the ART (and SCIENCE) of AUDIENCE MANIPULATION
Magic isn’t just a bag of tricks – it’s a finely-tuned technology for shaping what we see. Now researchers are extracting its lessons.
by Drake Bennett / August 3, 2008
“In September of 1856, in the face of a growing rebellion, Napoleon III dispatched Jean-Eugene Robert-Houdin to Algeria. Robert-Houdin was not a general, nor a diplomat. He was a magician – the father, by most accounts, of modern magic. (A promising young escape artist named Ehrich Weiss would, a few decades later, choose his stage name by adding an “i” to “Houdin.”) His mission was to counter the Algerian marabouts, conjurers whose artful wizardry had helped convince the Algerian populace of Allah’s displeasure with French rule.
A French colonial official assembled an audience of Arab chieftains, and Robert-Houdin put on a show that, in its broadest outlines, would be familiar to today’s audiences: he pulled cannonballs out of his hat, he plucked lit candelabra out of the air, he poured gallon upon gallon of coffee out of an empty silver bowl.
Then, as he recounted in his memoirs, Robert-Houdin launched into a piece of enchantment calculated to cow the chieftains. He had a small wooden chest with a metal handle brought onto the stage. He picked a well-muscled member of the audience and asked him to lift the box; the man did it easily. Then Robert-Houdin announced, with a menacing wave of his hand, that he had sapped the man’s strength. When the volunteer again took hold of the box, it would not budge – an assistant to Robert-Houdin had activated a powerful magnet in the floor of the stage. The volunteer heaved at the box, his frustration shading into desperation until Robert-Houdin’s assistant, at a second signal, sent an electric shock through the handle, driving the man screaming from the stage. The chieftains were duly impressed, and the rebellion quelled.
The story of Robert-Houdin’s diplomacy by legerdemain is well-established in magic lore, in large part because it is the only documented instance, at least since antiquity, in which a conjurer changed the course of world affairs. Stage magic, after all, isn’t statecraft, but spectacle and entertainment.
In the past year, though, a few researchers have begun to realize that magic represents something more: a deep and untapped store of knowledge about the human mind. At a major conference last year in Las Vegas, in a scientific paper published last week and another due out this week, psychologists have argued that magicians, in their age-old quest for better ways to fool people, have been engaging in cutting-edge, if informal, research into how we see and comprehend the world around us. Just as studying the mechanisms of disease reveals the workings of our body’s defenses, these psychologists believe that studying the ways a talented magician can short-circuit our perceptual system will allow us to better grasp how the system is put together. “I think magicians and cognitive neuroscientists are getting at similar questions, but while neuroscientists have been looking at this for a few decades, magicians have been looking at this for centuries, millennia probably,” says Susana Martinez-Conde, a neuroscientist at the Barrow Neurological Institute and coauthor of one of the studies, published online last week in Nature Reviews Neuroscience. “What magicians do is light-years ahead in terms of sophistication and the power of these techniques.”
As magicians have long known and neuroscientists are increasingly discovering, human perception is a jury-rigged apparatus, full of gaps and easily manipulated. The collaboration between science and magic is still young, and the findings preliminary, but interest among scholars is only growing: the New York Academy of Science has invited the magician Apollo Robbins to give a presentation in January on the science of vision, and a team of magicians is scheduled to speak at next year’s annual meeting of the Society for Neuroscience, the world’s largest organization of brain researchers. And in a world where concentration is a scarce resource, a better understanding of how to channel it would have myriad uses, from safer dashboard displays to more alluring advertisements – and even, perhaps, to
A great deal of the success of a piece of magic is simply getting the audience’s attention and sending it to the wrong place – to a right hand flourishing a wand while the left secrets a ball away in a pocket or plucks a card from a sleeve. Magic shows are masterpieces of misdirection: they assault us with bright colors and shiny things, with puffs of smoke and with the constant obfuscatory patter that many magicians keep up as they perform.
For years, cognitive scientists thought of perception as like a movie camera, something that reproduced the world in its panoply of detail. Over the past decade, though, that model has been increasingly questioned. For one thing, people have a pronounced tendency to miss things that are happening right in front of them. Daniel Simons, a psychologist at the University of Illinois, did a series of now-famous studies in the late 1990s that showed the extent of this cognitive blindness. In one, people were approached by someone asking them for directions, only to have, in the middle of the conversation, that person replaced by another. Only half noticed the change.
In another study, people were shown a movie clip of two teams, one in black shirts and one in white, each passing a basketball around. The subjects were asked to count the number of passes one of the teams made. Half said afterward that they hadn’t noticed the woman in a gorilla suit who, midway through the clip, strolled through, paused, and beat her chest.
Because of work like this, a new model has arisen over the past decade, in which visual cognition is understood not as a camera but something more like a flashlight beam sweeping a twilit landscape. At any particular instant, we can only see detail and color in the small patch we are concentrating on. The rest we fill in through a combination of memory, prediction and a crude peripheral sight. We don’t take in our surroundings so much as actively and constantly construct them. “Our picture of the world is kind of a virtual reality,” says Ronald A. Rensink, a professor of computer science and psychology at the University of British Columbia and coauthor of a paper on magic and psychology that will be published online this week in Trends in Cognitive Sciences. “It’s a form of intelligent hallucination.”
The benefit of these sorts of cognitive shortcuts is that they allow us to create a remarkably rich image of our environment despite the fact that our two optic nerves have roughly the resolution of cell-phone cameras. We don’t have to, for example, waste time making out every car on the highway to understand that they are, indeed, cars, and to make sense of how they are moving – our minds can simply approximate from the thousands of cars we have already seen in our lives.
But because this method relies so heavily on expectation – not only to fill in the backdrop around us but to determine where to send what psychologists call our “attentional spotlight” – we are especially vulnerable to someone who knows our expectations and can manipulate them, someone like a magician. “In magic,” says Teller, half of the well-known duo Penn & Teller and one of five magicians credited as coauthors of the Nature Reviews Neuroscience paper, “we tend to take the things that make us smart as human beings and turn those against us.”
Misdirection is, in a sense, the conjurer’s tool that is easiest to understand – we miss things simply because we aren’t looking at them. Martinez-Conde is particularly interested in misdirection, and the question of what it is about certain movements that attract and hold our attention. Robbins, a performing pickpocket and another of the magicians to coauthor the Nature Neuroscience paper, has found, he says, that semi-circular gestures draw people’s attention better than straight ones. “It engages them more,” he says. “I use them when I’m actually coming out of the pocket.”
Martinez-Conde is intrigued by this distinction, and has hypothesized that the particular magnetism of curved motions might spring from the fact that they don’t map as easily onto the quick, straight movements, or saccades, that our eyes instinctively use to focus on objects. As a result, she suggests, curved motions might require more sustained attention and concentration to follow.
Other effects, though, are more befuddling. Often eye tracking studies show that subjects can be looking right at an object without seeing it – car accident survivors report a similar paradox. Or, with just a little encouragement, a person can be made to see something where there’s nothing.
The vanishing ball illusion is one of the most basic tricks a magician can learn: a ball is thrown repeatedly into the air and caught. Then, on the final throw, it disappears in midair. In fact, the magician has merely mimed the last throw, following the ball’s imagined upward trajectory with his eyes while keeping it hidden in his hand.
But if the technique is easily explained, the phenomenon itself is not. If done right, the trick actually makes observers see the ball rising into the air on the last toss and vanishing at its apex. As Rensink points out, this is something more powerful than merely getting someone to look in the wrong direction – it’s a demonstration of how easy it is to nudge the brain into the realm of actual hallucination. And cognitive scientists still don’t know exactly what’s causing it to happen.
For the moment, the cognitive scientists looking at magic are confining themselves to these sorts of simple effects, and the fundamental questions they raise. Eventually, though, Rensink envisions a sort of periodic table of attention effects: methods for getting someone’s attention, methods for deflecting it, methods for causing someone to be blind to something they’re looking directly at. Such a taxonomy, he argues, wouldn’t just be helpful to magicians. The control and management of attention is vital in all sorts of realms. Airplane cockpits and street signs would be designed better, security guards would be trained to be more alert, computer graphics would feel more natural, teaching less coercive.
Still, even if none of this came to pass, there’s a value in simply coming to grips with the gaps and limits in our awareness. Like Robert-Houdin’s audience, awed by a magnet, we are more easily manipulated and more likely to put ourselves in compromising situations if we don’t know what we don’t know. “The main thing is knowing that you’ve got limitations,” says the cognitive researcher Daniel Simons. “Most people don’t understand the extent to which talking on a cellphone affects their driving.”
According to Teller, magic, more than anything else, serves as that reminder. And that explains why, despite its comparatively humble effects, it continues even in the age of Imax to attract practitioners and audiences. “The fundamental thing we do every day is ascertain what is reality, it’s this diagnosis of what the signals coming into our eyes are supposed to mean,” he says. “We say, ‘That’s a fence, I must not walk into it,’ or, ‘Is that a car coming around the corner? How much can I see of it? Oh, no, it’s only a bicycle.’ ” What draws people to magic, he believes, is an appreciation of how slippery that seemingly simple diagnosis can be. “They realize,” he says, “that the best way to grasp the power of deception is to do it themselves.”
email : smart [at] neuralcorrelate [dot] com
Nature Reviews Neuroscience, advance online publication, 30 July
2008 / doi:10.1038/nrn2473
Science and society: Attention and awareness in stage magic: turning tricks into research
by Stephen L. Macknik , Mac King, James Randi , Apollo Robbins, Teller, John Thompson & Susana Martinez-Conde 
“Just as vision scientists study visual art and illusions to elucidate the workings of the visual system, so too can cognitive scientists study cognitive illusions to elucidate the underpinnings of cognition. Magic shows are a manifestation of accomplished magic performers’ deep intuition for and understanding of human attention and awareness. By studying magicians and their techniques, neuroscientists can learn powerful methods to manipulate attention and awareness in the laboratory. Such methods could be exploited to directly study the behavioural and neural basis of consciousness itself, for instance through the use of brain imaging and other neural recording techniques.
Magic is one of the oldest and most widespread forms of performance art  (Fig. 1). It is also a discipline with a long legacy of informal experimentation. This informal research by magicians aims to determine what conditions allow for the maximum manipulation of human attention and perception. Much as early filmmakers experimented with editing techniques to determine which technique would communicate their intent most effectively, magicians have explored the techniques that most effectively divert attention or exploit the shortcomings of human vision and awareness. As such, magic is a rich and largely untapped source of insight into perception and awareness. Insofar as the understanding of behaviour and perception goes, there are specific cases in which the magician’s intuitive knowledge is superior to that of the neuroscientist. In this Perspective, we underline potential areas in which neuroscientists stand to reap great benefits from collaboration with the magic community (Box 1 highlights one such potential area of collaboration).
Using completely natural means, magicians create effects (magic tricks) that seem to be outside the laws of nature. One should note that, unlike so-called psychics, magicians do not claim to possess supernatural powers. The devices used by magicians can include one or more of the following: visual illusions (after-images), optical illusions (‘smoke and mirrors’), cognitive illusions (inattentional blindness), special effects (explosions, fake gunshots, et cetera), and secret devices and mechanical artifacts (gimmicks).
Visual illusions — and other sensory illusions — are phenomena in which the subjective perception of a stimulus does not match the physical reality of the stimulus. Visual illusions occur because neural circuits in the brain amplify, suppress, converge and diverge visual information in a fashion that ultimately leaves the observer with a subjective perception that is different from the reality. For example, lateral inhibitory circuits in the early visual system enhance the contrast of edges and corners so that these visual features seem to be more salient than they truly are [2, 3, 4, 5, 6]. Unlike visual illusions, optical illusions do not result from brain processes: they manipulate the physical properties of light, such as reflection (using mirrors) and refraction (a pencil looks broken when it is placed upright in a glass of water owing to the different refraction indices of air and water). Cognitive illusions can be distinguished from visual illusions in that they are not sensory in nature: they involve higher-level cognitive functions, such as attention and causal inference (most coin and card tricks used by magicians fall into this category).
The application of all these devices by the expert magician gives the impression of a ‘magical’ event that is impossible in the physical realm (see Table 1 for a classification of the main types of magic effects and their underlying methods). This Perspective addresses how cognitive and visual illusions are applied in magic, and their underlying neural mechanisms. We also discuss some of the principles that have been developed by magicians and pickpockets throughout the centuries to manipulate awareness and attention, as well as their potential applications to research, especially in the study of the brain mechanisms that underlie attention and awareness. This Perspective therefore seeks to inform the cognitive neuroscientist that the techniques used by magicians can be powerful and robust tools to take to the laboratory. The study of the artistic intuitions that magicians have developed about attention and awareness might further lead to significant new scientific insights into their neural bases.
Visual illusions in magic
Visual illusions are often used by neuroscientists to dissociate the neural activity that matches the perception of a stimulus from the neuronal activity that matches the physical reality. Those neurons, circuits and brain areas with activity that matches the physical stimulus rather than the subjective perception can be excluded from the neural correlates of consciousness. Visual illusions are also used by magicians to fool their audiences, often to enhance cognitive illusions. Here we discuss a few categories of visual illusions that have contributed to magic tricks, as well as their neural bases.
Spoon bending. In this illusion the magician bends a spoon, apparently by using the power of the mind. In one part of the trick, the magician holds the spoon horizontally and shakes it up and down. This shows that the neck of the spoon has apparently become flexible . The apparent rubberiness of the spoon is an example of the Dancing Bar (or Rubber Tree) illusion , in which an oscillating bar (or rubber tree) seems to bend when it is bounced rapidly. The neural basis of this illusion lies in the fact that end-stopped neurons (that is, neurons that respond both to motion and to the terminations of a stimulus’ edges, such as corners or the ends of lines) in the primary visual cortex (area V1) and the middle temporal visual area (area MT, also known as area V5) respond differently from non-end-stopped neurons to oscillating stimuli [8, 9, 10, 11]. This differential response results in an apparent spatial mislocalization between the ends of a stimulus and its centre, making a solid object look like it flexes in the middle.
The Retention-of-Vision Vanish. Persistence of vision is an effect in which an image seems to persist for longer than its presentation time [12, 13, 14]. Thus, an object that has been removed from the visual field will still seem to be visible for a short period of time. The Great Tomsoni’s (J.T.) Coloured Dress trick, in which the magician’s assistant’s white dress instantaneously changes to a red dress, illustrates an application of this illusion to magic. At first the colour change seems to be due (trivially) to the onset of red illumination of the woman. But after the red light is turned off and a white light is turned on, the woman is revealed to be actually wearing a red dress. Here is how it works: when the red light shuts off there is a short period of darkness in which the audience is left with a brief positive after-image of the red-dressed (actually white-dressed but red-lit) woman. This short after-image persists for enough time to allow the white dress to be rapidly removed while the room is still dark. When the white lights come back, the red dress that the assistant was always wearing below the white dress is now visible.
This same illusion is the basis for perceptual stability during the viewing of motion pictures (the image seems to be stable when in fact it is flickering). On a neural level, both turning on and turning off a stimulus generate responses in visual neurons that result in the perceptual visibility of the stimulus . The neural response that is generated by turning off a stimulus is called the after-discharge, and it has the perceptual consequence of a positive after-image that persists for approximately 100 ms after the termination of the stimulus [16, 17, 18].
Jerry Andrus’s Trizonal Space Warp. In this illusion the audience stares for several seconds at a spinning disk with three zones of expanding and contracting motion. They are then asked to look at a different object on stage that consequently seems to both expand and contract. Motion after-effects, more commonly known as The Waterfall Illusion, are the oldest-recorded visual illusions. First reported in his Parva Naturalia, Aristotle noticed that if one fixates a moving stream of water and then looks away, the rocks at the side of the stream will seem to move in the opposite direction to the water. This effect is caused by neural adaptation — that is, by the decrease in responsiveness of a neural system to a constant stimulus. In the Trizonal Space Warp illusion, adaptation to expanding and contracting motion occurs in three different parts of the visual field.
The above illusions are examples of magic tricks that could have been used to help elucidate the underpinnings of visual perception. There might be other fundamental visual processes that could be discovered by studying magic (Box 1). Further, we propose that there are cognitive processes that will be better understood as we learn more from magicians, as discussed in the next section.
Cognitive illusions in magic
Inattentional blindness and change blindness. Attended objects can seem to be more salient or to have higher contrast than unattended objects [19, 20, 21, 22]. These perceptual effects have well-documented neural correlates in the visual system . Magicians use the general term ‘misdirection’ to refer to the diversion of the spectator’s attention away from a secret action. Thus, misdirection can be defined as drawing the audience’s attention away from the ‘method’ (the secret behind the ‘effect’) and towards the effect (what the spectator perceives) [7, 24]. Misdirection can be applied in an overt or a covert manner. Here we use the term ‘overt misdirection’ to indicate cases in which the magician redirects the spectator’s gaze away from the method. In the more subtle ‘covert misdirection’, the magician draws the spectator’s attentional spotlight (which can be thought of as the spectator’s focus of suspicion) away from the method without redirecting the spectator’s gaze. Thus, in covert misdirection the spectators can be looking directly at the method behind the trick and yet be unaware of it because their attention is focused elsewhere.
The concept of covert misdirection is exemplified by the cognitive-neuroscience paradigms of change blindness and inattentional blindness. With change blindness, people fail to notice that something is different from the way it was before. This change can be expected or unexpected, but the key is that it requires the observer to compare the post-change state with the pre-change state. Change-blindness studies have shown that dramatic changes in a visual scene will go unnoticed if they occur during a transient interruption , such as a blink , a saccadic eye movement  or a flicker of the scene [28, 29, 30, 31], even when people are looking right at the changes. However, observers can also miss large gradual changes in the absence of interruptions . A dramatic example of change blindness is illustrated in the Colour-Changing Card Trick video by Richard Wiseman and colleagues (available online at YouTube.com). In this demonstration, the viewers fail to notice colour changes that take place off-camera.
With inattentional blindness, people fail to notice an unexpected object that is fully visible in the display. Thus, inattentional blindness differs from change blindness in that no memory comparison is needed — the missed object is fully visible at a single point in time. In a classic example of inattentional blindness, Simons and Chabris  asked observers to count how many times the members of a basketball team passed a ball to one another, while ignoring the passes made by members of a different team. While they concentrated on the counting task, most observers failed to notice a person wearing a gorilla suit walk across the scene (the gorilla even stops briefly at the centre of the scene and beats its chest!). In this situation no acute interruption or distraction was necessary, as the assigned task of counting passes was absorbing. Further, the observers had to keep their eyes on the scene at all times in order to accurately perform the task. Memmert showed, using eye-tracking recordings, that many observers did not notice the gorilla even when they were looking directly at it .
The magic community considers the covert form of misdirection to be more elegant than the overt form . Few studies have addressed their relative efficacy, however. Kuhn and Tatler  measured the eye movements of observers during the presentation of a magic trick (a magician made a cigarette ‘disappear’ by dropping it below the table). To our knowledge, this is the first study to have correlated the perception of magic with any physiological measurement. The goal of the experiment was to analyse the scan paths of subjects to determine whether observers missed the trick because they did not look at it at the right time or because they did not attend to it (irrespective of the position of their gaze). The results showed that the detection (or not) of the cigarette drop could not be explained at the level of the retina. That is, detection rates were not significantly influenced by blinks, saccadic movements or how far the cigarette was from the centre of vision at the time of the drop. The authors concluded that the magician primarily manipulates the spectators’ attention rather than their gaze, using similar principles to those that are used in inattentional-blindness studies. Thus, to overcome the magician’s misdirection, spectators should reallocate their attention — rather than their gaze — to the concealed event (that is, the cigarette drop) at the critical time . Recent studies have found that the directions of microsaccades can also be used as an indicator of the spatial allocation of covert attention [37, 38, 39]. Future research could aim to measure the microsaccade direction biases of spectators during successful and unsuccessful magic tricks.
A recent study of the Vanishing-Ball Illusion further supports the conclusion that the manipulation of gaze position is not critical for effective covert misdirection. In the Vanishing-Ball Illusion, a ball thrown by the magician vanishes mid-flight. To achieve this effect, the magician begins by tossing the ball straight up in the air and catching it several times without event; then, on the final toss, the magician only pretends to throw the ball. The ball is in reality hidden in the magician’s hand, but most spectators perceive it ascending and then vanishing mid-flight. During the execution of this trick, the magician’s head and eyes follow the trajectory of an imaginary ball being thrown upwards. Kuhn and Land  found that the magician’s use of such social cues was critical for making the spectators’ perceive the illusion (that is, the ball vanishing mid-flight). However, observers did not direct their gaze to the area in which they claimed to have seen the ball vanish, suggesting that the oculomotor system is not fooled by the illusion. Instead, the illusory effect is presumably caused by covert redirection of the attentional spotlight to the predicted position of the ball. This result is consistent with previous studies that suggested that there are separate mechanisms for perception and visuomotor control [41, 42, 43, 44, 45, 46, 47, 48]. For instance, the eye movements of blindsight patients are biased towards stimuli that the patients do not consciously perceive [49, 50, 51]. Kuhn and Land  further proposed that in the Vanishing-Ball Illusion the covert redirection of the attentional spotlight to the predicted position of the ball might be related to “representational momentum” (Ref. 52). That is, that the final position of a moving object that suddenly disappears is perceived further along the path of motion than its actual final position. The neural correlates of representational momentum might be located in the posterior parietal cortex in the primate . Observers of the Vanishing-Ball Illusion might also be tricked by the strong implied motion that is suggested by the magician’s moves. Recent studies have focused on the neuronal mechanisms that underlie the perception of implied motion (some examples of implied motion are the speed lines that are used by cartoonists, and still photographs of people running or dancing). Neurons that respond to implied motion are found in extrastriate visual areas of the dorsal stream, and they are thought to be also sensitive to real motion [54, 55]. Thus, implied motion might activate similar circuits to those that are active during the perception of real motion, and this might result in perceptual illusions. Another example of this might be when one pretends to throw a stick for a dog during a game of fetch.
How do magicians misdirect the audience’s attentional spotlight? Magicians can effectively control an object’s salience by manipulating the audience’s bottom-up and/or top-down attentional control mechanisms. Objects that are new, unusual, of high contrast or moving are salient, and the audience’s attention is more strongly drawn towards them. such object properties induce bottom-up control of attention (and are used to accomplish ‘passive misdirection’ in magic theory [7, 56] or ‘exogenous attentional capture’ in psychology) because the attention is driven by increased activity in the ascending sensory system. One way in which a magician might control bottom-up attention is by suddenly producing a flying dove. The spectators’ gaze and attention will focus on the dove’s flight, and this will give the magician a few unattended moments in which he or she can conduct a secret manoeuvre.
Another facet of bottom-up attention that magicians exploit is the fact that if more than one movement is visible, spectators will tend to follow the larger (that is, the more salient) motion . Hence the magician’s axiom, ‘A big move covers a small move.’ A neural process that might underlie this axiom is the low-level mechanism of contrast-gain control (or contrast-gain adaptation) . In contrast-gain control, the perceived contrast of a stimulus is affected by the contrast of surrounding stimuli (whereas in contrast-gain adaptation, the perceived contrast of a stimulus is affected by that of a preceding stimulus) . A large or fast-moving stimulus might therefore decrease the perceived salience of a small or more slowly moving stimulus that is presented either simultaneously (in contrast-gain control) or subsequently (in contrast-gain adaptation). Novel stimuli are known to produce stronger neural responses in the inferotemporal cortex (area IT), the hippocampus, the prefrontal cortex and the lateral intraparietal area [59, 60, 61, 62, 63]; these effects are attributed to bottom-up attentional processes.
The salience of an object can also be increased by actively directing attention to it. For example, a magician might ask a subject to perform a task that involves one specific object, so that any changes that are occurring in a second object are missed. Such techniques are considered to induce top-down attentional control (and are used by magicians to accomplish ‘active misdirection’ (Refs 7, 56) or by psychologists to accomplish ‘endogenous attentional capture’) because they modulate (increase or decrease) neural activity in low-level brain areas through feedback pathways from high-level brain areas that are involved in cognitive functions . One example of top-down attentional modulation is provided by recent work by Chen and colleagues , which shows that neural responses in the primary visual cortex, an early visual-processing area, are enhanced as a function of task difficulty during attentional tasks. Another example of top-down attentional control is when a magician asks the audience to watch carefully an object that is being manipulated in one hand, while at the same time conducting a secret action with the other hand.
The principles that underlie attentional capture and contrast-gain control and adaptation also apply to other sensory systems, for example the somatosensory system. Pickpockets use techniques similar to those that are used by magicians (for instance, sleight-of-hand manoeuvres) to manipulate the awareness and attention of their marks. One way in which pickpockets manipulate the somatosensory system by applying the axiom ‘A big move covers a small move’ is as follows. To steal a watch directly from the wrist of a mark, the pickpocket might first squeeze the wrist while the watch is still on  (invoking contrast-gain adaptation). This has two effects. First, it makes a high-contrast somatosensory impression that adapts the touch receptors in the skin, making them less sensitive to the subsequent light touches that are required to unbuckle and remove the watch. Second, the high-contrast impression leaves behind a somatosensory after-image, giving rise to the illusion that the watch is still on after it has been removed.
Another way in which magicians can alter an object’s salience is to split the audience’s attention by introducing several concurrent actions . If two actions start almost simultaneously, the one that begins first will usually attract more attention [7, 67]. Social cues, such as the magician’s gaze (for instance, in the Vanishing-Ball Illusion), their voice and verbal communication and their body language (pointing, tension/relaxation), also play an important part in manipulating the spectator’s attentional spotlight .
Misdirection occurs not only in space (what the audience looks at) but also in time (when the audience looks). Thus, magicians strive to redirect the audience’s attention away from the moment of the method and towards the moment of ‘magic’. Indeed, in many magic tricks the secret action occurs when the spectators think that the trick has not yet begun, or when they think that the trick is over. Many magicians use comedy and laughter as a way to reduce focused attention at critical points in time. The magicians’ term ‘time misdirection’ refers to the deliberate separation of the moment of the method from the moment of the effect. Usually a delay is introduced between method (that is, cause) and effect, preventing the spectator from causally linking the two .
Memory illusions and illusory correlations. Magic works in adverse circumstances: an important part of the entertainment is that spectators are naturally suspicious and will try to discover the method behind the trick. Thus, observers of a magic trick will often try to reconstruct events to understand what happened. However, a successful magician will either have made it impossible to discover the method, or will seem to have ruled out all possible methods (including the actual method) until magic is the only apparent explanation [7, 68] (see Supplementary information S1 (movie)). The magician can also influence the spectators’ recall of the performance by using misdirection: events that draw the spectators’ attention will be better remembered than less salient events [7, 24, 69]. An apparently natural or spontaneous action, such as scratching one’s head, will not be memorable (although it might be critical to the execution of the trick). Unspoken assumptions and implied information are also important to both the perception of the magic trick and its subsequent reconstruction . J.R. has observed that spectators are more easily lulled into eagerly accepting suggestions and unspoken information than into accepting direct assertions  (see Supplementary information S2 (movie)). Thus, in the process of reconstruction, implication can be remembered as direct proof. The magician can further influence future recollection by describing past events in a manner that will bias the reconstruction process . This is known in cognitive science as the ‘misinformation effect’ — that is, the tendency for misleading information presented after the event to reduce one’s memory accuracy for the original event. This effect can even lead to the creation of a ‘false memory’ for events that never took place . The famous Indian Rope Trick legend might have partially resulted from the misinformation effect. In the Indian Rope Trick, a boy climbs a magically suspended rope and disappears at the top. The magician follows the boy up the rope into the invisible area at the top and cuts him into pieces (evidenced by the bloody body parts falling from the invisible area down to the ground). The magician then descends the rope and magically reintegrates the boy with no harm done. In fact, the Indian Rope Trick has never been performed, despite numerous witness accounts [71, 72, 73].
Although the study of false memory and misinformation effects has become a mainstream topic in cognitive science over the past few decades, it is possible that the field would have advanced faster if scientists had looked at the magicians’ intuition of human memory earlier. Even today, despite the substantial progress that scientists have already made in this area, the misinformation effect as used by magicians could be robustly reproduced in the laboratory to study the neural underpinnings of memory mechanisms and, in particular, false-memory mechanisms.
Magicians can also make their audiences incorrectly link cause and effect. We all infer cause and effect in everyday life. When A precedes B, we often conclude that A causes B. The skilled magician takes advantage of this inference by making sure that event A (for example, pouring water on a ball) always precedes event B (in this case, the ball disappearing). However, A does not actually cause B: the magician only makes it seem so [74, 75]. This type of illusion — seeing a correlation that is not there — is termed an ‘illusory correlation’. Illusory correlations can arise from unequal weighting of information, from the participants’ expectancies (such as prior beliefs or stereotypical knowledge) and/or from selective attention and encoding. In this third possibility, illusory correlations arise when some events capture more attention or are more likely to be encoded in memory and remembered than other, less salient, events . Thus, the magician can effectively use misdirection techniques to draw illusory correlations between two unrelated events. Just as visual scientists use visual illusions to identify the neural mechanisms of perception, neuroscientists could use illusory correlations to identify the neural mechanisms that underlie the cognitive computations of cause and effect. In a recent study by Parris and colleagues , participants underwent functional MRI (fMRI) while watching films of magic tricks that involved apparent cause–effect violations. The brain activation that was induced by the watching of these films was compared with the activation that occurred in a control condition in which participants watched video clips of events that did not involve apparent causal violations. The results showed greater activation in inferior medial frontal areas during the viewing of magic tricks than during the viewing of the control videos.
The illusion of trust. Pickpockets rely heavily on social misdirection. Gaze contact, body contact  and invasion of the mark’s personal space  are effective misdirection techniques (see Supplementary information S3 (movie)). Further, magicians and professional pickpockets use established techniques of persuasion to manipulate the trust of their audiences/marks. Some of these principles are also used by confidence artists in various scams and frauds. Brain-imaging studies of subjects playing online trust-building games show that activation in the paracingulate cortex is critical to building a trusting relationship. This activation seems to be related to inferring the partner’s intentions so as to predict their behaviour . Once trust was established, activity in the ventral tegmental area, which is linked to the evaluation of expected and realized reward, was correlated with the maintenance of ‘conditional trust’ (Refs 79, 80). ‘Unconditional trust’ was correlated with activity in the septal area, which is linked to social attachment [81, 82, 83]. Future research will determine the role of conditional versus unconditional trust in confidence fraud schemes. Neuroscientists can take advantage of the persuasion techniques that are used by magicians and pickpockets to identify the neural circuits that underlie feelings of trust and mistrust.
Various principles of stage magic aim to manipulate attention and awareness. These principles have been identified by magicians and have been refined over the centuries to great effect. The time is now ripe to take them into the laboratory and use them to guide new and more powerful experimental testing and careful quantification. This would elucidate the mechanistic pathways in the brain that allow magic tricks to work and would also generate novel and robust laboratory techniques for studying attention and awareness. A number of magic principles were discussed during the Magic of Consciousness symposium during the 11th Annual Meeting of the Association for the Scientific Study of Consciousness (Box 2); they are reviewed below.
An action is a motion that has a purpose. During the execution of a magic trick, it is necessary to use unnatural actions. Thus, the magician needs to reduce the audience’s suspicion about such actions. One way to do this is to justify unnatural actions so that they seem natural . Teller  refers to this principle with the aphorism, “An action is a motion with a purpose.”
In everyday life we categorize the motions made by others by interpreting their intentions. If we see somebody pushing their glasses higher on the bridge of their nose, we assume that the glasses needed adjustment, and no further interpretation is made. A good magician makes use of such innocent actions to hide ulterior motions in a process called ‘informing the motion’. For instance, magicians with a mute on-stage persona, like Teller, can take advantage of the glasses-pushing action to discreetly hide a small object in their mouth (being mute, they have no lines to garble). A less clever magician might do the same motion (moving the hand over the mouth) without informing it with a purpose (adjusting one’s glasses). Such a motion will be subject to suspicion and scrutiny. In that case, even if the spectators have not seen exactly how the trick works, they might feel that something is amiss. The skilled magician informs every motion with a convincing intention (see Supplementary information S4 (movie)).
Apparent repetition, priming and ‘closing all the doors’. In everyday life, by repeatedly observing a process we are able to deduce its workings. Priming is a type of repetition effect in which the presentation of a stimulus that is similar to a target makes subsequent presentations of the target perceptually more salient . Priming is used experimentally, and by the magician, to affect the subject’s sensitivity to a later presentation of a particular stimulus. Moreover, repetition can be used to induce sensory illusions, as in the Vanishing-Ball Illusion described earlier. Spectators are more likely to perceive the illusory ball vanishing in mid-flight if an actual ball has been tossed several times first, so that they are primed to know what an actual tossed ball looks like . Thus, priming and repetition can be helpful in inducing some illusory effects. Magicians also use repetition to hide the method behind the trick: when observers see an effect repeated, they naturally assume that each repetition is done by the same method. But the magician can covertly change the method that underlies each apparent repetition of the effect. Indeed, when a good magician repeats an effect, the method is varied in imperceptible ways and in an unpredictable rhythm. That way, each time observers suspect one method is being used, they find their suspicion disproved by the subsequent repetition  (see Supplementary information S4 (movie) and S5 (movie)). The magician might even deliberately raise suspicion about a possible method and then show that suspicion to be unfounded . In this way, the magician closes the door on every possible explanation for the trick [68, 73, 86], until the only remaining possibility is ‘magic’. This tactic is referred to as Tamariz’s Theory of False Solutions (see Supplementary information S1 (movie)). The use of apparent repetition has the added benefit of confusing the spectators’ reconstruction process. Further, the specific weaknesses of each method will cancel each other out .
Never do the same trick twice. The corollary of the closing all the doors principle is that if the magician performs the same trick twice for the same audience, there is an increased chance that the audience will identify the method that is being used and figure out the trick  (see Supplementary information S5 (movie)). In several studies by Kuhn and colleagues, most observers caught the method when the trick was shown a second time [35, 36, 40, 88]. Similarly, most inattentional-blindness demonstrations are a one-time-only kind of effect. Observers are much more likely to see the gorilla the second time they watch the basketball video described earlier .
Magic combines multiple principles of attention, awareness, trust and perception to both overtly and covertly misdirect the audience. Whether they are used for performance art or as a means to illicitly separate victims from their money and valuables, the accomplished performer uses robust and intuitive manipulative devices that are of great interest to neuroscientists pursuing the neural underpinnings of cognition, memory, sensation, social attachment, causal inference and awareness. Among these devices, we would like to emphasize the use of misdirection as a means to generate cognitive illusions such as inattentional blindness, change blindness, memory illusions and illusory correlations. Magicians are able to obtain these effects under conditions of high scrutiny show after show. Some of the crucial principles one needs to take into account when designing a robust trick are the understanding that every motion should seem to have a purpose, that the magician should not perform the same exact trick twice, and that the most successful tricks use apparent repetition to close all the doors on every possible explanation of the trick except for ‘magic’ itself.
Cognitive neuroscience endeavors to reverse-engineer the entire spectrum of cognition by determining the neural correlates of the various cognitive processes that make up our lives. Magic techniques can provide methods and insights that could help to explain what happens in the brain when a spectator thinks he knows what happened on stage . The possibilities of using magic as a source of cognitive illusions to help isolate the neural circuits that underlie specific cognitive functions are endless. For example, the magicians authoring this article emphasize the use of humour as a critical aid to the successful implementation of many tricks. Their intuition is that when the audience is laughing it is as if time stops and the attentional spotlight is put on hold. That is, the magician can do virtually anything when the audience is laughing, and nobody will notice.
Recording neural activity (by fMRI, lectroencephalogram,
magnetoencephalography, et cetera) in someone who is watching magic tricks that are accompanied by humour might help researchers determine the potential interaction between the allocation of attention and the sensation of mirth. Further possibilities range far beyond the uses of magic that have already been tried experimentally in cognitive science, such as the employment of magic palming techniques to direct subjects into confabulating their reasons for choices that they did not actually make [89, 90] (Bx 3). Magical cognitive illusions are furthermore an outstanding method by which to dissociate the perceived contents of awareness from the actual physical events. That is, one primary purpose of magic is to segregate those events that the magician does not want the observers to be aware of from those that the magician does want them to be aware of. We propose therefore that magical techniques that manipulate attention and awareness can be exploited to directly study the behavioural and neural basis of consciousness itself, for instance through the use of brain imaging and other neural recording techniques. If neuroscience researchers succeed in adopting magical methods with the same alacrity as professional magicians, they too should be able to control sensory awareness precisely and in real-time, while at the same time assessing the neural activation that is associated with it.”