Sensation begins when sensory receptors respond to physical or chemical energy and convert it into neural signals. This conversion is called transduction. Perception is the organisation and interpretation of sensory input into meaningful objects, events, and experiences. Sensation therefore concerns the registration and coding of stimulation, whereas perception concerns what that stimulation comes to represent.
The sensory systems differ in their specialised receptors and pathways, but they share a common sequence: a stimulus reaches a receptor, the receptor transduces the stimulus, neural activity is transmitted through the nervous system, and central processing contributes to the resulting experience. Detailed receptor anatomy and sensory pathways are covered in the relevant Biopsychology notes; this note focuses on the general psychological principles of sensation and perception.
| Sensory system | Main stimulus | General receptor process |
|---|---|---|
| Vision | Electromagnetic light | Photoreceptors respond to light energy |
| Audition | Pressure waves | Mechanoreceptors respond to vibration |
| Somatosensation | Pressure, stretch, temperature and tissue-damaging stimulation | Specialised receptors respond to mechanical, thermal and nociceptive input |
| Taste | Chemicals dissolved in saliva | Chemoreceptors respond to tastants |
| Smell | Airborne chemical molecules | Chemoreceptors respond to odorants |
| Vestibular sense | Head movement and gravity | Mechanoreceptors respond to movement of vestibular structures |
Sensory systems must represent both what kind of stimulation is present and how intense it is. Receptor type and pathway contribute to sensory quality, while changes in neural firing help code stimulus intensity. Sensory experience is not a simple copy of physical energy: the relation between stimulus intensity and subjective magnitude is studied by psychophysics.
Sensory adaptation is a reduction in responsiveness during continuous or repeated stimulation. A constant smell may become less noticeable after several minutes even though the stimulus remains present. Adaptation helps the sensory system remain responsive to change. It should be distinguished from habituation, a broader learning process in which behavioural responding declines after repeated exposure to a stimulus.
Psychophysics studies the relationship between measurable physical stimulation and psychological experience. Its central questions include how weak a stimulus can be and still be detected, how large a change must be before two stimuli can be discriminated, and how perceived magnitude changes as physical intensity increases.
| Concept | Definition | Formula / Key Fact |
|---|---|---|
| Absolute Threshold | Minimum stimulus intensity detected on 50% of trials | A statistical detection point, not an all-or-none boundary |
| Difference Threshold / JND | Smallest detectable difference between two stimulus intensities | Also called the just noticeable difference |
| Weber's Law | JND is approximately a constant proportion of the starting intensity | ΔI / I = k; smaller k = finer discrimination |
| Fechner's Law | Sensation magnitude increases approximately with the logarithm of stimulus intensity | S = k log I |
| Stevens' Power Law | Perceived magnitude follows a power function of stimulus intensity | S = kIⁿ; the exponent differs across sensory continua |
Weber's law means that discrimination depends on proportional rather than absolute change. If a person can just detect a 2-unit increase from a 100-unit standard, the Weber fraction is 0.02; a larger standard would usually require a larger absolute change to preserve roughly the same proportion. The value of k varies across sensory dimensions.
Fechner treated successive JNDs as equal subjective steps and derived a logarithmic relation between physical intensity and sensation. Stevens later used direct magnitude-estimation methods and proposed a power function. For some continua perceived magnitude grows more slowly than physical intensity, whereas for others it grows more rapidly.
| Method | Procedure | Main use |
|---|---|---|
| Method of Limits | Stimulus intensity is increased or decreased in ordered series until the response changes | Efficient estimation of a threshold; vulnerable to anticipation and habituation errors |
| Method of Constant Stimuli | Several stimulus intensities are presented repeatedly in an irregular order | Produces a psychometric function and a stable threshold estimate |
| Method of Adjustment | The participant directly adjusts the stimulus until it is just detectable or matches a standard | Fast and intuitive, but relatively dependent on participant strategy |
A psychometric function plots the probability of a response, such as stimulus detection, against stimulus intensity.
The absolute threshold concerns detection: whether a stimulus is noticed at all. The JND concerns discrimination: whether a change or difference between stimuli is noticed. Both are usually estimated across repeated trials rather than treated as perfectly fixed boundaries.
The Weber fraction is k = ΔI/I. A smaller Weber fraction indicates finer discrimination because a smaller proportional change is sufficient for a JND. Weber's law is an approximation and works best over a middle range of intensities rather than at every possible stimulus level.
Traditional threshold approaches summarise detection performance using a threshold, but actual decisions also depend on uncertainty and response strategy. Signal Detection Theory (SDT) separates two factors: the observer's sensitivity (d-prime), which reflects the ability to discriminate signal from noise, and the response criterion, which reflects how much evidence is required before reporting that a signal is present. Criterion placement therefore expresses response bias: a liberal or conservative tendency to report that a signal is present.
Measures the separation between the signal-plus-noise and noise-alone distributions. A higher d-prime means the observer can distinguish signal from noise more reliably.
The decision cut-off for saying "yes, a signal is present." A liberal criterion increases both hits and false alarms; a conservative criterion reduces both. Motivation, expectations, payoffs and consequences can shift the criterion without changing sensory sensitivity.
Move the criterion, then change sensitivity. Only one of them improves detection.
| Signal Present | Signal Absent | |
|---|---|---|
| Respond "Yes" | Hit | False Alarm |
| Respond "No" | Miss | Correct Rejection |
Two observers can have the same sensory sensitivity but different hit and false-alarm rates because they use different criteria. A radiologist may adopt a relatively liberal criterion when the cost of missing a tumour is high, accepting more false alarms in order to reduce misses. SDT therefore separates perceptual discriminability from decision strategy.
Perception is constructive: the nervous system does not simply reproduce the physical stimulus but organises sensory evidence in relation to context, prior knowledge, goals, and the structure of the environment. A flat retinal image can therefore support a stable three-dimensional percept, and the same ambiguous input can be interpreted differently in different contexts.
Contemporary accounts generally treat perception as an interaction between stimulus-driven and knowledge-driven processes rather than as exclusively bottom-up or top-down. Rumelhart's Interactive Activation model, for example, allows letter-level and word-level information to influence one another and helps explain the word-superiority effect.
Stimuli can sometimes influence later responding even when they are not consciously identified. Such effects are commonly studied using priming. Subliminal stimulation may produce limited changes in processing or preference, but it does not support claims of complex learning or powerful behavioural control without awareness.
A predisposition to perceive stimuli in a particular way, shaped by expectation, motivation, and past experience. Bruner and Minturn's (1955) classic demonstration showed that the same ambiguous figure could be read as "B" in an alphabetic context and "13" in a numeric context.
Max Wertheimer, Wolfgang Köhler, and Kurt Koffka, working in Germany in the early 20th century, challenged the structuralist view that perception is built by adding up elementary sensations. Their central claim was that organised percepts cannot be understood simply as the sum of isolated sensory elements.
Gestalt psychologists described principles of perceptual grouping: recurring tendencies by which elements are organised on the basis of properties such as proximity, similarity, continuity, closure and common fate. Grouping often occurs rapidly with little deliberate effort. Prägnanz is the broader Gestalt tendency toward stable, coherent and relatively simple organisation.
Prägnanz describes the general Gestalt tendency to organise ambiguous or complex input into stable and coherent forms. The individual grouping principles describe recurring ways in which elements are organised; they need not be treated as literal derivations from a single rule.
Elements close together in space tend to be grouped. Distance determines grouping before other properties: the same 36 dots now read as three column-pairs.
In the examPrägnanz is the broader Gestalt tendency toward stable, coherent organisation. The six tabs illustrate distinct grouping principles that are consistent with this general tendency rather than literal applications of one single rule.
Figure-ground segregation is a fundamental aspect of perceptual organisation. The visual field is often organised into a focal figure that appears shaped and foregrounded and a less prominent ground that appears to continue behind it. Figure-ground organisation can interact with other grouping processes rather than functioning as a universally prior serial stage. Rubin's Vase (1915) is the classic demonstration: the same contour supports two interpretations, two faces or a vase, and either region can become figure.
Not a single pixel changes when the display flips. What changes is which side of the contour the visual system assigns the boundary to — the property called border ownership. The region that owns the border gets the shape; the region that loses it becomes background.
Because the same contour supports two competing organisations, the display isbistable: either the vase or the faces can become figure, but the percept does not ordinarily stabilise as both interpretations at once.
In the examRubin (1915) described properties that distinguish figure and ground. The figurehas definite shape and often appears nearer, whereas the ground is less clearly shaped and appears to continue behind the figure.
| Researcher | Year | Contribution | Key Term |
|---|---|---|---|
| Wertheimer | 1912 | Studied apparent motion; the commonly tested apparent-motion/phi interval is 30–200 ms | Apparent motion; phi phenomenon |
| Wertheimer | 1923 | Formal statement of grouping principles | Proximity, Similarity, Continuity, Closure, Common Fate |
| Rubin | 1915 | Figure-ground perception and reversible figures | Rubin's Vase |
| Köhler | Insight learning; critique of structuralism | Insight | |
| Koffka | 1935 | Applied Gestalt principles to development | Principles of Gestalt Psychology |
Two stationary lights flash in alternation, yet the sequence can be experienced as motion. This is the general class of apparent-motion effects associated with Wertheimer's early Gestalt work.
This demonstration should not be read as reproducing pure phi at every setting. In historical terminology, phi refers to objectless motion, whereas beta movementrefers to the impression that an object itself travels between positions.
In the examFor UGC NET, retain 30–200 ms as the commonly tested apparent-motion/phi association. In stricter terminology, phi is objectless motion and betamovement is the impression that an object itself travels between positions.
Gestalt principles are primarily descriptive: they identify regularities in perceptual organisation but do not by themselves specify the underlying mechanism. Prägnanz can also become circular if "simple" is defined only by what observers already perceive as simple. Later research has sought mechanistic explanations in neural processing and the statistical regularities of natural scenes.
Object recognition must explain how physically different views can be identified as the same object or pattern. A handwritten "A" and a printed "A", for example, differ in exact shape but are readily recognised as instances of the same letter.
Template accounts propose that an incoming pattern is recognised by comparison with a stored whole-pattern representation. Simple template accounts are inadequate as a general theory of human object recognition because they would require many stored variants for changes in size, orientation and style, and they do not readily explain recognition of novel instances. Template matching can nevertheless describe recognition in constrained situations with highly standardised patterns.
Feature-based accounts propose that objects are analysed into components such as lines, edges, angles and curves. Hubel and Wiesel demonstrated orientation-selective responses in visual cortex. Simple and complex cells differ in their receptive-field properties, while cells historically termed hypercomplex are now usually described as end-stopped cells. These findings provide biological support for feature-based accounts of early visual analysis, but they do not by themselves establish a complete cognitive feature-matching theory of object recognition.
Oliver Selfridge's (1959) Pandemonium Model represented recognition as a hierarchy in which image units provide input, feature units respond to particular features, cognitive units represent candidate patterns, and a decision unit selects the strongest candidate.
Feature-based accounts handle many variations in size and position better than a rigid whole-pattern template and help explain why patterns sharing more components are more easily confused. Their main limitation is that a list of features alone may not specify how those features are spatially related.
Vision is also sensitive to different ranges of spatial frequency, from coarse global structure to fine detail. Selective adaptation to one frequency range reduces sensitivity mainly around that range, supporting multiple partially independent channels for early visual analysis. Spatial-frequency processing is therefore relevant to how form information is encoded, although it is not by itself a full theory of object recognition.
David Marr described progressively richer visual representations:
The proposal moves from local image structure toward a representation of three-dimensional form that can support recognition across changes in viewpoint.
Biederman proposed that many objects can be represented as structured combinations of simple volumetric components called geons. The arrangement of geons provides a structural description that can remain recognisable across substantial changes in viewpoint.
Damage to junction information can impair recognition more than comparable damage to the middle of contours because junctions are especially informative about part structure. View-based models instead propose that recognition depends partly on stored views and interpolation among them. Evidence for viewpoint costs, especially with novel or structurally similar objects, challenges a wholly viewpoint-independent account.
Gibson proposed that sufficiently rich environmental information can specify perceptual structure directly, without the inferential reconstruction assumed by constructivist theories. His ecological approach emphasises invariant information in the optic array and the relation between perception and action.
Brunswik's probabilistic functionalism describes perception and judgment as the use of multiple imperfect cues. A cue's ecological validity concerns how reliably it relates to the distal property in the environment, while cue utilisation concerns how strongly the observer relies on it. The lens model therefore emphasises probabilistic cue weighting, in contrast with Gibson's proposal that sufficiently rich environmental information can directly specify perceptual structure.
| Theory | Main idea | What it contributes | Main limitation |
|---|---|---|---|
| Template Matching | Compare input with stored whole-pattern representations | Useful for constrained, highly standardised patterns | Simple versions generalise poorly across novel views and variants |
| Feature-Based Accounts | Analyse objects into component features | Explains feature-confusion patterns and links to early visual selectivity | Feature lists alone may omit spatial relations |
| Spatial Frequency Channels | Analyse coarse-to-fine image structure in parallel frequency ranges | Strong account of an important stage of early visual analysis | Does not by itself specify object identity |
| Marr's Computational Theory | Build progressively richer representations from image structure to 3-D form | Clarifies stages and computational goals of vision | Later view-based findings challenge wholly viewpoint-independent recognition |
| Recognition by Components | Represent objects as structured geon assemblies | Explains robust part-based recognition across many views | Less effective for fine within-category discrimination |
| Gibson's Ecological Approach | Perceptual structure can be specified by invariant environmental information | Emphasises natural information and perception-action relations | Less explicit about recognition from impoverished or ambiguous input |
Apperceptive agnosia involves impaired construction or discrimination of perceptual form despite adequate basic sensory input. In associative agnosia, a sufficiently coherent percept can be produced but cannot be linked normally to stored meaning or identity. Prosopagnosia is a selective impairment in recognising familiar faces and is commonly associated with occipitotemporal face-processing regions.
The retina is a two-dimensional surface, yet we perceive a three-dimensional world. The visual system uses several sources of depth information, including monocular, binocular, movement and oculomotor cues.
Monocular depth information includes pictorial cues that can be represented in a flat image, movement-produced information such as motion parallax, and oculomotor information such as accommodation.
Parallel lines appear to converge toward a vanishing point as they recede into the distance.
The rate of convergence provides depth information and is heavily used in pictorial representation. The Ames room exploits perspective information by constructing a distorted room that projects a misleadingly regular image from the viewing point.
In the examAll six displayed cues are monocular pictorial cues and can therefore be represented on a flat surface. Keep them separate from movement-produced information such as motion parallax, the oculomotor cue of accommodation, and the binocular cues of retinal disparity and convergence.
| Cue | Type | How it works |
|---|---|---|
| Linear Perspective | Pictorial | Parallel lines appear to converge toward a vanishing point as they recede into the distance |
| Texture Gradient | Pictorial | Surface texture becomes finer and more densely packed with increasing distance |
| Interposition (Occlusion) | Pictorial | An object that blocks part of another is usually perceived as closer |
| Relative Size | Pictorial | For objects assumed to be similar in size, the smaller visual image usually signals greater distance |
| Height in Visual Field | Pictorial | For objects on a ground plane, position relative to the horizon contributes to perceived distance |
| Aerial Perspective | Pictorial | Distant objects tend to appear hazier and less saturated because of atmospheric scattering |
| Motion Parallax | Movement cue | During observer movement, nearby objects move across the visual field faster than distant objects |
| Accommodation | Oculomotor cue | Changes in lens focusing provide useful distance information mainly for relatively near objects |
The eyes are about 6.5 cm apart, so each receives a slightly different image. The visual system uses this difference to compute depth, a process called stereopsis. Disparity is greatest for nearby objects and decreases with distance.
When fixating a near object, both eyes rotate inward. The degree of inward rotation provides an oculomotor distance cue, most useful for relatively near objects.
Gibson proposed that sufficiently rich structure in the optic array can specify surface layout directly rather than requiring inferential reconstruction. Texture gradients were especially important in this account because systematic changes in texture provide information about the orientation and extent of surfaces.
Historical theories differed over how much depth perception depends on inborn visual organisation and how much is acquired through experience. The contrast remains useful for identifying theoretical traditions, but it should not be treated as a choice between two complete modern explanations.
Perceptual constancy is the tendency to perceive objects as having relatively stable properties such as size, shape and colour despite changes in the proximal sensory image produced by distance, viewing angle or illumination.
We usually perceive an object as remaining roughly the same physical size even when its visual angle becomes smaller with distance. The size-distance invariance relation expresses the idea that perceived size depends jointly on visual image size and perceived distance. As an exam-oriented proportional shorthand: perceived size ∝ visual angle × perceived distance. Errors in perceived distance can therefore contribute to errors in perceived size.
An afterimage has a fixed retinal extent, yet it appears larger when projected onto a surface perceived as farther away. Emmert's Law therefore illustrates the relation between perceived distance and perceived afterimage size and supports size-distance scaling.
| Type | What stays relatively constant | Despite changes in |
|---|---|---|
| Shape constancy | Perceived shape of an object | Viewing angle |
| Colour constancy | Perceived surface colour | Illumination spectrum and intensity |
| Brightness / lightness constancy | Perceived lightness of a surface | Overall illumination level |
| Location constancy | Perceived position of stationary objects | Eye and head movements that shift the retinal image |
Perceptual adaptation is recalibration to systematically altered sensory input. In Stratton's (1897) inverting-vision studies, adaptation became more effective as participants moved and interacted with the environment. Active interaction strongly facilitates recalibration because visual changes can be related to self-produced movement. The exam association between inverted vision and active movement should be retained, but active movement should not be interpreted as proving that no adaptation can occur under passive conditions.
Visual illusions are systematic divergences between perceptual experience and a relevant physical property of the stimulus. They are useful for studying the cues and assumptions that contribute to perceptual organisation, but many illusions have more than one proposed mechanism.
Gregory's misapplied-constancy account interprets the fins as depth-related cues: one configuration resembles a receding corner and the other a projecting corner.
On this account, inappropriate size-distance scaling is applied to the flat figure, so two physically equal shafts are perceived as different in length. This is an influential explanation rather than the only proposed mechanism for the Müller-Lyer illusion.
In the examSegall and colleagues reported cultural variation in Müller-Lyer susceptibility, contributing to the carpentered world hypothesis. This finding is commonly used as evidence that environmental experience can influence the strength of the illusion.
The converging lines provide linear-perspective information, so the upper bar is commonly interpreted as lying farther away in the implied scene.
A standard size-distance account proposes that a bar interpreted as farther away is perceptually scaled upward. This account captures an important exam association without implying that a single mechanism has been conclusively established.
In the examIn Gregory's account, Ponzo and Müller-Lyer can both be understood through depth-related or constancy-scaling interpretations. Treat this as an exam-relevant theoretical account, not as proof that the two illusions share one uniquely established mechanism.
The centre circle is judged in relation to its immediate neighbours: small surrounds make it look larger and large surrounds make it look smaller.
This is usually described as a size-contrast or contextual effect rather than as a depth-based constancy error. It is also known as the Titchener circles illusion.
In the examContrast this plate with common depth-scaling accounts of Müller-Lyer and Ponzo. Ebbinghaus is usually treated as a contextual size-contrast effect rather than misapplied size constancy.
| Illusion | Phenomenon | Common explanation / account |
|---|---|---|
| Müller-Lyer | Two equal lines can look different in length depending on the fins attached to them | Gregory's account treats the fins as depth-related cues that invite inappropriate size-constancy scaling; other explanations also exist |
| Ponzo | Two equal bars placed between converging lines can look different in size | A common account treats the converging lines as linear-perspective information, so the bar interpreted as farther is perceptually scaled upward |
| Ebbinghaus (Titchener) | A circle surrounded by small circles can look larger than an equal circle surrounded by large circles | Usually described as contextual size contrast rather than misapplied depth scaling |
| Moon Illusion | The moon often appears larger near the horizon than high in the sky despite nearly unchanged angular size | One influential apparent-distance account proposes that horizon distance cues alter perceived distance and therefore perceived size; the Moon illusion has several competing explanations |
| Ames Room | People at opposite corners of a deliberately distorted room appear very different in size | Perspective cues support an incorrect room geometry, so size judgments are calibrated to a misleading spatial interpretation |
| Necker Cube / Rubin's Vase | Ambiguous figures reverse between alternative organisations | Multiple perceptual organisations are compatible with the same input, producing bistable perception |
Gregory proposed that some geometric illusions arise when depth-related cues in a flat figure trigger scaling processes that are normally useful in three-dimensional scenes. The account is highly relevant to Müller-Lyer and Ponzo questions, but it should be presented as a theoretical explanation rather than as the uniquely established mechanism for every geometric illusion.
A stationary object may appear to move when its surrounding frame moves. The visual system uses surrounding structure as a reference, so relative motion can be assigned to the smaller enclosed object. A familiar example occurs when a neighbouring train begins to move and one's own stationary train briefly appears to move in the opposite direction.
Adaptation level (AL) is a reference level formed from recent and contextual stimulation. Judgments are made relative to this reference: a moderate stimulus may seem weak after stronger stimulation and strong after weaker stimulation. The theory therefore provides a general framework for context and contrast effects in judgment and perception.
Speech perception must recover meaningful linguistic units from a rapidly changing acoustic signal. Coarticulation means that the acoustic form of a phoneme varies with surrounding sounds, so there is no simple one-to-one acoustic pattern for each phoneme. Continuous speech also lacks perfectly clear acoustic boundaries between words.
Categorical perception refers to the tendency for continuously varying speech signals to be experienced as belonging to discrete phoneme categories, with relatively sharp category boundaries. Context also shapes what is heard. In phoneme restoration, a missing or masked speech sound can be perceptually supplied from lexical and sentence context. The McGurk effect shows that visual information from a speaker's mouth can alter the phoneme that is heard.
| Theory | Main claim | What it helps explain | Main challenge |
|---|---|---|---|
| Motor Theory (Liberman et al.) | Speech perception involves recovering intended articulatory gestures rather than matching invariant acoustic patterns | Links perception to articulatory structure and has historically been used to account for categorical speech perception | Normal speech perception is possible despite substantial variation in speech-production ability; the necessity of a specialised motor decoder is debated |
| TRACE (McClelland & Elman, 1986) | Feature, phoneme and word levels interact, with activation flowing both upward and downward | Phoneme restoration, lexical/context effects and the Ganong effect | The role and timing of top-down feedback remain debated |
| Cohort Model (Marslen-Wilson & Tyler) | Word onset activates a set of lexical candidates that narrows as more speech arrives | Rapid spoken-word recognition and the importance of early phonemes | Early versions were too vulnerable to misperceived word onsets; later revisions allowed more flexibility |
Direct Realist Theory (Fowler) likewise proposes that listeners perceive the articulatory gestures that generate speech, but frames this as direct perception of real environmental events rather than operation of a specialised speech module. It is useful as a contrast with Motor Theory; the precise mechanisms and evidence for direct recovery of gestures remain debated.
Among the laws of perceptual grouping, the law of simplicity is a tendency to:
| Theorist(s) | Contribution | Key association |
|---|---|---|
| Weber | Proportional relation between standard intensity and JND | Weber's Law / Weber Fraction |
| Fechner | Quantitative psychophysics; logarithmic scaling | Fechner's Law |
| Stevens | Direct magnitude estimation; power-function scaling | Stevens' Power Law |
| Wertheimer | Apparent motion and Gestalt grouping principles | Phi / Gestalt Principles |
| Rubin | Figure-ground organisation and reversible figures | Rubin's Vase |
| Gibson | Ecological approach to perception | Direct Perception / Affordances |
| Gregory | Constructivist theory and illusion accounts | Perception as Hypothesis / Misapplied Constancy |
| Hubel & Wiesel | Orientation-selective visual cortical responses | Simple, Complex, End-Stopped Cells |
| Selfridge | Hierarchical feature-recognition model | Pandemonium Model |
| Marr | Computational account of visual representation | Primal Sketch / 2½-D / 3-D Model |
| Biederman | Recognition by Components | Geons |
| Brunswik | Probabilistic functionalism and cue weighting | Lens Model / Ecological Validity |
| Helson | Context-sensitive reference level | Adaptation-Level Theory |
| Liberman et al. | Motor Theory of speech perception | Articulatory Gestures |
| McClelland & Elman | Interactive model of speech perception | TRACE |
| Marslen-Wilson & Tyler | Spoken-word candidate narrowing | Cohort Model |
| McGurk & MacDonald | Visual influence on heard speech | McGurk Effect |
| Warren | Contextual restoration of masked speech sounds | Phoneme Restoration |
| Stratton | Adaptation to inverted/rearranged vision | Active Movement / Recalibration |