Australasian Science: Australia's authority on science since 1938

What We Can Learn from Pickpockets

Credit: Voyagerix/Adobe

Credit: Voyagerix/Adobe

By Jack Brooks

Scientists are using the perceptual trickery of pickpockets and magicians as a new tool to study perceptual processing in the brain.

A man is walking through a bustling marketplace. He can feel the heat of the midday sun beating down on his back. He stops to peruse the spice stalls; the range of colours creates an explosion on the retina at the back of his eyes. The shouts of people bartering over grocery prices ring in his ears, masking the chatter and footsteps of others wandering the market. Hungry, hot and tired, he stops for a snack – the scents of the street food have gotten to him. He pauses here to make his choice, his mind momentarily unaware of the overpowering sensory stimuli of the marketplace, before turning to walk back to his hotel room. He reaches into his pocket but to his shock his wallet is missing. In this case he was the mark of a street pickpocket. This practice is commonly associated with this kind of scene, but is performed equally well on stage by magicians.

Let’s venture into a pickpocketing act, with each move briefly deconstructed using our current understanding of perception.

A magician calls for a volunteer from the audience. An old man is selected; he is wearing a coat and slacks. The magician holds him by the wrist and guides him up to the stage. Unbeknown to everyone the magician has already taken the man’s watch. By the time the watch was removed he had created the illusion that the watch was still on the wrist. It will be produced later.

Why was the man unaware that his watch had been taken? Touch receptors on the skin respond to sustained stimulation by reducing how much they fire. This enables them to be more sensitive to the present environment.

Similarly my research has been able to change how participants perceive the heaviness of a stack of weights ( To lift the weights, the brain creates a motor command to contract the muscles in the right pattern. For instance, as your arms fatigue while carrying a suitcase, you have to increase the motor command to the muscles. Additionally the suitcase feels heavier. The scientific consensus was that this sense of heaviness arises from a corollary of the motor command, not from sensory input from receptors in the muscle.

In our experiments we used fatiguing exercise and muscle vibration to reduce the sensory input returning to the brain from these receptors. If these receptors contribute to heaviness then less sensory input from them should equate to the weight feeling lighter. Indeed, we found that weights lifted by participants after conditioning were perceived to be slightly lighter. Our experiments suggested that both the motor command and muscle receptors contribute to the sense of heaviness.

The bulk of sensory processing and integration of these signals occurs in the brain. For example, the majority of amputees experience a painful phantom limb even though the severed nerves are not the primary cause of the pain. A few decades ago neuroscientists attributed the pain to central causes.

Prior to this, in 1965 Roger Melzack and Patrick Wall proposed the gate control theory of pain in Science. This postulates that touch “closes the gate” so that painful stimuli no longer reach the brain. You may have experienced this yourself, bumping your knee and rubbing it to make it feel better.

Our stage magician makes use of a similar principle that the sensation from a “greater force supersedes that of a lesser force”. As the audience volunteer is not standing close enough to the stage, the magician pushes him with both hands on the chest to give the audience a better view. At around the same time the magician removes the man’s pen from his outer breast pocket.

A 2016 study in Cognition ( by Dr Lee Walsh of The University of NSW concurs that when two touches to the skin are applied at the same time, the greater force contributes more to the percept than the smaller force. However, the smaller force is still perceived.

There are multiple mechanisms through which such an effect could occur, but the data suggest that the most salient or intense stimuli makes the largest contribution to the percept. This makes sense, especially when one considers that processing of sensory input from the skin is more limited than our other senses. Although magicians do use misdirection in space, it is misdirection in time that is much more effective.

Dissociating events in time is at the very heart of magic. It allows magicians to create the illusion that the effect has been caused by something else, as the audience is not focused on the move when it occurs.

At this juncture the magician produces the man’s watch. The audience members wonder how he took the watch as he was just touching the man’s chest.

To differentiate events arising from self and those from the outside world it’s necessary to predict the sensation arising from self-generated movements. This ability is degraded in schizophrenia and other mental health conditions. In one experiment published in JAMA Psychiatry (, participants pushed with their right finger on an apparatus that transferred an equal force to their left finger at either the same time or after a delay. The experimenters used functional magnetic resonance imaging to scan the somato­sensory cortex, which is responsible for processing touch stimuli. In healthy controls there was less activity here when the force transfer was simultaneous, while in schizophrenics there was equal activity in both simultaneous and delayed conditions.

This group also completed a hallucination score to quantify the severity of their symptoms. This score correlated with their lack of sensory attenuation. These observations build on the literature to suggest that schizophrenics are less able to predict the sensation generated by their own movements.

Perceived causality can also be changed by later events after the effect, which can alter our recall of what the effect was and how it was done. This is a pertinent issue in eyewitness testimony.

The magician continues his act with a card trick. He stands to the right of the volunteer and hands him a deck of cards to shuffle. As the cards are shuffled, the magician reaches over with his right hand and begins procuring cards from the man’s breast pocket.

Later in the act the magician reaches into his own pocket and reveals a number of items he has taken from the man’s pocket. The man was fooled in this trick by change blindness, but how was the audience fooled?

The magician used each procured card to hide the pickpocketed items from the audience’s view. Belgian experimental psychologist Albert Michotte lumped these effects under the umbrella of amodal perception, where we perceive an object but do not “see” it.

An example of this was a study overseen by Dr Anne Aimola Davies of The Australian National University in which participants viewed a magician procuring and vanishing an invisible object ( Almost one in three participants reported seeing an object even though it didn’t exist. The authors interpreted this to be consistent with top-down processing, where we make inferences about what we perceive based on expectation. In the context within which the trick was shown, the participants’ expectation that the object existed outweighed sensory evidence from the stimulus that it did not exist.

Top-down processing may also drive phantom limb perception. Can it be exploited to attenuate the pain sensation? In some amputees the brain has a wealth of experience with a limb but only a small amount of sensory experience of it not existing. This makes treating phantom limb pain using illusions a difficult task.

In the rubber hand illusion, one arm is placed out of view, with a dummy arm placed above it. Viewing someone touching the dummy arm at the same time as feeling touch on the arm that is out of the view creates a robust illusion that the dummy arm is part of the participant’s arm.

A 2016 study published in Pain ( found that the rubber hand illusion can cause changes in the temperature and histamine reactivity of the hidden arm, suggesting that the illusion has potential to modulate pain. However, the study’s meta-analysis found that illusions of embodiment do not modulate pain.

The researchers also reviewed the therapeutic potential of mirror box therapy, where a mirror is placed above the affected limb (or where the amputated limb would be) in such a way that when the patient looks at it they see their other limb. If the unaffected limb is touched or moved about they feel touch or movement of the phantom limb. Mirror box therapy was invented by Vilayanur Ramachandran, who also found that when phantom limb patients were touched on their face they sometimes felt touch on the phantom limb. This has been interpreted as a reorganisation of the brain’s map of the body.

Mirror box therapy may work through a similar mechanism, but the Pain review concluded it only shows potential when multiple sessions are undertaken – the analysis found that one session of mirror box therapy has no effect.

For the first time scientists have begun studying sense of limb ownership in animals ( Further animal studies may help us learn how to treat neurological disorders and develop better prosthetic limbs.

But returning to our friend on the street: how was he pickpocketed? The magician surreptitiously grasped a hold of his wallet and waited. As the man turned and walked away he did all the work himself. He fell victim to tactile suppression: the reduced sensitivity of our skin during movement.

Modern pickpockets have built on these methods and created new ones, and have largely stuck to the rule in the magic community of nondisclosure of their methods. However, in the past decade world-class magicians have been divulging their methods to psychologists, working with them to design tricks to further our understanding of the brain. This generation of magicians is deeply aware that perception underlies their tricks. Renowned pickpocket Apollo Robbins says: “It doesn’t matter if people are aware of how I work, or even what I’m going to do. They still won’t catch it..

Jack Brooks is a PhD student investigating touch perception and body representations in the School of Medical Sciences at The University of NSW. The methods described are provided for the purposes of the article only.