Aesthetics for Birds

Aesthetics and Philosophy of Art for Everyone

AFB X JAAC Discussions #1: Leddington on Magic



Welcome to the inaugural AFB x JAAC Discussion!

We’ll be discussing Jason Leddington‘s paper “The Experience of Magic”. First, we’ll have a critical precís from James Hamilton, followed by a response. (Leddington’s paper is available here.)

The paper and comments touch on many different issues, and will be of interest to those thinking about perception, deception, performance, cognitive dissonance, beliefs and aliefs, movies, depiction, as well as a host of other related phenomena.

Thanks to Jason and James for participating in this, and for being available to talk about readers’ questions and thoughts. Come join the discussion!


Critical Precís (James Hamilton):

In the introduction to this paper, Jason gives four reasons why it is “unfortunate” that art critics, art historians, and philosophers have tended to ignore theatrical magic. One of these is that theatrical magic “offers a unique and distinctively intellectual aesthetic experience” (253). And this, as the rest of the essay bears out, is the most significant reason; for the burden of the paper is to deliver and defend a precise account of what that “distinctively intellectual aesthetic experience” is.

Jason forecasts the four parts to his essay at the end of the introduction: “Section 1 dispels two widespread misconceptions about the nature of magic and discusses the special sort of depiction it requires. Section 2 asks, ‘What cognitive attitude is involved in the experience of magic?’ and criticizes three candidate replies; Section 3 then argues that Tamar Szabó Gendler’s notion of ‘belief-discordant alief’ holds the key to a correct answer (Gendler 2008, 641). On this basis, Section 4 develops an account of the cognitive dimension of the experience of magic and explores some of its consequences” (254, emphases added).


My goal here will be to (a) rehearse some of the key points in the paper, and (b) make a few critical comments on Jason’s handling of those points. I do not disagree with the main account he gives of the “distinctively intellectual aesthetic experience” often available to spectators of theatrical magic, namely, (i) that magic is “a form of theater that apparently presents impossible events and at the same time represents them as impossible” (256), and (ii) that the experience of magic involves some kind of cognitive dissonance. I do not, however, think that he has made the case that this is a cognitive dissonance between two belief-like states each of which is rationally justified. Nevertheless, this is a novel, and I think, quite successful account of the nature and the experience of some theatrical magic and, moreover, of the kind of theatrical magic that should intrigue philosophers, art historians, and art critics. A central advantage of this view of theatrical magic and its experience is that it can readily explain failures in both the performance and the reception of particular presentations of theatrical magic. This is a point that Jason himself makes in the final section of the paper; and I think he is quite right. The trick, so to speak, lies in explaining the success conditions.


The highlights, by section

I – Jason uses empirical sources to underpin his claims that the primary goal of theatrical magic is not to fool the audience (so magic is not mere deception), and that theatrical magic is not charlatanry (an attempt to “convince the audience of the existence of supernatural powers”). Then he asks, “What is it?” His answer is that magic consists of depicting events the spectator knows are not possible as though they were really happening. Note that, by holding the magic show appears to present that which is known to be impossible, the view both leaves room for the commonplace belief that magic is deceptive but also shows why it is not merely so.

II – Next Jason considers three hypotheses concerning the experience of magic.

First up is “(H1) The experience of magic essentially involves willing suspension of disbelief.” (H1) is false, Jason remarks, because “the whole force of a magic performance consists in the fact that the audience knows that what they are apparently witnessing is, in fact, impossible. But if the impossible event is relegated to the realm of fantasy via suspension of disbelief, then it is no longer apparently witnessed at all” (256).

Next up is “(H2) The experience of magic essentially involves unwilling suspension of disbelief.” This view captures something right about the experience of magic, namely that our cognitive attitude towards magic is not something we voluntarily adopt. Rather it is a result of our unwilled perceptions of the world. Still, whether “it is willing or unwilling, suspension of disbelief relegates the impossible event to the realm of fantasy and, so, prevents us from apparently witnessing it at all” (257).

Finally, Jason considers this hypothesis: “(H3) The experience of magic essentially involves conflict of belief.” He thinks this, while much more promising, is still inadequate because, while “there is cognitive dissonance in [the experience, it is] not the sort that demands resolution on pain of contradiction.” The reason is that what is presented as really happening is known to be impossible. And he concludes that “the right account of the experience of magic must include an account of cognitive dissonance that is not a matter of conflicting beliefs” (257).

This may not be contentious. But, when I described this view to one of my colleagues, his immediate thought was the situation was a bit like the “preface paradox” which, according to its first describer (D. C. Makison, 1965. “The Paradox of the Preface,” Analysis, 25/6: 205-207), can best be resolved by distinguishing between individual beliefs within a set – each of which can be rationally held – and the entire set of such beliefs – wherein, if the set contains a contradiction, the set cannot be rationally held. However that may be, the kind of analysis Jason prefers for the sort of cognitive dissonance that is involved in theatrical magic is provided in the section III.

III – Here Jason turns to an analysis of one kind of cognitive dissonance made famous by T. S. Gendler, namely one involving “belief-discordant alief” (Tamar Szabó Gendler, 2008. “Alief and Belief.” Journal of Philosophy 105: 634–663). According to Gendler, in the classic case of the Grand Canyon Skywalk – a transparent cantilevered walkway extending out some seventy feet from the canyon’s rim and hanging over a thousand-foot drop – there is no conflict between the rational belief one is safe and a rational belief one might fall; instead there is a cognitive conflict between the rational belief one is safe and a belief-like state, “alief,” that one will fall. Jason notes that, while belief involves endorsement of the contents of the state, “in alief, a representational content is present in the subject’s cognitive system, but it is not endorsed” (257). And so, “we have a type of cognitive conflict that is passively incurred, has affective and behavioral consequences, and is not a matter of conflicting belief” (258). And then he uses this idea to formulate the hypothesis about the experience of magic that he thinks gets it right: “(H4) The experience of magic essentially involves a belief–discordant alief that an impossible event is happening” (258).

IV – In the final act, so to speak, Jason considers what it would be like to fail to present the magic in the right way and to fail to understand it. He begins by noting that “In general, suspecting that you know how a magic performance is accomplished is enough to ruin it. And since, when witnessing the apparent presentation of an impossibility, you typically will have some ideas about possible methods, the magician has to do more than conceal the actual method—namely, “cancel” all the methods that might reasonably occur to you” (258). Next, by exploring a particular case Jason is able to demonstrate that “the performance immediately produces cognitive dissonance in the spectator [to which the most “natural, immediate response … is to try to restore harmony. In this case, there are three options. First, the spectator can try to dislodge the alief … But since the illusion is robust, the alief refuses to budge. Second, the spectator can try to revise the belief that [what is depicted] is impossible. But this is rationally unacceptable—at least in part because it would generate more psychic dissonance than it would resolve. Third, if neither alief nor belief will budge, the spectator can at least try to mitigate their discord by devising a plausible explanation for the appearance of impossibility.” And this is what is aimed at: “the spectator struggles to minimize cognitive dissonance by explaining away appearances. But the point of the strategy of canceling methods is precisely to thwart this attempt and, so, to maximize the cognitive dissonance that spectators experience by depriving them of any means to mitigate it.” (259).


Comments on some aspects of the argument

The first concerns Jason’s contrasts among movies, theater, and theatrical magic. I am a bit worried that he is too quick on the history of magic and movies. The fact is that it may be no accident that magic emporiums were among the first locations of movie showings and that so-called “trick movies” were abundant in this same period. For many of the magicians of the early 1900s, magic and movies were seen to do similar things. (See Matthew Solomon, 2006. “Up-to-Date Magic: Theatrical Conjuring and the Trick Film, Theatre Journal 58/4: 595-615.)

A deeper worry is that he is too quick to characterize the differences between film and theater on the one hand and magic on the other in terms of depicting something as present and depicting something as really present. The idea is that “It is one thing to depict events as happening in a merely possible present—and so, in the world as it might have been—but a magic performance depicts events as happening in the actual present—and so, in the world as it is” (255). No doubt the distinction itself is reasonable, but it is dubious to align theater and movies with only the former and to align magic with only the latter. One wants to see the empirical evidence for that; and none is offered.

What is especially worrisome in Jason’s account is the explanation of what is wrong with the WSD [ed.: Willing Suspension of Disbelief] hypothesis about the experience of magic. Now, I quite agree that theatrical magic does not involve the willing suspension of disbelief. This is not the only reason the WSD hypothesis fails with respect to theatrical magic nor, if I am right about that, does it do as much to distinguish theatrical magic from theater as Jason thinks it does. For, some of the reasons against (H3) that he does not mention would actually seem to align theatrical magic with theater and some movies. (As Jason himself writes in another passage, in reference to Teller’s nearly correct definition of magic, “there is no mention of deception,” because magic is understood as a form of theater.) Clearly, theatrical spectators to any narrative performance perceive, things about situations, characters and events in the stories they watch. They perceptually recognize the characters, they come to discriminate among the options open to the characters, and their perceptions and discriminations lead them to hope for better outcomes than they believe are possible. Did they not do such things, we should think they simply had not understood the story they had witnessed. (See Alex Neill, 1993. “Fiction and the emotions,” American Philosophical Quarterly 30/1: 1-13.) If that is so, then something seems right about the idea that spectators perceive, think, and feel that those characters are really present before them.

All this is close to what is known as the “paradox of fiction.” That paradox is typically cast in terms of a conflict of belief, just as the third hypothesis about the experience of magic (H3) states. (All of which Jason knows, by the way: see his footnote #17.) The reason Jason rejects H3 (correctly, I think) is that what is presented in magic, what the spectator perceives as really happening is known to be impossible. But this just brings out the question about the other side of the coin – that is, what is the cognitive attitude that is in conflict with such knowledge?

Because whatever it may turn out to be is never simply given up, Jason thinks we should analyze that state in terms of Gendler’s cognitive notion of an “alief,” a belief-like state. But congenial as that seems to be to his thesis, it might actually be excessive. After all, it is not clear that theatrical magic actually has the problem for which turning to “alief” is the solution. Recall that what is in conflict with the knowledge one has that such events are impossible is the “appearance” that they are really happening, right now.

This suggests that the conflict is between something a spectator knows (and is supposed to know and hold firm to) and an equally “unbudgeable” perception the spectator is having. In this way, the situation appears to be like that a subject has when experiencing the Muller-Lyer illusion, in which the subject’s knowledge of the illusion does not prevent the lines from giving her the appearance that they are different in length. Her perception persists despite her knowledge the lengths are the same. At one point, Jason acknowledges this – see footnote #24 – but offers an analysis of the perception in the Muller-Lyer illusion in terms of a cognitive state, namely, “alief.” I am not at all sure why he does that.


Reply (Jason Leddington):

First, thanks very much to the site editors for the honor of having “The Experience of Magic” offered up for the first AFB x JAAC session. I’m really looking forward to the discussion—particularly since this article is part of a larger project that is still in its early stages.

Second, thanks to the editors of JAAC, both for working with AFB to make this possible, and, again, for the comments and suggestions that made the article better than it otherwise would’ve been.

Third, thanks to Jim Hamilton for his insightful critical précis. I met Jim for the first time earlier this year in Antwerp at a terrific conference at the Centre for Philosophical Psychology. We had some great conversations then—over dinner and walking through the city—and I’m very pleased to be able to continue my exchange with Jim here. I’ve got a lot to learn from him about the philosophy of theater and performance.

Finally, speaking of learning, I’ve got a lot to learn from the readers of this blog about pretty much everything. So, thanks in advance for your participation, AFB readers. I hope you find it worth your while.


Now, on to Jim’s “Comments on some aspects of the argument.” I want to keep my replies pretty brief. Mainly that’s because I need to think more about the issues; but I also don’t want to control the discussion. I’m very curious to see what in the article strikes you as important, naïve, contentious, (in)defensible, underdeveloped, suggestive, etc. (I’m also really interested in parallels with other arts and related aesthetic phenomena. So, please comment freely.)

Jim brings up four main issues in his comments:

  1. Magic vs. “trick movies” (¶1 of his comments);
  2. Modal features of theatrical and cinematic vs. magical depiction (¶2);
  3. Perceiving characters in narrative performance (¶3); and
  4. Why alief? Why not perception? (¶s 4–6).

I’ll take them in order.

First, Jim correctly notes that, “[f]or many of the magicians of the early 1900s, magic and movies were seen to do similar things.” (Consider, for instance, the work of magician and filmmaker Georges Méliès.) But does this threaten the rather strong contrast that I want to draw between magic and movies? Well, even if early movie audiences sometimes had quasi-magical experiences before the screen, most contemporary audiences do not experience movies as magical. The difference? We have an intuitive grasp of how film technology works. More broadly, film, photography, radio, TV, cell phones, digital special effects, computers, etc., have all become part of the “naïve physics” of contemporary culture. This makes it very hard to do magic with such stuff, since hardly anything they do strikes us as “impossible.” (See, for instance, this performance by Simon Pierro on Ellen. Even if it nicely integrates technology, what’s magical about it is precisely not the technology.) That our changing relationship to technology affects what we experience as magical is a special case of the point I make on p. 260 of “The Experience of Magic”—namely, that “spectators with different cognitive resources may have very different experiences of the very same performance.” (And recall Arthur C. Clarke’s famous remark: “Any sufficiently advanced technology is indistinguishable from magic” (1973: 21 n. 1); conversely, once the technology is no longer “sufficiently advanced,” it’s no longer potentially magical.)

Second, Jim says that I don’t provide any empirical evidence for aligning theater and movies “with only” a kind of depiction that does not purport to present events as though they are really happening, and magic “with only” a kind of depiction that does purport to present events as though they are really happening. But in neither case is the ‘only’ mine. I accept that theater and film can depict events as though they are really happening; however, I believe this sort of depiction is “typical of neither” (p. 255), and that this would be easy to confirm empirically. Also, magic performances often involve ordinary fictional theatrical depictions; however, these depictions function alongside the really-happening depictions that constitute the magic performance per se. For instance, a magic performance may be developed into a theatrical play, as in Maskelyne and Cook’s “The Mystic Freaks of Gyges” (Steinmeyer, 2003: 97). (Alternatively, magic may be injected into a conventional theatrical production, as in the 2015 Chicago Shakespeare Theater production of The Tempest, co-directed by Teller.)

Next, Jim puts further pressure on my claim that ordinary narrative theater doesn’t (typically) depict events as really happening by noting that “something seems right about the idea that spectators perceive, think, and feel that…[the] characters [in the play] are really present before them.” I agree that something seems right about this, but I’m not convinced that it is right, or (even if it is) that it creates a problem for my view. Engrossed in a game of make-believe I may, on seeing Rita, say that I see Hera; but even if she’s acting the part well, should we really say that I genuinely see Hera? Isn’t my “seeing Hera” as much a part of the make-believe as Hera herself? And even if it’s right to say that I see the character Hera, we might insist that what the narrative depicts is not the actions of a character, but the actions of a god (which I manifestly do not see). Rita depicts the god by playing the character (the “role”); and even if this counts as making the character visible, it doesn’t mean that the narrative purports to really present what it depicts—namely, the god. So, I still feel comfortable saying that ordinary narrative theater doesn’t (typically) depict events as really happening.

Finally, why think of the cognitive dissonance magic induces in terms of belief-discordant alief? Why introduce a new theoretical concept when we could just stick to “appearances” and talk about belief-discordant perceptual experiences? First, I should say that I’m surprised by the degree of resistance that the notion of alief provokes in philosophical audiences. As I understand it, the idea is pretty straightforward. Thanks to “features of the subject’s internal or ambient environment,” unendorsed yet emotionally- and behaviorally-significant representations occur in the cognitive system (Gendler, 2008: 644). These representations are aliefs. Note that, so far, this says nothing about the kind of content these representations have, and Gendler herself is catholic on this point (2008: 643). Thus, the account is compatible with the idea that the relevant representations might actually have the sort of non-conceptual structure some philosophers think perceptual experiences have. So, then, why aliefs rather than perceptual experiences? First, it’s not clear that this is an either/or matter. Second, I think that the problem with focusing strictly on the conflict between belief and perception is that the resulting account is insufficiently general. Not every magic effect is as visual as, say, Copperfield’s flying illusion. The best examples here come from “mental” magic, or mentalism. For example, while mind-reading acts seem impossible, their impossibility is arguably not perceptually apparent. Suppose that David Blaine asks you to think of any number from 1 to 1000, places his hand on your forehead, concentrates, and then names it. Suppose he does this a dozen times with different people choosing different numbers. This seems impossible, but its apparent impossibility is not perceptual; rather, Blaine’s performance uses what’s perceptible to give a strong impression of an otherwise imperceptible impossibility: mind-reading. This “strong impression” is, on my account, the alief that he is reading minds.

Works cited
Clarke, A. C. (1973). Hazards of Prophecy: The Failure of Imagination. In Profiles of the Future: An Inquiry into the Limits of the Possible (pp. 12–21). New York: Harper and Row.
Gendler, T. S. (2008). Alief and Belief. The Journal of Philosophy, 105(10), 634–663.
Steinmeyer, J. (2003). Hiding the Elephant. New York: Carroll and Graf.

Image credit: Maskelyne and Cooke’s “The Fakirs of Benares” at the Egyptian Hall, 1884, via the British Library Collection



  1. Thanks again to both of you for this! I’ll start the discussion off with a random thought I had that may be related (in the spirit of Jason’s request for parallels with other arts and aesthetic phenomena):

    Do you think there’s anything similar going on with jokes (or maybe humor more generally)? I thought this in relation to your thought – which seems very true – that knowing how the trick (or piece of technology) works ruins the special experience of magic. Lots of people think that explaining a joke destroys it. I wonder if these are related.

    Second, do you think there are different ways of knowing how the thing in question works? For example, it seems like we could know the broad strokes (it’s a sleight-of-hand trick) or the details (I hide it *here* by diverting your attention *here*). Or another example, with jokes – the physical or psychological mechanisms that underlie humor, the sociological background, having a theory of humor, or knowing what’s meant to make the particular joke funny. This seems plausible, and I wonder if you think it’s relevant. Maybe some ways of knowing ruin and others don’t?


  2. Hi Alex,

    Thanks for your questions. I think you’re absolutely right that there’s an interesting relationship between magic and humor.

    For one thing, laughter is one of the primary responses to good magic (though it’s usually just one part of a complex emotional response; e.g., see the first scene from this clip from David Blaine’s first TV special: Not surprising if, in fact, most humor depends on incongruity (see, e.g., John Morreall’s work on this). After all, what’s more incongruous than an apparent impossibility?

    But I hadn’t thought of *your* point—namely, that magic and humor are both, in some ways, hostile to explanation. The first thing I’d want to get clear on is the sense in which explanation can ruin humor. Is it just because it ruins the “drama” of the joke, takes the punch out of the punch line? (A joke typically has a weaker impact if you only “get” it once it’s been explained to you.) Because there does seem to be a clear difference between magic and humor: while being able to get a joke seems to depend on being *able* to explain it, having the experience of magic depends on being *unable* to explain the illusion. Thoughts?

    On your second point, about various degrees of knowledge of how the trick works: in the article I claim that having *any* sense of how the trick was accomplished—even if it’s as general as “it was sleight of hand” or “somehow there must have been wires”—is enough to ruin it. You might still admire the illusion, but you won’t have the experience of magic. This means that the best magic technique conceals not only its specific nature, but also its very *existence*. For this reason, even though they are highly skilled, many magicians will make efforts to *erase* the appearance of skill from their performances. (As a performer, the best thing to hear from a spectator after a magic performance is, “But you didn’t DO anything!”) I’m not sure, however, whether there’s a parallel in the case of humor. That said, as with humor, I don’t think that having a philosophical theory of magic, general knowledge of magic as a performance art (including its history and relevant cultural facts), or knowledge of relevant features of human cognition in any way threatens the experience of magic. On the contrary, I’ve found that when people acquire a basic understanding of magic and what it’s trying to accomplish, they actually become more open to, and interested in, magic performance—and so, presumably more susceptible to the experience of magic itself.

  3. Here’s something I’d like to offer up for discussion: what do you think the difference is between witnessing a good magic trick that completely baffles you and the experience of looking at drawings of impossible figures? Consider:

    And is the experience of looking at those figures different in any important way from the experience of seeing this sculpture by Hemaekers?

    The sculpture arguably creates a more powerful (and baffling) experience than the drawings, since only the former seems to present you with an impossibility. But probably it won’t take you long to recognize that the sculpture looks different from other angles:

    And of course you also know this about magic. Because you know it’s a trick, you know that there must be an “angle” from which it doesn’t look impossible at all. But the difference is that strong magic not only hides that angle, it leaves you with no sense of how there *could* be such a hidden angle. (Ideally, you’ll come away feeling like everything was out in the open, visible.) So, strong magic is like an impossible sculpture (a) that looks impossible from every perspective from which you can see it, and (b) on which there seem to be no other possible perspectives.

Leave a Reply

Required fields are marked *.