# Tactile illusions

Post-publication activity

Curator: Vincent Hayward

Tactile illusions are found when the perception of a quality of an object through the sense of touch does not seem to be in agreement with the physical stimulus. They can arise in numerous circumstances and can provide insights into the mechanisms subserving haptic sensations. Many of them can be exploited, or avoided, in order to create efficient haptic display systems or to study the nervous system.

## Contents

### All senses, including touch, are subject to illusions

It is sometimes assumed that vision is the main source of perceptual illusions and that, in contrast, touch is not subject to surprising perceptual phenomena. This belief is ancient. George Berkeley (1685-1753), Étienne Bonnot de Condillac (1714-1780), and others of this era frequently referred to touch as the provider of the 'truth' to the other senses. While this observation appears to be borne out frequently in everyday life, touch is similarly subject to ambiguous or conflicting sources of information, which, as in vision, audition, and other sensory inputs, provide circumstances in which touch-based illusions can arise. It could be that tactile illusions simply go unnoticed more frequently.

Like the systems subserving audition, vision, vestibular inputs, and taste/olfaction, (all of which are subject to illusions) the somatosensory system has evolved to solve perceptual problems quickly and reliably, subject to constraints that are physical (i.e., skin mechanoreceptors cannot be at a small distance from the surface comparatively to their size), physiological (i.e., neural computation is powerful but limited; mechanoreceptors must have a refractory period), and metabolic (i.e., only a proportion of afferent fibers from the periphery to the brain can be myelinated) in origin.

What is an illusion, visual, tactile, or otherwise? Gregory (1997) wrote that illusions are difficult to define. The commonly adopted definition of an illusion is that it is a discrepancy between perception and reality. This definition rapidly leads to the unsatisfactory conclusion that all percepts are illusions since a percept---which is a brain state---is always discrepant with a stimulus---which is a physical object. In fact these two notions cannot even be compared. For this reason, a simple operational definition which has the advantage of comparing things of the same nature was proposed (Hayward, 2008a). An illusion is a percept that arises from a stimulus combining two separable components. One component is fixed and the observer attends to it. What makes it an illusion is that the perception of this component is strongly contingent on the variation of a second component, perplexing the person made aware of the unchanging component of the stimulus. For instance, the moon disk appears larger when viewed close to the horizon than up in the sky. The fixed component is its angular size and the variable component is its elevation. The variable component can also be an internal state of the brain. The Necker cube illusion and other rivalry-based illusions are percepts that change through time according to brain states variations, although the stimulus is invariant.

Several surveys on tactile and haptic illusions were published recently (Hayward, 2008a; Bresciani et al., 2008; Lederman and Jones, 2011), describing several dozens of categories. The word 'haptic' is often used to refer to touch sensations that involve a motor component. According to Grunwald and John (2008) the term 'haptic' was coined by Max Dessoir (1867-1947), who was in need of a counterpart term for the terms 'optic' and 'acoustic'. In the past few years, the rate of discovery of new tactile and haptic illusions has increased greatly, indicating renewed interest in the subject, with more and better informed web-based resources than in the recent past, although much catching up remains to be done.

A particular aspect of haptic perception in humans, like in most animals, is that the whole body is a mechanically sensitive system. Of course, some species have developed specialized sensing organs: whiskers for rodents, pinnipeds, and other mammalians; dextrous fingers for primates, procyonids, and other families; scales for reptiles, crocodilians, and others; antennae, cuticles for insects and arthropods; spines in echinoids; and so on. In many cases these sensory organs are appendages with motor capabilities. Nevertheless, the whole body of an animal is, to some extent, mechanosensitive. Superficial mechanoreceptors are found in hair follicles, skin, scales, lips, cuticles; and deep receptors are found in muscles, tendons, ligaments, as well as other connective tissues, providing a great diversity of overlapping sensing options. What these sensing options have in common is that they all inform the brain of the mechanical state of the tissues in which they are embedded, but never provide direct information about the contacting objects. This information is always mediated by the laws of mechanics that govern the change of the mechanical state of tissues subject to internal and external loads.

Haptics and tactile sensing is thus the province of mechanics where the assumption of rigidity, which for simplification we are so easily inclined to adopt, is misleading even if we assume for the sake of analysis that the stimulated tissues take a quasi-static state (Hayward, 2011). The consequence is that we instinctively adopt finite-dimensional notions such as 'force' (which has meaning only for ideal 'point masses') or 'pressure' (which can be sensed only by compressible organs; since tissues are incompressible, the somatosensory system cannot sense pressure). These abstract notions are certainly convenient for helping our understanding of the tactile functions but have probably no significance for the brain. If we abandon these simplifications then the occurrence of haptic and tactile illusions---or surprising perceptual behaviors as defined earlier---can be expected. For example, a given load, which we normally express as a force, or a given displacement of a solid, which we normally express as a distance, do not correspond univocally to mechanical states of the body.

What follows is a selection of tactile perceptual phenomena that undoubtedly merit the status of being an illusion and that teach something specific about how the mechanical properties of objects are perceived by humans coarsely organized according to their likely attribution from more central to more peripheral haptic neural processing.

### Haptic perception interacts with other senses

When discussing haptic and tactile illusions it must always be kept in mind that senses rarely operate in isolation and that they all interact with each other in the formation of perceptual estimates and judgements. Many studies have shown that touch interacts with taste, vision, and audition, and thus specific perceptual effects are elicited when these interactions are strong. Here is a very classic and powerful example of such interactions which can easily be demonstrated in a classroom or elsewhere. Procure two graspable boxes of similar appearance but of different sizes as illustrated in Figure 1 and arrange them so they have the same mass. When asked to judge the relative heaviness of the two blocks, most people will be convinced that the smaller is heavier than the larger block. This effect has been known for more than a century and is frequently termed the Charpentier illusion (Charpentier, 1891), or more commonly the 'size-weight illusion'. The effect is not small (it can be of 20% or more of difference in judgement of heaviness) and has been, and continues to be, the subject of a very large number of studies that appear at a rate that does not seem to subside. Despite numerous attempts, a principled explanatory mechanism for its occurrence remains to be found.

Figure 1: A convenient setting to demonstrate the Charpentier Illusion.

Figure 1 shows two boxes or wood blocks that are easily graspable are arranged to have the same mass, viz. 200 g. Here they differ by one dimension only (say, 30 mm versus 90 mm) which make it possible to show, using the same blocks in the dark, that the two objects do cause a similar sensation of heaviness if the grasp is carefully executed in order to conceal information about their size difference. Conversely, the same blocks can be used to show that, in the absence of vision, the haptic finger-span size estimation method gives similar information to the brain as vision, which results in a similar illusion.

An uncontroversial aspect of the 'size-weight illusion' is the role played by prior experience. The majority of the available, numerous explanations that have been discussed for now a century give a central role to expectation based on prior information (Ross, 1966, Buckingham, 2014), an hypothesis that has received strong support with the demonstration that the effect can be inverted after sufficiently long practice (Flanagan et al., 2008).

There are numerous other haptic illusions that can arise from interactions between vision and touch, and this is also true of touch and audition. It is worth describing a representative demonstration of such interactions, as one example among many. Because frictional interactions between solids are generally accompanied by acoustic emissions that can be heard and because the vibrations of the source of emission can also be felt, audition and touch are in a position to collaboratively determine the mechanical characteristics of surfaces sliding against one another. Specifically, the glabrous skin of our hands---the skin inside the hands that we use to interact with objects---is covered by a layer of keratin, a material that has strong affinity with water. The mechanical properties of keratin change profoundly with hydration and so do its frictional properties (Johnson et al., 1993; Adams et al., 2013). As a result, the frictional sound made by rubbing hands is a direct function of their moisture content. If the sound emitted by rubbing hands is artificially modified, then the sensation of hand dryness is also modified (Jousmäki and Hari, 1998).

Figure 2: Equipment needed to observe audio-tactile interactions.

The set-up shown in Figure 2 can be used to demonstrate interactions between audition and touch. One needs a microphone (directional) to pick up an auditory scene, such as rubbing hands; a frequency equalizer (analog or digital); headphones (closed) to reproduce the modified scene. The high frequencies characteristic of frictional sounds can be enhanced or attenuated, affecting tactile perception. The perception of other frictional interactions will be affected similarly (Guest et al., 2002), most notably chalk against a blackboard, etc.

In this subsection, we have seen two examples, selected from many, where sensory information supplied by different senses interfered sufficiently to give the resulting percept an illusory quality, suggesting that a fundamental type of brain mechanism is the fusion of sensory information to extract a single object property such as weight, size, distance, numerosity, movement, mobility, wetness, softness, smoothness, and so on.

### Similarities of certain illusions across senses

In some cases of illusory perceptual phenomena there is a remarkable analogy between perceptual effects across senses, suggesting that certain brain mechanisms, even neural circuits, are shared by the senses, sometimes in surprising ways (Konkle et al., 2009). In vision there are many well-known effects arising from viewing certain line drawings (e.g. Delboeuf, Bourdon, Ebbinghaus, Müller-Lyer, Poggendorff, or Ponzo illusions). Interestingly, most of these visual illusions also operate in haptics when the figure represented as a raised drawing is explored with the finger (Suzuki, K. and Arashida, R, 1992).

The interpretations of these visual illusions frequently appeal to brain mechanisms engaged in resolving ambiguities introduced by optical projections (Howe and Purves, 2005, Wolfe et al., 2005). It is therefore surprising that these illusions also operate in touch (albeit not always as stably), since visual projections arise from the laws of optics and haptic projections come from self-generated movements (Hartcher-O'Brien et al., 2014). In contrast, explanations based on the anisotropy of fundamental sensory discrimination thresholds could apply in the two modalities (Heller et al., 1997; Mamassian and de Montalembert, 2010).

Figure 3: Geometrical visual illusions operate with touch.

The so-called visual vertical-horizontal illusion exemplified in the Figure 3 is a good representative example. For most people, the vertical segment appears to be longer than the horizontal one. They have the same length. Next, procure a page-size cardboard sheet and glue two 200 mm sticks on it, as indicated (chopsticks cut at length will do). Blindfolded exploration of the sticks will cause most people to feel, similarly to vision, that the vertical stick is longer than the horizontal one.

To mention another class of illusions that is common to all three non-chemical modalities, the so-called "tau effect" stands out. If two stimuli localized in time and in space are attended to, in all modalities: in visual space (Benussi, 1913), in auditory tonal space (Cohen et al., 1954), in auditory physical space (Sarrazin et al., 2007), on the skin (Gelb, 1914; Helson, 1930), the perceived distance between those stimuli depends on their temporal separation. A shorter time separation corresponds to a smaller perceived spatial separation. The reverse is also true and is called the "kappa effect" (Cohen et al., 1953). Numerous studies have been conducted about these and related phenomena, and the most commonly adopted approach to explain them is to evoke brain mechanisms aimed at coping with moving sources of stimulation in the presence of uncertainty (Goldreich, 2007). If the reader is interested in replicating any of these effects with electronically controlled stimuli, it is strongly advised to avoid employing the type of vibrator employed in consumer devices, particularly those based on eccentric motors, because their poor temporal resolution precludes the production of sufficiently brief stimuli.

Lateral inhibition is another neural computational principle that is shared by all senses and that can be invoked to explain universal interactions between intensity and proximity (von Békésy, 1959). Thus, apparent motion, which is tightly connected to the latter interaction, operates in touch as in other sensory modalities (Wertheimer, 1912; Bregman, 1990; Gjerdingen, 1994) by modulating the relative intensity of simultaneous stimuli that are separated in space (von Békésy, 1959). In the same vein, the permutability of amplitude and duration of short stimuli seems to be a general phenomenon (Bochereau et al., 2014). Perceptual rivalry can likewise be demonstrated in all three sensory modalities (Carter et al., 2008), so does the phenomenon of capture where the localization of a stimulus in space by one sensory modality is modified by synchronous inputs from other sensory modalities (Caclin et al., 2002), as well as the family of attentional and change blindness phenomena (Gallace et al., 2006).

The types of tactile and haptic illusions discussed so far (namely, interactions between sensory modalities, geometrical illusions, or space time interactions) share the quality of being classical in the sense that they have been known for a century or so. In the foregoing, haptic illusions that have been described more recently are described.

### Order of differences: the particular multi-scale nature of touch

Sensory processes must deal with scale differences because auditory, visual, and haptic scenes can be examined at different spatial and temporal scales. For example, when looking at a tree, the details of the venation of its leaves need not to be considered in assessing the shape of the whole tree. Visual information also frequently has a self-similar character when the scale varies. For example the fundamental process of the extraction of illumination discontinuities in an image is similar when examining leaf venation or the tree branch patterns. Visual objects are also self-similar when viewed from different distances. In audition, a musical melody exists independently from the timber of the sounds of each note. Sounds also often have a self-similar character in their spectral characteristics (Voss and Clark, 1975). The situation is more complex in touch because, unlike the other senses, the physics at play differs fundamentally according to the scale at which haptic interaction is considered, even though certain self-similarity characteristics can also be observed (Wiertlewski et al., 2011). Tactile mechanics begin at the molecular scale since touch clearly depends on friction-related phenomena that depend on microscopic-scale physics, and it ends at the scales covered during ambulation.

At the macroscopic scale the multi-scale character of haptic perception can be demonstrated by the following illusion. If a flat plate is made to roll on the fingertip, that is, if the observer is provided with no other information than the orientation of the direction of the normal to a solid object while exploring it as depicted by Figure 4a, then the resulting percept is comparable to that of exploring a real slippery object where the observer is given displacement, orientation, and curvature information as shown in Figure 4b (Dostmohamed and Hayward, 2005). Provided that appropriate precautions are taken such as averting vision and ensuring that the observer is not aware of the mechanical details of the stimulation, then observers feel as if they were touching a curved object.

Figure 4: Bent plate illusion.

Figure 4c shows a cam mechanism capable of generating the sensation of exploring a virtual object with two fingers obtained by combining two stimuli as in Figure 4a. This effect can be achieved by assembling two of the mechanisms described in (Hayward, 2008a) in mirror opposition as in Figure 4d. During exploration, the two fingers remain at a constant distance from each other, as indicated in Figure 4c by the two thin lines, but the sensation is that of exploring a round object.

Figure 5: Human curvature discrimination performance model (with permission of the IEEE).

The relationship of this illusion with the notion of scale can be established assuming that one of the fundamental haptic perceptual tasks is to assess the local curvature of solid objects. It may be accepted without proof that in the simplified case of a profile of constant curvature the measurement of three points on this profile is the minimum information required to determine its curvature. Figure 5a illustrates this necessity. Measurements are necessarily corrupted by errors which translate to discrimination thresholds. It can be intuitively seen that, ceteris paribus, the greater is the portion of the profile that is considered, parametrized by the length of the cord, $d$, the more accurate is the measurement of curvature (Wijntjes et al., 2009). Assuming the existence of an osculating circle to a shape, the estimation of its curvature requires the measurement of the relative position of at least three points (circles). For a given scale, $d$, the displacement, $h$, the slope, $\phi$, or the curvature, $c$, are all potential sensory cues. Measurement errors can be represented either by the relative change in height, $\Delta h$, by the relative change in slope, $\Delta \phi$, at its opposite ends, or by the relative change in curvature, $\Delta c$, everywhere. Figure 5a is an abstraction of the curvature sensing problem. Zero-order error ($\Delta h$), first-order error ($\Delta \phi$), and second order error ($\Delta c$) can be related to each other with simple algebra. Figure 5b shows the results of a weak fusion cue combination model where the weights attributed to each sensory cue increase according to the reliability of the corresponding cue (Wijntjes et al., 2009). Given the known discrimination thresholds for these quantities, the model predicts that in the small scales (approx. $d < 1.0$ cm) curvature is the most reliable quantity to be sensed, in the intermediate scales (approx. $1.0 < d < 75$ cm), slope has this role, and in the large scales (approx. $d > 75$ cm) it is displacement, as corroborated by numerous psychophysical studies.

It was thus found that in the range of scales comprised between the size of a finger and the size of an arm, first-order information---that is, orientation---dominates over the other sources. These numbers suggest that the anatomical sizes of the human haptic appendages impose strict limits on the type of features that can be felt. These quantities correspond to orders of differences of displacement: zero, one, and second order; reflecting physiological constraints which in turn reflect the scale at which processing is performed. Of course one could speculate that higher derivatives could be leveraged to discriminate smaller scale features. The change of curvature over space would then be characteristic of a surface with asperities where curvature changes over very small length scales, viz. 1.0 mm and less.

### On contact mechanics

One source of tactile illusions is clearly derived from contact mechanics effects. As alluded to earlier, extracting the attributes of a touched object from partial knowledge of one's own tissue deformation, is a noisy and ambiguous process. It occurs under the influence of internal and external loads, and is at the root of all effects described thus far. Contact mechanics, or the analysis of the deformation of solids in contact, is thus of immediate relevance in the perception of small-scale attributes such as surface details. Nakatani et al. (2006) described an intriguing effect where strips with different small-scale mechanical properties are juxtaposed to form a flush surface. When explored actively, such surfaces cause the sensation that they have raised or recessed geometries.

Figure 6: Fishbone illusion and variants.

In its original form, Figure 6a, the stimulus is a rigid surface textured as shown. A raised pattern (0.1 mm thick) has a 3 mm wide central spine with orthogonal processes extending on each side with a 2 mm spatial period. When rubbing the finger on the spine, it is perceived as a recessed feature compared to the sides. Variants of this stimulus can be realized by juxtaposing strips of different materials having different roughnesses, different frictional properties (such as metal and rubber), or even different mobilities (Nakatani et al., 2008). Figure 6b shows a variant that can be easily realized by drilling holes in a plastic or metal plate.

A rough explanation for this illusion involves the observation that, during sliding, surfaces with different frictional or mobility properties create different boundary conditions that cause a complex tissue deformation field to propagate inside the finger. Since the tactile system is by necessity capable of reporting a highly simplified version of the actual deformation field of the finger tissues, then peripheral or central neural processes provide their best guess of what the boundary condition could be. The difficulty of the inverse problem involved has been recognized by roboticists who noticed the inherent ambiguous nature of the corresponding computational problem (Ricker and Ellis, 1993; de Rossi et al., 1991).

In vision, it was found that the brain had preferences for certain solutions to ambiguous perceptual problems. As one instance among many, it is well known that the visual system prefers to accept motion over deformation to explain the raw visual inputs (Wallach and O'Connell, 1953). So we could conclude that the tactile system prefers to assign the possible cause of an effect to variations of geometry over variations of surface frictional properties or other factors that could affect an unknown boundary condition. This conclusion is supported by a number of related effects that are only briefly mentioned here (Wang and Hayward, 2008; Kikuuwe et al., 2005; Hayward and Cruz-Hernandez, 2000; Smith et al., 2009; Robles-De-La-Torre and Hayward, 2001) but which all point to the same conclusion.

### Mechanical regularities

It may be surmised that the mechanical world is considerably more complicated than the optical or the acoustic world. This argument rests on the observation that the diversity of mechanical phenomena that can take place is truly great for the reason, as alluded to earlier, that different physics apply at different scales. Moreover, a variety of nonlinear and complex mechanical behaviors take place when objects come into contact, slide on each other, are compressed, are collided with, and so on. Only a small subset of objects we interact with are simple, smooth, solid objects. Most other solid objects are aggregations of small scale structures like fabrics, soil, wood, or have multi-stable mechanics like retractable ball pens or keyboards, and so-on, multiplying the possible mechanical behaviors at infinitum. Yet, universal, environmentally driven regularities must exist that the brain can initially extract and later rely upon. In vision, instances of such regularities include the celebrated convexity, light-from-above, or object rigidity assumptions (Ramachandran, 1988, Gregory, 1980, Ullman, 1979). Surely, similar notions must exist in touch and haptics.

Crushing things. Many surfaces on which one steps are made of complex, inhomogeneous, aggregated materials. These include carpets, gravel, soils, underbrush, snow, which have a broadband mechanical response due to the nonlinear mechanics at play. Despite their variety, these materials all share the property of a stronger response when they are crushed faster. If this regularity is artificially reproduced by vibrating a rigid tile with a random signal modulated in amplitude, one experiences the strong sensation that the tile gives under the foot, as shown by Visell et al. (2011). A related effect was demonstrated by Kildal et al. (2010) when pressing on a rigid surface with a vibrating pen.

Gravity. A omnipresent regularity that the brain should have internalized is the movement of objects under the influence of gravity (McIntyre et al., 2001). Balls rolling down a slope of inclination, $\alpha$, accelerate according to $0.7 \sin(\alpha)$, no matter what is their size and what is the substance they are made of. (This regularity was discovered by Galileo circa 1638 in one of the most far-reaching experiments in the history of science (Settle, 1961)). If one holds a stick made to vibrate with an amplitude $f(t)\propto g[7.0 \iint \sin(\alpha(t)) \mathrm{d} t]$, where $g$ is a periodic function and $\alpha$ is the stick inclination angle, then the person holding the stick spontaneously experiences the irrepressible sensation that a ball is rolling inside the stick (Yao and Hayward, 2006). The coefficient 7.0 is the corrected acceleration of gravity to account for the rolling movement of a ball. Different functions $g$ give different levels of realism but the effect is highly robust. The perceptual problem is to determine the ball displacement, $x(t)$, knowing $f(t)$, a type of inverse problem that the brain solves effortlessly despite the fact that $g$ is unknown but periodic.

Contact mechanics. Another example of a regularity which is linked to what our body experiences when pushing against a stiff surface. Almost all solid objects in contact obey to a Hertzian law which states that the area of the surface of contact between the bodies increases with the load. The rate of increase is a function of the relative geometry of the two bodies but also of their material properties (Hayward, 2008b). Thus, softer materials correspond to a lower rate of increase of the contact area. If an apparatus is constructed to modify the finger contact surface as a function of the pressing force independently of the finger displacement, then the modification of the rate of increase of the area of the contact surface can induce an illusory sensation of finger motion (Moscatelli et al., 2014). A related effect appealing to similar principles is the sensation of heaviness induced by the lateral deformation of the fingertips in the absence of net loading (Minamizawa et al., 2007).

Absence of slip. The notion of mechanical regularity can be exploited in the opposite manner. What would the brain make of stimuli which, precisely, do not contain the regularities that can normally be relied upon? Here is an example of an illusory effect that could be interpreted in this light. The so-called 'velvet hand illusion' (Mochiyama et al., 2005) occurs when one moves the two hands in contact with each other without slip but with an interposed network of wires or thin rods in-between. It is a conflicting stimulus since, normally, moving the hands together in mutual does not generate any significant tactile sensation, but here, the thin objects sliding between the two hands do cause a powerful tactile input. To the violation of the aforementioned regularity, the brain responds by 'feeling' a film interposed between the two hands (Kawabe et al., 2010).

The nonlinear nature of small scale mechanics. There are very few natural mechanical phenomena of relevance to touch that could be said to have a "linear character". Moreover, there is no indication that linearity is a useful concept in the mechano-transduction to tactile inputs (see for instance Lamoré et al. (1986). Thus it comes as no surprise that complex signals used to drive somatosensation may create surprising effects if they deviate in specific ways from the natural signals that the somatosensory system has evolved to process. The somatosensory system has been shown to have evolved to optimize the detection of fast rate stimuli differently from slow rate stimuli (Iggo and Ogawa, 1977, Edin and Valbo, 1990). It is otherwise known that if a signal detection system exhibits this property, then periodic excitatory signals having an odd symmetry will cause the output of the detector to undergo a DC drift (a ratcheting behavior). There are many examples in biology of such behaviors including, for instance, the pupillary reflex or heart rate regulation (Clynes, 1962). In touch, odd-symmetrical stimuli do cause a sensation of a persisting external load on the limb or the finger (Amemiya et al., 2005; Amemiya and Gomi, 2014).

### Conclusion

In this note, only a small subset of known tactile and haptic illusions was discussed. They were used to point out the similarities and the differences of the putative perceptual mechanisms in other sensory modalities. In sum, touch exhibits a number of similarities to other perceptual systems, but touch has idiosyncrasies which can be understood from the observation that certain perceptual problems that touch faces cannot be related to those faced by other modalities.

It would be natural to ask whether tactile illusions are the expression of imperfections of the somatosensory system or if illusions are a necessity. In the opening paragraphs, the impossibility for the brain to gain perfect knowledge of the mechanical state of the body that it inhabits, let alone of the external objects that perturb its state, was made clear. Evolution has found methods able to expedite the resolution of these problems at speeds and accuracies that are compatible with the survival of the organism, such as quickly grabbing and evaluating the mass of an object, whether it is a 20 kg suitcase or a flimsy paper cup. These solutions are sometimes surprising and we call them illusions. So the answer to the question of the imperfection of the somatosensory system is rather a question of whether it could be improved. The answer is emphatically yes, through perceptual learning and other skill-based mechanisms.

## References

• Adams, M et al. (2013). Finger pad friction and its role in grip and touch. Journal of the Royal Society Interface 10(80): 20120467.
• Amemiya, T and Gomi, H (2014). Distinct pseudo-attraction force sensation by a thumb-sized vibrator that oscillates asymmetrically. In: M Auvray and C Duriez (Eds.), Haptics: Neuroscience, Devices, Modeling, and Applications, Part-II (pp. 88-95).
• Amemiya, T; Ando, H and Maeda, T (2005). Virtual force display: Direction guidance using asymmetric acceleration via periodic translational motion. Proceedings of the World Haptics Conference (pp. 619-622).
• Benussi, V (1913). Psychologie der Zeitauffassung. Heidelberg: Carl Winter’s Universitätsbuchhandlung.
• Bochereau, S; Terekhov, A V and Hayward, V (2014). Amplitude and duration interdependence in the perceived intensity of complex tactile signals. In: M Auvray and C Duriez (Eds.), Haptics: Neuroscience, Devices, Modeling, and Applications, Part I (pp. 93-100).
• Bregman, A S (1990). Auditory scene analysis. Cambridge, MA: The MIT Press.
• Bresciani, J P; Drewing, K and Ernst, M O (2008). Human haptic perception and the design of haptic-enhanced virtual environments. In: The Sense of Touch and its Rendering (pp. 61-106). Berlin Heidelberg: Springer.
• Buckingham, G (2014). Getting a grip on heaviness perception: a review of weight illusions and their probable causes. Experimental Brain Research 232: 1623-1629.
• Caclin, A; Soto-Faraco, S; Kingstone, A and Spence, C (2002). Tactile "capture" of audition. Perception & Psychophysics, 64(4): 616-630.
• Carter, O; Konkle, T; Wang, Q; Hayward, V and Moore, C I (2008). Tactile rivalry demonstrated with an ambiguous apparent-motion quartet. Current Biology 18(14): 1050-1054.
• Charpentier, A (1891). Analyse expérimentale de quelques éléments de la sensation de poids. Archives de Physiologie Normale et Pathologique 3: 122-135.
• Clynes, M (1962). The non-linear biological dynamics of unidirectional rate sensitivity illustrated by analog computer analysis, pupillary reflex to light and sound, and heart rate behavior. Annals of the New York Academy of Sciences 98(4): 806-845.
• Cohen, J; Hansel, C E M and Sylvester, J D (1953). A new phenomenon in time judgment. Nature 172: 901.
• Cohen, J; Hansel, C E M and Sylvester, J D (1954). Interdependence of temporal and auditory judgments. Nature 174: 642-644.
• De Rossi, D; Caiti, A; Bianchi, R and Canepa, G (1991). Fine-form tactile discrimination through inversion of data from a skin-like sensor. In: Proceedings of the IEEE International Conference on Robotics and Automation (pp. 398-403).
• Dostmohamed, H and Hayward, V (2005). Trajectory of contact region on the fingerpad gives the illusion of haptic shape. Experimental Brain Research 164(3): 387-394.
• Edin, B. B. and Vallbo, A. B. (1990). Dynamic response of human muscle  spindle afferents to stretch. Journal of Neurophysiology 63(6): 1297-1306.
• Flanagan, J R; Bittner, J P and Johansson, R S (2008). Experience can change distinct size-weight priors engaged in lifting objects and judging their weights. Current Biology 18: 1742-1747.
• Gallace, A; Tan, H Z and Spence, C (2006). The failure to detect tactile change: A tactile analogue of visual change blindness. Journal Psychonomic Bulletin & Review 13(2): 300-303.
• Gelb, A (1914). Versuche auf dem Gebiete der Zeit- und Raumanschauung, Bericht uber der VI. Kongress fur Experimentelle Psychologie (pp. 36-42).
• Gjerdingen, R O (1994). Apparent motion in music? Music Perception 11(4): 335-370.
• Goldreich, D (2007). A bayesian perceptual model replicates the cutaneous rabbit and other spatiotemporal illusions. PLoS ONE 2(3): e333.
• Gregory, R L (1980). Perceptions as hypotheses. Philosophical Transactions of the Royal Society of London B: Biological Sciences 290(1038): 181-197.
• Gregory, R L (1997). Knowledge in perception and illusion. Philosophical Transactions of the Royal Society of London B: Biological Sciences 352(1358): 1121-1127.
• Grunwald, M and John, M (2008). German pioneers of research into human haptic perception. In: Human Haptic Perception: Basics and Applications (pp. 15-39). Basel: Birkhäuser.
• Guest, S; Catmur, C; Lloyd, D and Spence, C (2002). Audiotactile interactions in roughness perception. Experimental Brain Research 146: 161-171.
• Hartcher-O'Brien, J; Terekhov, A V; Auvray, M and Hayward, V (2014). Haptic shape constancy across distance. In: M Auvray and C Duriez (Eds.), Haptics: Neuroscience, Devices, Modeling, and Applications, Part I (pp. 77-84).
• Hayward, V (2008a). A brief taxonomy of tactile illusions and demonstrations that can be done in a hardware store. Brain Research Bulletin 75(6): 742-752.
• Hayward, V (2008b). Haptic shape cues, invariants, priors, and interface design. In: M Grunwald (Ed.), Human Haptic Perception---Basics and Applications (pp. 381-392). Birkhauser Verlag.
• Hayward, V (2011). Is there a 'plenhaptic' function? Philosophical Transactions of the Royal Society B: Biological Sciences 366(1581): 3115-3122.
• Hayward, V and Cruz-Hernandez, M (2000). Tactile display device using distributed lateral skin stretch. In: Proceedings of the Haptic Interfaces for Virtual Environment and Teleoperator Systems Symposium, DSC-69-2 (pp. 1309-1314).
• Heller, M A; Calcaterra, J A; Burson, L L and Green, S L (1997). The tactual horizontal-vertical illusion depends on radial motion of the entire arm. Perception & Psychophysics 59: 1297-1311.
• Helson, H (1930). The tau effect---An example of psychological relativity. Science 71(1847): 536-537.
• Howe, C Q and Purves, D (2005). Perceiving geometry: Geometrical illusions explained by natural scene statistics. New York: Springer.
• Iggo, A. and Ogawa, H. (1977). Correlative physiological and morphological  studies of rapidly adapting mechanoreceptors in cat’s glabrous skin.  Journal of Physiology 266(2): 275-296.
• Johnson, S A; Gorman, D M; Adams, M J and Briscoe, B J (1993). The friction and lubrication of human stratum corneum. Tribology Series 25: 663-672.
• Jousmäki, V and Hari, R (1998). Parchment-skin illusion: sound-biased touch. Current Biology 8(6): 190-191.
• Kawabe, Y; Chami, A; Ohka, M and Miyaoka, T (2010). A basic study on tactile displays using velvet hand illusion. In: IEEE Haptics Symposium (pp. 101-104).
• Kikuuwe, R; Sano, A; Mochiyama, H; Takasue, N and Fujimoto, H (2005). Enhancing haptic detection of surface undulation. ACM Transactions on Applied Perception 2(1): 46-67.
• Kildal, J (2010). 3d-press: haptic illusion of compliance when pressing on a rigid surface. In: Proceedings of the International Conference on Multimodal Interfaces and the Workshop on Machine Learning for Multimodal Interaction, ICMI-MLMI’10 (pp. 21:1-21:8).
• Konkle, T; Wang, Q; Hayward, V and Moore, C I (2009). Motion aftereffects transfer between touch and vision. Current Biology 19(9): 745-750.
• Lamoré, P J J; Muijser, H and Keemink, C J (1986). Envelope detection of amplitude-modulated high-frequency sinusoidal signals-by skin mechanoreceptors. Journal of the Acoustical Society of America 79(4): 1082-1085.
• Lederman, S J and Jones, L A (2011). Tactile and haptic illusions. IEEE Transactions on Haptics 4(4): 273-294.
• Mamassian, P and de Montalembert, M (2010). A simple model of the vertical-horizontal illusion. Vision Research 50(10): 956-962.
• McIntyre, J; Zago, M; Berthoz, A and Lacquantini, F (2001). Does the brain model Newton’s laws? Nature Neuroscience 4(7): 693-694.
• Minamizawa, K; Kajimoto, H; Kawakami, N and Tachi, S (2007). A wearable haptic display to present the gravity sensation---Preliminary observations and device design. In: World Haptics Conference (pp. 133-138).
• Mochiyama, H et al. (2005). Haptic illusions induced by moving line stimuli. In: World Haptics Conference (pp. 645-648).
• Moscatelli, A et al. (2014). A change in the fingertip contact area induces an illusory displacement of the finger. In: M Auvray and C Duriez (Eds.),Haptics: Neuroscience, Devices, Modeling, and Applications, Part II (pp. 72-79).
• Nakatani, M; Sato, A; Tachi, S and Hayward, V (2008). Tactile illusion caused by tangential skin srain and analysis in terms of skin deformation. In: Proceedings of Eurohaptics, LNCS 5024 (pp. 229-237). Springer-Verlag.
• Nakatani, M; Howe, R D and Tachi, S (2006). The fishbone tactile illusion. In: Proceedings of EuroHaptics (pp. 69-73).
• Ramachandran, V S (1988). Perception of shape-from-shading. Nature 331: 163-166.
• Ricker, S L and Ellis, R E (1993). 2-D finite-element models of tactile sensors. In: Proceedings of the IEEE International Conference on Robotics and Automation, Vol. 1 (pp. 941-947).
• Robles-De-La-Torre, G and Hayward, V (2001). Force can overcome object geometry in the perception of shape through active touch. Nature 412: 445-448.
• Ross, H E (1966). Sensory information necessary for the size-weight illusion. Nature 212: 650.
• Sarrazin, J C; Giraudo, M D and Pittenger, J B (2007). Tau and kappa effects in physical space: The case of audition. Psychological Research 71(2): 201-218.
• Settle, T B (1961). An experiment in the history of science: with a simple but ingenious device Galileo could obtain relatively precise time measurements. Science 133: 19-23.
• Smith, A M; Chapman, C E; Donati, F; Fortier-Poisson, P and Hayward, V (2009). Perception of simulated local shapes using active and passive touch. Journal of Neurophysiology 102: 3519-3529.
• Suzuki, K and Arashida, R (1992). Geometrical haptic illusions revisited: Haptic illusions compared with visual illusions. Perception & Psychophysics 52(3): 329-335.
• Ullman, S (1979). The interpretation of structure from motion. Proceedings of the Royal Society of London B: Biological Sciences 203(1153): 405-426.
• Visell, Y; Giordano, B L; Millet, G and Cooperstock, J R (2011). Vibration influences haptic perception of surface compliance during walking. PLoS ONE 6(3): e17697.
• von Békésy, G (1959). Neural funneling along the skin and between the inner and outer cells of the cochlea. Journal of the Acoustical Society of America 31(9): 1236-1249.
• Voss, R F and Clarke, J (1975). 1/f noise in music and speech. Nature 258: 317–318.
• Wallach, H and O'Connell, D N (1953). The kinetic depth effect. Journal of Experimental Psychology 45: 205-217.
• Wang, Q and Hayward, V (2008). Tactile synthesis and perceptual inverse problems seen from the view point of contact mechanics. ACM Transactions on Applied Perception 5(2): 1-19.
• Wertheimer, M (1912). Experimentelle Studien über das Sehen von Bewegung. Zeitschrift für Psychologie 61: 161-278. [Excerpted and translated in: T Shipley (Ed.), Classics in Psychology. New York: Philos. Lib. 1961.]
• Wiertlewski, M; Hudin, C and Hayward, V (2011). On the 1/f noise and non-integer harmonic decay of the interaction of a finger sliding on flat and sinusoidal surfaces. In: Proceedings of World Haptics Conference (pp. 25-30).
• Wijntjes, M W A; Sato, A; Hayward, V and Kappers, A M L (2009). Local surface orientation dominates haptic curvature discrimination. IEEE Transactions on Haptics 2(2): 94-102.
• Wolfe, U; Maloney, L T and Tam, M (2005). Distortions of perceived length in the frontoparallel plane: Tests of perspective theories. Perception & psychophysics 67(6): 967-979.
• Yao, H-Y and Hayward, V (2006). An experiment on length perception with a virtual rolling stone. In: Proceedings of Eurohaptics (pp. 325-330).

• Grunwald, M (Ed.) (2008). Human Haptic Perception: Basics and Applications. Springer Science & Business Media.