About CA
Contact CA

Search Journal

Editorial Board

Permission to Reprint


Site Map


Philosophy with a Twist: La rivière du hibou

  Paisley Livingston
& Trevor Ponech

This paper explores the category of films known as “twist films” in relation to distinctions between different modes of epistemic access to works. With reference to the case of Robert Enrico’s short film, La rivière du hibou (1961), the philosophical significance of different sorts of twist films is explored. Twists are also discussed in relation to emotive responses, with special attention to the paradox of suspense.

Key Words
Ambrose Bierce, Robert Enrico, La rivière du hibou, “Occurrence at Owl Creek Bridge,” suspense, twist


In Aristotelian terms, a narrative twist is a conjunction of peripeteia and anagnorisis, where it is the spectator or reader, and not only the protagonist, who is meant to experience a combined reversal and discovery.[1] Small-scale twists are pervasive in narrative works but, in some cases, the entire plot of a work is organized in function of a major twist that is crucial to the understanding or appreciation of the work as a whole. Such cases are sometimes referred to as "twist films." In what follows, we argue that twist films may be philosophically significant in different ways. To that end, we say a few things about twists, in general, before we describe what we take to be some of the distinctive features of the twist in La rivière du hibou (directed by Robert Enrico, 1961).

1. Online offline: modes of cognitive access

We begin by mentioning a few distinctions pertaining to different modes of cognitive access to cinematic works. It is one thing to have an immediate perceptual experience of the audio-visual display of a work, and something else to think about a work in the immediate absence of such a perceptual experience. Some cognitive psychologists use computer terminology to name this distinction, contrasting online and offline responses. Although we are not generally fans of computer analogies in psychology, it may be acceptable to use the term ‘online’ to refer to a subject’s immediate perceptual, cognitive, and affective response to an audio-visual stimulus. We commonly distinguish between a subject’s first response and subsequent online responses to a work. The latter are normally influenced by comparisons and anticipations based upon memories of the prior online experience of that work, and also by recollections and reflections that take place after the first online viewing.

In contrast to online responses, offline responses are judgments or reflections about a cinematic work that are not part of an occurrent perceptual experience of the audio-visual stimulus. We can further distinguish between two kinds of offline responses to a work. One kind of offline experience is a matter of memories, critical reflections, and judgments referring to a work that the subject has previously experienced online. For the sake of convenience, we can call these “recollective offline” responses. A second kind of offline response is that of a subject who has never had an online experience of the work but instead thinks about it on the basis of testimony and other evidence. We call these “testimony-based” offline responses or attitudes. What is commonly referred to as a “spoiler” is a matter of testimony received prior to a first online experience, where this testimony is sufficient, if remembered and believed, to make certain kinds of surprises impossible.

2. Twist films

Returning now to the twist film, we can say that these are works designed to create a marked difference between what the spectator is normally led to infer about the story during much of the first online experience of the work, and what the spectator ends up believing about the content of the story during subsequent online and offline experiences of the work. More simply put, the spectator is induced to believe one thing about the story during a first viewing and is then led to recognize this belief as misleading.

Twists are different from suspense. Twists engineer a surprise for a target audience by propagating faulty inferences and beliefs in order to set the stage for the unforeseen disclosure of a veritas ex machina. Suspense generates and maintains an affectively charged anticipation of a course of events' outcome. This possible outcome is one the audience can easily foresee, though they might actively hope or wish that an alternate, equally foreseeable turn of events comes to pass instead. The twist's principal aesthetic effect contraindicates spoilers and recidivism. Spectators who rewatch the movie do not normally acquire the same mistaken, initial belief about the story events. That would only happen if they failed to remember what they learnt during their first viewing. On the other hand, recollective and testimony-based offline responses that remove uncertainty about outcome do not always preempt one's experiencing suspense.    

There are many straightforward examples of movies designed to get competent first-time viewers to adopt various narrative assumptions only eventually to disconfirm their audience's hitherto prevailing understanding of what the story is. The Usual Suspects (dir. Bryan Singer, 1995) is a good specimen. A man known to authorities as the petty con artist Roger “Verbal” Kint somehow survives a massacre aboard a ship rumored to carry $91 million worth of cocaine. Under interrogation, he recounts at length how he and four less fortunate criminal associates joined forces to rob and destroy the boat. Most of the movie consists of flashbacks supposed to present Verbal's narrative in cinematic form. After Verbal leaves the police station, the investigating agent, Kujan, is satisfied that he knows the identity and motives of the mastermind behind the heist and slaughter, disgraced police detective Keaton. While savoring his triumph over the pathetic Verbal and the arch villain Keaton, Kujan casts his eyes over the messy office borrowed for the interrogation. Assorted texts, pictures, and trinkets catch his eye. Only then does The Usual Suspects reveal that Verbal has, with much imagination, fabricated his account from an inspirational heap of random materials. Not even the cleverest spectators could be expected to realize prior to this scene that they, a little like Kujan, are being deeply misled by a bravura and far-reaching instance of unreliable narration.

There are, however, hybrid works where one category of spectators is meant to be surprised by the twist, while another category of spectators is not. Strictly speaking, La rivière du hibou is such a hybrid work. A spectator who is familiar with the Ambrose Bierce short story and aware that the film is an adaptation of this story should not be surprised when it turns out that the condemned man has hallucinated an elaborate escape in the moments before his death. Some spectators might also figure out well before the end of the film that the condemned man’s escape has an unreal quality and is part of some sort of dream or vision.

On the other hand, nothing in the film initially communicates to the audience that this is an adaptation of the short story by Bierce. The film's title isn’t a literal translation of the story’s title and so it doesn’t really point decisively in that direction. ‘Un événement au ruisseau de hibou’ would have come a lot closer. It is only in the credits at the end, after the twist, that the spectator finds an explicit indication of the film’s literary source. The situation is similar for the target audience of the version of the film presented on the American television series, The Twilight Zone. Only a small part of the American television public would have been familiar with this fairly obscure short story by Bierce, so the average American viewer in the 1960s could have been surprised by the twist on a first viewing of the film.

Our conjecture, then, is that this is a film that was intentionally designed to function as a twist film for many but not all first-time viewers. To make a claim about a work’s intended design raises questions about authorship, so perhaps a bit more should be said on this topic. We are not committed to the false idea that all authorship is individual authorship, and it could be more accurate to refer to the shared intentions or design that emerged in the collective production of a work. Sometimes there is joint authorship but, in some cases, there is a collaborative effort in which one or more persons play the role of the author. With regard to the authorship of La riviere du hibou, we know that the film was initially Robert Enrico’s idea and that he wrote the script, co-edited, and directed the film, but obviously he did not do everything that was artistically relevant to its making. For example, he wrote but did not perform the “Living Man” song that is an important part of the sound track. What Enrico tells us in his autobiography supports the hypothesis that he exercised sufficient decision-making control over the activities of other members of the team.[2] But this evidence may not be decisive, and, in what follows, we refer to Enrico et alia as the authors of the work.

3. Philosophical significance of the twist film

What is the philosophical significance of the twist film?  At first glance it may seem quite likely that, in general, the twist device has genuine philosophical potential. Perhaps it is not hyperbolic to say that liberation from error is one of the greatest promises of philosophy. One philosopher after the next has advised readers and interlocutors to become aware of the possibility that various opinions are erroneous. Pointing to the bases of our errors, the philosopher enjoins us to replace bad beliefs and bad reasoning with genuine knowledge or wisdom. Lucretius, for example, asks us to consider what it is like to watch other people struggling at sea or caught up in a gruesome battle. It’s not that we enjoy the sight of such misfortune, he comments, but we do enjoy the thought that we are removed from the trouble suffered by the others we observe.[3] Lucretius says it is similar with regard to philosophy since, if we accept the wisdom of the Epicurean system, we can enjoy the contrast between the security of our own beliefs and the errors and uncertainty suffered by others. Obviously it is not just Lucretius who made this sort of appeal for a better and more secure epistemic stance. Even the anti-dogmatic, skeptical philosopher has a policy of belief revision to recommend, such as a general suspension of potentially erroneous commitments.

It is tempting to think of twist narratives as a kind of practical philosophical tutorial involving the inculcation of better belief policies through the realization and correction of cognitive error. In order to get a point across about the springs of folly, the tutor leads the student into a cognitive trap, then releases him in the hope that he’ll draw a more general lesson about how to avoid such mistakes. On the hypothesis that people can sometimes learn from their mistakes, this could be a well-motivated narrative strategy. Even if doxastic voluntarism is false, because we have no direct, voluntary control over what we believe, we can indirectly influence our epistemic dispositions and thereby improve the ways our beliefs are formed and revised.          

But doubts can be raised about the idea that twists tend to promote an improvement of our belief policy. What is it, more specifically, that the spectator is supposed to learn from his or her experience of a narrative twist? If the cinematic evidence presented to the spectators was cleverly designed to deceive them with regard to important elements of the story, when the twist comes along they will have been taken in. At that point, the spectators can reflect that they had been correctly reasoning all along about the evidence that was given to them. They happened to have been wrong about the way the story turned out to be but, at the time when the evidence was first presented to them, they drew precisely the inferences they were supposed to draw. Since they correctly reasoned about the evidence they were given, and because there was no other evidence to which they ought to have had access, it would be wrong to say that they were guilty of any kind of epistemic irrationality. The spectator who is taken in by the twist has understood that part of the film competently and precisely in the way it was meant to be understood and, so, cannot be blamed for making some kind of cognitive error.

If that is how twists work, the lesson they convey is that sometimes we can be tricked and in error through no fault of our own. Rational, evidence-based reasoning isn’t sufficient to protect us from error here, and if we do happen to escape from some error, it is only through luck or external circumstances. Just as Peyton Farquhar has no control over the illusory vision he experiences before his death, the victim of a plot twist can only go where the narrative leads him. If this is right, then some kind of epistemic fatalism would be the implicit message of the twist, which is very bad news for both virtue epistemology and internalist models of justification. And if that is the only message of twist films, then spectators would not be given any reason or motivation to try to avoid being deceived when another twist film comes along.

This dire conclusion can be resisted. Whereas some twists are, indeed, designed to make epistemic puppets of the spectators, others aren’t. In this second kind of twist, when spectators look back at what happened, they have reason to believe that their reasoning about the evidence was somewhat faulty, and learning how this was so could be instructive for them.

4. A hybrid twist: La rivière du hibou

Let’s consider La rivière du hibou with the distinction between rational and irrational kinds of twists in mind, where a rational twist is one that deceives even a perfectly rational spectator, as in the example of The Usual Suspects. Early in La rivière du hibou, the spectators are led to align themselves with the condemned man. Point-of-view shots and various other devices encourage this alignment with the man the soldiers coldly prepare to kill. For example, the editing structure in the sequence on the bridge is designed to prompt the spectator to share in the prisoner’s attempt to look about in the hope of discovering some means of escape. Whereas the narrator of Bierce’s short story informs us that Farquhar is a wealthy slave owner who actively supports the Confederacy’s cause, the film conveys no such information but instead presents us with an apparently innocent civilian who is about to be executed. Since the bridge is visibly intact, this gentleman does not seem to be guilty of some major act of sabotage. Is the execution really justified? The spectator is meant to be sympathetic towards this man’s desire to escape.

When the cinematic image prompts us to imagine seeing that the rope has broken and that the man has some hope of survival, spectators may be relieved or even thrilled. When he appears to have gotten away and rejoices, we share in his appreciation of the value of life, his simple joy at being a “living man” who marvels at the light shining through the leaves and at the intricate web of a spider. We want to see this living man regain the embrace of his smiling wife. Yet when we learn that this has all just been a vision prior to death and that there is no real escape, the upshot is a striking vanitas meditation, a poignant reminder of the inevitability of death. One simple point that viewers may take away is that they ought to appreciate life more while they still have the chance, just as the man savours his existence most when its possible loss has become salient to him.

Twist films vary with regard to the rewards of or interest in subsequent viewings in which the surprise is lacking. In some cases, the surprise occasioned by the twist is the film’s most important effect. If consumed a second time, the work is like sparkling wine that has gone flat. Yet this is not true of all twist films. We want to argue that Enrico et alia’s work is a good example where the interest of the film is not exhausted by the twist. There is a very particular reason why that is so, a reason linked to the film’s philosophical point.

Spectators who see the film a second time, with the twist in mind, may be in a position to detect any number of previously unnoticed features of the cinematic presentation. For example, although the prisoner has been rather conveniently swept far downstream by the powerful rapids that suddenly turn up in a small creek, a shot from a cannon back on the bridge strikes impossibly close to where he is resting on the shore. When Farquhar is running along the road, we hear loud, obtrusive drum rolls on a snare drum. The camera, which tracks along in front of him, seems to play a cat and mouse game with the runner. Every time he appears to catch up with the camera, it pulls back out of reach again. This is repeated several times until he is exhausted. The visual rhetoric here seems to suggest that the man will never get where he’s trying to go, which turns out to be correct. Another unrealistic aspect of the film, which is mentioned by Enrico as having been planned in the early stages of the production, is that the wintry landscape at the beginning of the film stands in sharp contrast to the leafy, springtime landscape in which the dream takes place.[4]

When Farquhar finally seems to reach his plantation, a majestic gate magically swings open to let him in. This does not square well with a realistic escape but it is perfectly coherent with the entire escape being some kind of hallucination or dream. Spectators who notice such things only on a second viewing may wonder why they didn’t notice and take proper account of them during the first viewing. Why didn’t we see through the deception? To that question there is both a simple and a more subtle answer. The simple answer is that in the absence of overwhelming evidence for the subjective or hallucinatory status of the events, the spectator’s default or spontaneous assumption is that the represented events are what take place in the story conveyed by the work. The spectator is justified in going along with this on the first viewing since, after all, improbable things do happen in fictional stories, and we had no sure way to infer that this was not one of them. However, on a second viewing, we notice things that make our prior acceptance of the unlikely escape a lot more questionable. It is not only that the escape is an improbable fantasy. The magical opening of the gate and other evidence only really make sense when understood as parts of a representation of the character’s hallucination.        

The upshot is that, on a second viewing, we may recognize a difference between the twist in this film, where we are somewhat irrational to let ourselves be deceived, and other twists, where even the most alert and rational first-time spectator is tricked. And what is the error that we make if we are taken in by the twist in this film? Assuming that the target spectator is significantly aligned with the condemned man and with his desire to stay alive, it’s likely that a mild version of motivated irrationality is at work in our response to the audio-visual evidence. We discount the evidence that there is no real escape and that death is inevitable. And if that is what we realize, the film can function as a vanitas reminder, not only about our mortality, but also about our tenacious, motivated inclination to overlook the facts. The conclusion does not have to be the pessimistic thought that we are doomed by our own irrationality to make cognitive errors or that it is only a lucky twist of fate if one of our illusions gets corrected. Instead, the point could be that we should remain reflective and not let our wishful thinking lead us into error about our own mortality.

5. The paradox of suspense

Like many twist films, La rivière du hibou blends suspense into its planned effects on viewers. Enrico strongly prevails upon us to worry about whether the story's hero will reach safety. This facet of the movie serves its overall artistic design. Spectators emotionally worked up into a state of suspense are less apt to notice and ponder evidence that Farquhar is already a dead man. We might expect that those who already know or somehow manage to predict the twist will not merely be unsurprised by the turn of events but also relieved of any tension arising from the narration's gradual, strategic unfolding of the protagonist's fate. But this supposition is contradicted whenever viewers, despite their recollective offline knowledge of the narrative's outcome, report that they re-experience anxious thoughts and feelings when re-watching the depiction of Farquhar's escape attempt.

Here our discussion brushes past the philosophical literature on the paradox of suspense.[5] This puzzle derives from the incompatibility of three beliefs that people might be inclined to endorse individually: (1) Suspense requires uncertainty about an outcome. (2) Audiences who know this outcome in advance are deprived of the requisite uncertainty. (3) At least some repeat consumers of suspenseful narrative fictions re-experience suspense to varying degrees of intensity. Some philosophers solve the puzzle by positing minor lapses of the audience's rationality. This irrationality consists of, more or less, cold, unmotivated cognitive errors deriving from viewers' finite capacities for information storage and retrieval, judgment, categorization, and causal assessment. Hence, one way to solve the paradox is to deny premise (2) by surmising that informed spectators immersed in the narrative's flow temporarily forget what they know.[6] Episodes of “anomalous suspense” are caused by a cognitive mechanism, namely, an expectation that no two events or experiences thereof will be exactly alike. This subconscious information-processing bias generally optimizes agents' cognitive resources but impedes recall of facts about previous encounters with the movie. The notion that such a mechanism prevails in ordinary spectatorship is, however, contrary to the evidence, as repeat viewers often look for clues, notice details, and arrive at realizations because they are primed and enabled to do so by their recollective offline responses and attitudes.

An alternative solution rejects premise (3) by positing that those watching a film more than once are prone to confusion about the nature of their psychological condition.[7] The “raw feel” of an introspectable affective state can be difficult to describe and label. A viewer's top-of-the-head self-attributions and post hoc verbal reports can misidentify his or her emotions. One might mistake the qualitative features of one’s experience for those belonging to suspense when they are actually those of some other phenomenally similar emotion, such as looking forward to a known positive outcome or apprehensively awaiting a negative one. Emotions surely are complex. Live ones can be hard to precisely identify, and a priori conceptual distinctions difficult to formulate. However, it remains to be demonstrated that rational, competent spectators conversant with the full range of responses called “suspense” are typically as fallible as the misidentification thesis boldly predicts. A verdict on this question depends on how we conceive of suspense and, in particular, whether uncertainty is necessary for its occurrence.

Our proposal is that suspense is a cluster of familiar psychological responses involving active, emotionally aroused anticipation of a turn of events' impending outcome or resolution. Characteristically, this outcome is one over which the agent has no control or only weak influence. It also has a marked hedonic valence for the agent. We further suppose that the outcome's perceived imminence, like its degree of hedonic value and uncontrollability, influence the intensity of emotional arousal.

Many ordinary situations reliably trigger psychological episodes of this sort. Awaiting announcement of grant competition or medical test results; hoping the roulette ball stops at a particular number; watching and wondering if Farquhar escapes; and re-watching the unfolding of Farquhar's predeath fantasy could all elicit suspense. In some cases, the outcome that is the emotion's target is quite serious, insofar as how things turn out could significantly affect the agent's welfare or personal projects. The hedonic valences of other possible outcomes are more ludic and belong to categories of human interests associated with play and aesthetic experience. Some, like those of the roulette wheel, can go either way.   

Our distinction between surprise and suspense noted that suspense is comprised of a readily foreseeable prospective outcome, where the audience might hope or wish that an alternative, likewise foreseeable turn of events more to their liking transpires instead. Somebody in a state of suspense will usually have at least a rough idea of the different, unequally preferred ways in which the events they are more or less helplessly observing unfold might conclude. This agent's degree of uncertainty about how things will turn out is thus another factor in the emotion's emergence and strength. Second and subsequent viewings of La rivière du hibou won't be at all surprising—and might not be as suspenseful.

Yet foreknowledge need not entirely preclude suspense. Once our initial error has been corrected and we are alert to the cinematic display's presentation of Farquhar's subjective hallucination, we are not automatically freed of our self-serving desire for a happier, albeit decidedly out-of-reach, ending. For sympathetic spectators who wish that death would spare him, even though they know it will not, repeat experiences of the movie's harrowing representation of its protagonist's survival struggle might retain a measure of anxiety and frustration. The emotional arousal they experience belongs on the suspense spectrum.    

Diverse sorts of artistic displays, games, and other contrivances are meant to trigger and frustrate agents' desires that events progress toward and terminate in a certain preferred outcome. Frustration arises because these displays afford observers little, if any, opportunity to make things turn out the way they want. This claim seems consistent with a desire-frustration theory of suspense. According to this theory, rather than one's actual or entertained uncertainty about outcome, it is one's feeling an unrealizable desire to make a difference that generates and identifies suspense.[8] Successful narratives within the suspense genre are skillfully calibrated to make audiences urgently want to ensure or correct an event's imminent outcome, and often give audiences privileged knowledge of how characters could be saved from trouble or harm. But they are helpless to do anything with this knowledge and in furthering  their wishes.    

If viewers desire to assist characters or alter story events, this psychological disposition must be of a special sort. The rational spectator knows that the inaccessible, possibly fictive events and characters depicted by the narrative are not the kinds of entities that he or she can help or change. The supposition that a spectator genuinely and strongly wants to do so is therefore contentious. If we are to predicate an analysis of specifically narrative suspense on the idea that such vehicles of suspense typically provide gratifying experiences of frustrated desire, then care should be taken to differentiate between emotional responses to narrative artworks and, say, to impending news of whether you need knee surgery.

The viewer's wish to participate in and causally affect the narrative's outcome is superficial. It is but a mild, distracting taste of the serious, non-aesthetic anticipatory experience of helplessly observing a turn of events, where one is preoccupied by a concern with a subjectively unsatisfying prospective outcome. A potentially illuminating way of understanding our characteristic responses to narrative suspense is to regard them as artificial emotions.[9] Such emotions are real and involve no mental acts of making believe or pretending to be moved. They count as artificial in part because they are occasioned by artistic representations rather than by the represented events directly witnessed. Their artificiality also pertains to their transience and weakness. It is easier to turn our minds from such feelings, which lack the long-term and serious consequences that their analogous, actuality-inspired emotions would have.

The account of narrative suspense we favor might help explain how such a possibly negative emotion, inasmuch as it is distressing or painful, is readily enjoyable. Artificial emotions, even when unpleasant, are apt to provide pleasurable relief from boredom. Representations of violent, risky, and disturbing events afford the excitement of intense affect without the dangers to mind and body posed by real life crises. Moreover, proxies for negative emotions like suspense are frequently accompanied by a further compensation. No matter how frustrating one finds Farquhar's predicament, La rivière du hibou offers up for appreciation ample artistic merits. It likewise exhibits semantic and formal properties suited to play a heuristic role in our thinking about certain philosophical topics.

We have argued that some works with a twist, including La rivière du hibou, are best appreciated on a second viewing that reveals ways in which the twist was both prepared and subtly undermined. There are literary antecedents to this narrative strategy. Bierce’s short story, however, is not one of them. This is a case where a cinematic adaptation merits praise for some of the ways in which it diverges from and improves on the literary source. Bierce’s narrator is an omniscient liar whose propositions imply reliable access both to Farquhar’s mental states and to all sorts of relevant external facts, including facts unknown to Farquhar. It is only in the penultimate paragraph that this narrator gives the reader any reason at all to suspect that there is something strange or dubious about Farquhar’s supposed escape.

Simply put, the narrator tells us about the escape in order to trick us, abruptly taking back much of what he has said. Rereading the story, one doesn’t find foreshadowing or clues but bald deception on the part of the narrator, who tells the reader, for example, that Farquhar “knew that the rope was broken and he had fallen into the stream.”[10] If it is Farquhar’s neck, and not the rope, that is actually about to be broken, how could he “know” that the rope had already broken? He couldn’t, yet that is exactly what the omniscient narrator asserts. Bierce’s reader is, then, an epistemic puppet, and the rational first-time reader is to be tricked by the story twist. In the case of the film, a fully rational and informed spectator is initially tricked but gathers contrary evidence and corresponding doubts along the way, well before the shots depicting Farquhar’s demise.

A better antecedent for the alternative working of the twist in La rivière du hibou is Jane Austen’s Emma. Rereading the text of this novel, we realize that the narrator dangled all kinds of evidence before our eyes but that we let the narrator’s free indirect discussions of Emma’s vanity and conceit lead us to overlook this evidence. Similarly, in responding the first time to La rivière du hibou, the error many, but not all, spectators end up making is motivated and wishful. Spectators may subsequently be led to realize why they failed to draw a more lucid conclusion, and that’s a worthwhile and moving philosophical lesson. Many of us are like Farquhar at the end, rushing hopefully forward, never quite getting there, rushing hopefully forward, until the inevitable interruption.

Trevor Ponech

Trevor Ponech is Associate Professor of English, McGill University, Montréal.  His work on the philosophy of cinema appears in The Journal of Aesthetics and Art Criticism, The British Journal of Aesthetics, The Routledge Companion to Philosophy and Film (eds. Paisley Livingston and Carl Plantinga), and Cine-Ethics (eds. Jinhee Choi and Mattias Frey).

Paisley Livingston

Paisley Livingston is Chair Professor of Philosophy at Lingnan University, Hong Kong. He is the author of Art and Intention and Cinema, Philosophy, Bergman (OUP). His current research centres on the aesthetics of Bernard Bolzano.

Published on June 30, 2016.           


[1] Aristotle, The Poetics, trans. W. Hamilton Fyfe (Cambridge, MA: Harvard University Press, 1927), p. 1450a, 17.

[2] Robert Enrico, Au cœur de ma vie (Saint-Cyr-sur-Loire: Christian Pirot, 2005).

[3] Lucretius, De rerum natura, trans. R. C. Trevelyan (Cambridge: Cambridge University Press, 1937), II.1–8.

[4] Enrico, Au cœur de ma vie, p. 71

[5] See Aaron Smuts, "The Paradox of Suspense," The Stanford Encyclopedia of Philosophy (Fall 2009 Edition), Edward N. Zalta (ed.), URL = <http://plato.stanford.edu/archives/fall2009/entries/paradox-suspense/>.

[6] Richard Gerrig, Experiencing Narrative Worlds (Yale University Press, 1993),  pp.168-175; “Is There a Paradox of Suspense? A Reply to Yanal,” British Journal of Aesthetics, 37 (1997), 168–174

[7] Robert Yanal, “The Paradox of Suspense,” British Journal of Aesthetics, 36 (1996), 146–158.

[8] Aaron Smuts, "The Desire-Frustration Theory of Suspense," The Journal of Aesthetics and Art Criticism, 63 (2008), 281-290.

[9] Paisley Livingston, "Du Bos' Paradox," British Journal of Aesthetics, 53 (2013), 393-406.

[10] Ambrose Bierce, “An Occurrence at Owl Creek Bridge,” in Tales of Soldiers and Civilians (San Francisco: Steele, 1891), pp. 21-39, ref. on p. 29.