Weight Perception

Abstract

Weight perception is a form of perception: the sense of how heavy a lifted or supported object feels. It is not a direct read-out of physical mass but a construction the brain assembles from the forces a limb generates, the effort that generation costs, and prior expectations about how heavy the object should be. The psychophysics of heaviness follows Weber's law, the smallest detectable change in weight growing in proportion to the load, while the size-weight illusion — in which the smaller of two equally massive objects feels reliably heavier — shows the construction most vividly. Felt heaviness dissociates from the fingertip forces used to lift, and competing Bayesian and anti-Bayesian accounts contest how prior expectation and sensory evidence combine. This article sets out the psychophysics, the illusions, the cue integration, the perception-action split, and the theories.

Keywords: weight perception, size-weight illusion, heaviness, psychophysics

Pick up two boxes of the same mass, one large and one small, and the small one feels distinctly heavier — not slightly, but by a margin most people find hard to believe once told the two weigh the same. That everyday surprise is the entry point to weight perception, the sense of heaviness, which turns out to be one of the clearest cases in all of psychology where what is felt departs systematically from what is measured. Heaviness is not delivered to the mind by a scale in the hand; it is inferred from the muscular forces a lift demands, the sense of effort those forces cost, and expectations set by how the object looks, and because that inference can be fooled in lawful ways, weight perception has become a favoured laboratory for studying how the brain builds a percept from cues (Jones, 1986; Buckingham, 2014).

What Weight Perception Is

Weight perception is the perceptual estimate of how heavy an object is, formed when the object is lifted against gravity or supported in the hand. The physical quantity it tracks is mass — or, more precisely, the gravitational force an object's mass exerts — but the percept is not that quantity delivered raw. It is assembled from several signals: the load force the muscles must produce to move the object, cutaneous and proprioceptive afferents that report the resulting skin deformation and joint torque, and a centrally generated sense of effort that accompanies the motor command itself. Because these signals can be dissociated from true mass, felt heaviness can be pushed away from the physical value in reproducible ways, which is precisely what makes weight perception a useful window on perceptual construction (Jones, 1986).

The debate that runs through the field is how far heaviness rests on incoming sensory signals from the periphery versus on the outgoing motor command — an afferent account against an efference-copy account of the sense of effort. Both contribute: the felt heaviness of a load reflects the peripheral force and pressure signals it generates and the effort the lift is expected to cost, and the two can be teased apart experimentally (Jones, 1986). What unifies the modern treatment is the recognition that heaviness is an inference over cues, and that the illusions of weight are not malfunctions but the visible workings of that inference (Buckingham, 2014).

The Psychophysics of Heaviness

The oldest quantitative fact about weight perception is that discrimination is relative, not absolute. Ernst Heinrich Weber's early-nineteenth-century experiments on lifted weights established that the smallest detectable difference between two loads is not a fixed number of grams but a roughly constant fraction of the standard weight: a difference easy to feel against a light load becomes imperceptible against a heavy one. This proportionality — the just-noticeable difference growing with the standard — is the observation later generalised as Weber's law and made the founding datum of psychophysics. The Weber fraction for lifted weight is on the order of a few percent, so two loads must differ by roughly 2 to 5% before the difference is reliably felt, and this fraction is broadly stable across the middle of the usable range (Ross, 1969).

Relating the felt magnitude of heaviness to physical weight, rather than merely detecting differences, requires magnitude estimation, and here the governing relation is a power law. Asked to assign numbers to the apparent heaviness of lifted weights, observers produce judgments that grow as physical weight raised to an exponent greater than one, so apparent heaviness accelerates: doubling the physical load more than doubles the felt heaviness (Stevens & Mack, 1959). This places heaviness among the continua Stevens characterised with the power law, with an exponent near 1.45 that makes it one of the expanding sensory scales. The precision of weight discrimination is itself not fixed: it can improve when the discrimination is embedded in a task that supplies additional structure, and the same manipulations that distort felt heaviness can sharpen or blunt the ability to tell two weights apart (Wolf & Drewing, 2020). The first demonstration lets the reader vary a standard weight and read off the just-noticeable difference Weber's law predicts, watching the JND grow in proportion to the load.

The Size-Weight Illusion

The single most studied phenomenon in weight perception is the size-weight illusion: when two objects of identical mass but different size are lifted, the smaller feels substantially heavier than the larger. Augustin Charpentier described it formally in 1891, and it still carries his name as the Charpentier illusion; the priority and the original observation were reconstructed for modern readers by Murray and colleagues, who translated and analysed Charpentier's report (Murray, Ellis, Bandomir, & Ross, 1999). The effect is large and stubborn. It does not fade when the observer is told the two masses are equal, it survives lifting the objects repeatedly, and its magnitude can be a substantial fraction of the veridical weight, which is what marks it as a genuine perceptual illusion rather than a judgment error or a momentary surprise (Wolfe, 1898).

Two features of the illusion constrain any explanation of it. First, it is specifically an illusion of size relative to expectation: across the many weight illusions, the influence of size on felt heaviness is unusually strong and behaves differently from the influence of other object features, marking size out as a special cue rather than one contributor among equals (Saccone & Chouinard, 2019). Second, the illusion is robust and general across materials and conditions. A meta-analysis of the size-weight and the related material-weight illusions confirms a reliable, large effect and quantifies how its strength depends on the size and density contrasts involved, establishing that the phenomenon replicates far too consistently to be an artefact (Saccone, Landry, & Chouinard, 2019). The reliability of the illusion is exactly why it has become the standard test bed for theories of how heaviness is computed from cues (Buckingham, 2014). Figure 1 shows the arrangement that produces it.

Cue Integration

If the smaller object feels heavier, the natural question is which cue to size is doing the work — the seen volume, the felt volume, or something about how the object is grasped and moved. Ellis and Lederman separated visually specified volume from haptically specified volume and showed that both contribute: the illusion appears when size is given by vision alone and when it is given by touch alone, so the brain is combining volume cues from more than one channel in arriving at felt heaviness (Ellis & Lederman, 1993). This established weight perception as a genuine cue-integration problem rather than a quirk of the visual system. Table 1 sets out the cues the felt-heaviness computation draws on and the evidence that each contributes.

A stronger claim locates the driving variable more precisely. Amazeen and Turvey argued that heaviness in dynamic touch tracks the inertia tensor — the resistance an object offers to being rotated about the joints — rather than mass as such, tying felt weight to the mechanics of how the object is wielded (Amazeen & Turvey, 1996). Working from the opposite direction, Plaisier and Smeets equated the volumes of objects that differed in mass and found the illusion vanished, concluding that for the size-weight illusion it is the size contrast that matters and that, with size held constant, mass alone determines heaviness as it should (Plaisier & Smeets, 2012). Vision contributes even when it specifies not size but material: objects that merely look as though they are made of a dense material are lifted with more force and, initially, felt as heavier, showing that expectations about material feed the same heaviness computation (Buckingham, Cant, & Goodale, 2009). The second demonstration lets the reader hold mass fixed while changing an object's size, watching apparent heaviness invert with density, and then equate the volumes to make the illusion collapse.

Perception Versus Action

A decisive experiment split the felt heaviness of an object from the motor system's estimate of its mass. When a person first lifts an object, the grip and load forces they apply are scaled to how heavy they expect it to be, so a large object is gripped and lifted more forcefully than an equally massive small one. Flanagan and Beltzner tracked these forces across repeated lifts of the size-weight illusion stimuli and found that the fingertip forces corrected within a few lifts to the objects' true, equal mass — the motor system learned the real weight quickly — while the objects went on feeling as different in heaviness as ever. Perception and action had come apart: the sensorimotor system and the perceptual system were making independent estimates of the same object's weight, and only the motor one tracked reality (Flanagan & Beltzner, 2000).

This dissociation fits a broader division of somatosensory processing into a stream serving conscious perception and a stream serving the online control of action, each able to hold a different estimate of the same physical property at the same time (Dijkerman & de Haan, 2007). The finding also sharpens the puzzle the theories must solve: whatever produces the persistent illusion cannot be a simple failure to sense the true mass, because the motor system demonstrably senses it. The illusion lives in the perceptual estimate specifically, which is why accounts of it turn on how prior expectation is combined with sensory evidence to yield conscious heaviness rather than on the sensory signal alone (Buckingham, 2014). The third demonstration runs the lift sequence: load force adapts trial by trial toward the true mass while the perceived-heaviness trace stays stuck at its illusory value.

Bayesian and Anti-Bayesian Accounts

The most influential framing casts weight perception as Bayesian inference: the brain combines a prior expectation about an object's weight with the sensory evidence from lifting it, weighting each by its reliability. The trouble is that the size-weight illusion appears to run the wrong way for a simple Bayesian model. A large object carries a prior of being heavy, so if the percept were pulled toward that prior the large object should feel heavier — yet it feels lighter. The illusion is in this sense anti-Bayesian, biasing the percept away from the expectation rather than toward it, and Brayanov and Smith documented exactly this pattern, showing that the perceptual bias in the size-weight illusion is opposite in sign to the Bayesian bias seen in the motor system's force scaling (Brayanov & Smith, 2010).

The anti-Bayesian label was a challenge rather than a settlement, and the response was to build a Bayesian model that predicts the illusion in the right direction. Peters, Ma and Shams argued that the relevant prior is not over weight but over density: objects of a given size have an expected density, and a small dense object violates that expectation, so a hierarchical Bayesian observer that infers density and estimates weight from it reproduces the illusion without any anti-Bayesian machinery, making the effect a rational inference after all (Peters, Ma, & Shams, 2016). The two positions differ in what the brain is taken to expect — a weight tied to size, or a density tied to material and size — and the question is not closed. That the operative prior is learned rather than fixed is shown directly by training: after extended practice lifting a set of objects whose size and weight are made to correlate inversely — larger objects built to be the lighter ones — the illusion weakens and can reverse, so the size-weight prior is acquired from a lifetime of experienced correlations between how large things are and how heavy they prove to be rather than wired in (Flanagan, Bittner, & Johansson, 2008). What both positions share is the view that heaviness is the output of an inference combining prior and evidence, so that the same illusion which sharpens weight discrimination under some conditions is the trace of that inference at work rather than a defect in it (Wolf & Drewing, 2020).

Worked Example

The two central facts of weight perception become concrete when the numbers are worked, and the demonstrations reproduce this arithmetic. Take Weber's law first. If the Weber fraction for lifted weight is k = 0.08, then the just-noticeable difference is the fraction of the standard weight: JND = k times W. For a standard of W = 100 grams the smallest detectable increment is 0.08 times 100 = 8 grams; for W = 200 grams it is 0.08 times 200 = 16 grams; and for W = 400 grams it is 0.08 times 400 = 32 grams. The JND grows in strict proportion to the load, which is the whole content of Weber's law: a difference of 10 grams is easily felt against 100 grams but is below threshold against 400, because 10 is more than the 8-gram JND at 100 but less than the 32-gram JND at 400.

Now the size-weight illusion, where density makes the surprise quantitative. Density is mass divided by volume. Suppose three objects each have a mass of 500 grams but volumes of 1000, 500 and 250 cubic centimetres. Their densities are 500 / 1000 = 0.5, 500 / 500 = 1.0, and 500 / 250 = 2.0 grams per cubic centimetre. The three weigh exactly the same, yet the smallest is four times as dense as the largest, and it is that density contrast — the small object being far denser than its size leads one to expect — that drives it to feel the heaviest. Equate the volumes instead, giving all three objects the same size, and the density contrast vanishes along with the illusion, leaving mass as the only cue and the three feeling equally heavy, exactly as Plaisier and Smeets found (Plaisier & Smeets, 2012).

Discussion

Weight perception looks at first like the simplest of senses — a lift and a number — and turns out to be a constructed inference whose workings are exposed by its errors. The psychophysics fixes the ground: discrimination is relative, following Weber's law, and felt magnitude grows as a power of physical weight (Ross, 1969; Stevens & Mack, 1959). Onto that ground the size-weight illusion delivers its lesson, that felt heaviness is not physical heaviness but a function of cues to size, material and density, robust and large enough that it has resisted every attempt to explain it away as error (Murray et al., 1999; Saccone, Landry, & Chouinard, 2019). The perception-action dissociation then shows that the illusion is lodged specifically in conscious perception, since the motor system reads the true mass within a few lifts (Flanagan & Beltzner, 2000).

The live theoretical question is whether that construction is Bayesian. The illusion runs against a naive weight prior, which invited the anti-Bayesian reading, but a prior over density rather than weight recovers a rational inference, and the two accounts now compete on what the brain expects rather than on whether it infers at all (Brayanov & Smith, 2010; Peters, Ma, & Shams, 2016). Around this core sit further regularities the mature account must absorb — that heaviness carries a spatial connotation, the heavy being associated with downward and with the lower reaches of space, so that weight is bound up with representations beyond the purely mechanical (Vicovaro & Dalmaso, 2020). Read as inference rather than measurement, weight perception joins the other constructed senses, and its famous illusion becomes not a bug in the hand but a clear print of the computation the brain runs whenever it decides how heavy the world is (Buckingham, 2014).

Common Misconceptions

The size-weight illusion is just a momentary surprise that goes away once the weights are known to be equal.

It persists after the observer is told the masses are equal and after many repeated lifts, and its magnitude can be a large fraction of the true weight. That durability under full knowledge is exactly what qualifies it as a perceptual illusion rather than a corrigible judgment error (Wolfe, 1898; Flanagan & Beltzner, 2000).

Heaviness is simply the mass of the object read off by the hand.

Felt heaviness is inferred from muscular force, the sense of effort, and cues to size and material, not sensed as mass directly. Two objects of identical mass can feel very different, and an object that merely looks dense is felt as heavier, which a direct read-out of mass could not produce (Buckingham, Cant, & Goodale, 2009; Jones, 1986).

The illusion happens because the smaller object is lifted with too little force.

The fingertip forces correct to the objects' true, equal mass within a few lifts, yet the illusion in felt heaviness remains undiminished. Because the motor mis-scaling is gone while the illusion stays, the illusion cannot be a consequence of that mis-scaling; perception and action hold independent estimates (Flanagan & Beltzner, 2000).

Glossary

Anti-Bayesian bias.

A perceptual bias directed away from the prior expectation rather than toward it; the size-weight illusion is anti-Bayesian relative to a naive prior over weight, since the object expected to be heavy feels lighter.

Bayesian inference.

The combination of a prior expectation with sensory evidence, each weighted by its reliability, to form a percept; a leading framework for how the brain estimates heaviness from cues.

Charpentier illusion.

Another name for the size-weight illusion, after Augustin Charpentier, who formally described it in 1891.

Density.

Mass divided by volume; the contrast between an object's actual density and the density expected for its size is the variable that best predicts the size-weight illusion.

Efference copy.

An internal copy of a motor command; on the effort account of heaviness the brain gauges weight partly from the size of the command it issues to lift, not from afferent signals alone.

Inertia tensor.

The physical quantity describing an object's resistance to being rotated about an axis; proposed as the variable that dynamic-touch heaviness tracks when an object is wielded.

Just-noticeable difference.

The smallest change in weight that can be reliably detected; for lifted weight it is a roughly constant fraction of the standard, as Weber's law requires.

Load force.

The vertical force applied to lift an object against gravity; it is scaled in advance to expected weight and adapts to the true weight over the first few lifts.

Magnitude estimation.

A psychophysical method in which observers assign numbers to perceived intensity; applied to heaviness it yields a power-law relation to physical weight.

Material-weight illusion.

The illusion that, of two objects of equal mass and size, the one appearing to be made of a lighter material feels heavier; a cousin of the size-weight illusion driven by material rather than size cues.

Sense of effort.

The centrally generated feeling of exertion accompanying a motor command; one of the signals from which felt heaviness is constructed.

Size-weight illusion.

The reliable illusion that, of two objects of equal mass, the smaller feels heavier; the central phenomenon of weight perception.

Weber fraction.

The ratio of the just-noticeable difference to the standard stimulus; for lifted weight it is a few percent and roughly constant across the usable range.

Weber's law.

The principle that the just-noticeable difference is proportional to the magnitude of the standard stimulus; first established for lifted weights.

Key Researchers

Gavin Buckingham. University of Exeter (Department of Public Health and Sport Sciences); he studies the perception of heaviness and the sensorimotor control of lifting, and his review of weight illusions anchors the modern account of how visual cues to size and material bias both how objects are lifted and how heavy they feel. Faculty Page - Google Scholar - ORCID

Augustin Charpentier (1852-1916). University of Nancy; he formally described the size-weight illusion in 1891 — that of two objects of equal mass the smaller feels heavier — the founding observation of the field, still called the Charpentier illusion. Wikipedia

J. Randall Flanagan. Queen's University, Kingston (Department of Psychology and Centre for Neuroscience Studies); he showed that the perceptual and sensorimotor systems make independent predictions about weight, the fingertip forces adapting to an object's true mass even while it continues to feel illusorily heavy. Faculty Page - Google Scholar - ORCID

Lynette A. Jones. Massachusetts Institute of Technology (Department of Mechanical Engineering); she synthesised the psychophysics and physiology of the perception of force and weight, clarifying the roles of peripheral afferent signals and of the centrally generated sense of effort in judging how heavy a lifted object is. Faculty Page - Google Scholar - ORCID

Susan J. Lederman. Queen's University, Kingston (Professor Emerita of Psychology); a leading haptics researcher who dissected the cue combination behind the size-weight illusion, showing how haptically and visually specified volume each contribute to felt heaviness. Faculty Page - Google Scholar

Ernst Heinrich Weber (1795-1878). University of Leipzig; the founder of psychophysics, whose experiments on the smallest detectable difference between lifted weights established that the just-noticeable difference grows in proportion to the standard weight — the relation later formalised as Weber's law. Encyclopaedia Britannica - Wikipedia

Frequently Asked Questions

What is weight perception? Weight perception is the sense of how heavy a lifted or supported object feels. It is not a direct measurement of mass but an inference the brain builds from the muscular force a lift requires, the sense of effort it costs, and expectations set by how the object looks, which is why felt heaviness can differ systematically from physical weight (Jones, 1986; Buckingham, 2014).

What is the size-weight illusion? The size-weight illusion is the reliable experience that, of two objects of equal mass but different size, the smaller feels distinctly heavier. Augustin Charpentier described it in 1891, and it is still called the Charpentier illusion; it is the founding demonstration that felt weight is not the same as physical weight (Murray, Ellis, Bandomir, & Ross, 1999).

Why does the smaller object feel heavier? Because heaviness is judged relative to expectation. A small object is expected to be light, so when it weighs the same as a much larger one it is far denser than its size leads one to expect, and that density contrast makes it feel heavier. Equating the objects' volumes removes the contrast and the illusion disappears (Plaisier & Smeets, 2012; Peters, Ma, & Shams, 2016).

Does the size-weight illusion go away once the observer knows the objects weigh the same? No. It persists after the observer is told the masses are equal and after repeated lifting, which is what marks it as a genuine perceptual illusion rather than a mistaken judgment. Even the lifting forces adapting to the true mass do not dispel the felt difference (Wolfe, 1898; Flanagan & Beltzner, 2000).

What is Weber's law for weight? Weber's law states that the smallest detectable difference between two weights is a roughly constant fraction of the standard weight, so heavier loads must differ by more before the difference is felt. Ernst Heinrich Weber established this for lifted weights, making it the founding result of psychophysics (Ross, 1969).

How is felt heaviness related to physical weight? Felt heaviness grows as physical weight raised to a power greater than one, so apparent heaviness accelerates: doubling the load more than doubles how heavy it feels. This power-law relation places weight among the expanding sensory continua Stevens characterised (Stevens & Mack, 1959).

Do perception and action estimate weight the same way? No. When someone lifts the size-weight illusion objects, the fingertip forces correct to the objects' true, equal mass within a few lifts while the objects go on feeling different in heaviness. The sensorimotor and perceptual systems make independent estimates, and only the motor one tracks the real mass (Flanagan & Beltzner, 2000; Dijkerman & de Haan, 2007).

Is the size-weight illusion Bayesian or anti-Bayesian? Both descriptions have been defended. Against a naive prior that large objects are heavy the illusion runs the wrong way and looks anti-Bayesian, but a Bayesian model whose prior is over density rather than weight predicts the illusion in the right direction, so the debate now turns on what the brain expects rather than on whether it infers (Brayanov & Smith, 2010; Peters, Ma, & Shams, 2016).

References