Psychophysiologic Habituation

Abstract

Habituation is a response decrement that develops when a stimulus is presented repeatedly — the simplest and most widespread form of learning, seen from a sea slug's gill-withdrawal reflex to human orienting responses. Because a response can also wane through sensory adaptation or effector fatigue, habituation is defined not by the decrement alone but by a family of parametric properties — spontaneous recovery with rest, faster loss under weaker and more frequent stimulation, and dishabituation, the restoration of the waned response by a different stimulus — that its imitators do not share. Thompson and Spencer catalogued these properties, Groves and Thompson derived them from two opposed processes, decremental habituation and incremental sensitization, and Kandel traced the Aplysia decrement to a presynaptic depression of transmitter release. Three demonstrations develop the response-decrement curve, the dual-process net, and the synaptic mechanism.

Keywords: habituation, non-associative learning, dishabituation, sensitization, dual-process theory

Habituation is a behavioral response decrement that results from repeated stimulation and that does not involve sensory adaptation or motor fatigue (Thompson & Spencer, 1966). It is a form of non-associative learning: unlike conditioning, it requires no pairing of one event with another, only the repetition of a single stimulus, and for that reason it is treated as the most elementary case of learning and the one most broadly conserved across the animal kingdom. The phenomenon is easy to observe and easy to mistake for something else. A startle response to a sudden tone diminishes over successive tones; an orienting turn toward a novel light fades as the light recurs; a sea snail withdraws its gill less and less to a repeated touch. In each case the response wanes, but the waning is of interest only because it is not the trivial exhaustion of a sensor or a muscle. The evidence that habituation is a genuine change in the nervous system's disposition to respond, rather than an inability to respond, is that a change of stimulus restores the response at once — the operation of dishabituation — and that the response returns on its own after a period of rest. The sections below define habituation against its imitators, set out the parametric characteristics that Thompson and Spencer used to give it operational content (Thompson & Spencer, 1966; Rankin et al., 2009), develop the dual-process theory that derives those characteristics from two opposed processes (Groves & Thompson, 1970), and trace the phenomenon down to the orienting reflex (Sokolov, 1963) and to the synapse (Castellucci & Kandel, 1974).

What Habituation Is

A response can decline for uninteresting reasons. If a receptor is stimulated without pause its transduction machinery adapts and its output falls, and if a muscle contracts repeatedly it fatigues; in both cases the organism responds less because its periphery can no longer deliver the response, not because its nervous system has learned anything. Habituation is the response decrement that remains after these peripheral explanations are excluded (Thompson & Spencer, 1966). Harris had already drawn the distinction in an early review, insisting that a habituatory response decrement in the intact organism be separated from the receptor and effector changes with which it is easily confused (Harris, 1943). The exclusion is not a matter of assumption but of test. Sensory adaptation is ruled out because habituation is often stimulus-specific in a way adaptation is not, and because it depends on the interval between stimuli rather than on their total energy. Fatigue is ruled out decisively by dishabituation: after the response to a repeated stimulus has waned, presenting a different and usually stronger stimulus restores the original response immediately, which a fatigued effector could not produce. The response was therefore available all along; what changed was the organism's readiness to emit it to that particular stimulus.

This places habituation among the forms of non-associative learning, alongside sensitization, and separates it from the associative learning of classical and operant conditioning. Its interest is partly its ubiquity and partly its tractability. Because it is present in animals with nervous systems small enough to map neuron by neuron, habituation became the phenomenon through which the cellular basis of a behavioral change could first be pursued directly (Thompson & Spencer, 1966; Thompson, 2009). It is also of methodological importance far beyond the study of learning: the habituation of looking time is the workhorse measure of infant cognition, where the recovery of looking to a changed display — dishabituation — is read as evidence that the infant discriminates the change. In the National Library of Medicine's Medical Subject Headings this psychophysiological phenomenon is indexed as Habituation, Psychophysiologic (descriptor D006185), defined there as the disappearance of responsiveness to a repeated stimulation and explicitly not including drug habituation; the qualifier separates the response decrement studied here from the pharmacological tolerance the same word can name.

The Parametric Characteristics

Thompson and Spencer gave habituation operational content by listing nine parametric characteristics — regularities in how the decrement depends on the stimulus and its scheduling — that any adequate theory would have to reproduce, and that together distinguish habituation from adaptation and fatigue (Thompson & Spencer, 1966). The most important are as follows. Repeated stimulation produces a negatively accelerated decrease in the response, rapid at first and then slowing toward an asymptote. When the stimulus is withheld the response recovers over time, a spontaneous recovery that shows the decrement to be temporary. Repeated series of habituation and recovery make the habituation progressively more rapid, a potentiation of the effect. Habituation is faster and deeper when stimulation is more frequent — when the interstimulus interval is shorter — and when the stimulus is weaker; a sufficiently intense stimulus may show no habituation at all. The decrement generalizes to other stimuli in proportion to their similarity, an instance of stimulus generalization. And the presentation of another, typically stronger, stimulus restores the habituated response — dishabituation — an effect that itself habituates on repetition.

Rankin and an international group of investigators later revised and extended this list, sharpening several definitions and adding the distinction between a short-term habituation that recovers within seconds to minutes and a long-term habituation that persists for hours to days and follows different rules (Rankin et al., 2009). The revision matters because the two forms dissociate: massed, frequent stimulation yields rapid short-term habituation but poor long-term retention, whereas spaced stimulation habituates more slowly within a session yet produces durable long-term habituation — the spacing effect that recurs throughout the study of memory. The demonstration below realizes the short-term characteristics as a single decrement curve whose rate responds to stimulus intensity and interstimulus interval, with rest producing spontaneous recovery and a novel stimulus producing dishabituation. Table 1 lists the principal characteristics and what each one rules out or implies.

Table 1

Principal Parametric Characteristics of Habituation

Note. Adapted from the characteristics catalogued by Thompson and Spencer and revised by Rankin and colleagues (Thompson & Spencer, 1966; Rankin et al., 2009).

Dual-Process Theory

The characteristics do not all pull the same way. Weak stimuli habituate monotonically, but strong stimuli often make the response grow over the first few presentations before it declines, and a strong extraneous stimulus can raise responding to everything for a while. Groves and Thompson resolved this by proposing that repeated stimulation engages two independent processes whose algebraic sum is the observed behavior (Groves & Thompson, 1970). One is habituation proper, a decremental process that builds up in the specific stimulus-response pathway. The other is sensitization, an incremental process that reflects a transient rise in the organism's general state of excitability or arousal and is not specific to the stimulus that produced it. The response actually emitted is the baseline modified by the difference between them.

The theory earns its keep by deriving the intensity effect that a single-process account cannot. For a weak stimulus the sensitization process is negligible, so the net function is simply the habituation decrement and the response falls from the start. For a strong stimulus the sensitization process is large but transient: it rises quickly over the first few presentations and then decays, while the habituation process grows more slowly but is enduring. Early on, sensitization outruns habituation and the net response climbs above baseline; as sensitization fades and habituation accumulates, the net response turns and falls. The result is an inverted-U time course — an increment followed by a decrement — that appears only for stimuli strong enough to recruit sensitization. Figure 1 shows the two component processes and their sum, and the demonstration lets the stimulus intensity be varied so that the net curve passes from a monotonic decrement to the inverted-U as sensitization is recruited.

The Orienting Reflex and the Neuronal Model

Habituation of the orienting reflex — the complex of attentional and autonomic adjustments a novel stimulus evokes — gave Sokolov the material for an influential cognitive theory of the phenomenon (Sokolov, 1963). Sokolov observed that the orienting reflex habituates to a repeated stimulus but returns whenever any parameter of the stimulus changes, including a change to a weaker stimulus or the omission of an expected stimulus. A simple decrement in a reflex pathway cannot explain a response to the absence of a stimulus. Sokolov therefore proposed that the nervous system builds an internal neuronal model of the repeated stimulus — a stored representation of its properties — and compares each incoming stimulus against that model. So long as the stimulus matches the model the orienting reflex is suppressed; any mismatch, whether the new stimulus is stronger, weaker, or simply absent, releases the reflex. Habituation on this account is not the weakening of a pathway but the construction of an expectation, and dishabituation is the detection of a discrepancy from it.

The comparator idea reframes habituation as an active, model-building process and connects it to attention and prediction rather than to fatigue. It anticipates the modern view of the brain as an organ that predicts its input and responds to prediction error, and it explains why habituation is so sensitive to the fine structure of the stimulus: the richer the model, the more kinds of change can violate it. Where the parametric approach describes habituation from the outside and dual-process theory decomposes it into opposed tendencies, the neuronal-model theory specifies what the decremental process is computing — the match between a stimulus and a learned representation of it.

Neural Mechanisms in Aplysia

Because the marine snail Aplysia has a nervous system of large, identifiable neurons, its gill-withdrawal reflex became the preparation in which the cellular basis of habituation was established. A light touch to the animal's siphon makes it withdraw the gill; when the touch is repeated the withdrawal habituates, and it shows the parametric signature of genuine habituation, including spontaneous recovery and dishabituation by a strong stimulus to another site (Pinsker et al., 1970). Kandel and colleagues then localized the change. Recording from the sensory neuron that carries the siphon stimulus and the motor neuron that drives the gill, Castellucci and Kandel found that the decline in behavior tracked a decline in the size of the excitatory postsynaptic potential the sensory neuron produced in the motor neuron, and a quantal analysis showed the cause to be presynaptic: the sensory neuron released progressively fewer quanta of transmitter with repetition, while the motor neuron's sensitivity to transmitter was unchanged (Castellucci & Kandel, 1974).

This homosynaptic depression — a decrement intrinsic to the repeatedly active synapse itself — is the mechanism of short-term habituation, and it locates the memory of the repeated stimulus in a specific, mapped synaptic connection. Dishabituation and sensitization, by contrast, are heterosynaptic: a strong stimulus elsewhere recruits modulatory interneurons that facilitate transmitter release at the same terminals, which is why an independent process can restore the depressed response, exactly as dual-process theory requires. The distinction between short-term and long-term habituation also has a cellular counterpart, the long-term form depending on altered gene expression and a lasting reduction in the number of synaptic connections rather than a transient change in release. The demonstration below traces the sensory-to-motor synapse across repeated activations, showing the postsynaptic potential shrink as release falls and recover when a sensitizing input is applied.

Habituation, Memory, and Priming

If habituation is a form of memory, its properties should follow from the properties of the memory that stores the repeated stimulus. Whitlow tested this directly in the vasomotor response of rabbits, showing that short-term habituation has the signature of a short-term memory: the decrement to a tone depended on the recency of the previous tone and was disrupted by an interpolated distractor, decaying over the same interval as other short-term memories (Whitlow, 1975). Habituation, on this evidence, is not a fixed drop in synaptic efficacy but the momentary consequence of the stimulus being already represented in an active memory.

Wagner built this insight into a general theory of habituation and memory (Wagner, 1979). In his account a stimulus is processed less — and evokes a smaller response — to the degree that it is already represented in a limited-capacity short-term store, that is, to the degree that its representation is primed. A stimulus can be primed in two ways: self-generated priming, when a recent presentation of the same stimulus still occupies the store, which yields the short-term, recency-dependent habituation Whitlow measured; and retrieval-generated priming, when associative cues call the stimulus representation into the store in advance, which yields a longer-lasting, context-dependent decrement. The theory thereby unifies non-associative and associative learning under one mechanism — the priming of representations — and derives the difference between short-term and long-term habituation from the difference between the two routes to priming. It also returns habituation to where Sokolov placed it, among the processes by which the nervous system anticipates its input, differing chiefly in supplying the memory mechanics that make the anticipation work.

Worked Example

The short-term characteristics can be captured by a single response-decrement model, and working it through shows how the parametric properties fall out of it. Let the response to the nth presentation be R(n) = A + (R₀ − A)e^(−λn), where R₀ is the response to the first presentation, A is the asymptotic level to which the response falls, and λ is a habituation rate that increases as the stimulus becomes weaker or more frequent. Take R₀ = 100, A = 15, and a reference rate λ = 0.30. The response then declines in a negatively accelerated way: 100 at the first presentation, 77.97 after one repetition, 61.65 after two, 49.56 after three, 33.97 after five, and 19.23 after ten, the steps shrinking as the curve bends toward its asymptote of 15. The decrement is half complete — the response is halfway from R₀ to A — at n = ln 2 / λ = 0.693 / 0.30 = 2.31 presentations, a compact index of habituation rate.

The frequency characteristic is the rate's dependence on the interstimulus interval. Holding the presentation count at five, a long interval (λ = 0.15) leaves the response at 55.15, a medium interval (λ = 0.30) at 33.97, and a short interval (λ = 0.60) at 19.23: more frequent stimulation habituates faster and deeper, exactly the Thompson–Spencer regularity. Spontaneous recovery is then a relaxation of the decrement during rest. After the tenth presentation the response sits at 19.23; withholding the stimulus recovers it toward R₀ by a factor 1 − e^(−μD) for a rest of duration D, so with μ = 0.5 a rest of D = 5 recovers 91.8 percent of the lost response, returning it to 93.37, while a longer rest returns it essentially to 100. Dishabituation short-circuits the wait: a strong different stimulus resets the effective presentation count, so the next presentation of the original stimulus again evokes near-R₀ responding without any rest at all — the operation that distinguishes this curve from the flat exhaustion of fatigue.

The dual-process theory adds the one behavior this single decrement cannot show. Modelling the net response as a baseline plus a transient sensitization process minus the growing habituation decrement, a weak stimulus (little sensitization) peaks at the first presentation, n = 0, and declines monotonically, whereas a strong stimulus (large, transient sensitization) peaks two presentations later, at n = 2 — an interior maximum above baseline — before habituation drags it down. That shift of the peak away from the first presentation, obtained here from the same subtraction Groves and Thompson proposed, is the inverted-U that marks a stimulus strong enough to sensitize (Groves & Thompson, 1970).

Discussion

Habituation occupies a peculiar position in the science of learning: it is the simplest thing an organism can learn and, partly for that reason, the one whose mechanism has been pursued furthest. Its history is a steady sharpening of a single question — what makes a response wane — from a behavioral definition that had to be defended against adaptation and fatigue, to a catalogue of parametric characteristics with operational teeth (Thompson & Spencer, 1966; Rankin et al., 2009), to a decomposition into opposed processes (Groves & Thompson, 1970), and finally to a mapped synapse whose transmitter release declines with use (Castellucci & Kandel, 1974). What holds the levels together is dishabituation. It is the behavioral test that excludes fatigue, the phenomenon that forces dual-process theory to posit an independent incremental process, the discrepancy that releases Sokolov's model-matched orienting reflex (Sokolov, 1963), and the heterosynaptic facilitation that restores the depressed synapse (Pinsker et al., 1970). A single operation therefore does the same explanatory work at every level of analysis, which is why habituation has served as a bridge between behavior and its cellular basis. The later convergence with memory theory (Whitlow, 1975; Wagner, 1979) completes the picture by making habituation continuous with the rest of learning: a response wanes because its stimulus is already represented, whether by recent occurrence or by associative retrieval, and the organism that has built such a representation is, in the plainest sense, one that has learned (Thompson, 2009).

Glossary

Below-zero habituation.

Continued stimulation after the response has reached its asymptote, which produces no further visible decrement but slows later spontaneous recovery, showing that the underlying process keeps building.

Dishabituation.

The restoration of a habituated response by the presentation of a different, usually stronger, stimulus; the defining test that separates habituation from sensory adaptation and fatigue.

Dual-process theory.

Groves and Thompson's account in which observed behavior is the algebraic sum of an independent decremental habituation process and an incremental sensitization process.

Habituation.

A behavioral response decrement produced by repeated stimulation that is not attributable to sensory adaptation or motor fatigue; the simplest and most widespread form of learning.

Homosynaptic depression.

A decline in synaptic transmission intrinsic to the repeatedly active synapse itself, caused presynaptically by reduced transmitter release; the cellular mechanism of short-term habituation in Aplysia.

Interstimulus interval.

The time between successive presentations of the stimulus; shorter intervals produce faster and deeper short-term habituation but poorer long-term retention.

Long-term habituation.

A durable decrement lasting hours to days, favored by spaced stimulation and, at the synapse, by altered gene expression and a lasting reduction in synaptic connections.

Neuronal model.

In Sokolov's theory, an internal representation of the repeated stimulus against which incoming stimuli are compared; a match suppresses the orienting reflex and any mismatch releases it.

Non-associative learning.

Learning that results from repeated exposure to a single stimulus, requiring no pairing of events; habituation and sensitization are its two forms.

Orienting reflex.

The complex of attentional and autonomic adjustments elicited by a novel stimulus; its habituation and change-driven recovery motivated Sokolov's comparator theory.

Potentiation of habituation.

The finding that repeated series of habituation and recovery make each successive habituation more rapid and pronounced.

Sensitization.

An incremental, non-specific increase in responsiveness produced by a strong or noxious stimulus; the process opposed to habituation in dual-process theory and the basis of dishabituation.

Sensory adaptation.

A decline in receptor output under continued stimulation, a peripheral change distinct from habituation and excluded by dishabituation and by habituation's dependence on interval rather than energy.

Spontaneous recovery.

The return of a habituated response over time when the stimulus is withheld, demonstrating that the decrement is temporary.

Stimulus generalization.

The transfer of the habituated decrement to other stimuli in proportion to their similarity to the repeated one, indicating a central rather than a receptor-level change.

Stimulus specificity.

The property that habituation attaches to the particular repeated stimulus, so that a change in the stimulus restores the response; a hallmark separating it from general fatigue.

Key Researchers

Eric R. Kandel (b. 1929). Professor at Columbia University and Nobel laureate; with his colleagues he established the cellular and synaptic mechanisms of habituation and dishabituation in the Aplysia gill-withdrawal reflex. Faculty Page - ORCID - Wikipedia

Catharine H. Rankin. Professor of Psychology at the University of British Columbia; she led the international consensus that revised and extended the behavioral characteristics of habituation and studies its genetics in C. elegans. Faculty Page - ORCID - Google Scholar

Evgeny N. Sokolov (1920-2008). Psychophysiologist at Lomonosov Moscow State University who formulated the orienting reflex and the neuronal-model, or comparator, theory of habituation. Wikipedia

Richard F. Thompson (1930-2014). Behavioral neuroscientist at the University of Southern California who, with W. A. Spencer, specified the parametric characteristics of habituation and, with P. M. Groves, the dual-process theory. Google Scholar - Wikipedia

Allan R. Wagner (1934-2018). Psychologist at Yale University who developed the SOP priming theory of habituation and memory, uniting non-associative and associative learning under the priming of stimulus representations. Faculty Page - Wikipedia

Frequently Asked Questions

What is habituation?
Habituation is a behavioral response decrement that develops when a stimulus is presented repeatedly and that is not caused by sensory adaptation of the receptors or fatigue of the effectors; it is regarded as the simplest and most widespread form of learning (Thompson & Spencer, 1966).

How is habituation different from sensory adaptation and fatigue?
Adaptation is a decline in receptor output and fatigue a decline in effector capacity, whereas habituation is a central change in the readiness to respond, shown by the fact that a different stimulus immediately restores the response, which a tired sensor or muscle could not do (Thompson & Spencer, 1966).

What is dishabituation?
Dishabituation is the restoration of a habituated response by presenting a different and usually stronger stimulus; it is the operation that proves the response was still available and, in dual-process terms, reflects a separate sensitization process (Groves & Thompson, 1970).

What are the main characteristics of habituation?
They include a negatively accelerated response decrement, spontaneous recovery during rest, faster and deeper habituation under weaker and more frequent stimulation, stimulus generalization, and dishabituation, a set later revised and extended by an international group (Rankin et al., 2009).

What is dual-process theory?
Dual-process theory holds that repeated stimulation drives two independent processes, a decremental habituation process and an incremental sensitization process, whose sum is the observed behavior; it explains why strong stimuli can rise before they fall (Groves & Thompson, 1970).

What is the orienting reflex and how does it habituate?
The orienting reflex is the set of attentional and autonomic adjustments a novel stimulus evokes; Sokolov argued that it habituates as the brain builds an internal model of the stimulus and returns whenever an incoming stimulus mismatches that model (Sokolov, 1963).

What happens in the brain during habituation?
In the Aplysia gill-withdrawal reflex the behavioral decrement reflects a presynaptic reduction in the transmitter released by the sensory neuron onto the motor neuron, a homosynaptic depression localized to the repeatedly active synapse (Castellucci & Kandel, 1974).

Is habituation a form of memory?
Yes; short-term habituation depends on the recency of prior stimulation and is disrupted by distractors like other short-term memories, and Wagner's theory treats it as the reduced processing of a stimulus already represented in an active store (Whitlow, 1975).

References