarily older unmyelinated vagal pathway from a newer myelinated one that supports the calm, socially engaged state and the rapid regulation of the heart (Porges, 2007). Whatever the fate of the specific evolutionary claims, the shift from one arousal dial to a structured autonomic space reorganized how peripheral signals are interpreted, and the demonstration below lets the reader see how the two branches jointly set the heart rate.

Electrodermal Activity

Electrodermal activity, the variation in the electrical conductance of the skin produced by the sweat glands, is the oldest and one of the most widely used autonomic measures. Its great analytical convenience is that the eccrine sweat glands of the palms and soles are innervated almost purely by the sympathetic branch, so that skin conductance offers a comparatively clean read-out of sympathetic activation uncontaminated by parasympathetic influence. The signal has two components. The slowly drifting tonic level, the skin conductance level, tracks general arousal over minutes; superimposed on it are fast phasic bursts, the skin conductance responses, that follow discrete stimuli within one to three seconds. A response that follows an identifiable stimulus is event-related, while spontaneous responses that arise without an external trigger index the tonic state of arousal. Because the measure is sensitive, cheap, and easy to record, it became the backbone of applied psychophysiology, from the orienting-response studies of the mid-century to the arousal component of the polygraph. That very sensitivity is also its weakness: skin conductance rises to novelty, effort, threat, and significance alike, the paradigm case of a signal whose interpretation depends wholly on context. The field's response has been to standardize how the signal is recorded and scored, so that a skin conductance response measured in one study is commensurable with one measured in another, a concern formalized in successive sets of publication recommendations (Fowles et al., 1981; Boucsein et al., 2012).

Cardiovascular Measures and Heart Rate Variability

The heart is dually innervated, receiving both sympathetic and parasympathetic input, which makes cardiovascular measures rich but harder to interpret than electrodermal ones. The most informative of them has proved to be not the average heart rate but its variability. Heart rate variability is the beat-to-beat fluctuation in the interval between successive heartbeats, and it is far from noise: a healthy heart at rest speeds slightly on inhalation and slows on exhalation, a rhythm called respiratory sinus arrhythmia that is mediated by the myelinated vagus. Because the vagal brake can adjust the heart within a single beat while sympathetic effects unfold over seconds, the fast, respiration-linked component of variability is a relatively specific index of parasympathetic control (Task Force, 1996). Variability is quantified in two families of measures. Time-domain indices work directly on the series of interbeat intervals: the standard deviation of those intervals captures overall variability, while the root mean square of successive differences, RMSSD, isolates the rapid beat-to-beat changes that reflect vagal activity. Frequency-domain indices decompose the same series into a high-frequency band, aligned with respiration and vagally mediated, and a low-frequency band of more mixed origin (Shaffer & Ginsberg, 2017). Thayer and Lane placed these measures in a theory: their model of neurovisceral integration holds that a network linking the prefrontal cortex to the heart supports flexible self-regulation, so that resting vagal tone indexes the capacity to regulate emotion and attention, and low variability marks a system stuck in defensive rigidity (Thayer & Lane, 2000). The interpretation is not without dispute, and the low-frequency band in particular has resisted a clean autonomic reading, which is why the field has worked to standardize both the computation and the reporting of these indices (Quintana et al., 2016). The demonstration below builds heart rate variability from a stream of interbeat intervals so that the reader can watch the time-domain indices change with vagal tone.

Event-Related Potentials

Where autonomic measures resolve psychological events over seconds, the electroencephalogram resolves them over milliseconds, and the event-related potential is the technique that extracts that resolution. The raw electroencephalogram recorded from the scalp is dominated by ongoing rhythms that swamp the small voltage changes evoked by any single stimulus. Averaging many trials time-locked to a repeated event cancels the uncorrelated background and leaves the event-related potential, a sequence of positive and negative deflections labeled by polarity and latency. The most studied of these is the P300, a large positive wave arising around 300 milliseconds after a rare, task-relevant stimulus, discovered when Sutton and colleagues showed that the amplitude of a late positive component grew with the uncertainty an observer resolved (Sutton et al., 1965). In the standard oddball task the participant monitors a stream of frequent stimuli for occasional targets; the rarer and more relevant the target, the larger the P300 it evokes, which is why the component is read as an index of the attentional and memory operations that update a model of the context (Polich, 2007). Other components probe other processes. The error-related negativity, a sharp frontal negativity appearing within about 100 milliseconds of a mistaken response, revealed a neural system that monitors performance and signals errors before the person is even aware of them (Gehring et al., 1993). Because these components are precise, replicable, and tied to specific operations, they have become clinical as well as experimental tools, with blunted or exaggerated components serving as candidate markers of psychopathology (Hajcak et al., 2019). The frequency composition of the ongoing electroencephalogram carries information too: the relative activation of the left and right frontal cortex, frontal alpha asymmetry, has been studied as a correlate of approach and withdrawal motivation, though with the same caveats about reliability that attend any difference score (Coan & Allen, 2004). The demonstration below reproduces the oddball paradigm and its central finding, that P300 amplitude grows as the eliciting event becomes rarer.

The Orienting Response and the Startle Reflex

Two reflexive responses have been especially productive because their modulation by psychological state is lawful enough to serve as a probe. The orienting response is the constellation of changes, a phasic skin conductance response, a brief cardiac deceleration, a turn of the receptors toward the source, that a novel stimulus evokes and that wanes as the stimulus is repeated and found inconsequential. Sokolov analyzed this waning as evidence that the nervous system builds an internal model of the expected stimulus and responds only to the mismatch between input and model, so that habituation is not fatigue but the updating of a neuronal representation (Sokolov, 1963). The startle reflex, the fast protective blink and body flexion evoked by a sudden intense stimulus, is valuable for the opposite reason: its magnitude is systematically tuned by the affective state in which it occurs. Lang and colleagues showed that the startle blink is potentiated when the person is in an unpleasant emotional state and attenuated when the state is pleasant, because a defensive reflex is primed by an already aversive context and inhibited by an appetitive one (Lang et al., 1990). Startle modulation thus provides a rare thing in psychophysiology, a measure whose direction, not merely its magnitude, reports the valence of an emotional state, and it became a workhorse for studying fear, anxiety, and their disorders.

Worked Example

The computation of heart rate variability from a short recording makes the logic of the time-domain measures concrete. Suppose a segment of electrocardiogram yields eight successive interbeat intervals, in milliseconds: 800, 850, 810, 790, 830, 800, 770, 820. The mean interval is their sum, 6470 milliseconds, divided by 8, which is 808.75 milliseconds; converting to rate, 60000 divided by 808.75 gives a mean heart rate of about 74.2 beats per minute. The RMSSD, the index that isolates vagally mediated beat-to-beat change, is built from the differences between consecutive intervals rather than from their spread about the mean. Those successive differences are 50, then negative 40, then negative 20, then 40, then negative 30, then negative 30, then 50 milliseconds. Squaring each removes the sign and weights the larger jumps: the squares are 2500, 1600, 400, 1600, 900, 900, and 2500, which sum to 10400. There are seven differences among eight intervals, so the mean squared difference is 10400 divided by 7, or about 1485.7. The root mean square of successive differences is the square root of that mean, which is about 38.5 milliseconds. The separate calculation of overall variability, the standard deviation of the eight intervals themselves, comes to about 24.8 milliseconds. That the RMSSD exceeds the standard deviation here reflects the rapid alternation in the series, exactly the fast beat-to-beat change the vagus produces; a recording dominated by slow drift would show the opposite ordering. The single number 38.5 milliseconds is what a study would carry forward as its index of parasympathetic control, and the worth of the measure rests entirely on the standardization of how the interbeat series was cleaned and computed (Task Force, 1996; Shaffer & Ginsberg, 2017).

Discussion

Psychophysiology matters first because it made the internal states of the intact, awake person measurable without opening the skull, giving psychology a set of dependent variables that are continuous, involuntary, and recorded in real time as a mental process unfolds. That access is the discipline's enduring contribution, from the arousal theories of the mid-century to the affective neuroscience of the present. It matters second because the field's hardest lesson, that a physiological signal underdetermines its psychological cause, is a general truth about inference from measurement that reaches well beyond the laboratory; the disciplined move from an ambiguous marker to a specified psychological claim is the methodological core that distinguishes the science from the folk belief that the body simply reveals the mind (Cacioppo & Tassinary, 1990). The theoretical advances that reshaped the field, the replacement of a single arousal dimension by a two-dimensional autonomic space and the linking of vagal control to self-regulation, both consist in refusing an oversimple mapping and specifying the structure that connects mind to body (Berntson et al., 1991; Thayer & Lane, 2000). The open questions are correspondingly about specification. Whether resting heart rate variability is a stable trait index of regulatory capacity or a labile state measure, what the low-frequency band of the cardiac spectrum actually reflects, and how far peripheral signals can be pushed toward diagnosing psychopathology are all live (Hajcak et al., 2019; Quintana et al., 2016). What is settled is the founding premise, now so ordinary that its novelty is forgotten: that the traffic between mind and body leaves signals on the surface of the organism, and that those signals, read with care, are evidence.

Current Directions

The most active front in contemporary psychophysiology is the migration of measurement out of the laboratory and into daily life, driven by wearable sensors and the software that makes their signals tractable. Ambulatory assessment records the electrocardiogram, skin conductance, and movement continuously over days, allowing psychophysiological relations established in the controlled setting to be tested against the noise of real behavior; the enterprise has required new analytic strategies to separate a psychological effect from the metabolic and postural demands that also drive the heart and the sweat glands outside the seat of an experiment (Wilhelm & Grossman, 2010). The tractability of these large, messy datasets has come to depend on open, validated analysis pipelines: toolkits such as NeuroKit2 implement the standardized cleaning, peak detection, and index computation that the field's guidelines specify, so that a heart rate variability value is computed the same way across studies and laboratories (Makowski et al., 2021). Reviews aimed at psychologists now pair the theory of these indices with explicit computational tutorials, an acknowledgment that reproducibility in the field is as much a matter of shared code as of shared apparatus (Pham et al., 2021). A second current runs inward rather than outward, toward the brain's representation of the body's own signals. Work on interoception and the neural monitoring of visceral state has shown that the phase of the heartbeat and the rhythm of the stomach modulate perception, attention, and even self-related processing, reframing the viscera not as a downstream read-out of emotion but as an active input that shapes cognition (Critchley & Harrison, 2013; Azzalini et al., 2019). Together the two directions are pulling psychophysiology toward continuous, naturalistic recording on one side and toward the interoceptive integration of bodily signals in the brain on the other.

Common Misconceptions

A polygraph detects lies.

The polygraph records autonomic arousal, chiefly skin conductance, heart rate, and respiration, and arousal accompanies fear, anger, and surprise as readily as deception, so there is no physiological signature of a lie as such. Meta-analysis finds that a better-founded method, the concealed information test, detects the recognition of guilty knowledge with reasonable accuracy, but this measures memory for crime-relevant details, not lying itself (Ben-Shakhar & Elaad, 2003). The persistent belief in lie detection is the clearest cost of reading an arousal signal as though it were an invariant.

Each emotion has its own unique bodily fingerprint.

The evidence for emotion-specific autonomic patterns is real but partial: discrete emotions are distinguished by patterns across many measures, with substantial overlap, rather than by any single diagnostic signal (Kreibig, 2010). This is the many-to-one problem in concrete form; a raised heart rate is common to fear, anger, and exercise, and it is the configuration in context, not the lone measure, that carries emotional information.

Heart rate is controlled by a single arousal dial.

The heart receives independent sympathetic and parasympathetic input, and the same rate can result from reciprocal, coactive, or uncoupled combinations of the two branches; a single number for arousal therefore discards the information that autonomic space was formulated to preserve (Berntson et al., 1991). This is why heart rate variability, which reflects the branches' distinct time courses, carries information that mean heart rate does not.

Glossary

Autonomic nervous system.

The division of the nervous system that regulates the viscera, glands, and vasculature, comprising the sympathetic and parasympathetic branches.

Autonomic space.

The model representing sympathetic and parasympathetic activity on two independent axes rather than one arousal continuum, allowing reciprocal, coactive, and uncoupled modes of control.

Electrodermal activity.

Variation in the electrical conductance of the skin driven by the sympathetically innervated sweat glands, comprising a tonic level and phasic responses.

Electroencephalogram.

The record of the brain's electrical activity measured from the scalp, the substrate from which event-related potentials are extracted.

Error-related negativity.

A sharp frontal negative deflection appearing within about 100 milliseconds of an erroneous response, reflecting a performance-monitoring system.

Event-related potential.

A voltage deflection extracted from the electroencephalogram by averaging many trials time-locked to a repeated event, labeled by polarity and latency.

Heart rate variability.

The beat-to-beat fluctuation in the interval between heartbeats, whose fast, respiration-linked component indexes parasympathetic control.

Orienting response.

The set of autonomic and behavioral changes evoked by a novel stimulus, which habituates as the stimulus is repeated and found inconsequential.

P300.

A large positive event-related potential arising around 300 milliseconds after a rare, task-relevant stimulus, indexing attention and context updating.

Parasympathetic nervous system.

The autonomic branch associated with rest and restoration, whose vagal output can slow the heart within a single beat.

Polyvagal theory.

Porges's account distinguishing a myelinated vagal pathway supporting calm social engagement from an older unmyelinated pathway, offering a functional reading of vagal control.

Respiratory sinus arrhythmia.

The natural speeding of the heart on inhalation and slowing on exhalation, a vagally mediated rhythm and the basis of high-frequency heart rate variability.

RMSSD.

The root mean square of successive differences between interbeat intervals, a time-domain index that isolates rapid, vagally driven beat-to-beat change.

Skin conductance response.

A phasic rise in electrodermal conductance following a discrete stimulus within one to three seconds, a sensitive but context-dependent sign of sympathetic activation.

Startle reflex.

The fast protective blink and flexion evoked by a sudden intense stimulus, whose magnitude is potentiated in unpleasant and attenuated in pleasant emotional states.

Sympathetic nervous system.

The autonomic branch that mobilizes the body for action, driving the sweat glands and, with a slower time course than the vagus, the heart.

Vagal tone.

The level of parasympathetic influence carried by the vagus nerve to the heart, estimated from high-frequency heart rate variability and linked to self-regulatory capacity.

Key Researchers

John J. B. Allen. Professor of psychology at the University of Arizona; his work on frontal electroencephalographic asymmetry and cardiac measures has advanced the use of psychophysiological signals as markers of emotion and depression. Faculty Page - Google Scholar

Gary G. Berntson. Professor emeritus of psychology at The Ohio State University; with John Cacioppo he formulated the doctrine of autonomic space, replacing the single arousal continuum with a two-dimensional model of autonomic control. Faculty Page - Google Scholar - Wikipedia

Margaret M. Bradley. Scientist at the University of Florida's Center for the Study of Emotion and Attention; with Peter Lang she developed the affective startle-modulation paradigm and the standardized stimulus sets that underpin much emotion research. Faculty Page - Google Scholar

John T. Cacioppo (1951-2018). Psychologist at the University of Chicago and a founder of social neuroscience; his relational analysis of psychophysiological inference and his work on autonomic space set much of the field's theoretical agenda. ORCID - Wikipedia

Hugo D. Critchley. Chair of psychiatry at Brighton and Sussex Medical School; his research on interoception and the brain's monitoring of visceral state has reframed bodily signals as active inputs to cognition and emotion. Faculty Page - ORCID - Google Scholar - Wikipedia

Greg Hajcak. Professor of psychology at Santa Clara University; his work on the error-related negativity and other event-related potentials has developed them as reliable neural markers of risk for anxiety and depression. Faculty Page - ORCID - Google Scholar

Peter J. Lang (b. 1930). Professor emeritus at the University of Florida; his three-system analysis of emotion and the discovery that the startle reflex is modulated by affective state made emotion physiologically measurable. Faculty Page

Dominique Makowski. Researcher at the University of Sussex and lead author of NeuroKit2; his open-source tools have standardized the computation of psychophysiological indices across the field. Faculty Page - ORCID - Google Scholar

Stephen W. Porges. Distinguished scientist at the Kinsey Institute, Indiana University; he formulated the polyvagal perspective on the parasympathetic nervous system and its role in social engagement and regulation. Faculty Page - Google Scholar - Wikipedia

Evgeny N. Sokolov (1920-2008). Physiologist at Moscow State University; his neuronal-model account of the orienting response and its habituation became a foundation for the psychophysiology of attention. Wikipedia

Louis G. Tassinary. Professor at Texas A&M University; with John Cacioppo he authored the foundational analysis of how psychological significance is inferred from physiological signals. Faculty Page - Google Scholar

Julian F. Thayer. Professor of psychological science at the University of California, Irvine; his model of neurovisceral integration links heart rate variability to prefrontal regulation of emotion and attention. Faculty Page - ORCID

Frequently Asked Questions

What is psychophysiology?
Psychophysiology is the branch of science that infers psychological states and processes from the physiological signals of the intact, unrestrained organism, such as brain waves, heart rate, and skin conductance. It manipulates a psychological variable and reads its consequences in a bodily measure recorded noninvasively, making the internal states of the awake person measurable without surgery (Cacioppo & Tassinary, 1990).

How is psychophysiology different from physiological psychology?
The two differ mainly in direction and method. Physiological psychology manipulates the nervous system, often invasively in animals, and observes the behavioral result, whereas psychophysiology manipulates a psychological state in the awake human and observes the physiological result through noninvasive recording (Cacioppo & Tassinary, 1990).

Why can a physiological signal not simply reveal what someone is feeling?
Because the mapping between mental states and bodily signals is many-to-many: one state produces many signals and one signal arises from many states. A rise in skin conductance follows startle, effort, threat, and arousal alike, so only the experimental context licenses an inference from the signal to its cause (Kreibig, 2010).

What is heart rate variability and why does it matter?
Heart rate variability is the beat-to-beat fluctuation in the interval between heartbeats. Its fast, respiration-linked component reflects parasympathetic control through the vagus nerve, and resting variability has been linked to the capacity to regulate emotion and attention, making it a widely used marker of self-regulation (Thayer & Lane, 2000).

What does the P300 event-related potential indicate?
The P300 is a large positive brain-wave component arising around 300 milliseconds after a rare, task-relevant stimulus. Its amplitude grows as the eliciting event becomes rarer and more significant, so it is read as an index of the attention and memory operations that update a model of the current context (Polich, 2007).

What is autonomic space?
Autonomic space is the model that represents sympathetic and parasympathetic activity on two independent axes rather than as opposite ends of one arousal dimension. It allows for reciprocal control, coactivation of both branches, and patterns in which one branch changes while the other holds, so the same heart rate can reflect very different autonomic states (Berntson et al., 1991).

Can a polygraph really detect lies?
No. The polygraph records autonomic arousal, which accompanies many emotions besides deception, so there is no physiological signature of a lie as such. A better-founded approach, the concealed information test, detects recognition of crime-relevant details rather than lying itself, and even it is not infallible (Ben-Shakhar & Elaad, 2003).

What is the orienting response?
The orienting response is the set of autonomic and behavioral changes, including a skin conductance response and a brief cardiac slowing, that a novel stimulus evokes and that fades as the stimulus is repeated and found inconsequential. Sokolov analyzed this habituation as the nervous system building an internal model and responding only to mismatches with it (Sokolov, 1963).

References