Perception, 2010, volume 39, pages 1286 ^ 1289
doi:10.1068/p6766
SHORT AND SWEET
The exploitation of Gestalt principles by magicians
Anthony S Barnhart
Department of Psychology, Box 871104, Arizona State University, Tempe, Arizona 85287-1104, USA;
e-mail: [email protected]
Received 2nd July 2010, in revised form 11 August 2010
Abstract. Magicians exploit a host of psychological principles in deceiving their audiences.
Psychologists have recently attempted to pinpoint the most common psychological tendencies
exploited by magicians. This paper highlights two co-occurring principles that appear to be the
basis for many popular magic tricks: accidental alignment and good continuation.
In recent years magic tricks have received attention as a means of studying psychology
and neuroscience. To this end, authors (Kuhn et al 2008; Macknik et al 2008) have
cited lists of psychological tendencies most often exploited by magicians to successfully
deceive audiences (eg inattentional blindness and priming), but two phenomena were
notably absent in both analyses, specifically the concept of accidental alignment and the
Gestalt grouping principle of good continuation. These two principles, often employed
in tandem, are important psychological tendencies that magicians regularly exploit.
In order to understand how magicians utilize these phenomena, imagine the following
scenario. A magician presents one long length of rope. After folding the rope in half
to identify its center, the rope is cut at what appears to be its center point. The
resulting two ropes are then displayed as in figure 1a. Later, the ropes are `magically'
restored to their original, uncut state. A video of the cut-and-restored rope trick can be
seen in the supplementary materials (http:/dx.doi.org/10.1068/p6766). The cut-and-restored
rope is one of the oldest themes in magic, and nearly every instantiation of the illusion
relies on the audience's inclination to form faulty assumptions that guide their perceptions. In this case, the faulty assumption that the rope was cut at its center point leads
audience members to perceive the arrangement in figure 1a as being two equal-length
ropes continuing unimpeded, behind the visual obstruction of the magician's hand (see
figure 1b for the assumed grouping). In actuality, before the ropes were cut, a subtle feat
of sleight-of-hand was performed that caused the ropes to be cut at a different location,
just inches from the end of the rope. The true grouping of the ropes can be seen in
figure 1c.
(a)
(b)
(c)
Figure 1. (a) The audience's view of
the cut rope. (b) The assumed grouping of the ropes (hand removed).
(c) The true grouping of the ropes
[see also the movie ‡ swf movie
file (20 Mb)].
Gestalt principles in magic
1287
Given an ambiguous grouping, as that imposed by the obstructing hand in figure 1a,
viewers make spontaneous assumptions about the state of the grouping behind the
obstruction. These assumptions are likely to be constructed from a few sources. First,
when presented with an ambiguous input, the visual system is thought to select either
the most likely (Helmholtz 1867/1962) or the simplest (Pomerantz and Kubovy 1986)
available interpretation. Not only is an arrangement with two continuous, equal-length
ropes more likely (the arrangement in figure 1c would be improbable in any situation
other than a magic show), it is also a simpler arrangement than the two unequal lengths
in figure 1c.
The second factor that influences the audience's interpretation is their perception
of the obstruction itself. The illusion of continuity is strengthened if spectators perceive
the positioning of the magician's hand to be irrelevant, rather than purposeful. Specifically, the audience should assume that the magician's hand is not specially placed to
occlude some anomaly in the ropes. Luckily, the likelihood principle again suggests that
an accidental alignment is improbable. We regularly view occluded objects and have
come to learn that accidental alignment of occluders with environmental anomalies
is exceedingly rare. Rather , we typically assume that our perceptions are viewpointgeneral. When viewing some feature, we assume that a person standing in a different
location is unlikely to have perceptions substantially different from our own (Cowie 1983).
Yet for spectators at a magic show, these accidental alignments that are rare in the world
are integral to the show. For example, the same two principles apply when the rope is later
`restored' to its original state. What the audience perceives as a long, continuous rope is
actually the short rope from the top of figure 1c positioned at the end of the longer rope,
with the junction occluded by the magician's hand (see figure 2).
Figure 2. The short rope is held
at the end of the long rope to
create the illusion of restoration.
These two tendencies in conjunction (the assumption of non-accidental alignment
and good continuation) reinforce the perceptual `filling-in' of objects behind an
occluder. The automaticity of this filling-in process is the likely reason behind its
common exploitation by magicians. Beyond the rope trick, other examples include the
famous Chinese Linking Rings illusion (wherein a set of seemingly solid metal rings
are linked together) and almost any illusion where an object like a spoon is bent,
supposedly via psychokinesis. In the case of the rings, subtle methods of presentation
facilitate the audience's faulty perception that the rings continue unbroken through the
magician's hand. In most spoon-bending illusions, the fulcrum of the bend is covered by
the fingertips, so the illusion only works if the audience assumes that the pieces of the
spoon on either side of the hand were attached and continuous to begin with.
These psychological principles are also employed in large-scale stage illusions.
The classic `Sawing a Woman in Half' illusion is rendered ineffective if the audience
perceives that the head protruding from one end of the box belongs to a different
1288
A S Barnhart
individual than the feet protruding from the opposite end. The trick only works if the
audience `fills-in' the occluded body linking the head and feet. On a large scale,
the illusion of continuity can be harder to elicit. As Shore and Enns (1997) demonstrated, completion of this kind is hindered as occluders increase in size. Magicians
have developed novel ways to combat this problem. The `Zig Zag Girl' illusion, developed
by Robert Harbin in the 1960s, employs a visual stimulus that makes the perception
of good continuation irresistible. In this illusion, an assistant's midsection appears to
be removed from her body. To facilitate the perception of continuity between the
assistant's exposed head, hand, and foot, an abstract silhouette of the assistant is
painted on the exterior of the box (see figure 3a). This imposes a specific interpretation
of the orientation of the assistant's body within the box. With the silhouette removed
(figure 3b), it is much easier to generate an alternative interpretation that can account
for placement of the assistant's extremities. In this case, the assistant could position
her body sideways within the contraption while still placing her face, hand, and foot
through the openings. The availability of this alternative interpretation renders the
illusion much less compelling.
(a)
(b)
Figure 3. (a) The `Zig Zag Girl' illusion performed by Ken Levy (http://www.amazingken.com).
The abstract silhouette facilitates the perception of good continuation. (b) The illusion becomes
much less compelling with the silhouette removed.
The prominence of these two principles in magic suggests that many illusions are
effective because they require non-Bayesian thinking in order to reach an accurate
solution. Chater (1996) noted that the simplicity and likelihood principles (which
explain good continuation) could be demonstrated within a Bayesian framework,
and that the assumption of non-accidental alignment is based on perceptual priors.
Similar non-Bayesian strategies would be necessary to avoid falling prey to Kuhn and
Land's (2006) vanishing-ball illusion, which is driven by susceptibility to joint attention.
The tendencies outlined here should be added to the growing list of heuristics that
spectators use when attempting to make sense of magic illusions, with future research
examining the automaticity of Bayesian assumptions in magic. Such a taxonomy
could create a firm foundation for establishing a `science of magic' (Kuhn et al 2008;
Macknik and Martinez-Conde 2009).
Gestalt principles in magic
1289
Acknowledgments. This paper is dedicated to the memory of Sheila Levy. Special thanks to Ken
Levy for permission to use his image. Thanks also to Gustav Kuhn, Steve Goldinger, and Mike
McBeath for advice on a previous version of the manuscript.
References
Chater N, 1996 ``Reconciling simplicity and likelihood principles in perceptual organization''
Psychological Review 103 566 ^ 581
Cowie R I, 1983 ``The viewer's place in theories of vision'' Proceedings of the Eighth International
Joint Conference on Artificial Intelligence volume 2 (San Francisco, CA: Morgan Kaufmann)
pp 952 ^ 958
Helmholtz H von, 1867/1962 Treatise on Physiological Optics volume 3 (New York: Dover, 1962);
English translation by J P C Southall for the Optical Society of America (1925) from the 3rd
German edition of Handbuch der physiologischen Optik (Hamburg: Voss 1910; first published
in 1867, Leipzig: Voss)
Kuhn G, Amlani A A, Rensink R A, 2008 ``Towards a science of magic'' Trends in Cognitive
Sciences 12 349 ^ 354
Kuhn G, Land M F, 2006 ``There's more to magic than meets the eye'' Current Biology 16
R950 ^ R951
Macknik S L, Martinez-Conde S, 2009 ``Real magic: Future studies of magic should be grounded
in neuroscience'' Nature Reviews Neuroscience 10 241
Macknik S L, King M, Randi J, Robbins A, Teller, Thompson J, Martinez-Conde S, 2008 ``Attention
and awareness in stage magic: Turning tricks into research'' Nature Reviews Neuroscience 9
871 ^ 879
Pomerantz J R, Kubovy M, 1986 ``Theoretical approaches to perceptual organization: Simplicity
and likelihood principles'', in Handbook of Perception and Human Performance volume II
Cognitive Processes and Performance Eds K R Boff, L Kaufman, J P Thomas (New York: John
Wiley) pp 36 ^ 45
Shore D I, Enns J T, 1997 ``Shape completion time depends on the size of the occluded region''
Journal of Experimental Psychology: Human Perception and Performance 23 980 ^ 998
ß 2010 a Pion publication
N:/psfiles/banners/
final-per.3d
ISSN 0301-0066 (print)
ISSN 1468-4233 (electronic)
www.perceptionweb.com
Conditions of use. This article may be downloaded from the Perception website for personal research
by members of subscribing organisations. Authors are entitled to distribute their own article (in printed
form or by e-mail) to up to 50 people. This PDF may not be placed on any website (or other online
distribution system) without permission of the publisher.