Thoughts on suspense and uncertainty

This entry is more of a condensation, review, and reworking of some material I want to remember than it is an analytic response paper; the main point for me, for now, is to practice thinking about what these things have to do with one another.

A fair helping of my literature list is novels that might fairly be categorized as novels of suspense. There are the books that are often marketed under that rubric, such as the commercial fictions of Graham Greene, John le Carré, and Eric Ambler. There are the books of other genres that also prominently feature suspense as a dominant emotional constituant, organizing principle, or what have you: the detective novels, sensation novels, gothic tales, maritime adventures, and impressionistic political romances -- here I am thinking of The Secret Agent and its parodic counterpart (at least in the first two thirds) The Man Who Was Thursday. And beyond that, many of the novels that do not seem overall to be novels of suspense feature suspense at their hearts: the agony of indecision at the absent center of Lord Jim, the constantly deferred and Kafkaesque search in Ice, even the interpersonal agony that comes to a head in the caves outside Chadrapore in A Passage to India.

Meanwhile, literary critics, film scholars, and popular entertainment buffs are not the only people to have noticed suspense; psychologists want to know about it, too. Much of the experimental work on the subject is organized around a quest to determine the necessary and sufficient conditions under which it occurs. The empirical research has been fairly modest, but it has generated several concrete hypotheses about the conditions that give rise to suspense for readers and viewers of narratives.

William Brewer and his associates (e.g. Brewer and Lichtenstein 1981; Brewer 1996) have attempted to identify the basic narrative structures that typically engender the experience of suspense. Their theory, based on experimental research with both artificially constructed texts and extant short stories and fables, is that suspense relies upon differences between the underlying event structure of a narrative (hey, it's the fabula again) and the presentation of those events in the discourse structure of a text (e.g. the sujet). They hypothesize that suspense, curiosity, and surprise are each produced by distinctive arrangements of the discourse structure with respect to the underlying event structure:

While curiosity and surprise structures require that the discourse structure diverge from the event structure, they claim, suspense does not. Suspense occurs, according to this model, any time an "initiating event" (one that has the potential to lead to some significant outcome for a major character) is included in the event structure of a text. The discourse structure is typically, though not necessarily, organized in parallel with the event structure, with the initiating event occurring early in the text, other material intervening between initiating event and outcome, and finally the outcome occurring late in the text to resolve the suspense. Brewer and Lichtenstein (1981) claims that the relevant events must be "significant" in order to generate suspense.

Some other research, by Dolf Zillmann and collaborators, has concentrated less on determining the kinds of narrative structure that engender suspense than on the question of what kinds of emotional states viewers and readers must engage in before they can have the suspense experience. Along these lines, Zillman proposes the empathy theory of suspense, which claims that readers and viewers experience suspense by virtue of feeling empathy for endangered -- or, perhaps, to combine this thesis with the research of Brewer and Lichtenstein (1981), plausibly fortunate -- characters, and by virtue of feeling themselves to be participating as witnesses in the narrative events. The conditions, he suggests, for the kind of empathic distress required for suspense are (a) liking, or at least a failure to dislike, the threatened protagonist (Zillman 1980); and (b) the perception that the harm (or, again, presumably, benefit) is likely but not absolutely certain (Zillman 1991).

This relationship between suspense and readers' assessments of the relative likelihood of a particular outcome is a point of some contention in the field. While Zillman and Brewer believe that suspense is maximal when the perceived likelihood of good outcome is very low, other researchers disagree. Carroll (1984), for example, suggests that any kind of extreme perceived likelihood, whether very high or very low, will heighten suspense; others (cited in Brewer 1996) predict that suspense should be maximal when the results are felt to be perfectly indeterminate. If we were a little better at saying exactly what suspense was or how we could measure it, we might test these different hypotheses properly. But in any case, it seems to me that the intensity of suspense relies more on the reader's investment in the narrative than on the degree of her certainty about the outcome of the story.

Suspense in formulaic, or conventional, circumstances seems to back up my intuition. Many genres include constraints on the assignment of outcomes to good characters, which make the odds quite high that good characters, particularly in formulaic works, will come to a good end. Brewer (1996) argues that this datum indicates that readers suspend their knowledge of this genre-based information in calculating the likelihood of a given outcome in a given reading experience; Zillman (1991) argues instead that this knowledge is invoked in calculations underlying suspense, and that readers do not feel suspense with respect to formulaic narratives. Experimental data on this question are still relatively few, and not, in my view, currently sufficient to establish what level of uncertainty maximizes suspense.

And in fact, as I discussed in part last week, the very calculation of probability at all, though a appealing candidate for being the crucial precursor to suspense, is a less certain element of reading and inferencing than it would appear. Research on the degree to which readers flesh out their mental models of events portrayed in a narrative suggests that it is often far less elaborate as a reader goes through a text than it might appear to a scholar who is considering that text in preparing an analysis, or even to a beginning student who is being given test prompts after completing a story or novel. Meanwhile, other research on the ways people draw conclusions about scenarios in which there is uncertainty (e.g. the work of Daniel Kahnemann and Amos Tversky, or the research on the illusory transparency of intention by Boaz Keysar and his associates, or even work on the way young children do and don't fail various theory of mind tasks) reminds us that whatever people are doing when they are asked -- and in these cases, it is important to remember, they are asked, making the experimental situations markedly different from the circumstances in which the subject is making inferences online, without additional prompting -- it is not acting as a literal-minded calculator of probability. We use rules of thumb, rough heuristics, broad pattern-matching, and other shortcuts, instead. Knowing what those mental shortcuts are can help researchers predict all kinds of things about the kind of interpretive behavior people can and do engage in.

One of the habits of mind identified by Tversky and Kahneman is the one lying behind the so-called "illusion of validity," in which people’s confidence in their judgment depends primarily on the closeness of the match between evidence and prediction, regardless of factors that could limit the predictive strength of that evidence. The sheer fact of consistency or repetition is enough to build our confidence in the likelihood that something of the same sort will happen again. The availability of similar examples in working memory also has a strong effect on the predictions people make. When people can pull an example of some event from memory easily, they perceive it as a more likely outcome.

These "habits of mind" might help to explain what's going on with the way suspense works in formulaic contexts. A narrative in which violent or otherwise adversive events are plentiful, for example, would be likely to make a reader more confident in her prediction of a similar outcome as the conclusion of a suspenseful sequence and, by association, more likely to judge a sequence as suspenseful. Similarly, this tendency supports readers' anticipations in circumstances where a text evokes conventions under which she has experienced suspense before. I wonder, in fact, if there is some way in which this tendency may short-circuit the need for online appraisal of the relationship between events in search of global coherence. The mere association of circumstances may go a long way towards triggering the kind of heightened arousal we associate with suspense.

Enough of this for now; I'll try to follow up with a discussion of the role of simulation and play behavior in the affective responses we have to fictional representations this weekend.

--
Brewer, W. F. (1996). The nature of narrative suspense and the problem of rereading. In P. Vorderer, H. J. Wulff, and M. Friedrichsen (Eds.), Suspense: Conceptualizations, theoretical analyses, and empirical explorations. Mahwah, NJ: Lawrence Erlbaum Associates. 107-127.

Brewer, W.F., and Lichtenstein, E. H. (1981). Event schemas, story schemas, and story grammars. In J. Long and A. Baddeley (Eds.), Attention and Performance IX. Hillsdale, NJ: Lawrence Erlbaum Associates. 363-379.

Kahneman, D., Slovic, P. and Tversky, A. (1982), Judgment Under Uncertainty: Heuristics and Biases, New York: Cambridge University Press.

Zillman, D. (1980). Anatomy of suspense. In P. H. Tannenbaum (Ed.), The Entertainment Functions of Television. Hillsdale, NJ: Lawrence Erlbaum Associates. 133-163.

Zillmann, D. (1991). Empathy: Affect from bearing witness to the emotions of others. In J. Bryant and D. Zillmannn (eds.), Responding to the screen: Reception and reaction processes. 135-167. Hillsdale, NJ: Lawrence Erlbaum Associates.

Zillmann, D. (1996). The psychology of suspense in dramatic exposition. In P. Vorderer, H. J. Wulff, and M. Friedrichsen (Eds.), Suspense: Conceptualizations, theoretical analyses, and empirical explorations. Mahwah, NJ: Lawrence Erlbaum Associates. 199-231.

(literature) | Comments (1)

all about the log | discussion | genre references | language | literature | narrative gaps | planning |