Making Sense of Genre
Deborah Knight

For those who consider genre to be too simplistic or formulaic for academic consideration, there is the question whether anything like serious comprehensional activities are involved in reading or viewing generic texts. Are we just passive consumers of stories that, in an important sense, we already know? Or are we in some respect active participants, constructing meaning from the narrational cues of a genre, so that familiarity with a generic form turns out to be an advantage to us, helping us to sort expediently between salient and non-salient information as it is presented to us by the generic narrative?<1> Is it just irrational for spectators and readers to return, again and again, to consume new installments of familiar generic fictions? This question seems to apply equally to those who prefer the successive consumption of sequences of autonomous but generically similar texts (devotees of action or horror films, for example), to those who prefer to consume serialized texts where each has a more-or-less autonomous structure (P.D. James Adam Dalgleish mysteries, television series which are presented as more-or-less independent story units), to those who prefer unclosed generic texts, like soap opera, where particular storylines develop over long periods, interwoven with many others, and thus seldom coming to anything like closure, and to those (doubtless the majority) who mix and match.

In this paper, I first examine a view defended by, among others, Paul Ricoeur, Northrop Frye, Roland Barthes and Peter Brooks, which suggests that active interpretation is a hallmark feature for understanding any narrative, including generic ones. This view is, broadly, Aristotelian; it is concerned with narrative form and structure, and with the sorts of comprehensional activities required to understand and thus to retell and interpret a story. Next, I turn to N”el Carroll s recent argument<2> which purports to make sense of the repeated consumption of successive generic fictions by showing that consumers of generic fictions are active followers of the particular stories they watch or read. With this argument, Carroll confronts and resolves what he calls the paradox of junk fiction. <3> In the final sections of the paper, I raise questions about Carroll s strategy for making sense of genre. I distinguish between what is involved in making sense of some behaviour and the question whether that behaviour is or is not rational. Further, if it is to be philosophically persuasive, Carroll s account must work for the full range of generic subcategories. I suggest that it does not. Finally, I argue that Carroll s view of the active reader as primarily concerned with making guesses about future occurrences in any particular generic story fails to acknowledge the important role of retrospection that is at the heart of the hermeneu- tical account of narrative comprehension. I suggest that some genres do not depend to any significant extent on retrospection, on synthesizing the elements of an ongoing story, or on discovering the point or purpose of the events that make up the story. If this is correct for some genres, then there is a real question whether some generic stories might be consumed without being understood. If this is plausible, then we unexpectedly wind up face-to-face again with the paradox Carroll had hoped to dissolve.

(1) GENRE: FORM AND FORMULA

Generic fictions are, first and foremost, identified in terms of familiar, codified, conventionalized and formulaic story structures.<4> Plot action is a main focus of generic fictions; the answer to the question, What fictional genre is this? is standardly given by a key term which figures the line of plot action to be found in the particular story, as in mystery or thriller or horror or family melodrama. Generic fictions are also associated with highly conventionalized characters: detectives or gangsters, Westerners or lovers-on-the-run. Genre characters are identified functionally, in terms of their role in a particular story structure, rather than psychologically. Indeed, story is the main focus of genre. As Frank Kermode remarks, generic fictions encourage underreading.<5> They are fictions of easy access, not usually the sort to make many demands on the reader s or viewer s breadth of cultural knowledge. However, generic fictions do make demands on the reader s or viewer s knowledge of other generic fictions, and the sorts of story patterns that are to be associated with the particular generic category of the text in question.

One might argue that, because they encourage a certain kind of underreading one which focuses on developing plot action rather than, say, the beauty of the prose or the cinematography that generic fictions are ideal places for the rehearsal of a certain kind of interpretation. Generic fictions, we might suppose, allow us to practice our skills of narrative comprehension. Let us consider a generally held view that emphasizes the activity of the reader or viewer in her engagement with narratives. I associate this view with, among others, Paul Ricoeur, Northrop Frye, Roland Barthes and Peter Brooks.

In his work on the hermeneutical interpretation of narrative texts, Ricoeur has emphasized the deep interconnection between following and understanding a story.<6> And that surely this seems plausible. If someone insisted that she had followed all the actions of a novel or film but had not understood it, one might want to suggest that perhaps she had not quite followed the text after all. Similarly, if someone said she understood the story but then, in answer to further questions, revealed that she hadn t followed the events of the story, we would be doubtful that she had actually understood it. We might guess that she had had the point of the story explained to her, but that she had not herself managed what Ricoeur talks about when he talks about extracting a configuration from the mere succession of events; one occasionally reads exam answers like this. In both cases, we would be inclined to say that the reader or viewer hadn t quite got the point, and we might be able to show her where she went off-track in following the story, or how her understanding of the story missed its mark if, under consideration, she had not seen how the course of action had unfolded.

Many others have emphasized the conjunction between following and understanding. Northrop Frye speaks of the double articulation of literature, the reciprocal interconnections of plot and theme, between what is happening and what it means for the text as a whole.<7> Peter Brooks<8> makes a similar point with reference to two of the five codes popularized by Roland Barthes in S/Z,<9> drawing attention to the connection between the proairetic the code of actions and events, the one we refer to when we want answers to questions like, What is going on (now)? and the hermeneutical the code of mystery and enigma, the one we refer to when we want answers to questions like, Why is this happening? (particularly if the answer to this question is withheld by the text, where the answer is not immediately apparent). Generic fictions are, admittedly, highly conventional in terms of the story-forms they adopt. If we follow Ricoeur, Frye, Barthes and Brooks<10> in stressing the centrality of plot to this process of following and understanding narratives and in particular, if we accept the idea that, as readers or viewers, we are constantly engaged in an activity of sorting, synthesizing and ordering particular plot elements to extract from them a sense of the story as a whole then we can see that reading for the plot, or looking for it, in the case of cinema, is a skill which we can develop and practice with generic texts as well as with non-generic texts.

So here is one means of countering the idea that consuming generic texts is in some sense irrational. On the formalist line exemplified by Ricoeur, Frye, Barthes and Brooks, readers and viewers engage in the development and refinement of a significant cognitive skill. But perhaps the claim of the irrationality of genre can still be advanced. Take a Wittgensteinian analogy: How much practice do we need with the arithmetical rule for adding two before we know how to go on with adding two?<11> Comparably: how many Westerns or gangster films do we have to watch before we know (can reasonably expect, can predict, can infer) that there will be a confrontation between individual interests and the socio-moral requirements of the community; how many episodes of Inside the Line or Law and Order do we have to watch before we know that within the institutions that ostensibly define and defend justice, there is only so much that can be done about the malefactors who get away with it ; how many weeks or months must we invest in daytime soaps before we know that a central woman character, deciding to act on her own (to protect her family, say, or to investigate a mystery) will be at serious personal risk and need rescue by her lover/husband/father. (Indeed, this conventionalized association of serious risk and the independent actions of seemingly resourceful, competent women characters is hardly restricted to the soaps. In the action film Speed, when Keanu Reeves leaves Sandra Bullock alone in an ambulance after they have finally survived the bomb-rigged bus, any viewer ought to know she is at risk without him.) One might still ask, isn t this repeated consumption of stories which, in an important sense, we already know, just about as irrational as continuing to practice the rule for plus two just in case we haven t gotten it right yet?<12>

To take the formalist line exemplified by Ricoeur, Frye, Barthes and Brooks, we could reply that the apparent repetitiveness of generic fictions testifies to the pleasures viewers derive from rehearsing and refining their interpretational skills through engagement with particular texts that are, as individual examples of generic works, new to them. For all virtual Aristotelians which is to say, for nearly all of us whose basic idea of a story is that it should have a beginning, a middle, and an end there may be a kind of pleasure derived from the repeated engagement with generic plots.<13> The formalist line emphasizes Aristotelian form: the wholeness of the narrative text, the beginning-middle-end structure, the priority of action over character, and the requirement that, to understand any story, we must be able, as Ricoeur reminds us, to derive a configuration from a succession of storied events. The formalist line will, then, stand as a background assumption for the discussion of genre. It is not yet obvious that the successive consumption of generic fictions is any less rational than the successive consumption of any other text-type.

I turn now to discuss a particular strategy for making sense of genre recently proposed by N”el Carroll, one which I take to be an example of what I call the formal-cognitivist gambit. Here, the term formal stands for the formalist line and its Aristotelian background. The term cognitivist marks the preference for making sense of the ongoing consumption of generic fictions in terms of the opportunities these fictions provide for cognitive activity and in terms of the pleasure derived by spectators from those activities: chiefly, the pleasures afforded by the opportunity to guess or infer, often correctly, what is going to happen next in an ongoing course of narrative events, as well as the opportunity to make judgements, including moral judgements, about those actions. While it should be evident that I am sympathetic to Carroll s general project, I will raise some concerns about whether Carroll s strategy for making sense of genre works.

(2) CARROLL AND THE FORMAL-COGNITIVIST GAMBIT

Noell Carroll opts for the formal-cognitivist gambit in his recent consideration of why it might be rational, or at least not irrational, to read or view junk fictions. By junk he means the sorts of best-sellers that line the stalls at airport gift shops as well as things like Harlequin romances; sci-fi, horror, and mystery magazines; comic books; and broadcast narratives on either the radio or TV, as well as commercial movies (225). The general hallmarks of genre, especially the formal and the formulaic, are central to Carroll s account. What Carroll calls junk are, as he puts it, narratives [whose] story dimension is the most important thing about them, narratives which aspire to be page-turners (225). They also generally belong to well-entrenched genres and manifest only a limited repertoire of story-types (226). Carroll s leading question concerns the paradox of junk fiction, to which I have already alluded in my introduction: How could it be rational for us to be interested in consuming stories that we already know (227), given that most consumers of junk fictions are already well acquainted with the formulae of these genres and thus will have a good sense, even before beginning to read, view or hear any particular example of the genre, pretty much how it will go (226-27)?<14>

Carroll's solution to or more correctly, his dissolution of the paradox of junk fiction is to make sense of this behaviour. To make sense of the successive consumption of generic fictions, he introduces the notion of the transactional value that accrues to the reader or viewer. Junk fictions allow for, and even encourage, the practice of interpretation within the constraints of genres: the reader or viewer gets to practice making and testing hypotheses about how a course of events will play out. The solution to the paradox seems at first glance to be of a piece with the account of narrative understanding which we have already considered as the formalist line exemplified by Ricoeur, Frye, Barthes and Brooks. As Carroll remarks, the ultimate union of hero and heroine in a Harlequin Romance or in Sleepless in Seattle may never be in serious doubt; but the question remains how will that union finally be realized, and the junk reader or viewer can remain engaged in, and derive a certain satisfaction from, observing and anticipating the particular developments of each plot (234-5).

Carroll contends that, as consumers, we experience a sense of satisfaction when our inferences and interpretations are correct (235). (He might add that we may experience surprise when our inferences are incorrect, and that this surprise might also be satisfying if it encourages us to attend more closely to the unfolding action.) So the transactional value for the consumer of genre is one that leads to, and ideally supports, the opportunity for self-rewarding cognitive activity (235). Now it is important to note here that Carroll does not restrict the range of self-rewarding activities to cognitive judgements, but also acknowledges the satisfactions derived from moral and emotional judgements (236).<15> The sort of satisfaction in question then is basically a text-directed variety of the day-to-day sorts of practical reasoning which (many believe) form the basis for our decision making, for our predictions about what decisions others will make, and thus for our explanations about what we and others are up to. From Carroll s perspective, the promotion of self-rewarding, readerly activities explains the apparent paradox of junk fiction (238); in fact, insofar as junk fictions are structured to encourage this sort of transactional activity between consumers and generic texts, these self-rewarding activities explain the paradox away.

If we accept his argument, Carroll s dissolution of the paradox of junk fictions explains the successive consumption of generic texts by any given reader or viewer. Even with formulaic fictions, the reader or viewer is active in following the story. The reader or viewer moreover derives satisfaction from being able to make guesses, often correctly, about just what is going to happen next in the ongoing course of action of the particular fiction. I call this a formal-cognitivist gambit, since it accepts the general formalist line about the centrality of the ongoing interpretational activities emphasized by Ricoeur, Frye, Barthes, and Brooks, but adds the twist that foregrounds cognition and the application of practical reasoning skills, principally prediction, to the understanding of character motivation and plot action. Indeed, as we see from several of Carroll s examples, the hypotheses about what will happen next in the plot are often hypotheses about what a character is going to do next, and these hypotheses in turn depend upon being able to explain the character s actions in terms of her goals and reasons for acting, as well as in terms of her circumstances.

(3) PROBLEMS WITH SENSE-MAKING AND QUESTIONS OF RATIONALITY

Carroll s strategy is intended to show that the consumption of generic fictions is rational because he can demonstrate that it makes sense. Showing that something is rational by showing it to be intelligible is not to be differentiated from showing that something is not irrational by showing that, appearances to the contrary, it can be made sense of. I am very fond of this strategy; I think it is central to the whole practice of intentional psychology, of which the dissolu- tion of the paradox of junk fiction is a part. But of course there is an equivocation here with the terms rational and irrational. It is one thing to make sense of suboptimal behaviour, including pathological behaviour, by showing how the agent s beliefs and desires, hopes and fears, memories and goals combine to lead her to do something that, all things considered, is not obviously in her own best interests. We can, in one sense of the terms rational and irrational, show that the behaviour can be explained by reasons and is not, on this construal, irrational. But on a rather stricter construal of rational and irrational, just showing the agent s reasons for acting as she did can still be of a piece with a conclusion that her actions are, after all, irrational.

We can make sense of the most peculiar sorts of action and behaviour by means of attributing a very strange desire to the agent, coupled with otherwise quite mundane beliefs. We can also make sense of behaviour by attributing to the agent a belief that we find very peculiar, but which seems on all the evidence nevertheless to be a belief held true by the agent, coupled with fairly straightforward desires that are consistent with that belief. For instance, we can make sense of the anorexic s behaviour in terms of her desire to be thin and her belief that by eating fewer than 1,000 calories a day she will become so. We can also make sense of her behaviour in terms of her belief that she is overweight, and her desire to be thinner. We can, in short, understand her behaviour we can make sense of it, it is intelligible to us. But sense-making explanations do not make this behaviour rational: the anorexic s unshakeable belief that, at five foot five and 87 pounds she is far too fat, as well as her desire to be thinner, given her weight and height, are not easy to rationalize, though they certainly combine to make (a certain kind of) sense of the pattern of behaviour which has her starving herself to death. So there is the question whether, because he can make sense of the successive consumption of generic fictions, Carroll can show that such behaviour is rational.

One might in fact raise the question whether we need to show that this behaviour is rational in any strict sense. There are lots of repetitive activities one can engage in, where one knows pretty certainly how things will go before one begins them, which have positive transactional values for the individual engaged in them, and which have not yet, to my knowledge, been challenged as to their rationality: gardening, cooking, knitting, taking photographs, phoning or visiting friends and family, having medical check-ups, engaging in sex or exercise, and so forth. The question about rationality and generic fictions is largely tied, I think, to two prejudices, both of which Carroll speaks to: the high art/mass art prejudice, which holds, roughly, that serious attention should properly be paid to serious art and literature, that the very idea of mass art is an oxymoron, and that the only explanation which could make sense of the ongoing consumption of junk is an explanation in terms of suboptimal or irrational behaviour. But this is no explanation, since it presupposes the irrationality of the masses and the behaviours which they pursue. The second prejudice is that the consumption of junk does not demand any sort of reasoning, and the formal-cognitivist reply to this is to show that the consumption of generic fictions demands at least the practice of practical reasoning. Yet hidden in these two prejudices is a fear or anxiety not addressed by the formal-cognitivist gambit: roughly, that the successive consumption of generic fictions is motivated, not by the pleasures of reason and reasoning, but by the pleasures of the emotional responses that one can anticipate and which will genre being genre be gratified. Carroll s formal-cognitivist approach to genre, despite its recognition of the idea of self-rewarding activities, has not, I think, gone far enough to deal with the pleasures of genre that are not the pleasures of reasoning.

Much as I'd like to be persuaded by the argument about transactional value, I m not convinced that what makes the successive consumption of generic fictions intelligible is that any generic text provides a forum for the self-rewarding cognitive activities associated with making inferences and testing hypotheses about what actions are likely to happen next in that particular story. I m unpersuaded that the account is properly transgeneric because I m doubtful about the adequacy of the emphasis Carroll places on formulating hypotheses concerning expectations, concerning what is going to happen next. A refinement from the formalist line exemplified by Ricoeur, Frye, Barthes and Brooks could solve the adequacy problem, but once introduced it might confirm that the repeated consumption of fictions from some generic subcategories just isn t quite rational. This would follow if it turned out that not all consumers of genre follow or understand the particular story in the way required by the formalist line.

(4) IS CARROLL'S FORMAL-COGNITIVIST GAMBIT PROPERLY TRANSGENERIC?

Certainly Carroll s dissolution of the paradox of junk fiction works for some generic fictions, and for some story structures. As Carroll remarks, the paradox disappears when we are thinking of what is called classical detective fiction ; I am inclined to think that it disappears not just for classical detective fiction but also for hard-boiled and other detective fictions, whether those that feature professional investigative-detectives (Sam Spade, V. I. Warshawski ), amateur detectives (remunerated or otherwise, for example Holmes, C. August Dupin, Miss Marple, Kate Fansler ), or simply characters who stumble upon something they take to be a mystery and set about to solve it. Indeed, I think the paradox disappears for mystery stories in general, and might even disappear for stories centered around any mystery or enigma, whether a traditional mystery story or not: some romances and a fair amount of soap opera work on the principal of an enigma or mystery, for instance a secret from the past.

The crossroads between mystery and detection is so clearly tied to the interconnection of what Barthes calls the proairetic and the hermeneutic, the codes of action and enigma, that it is hard to see how any reader or viewer could simply, passively run her eyes over the words or images of such a fiction and not form hypotheses about how things will go, hard to imagine how she could fail to participate in a continual process of constructing a sense of where the story is headed by envisioning or anticipating the range of things that are apt to happen next (235). Indeed, for those who enjoy fictions which centre on resolving a mystery or explaining an enigma, the idea of ongoing cognitive activity seems essential to following the story at all. Frank Kermode notes that detective stories often exhibit what he calls a specialized hermeneutic organization <16> which focuses attention on clues and the possibility of ordering them so as to produce a solution; but this is indeed a specialized organization, one not necessarily employed by other generic subcategories.

A major limitation of Carroll s version of the formal-cognitivist gambit emerges right here in what even he concedes to be a paradigm case of a genre demanding ongoing cognitive activity. It is this: the sort of cognitive activity demanded of those who want to follow and understand a mystery or detective fiction is one that is not merely anticipatory, not merely organized prospectively, and certainly not focussed exclusively on the question What will happen next? There is a fundamentally retrospective, synthetic and synoptic aspect to understanding mystery or detective fictions. Mystery and detective fictions exhibit the close interconnection emphasized by Ricoeur, Frye, Barthes and Brooks between the proairetic and the hermeneutic, between the question What is happening? (and the related question, What has happened? ) and the question Why? . This is because we cannot be certain what the mystery is until we are able to explain just what has happened, and such an explanation depends upon discovering the answer to the question Why? and well as the question Who? . In fictions centred around the mysterious or the enigmatic, our understanding is always subject to revision.

For something to be a mystery or an enigma at all requires that it be incompletely or even incorrectly understood in the beginning often by the detective, and almost always by the audience as well. The mystery can only be solved when we discover what actually happened. Standardly, this will involve the reinterpretation, or at the very least, the redescription of a prior situation or set of events. It will involve seeing that what we had formerly accepted or simply taken for granted ought not to have been taken for granted, that information we thought we had understood we had not understood correctly after all. Understanding a mystery is not simply a prospective or projective undertaking, one that requires us to make inferences about what will happen next. Rather, it involves retrospection, rethinking not only what we think has happened, but also rethinking what counts as salient or relevant to our understanding of what has happened. So in order to follow the plot of a mystery or a detective fiction, the consumer does not merely hypothesize about the range of possible future actions, she has to hypothesize about various possibilities to explain what seems to have happened in the past. Those different possibilities about what the past events and actions were will produce different views about what the course of events actually was, and will have different implications for what we anticipate happening in the future.

It is much harder to see that anything quite so involved is needed to understand fictions from other generic subcategories. Carroll's example of Sleepless in Seattle is a case in point. We can agree that when the heroine finds the boy s knapsack, the viewer tracks the action in terms of the question of whether our heroine and our heroes will meet or pass each other on the elevators (235); but this question is hardly of the same order as the question that focuses the early action of, for example, Manhattan Murder Mystery, the question did Larry and Carol s neighbour murder his wife? There is no Why? tied to the issue of the backpack; whereas Why? has to be answered in the latter case. Larry and Carol learn of the unexpected death of their neighbour s wife. Yet Carol is suspicious. The wife was, after all, in good health. Though there was a joint burial plot, it appears that the wife has been cremated rather than buried. Carol begins to wonder whether it is death by natural causes or murder. If murder, then an explanation has to connect the murderer to the action. The very idea of motive underlines this. For something to be a murder, we have to produce a murderer and a motive: to identify the murderer, we have to be able to tell a story or produce an explanation that shows the reasons why the action was committed. To have any idea what is going on in a mystery or detective fiction, one must always couple the question What? with the question Why? There is no need at all to ask Why? about the main line of action in Sleepless in Seattle. If the heroine does not meet the heroes now, we guess that she will meet them later; it is just a matter of what causal nexus of actions and events will produce the inevitable meeting.

The dependence of narrative understanding on both anticipation and retrospection is evident in the mystery and detective subgenres, as is the connection between the question What is happening (what has happened, what is going to happen)? and Why? On the other hand, action films provide an example of a genre where the question Why? is hardly relevant to whatever inferences and interpretations spectators might make about what will happen next. They also exemplify the comparative insignificance of retrospection on our successful understanding of an unfolding course of events. Whether we are talking about True Lies or any of the Die Hard films or Speed, it would be singularly pointless for a viewer to be concerned with hypothesizing the range of things that are apt to happen on the basis of the particular actions of the particular films. Partly this is because a central concern of action films is suspense, and suspense involves the regulated introduction of the unexpected.<17> So little wonder that in Speed, just when it looks like things are going to be okay, when the bus has made the near-impossible right-hand turn onto the freeway and the characters, along with the audience, experience the euphoric release of tension typical of such moments when success has been torn from the teeth of what ought reasonably to have been disaster we discover what we could not have expected or even guessed: that the freeway hasn t been completed, that there is (yes!) an unfinished overpass which the bus must jump, since it has been rigged to blow up if it slows below 50 mph. Yet even at that moment of pleasurable release of tension, anyone who is at all familiar with the genre ought to be able to say that at any second now something is bound to go wrong.

So if by cognitive activity Carroll means that, once the bus reaches the apparent security of a long stretch of highway, we hypothesize that something bad is going to happen, then there would be general agreement. I take it, however, that Carroll means something much more strictly keyed to the particular film, something much less generic than the guess that something bad is going to happen. He has to mean something more keyed to the particular film, since otherwise we are back at the point of paradox, watching a film where, because of our familiarity with the genre, we know what is going to happen next. So if what is at stake here is some prediction about what is going to happen next in this particular plotline, there is the question just how many of us would have guessed that there would be a great gaping hole in the highway? And even if someone did guess that, aside from a momentary sense of superiority and self-congratulation, does it really matter?

The point here is that in action films, you aren't expected to guess just what is going to happen next. Surely it would detract from the pleasure of action films if, in a regular fashion, you could guess what is going to happen next; it would deprive you of the pleasure of anticipating further calamities, a pleasure keyed to the certainty that there will be more of them, and to the surprise that you can t always foresee what they will be. All you need to know is that something is going to happen; and given the escalatory plot structure of action films, you can also guess (correctly) that you won t have to wait long for whatever it is to happen. So I m not persuaded that correct guesses about just what is going to happen next are what characterize the self-rewarding cognitive activities of genre viewers in general. The generic guess that something bad is about to happen will serve quite well to fuel our anticipation, and is bound to be confirmed. A generic guess will be splendidly self-rewarding, without being keyed to the actual development of the particular generic plot at least in cases like the action film.

Nor do action films depend upon an ongoing process of revising our sense of what has already occurred in order for us to go on to make new guesses about future events and actions in the plot. In Manhattan Murder Mystery, when Carol thinks she has caught a glimpse of her neighbour s presumably dead wife going past on a city bus, it causes her (and us) to rethink whether or not the neighbour s wife is dead leave alone whether or not she has been murdered! Action films like Speed just don t demand any comparable ongoing reconsideration of what we take to have happened.

In the context of the action film, I'd venture that the sorts of guesses that deal with the range of things that could plausibly happen are themselves as generic and formulaic as the film in question. The range of things that are apt to happen are things that are only apt to happen because this is an action film and not a mystery or a screwball comedy. As soon as Carroll gives the nod to the idea that spectators are busy making inferences about the range of possible occurrences, then I think he winds up espousing a view of the interpretation of particular generic texts that squares with the reading within a system approach from which he wishes to distinguish his own view. What counts as the range of possibilities of action is one that is largely determined by the genre rather than within the constraints of the particular text.

A sketch might be helpful. Watching Speed, my guesses went more or less like this: there will be progressively more dangerous and improbable situations; an innocent or two will die along the way, but probably because they have endangered the group; there will be progressively more focus on the romantic-sexual attraction between our hero, Jack (Keanu Reeves) and our heroine, Annie (Sandra Bullock); either the hero s partner or a close friend of his will die at the hands of the villain, played by Dennis Hopper confirming for those who need to be reminded the threat posed by the villain and justifying the eventual confrontation between hero and villain; when the great rescue occurs, everyone will, however improbably, get off the bus with nothing worse than scrapes and bruises; once the bus blows up the police will have to catch the villain; the villain will abduct the heroine (it would be such a waste of Dennis Hopper not to have him abduct the heroine); the hero will confront the villain; the villain will die; but the death of the villain will leave the heroine in a life-threatening situation; things will seem hopeless; having gone through so much to try to rescue Annie, Jack will not abandon her but rather stay with her in the face of what looks like certain death; there will be one final action sequence that will destroy any number of small-scale models of subway cars; Jack and Annie will miraculously live through it all; they will kiss; a crowd will applaud; the closing credits will roll. These seem to me to be guesses keyed to expectations about the genre, rather than expectations about the particular story.

The range of possibilities of action in the Western or in action films, in horror or in family melodrama, are plausible only given the constraints of the genre, and certainly have little to do with what it would be plausible to expect if the represented course of action were unfolding in our real world. Thomas Sobchack makes this point when he remarks that genre characters can do what we would like to be able to do. They can pinpoint the evil in their lives as resident in a monster or a villain, and they can go out and triumph over it. We, on the other hand, are in a muddle. We know things aren t quite right, but we are not sure if it is a conspiracy among corporations, the world situation, politicians ; but whatever it is, we can t call it out of the saloon for a shoot-out or round up the villagers and hunt it down. <18> The beauty of genre is its comparative simplicity: what is plausible or possible is so only within the stripped down, economical, yet well-ordered lines of the formula in question.

Do we watch action films like Speed just to guess what is likely to happen next, or to feel some sort of self-rewarding cognitive gratification when we do? I doubt that, in the main, viewers watch Speed to make cognitive, moral or emotional judgements at all. We don t need to make them, the genre film makes them for us. These are, after all, not fictions of narrative complexity or ambiguity, any more than they are fictions of moral complexity or ambiguity. With generic fictions, by and large we are not in doubt about moral questions. We watch Speed because it delivers progressively escalating action, suspense and romance. We watch for the pleasure of the overt action ; some will, in addition, watch for the pleasure of anticipating the union of the romantic couple. We may also watch for the pleasure of built-in moral judgements: for example, for the opportunity to see the villain perform a sequence of progressively more heinous acts. These are the sorts of expectations that action films reward.

Given the page-turning character of junk fictions and the equivalent in cinematic terms, which Thomas Sobchack describes as the genre film s tendency to feel shorter than it is the actual range of relevant inferential options is pretty clearly limited, and arguably so are the sorts and range of cognitive activities required of the reader or viewer. If, as Carroll claims, the transactional value for readers and viewers of junk fictions derives from the self-rewarding character of their interpretive activity that so often the interpreter can guess correctly about how things will unfold it seems at least worth emphasizing that what counts as a reasonable guess in the context of particular generic fictions is a guess keyed to our expectations about the genre, and the likelihood of those guesses being seriously mistaken is only evidence of the degree to which we are unfamiliar with the conventions of the generic subcategory in question. However, as soon as we admit guesses that are keyed to expectations formed thanks to our familiarity with each subgenre's preordained forms, known plots, recognizable characters, and obvious iconographies ,<19> then we realize that most of our general, generic guesses will be pretty much on the money most of the time. In the context of the action film, our guess that something will happen next is bound to be rewarded. And even only marginally more sophisticated or context-specific guesses about future action will often turn out be right in the long term even if they are wrong in the short term. Jack might not actually kiss Annie until the final scenes of Speed, so I might have to wait a while for my guess that they will kiss to be rewarded. He didn t, for instance, kiss her after she successfully pulled the bus around the near-catastrophic right-hand turn onto the highway. Nor did he kiss her after they escaped from the bus (an exploding 747 interrupted the otherwise romantic moment). In one sense, if I had guessed at each point that they would kiss, I would have been disappointed, or at least not rewarded . But satisfaction is guaranteed with genre; the deferral of the inevitable provides the additional pleasure of prolonged anticipation.

So it seems that the formal-cognitivist gambit, at least as we see it exemplified in Carroll s dissolution of the paradox of junk fiction, might not succeed in showing that the successive consumption of generic fictions is rational in the way Carroll had hoped. It is not even obvious that it is largely supported by the practice of practical reasoning skills. Indeed, Carroll s account, based on the notion of transactional value and self-rewarding cognitive activities seems to bring the discussion of genre back to the point of the original paradox. For once we assume that the sorts of guesses and hypotheses about how generic fictions will proceed are, after all, largely keyed to our expectations about the genre or subgenre, then we find we are, again, dealing with the pleasures derived from what we already know about the narrative form and conventions of generic fiction.

OBSERVATIONS IN CONCLUSION

I have three concluding observations about the formal-cognitivist gambit. The first goes back to the notion of wholeness or formal unity which underlies the formalist account of narrative comprehension which I have sketched with reference to Ricoeur, Frye, Barthes and Brooks. Let us grant that Ricoeur is correct when he suggests that following and understanding a story are absolutely interconnected activities, that you don t get one without the other. This still makes it an open (and indeed an empirical) question whether there might not be many otherwise perfectly content consumers of generic fictions who do not follow the story, who do not extract a configuration from a succession of events, and who thus may well not understand generic fictions. This would be heresy from the formalist point of view, but I have experienced it myself and have heard it reported in classrooms and in other discussions. I think this situation is exacerbated by cinematic and televisual generic narratives, where the speed of delivery is such that some viewers might at no time hold the whole narrative together in any formal configuration at all. Such viewers might have little ongoing sense of how the successive segments of the narrative are part of the narrative as a whole. And this might not seem to be a problem for such viewers, for two reasons: because, in a sense, they already know how things will go, and because the pleasure or transactional value they seek and derive from these fictions is somatic rather than cognitive. They are fine if they can keep each sequence of action together, and if they can follow shifts between sequences, but have no particular need, even comprehensionally, to hold the story-as-a-whole together as they watch it. And because generic fictions, as Carroll rightly remarks, aspire to be page-turners (or their cinematic equivalents), the attention span needed to believe that one is following the story (even if one isn t) is quite limited. The narrative provides what cues are necessary to keep the viewer oriented now; the actual configuring of a whole story is unnecessary, and the absence of such a configuration does not strike the viewer as a loss.

Obviously we can be trained to attend to the whole structure of a narrative this is the sort of training that goes on in classrooms all the time. And it takes a lot of training to get viewers to the point where they can make a serious attempt to hold even a generic film together on one viewing. But it may turn out to be debatable that any actual attentive concern with the story as a whole, and thus with integrating the various narrational components of the story into a whole, is indeed at issue for all consumers of generic fictions. If this is true, then the formal-cognitivist gambit is in real jeopardy, both as a formalist account and as a cognitivist account, of the consumption of generic fictions. It will then only pertain to those consumers who do attend to the formal and narrational cues of the fiction. And we might discover that, at the end of the day, there are fewer virtual Aristotelians than might have been imagined.

A second observation, following from the first, concerns the need to get clearer about what we might broadly call the somatic responses to generic fictions, and the question just how these somatic responses interact with cognitive and moral activities on the part of viewers and readers. Here I am not suggesting that body-based and emotional responses to generic texts (or to anything else, for that matter) are in radical contradiction to intellectual, cognitive, or reason-based responses. Philosophers such as Ronald de Sousa and neurologists such as Antonio Damasio have made persuasive cases for the rationality of emotion and for the somatic basis of practical reasoning.<20> Somatic responses were acknowledged by Aristotle, who draws our attention to the fundamental role played by the emotional responses proper to spectators. There is a somatic aspect, and not merely a cognitive one, to at least two ideas that are central to Carroll s own account of genre: anticipation and gratification. Perhaps in order to make sense of the behaviour of the consumer of generic fictions we must consider the somatic as well as (or perhaps even instead of) the cognitive.

Turning our attention from the cognitive to the somatic, or turning our attention so as to include the somatic, does not entail that the consumption of generic fictions slides from the realm of the rational to the realm of the irrational. Responses to works of high art have a somatic dimension; this does not make our appreciation of these arts irrational, so there is little reason to conclude that a significant somatic dimension to one s response to genre entails irrationality. What wants investigation is whether the self-rewarding transactional value of generic fictions might turn out to be profoundly somatic. If so, the repeated consumption of generic fictions might require at the very least discussion of changes in metabolic and biochemical levels to augment the current focus on the cognitive pleasures of interpretation.

My third concluding observation returns me to the question of the other sorts of cognitive activity that are cited as reasons that could make sense of the consumption of generic fictions. Carroll, as I have mentioned, suggests that moral judgements as well as cognitive judgements are ways of achieving the transactional value that he suggests marks the relationship between consumer and text. Here I must confess real scepticism. I think it is far from obvious that moral judgements are made i.e., deliberated over, reflected upon, contemplated by viewers or readers of generic fictions. Rather, they are made for us by the generic text. As Thomas Sobchack has shown persuasively, for anything to be generic, it has to have at least two moral dimensions that are part of the form and formula of genre. The formal component has to do with the moral structure of the plot of any generic work. For something to be genre, it must achieve a particular sort of aesthetic and moral order. The formulaic component concerns the deployment of characters with immediately identifiable moral qualities, where this is supported by the iconographic aspects concerning character and setting. In genre, moral issues are easy to schematize. Carroll might well argue that in genre the viewer can test hypotheses about the moral worth of an action or a character. But genre just wouldn t be genre if this were the general case. Rather, characters and situations are their moral significance, their moral value: if we recognize this, then the fiction will ceaselessly confirm it for us.

In conclusion, the formal-cognitivist gambit might provide an explanation of the ongoing consumption of generic fictions by those already familiar with the various generic subcategories, but only under the assumption that the reader or viewer in question already thinks like a virtual Aristotelian, and does indeed (attempt to) understand any particular fiction as a whole. The formal-cognitivist gambit does not seem to explain the consumption of generic fictions by those who do not work toward the sort of holistic understanding that is the basis of the formalist part of the gambit. As for the cognitivist part of the gambit, I have raised three issues that suggest Carroll hasn t succeeded in dissolving the paradox of junk fiction after all. First, it isn t clear that making sense of this behaviour goes any particular distance toward making it rational. Second, it isn t clear that the cognitivist account is persuasively transgeneric: it works very well with some genres, yet seems unnecessary for the explanation of others. Third, it isn t clear that the attention Carroll gives to the prospective, forward-looking engagement with particular developments in an ongoing course of events is adequate as an account of what is involved in following and understanding a story. In extreme cases, it may well not be necessary for readers whose attention is keyed primarily to what is happening now. And if I am right in my suspicion that the best guesses about what sorts of actions one should expect are properly guesses keyed to the genre rather than to the particular text, then we find ourselves right back at the original paradox. What I think is needed to dissolve the paradox of junk fiction is an account that explicates the relation- ship between making sense of fictions, including generic ones, and making sense of agents behaviour. Carroll s account needs to say more about the sorts of practical reasoning skills involved in these two sense-making activities, and in particular, more about the extent to which practical reasoning makes a difference to the understanding of genre.<21>

Notes

<1> In the case of cinema, the argument in favour of the connection between narrational cues and comprehension an argument that certainly extends to generic films can be found in David Bordwell, Narration in the Fiction Film. Madison, Wisconsin: University of Wisconsin Press, 1985; see also Edward Branigan s Narrative Comprehension and Film. London: Routledge, 1992.

<2> No‰l Carroll, The Paradox of Junk Fiction, Philosophy and Literature Volume 18: 225-241.

<3> The concern to show that the consumption of literature has a rational basis is apparent also in, for example, Paisley Livingston, Literature and Rationality. Cambridge: Cambridge University Press, 1991.

<4> In Genre Film: A Classical Experience, Film Genre Reader, ed. Barry Keith Grant (Austin: University of Texas Press, 1986), pp. 102-113, Thomas Sobchack has argued that we think about the fictional genre film as a single category that includes all that is commonly held to be genre film (102). Sobchack s line is formal and, broadly speaking, Aristotelian. He suggests that despite the obvious differences between the various generic subcategories (between thrillers and screwball comedies, for example), the fundamentals of story structure, characterization, theme and iconography, as well as the fundamental mimetic idiosyncracy of such fictions, are common to generic films as such. Attention to these formal or classical aspects of genre will give us a better understanding of how genres work, and how spectators relate to them.

<5> Frank Kermode, Secrets and Narrative Sequence, in On Narrative, ed. W. J. T. Mitchell. Chicago: University of Chicago Press, 1981, pp. 79-97.

<6> See the three volumes of Time and Narrative, trans. David Pellauer and Kathleen Blamey. Chicago: University of Chicago Press, 1984, 1985, 1988; Life in Quest of Narrative, On Paul Ricoeur, ed. David Wood. London: Routledge, 1991, pp. 20-33; The Narrative Function, Hermeneutics and the Human Sciences: Essays on Language, Action and Interpretation, ed. John B. Thompson. Cambridge: Cambridge University Press, 1981, pp. 274-297; Oneself as Another, trans. Kathleen Blamey. Chicago: University of Chicago Press, 1992.

<7> Northrop Frye, The Anatomy of Criticism: Four Essays. Princeton: Princeton University Press, 1957.

<8> Peter Brooks, Reading for the Plot, in Reading for the Plot: Design and Intention in Narrative. Cambridge, MA: Harvard University Press, 1992, p. 18. Others have drawn attention to this doubled feature of narrative; Brooks cites, for example, Jonathan Culler s notion of the double logic of narrative, p. 28. Both writers are mentioned in relation to this notion in Gerald Prince s A Dictionary of Narratology. Lincoln: University of Nebraska Press, 1987.

<9> Roland Barthes, S/Z, trans. Richard Martin. New York: Hill and Wang, 1974. The other three codes concern the semes, the symbolic, and the cul- tural.

<10> It may strike some as just mistaken to see Roland Barthes as a team player in the ongoing game of textual hermeneutics. I certainly see where this objection comes from. But however unorthodox, however seemingly anti-holistic Barthes interpretive work might be, it is nevertheless strongly thematic in motivation, and is drawn upon by many more holistically motivated theorists and philosophers. His work is a rule-proving exception, an exception that tests the rule.

<11> This example comes from 185 of Ludwig Wittgenstein s Philosophical Investigations. Oxford: Basil Blackwell, 1958.

<12> Carroll discusses the idea of practice in The Paradox of Junk Fiction, and notes that many skills are profitably and enjoyably practiced without repetitive practice being seen to be irrational.

<13> The phrase is Brooks .

<14> This question is to be distinguished, Carroll insists, from other questions about genre. Carroll is not pursuing the question of how to make sense of why some consumers of genre take such delight from particular subcategories of genre. While he acknowledges the idea of genre as a system of systems, but the paradox that is of interest to him here is not one that is appropriately answered in terms of anything like reading in a system (231). Carroll also distinguishes his question from the question why it is rational to return to any particular text (generic or otherwise) which he calls the paradox of recidivism (227). I believe Carroll would ultimately deny that recidivism is a paradox, and would thus deny that it is irrational to read or watch fictions again. Still, recidivism is not unconnected to the issue of genre (as Carroll notes); it would certainly be worth investigating for the light it might shed on the selection and function of classical or canonical texts, and the role they play in any sys- tem-of-systems account of generic fictions.

<15> I will return to the question of moral judgements, but I confess that I am uncertain just what is involved in emotional judgements. Perhaps Carroll is speaking of emotional responses such as fear, hope, excitement though I doubt he uses the term judgement so loosely. Perhaps he is speaking of judgements that are in part motivated by emotional response: if so, I think this needs more spelling out.

<16> Kermode, Secrets and Narrative Sequence, p. 83.

<17> This is a point Carroll has already developed in his article Toward a Theory of Film Suspense, Persistence of Vision no. 1 (Summer 1984): 65-89, where he proposes that suspense develops in relation to two possible, logically opposed outcomes to a set of events, such that one outcome is morally correct but unlikely, while the other is evil and likely.

<18> Sobchack, Genre Film: A Classical Experience, p. 108

<19> Sobchack, Genre Film: A Classical Experience, p. 105.

<20> Ronald de Sousa, The Rationality of Emotions. Cambridge, MA: The MIT Press, 1991; Anthonio Damasio, Descartes Error: Emotion, Reason, and the Human Brain. New York: G. P. Putnam s Sons, 1994.

<21> I am grateful to the Social Sciences and Humanities Research Council of Canada for their ongoing support of my research, and to Queen s University for an ARC grant that supported a research project on genre, gender and narrative. Profound thanks to George McKnight (Film Studies, Carleton University), with whom I have collaborated on the topic of the narrative conventions of mystery-detective fiction, for his comments and suggestions on this paper.

from www.hanover.edu/philos/film/vol_02/knight.htm