I
Suspense is well known for sparking the eponymous paradox of suspense, which centres on the very possibility of experiencing suspense concerning what one already knows.1 In this paper, we suggest that suspense is apparently paradoxical in yet another way. Put briefly, suspense involves an inconsistency such that one does and does not want to know the outcome of the story. Call this the desire paradox of suspense. In what follows, we use this apparent paradox to carve out what may be special about suspense as a particular mode of engaging with narratives, and we flesh out some details of this experience.
The desire paradox of suspense consists of two claims that appear to be contradictory. Let us refer to the first claim as the desire claim (henceforth DC):
DC Readers who experience suspense want to know the outcome of the story.2
What speaks in favour of DC?
Current definitions of suspense either entail DC or at least suggest something like it. Thus, Thomas Anz straightforwardly defines suspense in terms of an ardent desire to know the proceedings of the story.3 For other theorists, suspense comprises something that should at least come along with, or motivate, wanting to know the outcome of the story. The ingredients of suspense for which this holds true include apprehension, hope, and fear.4 Apprehension, hope, and fear are forward-looking states that involve a motivation to advance what you want (hope) or evade something unwanted (apprehension, fear). Since any act of advancement or evasion should come with some forward-looking epistemic base (you need to know what to advance or evade and how to do it), it should also come with some motivation to know about these things. Therefore, if you think that suspense involves apprehension, hope, or fear, you should happily embrace DC. Also, if suspense heightens reading motivation (readers do not want to put away the book, and so on), then this is easily explained by DC.
If you believe that uncertainty concerning story outcome is part and parcel of suspense, then DC suggests itself as a natural concomitant of your theory. Uncertainty is described in the psychological literature as something that comes with a motivation to reduce uncertainty, hence a desire to know what is the case.5 This is especially plausible if there is something that is at stake; that is, the uncertainty concerns states of affairs such that readers have an emotional investment in them, as may be the case concerning outcome states in suspenseful stories.
The second constituent of the desire paradox of suspense is the negation of DC. Let us refer to this as the non-desire claim (NDC):
NDC Readers who experience suspense do not want to know the outcome of the story.
What speaks in favour of NDC?
Suppose you are in the grip of a suspenseful story and you are simply given the outcome of the story right away. This is a spoiler. NDC thus accounts for the preferences of readers who do not like spoilers but rather prefer to experience the story as it is presented.6 Also, we may safely assume that authors do not want their readers to skip the main (suspenseful) part of the story and jump right to the end. After all, authors take great pains to write their stories such that readers’ desires concerning story outcomes are built up, maintained, and shaped as the story progresses. The ‘intended’ reader thus refuses to be told the outcome before it is due. And so would the ‘ideal’ reader, that is, a reader who realizes the suspensefulness as part of the aesthetic qualities of the story in an optimal or even best way. In sum, then, while NDC as it stands is an empirical hypothesis, it also has a simple explanatory role in a both simple and plausible theory of artwork reception, and it may thus participate in the credibility of this very theory.
II
A paradox indicates that something is wrong with your assumptions about some subject because there is an unacceptable consequence that follows from these assumptions. However, the desire paradox may not be of this sort. After all, the apparent contradiction between DC and NDC stems from (the content of) what a person desires. Conflicting desires appear frequently. Sometimes one does and does not desire something (‘Peter wants Anna to leave and he does not want Anna to leave’), or one may desire things that are incompatible (‘Peter wants to buy the cake and save the money to buy something else instead’), or one may desire something that one knows to be impossible (‘Anna wants to meet Napoleon although she knows that he is dead’), or one may desire things against one’s all-things-considered best judgement (‘Peter wants another beer although he thinks this is bad for him’). The desire paradox, then, could be said to be just another instance of our all-too-common propensity to desire what is impossible in one way or the other, or what is against one’s better judgement.
However, we must not be too quick here. First, note that the desire paradox is not about some accidental mishap; rather, it describes the common structure of the reader’s desires. Both DC and NDC are general claims. The conflicting desires in question are said to be regular concomitants of reading a suspenseful narrative for any reader. In this way, they seem to differ fundamentally from the examples of conflicting desires mentioned above.
Second, as our brief list of examples indicates, there are different types of conflict between desires as well as neighbouring cognitive states, and, if reading a suspenseful narrative involves such conflicts, we may want to know what kinds we are dealing with. For instance, if the claims of the desire paradox are to be read analogous to ‘Peter wants Anna to leave and he does not want Anna to leave’, then they seem to express some sort of ambivalence on the reader’s part. Peter’s situation is one in which he does not know what he (really) wants. But does that sound adequate as a description of the reader’s predicament? Relatedly, we may wonder whether the formulation of either DC or NDC in terms of ‘wants’ or ‘desires’ is adequate. Compare again: if Peter wants another beer contrary to his all-things-considered best judgement, then we may wonder, on conceptual grounds, whether this really counts as ‘wanting’ another beer or whether it is more adequate to say that he experiences a desire for a beer but does not want it.7
Third, things somehow do not go smoothly for Peter if he wants Anna to leave and does not want Anna to leave, if he wants to buy the cake and save the money, or if he wants another beer although he thinks this is bad for him. (This is true even more so if poor Peter experiences such states regularly.) Apparently conflicting states such as these are less than optimal epistemically or practically, and we may wonder whether such a negative appraisal suggests itself for the state we are in when reading a suspenseful narrative. And, if it should turn out that reading a suspenseful narrative involves states that are, say, irrational or acratic, then we may want to relate this finding to received assumptions about the highbrow qualities of art consumption.
The foregoing considerations should suffice to show either that the desire paradox cannot be understood along the lines of ordinary conflicts of the will, or else that, if it should turn out to be nothing but an instance of such conflicts, then this is a notable discovery that needs some arguing on its behalf. In any case, more needs to be said about the apparent contradiction between DC and NDC. Let us try to take the desire paradox seriously, then, and see what it can teach us about the psychology of suspense.
Standard procedure has it that in solving a paradox, one can reject at least one of its claims, or one can offer a reading of at least one of the claims such that the apparent contradiction disappears. A natural starting point for rejecting DC is the assumption that uncertainty is not a defining part of suspense.8 If uncertainty is not necessary for suspense, then you would expect that there are cases of suspense where you know the outcome of the story, and if you know the outcome of the story, then it does not make much sense to claim that you want to know the outcome of the story – or so the argument goes.
We do not think that an argument against DC along these lines is promising. As we have just stressed, DC is not only about a reader’s epistemic state but it is primarily a claim about what a reader wants. If you have established that uncertainty is not mandatory for suspense, then you have not thereby established that there is no relevant desire in suspense. Wanting to know the outcome need not be explained in terms of an epistemic deficit (uncertainty concerning the story outcome) that readers want to overcome. You may want to know the outcome just because you feel that otherwise things are not rounded off and accounted for. Maybe you want it because you expect the pleasing experience that comes with closure. We could rephrase DC as ‘Readers who experience suspense want to be told the outcome’ (and then rephrase the second claim of the paradox to be explained below accordingly).9
Let us now turn to a reading of DC such that the contradiction disappears. Both epistemic deficits and desires are states that a person has at certain points in time. It may be true of Anna that she wants to know the name of her neighbour’s cat today but this did not occur to her yesterday. Yet there is no contradiction involved. The sentences ‘Anna wants to know the name of her neighbour’s cat’ and ‘Anna does not want to know the name of her neighbour’s cat’ should be read as involving a temporal index that anchors the respective desires to different temporal intervals.
So, let us add temporal indices to the claims of the desire paradox. DC could be given a reading such that the desire to know the outcome of the story obtains only when it is due – namely, when the suspense arch is about to be closed and the storyteller is (almost) done. Let us refer to this point as ‘t2’ and to some earlier point within the suspense arch as ‘t1’:
DC-1 Readers who experience suspense want to know the outcome of the story (only) at t2.
NDC in turn now reads thus:
NDC-1 Readers who experience suspense do not want to know the outcome of the story at t1.
Simply put, readers develop different desires in the course of their reading activity. At first, they do not want to know the outcome of the story (but rather enjoy the suspense). Later on, however, they want to be given the playfully withheld information.
As it stands, this solution to the desire paradox of suspense has some drawbacks. For the contradiction to vanish readers need to deny information concerning the outcome of the story at t1. This is basically what NDC-1 claims. However, NDC-1 flies in the face of all the evidence that we have accumulated above in favour of DC. If suspense comprises something that regularly comes along with, or motivates, wanting to know the outcome of the story (such as apprehension, fear, or uncertainty), then one needs to explain how and why these motivational powers are regularly blocked. (Note that it may be difficult to maintain that states such as apprehension, fear, or uncertainty are present in the first place if one needs to argue that they do not exert their regular motivational powers.) Also, one needs to produce an alternative explanation for the overt behaviour of readers – namely, their propensity to eagerly keep reading until the outcome of the story is provided.
Next, suppose that a reader does not want to know the outcome of the story at t1. It is implausible to assume that wanting to know the outcome of the story comes as a surprise towards the end of the story (at t2). Rather, if readers do not want to know the outcome of the story at t1, then this should come with knowing that they will want to know the outcome of the story at t2. In other words, a reader who enjoys a suspenseful passage of text and who does not want to receive a spoiler (at t1) knows perfectly well both what a spoiler amounts to (namely, untimely information concerning the story outcome) and that they can expect – or at least hope for – the relevant information at the right time (at t2). For suppose that you tell the reader (at t1) that they will never get to know the outcome of the story. That surely amounts to a spoiler, too. Accordingly, the reader’s state (at t1) should be something like this: ‘I do not want to know the outcome of the story (now) but I will want to know it later (at t2)’. An accordant reading of the first claim of the desire paradox would be this:
DC-2 Readers who experience suspense anticipate (at t1) that they will want to know the outcome of the story (at t2).
This characterization has it that suspense involves higher-order attitudes (at t1). The hierarchical structure involved is quite common. Buying groceries, for instance, need not involve a desire to eat now; rather, you buy groceries now because you know that you will want to eat them later. And just as the (at t1) anticipated desire to eat (at t2) may motivate buying food (at t1), we could say that the (at t1) anticipated desire to know the outcome of the story (at t2) motivates reading on (at t1).
However, we think that it is implausible to attribute this psychological structure to someone who is in the grip of a suspenseful story. Remember that, while the phrase ‘want to know’ as part of DC need not entail an epistemic deficit, it does entail a desire. There are different sorts of desire.10 While all desires involve a pro-attitude towards something and a motivation to pursue it in some way, we think that, as part of DC, we should understand ‘desire’ to also imply some sort of a felt urgency. There is some way it is like to have this (kind of) desire, a phenomenal quality, which marks the desire as an occurrent mental state. Arguably, this felt urgency gives feelings of suspense at least some of their characteristic phenomenal flavour. In contrast, an anticipated desire – such as the anticipation of desiring to eat some fruit tomorrow that you buy today – lacks this phenomenal aspect. (Compare: anticipated pain is phenomenologically very different from actual pain.) Accordingly, it is implausible to assume that the desire that is mentioned in DC is an anticipated desire as DC-2 would have it.
However, while the grocery-buying analogy breaks down and DC-2 does not map the reader’s psychological state, the idea of a second-order desire might be on the right track. We think that what is characteristic of experiences of suspense is that you do want to know the outcome of the story and you do not want your experience to be spoiled. That in turn means that you do not want to give in to temptation and, say, skip a few pages in your book and jump to the outcome of the story immediately. In other words, you do not want to act on your desire to know the outcome of the story. In the suspenseful part of the story, you instead distance yourself from your desire to know the outcome of the story. Only when you have lived through the suspense arch that the story provides may you feel that the time is ripe for giving in to temptation, that is, to embrace your desire to know the outcome. Accordingly, what changes in the course of your reading is not the desire to know the outcome – this desire is present all along. What does change is a higher-order desire that is directed at the first-order desire to know the outcome. A more precise version of the second claim of the desire paradox thus reads:
NDC-3 Readers who experience suspense do not want to want to know the outcome (at t1).11
Accordingly, experiences of suspense do involve conflicting desires. But there is no clash of first-order desires such that you do and do not want a particular thing. Rather, the desires that are present (at t1) feature different objects. If you say that, while reading a suspenseful passage of text, ‘you do not want to know the outcome’, then this is shorthand for a more complex mental attitude involving a higher-order desire that is directed at your desire to know the outcome. More specifically, the higher-order desire is not a desire not to have the first-order desire. Rather, it is a desire not to act upon it (in spite of its felt motivational power).
In insisting that suspense involves a conflict of desire, the account given so far preserves the idea that experiences of suspense involve experiencing an urge – namely, the urge to know the outcome of the story (DC). However, the account also explains the internal (felt) tension that is part and parcel of suspense. This tension results from a conflict of desires and is captured by NDC-3.
The account may also explain some of the enjoyment of suspense, since experiencing suspense not only involves suffering from an occurrent desire (to know the outcome of the story) that is (temporarily) unfulfilled but also involves exerting some self-control. Such self-control may be enjoyed in itself. We may also speculate that the suspense experience is the better the stronger the first-order desire to know the outcome of the story is. As a consumer of suspenseful narratives, you may thus want your desire to know the outcome to be especially strong. In sum, we suspect that the higher-order structure as introduced is an important factor in suspense with its phenomenal profile involving an urge, tension, and enjoyment, and possibly distinguish suspense from states such as fear or apprehension simpliciter.
Could our analysis be challenged as unnecessarily complex? How about we understand DC such that it encompasses the aversion to spoilers as expressed in NDC-3 in the first place? Here is an accordant reading:
DC-3 Readers who experience suspense want to know the outcome of the story only by following the story as presented by the artwork.
DC-3 gives a restriction on the reader’s desire by way of qualifying its content (rather than by way of introducing a higher-order desire as is indicated by NDC-3). Moreover, DC-3 implies that readers who experience suspense do not want to know the outcome of the story in any other way (other than by following the story as presented by the artwork, that is), which in turn offers itself as a reading of the very NDC. Thus read, there is no contradiction between the two claims.
We think that something like DC-3 is indeed true of readers who experience suspense. In order to see what is wrong with this as a critique of our analysis of the suspense experience, however, we need to introduce yet another distinction. There is a difference between what a reader desires regarding the story content and what a reader desires regarding his or her experience. Let us illustrate this distinction by way of example and fill in some details of DC and DC-3 respectively:
DC* Anna wants to know whether Van Helsing defeats Dracula.
DC-3* Anna wants to know whether Van Helsing defeats Dracula only by following the story as presented by Bram Stoker’s Dracula.
We think that something like DC* is an indispensable part of Anna’s reading experience if it is to count as suspenseful. The novel brings up this question once Van Helsing sallies forth to defeat the monster, and much of the unfolding story centres around the question whether the hero succeeds.12 As indicated above, the prolonged desire to know whether Van Helsing defeats Dracula is what keeps Anna eagerly reading and grounds the characteristic ‘urge’ that is part of suspense. DC* thus explains both Anna’s behaviour and an important part of the phenomenology of her experience. DC-3*, in contrast, does nothing of this sort. DC-3* is neither licensed by the narrative as part of a proper engagement with its plot, nor does it explain her behaviour or the flavour of her experience. We conclude that DC-3* is not part of Anna’s experience of suspense but rather reflects her conceptualization about this experience.13 Indeed, if we did claim that DC-3* were part of Anna’s experience of suspense, then we might have a hard time maintaining that Anna is ‘caught up’ in the story, as seems to be the case in much suspenseful reading. For being ‘caught up’ in the story apparently just means that one centres one’s attention on the unfolding of the story (rather than on ‘success conditions’ of one’s experience, as DC-3* has it).
III
Two concluding remarks. First, when we claim that experiences of suspense involve us in inconsistency, we do not want to say that this involves any sort of irrationality. There is nothing wrong with your belief system if you undergo experiences of suspense, hence there is no charge of theoretical irrationality. But there is no practical irrationality either, such that you fail to act on what you think you should do. If you do not want to want to know the outcome of the story, then this may reflect a perfectly rational all-things-considered judgement to the effect that enjoying the suspense (rather than spoiling it) is what you should do.
Second, a description of an inconsistency is something very different from an inconsistent description. While the latter shows that there’s something wrong with your theory, the former is not something to be explained away but rather something that, in our case, should be embraced as part of an explanation.
Notes
[1] Aaron Smuts, ‘The Paradox of Suspense’, in Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, winter 2021 ed. (Stanford University, 1997–), https://plato.stanford.edu/archives/win2021/entries/paradox-suspense.
[2] For the sake of simplicity, we shall refer to those who consume a suspenseful narrative as ‘readers’, but it should be clear that our considerations apply to other media modalities as well. Also, we use ‘outcome of the story’ as covering any kind of plot-directed future proceedings of the story that are the focus of a reader’s suspenseful anticipation.
[3] See Thomas Anz, ‘Spannung’, in Reallexikon der deutschen Literaturwissenschaft, ed. Jan-Dirk Müller et al., vol. 3 (Berlin: de Gruyter, 2003), 464.
[4] See among many others Andrew Ortony, Gerald L. Clore, and Allan Collins, The Cognitive Structure of Emotions (Cambridge: Cambridge University Press, 1988); Ed Tan and Gijsbert Diteweg, ‘Suspense, Predictive Inference, and Emotion in Film Viewing’, in Suspense: Conceptualizations, Theoretical Analyses and Empirical Explorations, ed. Peter Vorderer, Hans J. Wulff, and Mike Friedrichsen (Mahwah, NJ: Erlbaum, 1996), 149–88; Christy Mag Uidhir, ‘An Eliminativist Theory of Suspense’, Philosophy and Literature 35 (2011): 121–33.
[5] See Arie W. Kruglanski and Shira Fishman, ‘The Need for Cognitive Closure’, in Handbook of Individual Differences in Social Behavior, ed. Mark R. Leary and Rick H. Hoyle (New York: Guilford Press, 2009), 343–53. The need for cognitive closure varies among individuals. Presumably, those readers who do not feel the need to reduce uncertainty concerning the story outcome are those who do not find the story suspenseful.
[6] Interestingly, there is empirical evidence that spoilers do not necessarily diminish reading enjoyment, so from a hedonic standpoint it may be unwise to reject them; see Jonathan D. Leavitt and Nicholas J. S. Christenfeld, ‘Story Spoilers Don’t Spoil Stories’, Psychological Science 22 (2011): 1152–54. We wish to remain neutral as to how an aversion to spoilers is to be evaluated (as irrational, for instance) and merely assume that it is a widespread attitude among readers.
[7] See Harry Frankfurt, The Importance of What We Care About (Cambridge: Cambridge University Press, 1988), 14, for the accordant distinction.
[8] See for example Aaron Smuts, ‘The Desire-Frustration Theory of Suspense’, Journal of Aesthetics and Art Criticism 66 (2008): 281–91; Pablo Delatorre et al., ‘Confronting a Paradox: A New Perspective of the Impact of Uncertainty in Suspense’, Frontiers in Psychology 9, no. 1392 (2018).
[9] Smuts argues that suspense need not be explained in epistemic terms at all but is the frustration of a strong desire to affect the outcome of an imminent story event (Smuts, ‘Desire-Frustration Theory’). Arguably, the desire paradox of suspense could be rephrased to align with this theory such that readers apparently do and do not have the relevant (non-epistemic) desires. We thank an anonymous reviewer for pointing this out to us.
[10] On the variety of desire, see Frankfurt, Importance of What We Care About, and Tim Schroeder, ‘Desire’, in Stanford Encyclopedia of Philosophy, ed. Edward N. Zalta, summer 2020 ed. (Stanford University, 1997–), https://plato.stanford.edu/archives/sum2020/entries/desire/.
[11] The phrase ‘do not want’ in NDC-3 allows for two different readings – namely, (1) ‘readers have the higher-order desire not to have the first-order desire’ and (2) ‘readers do not have the higher-order desire to have the first-order desire’. While (1) is about a proper desire, (2) is about its absence. Arguably, both (1) and (2) are part of experiences of suspense. However, we think that the stronger reading (1) is adequate, for it preserves the idea of conflicting desires and a resulting ‘tension’ (to be detailed below). We may therefore want to rephrase NDC-3 as ‘Readers who experience suspense want to not want to know the outcome (at t1)’.
[12] In other words, whether or not Van Helsing defeats Dracula thereby forms an important part of what has been termed the erotetic structure of this narrative. For background, see Noël Carroll, Theorizing the Moving Image (Cambridge: Cambridge University Press, 1996), esp. 95–100.
Acknowledgements
Previous versions have been presented to audiences at a symposium in honour of Harry Müller in Hamburg, at Humboldt-Universität zu Berlin, and at the University of Göttingen. We thank these audiences and two anonymous reviewers for Estetika for various helpful suggestions.