Literary Discourse: Aspects of Cognitive and Social Psychological Approaches [Reprint 2019 ed.] 9783110864236, 9783110106855

Citation preview

Literary Discourse

Research in Text Theory Untersuchungen zur Texttheorie Editor Jänos S. Petöfi, Bielefeld Advisory Board Irena Bellen, Montreal Antonio Garcia Berrio, Madrid Maria-Elisabeth Conte, Pavia Teun A. van Dijk, Amsterdam Wolfgang U. Dressler, Wien Nils Erik Enkvist, Abo Peter Hartmann f , Konstanz Robert E. Longacre, Dallas Roland Posner, Berlin Hannes Rieser, Bielefeld Dieter Viehweger, Berlin/DDR Volume 11

w DE

G Walter de Gruyter • Berlin • New York 1987

Literary Discourse Aspects of Cognitive and Social Psychological Approaches Edited by Làszló Halàsz

w DE

G Walter de Gruyter • Berlin • New York 1987

Printed on acid free paper (ageing-resistant - p H 7, neutral) Library of Congress Cataloging in Publication Data Literary discourse. (Research in text theory = Untersuchungen zur Texttheorie; v. 11) Includes index. 1. Discourse analysis, Literary. 2. Literature—Psychology. I. Haläsz, Läszlo, 1933 . II. Series: Research in text theory; v. 11. P302.5.L57 1987 401.41 86-32754 ISBN 0-89925-325-3 (U.S.)

CIP-Kurztitelaufnahme

der Deutschen Bibliothek

Literary discourse : aspects of cognitive and social psycholog. approaches / ed. by Laszló Halàsz. - Berlin ; New York : de Gruyter, 1987. (Research in text theory ; Vol. 11) ISBN 3-11-010685-X NE: Halàsz Laszló [Hrsg.]; GT

ISSN 0179-4167 © 1987 by Walter de Gruyter & Co., Berlin 30 Printed in Germany Alle Rechte des Nachdrucks, der photomechanischen Wiedergabe, der Herstellung von Photokopien - auch auszugsweise - vorbehalten. Satz: Utesch Satztechnik GmbH, Hamburg; Druck: Hildebrand, Berlin Bindearbeiten : Lüderitz & Bauer, Berlin

Contents Laszlo Halasz Cognitive and Social Psychological Approaches to Literary Discourse. An Overview

1

Robert P. Abelson Artificial Intelligence and Literary Appreciation: How Big Is the Gap? . .

38

Robert de Beaugrande Schemas for Literary Communication

49

Csaba Pleh On Formal- and Content-Based Models of Story Memory

100

Janos Laszlo Understanding and Enjoying

113

Martin S. Lindauer The Short Story: its Place in the Psychology of Literature

125

Laszlo Halasz Social Perception and Understanding of Interaction in the Short Stories Entitled 'Everything that rises must converge' and 'Brutes' (Barbarians) .

140

Colin Martindale Narrative Pattern Analysis: a Quantitative Method for Inferring the Symbolic Meaning of Narratives

167

Ravenna Helson A Personality Assessment Approach to the Study of Literature

182

Dean K. Simonton Shakespeare's Plays and Sonnets: Correlates of Differential Greatness . .

193

Antal Bokay Psychoanalytical Text Theory and Literary Interpretation

201

Larry Gross Life vs. Art: the Interpretation of Visual Narratives

215

List of Contributors

232

Author Index

233

Subject Index

239

LÀSZLÓ HALÀSZ

Cognitive and Social Psychological Approaches to Literary Discourse An Overview As used by structuralists, literary discourse means, in opposition to the story which comprises the actions, happenings, characters, setting, that part of the narrative which consists of the expressions, of the way how the story is communicated. The same distinction can be virtually found in the fabula-sujet pair of the Russian formalists (see Chatman, 1978). The adjective and the noun 'literary discourse' in the title of our volume, however, is to be interpreted in the original meaning of 'discourse', in a wider sense than just defined above. Discourse means speech, talk, conversation, lecture, sermon, treatise. So 'discourse' is to say things, speak to give information, discuss something, talk or address given in public, talk to an audience for the purpose of teaching, address on moral subject, and so on. That is, by using the words 'literary discourse', we simply wanted to indicate that we consider literature as a peculiar kind of discourse. When we approach it from a cognitive and social psychological point of view, our attention is focussed on how this discourse as a dynamic and communicative event takes place. Though we are mainly concerned with how this sort of discourse produces an effect and how the reader experiences it; moreover how the reader perceives, remembers, comprehends the literary text or texts as literary discourse, the cognitive and social psychological study of the processing actions cannot be made without having a closer look at certain characteristics of the text itself.

Literary and psychological cognition Once an American social psychologist laboured under the conviction that his students could not really recognize social psychological implications in their own lives, i. e. outside the walls of the lecture-hall. He looked for and found a way how to overbridge the gap between science and life: he compiled a Reader of literary texts (Fernandez, 1977). 1 1

Similar selections have been published in the field of developmental psychology (Haimowitz and Haimowitz, 1966; Kiell, 1974) and that of psychopathology (Rabkin, 1966; Stone and Stone, 1966).

2

Lâszlô Halâsz H e illustrated basic social p s y c h o l o g i c a l p r o b l e m s w i t h e x c e r p t s f r o m the

w o r k s o f o u t s t a n d i n g a u t h o r s . N a m e l y : m o t i v a t i o n in social b e h a v i o u r ; a c q u i r ing m o t i v e s and a t t i t u d e s ; identity and a n x i e t y ; social roles and social n o r m s ; differences in class, caste, and religion; r e f e r e n c e g r o u p s ; the g r o u p as a s y s t e m ; c o m p e t i t i o n and p o w e r ; leadership; g r o u p c o n f l i c t ; social c h a n g e , social d e viance. F e r n a n d e z , as a social p s y c h o l o g i s t , and the o v e r w h e l m i n g m a j o r i t y o f r e a d ers, t o o k f o r g r a n t e d t h a t literature c a n bring i n t o life w i t h m e r e w o r d s all the f o r m s o f c o m m u n i c a t i o n , a p p r o a c h and d e p a r t u r e , physical c o n t a c t , t o n e o f v o i c e , r a t e o f s p e e c h , f r o w i n g , flutter o f an eyelid. It can p r e s e n t all t h e strata o f h u m a n relations t h r o u g h the single n a r r o w e d channel o f verbal c o m m u n i c a t i o n . It is this that the great talents o f w o r l d literature m a k e use of, t h o s e w h o c o u l d p r o d u c e in their u n i q u e l y rich d e p i c t i o n o f h u m a n b e h a v i o r p s y c h o l o g i c a l a c h i e v e m e n t s w h i c h science, w i t h its o w n m e a n s , c o u l d d i s c o v e r in m o s t cases o n l y m u c h later. 2 2

There is also such a selection which shows a sort of reverse situation. That is, it reviews literary works which in their own way also illustrate how psychology repays its debt to literature. Katz, Warrick, and Greenberg (1974) present several short stories in the subjects of psychobiology, learning process, sensation and perception, social processes, developmental processes, personality, abnormal processes and therapy. What renders them so suitable for introducing the reader to psychology is that they might not even have been written without this or that discovery of psychology. In this connection we must lay special emphasis on the relation between psychoanalysis and literature. If we made a selection entitled Literature through psychoanalysis not only one but even ten volumes would not suffice. Before the appearance of psychoanalysis the psychological insight of many a writer led to the description of remarkable psychoanalytical situations. In it the psychoanalysts saw the sign of interweaving between their theory and literature and, at the same time, an imperative challenge to the psychoanalytical study of art. The example in this regard too was set by Freud. Of several of his works of this vein the bulkiest one analyzes Jensen's long short story entitled Gradiva (Freud, 1973 (1907)). The work - as is proved by the correspondence between Freud and the author - sprang up independent of the just budding psychoanalysis and offered Freud an excellent opportunity to demonstrate his basic idea, according to which the author's own complexes appear in his works, consequently the presentation of the relation between the turning-points in the life of the author and the contentual characteristics of the work offers valuable contributions to the understanding of the literary work. Simultaneously it demonstrates, or even it can be said, teaches psychoanalysis. Gradiva is built on the dreams, wishful thinking of the protagonist, and Freud, similarly to a detective, unrevels happily the offered themes about dream-life and unconscious processes. After the completion of the analysis it came to light from the correspondence between Freud and the writer, and through the services of Jung, the then most intimate collaborator of Freud, that, transferred and condensed, the decisive young-age experiences of the writer appear in the protagonist's dreams and desires. Freud's analysis is followed by a great flow of similar psychoanalytical interpretations of novels, short stories, and dramas. In the writing of these studies an important role was played by the sweep of free associations, the description, differing from that of the classics, of space and time befitting psychological events, the Freudian view of childhood, psychoanalytical dream symbols, sexual repressions, death instinct, various pathopsychological motives, all of them as organic parts of the contemporary literature. Here it is not our task to eulogize what new possibilities psychoanalysis opened up and how fruitful its import is. Let is suffice to say that we

Cognitive and Social Psychological Approaches

3

Naturally psychological cognition is only one, and as such not an inevitable, concomitant of literature. Its perfection to life cannot be, of course, narrowed down to any artistic representation of characters, states of mind. It is also true that character-portrayal, psychography is not realized through coming across scientific discoveries, new scientific truths. Even if some authors do come near such truths, which is not a precondition but a new proof of their greatness, it should be treated as a sort of artistic by-product. At any rate, such by-products also illustrate clearly what each and every literary work has in common: the artistic form cannot be separated from the thoughts and sentiments present in its very medium. Consequently, we cannot evade getting permeated with the happenings of the work, in the same way as they are reflected in the mirror of the mind of each character (Vygostsky, 1965). Radically opposed to this is the concept most markedly professed by Eichenbaum (1924) of the Russian formalists. It says that only "boring text-books and papers" seek the explanation of the protagonists' characters and acts in psychological laws. The deeds of the protagonist are determined by the characteristics of the artistic form. Beneath Hamlet's hesitation are solely hidden the laws of the form of artistic construction. It is they and not the psychography of the protagonists that should be tackled and grasped, since artistic creation, by its nature, is supra-psychological. Formalists, and partly structuralists as well, reason by saying that characters are products of plots. They are participants or actants, and it is a mistake to take them for real beings. They stand apart from any psychological and moral measures. Chatman (1978: 113), however, points out that other structuralists advocate "a more open, afunctional notion of character." For example, Todorov differentiates plot-centered or apsychological and character-centered or psychological narratives. Actions in a psychological narrative are the manifestations or symptoms of personality. would maim ad absurdum the literature of the 20th century if we eliminated from it everything which may be "suspected" of this impact. (Because of the extraordinary richness of the related literature here we make mention of only two works: Keill, 1963; Grimaud, 1976). As much, however, we must remark that the appearance in literature of the motives in question is not an unconditional security of an artistically more authentic human portrayal, as it is also true that the psychoanalysis of works does not necessarily enrich their understanding either. As in the case of many other far-reaching theories, in that of psychoanalysis, too, it is proved: when a problem solution model is so much widened that it becomes omniscient and especially predestined to explain away everything, it is on the best way of rendering its possibilities ineffectual. The results of psychoanalysis applied like that in the details between the text and its writer may fall short of the targets. Great is the lure that the cognition of the given (life-) work should become the self-justification of the psychoanalytical tenets, the illustration of the Oedipus complex (cf. Weber, 1958). There loom up great dangers of one-sidedness if we see nothing else but a perambulating Oedipus complex in the personality of the characters. It, however, shows the weight of the psychoanalytical approach to literature that even those have difficulties in getting rid of it, who, in accordance with their criticism, reject fundamental psychoanalytical theses, wanting to exchange them for others, who emphasize the insufficiency of psychoanalysis.

4

László Halász

Barthes - Chatman (1978: 116) goes an - originally also professed a narrow functionalism, the secondariness of character. However, when analyzing Balzac's Sarassine, he no longer reasons by saying that character is subservient to action. And besides emphasizing the justification of terms like trait and personality, he also states that "reading narratives is nothing less than a 'process of nomination' and that one element to be named is the trait". Chatman is obviously right in saying that modern characters have not only numerous traits, but, in their exceptional roles, their heterogenity, the fact that they cannot be reduced to a single aspect, is also essential. We can add that modern taste itself is complex as well. Thus the contemplation of a character in a not modern literary work gives pleasure even to the reader of today because in him, too, are several, and not infrequently contradictory, features.

The reader's biases; perception of the literary character All this will become clearer even through roughly sketching the psychological processes evoked by literary discourse. When selecting, we match (collate) and make a pre-estimate on the basis of our past experience. We weigh and measure whether the work comes up to our expectations concerning the given literary form. So to say, the reader starts processing information, often subliminally, before laying a hand on the text. He sifts them, arrives at judgements, draws conclusions, while building up expectations. His biases, positive and negative feelings are awaken. He carries this on when reading the title. 3 The information network activated by the writer's name is also to be reckoned with. It is enough to know only by hearsay a few characteristics of the time, space, aspect, and atmosphere of the writer's world to involuntarily start shaping preliminary impressions.4 3

4

Pratt (1977) points out how certain conventions help with the interpretation of titles. For example, there are titles which directly characterize the content (An American Tragedy). Novels with twofold titles are the most revealing of topic (Pamela, or Virtue Rewarded). If the title is the name of a person, the reader expects him to be in the focus as protagonist. Even, there is more to it than that. The name of the literary protagonist has a special effect of its own. As Lindauer (1976) says, independent of their literary setting names connote various other meanings. One of the sources of this is the popularity of the name (names of historical figures or those of pop stars). Another source is the name's inherent physiognomic structure. "The letters of a name as shapes (and sounds) can arouse positive or negative effect (Werner, 1955)." Bodman et. al. (1976) exposed in isolation the names of unknown literary characters for the subjects. Following the positive or negative ratings of the subjects, they were compared with the connotations of the protagonist in the text. They found a "relatively weak but still discernable match" between them. That is, the reader "digs" the names themselves. Sherif's study (1935) shows the author's prestige effect salient. The experiments of Michael, Rosenthal and De Camp (1949), Das, Rath and Das (1955), Frances (1963), Aronson, Turner and Carlsmith (1963) provide more subtle and contradictory information (see further Child, 1 9 6 8 1969, Crozier and Chapman, 1981). However, the findings of Asch (1952) are likely to have been the

Cognitive and Social Psychological Approaches

5

Anticipating what the author and the title suggest together, the reader is extraordinarily open to receive information in harmony with his expectations. A glance at the opening lines is a further significant step towards making the contours more contrastive. We endeavour to find out as early as possible where, when, and in whose company we are. We make clear the approximate age(s), sex, and some other particulars of the character(s). As we produce in the text, our relation becomes more and more personal. This way of person perception lies farthest from the one we use when judging the members of our family, our colleagues, and friends. Regular contact, the flow of reciprocal interactions give us a great many pieces of information, each of which to the very last is a part of everyday reality. And, of human beings in the flesh and blood we form impressions even if we have little information and our relation is superficial and limited. Even then we have some data as regards the appearance, expressive movements, and a few verbal utterances of the other party. Sometimes the probable reactions of the person in question are so important that we have to assess, even in possession of limited information, who he is, better to say, who he may be. We may have seen our man a couple of times, or not, and we may have heard his voice only. All these possibilities have been most thoroughly examined (see summary review Secord-Backman, 1972). In the typical instance of person perception we perceive the other party for the first time. Our imagination, thinking is put into motion by our visual, auditive, most decisive up to now. They arenotbasedon experiments with literary texts, yet they are valid also to the reception of them. Accordingly, when in the experiment not the real but another author belongs to the text, it transforms the cognitive content of the text, pushing it against the background in which the latter worked. Different is the meaning of the same text in different context. In our own experiment (Halisz, 1980) we did not resort to misleading the subjects, to exchanging authors, giving false or invented authors. Half of the subjects were given the real name of the author; the other half were given no name at all, and they all were to judge the artistic value of a great number of poems. When the author's name was given the subjects distinguished them better and more sharply. Whatever they accepted or refused they did so more unambiguously than in lack of the names. It is obvious that the name of the author helps the recipient to bring his evaluation closer to his declared value system. Comparing the results with those of the experts who always judged the works in no-name condition showed that the subjects who got the names judged the works more similarly to them than the subjects who were not given the names did. This is only partly explained by the fact that the experts recognized some of the poems or their authors. It is more decisive that, owing to their competence, they were able to recall several adequate factors of the structure and the special-time co-ordinates of the work on the basis of the work itself. But according to the high-standard deviations in some cases this was not unambiguous even for them. The network associated to the name of the author of the literary work, as to a cue, shows the place which the text occupies in the formations furnishing the individual with pro or con reference data: in text books, critiques, with teachers and lectures, in papers, local opinion leaders, the circle of friends, and in that of the family. The evaluation of works can be imagined as the formation of a peculiar tension system on the basis of former information about the author and on that of the information gained about his given work with the flow of the two shaping each other. Upon hearing or seeing the name of the author, a prestige hierarchy, an order of value is set into operation which is part of the reference-information network that helps with the perception of the literary work through the mobilization of an arsenal of real factual knowledge. At the same time we can say that obeying the delusive effects of prestige (Latin prestigiae =

6

Laszlo Halasz

and other impressions. It can also happen that not we but a relative of ours knows the given person, of whom we are to form a picture on the basis of a character sketch which can consist of the enumeration of traits or only of a few connected sentences. As opposed to this sketchy information, in literary discourse we have plenty of data as to the physical appearance, expressive movements, words, and acts of the given person, all of them coming from somebody (i. e. from the writer) who "knew" him, yet the writer is not directly instrumental in making us reach him. It is the protagonist that feels, thinks, says something, gets into interactions with others, and not rarely he is spoken of. Unlike real people, the literary protagonist does not react to us. Thus we cannot be influenced in one way or another whether he likes us, finds us attractive, or not, and what he thinks of us at all. He cannot be expected to return our compliments. True, it cannot happen either that he becomes estranged from us, contrary to our expectations. In short, the literary protagonist is much more at the mercy of the reader, who can treat him more arbitrarily than fellow-beings in everyday social contacts, which are always accompanied by practical consequences, even though momentarily stress is placed on getting acquainted with and/or seeing through the other. In the perception of the literary figure this is possible only in a round-about and symbolic way. But the reader can be more tolerant and flexible with him, not rushing judgements but waiting freely for events to develop. Deprived of the role of controlling actions, the reader can only maneuver with the protagonist notably in the interest of comprehending the meaning of the work. delusion) means more than mere "worthless enthusiasm of worthless superiority" (Thackeray), or voracious imitation of the norms of a reference group. As readers, all of us somehow use, or at least have used snobbery as a pattern of evaluation. The prestige of highly valued authors might help us to select, i. e. to make the valuable work - so far appreciated without due internalization and without becoming part of the personal way of life - a future source of genuine literary enjoyment and the basis of a judgment recognizing the qualities characteristic of the work. Manifestations of literary (and other sort of) snobbery often reflects the snags in the way leading to sophisticated appreciation. According to this view, it is hardly comprehensible why snobbery has such a bad press, and why it appears as a conglomeration of particularly harmful, repugnant features, almost sins, as a sign of mendacious hypocrisy, total lack of individuality, opportunist backwardness or silly superficial imitation, sensation-seeking, idle semi-education, impoverished affections, bowing and scraping to authority. An evaluation of snobbery as a process, and its judgment with more subtlety, less affection and more self-criticism is rarely found. A strong defensive mechanism underlying this must be assumed. In the wake of Freud the type of tension-reducing projection, when one assigns unconsciously to others or to the external world his own characteristics undesirable even to himself, has become a well-known phenomenon. The more stringent, merciless and heated (hence the hot temper) the verdict, the more successful this relieving exorcism, since refusing something with such vehemence proves strikingly (mainly for ourselves) that it cannot be characteristic of us. Becoming aware of our suggestibility by prestige is one of the most serious losses of prestige. It is particularly hard to face the effects of prestige and context when a value which we believe to be so important and especially our own (De gustibus non est disputandum) turns out to be controlled externally, i. e. our ability to judge literary works, the ability which compromises our intellectual strength and taste.

Cognitive and Social Psychological Approaches

7

Empathy and identification, projection and attribution during reception As is pointed out by Kreitler and Kreitler (1972), literature evokes empathy in a way different from other forms of linguistic communication, due to the contents, the plot, the described events, situations, or processes, but mainly because of character-portrayal. As a result of the inner dynamics and sentiments of the protagonists, the cues of empathy are similar to those in real life, though weaker, for movements arrested verbally are not so forcible as those registered visually. And a verbal account describes, broken up and at length, expressive movements, comprehensive action patterns taking place simultaneously and momentarily. Notwithstanding, verbal formulae accompanied by definite gestures, as well as the suggestion of the non-verbal action, and the rich depiction of movements with words loaded with emotion and sentences full of rhythm, arouse kinesthetic imitation, which serves as the background of the reader's arousal. By means of scenes and events described in the work, the reader interprets his own physiological state on the basis of his former experience. The recipient transgresses the given information, that is, he elaborates and expands, in a disconnected way, the depicted facts. At the same time he identifies himself with the author, which helps uphold the continual experience. Since the author cannot present all the details, he endeavours to make the reader fill in the gaps, do gestalt-formation. The activation of the closure tendency is probable as it befits how we act under everyday circumstances. Under such circumstances, however, we can hardly afford to make imaginary reconstructions about others. The reader's reconstructions increase empathic answers. Imaginary representations formed in the intervals of, or following reading can be the nuclei of a stronger identification with the literary character. Introjection, as well as projection, can have a part in identification which is a frequent, even if not necessary, outcome of empathy. Introjection is a means of incorporating environmental effects and characteristics, especially the traits of other persons who are the subjects of identification. Projection diverts from identification, if one projects on the other person such motifs that really exist only in one. But it can be a part of the identification if the real similarity is chanced upon in a direction that leads from the subject towards the object. There is no doubt that it is easier for the reader to identify himself with literary characters whose self-concepts and personality traits are similar to his own (Kingston and White, 1967). It should not be forgotten, however, that the systematization of Holmes (1968) directs attention to the fact that we can project traits and feelings onto others in total contradiction to our own. This can be fraught with arbitrary interpretations, distortions. 5 It is less

5

Even the attribution processes are not free of these sources of mistake. Attribution serves the cognition of the relation between (distant) things and their direct effects, between the source and

8

Laszlo Halasz

o b v i o u s , that the inclination itself t o a s t r o n g i m a g i n a r y identification can likewise b e the s o u r c e o f s t r o n g defense reactions. A s is stressed b y Stotland (1978),

the relation b e t w e e n i m a g i n a r y identification,

e v o k i n g similar life

e x p e r i e n c e and e m p a t h i c r e s p o n s e is n o t so u n a m b i g u o u s and d i r e c t as he s u p p o s e d in his f o r m e r w o r k s . O f this an o b v i o u s lesson c a n b e d r a w n . A s o u r c e o f stimulus challenging i m a g i n a t i o n w i t h the f o r c e o f literary d i s c o u r s e can be t h e s o u r c e o f varied defense r e a c t i o n s as well. 6 Since e m p a t h i c p r o c e s s e s - stress K r e i t l e r and K r e i t l e r - c o n t a i n elements o f satisfying f o r b i d d e n desires, t h e y c a n a r o u s e anguish, d i s t u r b a n c e . T h e recipient is inclined t o p r o j e c t t h e m o n t o the c h a r a c t e r s in o r d e r t o p a r t l y shift the b u r d e n o f guilty feeling and responsibility. It is also a lure f o r p r o j e c t i o n that the recipient has a certain f r e e d o m w h e n defining the t o p i c and o r g a n i z i n g the material. A n d n e v e r all the m o t i v e s f o r the actions o f the p r o t a g o n i s t

6

are

the event, one factor coming from or being attributed to the other (Heider, 1958). Thus it renders possible for one to understand the world, to control events influencing oneself or others, and to predict (foresee) those to come later. That is why attribution shows up invariances, the lasting dispositional - features of objects, events, and people; those that are characteristic of them. Besides his mentioned work, Heider also says in a later interview (Heider, 1976) that he analyzed many tales, short stories, and novels when elaborating the basic idea of his work, as they enshrine "naive psychology" which so much deserves the attention of the expert. For example, in Aesop's The Fox and the Grapes the fox pretends as if he did not want the grapes, perhaps he is even convinced of it, and he does not act as one who cannot get them. He attributes frustration to motives such as "I don't want to", "I have no intention to", "I won't try to" instead of "not able to", since the latter would undermine his self-esteem. Behold the egocentric atrribution. - In Racine's Andromache, Hermione says that her revenge will be in vain if her victim does not get to know at the moment of his death that his destruction comes from her hand. In other words, the joint attribution of the source and the intention is of paramount significance in the course of retribution. Revenge is complete if the target person knows to whom he is to attribute the harm he suffers in return. So the recipient comes across in literary discourse such attributions with which he might have something to do in everyday practicer as a naive observer or actor. Yet here again he must reckon with the defense mechanism, with self-justification. When weighing the attributional bases of person perception we must take into consideration, as opposed to the ideas of Jones and Nisbett (1971), the cognitive and motivational processes together (Stephan, 1975; Taylor and Koivumaki, 1976; Ross, Greene and House, 1977). The observer's (recipient's) attributions can hardly be separated from his projections. With the setting into operation of the self-defensive processes projection takes part in attribution. So the significance of the question to what an extent the recipient's impression about the other (the literary character) reflects that person's traits and to what an extent his own. This very specifity of literary discourse in the centre of the studies of Holand (1975). He analyzes the processing of a few American short stories in a several-hour interview with university subjects. He shows how the readers' characteristic patterns of defensive and adaptive strategies operate in order to shape the text that their whishes should be gratified. Holland's key term is the DEFT, that is, defense, expectations, fantasy and transformation. Expectation puts the literary work in the sequence of a person's wishes in time; transformation endows the work with a meaning beyond time. Defense shapes what the individual lets in from outside; fantasy shapes what the individual puts out from himself into the outside world. After all, by means of our D E F T we read recreating our identity.

Cognitive and Social Psychological Approaches

9

revealed, which again invites supplementations, additions. Though identification in the reception of the work takes much shorter a time and is less forcible than it is in an everyday situation, the motives are similar. The recipient of the work adopts the role of the character cognitively and emotionally to realize, at least in imagination, the aims fulfilled by the character. For this to become a model-imitating identification, more expansive cognitive processes are to be put into operation, which perhaps takes place only later in the course of daydreaming which is part of the function of desire-satisfaction. Though our thoughts have several times run along parallel lines with those of Kreitler and Kreitler, our reservation as to the literal interpretation of identification might even be greater than theirs. Due to the difference between the position (mode of life) of the recipient and that of the character, the reader cannot perceive, feel, and do the same things to the same extent as the protagonist can. Yet he can reach congruences with the protagonist, which results in a lasting understanding even if the process itself is partial and transitory. But making contact does not bring into action only what directly belongs to us. Literary characters can revive the lives and traits of real persons, without our experiencing them as if they had happened to us. They have meanings as embodiments of definite experience, in the light of the analogy perceived by us.

Expectations, first impressions; the relation between the reader's and the protagonist's point of view It is superficial to think that the reader knows what to expect, when he has already got to know with whom, where, and how he is confronted. It would indeed be daring to say this about a person we have been meeting every day for many a month. And not only the observer is to blame. We may have had little information about the very thing that is the most important from the point of view of his probable actions, fate. Especially so, when a change takes place in his life. Though it is true that one prepares with one's past and character the situations one gets into, at least the way one reacts to them, a new situation can bring to light many things, so far hidden or undeveloped. This possibility is so closely related to literary discourse that the reader takes for granted the significant extraordinary change. And he awaits it not idly, but anticipates what may happen, which are the - not infrequently contradictory alternatives that can occur. In this sense he is prepared for the unexpected. His near guesses (surmises) about the formation of the fate of the character, about the outcome of things, however, remain fragmentary, incomplete, and hazy in details, even if it turns out in the long run that there were some among them which were roughly relevant. As the reader progresses in the text, he revises or refines his former expectations. Occasionally a radical reorganization becomes necessary. Past information, already processed are put in another light. They are reviewed again. The

10

Lâszlô Halâsz

new altered meaning arrived at now gets superimposed, covering entirely what was formerly attributed to the work and/or the character(s). We might believe that the short story, contrary to the novel, does not necessitate this. The novel, as is known, describes the character of the protagonist in general, in its development, in the sweeping arch of the process of its formation, whereas - at least as compared to this - the short story presents a protagonist with a developed character, as if carved out of the process. The reader, however, gets his bearing as to the character of the protagonist(s) of the short story, too, when processing the continual, successive information. The first impression about the protagonist is relatively soon formed, even in possession of insufficient data. Later on, the protagonist shows other facts as well, consequently he undergoes a change from the point of view of the recipient. If the author had said about the protagonist everything which he deemed important by the time the reader formed a first impression the short story might as well have been brought to an end. In the course of forming a more final judgement about the protagonist, later impressions about him inevitably clash with earlier ones. Not rarely it is indeed a sort of wrestling, for the resistance of the first impression is not to be belittled; it acts as a basis which tries to assimilate all the further information to itself (on first impression in general, see Asch, 1952, Luchins, 1957,1960, Hovland, 1974, Anderson, 1965; in the analysis of literary works Perry, 1979). So we transfer several elements of the happening of our reading into the present time of reception. Simultaneously we also cut a path in the world of the work, in the direction of the past. By means of empathy, identification, and analogical recognition the picture in the making about the protagonist gets supplemented with a few probable elements from his past. As we proceed we realize that we must, again and again, establish the situation and our relation to it. The experience of the reader should be sharply demarcated from that of the protagonist(s). The most fundamental version of this is when the run of literary discourse progresses towards dénouement similarly to that of reading. Despite the coincidence of the directions, the units of the two time dimensions are greatly different. This, in itself, would not prove to be a difficulty for the recipient of the difference between the units of time in the short story or novel and those of reception remained constant from beginning to end, that is, if the given units of the text were always allotted happenings of the same duration. This, however, is not the case, for the given units of the text can mean the most varied amounts of time in the life of the protagonist, in sequence of events. Thus the experience of the reading is also dependent on the way how we process the seeming continuity of the text, its flow of time which once trickles, once gushes forth - feeling it truer than that of the reading. Different is the case when we get information from two or three directions. Once from the time and space where the protagonist happens to be momentarily, sometimes from his past, with an occasional jump into the far future. What renders difficult for us to connect really related elements is that we are informed

Cognitive and Social Psychological Approaches

11

successively about motifs which as a matter of fact are significantly separated in time and space; or we get information much later about motifs which took place much sooner. The run of literary discourse - not infrequently - does not coincide even in its direction with that of the reader. N o w reception also includes how the accumulation of information from beginning to end, from the present to a definite point in the future, collides with the giving out of information which does not regularly alternate the present, the past, and the future. The processing of the way of the giving out of information itself becomes an organic part of the experience. When we get to know the relation between the narrator and the protagonist, we find ourselves on a new road of the repeated determination of the situation of the protagonist and that of our own. We may have the impression that the text runs so smoothly and automatically as if everything that happened to the protagonist and the others had been simultaneously registered by a device which can observe and register everything in proportion of its importance without those concerned being aware of it. This communicating apparatus is, of course, the writer himself, who, however, cannot be seen or heard at all and who did not even place a narrator between the text and himself. The reader has no other means at his disposal but what and how the author deemed important to commit to paper, when trying to find out the authors' intentions, judgements, attitudes. There is no information whatsoever, indicating which statement, question, or action of the protagonist could be emphatically decisive, while the reader is unable to assess all the manifestations. In his uncertainty, he falls back on his overall impression. And this undoubtedly is formed dependent on how he sees the protagonist. So he finds himself in a paradoxical situation where he tries to interpret the writers position, his relation to the protagonist and to the world of the work with the help of the pattern of information into which he has already encoded his own relation. At a first glance it may appear that the reader can get better guidance when the protagonist is the narrator himself. There is no room for misunderstanding: the protagonist-narrator relation to himself, to the others, to his companions leaves a trace on everything whith we read. The weighing of information about the protagonist directly or about the external world always coincides with the weighing of the protagonist (his ability to observe, prejudice, morality). And it cannot be lastingly forgotten that behind everything which the protagonistnarrator says is the author. Thus the uncertainty of the reader becomes even greater. Now he tries to determine the author's position, his relation to the world through the mediation of the protagonist-narrator. That is, on the basis of a pattern of information into which the protagonist, from beginning to end, encoded his own relation which can be comprehended only from that pattern. And this time it is even more inevitable that it should also be the result of interpretation how the reader turns this complex series of information into a pattern for himself. Various dimensions are to be fitted together, collated, or transgressed. In the same way as the

12

László Halász

reader sees the protagonist and as the protagonist sees himself and/or the world, as the protagonist thinks to be seen by others, as the writer sees the protagonist, and as the reader sees the writer. When the reader is told the happenings by an independent narrator, the former's task continues to increase: he has to clear up the point of view of the narrator and the relation between him and the narrator. 7 The reality of the imaginary; relation between the world outside the work and the world of the work Meanwhile, the reader compares, draws a parallel, supplements - in accordance with empathy, identification, and analogical recognition - the knowledge about the protagonist on the basis of his knowledge about himself and acquaintances. When he makes contact between the world of the work and the world outside the work he interprets one related to the other. At the same time, it is also obvious that the protagonists are the products of the author, even if the episodes which could play a role in the writing of the work happen to be known and were real. The collision between, the weighing and measuring the varied factors evoked by literary discourse becomes overloaded with the matching of the imaginary actions of imaginary characters to the real actions of real characters. From the first part of this rather summary statement it is evident that the imaginary actions of imaginary characters are derived from a real human being, from the author. He plants into the fiction several elements of reality as accurately as is rendered possible by human perception and memory. So in the first place the elements directly taken from reality without any change and the modified ones collide and dovetail within the presented imaginary world itself. The recipient, consciously or subconsciously, immediately alters them by means of his imagination. Not only because he arrives at the imaginary actions of imaginary characters through decoding inanimate letters, abstract symbols. And not only because he enriches the experience with the motifs of other artistic creations, that is, with the recalling of events, situations reproduced and reconstructed in the mind. He does not use only exact data either when mobilizing knowledge concerning the real actions of real characters. This becomes the most apparent when he draws a comparison between the protagonist and himself. In vain is he a flesh and blood being, he would not have a picture of himself without imagination which unifies, fills in gaps, condenses, and alters. 7

Uspensky (1973) analyzes in detail the numerous variations of the point of view: the phraseological point of view, the author's position, the point of view in space and time, the psychological point of view as the problem of the author's knowledge, the so-called composite point of view: the interweaving of several points of view (phraseological, psychological, ideological), and the pragmatical aspects of the point of view.

Cognitive and Social Psychological Approaches

13

N e v e r t h e l e s s the recipient is inclined t o take as if it w e r e real w h a t h e is a b s o r b e d in. 8 H e r e w e are r e m i n d e d o f the legend a b o u t the C h i n e s e painter w h o after painting his m a s t e r p i e c e stepped i n t o the landscape, w a l k e d a w a y along t h e p a t h , a n d disappeared in the b e f o g g e d m o u n t a i n s , in the p r e s e n c e o f t h e imperial s p e c t a t o r , w h o also feels like f o l l o w i n g h i m . A n d as t o the recipient o f literary d i s c o u r s e , h e prefers t o r e g a r d e v e r y event, situation, conflicts as real, p u s h i n g aside t h e fact t h a t it is o n l y fiction. H e c a n n o l o n g e r u p h o l d this attitude w h e n it is already certain t h a t the d r e a m s , c h i m e r a s , and hallucinations get m i x e d w i t h his " u s u a l " visual a n d auditive sensations, w i t h o u t the p r o t a g o n ist being a w a r e o f this d i s t u r b a n c e . A c h a n g e in the attitude b e c o m e s even m o r e n e c e s s a r y w h e n t h e r e a d e r , in lack o f i n f o r m a t i o n , is all the t i m e u n c e r t a i n a b o u t w h a t t h e p r o t a g o n i s t " o n l y " imagined and w h a t - w i t h i n the logic o f fiction m i g h t really take place. A f t e r all, t h e recipient imagines the p r o t a g o n i s t as did the w r i t e r , w h o m o v e s t o a n d f r o m b e t w e e n i m a g i n a t i o n and reality w i t h o u t e n c o u n t e r i n g a n y o b s t a -

8

So it is not surprising at all when Koestler (1949) finds that "the images of real people in our memory are not as different from our images of fictional characters as we generally believe". Kreitler and Kreitler (1972), who cite Koestler, have pointed out the concepts and techniques we usually use to describe people do not greatly differ from the ones with which we describe specifically literary figures. There are also data to indicate that the writer forms his characters in accordance with the implicit theory of personality of his own, that is, his system of biases operating in the course of his person perception (Bruner and Tagiuri, 1954). Rosenberg and Jones (1972) have studied Dreiser's work entitled Gallery of Women. They collected each and every trait as many times as it occurred in the case of each and every character. Then they checked how often and what pairs of traits appeared together. This measure shows the similarity observed between the traits in a pair, the probability that one comes from the other. They found that only some personality dimensions had a determining role in the interpretation of the traits. Namely: conform = successful, full of repression, conventional; non-conform = radical, liberal, lonely; masculine = hard, sincere, great; feminine = soft, sensual, charming, clever. These were the main dimensions of Dreiser's implicit theory of personality. When he knew one's sex and conformity, he supposed implicitly that he would also know all the other remarkable traits closely connected with them, so he would know the man himself (or he was able to characterize him). Helson (1970,1977) has examined, besides the authors, the books themselves with a personality assessment questionnaire. She asked judges to rate a sample of works of fantasy. So formal literary dimensions, needs (achievement, accomplishment, aggression, order, exploration or discovery, contact with the unreal emotion, wish-fulfilment, independence), relations among characters, properties of style can be described. Whereas Simonton (1978) elaborates a version of the biographical approach. It shows to what an extent the characteristics of a given work are the consequences of biographical and historical events taking place at the time of the writing of the work. Roughly speaking we might say that he considers the work as a living manifestation of the author and studies the extent to which it is the consequence of the autobiography. Moffet (1975:216) goes as far that he thinks valid not only for literary discourse but for other arts as well that, in opposition to other inanimate objects, "they are specifically structured to facilitate projection of human life and are thus readily personified". He suggests that the research in psychology of arts should regard "art objects as people" which phrase "is a concise though imprecise way of saying that art works embody qualities of persons".

14

Laszlo Halasz

cle. This does not merely add a new plane to the relation between the imaginary and the real, but multiplies their intertwining and collision. The reader of lyric poems is not expected to deal with the actions, fates of fictitious persons. He may come across stylization, condensation, symbolization; but never the mode of removal and transfer usual in narrative epic. The lyricist, says Becher (1955) is his own creator, he is the protagonist. The mechanisms, however, of empathy, identification, and analogical recognition are less forcible, if the character is not sharply marked from other figures of the world. It is a special characteristic of the lyric poem, which depicts both the external world and the memories, thoughts of the poet as part of his "sensitive mood" (Hegel), that the author becomes entirely one with what and how he relates. Progressing step by step, that is, proportioning, cannot be dispensed with in the mental activity of the lyricist either, but he is characterized by a true-to-life identification with existence and phenomena. So the reader has very limited access to data indicative of the character and circumstances of the protagonist with which to qualify, interpret him and his own relation. This may be the main reason of the fact that sensations generated by emotionally emphatic vocalverbal formulae are not as clean cut as those in literary narratives, they cannot be linked to the behavior of the protagonist(s). They are more shoreless, more diffuse.

New, surprising, incongruent; familiar, customary, congruent components and the reader's catharsis The perception of literary discourse of every kind presupposes certain intellectual-esthetic emotions. The collation of the past and present, leads us to the experience of novelty; that of expectation and outcome to surprise. Quite understandably, uncertainty is due to the collation of simultaneously incongruent expectations, the comparison of simultaneously incongruent response tendencies gives rise to conflict experience, and relating the parts of a single experience to one another leads to the most intimate experience of complexity (Berlyne, 1960, 1971). Again Berlyne points out that, when we are exposed to new, surprising, ambiguous, problematic and/or contradictory impressions, to so-called collative variables, an explorative behavior grows on us, accompanied by strong curiosity. It goads us to find further information to directly make uncertainty or conflict disappear, or to trigger off the process of thinking which leads to the solution of the problem. So the energy to resolution is given by the very tension which is aroused by the multilevelled pattern of stimuli.9 9

The measure of the novelty, complexity, and surprisingness of the work is connected with the recipient's level of arousal. The more novel, the more surprising a work appears the higher level of arousal accompanies it; the more familiar, the more customary the work, the lower the level of arousal. In general the reception of works having either too high or too low level of arousal is unfavourable.

Cognitive and Social Psychological Approaches

15

However, it is the perception of multilevelledness the amassing, confrontation of information that directs attention to what the information says about itself (about how "it is written") (Jacobson, 1960). With this, of course, a new source of tension is opened up. While resolving the accumulated tension, we are simultaneously arousing another one. Yet, it would be more than inaccurate to think that the main source of the reader's enjoyment is the chaos derived from multilevelledness, the drastic upsetting of the state preceding reading. He is attracted by the challange finally to find, upon discovering the meaning(s) of the text, new principles of organization. Besides, the reader does not easily give up looking for information which is in perfect harmony with one another and with his former knowledge. Thus he sees anew in literary discourse what he already knows, thinks, and is accustomed to. Naturally this has the effect of self-reinforcement, which is not an insignificant source of pleasure. As a matter of fact, the aroused dynamics of tension is multiple in one more respect. It is the clash and interaction of cognitive motifs, inherently upsetting (multiple, contradictory) or increasing (congruent to one another) balance. That is, as a result of the reciprocally counterbalancing effect of the two groups of motifs, the complexity of the former keeps growing The consequences of the reduction through repeated reception of the greatness or effect evoked by the work, in other words that the work gradually loses its arousal potential, have been the most concisely and radically made by Martindale (1975) in his undertaking to work out the psychology of literature. Its essence is that the interest in, preference, and attention to works of identical, or similar, type by a string of consecutive authors, will flag in time. To counterbalance the trend pointing in the direction of customariness, consecutive works are to have greater and greater arousal potential. And in this, novelty, incongruity, complexity, unpredictability, and other collative variables have an outstanding role. These are created by authors in the framework of a biphasic process, says Martindale relying on the idea of Kris (1952). The beginning inspiration stage - which is accompanied by regression in the service of the ego - is followed by another phase with a relatively less regressed way of thought. Regression is a return of the secondary thought process of the primary thought process. In Martindale's opinion the secondary and primary thought processes form a continuum which is the axis of the changes in the various states of consciousness and types of thought. As Freud (1963 (1900)) writes, cognition characterizable with the secondary process is abstract, logical, and reality-oriented. The primary process, however, is concrete, irrational, and autistic. The latter is the way of thought of dreams and disturbed psychological life. B y means of his content-analytical procedure Martindale proves that the authors create novelty mainly with the primary process, with the increase of the depth of regression, with the growth of free association, with that of the more and more remote analogies. With all this, however, literary discourse becomes fraught with new contradictory challenges. The recipient does not pay any attention to and rejects texts which are so familiar that he finds nothing in them (they have too low level of arousal) and those which are too complex (they have too high level of arousal) respectively. Whereas the novelty, complexity, incongruity the author deems necessary for middle level of arousal can for a shorter or longer time prove too high level of arousal even for a more educated recipient. Its effect, in turn, is that he eliminates some of the elements he finds too new, surprising, and incongruent, that is, he deprives the work of a certain proportion of its complexity. As a matter of fact he is able to tolerate, to receive it only like that, for the arousal evoked by the seemingly incomprehensible shocking moments reduces his processing capacity as well (Konecni, 1976-1977).

16

Laszlo Halasz

and the latter, because of their influence, also continue to lose of their unambiguity.10 Though not without precedent, we have circumscribed, in a somewhat unaccustomed way, catharsis as evoked by literary discourse. Richards (1926) emphasizes that Coleridge characterized poetic and every valuable experience as something whose essence manifests itself in the equilibrium or reconciliation of contradictory or incongruent qualities. Tragedy is the best example for this, because the differring impulses create a harmony, more perfect than that of any other artistic experience. Their unification into a single organized reaction is the catharsis. Vygotsky (1965), in analyzing a short story, and a tragedy as well, stresses that all the moments the work is made up of are aimed at extinguishing the direct emotional effect of the narrated event, and transforming it into something which is the very opposite of the former. Thus catharsis is the outcome of the collision 10

The responses to the challenge of the recipients differ dependent on their cognitive styles. Rittelmeyer (1969) exposed excerpts of traditional and modern musical compositions, reproductions of modern paintings, and a modern literary text to university subjects who judged each work and author, then they filled in a scale to rate the degree of authoritarianism (Adorno et al., 1950). The results prove that the moderner the work is the more unfavourable its reception is, the unfavourable judgement of the work is also extended to the author; closed-minded (Rokeach, 1960) and authoritarian subjects judge modern works more unfavourably, but in the case of traditional ones they do not differ from egalitarian and open-minded subjects; closed-minded and authoritarian subjects' opinion about modern arts, let them be fine arts, music or literature, is ab ovo more unfavourable. Pyron (1966) made his subjects judge reproductions of popular-commercial, classical, and modern paintings; literary and musical works (or their excerpts), and fill in various attitude scales as well. Those who have difficulties in tackling unexpected arrangement of the elements, easily reject foreigners, have more prejudices, are more closed-minded and irritated by any kinds of change, and often need outer backing, are markedly inclined to reject modern literature and other modern arts. The basic idea of the consistency theories (Heider, 1958; Festinger, 1964; Osgood et al., 1957; Rosenberg, 1960) is that the decisive trend of organism is the effort aimed at balance, harmony, and the strengthening of the expected, of the familiar. The complexity theories shift the stress just on the very opposite. The main motivation is not the reduction at any price of inconsistency, but the search for surprise, novelty, the unexpected. McGuire (1968) points out that under certain conditions a dynamic balance of the opposite forces may be struck. It is interesting that the opposite theories attempt to explain entirely different sectors of human behavior. The consistency theories deal with the "deadly serious" problems of economic welfare, self-integrity, the relation with others; the complexity theories examine man in his entertainment, play, aesthetic activities. Of this, however, - adds McGuire - the conclusion is not to be drawn that our need for variety is less fundamental than our aspiration after consistency. Notwithstanding we can say it unambiguously shows why the study of complexity is given a much greater role in the examination of reception. To generalize it: those who are more complex (Barron, 1963; Bieri, 1968), those with a more abstract conceptual system (Harvey, 1967) are more inclined to receive any kinds of (literary and non-literary) text differing from their knowledge, their beliefs, due to their greater ability to grasp and comprehend the essence, that is, due to the characteristics of their information processing. As opposed to them, reverse reception can be expected of authoritarian and more closed-minded individuals.

Cognitive and Social Psychological Approaches

17

between the structure (form) of the text and the structure forming the basis of the material (contents), in which the tensions destroy one another. In reality, these tensions fall into more manifold proportions. The contradictions within the structure of the text, within the structure of the narrated material, and the ones emerging between the elements of these structures and the reader's reactions, are also decisive - and not only in the form of destroying one another at all. Catharsis is sometimes the outcome of the emotional potential changes which are inseparable from cognitions and recognitions taking place, in the course of reading; another time, they themselves prepare the sole recognition, shocking and extraordinarily significant. Again, reference is to be made to the relations between the reader and the protagonist. Even doubly so. To wit, the protagonist himself, due to the effect of the conflicting events which may radically change his fate, also undergoes catharsis. This way the reader can enrich his experience that, while he determines his relation to the protagonist through the confrontation of his position, the protagonist's catharsis also comes into interplay with his own, better to say, with the emotions engendered in him how he empathizes with the protagonist through interpreting it.

The polyvalence of literary discourse as challenge for the reader. Channel capacity problems It has already been mentioned what biases, distortions one-sidedness can be the outcome of the prestige and contextual effect associated with the writer's name, referential-information network (foot-note 3), the projection and attribution processes of the recipient (foot-note 4), and his cognitive styles (foot-note 7). We have seen as well how the new, unusual structure of the work can be an obstacle - at least for certain recipients for a certain time - in the way of educated reception (foot-note 6). Though undoubtedly much is dependent of the competence of recipients, we cannot allow this fact to conceal those psychological characteristics of literary discourse that bring about the impoverishment or limitation of its peculiarly enriching possibilities. In Sherrington's classical metaphor (1906), our nervous system is a funnel turned with its wide opening to the outer world and with its narrow opening to our behavior. The amount of information entering the central nervous system is much greater than that could get through the narrow opening. Vygotsky (1965) thinks that somehow we need to gratify a considerable part of the "attractions and calls" which cannot get through the narrow opening. And to do so, art helps just through bringing into play the otherwise eliminated pieces of life, through receiving the existing but not realized patterns of behavior. When receiving art, we supplement information we otherwise cannot have access to, and counterbalance our one-sidedness (Simonov, 1966).

18

Laszlo Halasz

Let us compare - strictly only from this aspect - literary discourse with scientific one. Literature and science can be termed such systems which contrarily change the utilization of the recipient's channel capacity. In the first approach, I use this notion in a strictly cybernetic sense. Thus channel is an operational interconnection between two points, irrespective of there being a physical connection between them - as in technical communication systems - , or not (Ashby, 1964). Scientific discourse endeavours to give, as exactly as possible, co-ordinates in space and time, to determine strictly the reality and truth value of what is stated, further its sphere of validity (with the relation between reality and imagination in it). N o matter how complex and contradictory its conceptual contents are, they are conveyed by meanings, unambiguously attached to words and sharply separable from other words. Scientific discourse attempts to convey universal meaning, unchanged in space and time. Its conciseness is derived from eliminating everything which contradicts to the statements and the principles of logic necessary for verification (Marcus, 1974). Scientific discourse brings the recipient up against a clearly defined problem. The given information, the imaginary-thinking operations, and the aim in view are perfectly specified (cf. Reitman, 1965). Scientific discourse, in the interest of the most confinable and verifiable information processing - dependent on the peculiarities and development of its field - , makes the recipient filter out of his impressions to be produced everything which does not comply with the objective criteria of pure cognition. In this sense, the emotions, memories, and thoughts of the processor are burdened with narrowed channel capacity. As opposed to this, literary discourse is a text full of intentionally organized connotative meaning as Ingarden (1931) and Hartmann (1966) have described the different strata. The consequence of intentional use of connotations is that the contextual meanings are overwhelming and the sum total of the meanings cannot be computed (Marcus, 1974). So literary discourse is polyvalent. While in the case of polysemy in real contexts "the intended meaning is usually obvious", "polyvalence is a property of a context which signals that multiple meanings are indeed intended and recoverable, i. e. that polysemy is not to be eliminated" (Schmidt, 1982: 90). Beaugrande (1980: 292) stresses that literary texts are "vehicles for alternative organizations of the world and of discourse about the w o r l d . . . " According to the alternativity principle it is possible for us to imagine other worlds than the given reality (cf. Iser, 1978). All these offer a large scope to the recipient's emotions, memories, and thoughts. It is doubtless that compared to scientific, even to everyday discourse, processing literary discourse has a great degree of freedom. Great but not endless. That is, the degree of freedom of literary discourse cannot be compared to a projective test without having any sort of structure. A literary work is polyvalent as the poetic function is superior to the referential one but the reference itself is not eliminated by this (cf. Jakobson, 1960). Provided that the subject under-

Cognitive and Social Psychological Approaches

19

stood the instruction, with a projective test the question does not make sense how the subject could misunderstand the meaning so much or how he could catch it so deeply. With a literary work similar questions are fundamentally relevant. Undoubtedly the situation would be simpler if, during a given unit of time, literary discourse furnished the reader with less information. The limitedness of our channel capacity is fundamentally not dependent on the quantity of signs reaching our senses, but on the predictability of the signs, that is, on the number of expectable possibilities at given points. From this point of view, literary discourse is a further challenge. It taxes very hard the filter system called attention. The first stage of filtering generally refers to a physical feature of impressions contesting one another. The subject is set to receive stimuli from a given channel. In the second stage the message in question must be recognized before the filter selects or refuses it. The subject is set to respond to stimuli, which has a definite meaning (Broadbent, 1958, 1971). Filtering is not complete, the refused moments will be attenuated only (Treisman, 1960). And the extent of the limitedness of attention is not constant once and for all, but it changes from moment to moment. The effects, however, of a too multiple task cannot be counterbalanced through the increase of mobilized efforts (Kahneman, 1973). Attention becomes more and more indivisible, unidirectional. This happens in processing literary discourse as well. Dependent on his attitudes and motives, the reader literally does not pay attention to several factors. Ingarden exactly described it as early as 1931. The complexity of the grasping of the work is caused by too many simultaneous tasks encumbering the recipient, of which he does only a few centrally. So certain parts and strata of the literary text are lit up, others become obscure. We never grasp the work in its entirety, but always only partially, in perspective shortening. This shortening is a contrast between the richness of literary discourse and the relatively narrow consciousness of the reader. It may change from instance to instance, but can never surpass an upper limit. The work, however, is always above that level. Even a highly prepared and talented reader cannot perceive it in its entire richness. In the course of reading, distortion, falsification, and imperfect, insufficient imagination is always significant.11 11

The measures of this can be astonishing. Berlyne (1972) together with one of his collaborators took excerpts from poems and turned them into prose versions (with common non-poetic words, word order, sentence structure) for university subjects to listen to. "A little disturbing finding" of the experiment, says Berlyne, was that the researchers were unable to produce prose texts which the subjects, one day after the reading out, judged less favourably than the poems of three celebrated poets! In an experiment of our own (Halász, 1980) I made several versions of a conventional and a modern short story on the basis of various value-depriving points of view (unmotivated happy ending, cancellation of essential motives, depersonalization, explanatory remarks, addition of fashionable stereotypes). The subjects (well-educated grammar school students, aged 17 and 18) were given all the variations including the original. Immediately following reading they evaluated

20

Laszlo Halasz T h e w e l l - v e r s e d reader, especially the e x p e r t - as evidenced b y p r a c t i c e - is

fully a w a r e o f the situation. W h e n s u c h a p e r s o n aims at a t t e m p t i n g t o e x p l o i t t h e richness o f i n f o r m a t i o n o f a w o r k , he reads it several times t o take his t i m e at t h e v a r i o u s m o m e n t s . A n d y e t he k n o w s that even if he has s u c c e e d e d in getting n e a r e r t o t h e w o r k its e x p l o i t a t i o n is o n l y relative (cf. M o l e s , 1 9 5 8 ) . 1 2

Impoverishment (author)

and enrichment.

and the

Contradictory

relation

between

the

text

recipient

T h e t w o stages o f t h e o p e r a t i o n o f o u r attention can also be d e s c r i b e d in a w a y different f r o m the a b o v e . T h e first stage is the p r e a t t e n t i v e o n e w h i c h is c h a r a c t e r i z e d b y t h e parallel ( m u l t i - c h a n n e l ) p r o c e s s i n g o f i n f o r m a t i o n . W e t e m p o r a rily s t o r e the s i m u l t a n e o u s i n f o r m a t i o n t o avail t h e m for f u r t h e r analysis. T h i s activity is " w a s t i n g " . I n t h e s e c o n d stage, o u r attention is f o c u s s e d t o o n e p o i n t . Parallelism ceases,

12

at a given m o m e n t ,

w e can o n l y

observe

one

thing.

them as the versions of given themes. Though in the case of the conventional short story the original ranked first, the version turning the artist's intention to the contrary through the unmotivated happy ending was a close runner-up. With the more transformed, more complex work, however, the order of value sort of made a full turn due to the difficulties (distortions, incomplete imagination) of processing: what ought to have been dropped was selected, and what should be on top was pushed to the bottom. Even a rough falsification has a chance to appear more valuable than the original. Konecni (1984) gave undergraduates original excerpts and modifications from outstanding writers. Though the modifications were rather drastic stylistically and violated the conventions which according to outstanding critics are fundamental, the subjects gave no sign at all that enjoyed or comprehended them less than the original ones. Finally we remind the reader of the experiment by Nisbett and De Camp Wilson (1977) with a part of Rabbit, Run written by I. Updike. The original text and three rather shortened ones were used. The subjects read only one variation; then they reported on the emotional influence. These reports did not much differ from one another, that is, it was of no consequence whether the text in question was omitted or not. Later the subjects got all the other variations. Having read the original text the subjects declared that the passages which were omitted from the shortened variations increased the emotional influence significantly, that is, the original work was better. And having read some of the incomplete texts the subjects regretted lack of the missing passage and said how much greater the influence would have been if they had read it as well. All they said in details were quite similar to the opinion of the subsequent subjects who read the original text. So the subjects reported not on the real influence but on the things and the great influence they thought they ought to have exerted. But this can be explained not only by the tendency "telling more than we can know", not only by our causal inference that specific information obviously has to produce a specific answer. It is probable that this is more directly connected with the specific nature of the information. Otherwise during rereading the dynamics of cognition does not get wasted. Not primarily due to the fact that we perceive the structural properties which remained hidden in the course of the first reading. Rather because we already know what happens to the protagonist, and yet we weigh newer and newer possibilities, experiencing new facets of uncertainty and resolution in accordance with the alternativity principle (Beaugrande, 1983).

Cognitive and Social Psychological Approaches

21

Meanwhile, on the basis of incoming information, past experiences, attractions, and expectations, an interpretative - constructive process takes places (Neisser, 1967). Psychology discovered long ago the two kinds of cognitive process: images, thoughts can also be self-generated, not only the result of purposeful thinking. Putting the transitions into brackets: intuitive-rational, spontaneous-constrained, prelogic-logic, autistic-realistic, primary-secondary process. The mental activity labelled with the first element of the pairs is rich and chaotic, it generally only happens to one. The other kind, in turn, is deliberate, effective, and obviously purposeful, usually under self-control, says Neisser. These two kinds of psychological organization remind us of the difference between parallel and sequential processing. A sequential program constructs only one thing during a given time. As well as rational and logical thinking, it suggests that every image, thought or action is related to the previous one. A parallel program, however, carries out many activities simultaneously, or at least independently from one another. Its result may be useful, but not inevitably at all. Since unconscious trains of thoughts are really not parallel lines, which never meet, Neisser (1963) calls this activity multiple thought. The most striking analogy between preattentive processes and multiple thought is offered by dreams and the operation of imagination in general. Dreams and images can only be roughly defined, they quickly disappear. Unless they are grasped and elaborated through secondary process, they have little effect on further thinking and behavior. In solving problems, selective thinking is indispensable, which, in turn, is indicative of the characteristics of attention focussed to one point. Using Neisser's outlined argumentation, we can say that the processing of literary discourse is a peculiar situation because the topic, the pattern of stimulus itself, is entirely the result of the interpretative-constructive processes of another person. In reality, it is construction at its best, which is extraordinarily complex and multiple. The consequence of this is the waste which is characteristic of the preattentive process of the recipient, and, at the same time, the too free construing, that is, significant distortion. Being conscious of this, we ought to rephrase, the other way round, our fundamental question. What renders it possible at all that we can speak of reception of any kind? This is what would need explanation. On account of literary discourse being construction, it maps out limits to our mind, guides our purposeful thinking. It gives us the "main sequences" which our mental activity follows step by step. This is how thinking usually takes place. The main sequence equals to the operation of consciousness directed to a determined object, which is congruous to behavior and control motor manifestations. When, however, it is more or less divorced from the realization of actions, from motor activity, wishful processes, logically not fitting, also appear. While one of the processes analyzes the pattern of something seen (or recalled), the other one attributes the emotional meanings of personal needs to

22

Laszlo Halasz

certain features. Thus, occasionally, the direction of the main sequence is also modified. Literary discourse often shows this duality. On the one hand, information processing is separated from motor activity, from practical purposes, from adaptation to the outside world. With the already known simile: the funnel is turned not towards action, not towards overt behavior. This amplifies the moments which disturb the success of the main sequence with bizarre, irrational elements. On the other hand, in the course of processing, especially in that of the imaginary process, non-realized forms of behavior come to life as well. The indirect relation between them and the action may more clearly clarify the main sequence. Imageries evoked by literary discourse, however, do not entail an extraordinary flow of "inner images". The readers can well understand the position, behavior, and traits of the protagonists yet vividly descriptive imageries surface in them only accidentally. Imagination originally did not use to mean a metaphor as the "inner image", but an action, an active form of behavior, the imitation of a model, present or absent. Therefore imagination acts "as if behavior", a muted role-taking (Sarbin, 1972). It seems that the richness of muted role-taking of the recipient, how much he shares the fate of the protagonists, is dependent on the delicate ratio of proportions of his parallel and sequential processing, together with ego-involvement, and the increase of level of arousal. Obviously the reader who expects literary discourse to put into the centre of interest a protagonist, a world, and in a way which does not deviate from his (narrow and rigid) expectations and processing, drastically violates these relations. For him, catharsis is a smooth self-reinforcement generated in the course of following the protagonist's fate. The one-sided "material-contentual" attitude (Hartmann, 1966) becomes exclusive, which hinders the grasping of the artistically moulded strata of character and fate. Reception is built on the most probable and solid psychological supports. These are slightly sensitive to the nature and qualities of literary discourse. The examined contradiction appears here the most excessively, extremely magnified. The very reason why the recipient turns to a short story, a novel or a drama is to find peculiar surplus possibilities which increase the utilisation of his channel capacity. However, the intricacy, multilevelledness of works are fed into a system of information processing whose outlet (let it be action or conscious cognition) is of limited capacity. Even if literary discourse can be regarded as something counterbalancing the one-sidedness, limitidness of information processing, the reader can be regarded as someone compensating the obstacles of processing with one-sidedness. The more forcibly so, the less competent the reader, better to say, the more original, newer or surprizing the text. Lotman (1972) points out that the reception of literary discourse is always a fight between the author and the recipient. The reader makes an effort to

Cognitive and Social Psychological Approaches

23

squeeze the text into the limits of his familiar images, ideas. He is interested in getting the necessary information with the least possible effort. While the author aims at increasing the versatility and complexity of the code system through the structure of his work, the reader is inclined to reduce it to the minimum satisfactory for him (see foot-note 6). All this helps to understand the cause of the success of trivial literature in the circle of less competent recipients. One-sidedness-impoverishment which compensates the obstacles of processing is the essence of such a work itself. The author of pseudo-literature himself "fits" the simplification of the recipient into the structure in advance. By doing so, among others, he spares the reader the burden of greater dissonance and satisfies his strong need for security (cf. Groeben and Scheele, 1975). We must go in even further in the examination of interaction between literary discourse and the reader. We are to take into consideration not only the consequence of the reduction caused by the recipient by his trying to awake consciousness of literary work but also by - from opposite direction - the not necessarily "illiterary" attitude by which one processes non-literary texts. We cannot forget about the lesson of some literary tendencies. The French new novel as specific "non-fictive" literary discourse gives up the romantic elements and its effects is in the heroic epos originating from all the written individual or collective happenings. So - as it was stated by Sarraute (1956) nowadays the reader prefers a documentary-based story (or at least it seems to be based) on real happenings to a novel. Or let us think of the so-called "white writing". This neutral form is the "zero degree of writing" (Barthes, 1972). Wishes, calls and judgments are not to be found in it. It tries to eliminate all the metaphorical and anthropomorphizing descriptions. Though we have tried to draw a sharp line of demarcation between everyday, scientific or literary discourse and to furnish the criterion for literariness, the theory of literature is not limited only to fiction but it includes some part of everyday and scientific texts from thematic and stylistic point of view - at least in some ages. N o doubt it is like this nowadays as well. Cyzars (quoted by Wehrli, 1951) remarked some decades ago that he had found the released aesthetic energy in none of modern literary works to such a great extent than in a given ophtalmological text-book. There is no literary object, there is only literary function that any sort of written text can have - said Genette (1972). And as Chomsky (1970) emphasized literariness depends not on the text but on the reader. That is, not the work in itself is literary but the way how it is read. Practically all the kinds of texts can be read as literature or non-literature. Pratt (1977) also put down how important the readers' decision is whether to read a work of literature or not. Here fits in Schmidt's (1981) view according to which "literariness" and "meaning" are the results of the recipient cognitive operation. When we point out that - in accordance with the view and traditions of our narrative culture - the reader can process as literary discourse a non-fictive text obviously not written with literary aim, it (also) means, at the same time, that he

24

László Halász

perceives it as multilevelled and polyvalent. He gives emotions, memories, and thoughts of his own a large scope. In this sense, consequently he enriches the text by all means. When, however, we speak about the contradiction between the extraordinary complexity of literary discourse and the recipient's limited capacity of processing, we bring into the centre of attention the reduction as a result of which the recipient impoverishes literary discourse. At any rate, this must be added that in the preattentive period a global processing occurs and by the quick comprehension of the visual associations their meaning as "tacit" knowledge is available for our perceptive system before verbalization (cf. Posner, 1973, Turvey, 1973). The enrichment and operation of this peculiar treasure of knowledge are inseparable from how the reader enters into the spirit of the work - even if there is no trace of this in the verbal indices of reception. So the reader's impression may be correct when receiving a literary work he feels getting and understanding a lot of things. At the same time it is also understandable why a significant part of the work, even of the experience slips out when during recall he tries to denominate what he has read which requires the attention of limited capacity.

Knowledge of world events and knowledge Schemata in (literary) text processing

of structure of stories.

Simplification, distortion, narrowing, impoverishment, falsification - all are moments which are indicative of the limits of remembering and, simultaneously, of those of comprehension. Experts in folk tales (cf. Propp, 1968 (1928), Lüthi, 1960) have pointed out that what is too complicated or subordinated, not homogeneous at the first hearing will fall out from memory. So if symbols of a fairy tale are too unusual or complicated, due to the elimination during reception the tale will become a sentimental and banal story. Speaking explicitly about the reader of a literary text, Ingarden (1968) emphasizes that he condenses, summarizes certain sentences, groups of sentences. He commits to his mind only the salient of events, the most important character, or facts important for him. Knowing Bartlett's (1932) life-work and the matchless career of the notion of schema, all this is more than self-evident. Obvious is the significance of the examinations carried out with brief, simple stories, aimed at studying their structure and text processing, the reader's remembering and comprehension. Of the rich story understanding literature we refer only to a few whose connections and/or lessons concerning literary text processing are markedly relevant. Kintsch and van Dijk (1975) have pointed out that "semantic reduction rules" operate in the course of text processing. We are strongly inclined to delete and summarize the description of the reasons, intentions, and inner consequences of actions; the description of the events and sequences of actions forming the

Cognitive and Social Psychological Approaches

25

possible alternatives; the secondary conditions or consequences of the main actions; the descriptions of place, time, and circumstances not strictly connected with the action; the description of persons and objects, unless they are the conditions of special actions; repeating, commenting, and summarizing descriptions, and finally that of the dialogue. At the first glance we would be apt to take them for the most essential criteria of literary discourse. Even if we immediately admit our inaccuracy, it is doubtless that literary discourse surpasses every other kind of discourse due to the delicate nuances, richness and profusion of details of the descriptions mentioned. So literary discourse puts the reader's memory to an extremely difficult test. The reduction of intentions, reasons, inner consequences significantly influences the processing of even the most traditional and prominently coherent short story, even if the short story is really short and the text "hangs together" which helps the recipient's memory. Coherent text is connected discourse which helps the recipient to relate text to what he already knows. Each sentence or clause is subject to contextual interpretation. That is why a reader has to relate new incoming information to the information he already has, either from the text, the context, or from his general knowledge system (Kintsch - van Dijk, 1978). So comprehension is a continuous modification of the initial knowledge structure. The importance of the reader's knowledge in the comprehension and memory of the text explains how the recall and summarizing of familiar or well-structured events take place. Kintsch and van Dijk (1975) have found that the summaries of a conventional story were significantly similar among the readers; the organization of this story matched their expectations. But the summaries of an "unconventional" story did not show such similarity in the least; the organization of this story did not conform to the readers' expectations. Kintsch and Greene (1978) have pointed out that the recall of a conventional story (a Grimm fairy tale) was more successful than that of an unconventional story (an Apach Indian tale). Kintsch, Mandel and Kozminsky (1977) used a well-structured and a less wellstructured story in usual and scrambled paragraph order. The readers of the wellstructured story recognized it successfully even when they were given it in scrambled paragraph order. Their success was so significant that they made basically similar summaries with those who were given the text in normal order. Compared with a classic-realistic literary work a lot of unconventional works seem to be disconnected texts giving the impression of unrelated collection of sentences or lists of words. An unconventional literary work is unknown not only as its subject, but as its structure, its unusual way of expression and composition. Compared with conventional classic-realistic texts unconventional literary works seem to be less well-structured, even sometimes definitely scrambled. One of the most important consequences of this is that the connection between the text world and the average recipient's world becomes less and less. And when the contact with existing knowledge is minimal a text will be difficult to understand and remember.

26

Laszlo Halasz

The previously mentioned experiment by Kintsch, Mandel and Kozminsky (1977) and similar experiments by Mandler (1978), Stein and Nezworski (1978) show that the subjects reorganize the input and reconstruct the stories in the canonical sequence by the aid of their knowledge of the structure. The description of structural regularities (mainly in folk tales and fables from the oral tradition) is realized by story grammar. As Mandler (1982: 307), says: " . . . a story grammar is a rule system which specifies canonical sequences of units occurring in stories and the conditions under which they can be changed, deleted or moved. A story schema based on the grammar is a mechanism which has incorporated some of all of these regularities and makes use of them during processing." Van Dijk (1972) thinks that the narrative structure is based on a series of represented events and that a theory of action can be an appropriate basis for a theory of narrative. Lichtenstein and Brewer (1980) point out that the schemata, which according to Mandler and Johnson (1977), Rumelhart (1975, 1977), Stein and Glenn (1977), Thorndyke (1977) might underlie our understanding of one type of narrative, simple stories, have common properties. Namely: all of them are some types of category for encoding a goal-directed behavior event. All this does not mean that Lichtenstein and Brewer doubt that purely literary or linguistic components have an effect in text understanding. For instance, the common convention that the information about the location and major characters occur at the beginning of the story, makes the expectation that this information is the first to appear. It can also be known that the moral can be found at the end of the story if it can be found at all. These sorts of expectations are important components of one's story schema and help in reconstructing the canonical sequence. The experiments of Lichtenstein and Brewer, however, have proved, too, that in describing the behavior of characters subjects "rely on their knowledge of the structure of naturally occurring sequences of behavioral events... The ability to reconstruct the canonical sequence of story information may be as much a function of subjects' knowledge of the structure of naturally occurring events as of information which is specifically a part of their knowledge of the structure of stories" (Lichtenstein and Brewer, 1980: 442). So the success of trivial literature, that is, the fact that in a wide circle of recipients it is the literature, is due not at all only to motivation-energetic factors. In other words, not merely the attitude of escaping from reality has an effect, not merely the directly paid gratification of the reduction of the unpleasant inner tension is decisive. In the populous camp of such recipients is dominant a rather sketchy image of reality, as well as of world events, and they have most simplified impressions. Whereas "Individual can use those schemata which embody their extensive knowledge of world events to understand and later to reconstruct from memory, narrative descriptions of events" (Lichtenstein and Brewer, 1980: 443). So it is obvious that even from the strict point of

Cognitive and Social Psychological Approaches

27

view of text processing they are not able to receive literary discourse for which these schemata are rather rough and inadequate.13 Herewith, at the same time, we have been given an essential proof that our knowledge about how human actions take place, how the subjects of these actions proceed, is a sort of knowledge on which we can also rely when making expectations about how the protagonist attempts to solve the problem in a literary text (see foot-note 8). In this sense it is palpable that the results of text processing studies can best be turned to use at literary texts which themselves have traditional problem solving structure. And to the least use at those which build upon different structures: some modern forms of novels, short stories, dramas (cf. Winner, 1982). Winner goes on saying that the story grammar does not deal with the point of view. Though in some stories everything is shown through the omniscient eyes of the author, in others through the perspective of one of the characters. However, some researchers dealing with story understanding have taken the first steps in this respect as well. In the light of the studies by Abelson (1976), Pichert and Anderson (1977), Anderson and Pichert (1978), Bower (1978), we have a clearer picture of how essential the role of perspective in recall is. It has been proved that even a brief, simple text can be differently interpreted based on alternative schemata. If even in such texts it was dependent on the perspective what the readers found important in them, we cannot have any doubt about the relevance of the alternative schemata of the reader of literary text. The polyvalence, the alternativity of literary discourse can essentially be regarded as a challenge for the reader to alternate his perspective, that is, to observe the protagonists, to rehearse the events from the point of view of a neutral observer; from that of the protagonist and several other characters, etc. At the same time, it must be stressed that the stronger the reader's inclination is to identify himself only with a single character he finds congenial, the more likely he is to give up the multilevelledness of interpretations, the significant use of alternative schemata offered by literary discourse. Naturally the special laboratory conditions themselves among which the experimental investigation of text processing takes place also limit their validity; the recipient of the literary text treats a different text in a different way (cf. Neisser and Hupcey, 1974/75, Dillon, 1980, Spiro, 1982). But Mandler or Kintsch has the same standpoint. According to the story schema theory the top-down or conceptual-driven theory plays a role in understanding and memory. This, however, is not "the sole explanation" not only in literary discourse, but not even in the story processing of the simplest story. "A 13

Brewer and Lichtenstein (1982) point out that the popular literature (the detective story, the Western, the romance, etc.) the overall purpose of which - as opposed to the "serious literature" - is entertainment, which is built on a discourse structure producing surprise and resolution, suspense and resolution, curiosity and resolution.

28

Laszlo Halasz

schema provides a type of information that is useful in guiding comprehension and a base of information upon which to construct the inferences that are necessary to understand connected text. A schema also provides a retrieval mechanism in the form of temporally guided search plan, as well as a reconstructive mechanism to fill in gaps in memory. But it was never meant to exlude or to account for the effects of sentence structure, word frequency, emotional value of material, rhymes, or a host of other kinds of information available to the remember" (Mandler, 1982: 309-310). Kintsch (1980: 98) also stresses that " . . . the comprehension process is controlled not merely by a single (conventional) schema that is imposed on the process in top-down fashion, but several different schemata..." They "are activated in a bottom-up fashion from the text itself, or that are indeed constructed while reading a story from the information so far processed... Different control schemata allow for different levels of comprehension..." That is, the reader's processes of reduction are in inverse ratio to the degree of their competence as they possess many sorts of schemata and ability to create new schemata. While only certain strata can be grasped with certain schemata, their one-sidedness is also counterbalanced by different schemata to grasp other strata. The more one is dealing with more varied schemata on the world events on the one hand, and more varied literarily-aesthetically authentic schemata on the other hand, the more probable it is that whereas one will impoverish the work of art in some respects, in others one will be able to employ his channel capacity better and this increases the possibilities of richer processing.

The circumstances giving rise to the volume, and its structure Our volume is compiled on the basis of the papers presented at the Conference on Social Perception and Understanding of Interaction in Literature. (The conference was held in Budapest, September 1983, in the framework of the scientific agreement between the American Council of Learned Societies and the Hungarian Academy of Sciences.) As expounded above, we have already seen that one of the most specific and probably the most complicated forms of discourse processing is that required by literary discourse. To take specifities may be an important step for getting to know human information processing. Besides these we have to take into consideration the centrality of social perception in human affairs in general and relationships between different cultures in particular, moreover the importance of literary discourse as a symbolic mode of communication in formal and informal learning and in shaping cultures. If these are confronted with the psychological findings at our disposal it is obvious why American and Hungarian social psychologists and psycholinguists dealing with the arts, especially with (literary) texts, co-operate in cognitive and social psychological approaches of literary discourse.

Cognitive and Social Psychological Approaches

29

The long-term aim of the co-operation is to content-analyze and test a wide range of short stories (and other literary phenomena) representing the two cultures, moreover to use a narrower range of the same texts in order to examine the social perception during reception with American and Hungarian subjects selected from the same points of view. Before beginning this specific cross-cultural research, as a first step mainly due to theoretical-methodological reasons it was the presentation of psychological approaches of literary discourse itself suitable, moreover its reception from the point of view of social perception and understanding of interaction in as many-sided a way as it was possible under the given circumstances. (As a common set of demonstration the authors could use an American, a Hungarian and a Japanese short story, the last one being equally different from both other cultures.) The studies comprehend a rather varied spectrum. From the artificial intelligence to the psychoanalytical text-theory, from the literary use of personality assessment to the examination of the reader's perception as hypothesis-making, from the psycho-linguistic understanding of literary discourse to "reading" stories in pictures, from the examination of the symbolic meaning of the short stories to the description of psychological characteristics of the short stories. So the studies give a good opportunity to make it perceptible how far a subject which seems to be rather limited an esotheric is thoroughly interwoven with the "great questions" of psychology and with the practical problems like the communication between different cultures. The first part of the volume deals, from the point of view of literary discourse, with the questions of artificial intelligence, processing schemata, story memory, understanding, and enjoyment. R. Abelson analyzes how big the gap is in the relation of artificial intelligence and literary appreciation. He enumerates five possible types of limitation, namely: analysis of plots, sensory-motor experience, moods, problems of aboutness, and the task of critical evaluation. He is of the opinion that it is common in the criticism of the capabilities of computers that computer programs for text understanding lack the background of sensory and "life" experience that would provide them suitable information for carrying out the appreciation and evaluation of literary texts. The author emphasizes the emerging difficulties, and the fact that, as the machine by imitations of human functions has considerably surprised us so far, further surprises cannot be excluded either. R. Beaugrande deals with the schemas for literary communication. He writes up literature significance for language psychology, the return of schema, the relation between communication and literary communication, the relation fictionality and historicity in literary communication, the lessons of the "found poetry", the question of super-coherence, and schemas for literature, moreover the verbalized responses of the ordinary reader. He thinks that contextualizing, complexity, familiarity, abstractness, and similar factors may occur in all kinds of the examined processes. Through getting acquainted with the processing

30

Laszlo Halasz

schemata promoting the agreement of some fundamental questions about complexity and literary texts, the psychology of literature contributes to the thro roughest possible exploration of the human processing capacities. Cs. Pleh explains the formal and content-based models of story memory compared in the same empirical material, investigating their predictive force. His main result is that the more "subjective" a model was, the better its predictions for recall were. He suggests a model of story comprehension and recall in which the reader tries to actively reconstruct the interpersonal action of the hero, using up all his social knowledge. A peculiar new task would be the psychological analysis of the relevant textual devices. As for literary discourse the author finds that emphasis lies on what is beyond the level of social action schemata. The selection of a given "action perspective" is a much more openminded process with a literary text than with brief, simple stories (traditional folk tales) in which the role of cultural schemata is decisive. J . Laszlo gives an account of the lessons of an experiment failure, analyzing the relation of the understanding and enjoying of literary discourse. He finds that paid and generally "obedient" subjects did not follow the instructions requesting a relatively superficial processing of a literary text (in the sense of "levels of processing" paradigm), which indicates that a literary text indeed generates a peculiar, deautomatized reading. From the point of view of information processing this means that enjoying a text is qualitatively different from understanding a text. Aesthetic reading can be regarded as a common cognitive style. It is probable that not only conversational styles or texts can be typified, but also the ways of processing language materials. The psychological study of literary discourse can contribute empirical data about "aesthetic reading" to the scientific description of this typology. The second part of the volume deals with the place of the short story within psychology of literature and contains the studies carried out with the given short stories on the reader's social perception and on the possibility of the application of narrative pattern analysis and personality assessment. M. Lindauer points out that due to its shortness and other consequences deriving from it the short story offers the researcher methodological advantages which other literary forms do not possess. In it the effective stimulus, intervening processes and the literary responses are easier to identify, trace, and measure. The role of the brevity of the short story is fundamental in carrying out regular studies under controlled circumstances. The reader of the short story immediately finds himself in the centre of the events, has hardly any information about what happened previously and will never get to know what happens to the characters after its end. Therefore the reader has to work hard for comprehension, fill in the gaps, and "search for the point". So the reader is in an intim and close relation with what he reads. L. Halasz studies partly with explorations and partly with experiments the social perception of literary characters and the understanding of their interaction. He thinks that hypothesis-making connected with social perception is

Cognitive and Social Psychological Approaches

31

rather strongly directed by any part (segment) of the short story - basically in accordance with the whole short story. From this point of view the beginning part of the story has no inherent effect. The secret of its key role is its position in the text. At the same time, the parts (segments) "behave", within certain limits, differently despite the coherence penetrating all the parts. This opens a wide enough gate to the unexpected and unpredictable; the role of early hypotheses directing processing has a remarkable degree of freedom. C. Martindale's starting-point is that if the narratives indeed have implicit symbolic themes the content relevant to these has to follow the coherent trend throughout the course of the narratives. For example, the night journey narratives, which refer to the widespread mythic and literary themes of a visit to hell or some analogue of hell, can have some different possible — literal, moral, religious, psychological — interpretations. These are tested by computed-aided content analysis. The statistical procedures are used to examine trends in primary process content and evaluative connotations of text words. The method is rather general: any sort of content can be examined to look for any sort of trend. The analyses of the given short stories verify this. R. Helson demonstrates the possibilities of the personality assessment approach. She shows how a possible tool adjective test list and Q-sort were employed in order to provide comparable and quantifiable descriptions of literary characters, moreover historical events on the one hand, imaginative literary works for children on the other hand. Then she reports on the comparison of several Hungarian short stories, stories written by women, stories by "famous entertainers" and some "world's great stories". In this comparison figure the dimensions of social interaction, pressures, needs, attributes of characters, moral attitudes, structural variables, stylistic characteristics, and reader reactions. She finds remarkable differences between Hungarian short stories and others from the point of view of co-operation, equality and task orientation of social interaction. The third part of the volume examines the questions of the measuring of the popularity of literary works, those of the psychoanalytical literary interpretation and the interpretation of usual narratives. D.Simonton contributes to the comprehension of the bases of successful literary (more precisely dramatic) communication. He starts from the obvious observation that some works of literature are much more successful than others. In the case of a successful product the author's intentions are realized in the appreciator's reactions. In the centre of his interest is the relation between the contentual and formal characteristics influencing success and the "greatness" of the work. It presents a paradigm which analyzes groups of plays in a two-stage causal model. "Differential greatness" is a function of given aesthetic attributes and the characteristics of form and contents are the consequences of biographical and historical events which took place at the time of the writing of the work. Though the case-study analysis deals with Shakespeare's plays, the procedure itself can be applied to other literary forms as well.

32

Laszlo Halasz

In the opinion of A. Bokay, Freud's early work on jokes has in its background an interpretation system which together with our contemporary linguistic knowledge can significantly expand our theories explaining the understanding and existence of texts of a certain type. Specific texts can be defined only in their characteristic, subjective functions. So we simultaneously need such a text, a system of social criteria guaranteeing, rendering possible the existence of the text, and a personal sphere representing the need. The three levels define one another, thus the phenomenon itself, the joke as such can be produced only through plying the threefold system. To establish a literary text theory we must modify our starting-point; it must be further developed in the light of understanding, that is, in that of the central idea originating from the nature of the text. L. Gross analyzes the interpretation of the mediated events in the symbolic strategies. He distinguishes four categories: invisible mediation, unobtrusive mediation, media-events, and scripted, staged events, and classifies them along three dimensions: observer's assessment, status of event, and interpretative strategy. The author thinks that the process of the understanding and evaluation of the visual narratives is fairly similar to the attribution process, which Heider ascribed to the "naive psychology of action". The author analyzes how the interpretation and aesthetic evaluation of the visual narratives include the attributes which refer to the director, containing the judgement of the director's intention on the one hand and his capability to realize the intention on the other hand. Remark of the editor This volume is being published in English in order to allow the widest possible access to these contributions, but since English is not the native language of some of the authors, there may be some stylistic rough edges. They have not been smoothened out on purpose, for which the editor hopes of the understanding of the reader. References Abelson, R. P. 1976 "Script, Processing in Attitude Formation and Decision Making" in: Carroll and Payne (eds.) (1976: 3 3 - 4 7 ) . Abelson, R. P., E. Aronson, J. W. McGuire et al. (eds.) 1968 Theories of Cognitive Dissonance: A Sourcebook (Chicago: McNally). Adorno, Th., E. Frenkel-Brunswik, N. Sanford and M. Levinson 1950 The Authoritarian Personality. (New York: Harper). Alloway, Th., L. Kramers and P. Pliner (eds.) 1972 Communication and Affect (New York: Academic Press). Anderson, N. H. 1965 "Primacy Effects in Personality Impression Formation Using a Generalized Order Effect Paradigm", Journal of Personality and Social Psychology 2, 1 - 9 . Anderson, R. C. and J. W. Pichert 1978 "Recall of Previously Unrecallable Information Following a Shift in Perspective", Journal of Verbal Learning and Verbal Behavior 17, 1-12.

Cognitive and Social Psychological Approaches

33

Aronson, E., J. A. Turner and J. M. Carlsmith 1963, "Communicator Credibility and Communication Discrepancy as Determinants of Opinion Journal of Abnormal and Social Psychology 67, 31-37. Asch, S., 1952 Social Psychology (Prentice Hall: Englewood Cliffs). Ashby, W. R. 1964, An Introduction to Cybernetics (London: Methuen). Barron, E 1963 Creativity and Psychological Health (New York-Toronto-London: Nostrand). Barthes, R. 1972 Le degré zero de l'écriture (Paris: Seuil). Bartlett, E 1932 Remembering (Cambridge University Press). Beaugrande, R. 1980 Text, Discourse and Process (New Jersey, Norwood: Ablex). Beaugrande, R. 1983 "Surprised by Syncretism: Cognition and Literary Criticism Exemplified by E.D. Hirsch, Stanley Fish, and J. Hillis Miller", Poetics 12. 83-137. Becher, J. R. 1955 Macht der Poesie (Berlin: Aufbau Verlag). Berlyne, D.E. 1960 Conflict, Arousaland Curiosity (New York: McGraw-Hill). Berlyne, D.E. 1971 Aesthetics and Psychobiology, (New York: Appleton). Berlyne, D.E. 1972 "Affective Aspects of Aesthetic Communication" in: Alloway et al. (eds.) (1972:97-118). Bieri, H. 1968 "Cognitive Complexity and Judgment of Inconsistent Information" in: Abelson et al. (eds.) (1968: 633-640). Blumensath, H. (ed.) 1972 Strukturalismus und Literaturwissenschaft (Köln: Kiepenhauer und Witsch). Bobrow, D. G. and A. Collins (eds.) 1975 Representation and Understanding: Studies in Cognitive Science (New York: Academic Press). Bodman, P. L., Micher, M., William Ch. and Lindauer M. 1976 "Judgments of the Positive and Negative Qualities of Names in Literature" in: Lindauer (1976: 34 - 40). Bower, G. H. 1978 "Experiments on Story Comprehension and Recall" Discourse Processes 3, 211-232. Brewer, W.F. and E.H. Lichtenstein 1982 "Stories Are to Entertain: a Structural-Affect Theory Stories", Journal of Pragmatics 6, 473 - 486. Broadbent, D.E. 1958 Perception and Communication (New York: Pergamon Press). Broadbent, D.E. 1971 Decision and Stress (London: Academic Press). Brunner, J. S. and R. Tagiuri 1954 "The Perception of People" in: Lindzey (ed.) (1954: 634-654). Carroll, J. S. and J. W. Payne (eds.) 1976 Cognition and Social Behavior (New Jersey, Hillsdale: Lawrence Erlbaum Associates). Chase, W. (ed.) 1973 Visual Information Processing (New York: Academic Press). Chatman, S. 1978 Story and Discourse (Ithaca and London: Cornell University Press). Cherry, S. (ed.) 1974 Pragmatic Aspects of Human Communication (Dordrecht: Holland). Child, I. L. 1968 -1969 "Esthetics" in: Lindzey and Aronson (1968 - 69: 853 -916). Chomsky, N. 1970 Sprache und Geist (Frankfurt am Main: Suhrkamp). Crozier W.R. and A.J. Chapman (eds.) 1984 Cognitive Processes in Art (Amsterdam: North Holland Publishers). Crozier, W. R. and A. Chapman 1981 "Aesthetic Preferences: Prestige and Social Class" in: O'Hare (ed.) (1981:242-278). Cysarz, H. quoted by Wehrli, M. 1951 Allgemeine Literaturwissenschaft (Bern: Francke). Das, J. P., R. Rath and R. S. Das 1955 "Understanding versus Suggestion in the Judgment of Literary Passages", Journal of Abnormal and Social Psychology 51, 624-628. Dillon, G. L. 1980 "Discourse Processing and the Nature of Literary Narrative", Poetics 9, 163-180. Eichenbaum, B.M. 1924 Szkvozj Literaturu. Szbomyik sztatyej (Through Literature. Selected Papers) (Leningrad: Academia). Fernandez, R. 1977 Social Psychology through Literature (New York: Wiley). Festinger, L. 1964 Conflict, Decision and Dissonance (Stanford: Stanford University Press). Francès, R. 1963 "Limites et nature des effets de prestige; II. Notoriété de l'auteur et jugement de l'oeuvre", Journal de Psychologie normale et pathologique 60, 437-456. Freedle, R. (ed.) 1979 New Directions in Discourse Processing (New Jersey, Norwood: Ablex).

34

Lâszlô Halâsz

Freud, S. 1973 Der Wahn und die Träume in W. Jensen "Gradiva" (Frankfurt am Main: Fischer Verlag). Freud, S. 1963 Die Traumdeutung. Gesammelte Werke, chronologisch geordnet. Bd. II-III. (Frankfurt am Main: Fischer Verlag). Genette, G. 1972, Strukturalismus und Literaturwissenschaft, in: Blumensath (ed.) (1972: 71-88). Göpfert, H . G. et al. (eds.) 1975 Lesen und Leben (Frankfurt: Buchhändler-Vereinigung). Grimaud, M. 1976 "Recent Trends in Psychoanalysis: a Survey with Emphasis on Psychological Criticism in English Literature and Related Areas", Sub-Stance 13, 136-162. Groeben, N . and B. Scheele 1975 "Zur Psychologie des Nicht-Lesens" in: Göpfert et al. (eds.) (1975: 82-114). Haimowitz, M. L. and N . R. Haimowitz (eds.) 1966 Human Development Selected Readings. 2nd ed. (New York: Cromwell). Halâsz, L. 1980 "Organization of Judgments of Works of Art and Psychological Characteristics of Value Recognition" in: Kardos, Pléh and Hunyady (eds.) (1980: 127-149). Hartmann, N . 1966 Aesthetik zweite, unveränderte Auflage. (Berlin: Gruyter). Harvey, O . J . 1967 "Conceptual systems and attitude change" in: Warr (ed.) (1967: 315-333). Harvey, J., W.J. Ickes and R. F. Kidd (eds.) 1976 New Directions in Attribution Research I (New Jersey: Hillsdale). Heider, F. 1958 The psychology of Interpersonal Relations (New York: Wiley). Heider, F. 1976 "Conversation with Fritz Heider" in: Harvey, Ickes and Kidd (eds.) (1976: 3-18). Helson, R. 1970 "Sex-Specific Patterns in Creative Literary F a n t a s y J o u r n a l of Personality 38, 344-363. Helson, R. 1977 "The Creative Spectrum of Authors of Fantasy", Journal of Personality 45, 310-316. Himmelfarb, S. and D. H . Eagly (eds.) 1974 Readings in Attitude Change (New York: Wiley). Holland, N . H . 1975 i Readers Reading (New Haven; Yale University Press). Holmes, D.J. 1968 "Dimensions of Projection", Psychological Bulletin 69, 248-268. Hovland, C . I . 1974 "The Role of Primacy and Recency in Persuasive Communication" in: Himmelfarb and Eagly (eds.) (1974: 558-575). Hovland, C . I . (ed.) 1957 The Order of Presentation in Persuasion (New Haven: Yale University Press). Hovland, C. I. and M.J. Rosenberg (eds.) 1960 Attitude Organization and Change (New Haven: Yale University Press). Ingarden, R. 1931 Das Literarische Kunstwerk (Tübingen: Niemeyer). Ingarden, R. 1968 Vom Erkennen des Literarischen Kunstwerks (Tübingen: Niemeyer). Iser, W. 1978 The Act of Reading (London and Henley: Routledge Kegan Paul). Jakobson, R. 1960 "Linguistics and Poetics" in: Seboek (ed.) (1960: 350-377). Jones, E. E. and E. E. Nisbett 1971 The Actor and the Observer: Divergent Perceptions of the Causes of Behavior (Morristown: General Learning Press). Kahneman, D. 1973 Attention and Effort (New York: Englewood Cliffs, Prentice Hall). Kardos, L., Cs. Pléh and Gy. Hunyady (eds.) 1980 Attitudes, Interaction and Personality (Budapest: Academic Publishing House). Katz, H . A . , P. Warrick and M . H . Greenberg 1974 Introductory Psychology through Science Fiction (Chicago: Rand McNally). Kiell, N . 1974 The Adolescent through Fiction (New York: International University Press). Kiell, N . 1963 Psychoanalysis, Psychology and Literature. A Bibliography. (Madison: Wisconsin University). Kingston, A.J. and W. F. White, 1967 "The Relationship of Readers' Self-Concepts and Personality Components to Semantic Meanings Perceived in the Protagonist of a Reading Selection", Reading Research Quarterly 2, 107-116. Kintsch, W. and van T. A. Dijk 1975 "Comment on se rappelle et on résume des histoires", Langages 40, 98-116. Kintsch, W., 1980, "Learning from Text, Levels of Comprehension, or Why Anyone Would Read a

Cognitive and Social Psychological Approaches

35

Story Anyway", Poetics 9, 8 7 - 9 8 . Kintsch, W. and T. A. van Dijk 1978 "Toward a Model of Text Comprehension and Production", Psychological Review 85, 363-394. Kintsch, W. and E. Greene 1978 "The Role of Culture Specific Schemata in the Comprehension and Recall of Stories", Discourse Processes 1, 1-13. Kintsch, W. T. S. Mandel and E. Kozminsky 1977 "Summarizing Scrambled Stories", Memory and Cognition 5, 547-552. Koestler, A. 1949 "The Novelist Deals with Character", Saturday Review of Literature 1. January. Konecni, V.J. 1976-77 "Quelques déterminants sociaux, émotionels et cognitifs des préférences esthétiques à des mélodies de complexité variable", Bulletin de Psychologie 14-16, 688 - 7 1 5 . Konecni, V.J. 1984 "Elusive Effect of Artists' Messages" in: Crozier and Chapman (eds.) (1984: 71-93). Kreitler, H. and Kreitler, S. 1972 Psychology of Arts (Durham: Duke Univ. Press). Kris, E. 1952 Psychoanalytical Explorations in Art. (New York: International University Press). LaBerge, D. and J. Samuels (eds.) 1977 Basic Processes in Reading Perception and Comprehension (New Jersey: Hillsdale Lawrence Erlbaum Associates). Lichtenstein, E. H. and W. E Brewer 1980 "Memory for Goal-Directed Events", Cognitive Psychology 12, 4 1 2 - 4 4 5 . Lindauer, M. 1976 (ed.) Research in Psychological Aesthetics and the Arts: a Personal Collection (New York: State College, Brockport). Lindzey, G. and E. Aronson (eds.) 1968 - 69 Handbook of Social Psychology III (Massachussetts : Addison-Wesley ). Lindzey, G. (ed.) 1954 Handbook of Social Psychology (Cambridge, Massachussetts: AddisonWesley). Luchins, A. S. 1957 "Primacy-Recency in Impression Formation" in: Hovland (ed.) (1957:33 - 61). Luchins, A. S. 1960 "Influence of Experience with Conflicting Information on Reaction to Subsequent Information " , Journal of Social Psychology 51, 367-385. Lotman, J. M. 1972 Die Struktur Literarischer Texte (München: UTB 103). Lüthi, M., 1960 Das Europäische Volksmärchen. Form und Wesen (Bern und München: Francke). Mandler, J . M. 1982 "Some Uses and Abuses of Story Grammar" Discourse Processes 5, 305 - 3 1 8 . Mandler, J . M. 1978 "A Code in the Mode: The Use of a Story Schema in Retrieval", Discourse Processes 1, 1 4 - 3 5 . Mandler, J. M. and N. S. Johnson 1977 "Remembrance of Things Parsed: Story Structure and Recall", Cognitive Psychology 9, 111-151. Marcus, S., 1974 "Fifty-two Oppositions between Scientific and Poetic Communication" in: Cherry (ed.) (1974: 8 3 - 9 6 ) . Martindale, C. 1975 Romantic Progression: The Psychology of Literary History (Washington: Hemisphere). McGuire, W.J. 1968 "Résumé and Response from the Consistency Theory Viewpoint" in: Abelson et al. (eds.) (1968:275 - 297). Michael, W. B., B. G. Rosenthal and M. A. DeCamp 1949 "An Experimental Investigation of Prestige-Suggestion for Two Types of Literary Material", Journal of Psychology 28, 303 - 3 2 3 . Moffet, L. A. 1975 "Art Objects as People: A New Paradigm for the Psychology of A r t J o u r n a l for Theory of Social Behavior 5, 215 - 2 2 3 . Moles, A.A. 1958, Théorie de l'information et perception esthétique (Paris: Flammarion). Neisser, U. 1963 "The Multiplicity of Thought", British Journal of Psychology 54, 1-14. Neisser, U. 1967 Cognitive Psychology (New York: Appleton Century Crofts). Neisser, U. and J. A. Hupcey 1974/75 "A Sherlockian Experiment", Cognition 3 (4), 307-311. Nisbett, R. E. and T. DeCamp Wilson 1977 "Telling More Than Know: Verbal Reports on Mental Processes", Psychological Review 84, 231-259. O'Hare, D. (ed.) 1981 Psychology and the Arts (Brighton, Sussex: Harvester, Humanities Press). Osgood, Ch., G . J . Suci, P. H. Tannenbaum, 1957 The Measurement of Meaning (Urbana: University Illinois).

36

Lâszlô Halâsz

Perry, M. 1979 "Literary Dynamics: How the Order of a Text Creates its Meaning", Poetics Today 1,35-64,311-361. Pichert, J.W. and R. C. Anderson 1977 "Taking Different Perspectives on a Story", Journal of Educational Psychology 69, 309-315. Posner, M. 1973 "Coordination of Internal Codes" in: Chase (ed.) (1973: 3 5 - 7 3 ) . Pratt, M. L. 1977 Toward a Speech Act Theory of Literary Discourse (Bloomington-London: Indiana University Press). Propp, V.J. 1968 Morphology of the Folktale (Austin: University of Texas Press). Pyron, B. 1966 "Rejection of Avant-garde Art and the Need for Simple Order", Journal of Psychology 63, 159-178. Rabkin, L. Y. (ed.) 1960 Psychopathology and Literature (San Francisco: Chandler). Reitman, W. 1965 Cognition and Thought (New York: Wiley). Richards, I. A. 1926 Principles of Literary Criticism (London: Routledge, Kegan). Rittelmeyer, S. 1969 "Dogmatismus, Intoleranz und die Beurteilung moderner Kunstwerke", Kölner Zeitschrift für Sozialpsychologie 21, 93 -105. Rokeach, M. 1960 The Open and the Closed Mind (New York: Basic Books). Rosenberg, M. J. 1960 "An Analysis of Affective-Cognitive Consistency" in: Hovland and Rosenberg (eds.) (1960: 1 5 - 6 4 ) . Rosenberg, S. and R. A. Jones 1972 "A Method for Investigating a Person's Implicity Theory of Personality: Theodore Dreiser's View of Peopl e", Journal of Personality and Social Psychology 22,372-386. Ross, L., D. Greene, P. House, 1977 "The False Consensus Effects: an Egocentric Bias in Social Perception and Attribution Processes", Journal Experimental Social Psychology 13, 279-301. Rumelhart, D . E . 1975 "Notes on Schema for Stories" in: Bobrow, and Collins (eds.) (1975: 211-236). Rumelhart, D. E., 1977 "Understanding and Summarizing Brief Stories" (1977: 265 -304). Sarbin, Th. 1972 "Imagining as Muted Role-Taking: a Historical Linguistic Analysis" in: Sheenan (ed.) (1972:333-354). Sarraute, N., 1956 L'ère du soupçon (Paris: Gallimard). Schmidt, S. J. 1981 "Empirical Studies in Literature: Introductory Remarks" Poetics 10, 317-336. Schmidt, S.J. 1982 Foundation for the Empirical Study of Literature (Hamburg: Buske). Secord, P. F. and C.W. Backman 1964 Social Psychology (New York-San Francisco-Toronto-London: McGraw Hill). Seboek, T. A. (ed.) 1960 Style in Language (Cambridge: MIT Press). Sheenan, P. W. (ed.) 1972 The Function and Nature of Imagery (New York: Academic Press). Sherif, M. 1935 "An Experimental Study of Stereotypes", Journal of Abnormal and Social Psychology 29, 371-375. Sherrington, C. S. 1906 The Integrative Action of the Nervous System Yale University Press: New Haven. Simonov, P. V. 1966 Sto takoje emocija? (What is Emotion?) (Moscow: Nauka). Simonton, D. K. 1978 "Time-Series Analysis of Literary Creativity: a Potential Paradigm", Poetics 249-259. Spiro, R. 1982 "Long-term Comprehension: Schema-Based Versus Experiental and Evaluative Understanding", Poetics 11, 7 7 - 86. Stein, N. L. and M. T. Nezworski 1978 "The Effect of Organization and Instructional Set on Story Memory", Discourse Processes 1, 177-193. Stein, N. L. and C. G. Glenn 1979 "An Analysis of Story Comprehension in Elementary School Children" in: Freedle (ed.) (1979: 53-120). Stephan, W. G. 1975 "Actor vs Observer: Attributions to Behavior with Positive or Negative Outcomes and Empathy for the Other Role", Journal Experimental Social Psychology 11, 205-214. Stone, A.A. and S.S. Stone (eds.) 1966 The Abnormal Personality through Literature (New York-Prentice Hall; Englewood Cliffs).

Cognitive and Social Psychological Approaches

37

Stotland, E. 1978 Empathy, Fantasy and Helping (Beverly Hills-London: Sage). Taylor, S. E. and J . H . Koivumaki 1976 "The Perception of Self and Others: Acquaintanceship, Affect and Actor-Observer Differences", Journal of Personality and Social Psychology 33, 403-408. Thorndyke, P. W. 1977 "Cognitive Structures in Comprehension and Memory of Narrative Discourse", Cognitive Psychology 9, 77-110. Treisman, A. M. 1960 "Contextual Cues in Selective Listening", Quarterly Journal of Experimental Psychology 12, 242-248. Turvey, M. 1973 "On Peripheral and Central Processes in Vision: Inferences from an InformationProcessing Analysis of Masking with Patterned Stimuli", Psychological Review 80, 1 - 5 2 . Uspensky, B. 1973 A Poetics of Composition (Berkeley: University of California Press). van Dijk, T. A. 1972 Some Aspects of Text Grammars (The Hague: Mouton). Vygotsky, L. S. 1965 Pszihologija iszkusztva (Psychology of An) (Moskva: Izdatyelsztvo Iszkusztvo). Warr, B. P. (ed.) 1967 Thought and Personality (Harmondsworth: Penguin). Weber, J. P. 1958 La psychologie de I'art (Paris: P. U. E). Werner, H. (ed.) 1955 On Expressive Behavior (Worcester, Massachussetts: Clark). Winner, E. 1982 Invented World. The Psychology of Arts (Cambridge, Massachussetts: Harvard University Press).

ROBERT P. ABELSON

Artificial Intelligence and Literary Appreciation: How Big Is the Gap? In the past ten years, there have been many new developments in the field of artificial intelligence (AI), wherein computers have been programmed to perform in ways somewhat similar to intelligent human performances. Of special interest to analysts of literature are developments in naturalistic information processing, in which the machine deals with typed inputs expressed in ordinary language, responding by printing ordinary language outputs. Programs have been developed which to some degree "understand" simple story fragments or other text materials. They demonstrate their understanding by summarizing the text or by answering questions about it (Schank and Abelson, 1977; Schank and Riesbeck, 1981). Other programs exist which carry on conversations or which engage in argumentation. Thus far, these AI programs are not very powerful. But their existence raises in principle the question of whether a computer could ever be programmed to understand or appreciate literature. This question is more philosophical than it is practical. (Who would want a computer to stand around reading books?) Fundamental issues are raised of what constitutes intelligent understanding of a piece of literature.

What some simple existing programs can do First, consider the text fragment below, and some questions that might be asked about it: J O H N WENT TO A RESTAURANT. HE TOLD THE WAITER HE WANTED STEAK. HE ATE QUICKLY, AND LEFT A LARGE TIP. Q : WHAT DID J O H N EAT? Q : W H O SERVED IT TO HIM? The problem here is that it is obvious to a human reader that John ate steak, and that the waiter served it to him, but those facts are not explicit in the story. Where, then, does the reader get this information? The compelling answer is that the reader invokes world knowledge, here a set of facts about what typically happens in restaurants: Typically a customer eats what he orders, and it is served

Artificial Intelligence and Literary Appreciation

39

by someone in a waiter role. Given this general schematic knowledge, it is easy for a person or a program to fill in, or "instantiate" the particulars from a specific text, such as the one above. At Yale we called the type of schema involving stereotyped event sequences a "script". We worked out computer programs which contained script knowledge, and which therefore could cope with texts concerning scripted events such as restaurant trips (Cullingford, 1978). A large set of human activities are scripted - doctor visits, airplane rides, wedding ceremonies, grocery shopping, and so on and so on - and thus a lot of prepackaged human knowledge can be made available to computer programs. Of course scripts are inflexible, and do not take account of spontaneous planning in novel situations, or expressions of feeling, or moral conflict, or many other elements which characterize literature. O f that, more later, but meanwhile, let us touch upon some other applications of script knowledge. Scripts occur in the unfolding of political events. Thus the understanding and possibly also the ideological interpretation of accounts of such events is simplified by script knowledge. There is an AI program called P O L I T I C S (Carbonell, 1978), which when given descriptions of political events responds by offering explanations of why they might have happened. This program has available a set of political scripts, and also some simple assumptions about the likely goals and plans of political actors. A somewhat related program named F R U M P (Dejong, 1979), reads the output of press stories about the news of the day, and then offers short summaries of types of stories it has been programmed to regard as interesting. F R U M P is internally provided with scripts concerning earthquakes, diplomatic negotiations, airplane accidents, changes of government, military campaigns, space launchings, police activities, etc. Figure 1 shows a F R U M P summary of a very complicated news story about the arrest of a man charged with murder. I will not take the time to try to explain how the program makes inferences about the dates when the various events occurred, but the reader may enjoy trying to puzzle this out. The performance of the program on this particular story was rather impressive. However, on some other stories, the program made ridiculous errors. Once it was given a news item which started with, "The death of the Pope shook the world." In its summary F R U M P said, "There was an earthquake in Italy. One person died." What the program did was to interpret the metaphorical phrase "shook the world" as a literal reference to an earthquake. This type of mistake was later repaired, but the possibility of such misinterpretations exposes the fragile nature of AI text understanding. NEW YORK, MAY 30 - A FEDERAL GRAND JURY YESTERDAY CHARGED LUIS SEREZ WITH FIRST DEGREE MURDER IN CONNECTION WITH THE MAY 16 SLAYING OF A NEW YORK CITY POLICE OFFICER. OFFICER FREDRICK GOVEL WAS ALLEGEDLY SHOT BY SEREZ WHEN HE ATTEMPTED TO QUELL A BARROOM DISPUTE IN A LOWER MAN-

40

Robert P. Abelson

HATTAN TAVERN. AFTER THE SHOOTING SEREZ ESCAPED THROUGH AN ALLEY SERVICE ENTRANCE. POLICE APPREHENDED SEREZ LAST THURSDAY WHEN HE SUCCUMBED TO A POLICE RUSE. HE WALKED OUT OF HIS WEST 19TH STREET APARTMENT BUILDING INTO THE HANDS OF WAITING DETECTIVES. A POLICE DETECTIVE REPORTEDLY PHONED SEREZ SAYING "SOME GUY SAID CALL THIS NUMBER AND SAY THE POLICE ARE O N THE WAY TO GET YOU." AND THEN HUNG UP. WITHIN MOMENTS SEREZ CAME RUNNING OUT OF THE BUILDING. ENGLISH SUMMARY: LUIS SEREZ MURDERED FREDRIC GOVEL. POLICE ARRESTED HIM O N MAY 22, 1975 AND CHARGED HIM WITH MURDER. HE WAS INDICTED FOR MURDER O N MAY 29,1975. Figure 1.

A Frump summary of a news story.

Around the time these programs were coming out, there was a surge of philosophical criticism of artificial intelligence. Dreyfus (1979) argued that computers could never successfully imitate human mental life, primarily on the grounds that computers lack the sensory-motor experience that humans have. Searle (1980) made a more sweeping accusation. He said that since computers lack the capacity for having hopes, fears, desires and other human "intentional states", they were forever incapable of having access to the understanding of texts. Instead, the matching of inputs to prepackaged scripts was nothing more than a stereotyped mechanical performance with no insight into meaning. These criticisms were made without real appreciation of the possible improvements in AI capability beyond the level of merely dealing with scripts. Nevertheless, the issues raised are general, and are quite relevant to the question of AI appreciation of literature.

Possible limitations on AI appreciation of literature As I see it, there are at least five possible types of limitations on AI appreciation of literature, that might be labeled as follows: 1. 2. 3. 4. 5.

Plots Sensory-motor experience Moods Aboutness Evaluation

I consider each of these in turn.

Artificial Intelligence and Literary Appreciation

41

Plots Texts about people who go to restaurants, eat and leave, are obviously unrepresentative of real stories. They are dull and pointless. Nothing interesting happens. For people to be willing to call something a story, it usually must contain suspense (Brewer, 1981), conflict, drama, the unexpected - some sort of structure that constitutes a "plot". One type of theory that attempts to capture how stories are structured is "story grammar" theory (cf. Rumelhart, 1977). In this view, stories are like sentences in that they can be decomposed into constituents, which in turn can be further decomposed, all according to a set of constrained options constituting a "grammar". Thus a story might be decomposed into a Setting plus a series of Episodes plus a Resolution, and an Episode might be further constituted as a Stimulus plus an Action plus an Outcome, and so on. This type of theory has been controversial (Black & Wilensky, 1979; Mandler & Johnson, 1980). I myself agree with the argument put forward recently by Wilensky (1983b). He argues that stories are not grammatical structures in the same sense that sentences are, because the component elements are not textual entities, they are conceptual entities. Thus the same story can be told both in words and on film with quite different surface constituents. Furthermore, it is possible to say of a story that it is peculiar, but it is not possible to say that it is "ill-formed" or ungrammatical in the same sense one can say that about a sentence. Wilensky concludes that the story grammar approach is fundamentally misguided. I will not therefore cover its details here, but will go on to other approaches oriented more toward conceptual rather than syntactic structuring. One such theory is Wilensky's (1983b) own theory of story "points". Characteristically, stories have points. There may be many types of points Wilensky doesn't tell us how many - but one common type involves problems surrounding the pursuit of goals by the main character. There can be goal conflict within an individual, for example, a conflict between self-interest and ethical values; or goal conflict between individuals, as in romantic rivalries; or what Wilensky calls a "goal subsumption termination", wherein an arrangement satisfying many goals for an individual, such as marriage or holding a job, is terminated. To make a point, and therefore a story, out of a goal problem, it is sufficient that the problem be clearly introduced, that its consequences be elaborated so that the reader appreciates that there is no easy solution, and finally a resolution given in which the problem is resolved. These components will not be news to literary analysts, but the importance of Wilensky's system is that each of the point components can in principle be identified by a computer program. (A good deal of effort has gone into programs which identify and make inferences about common goal states [Schank & Abelson, 1977; Wilensky, 1978, 1983a]). Therefore, computers can presumably capture the points of stories. O f course, many stories involve complex interconnections of points. The resolution of one

42

Robert P. Abelson

problem state may create new problem states, for example. What we usually mean by "plot", I think, is a network of simple points in the service of some larger point about points. Again, this kind of analysis is probably achievable by computer, with some qualification to be noted later. There are other AI theories which address plot problems. Dyer (1983) focuses on stories teaching an object lesson about faulty plans. These object lessons are coded into what Dyer calls Thematic Affect Units, or TAUs. These abstract units are very similar to proverbs. For example, one of Dyer's TAUs refers to "locking thebam door afterthe horse isstolen".Init,acharacterisfacedwithathreattoagoal ofmaintainingpossessionofaparticularobject;thereisaplantomeetthethreat,but the plan is not put into effect until afterthe threat has been consummated, so that the object is lost. This could be exemplified by a farmer who doesn't lockhis barn door until after his horse has been stolen, or a scientist who doesn't raise his assistant's salary until the assistant has agreed to take a job elsewhere, and in many other ways. Readers' responsiveness to TAU elements in stories has recently been studied experimentally. Seifert et al. (1986), using semantic priming techniques, have shown that under certain general conditions, a story with a given TAU will tend to remind the reader of other stories with the same TAU. The TAU analysis was developed for use with a story understanding program named B O R I S (Lehnert et al., 1983). Another plot analysis system growing out of that same program is Lehnert's (1982) theory of Plot Units. Plot Units are clusters of Intentions, Actions, and Outcomes simultaneously affecting a pair of characters. For example, the Plot Unit "Broken Promise" consists of a need by Actor A for which help is solicited from Actor B, who agrees, but for his own purposes fails to deliver; thereupon A's need is frustrated but B's ulterior purpose is satisfied. Stories are analyzed as networks of overlapping Plot Units. That the system has relevance for human readers has been shown in an experiment by Reiser, Lehnert & Black (1981). In their study, readers given sets of miniature stories to sort into groups of similar stories tended to put all the Broken Promise stories together, all the Rectified Mistake stories together, and so on, despite considerable variation in surface content. These plot analysis systems all have their strengths and weaknesses. A detailed discussion of these is not so important here as simply to note that they suggest that progress has been made in computer-oriented analysis of plots, and that while the complexities of real plots have by no means been mastered, this is one of the less problematic of the unsolved areas of computer analysis of literature.

Sensory-motor experience While a good deal of progress has been made in the field of robotics, to my knowledge no device with mechanical eyes, arms, hands, and so on has been interconnected with a program whose main purpose is the general understand-

Artificial Intelligence and Literary Appreciation

43

ing of prose. N o text understanding program to date has been able to comprehend references to sensory and motor experiences except by dry semantic summarization of their aspects. Thus a program told that a character had sustained a sudden blow to the back of the head might infer that the character was hurt and would fall and might lose consciousness, that someone with a sinister motive might have done it, etc. But the computer would not have the option of imagining a blow to its own head to experience the "feel" of the action. How important is this? At a minimum it exposes an inefficiency in disembodied semantic analysis. Hundreds of implications have to be stored individually in memory, given the lack of a body with which to replay sensations. For example, in the short story Everything that rises must converge, we are told that after Julian's mother has said twice that it was a mistake to have bought a new hat, Julian "raised his eyes to heaven". How is this phrase to be understood? One could have a lexicon of idioms, which when duly consulted would note that raising the eyes upward is a sign of exasperation. Exasperation in turn would be defined elsewhere as an emotional state with certain interpersonal properties. Alternatively or additionally, a human reader can mentally or even physically roll the eyes upward and note the consequent feeling. Beyond this "big dictionary" problem, it might often be an important aspect of literary appreciation that the reader achieve something deeper than a semantic package connoting a given experience (cf. Spiro, 1982). Among the thousands of examples of powerful literary descriptions of internal sensations, one of my favorites is Hemingway's account in The old man and the sea of how the old fisherman struggles to land his last and greatest fish, hanging on to the line itself for hours with his bare hands. Are we to come away with a mental representation of this story which aggregates the facts that fishing lines cut flesh, that the old man's hands were raw, that he was fatigued and that it was difficult to land the fish, etc., or are we also to feel the line cutting the flesh of our own palms and wince with the old man's pain? The latter, I should think. Score one against the computer...

Moods Related to sensory-motor experience is the question of mood and other emotional experience. Here the same issues are raised again, along with two new ones, one a side issue, the other a main issue. The side issue is that human response to literature might well depend upon the mood of the reader at the time of sitting down to read. Thus if sitting alone on a dark night in a spooky mansion, a reader might feel so apprehensive as to be especially susceptible to every threatening event in a mystery story. I mention this reading context phenomenon because it could in its own right be interesting, but I think it unfair to include in the psychological analysis of literature arbitrary influences on the reader at the time of reading.

44

Robert P. Abelson

The more major issue is that moods by their nature are cumulative. A little nervousness plus a little more nervousness plus a little more may add up to a lot of nervousness. Good literature seems to have this aspect of cumulative impact. In the Everything that rises... story, Julian gets progressively more and more exasperated with his mother, and this makes his peculiarly defiant actions on the bus intelligible. But existing computer programs for story understanding do not usually include provisions for the progressive deepening of moods and impressions. Programs that structure semantic propositions do not in general have counters, as it were, that index the degree of various affective states. Is this a fundamental limitation on computer capability? Well, probably not. The psychiatrist Kenneth Colby has devised a program to simulate the conversation of a paranoid personality named PARRY (Colby, 1975), and it has provision for various degrees of anger, fear, and so on, which in turn color its interpretations of various inputs. This sort of thing from a conversation program could presumably be applied to a text reading program. Admittedly, the PARRY program is very rudimentary and specialized, but it is a beginning.

Aboutness An important aspect of literature is that the surface text may, when analogically or abstractly interpreted, refer to a level of meaning quite removed in manifest content. A story may therefore be "about" something other than what it seems to be about. Thus James Joyce's short stories "The Dubliners" seem to be about the general atmosphere of social paralysis in the life of Dublin, even though each story could be analyzed more simply in terms of goal conflicts within and between particular individuals, a la Wilensky's (1983) theory of story points. Ionesco's play, "Rhinoceros", in which characters one by one turn into rampaging rhinoceroses, is presumably about the snowballing effects of conversions to an evil cause, as in Nazi Germany, even though nothing whatever of this interpretation is given explicitly in the text. How is the reader to know what a story is "about"? As the psychologist Holyoak (1982) points out, stories can be about the human condition as personally or historically recognized; about other stories, which the given story embellishes or satirizes; or parts of a story can serve as commentary on other parts of the same story, so that the text is in some sense about itself. This array of possibilities is bewildering, and it is hard to imagine how a program could search the space of possibilities, much less locate anything relevant from such a search. Of course people may have difficulty, too. This is part of the common lament against modern playwrights such as Albee, Beckett, and Pinter. We wonder what their plays "mean". For the computer, the major limitations are how to cumulate enough background experience so that many meanings would be available at analogical levels, and how to decide or "compute" which possible analogic meaning is

Artificial Intelligence and Literary Appreciation

45

appropriate in a given instance. The computer would have to be familiar with historical incidents, with "personal" experiences, with other stories, and so on, in order to move beyond the textual surface to the analogical level. There is nothing in principle that says the computer couldn't be provided with lots of analogies. The trouble is that there is no natural way to do this, because the computer doesn't have a "life of its own". Oh, yes, we could give it a lot of stories to read, and it is wildly conceivable that ways might be devised for it to relate new stories to old stories it remembered, but it would take a major effort to achieve anything of this sort that would be very interesting. And this brings me to my final set of points. Evaluation Could a computer make a critical evaluation of a piece of literature? Could it conclude that a plot was unimaginative, or a character unbelievable? One senses a certain difficulty here - the computer might not have enough background to do this. Part of what is lacking, surely, is the life experience factor we referred to previously. There is no opportunity for the program to tell itself that it has known people like a character described in the story, but that they behaved in some critical respects differently from the author's character, and therefore that the character is unconvincing. Another problem might be the computer's tendency to take everything it is told too seriously and too literally, all inputs having equal credibility. Unless some provision is made otherwise, a program might be willing to accept any description whatever of the attributes of a character, without questioning them. The philosopher Searle (1980) regards this failing as intrinsic to computer programs, and blames it on their lack of intentionality, that is, the capacity to distinguish truth from falsity, wanting from having, and so on. Among the many rebuttals to Searle, my own (Abelson, 1980) concurs in the accusation that present programs are gullible, but does not regard this as an intrinsic limitation. The trick to overcoming the gullibility limitation lies in the computer science concept of multiple einvironments for representing belief systems (Wilks & Bien, 1983). In a multiple environment, it is possible to distinguish between what an actor believes, what some other actor believes the first actor believes, what the first actor believes about what the second actor believes about what the first actor believes, and so on. Such embedded belief structures have in fact been used in the analysis of fables (Bruce, 1980). In one, for example, a fox tries to trick a rooster to come down from a tree, with a flattering invitation. The rooster pretends to be deceived, suggesting that he has a friend who would also like to come down. When the fox follows the rooster's directions to find the friend, the fox is trapped by a dog whose hiding place had been known to the rooster. In order to understand this fable, the reader must construct a representation of the rooster's belief that the fox will think the rooster trusts the fox's intentions.

46

Robert P. Abelson

What is important in recent analyses of embedded belief structures is not that computers can master fox and rooster fables, of course. In the present context it is the function multiple structures can serve in literary criticism. The developer of the fable analysis indeed understood the necessity for readers of stories to escape a gullible acceptance of stories as facts. Bruce (1980) puts it as follows: "There are, in fact, no 'facts' in a story. Instead, we have sentences produced by an author that are used by a particular reader to manufacture a model of events, a model that includes models presumably manufactured by the characters. The reader must also consider t h a t . . . the author is actively choosing what to say and when to say it. Thus we need to consider the belief spaces of the implied author and of the implied reader as well as those of the characters." To repeat, the fundamental limitation on the computer program as literary critic is not that it couldn't in principle stand at a distance from the text and regard it with varying perspectives, but rather, that those perspectives would not be likely to be very rich because there is no natural way at present to give the computer a sufficient depth of experience. Summing up Let me recapitulate my conclusions for each of the five possible types of limitations on AI appreciation of literature. For the analysis of plots, there are many recent AI developments, giving grounds for optimism that realistic plot representations could in time be constructed by computer. With sensory-motor experience, there are the problems that a big experiential dictionary would be necessary, and that even at that, the computer would not have the benefit of the raw "feel" of the experience. For moods, a similar thing could be said, along with the additional problem that a calculus would need to be provided for the manners in which moods and other internal states cumulate, since the cumulation of small impressions often embodies considerable literary impact. With the problem of aboutness, an overwhelming difficulty looms that the computer is not forseeably likely to have accumulated the depth and variety of experiences to provide analogues that a text might be "about". Thus even the problem of how to choose among several possible analogic interpretations is moot. Finally, the task of critical evaluation is also made difficult by the lack of breadth of experience. However, it does not seem appropriate to criticize the computer as intrinsically unable to construct a model of its own beliefs as distinct from the beliefs promulgated in a story. A common theme runs through all the effective criticisms of computer capabilities. It is that computer programs for text understanding have lacked a background of sensory and "life" experiences to inform their appreciation and evaluation of literary materials. This conclusion seems straightforward enough; what is unclear is whether computers could ever effectively be given such experience.

Artificial Intelligence and Literary Appreciation

47

H e r e , I t h i n k t h e wisest a n s w e r is t h a t w e d o n ' t k n o w . C e r t a i n l y it w o u l d have t o b e a m a j o r p r o j e c t , involving a v a r i e t y o f s e n s o r y and r e g u l a t o r y a t t a c h m e n t s t o the t e x t u n d e r s t a n d e r , a n d an e n v i r o n m e n t in w h i c h the m a c h i n e c o u l d a c c u m u l a t e , learn f r o m , and r e m e m b e r e x p e r i e n c e s . T h e difficulties w o u l d b e e n o r m o u s . O n t h e o t h e r h a n d , it seems g r a t u i t o u s t o dismiss these possibilities as s o m e h o w impossible. W e have been surprised b e f o r e b y m a c h i n e imitations o f h u m a n f u n c t i o n s , and I imagine w e m a y be surprised again.

References Abelson, R. P. 1980

"Searle's Argument is Just a Set of Chinese Symbols", Behavioral and Brain Sciences 3, 424-425. LaBerge, D. and A. Collins (eds.) 1977 Basic Process in Reading: Perception and Comprehension (Hillsdale N . J . : Erlbaum). Black, J. B. and R. Wilensky 1979 "An Evaluation of Story Grammars", Cognitive Science 3, 213 - 2 2 9 . Brewer, W. 1981 "Literary Theory, Rhetoric, Stylistics: Implications for Psychology" in Spiro, Bruce and Brewer (eds.) (1981: 221-239). Bruce, B. 1980 "Analysis of Interacting Plans as a Guide to the Understanding of Story Structure", Poetics 9, 295-311. Carbonell, J. G., Jr. 1978 "Politics, Automated Ideological Reasoning", Cognitive Science 2, 2 7 - 5 2 . Colby, K.M. 1975 Artificial Paranoia: A Computer Simulation of Paranoid Processes (Elmsford, N. Y.: Pergamon Press). Cullingford, R. E. 1978 "Script Application: Computer Understanding of Newspaper Stories" Technical Report No. 116. (Department of Computer Science, Yale University). Dreyfus J . H. 1979 What Computers can't Do (New York: Harper and Row). Dyer, M. G. 1983 "The Role of Affect in Narratives", Cognitive Science 7, 171-190. Holyoak, K.J. 1982 "An Analogical Framework for Literary Interpretation", Poetics 11, 105 -126. De Jong, G. 1979 "Prediction and Substantiation: A New Approach to Natural Language Processing", Cognitive Science 3, 177-212. Lehnen, W. G. 1982 "Plot Units : A Narrative Summarization Strategy" in Lehnert and Ringle (eds.). Lehnert, W. G. and M. H. Ringle (eds.) 1982 Strategies for Natural Language Processing (Hillsdale: Erlbaum). Lehnert, W. G., M. G. Dyer, P.N. Johnson, C . J . Yang, and S. Harley 1983 "BORIS: An In-depth Understander of Narratives", Artificial Intelligence 20, 1. Mandler, J . M. and N. S. Johnson 1980 "On Throwing Out the Baby with the Bathwater: A Reply to Black and Wilensky's Evaluation of Story Grammars", Cognitive Science 4, 305 -312.

48

Robert P. Abelson

Reiser, B. J., W. G. Lehnert, and J. B. Black 1981 Recognizing Thematic Units in Narratives (Paper presented at the Third Annual Meeting of the Cognitive Science Society, Berkeley). Rumelhart, D.E. 1977 "Understanding and Summarizing Brief Stories" in LaBerge and Collins (eds.) (1977: 265-304). Schank, R. C. and R. P. Abelson 1977 Scripts, Plans, Goals, and Understanding (Hillsdale: Erlbaum). Schank, R. C. and C. K. Riesbeck 1981 Inside Computer Understanding: Five Programs Plus Miniatures (Hillsdale: Erlbaum). Searle.J. 1980 "Minds, Brains, and Programs", Behavioral and Brain Sciences 3, 417-457. Seifert, C. M., G. McKoon, R. P. Abelson, and R. Ratcliff 1986 "Memory Connections Between Thematically Similar Episodes", journal of Experimental Psychology: Learning, Memory and Cognition 12, 2. Spiro, R.J. 1982 "Long-term Comprehension: Schema-Based vs. Experiential and Evaluative Understanding", Poetics 11, 77- 86. Spiro, R. J., B. C. Bruce, and W. E Brewer (eds.) 1981 Theoretical Issues in Reading Comprehension: Perspectives from Cognitive Psychology, Linguistics, Artificial Intelligence and Education (Hillsdale: Erlbaum). Wilensky, R. 1978 "Why John Married Mary: Understanding Stories Involving Recurring Goals", Cognitive Science 2, 235 -266. 1983a Planning and Understanding (Reading, MA: Addison-Wesley). 1983b "Story Grammars vs. Story Points", Behavioral and Brain Sciences, 6, 579 - 623. Wilks, Y. and J. Bien 1983 "Beliefs, Points of View, and Multiple Environments", Cognitive Science 7, 95 -119.

ROBERT DE BEAUGRANDE

Schemas for Literary Communication 1. The significance of literature for language psychology 1.1 N o one can deny that the discipline of psychology in America has undergone many fundamental changes in its outlook over the last three decades. Few of the standard axioms of the discipline have gone unrevised. If I had to single out what I view as the most important change, it would be: the recognition of H U M A N P R O C E S S I N G D I S P O S I T I O N S as determinants of behavior and communication. 1.2 In early research, humans were uniformly regarded as 'organisms' whose behavior was directly controlled by 'conditioning' from the environment (e.g. Pavlov, 1927; Skinner, 1938). Manipulations of the environment were considered a sufficient causality to control and explain human behavior. Standard experimental designs paid little attention to people's prior knowledge and predispositions as 'information processors' (to borrow a term no one would have used in that early period). This attitude pervaded research on language as well. Many instruments in the classical arsenal of experimental materials owed their popularity to their ostensible remoteness from relevant communicative contexts - nonsense syllables and word lists. 1.3 This approach was more than just a fashion of the times. It was a reaction to what I would consider the most fundamental problem for language research of any kind. On the one hand, language as an empirical fact is never found in a vacuum, but always in N A T U R A L C O N T E X T S O F C O M M U N I C A T I O N , where the elements and structures of the language as a system are assigned concrete functions to meet the needs of ongoing human communication. On the other hand, a theory of language is obtained by abstracting away from specific contexts to distill out what is typical, general, and repeatable (Hartmann, 1963). Thus, the main problem is: which aspects of context should in fact be leveled or discounted, and which aspects should be retained in a theory? If language categories remain undefined or at least underdefined until put to use in context, is it possible that removing contexts can obscure the essentials of language as a communicative manifestation? 1.4 This question has never been conclusively answered; many theories have barely addressed it at all. The response of language psychology (and of early linguistics as well) was to assume that human contexts were not essential factors for research. The typical exyperimental design was based on L O W - C O N T E X T

50

Robert de Beaugrande

TASKS with no direct resemblance to the H I G H - C O N T E X T PROCESSES of normal human communication and interaction in the everyday world. We now need to clarify the extent to which low-context tasks are in fact able to provide an account of high-context ones. Above all, we should determine the ratio between CONTEXTUALIZATION and COMPLEXITY. Three postulates, which I have tried to represent in Figure 1, are readily evident:

Fig. 1 Context versus complexity

1.4.1 A low-context task is less complex than a high-context one, because contextualizing increases complexity. Each contextual factor would add incrementally to the complexity of the task, so that a minimal context, e. g. working with simple nonsense syllables, should be the easiest. This postulate would probably be favored by traditional language psychologists, because it suggests that their research was conducted on the most auspicious level. 1.4.2 A low-context task is more complex than a high-context one, because contextualizing decreases complexity, thanks to contextual constraints. Contexts show people which of their already rehearsed and expertly skilled processing strategies should be used for the task. Such strategies would probably not be available or decidable for a novel, low-context task. This postulate should be unwelcome to traditional psychologists, because it suggests that they made their own work much more abstruse and difficult than necessary. However, some

Schemas for Literary Communication

51

current researchers in cognitive psychology seem to suggest such a postulate (e.g. Anderson, 1976; Kintsch, 1977). 1.4.3 A low-context task can be either more or less complex than a highcontext one, because contextualizing can either decrease or increase complexity, depending on the familiarity of the task in respect to one's prior knowledge and experience. Contextual constraints simplify a task if they correspond to the constraints of natural communication. But artificial, non-natural constraints, e. g. having to pair a real word with a nonsense word, make the task harder. This postulate recommends an orientation of research toward experiments whose constraints are modeled more closely on the familiar constraints of everyday processing (cf. Neisser, 1976; Spiro, 1977). There would be a major shift of emphasis away from skills or knowledge acquired during the experiment itself over to skills or knowledge acquired during people's prior history as social and psychological agents (cf. 1.2). 1.5 More is at stake here than just rounding out our incomplete picture of human language processing. In the coming years, our concern will be nothing less than the status and validity of all established findings in language psychology. We have inherited a large array of specific findings, some of them quite robust, but no comprehensive theory of discourse processing that could account for the wider significance of the findings, or that could reconcile disparate findings. Psychologists and psycholinguists are naturally language users themselves; as such, they have been decisively affected by their own experiences with language in context. This corpus of experience necessarily creates rich presuppositions about what language is and does. These presuppositions are in turn entailed, often without acknowledgement, in theories and interpretations of experimental findings. Taken seriously, this fact means that all extant theories and findings need to be validated by research on the origin and application of cognitive and communicative predispositions. In the meantime, even seemingly 'hard' findings of tightly controlled experiments may in fact be 'soft', because we can't see the hidden presuppositions and their consequences. 1.6 I shall argue that a comprehensive psychology of literature can be a leading field in this quest. Literary communication is a prime example of a highcontext domain that commands widespread participation in human cultures (3.3). Moreover, I shall try to show that a psychological definition of literature is the only workable kind; hence literary studies should be strongly motivated to clarify its own foundations by participating in this research, once a workable fundamental framework has been provided.

2. The return of the schema 2.1 In recent years, psychology has finally moved away from the view that language comprehension involves merely the 'abstraction of traces' from the input (Gomulicki, 1956; Gibson, 1971) - a descendant the traditional 'stimulus-

52

Robert de Beaugrande

response' approach. Notions that had long been unfashionable, such as the 'schémas' described by Bartlett (1932) and the 'advance organizers' proposed by Ausubel (1960) are now under intense discussion. Properly elaborated, schema theory can be the mainstay of the kind of research program I described in Section 1, i. e., incorporating human processing dispositions and natural contexts into theories and experimental designs. However, before rapid progress can be made, we shall need to address some readily evident problems. 2.2 First, there is no consensus on the quantity of schémas people use. Some researchers seem to envision an enormous number, one for every low-level or high-level structure (e. g. Adams & Collins, 1979). That way, we simply trade the old problem of 'response selection' for an even more intractable new problem of 'schema selection'. Obviously, people couldn't process discourse if they had to spend a lot of time and effort deciding which schémas to apply from among a staggeringly vast store. Other researchers envision a fairly small numer of schémas (e. g. Schank et al., 1975), abstracted from typical sequences of actions, events, and states in human interaction. Yet it is not clear how inclusive the schémas proposed so far might actually be. 2.3 Second, there is not consensus on the specificity of schémas. They might be very general, with all the details left as unbound variables to be supplied in each context. Or, they might be very specific, tailored to each context the person has encountered or expects to encounter. There are disadvantages to both extremes. Unduly general schémas would be too vague to be very helpful in actual use. Unduly specific schémas would trap people in wrong predictions, and make it impractical to deal with innovation. 2.4 Third, there is no consensus on the origin and development of schémas. Surely people don't set up a schema after every experience - that would defeat the whole purpose of schémas to cover broad types of situations. Also, it stands to reason that on at least some occasions, people revise or discard their schémas when the evidence is unfavorable or unaccountable; yet if every bit of adverse evidence led people to discard a schema, once again, the purpose would be defeated. N o schema could be expected to contain or foresee every variation. 2.5 Empirical research on all these issues is extremely pressing. As I said before, we must be concerned not just with doing new experiments on language, but with establishing the fundamental validity of already available findings in the entire discipline of language psychology (1.5). Schémas of some order must be responsible for people's abilities to perform such classical experimental tasks as using 'associations' among items on word-lists, or applying 'tags' or 'cues' during the recall of words. The least desirable conclusion would be that people develop special, unique schémas for being test subjects in psychological experiments. However, this conclusion is quite implausible: people doing word tasks aren't going to forget that they use words in everyday comunication as well. Still, participation in experiments could modify people's processing schémas somewhat to fit the occasion, especially when a large number of trials is involved (e. g. in Shiffrin & Schneider, 1977).

Schemas for Literary Communication

53

2.6 Let's consider for example some standard experimental designs using word lists as language materials. Test persons would study a list of words and then try to recall or recognize as many items as possible. Or, each item was 'paired' with an arbitrarily 'associated' item, and people had to respond to each item with the correct 'associate'. Other times, people were allowed to supply their own responses in 'free association'. These paradigms were popular because subjects were often highly consistent in how they performed on these tasks. For instance, given the word 'needle' on a free association task, 'thread' and 'pin' were extremely common responses, whereas 'pincushion' and 'diligence' were extremely uncomon (Kent & Rosanoff, 1910). Further studies showed that some 'associations' between word pairs are consistently stronger or weaker for most people (e.g. Jenkins, Mink & Russell, 1958; Deese, 1961; Rothkopf & Coke, 1961). 2.7 If people consistently behave this way, then they must have some kind of stored schemas that stipulate what is or is not normally 'associated'. These schemas ought to be active not just during word-list experiments, but also during ordinary communication, e. g. a context involving the topic of 'needlework'. The well-known power of associates to act as distractors during experiments (cf. Anisfeld & Knapp, 1968; Underwood & Freund, 1968) indicates that the schema become automatically activated, whether a person consciously requests them in a single context. We now need to investigate how far the same schemas (or type of schemas) apply to the two settings: the low-context experiment versus the high-context situation in real life. Classical stimulusresponse theorists didn't do much to resolve this question; they measured the level of specific associations among groups of test subjects without offering an elaborated processing theory of what it means for words or concepts to be associated. Schema theorists might derive the 'strength of association' phenomenon from the nature of relevant schemas. The relative frequency of associations across significantly large classes of contexts would be one factor determining the structure of a schema and thereby the strength of associations among its components. For instance, a great many people, including myself, have occasion to use a 'needle' with 'thread'; but only skillful or habitual sewers are likely to own a 'pincushion'; I never had one, and I sewed not with 'diligence', but with clumsiness and resentment. 2.8 Another factor might be the differential links within a schema. Some components could be necessary, some could be merely typical, and some could be purely accidental (Beaugrande, 1980). A necessary association would on the average elicit the most common responses, a typical one the next most common, and an accidental one the least comon. For sewing, a needle is necessarily accompanied by thread; and for some tasks at least, pins must be used to hold the materials in place while they are sewed. But the success of sewing doesn't depend on having a pincushion, or working with diligence. 2.9 Still, an account in terms of frequency or linkage need not by itself yield uniform predictions for observable results. The prior experience of test subjects

54

Robert de Beaugrande

is not likely to be uniform, so careful controls would be needed. Moreover, tests indicate that people in a culture don't always agree about what's necessary or typical, not even for commonplace concepts (Rosch, 1977). Thus, we could have a unified theoretical account of schema application, but a range of variable results from case to case. The processes might be uniform even if the observed outcomes are not. I suspect that such is often the case in literary communication (cf. Section 7). 2.10 In essence, schema theory is being re-aligned, going from a theory of application to a theory of development. The structure of schemas is traced back to the ways people understand and remember their experiences. Here also, it remains to be shown whether the standard experiments can provide and test workable hypotheses. For instance, the 'tagging' models (e. g. Kintsch, 1970; Anderson & Bower, 1972) assume that people already know the words on a test list; what is learned is which words were on a particular list, i. e. which words are 'tagged' in memory as members of that list. In general, then, a given experience would be reconstructed as a set of tagged nodes drawn from the categories shared by prior experiences in one's store of world-knowledge (disregarding the case of very young children). On the other hand, the 'encoding specificity' model (Tulving & Thomson, 1973; Watkins & Tulving, 1975) assumes 'retrieval cues' have to be specifically stored at the time of the learning event. In that case, each individual context becomes the primary determinant, rather than a secondary or derived effect of tagging diverse nodes in a general store of world-knowledge. We could of course attenuate the force of the 'encoding specificity' model by assuming that in every situation, a staggering number of cues contact the human memory, whether or not the person is aware of them or can act upon them (cf. Keele, 1973). What turns out to be relevant for a later context can then be selectively retrieved on demand (cf. Anderson & Pichert, 1978). In this view, schemas can be created at any time, not just taken ready-made from a stable storehouse (cf. Section 7). 2.11 As we know, schema theory has opened up some of its own experimental approaches quite unlike the classical tests with word lists. One of the best known of the new trends is the 'story grammar' approach. 1 Its success and wide recognition has been due to robust findings on the effects of story structure upon comprehension and recall. In particular, if people are told a story that doesn't conform very well to conventional story structure, they tend to recall the story in a more conventional version (cf. Kintsch et al., 1977; Thorndyke, 1977; Mandler, 1978; Stein & Glenn, 1979; Johnson & Mandler, 1980). Hence, these experiments do appear to activate from everyday experience prior knowledge structures (schemas) that directly influence the observable results. 2.12 However, we need to consider how far the stories being used are really representative of the stories people encounter in everyday life. In the studies 1

In early sources, the terms 'story grammar' and 'story schema' were defined in much the same way; more recently, the terms have moved further apart and eventually become designations for rival approaches. See Beaugrande (1982a) for discussion.

Schemas for Literary Communication

55

most often cited, the experimental stories were either invented by the experimenters (e.g. Mandler, 1978; Stein & Glenn, 1979) or adapted by grossly simplifying traditional tales (e.g. Rumelhart, 1975; Mandler & Johnson, 177; Thorndyke, 1977). As I have recently argued at some length (Beaugrande, 1982a), the tendency is to exaggerate the contrast between 'well-formed' stories versus 'ill-formed' ones. I suggested that the most famous 'literary' stories those that survive, like the Arabian Nights or the Decameron - are those which offer stimulating mixtures of confirmations and violations of what people expect (see also Beaugrande & Colby, 1979). If so, these experimental stories are missing some of the dynamics of natural story-telling in respect to precisely the question that matters: how powerful and compelling prior schemas can be, and how departures (innovations) occur. 2.13 I have kindred reservations toward the 'story schema' approach taken by the Yale group. Here, the focus is less on the surface text of a story (as it is in the story-grammar approach), than on the underlying contend, i. e. the classes of states and events. And the main procedure is not to test humans, but to build story-understanding programs for computers. Still, the problems of quality and predictability for the stories themselves have not been fully resolved. In early work, the stories were expressly invented by the investigators, usually to portray the exploits of John and Mary, that prototypical American couple, who, after an extended residence in Cambridge, Massachusettes and Austin, Texas, settled down to life in New Haven, Connecticut, where the most popular activities are apparently going to restaurants and committing acts of violence on one's spouse (cf. Schank et al., 1975; Schank & Abelson, 1977). In later work, stories were taken from newspapers, but still as standardized types, such as reports about automobile accidents (Cullingford, 1978) or terrorist outbreaks (Schank, Lebowitz & Birnbaum, 1978). Nobody has shown that these schemas would work for highly innovative texts, such as the later novels of James Joyce or William Faulkner. 2.14 N o doubt the conditions of pioneering research have encouraged these limitations. The stories tend to get reduced to whatever the theory views as the essentials. As Jean Mandler (personal communication) points out, psychological evidence is easy to gather for the top-level structures in a story grammar, but not for the more specific ones, at least not with the standard experimental procedures. If people know they're going to hear a story, they expect an initiating event, a problem or conflict, a resolution, and an ending; but apparently people differ considerably in how they treat the detailed organization of the story - possibly, a case of unified processes having diverse results (cf. 2.9). Similarly, innovative stories are laborious to anticipate in a working program for computers, where basic, everyday knowledge already makes the programs slow and bulky. Still, the question of what would in fact be involved in making a program for innovative stories is surely worth considering. 2.15 Recently, there has been vehement controversy between advocates of 'story grammars' versus advocates of 'story schemas', for instance, in the last

56

Robert de Beaugrande

1982 issue of the Journal of Pragmatics. The point of contention seems to be whether a 'story' is best described as a type of structure or as a type of content. Psychologically, I consider this dichotomy to have been misleadingly stated. Both views suggest that the text is indisputably there with its inherent (predetermined) structure and content. But that is what we set out to determine, not what we can simply presuppose. A text is not 'there' in any psychologically interesting sense until a human performs some kind of processing action upon it. Thus, a text type such as 'story' can exist as a type only of people routinely apply to it a consistent set of processing schemas that constitute both structure and content (cf. 3.11). If the investigators act like they already know what a story is - e. g. by writing their own samples - they are presupposing what is in fact yet to be determined by the research. In effect, the investigators are relying on their own schemas and pursuing motives that might be quite remote from the schemas used by famous story-tellers (e. g. rating well-formedness over interest; cf. 2.12).

2.16 Empirically at least, it would be very hard to find in spontaneous communication a language structure occurring without content, or content occurring without structure. Even a simple word-list, an atypically loose and low-context structure, cannot cancel the fact that people normally use the words to activate content. Conversely, content cannot be in active storage without some structure, even if it's only the atypically unsupervised chaining of 'free association'. As noted in 2.7, schema activation appears to be largely automatic; any separation between structure and content would probably require an unnaturally intense concentration of attention. 2.17 I submit that the advance of schema theory can also be supported by a comprehensive psychology of literature. In essence, literature is a communicative domain in which certain top-level schemas (which I shall term 'constitutive schemas') control the selection, activation, or formation of lower-level ones (cf. Section 7). Literature offers important advantages scarcely found elsewhere. We have a vast supply of undisputed, naturally occuring samples, and a detailed record of their history and interpretation (3.3). Literary texts persist over time and pervade many otherwise disparate cultures. Thus, literature promises both many challenges and many insights to a language psychology that now recognizes the fundamental role of processing schemas in human communication.

3. Communication and literary communication 3.1 I shall now examine some merits of my contention that by studying the conditions of literary communication, we can broaden our outlook on the complex problems of contexts and schemas in human communication. So far, most psychologists who have used texts rather than word lists or isolated sentences have preferred non-literary samples of narrative and expository dis-

Schemas for Literary Communication

57

course (e.g. Bransford & Johnson, 1973; Meyer, 1975; Kintsch & van Dijk, 1978; Kintsch & Vipond, 1979; cf. the sources cited in 2.11 ff for story research). 3.2 No doubt literary texts tend to appear unmanageably complex, at least in relation to embryonic theories of text processing. But as I suggested in 1.4.3, complexity is a function of one's prior experience with contexts comparable to the context of the task at hand. If we find large numbers of people producing and comprehending literary texts, then we can conclude that people must have for this purpose highly skilled processing schemas that restrict complexity (cf. 7.14). Those schemas deserve investigation as much as the schemas that apply to the learning of word lists (cf. 2.6 f). In all probability, literary communication will impel us to broaden and refine our theories of language processing. 3.3 To begin with, it is striking that literature as an institution appears to be a cultural universal. So far, almost every documented culture seems to have (or have had) a mode of literary discourse, though sometimes with admixtures from other domains, such as ritual or religious discourse. In this sense, literature is simply a given, an empirical fact: a corpus of texts that people in the culture have agreed to consider literary, for whatever motives. 3.4 But it is far from simple to determine what actually qualifies a text as literary, i. e. what properties a text might be assigned in order to increase or decrease its probability of being judged 'literary' in a particular cultural setting. The null hypothesis advanced by some scholars (e. g. Hirsch, 1978), is that there is no text property shared by all literary texts. Literature is simply whatever people think it is and admit to the 'literary canon'. But this hypothesis doesn't explain the fact of literature - it only states the fact and leaves us right where we started. People must have motives for picking out a text as literary. 3.5 Linguistically oriented approaches, notably that of Roman Jakobson (1960, 1968) have sought to define some domains of literature in terms of the distinctive surface features in literary texts. To be recognizable as distinctive, these features would have to be deviant with respect to 'ordinary language' (cf. Levin, 1963, 1965; Koch, 1978). Either the features themselves would be rare, or else the patterns of features. This account has enjoyed some currency in recent years. But I believe that deviant surface features are an epiphenomenon accompanying a more basic factor - a by-product, not a explanatory cause, of literary communication. 3.6 There are two main senses for the concept on 'deviance', a broad one and a narrow one. The narrow sense concerns the highly salient and unconventional patterns typically found in poetic discourse, such as that of e. e. cummings, or in the prose works of Gertrude Stein and the later James Joyce. However, scholars would hesitate to argue that the 'most deviant' text is also the 'most literary' text, and for at least two good reasons. One reason is that many texts we recognize as 'literary' can hardly be called 'deviant' in the narrow sense, e. g. the novels of Theodor Fontane or Ernest Hemingway. The other reason is that similar

58

Robert de Beaugrande

'deviant' surface features can and do appear in texts people don't consider literary, such as advertisements. 3.7 In the broad sense, 'deviance' includes any distinctive individual touch of style (Enkvist, 1973). At least, this sense doesn't exclude recognized 'literary' texts. But it fails emphatically as a sufficient criterion, since stylistic specifics are certainly not limited to 'literary texts'. Moreover, 'deviance' threatens to become the standard state of affairs - hardly something that could count as 'deviant' by definition. Most texts are distinctive in at least some ways, so which ones are to be judged 'literary?' 3.8 Another possible account would be to define the 'literary text' in terms of its content - the topics it deals with. For example, cavalier, pastoral, and sacred poetry, such as that of Michael Drayton, John Donne, and George Herbert, respectively, each focuses on specific distinctive topics. But literary content is an even more obvious epiphenomenon than surface features. For one thing, the same topics appear in non-literary texts, such as festive speeches and church sermons. Besides, many acclaimed literary texts specifically address topics few of us would consider 'inherently literary', witness historical dramas and naturalist novellas. Apparently, there is something in the particular treatment of these topics that invites us to consider the text a literary work. 3.9 Thus, the two commonplace accounts of literature - as a type of texts being assigned distinctive surface features or distinctive content - are neither necessary nor sufficient conditions for 'literariness'. They fail to distinguish between literary and non-literary, so they can at best be CUES people use to postulate a particular mode of discourse. I shall offer some evidence for this contention later on. 3.10 I surmise that here too, the problem is aggravated by trying to define a text type according to independent text properties. As I stressed in 2.15, there are no text properties in a vacuum: someone has to process the text and constitute the properties. We tend to attribute to 'literary text properties' an independent existence because we are habituated toward constituting them, and because people in the same culture - particularly specialists - tend to agree about the 'literary properties' they constitute. But if text properties are a result, not a cause, they cannot serve to define or explain what makes something a 'literary work'. Our account must be in terms of how and why these properties are constituted. 3.11 Therefore, the solution of adopting a psychological definition should be preferred (cf. 1.6). My proposal is to define literature as a communicative domain dominated by certain constitutive processing schemas (2.17; 3.13; 7.2 ff). For literature, the most powerful and high-level schema might be called ALTERNATIVITY. By this I mean that the participants in literary communication are free and willing to contemplate other worlds besides the accepted 'real world' (Iser, 1978). Now I don't mean that the text must be 'fictional' in the sense that people agree it doesn't give an account of things that actually happened at some historical time and place. At best, the non-historical aspect is

Schemas for Literary Communication

59

one possible cue for the literary experience, but by no means a prerequisite. We find literary works covering a wide range from the fantastic to the historical to the documentary (cf. Section 4). 3.12 In POETIC communication, the alternativity principle is extended to the negotiability of language in discourse. Once again, there is no requirement that this language must be obviously 'deviant' from ordinary discourse (cf. 3.6 f). Some poets, such as W.H. Auden or Carl Sandburg, deliberately wrote poems after the model of everyday language. However, to process poetic discourse, we must appreciate how language itself is organized, and how it can be reorganized for particular motives. The poet has that option of reorganizing discourse, no matter how little or how much that option is utilized in a particular case. Of course, the poetic function is not limited to poetry, though the latter is the most common and important genre for it. We can find poetic alternativity in the language of folktales, dramas, novels, and so forth, provided we are willing to process it accordingly (cf. Section 5). 3.13 In the most general perspective, alternativity DE-AUTOMATIZES text processing, at least to some degree. Mukarovsky (1964) suggests that it is the function of deviant language to block our normally 'automatic' processing and to force us to perceive and deal directly with the specific occurrences we find in the text at hand. But my contention is that alternativity is a CONSTITUTIVE SCHEMA - that is, a schema that defines the domain by controlling the selection and application of other schemas2 - which, in the act of its acceptance, shifts the routine boundary between automatic versus attentional processing. Specialized text features and text content act as cues and reminders, but they are not necessary. We can increase attention and thereby alter processing of any text by registering it as one among several alternative ways of organizing worlds and discourses.

DOCUMENTARY

• FANTASTIC

HIGHLY DEVIANT LANGUAGE < Fig. 2

• EVERYDAY LANGUAGE

Range of literature and poetry

3.14 We accordingly have the scheme shown in Figure 2. The 'literary' ranges on the content parameter from documentary to fantastic, with all possible degrees or mixtures in between. The poetic ranges on the language parameter from everyday to highly deviant, again with all possible degrees or mixtures in between. If we grant that alternativity is the constitutive schema of the 2

In the sense of Searle (1969:33ff), a 'constitutive' schema would be one whose application defines the domain.

60

Robert de Beaugrande

enterprise, then such wide ranges are exactly what we should predict. If the literary or the poetic were to be fixed at just one end or one point on these parameters, the potential for other alternatives would be defeated. And in fact, I think critics tend to consider literature or poetry 'bad' if it accepts its own conventions as laws and final goals, and abandons all search for further possibilities. Common critical terms such as 'trite', 'trivial', 'unoriginal', 'derivative', and 'epigonic', all suggest this view. 3.15 This psychological definition of literature and poetry as determined by the alternativity schema - and by a shift away from automatic toward attentional processing - at least has the merit of escaping the problems created by the traditional definitions in terms of surface features or content. Those definitions are not 'objective' in any case, since we must admit sooner or later that features and content are themselves outcomes of psychological processes. The question is not longer: 'What features does the literary text have?', but rather: 'What schemas do text producers or receivers apply in order to constitute literary texts from linguistic artifacts?' 4. Fictionality and historicity in literary

communication

4.1 A literary text may be called 'historical' in at least three senses: 4.1.1 The content assigned to the text may include historical elements, i. e. elements believed to have actually occurred at some earlier time in history. 4.1.2 The text as an artifact can persist in historical time, even when the culture wherein it was originally produced has been subjected to change or has disappeared entirely. 4.1.3 The text as an artifact was produced at some point in historical time. This point sets the limits of what conditions can have obtained before or during production.3 4.2 The first sense brings up the complex opposition of FICTIONALITY versus FACTUALITY. Nearly all works contain at least some elements that have correlates in the accepted model of reality. Even the fantastic world of Shakespeare's Tempest has a 'Duke of Milan' and a 'King of Naples'. Measure for Measure is set in Vienna (though Shakespeare gave the main characters Italian names and mentioned no landmarks of the city); and the basic incident of a governor threatening a sister with her brother's execution was reported in 1547 as a recent historical event. Apparently, factuality offers one source among many for building a literary world (4.5, 11). 4.3 FANTASTIC LITERATURE is an extreme end of the scale. According 3

Of course, text reception is also a historical act, though much less attention has been devoted to this fact than to the historicity of text production (authorship). Literary history would have to be completely rewritten from the standpoint of the reader (cf. Jauss, 1970; Weinrich, 1971). Psychological evidence for historically remote acts of reading can be only indirectly gathered.

Schemas for Literary Communication

61

to Todorov (1970), the fantastic presents deliberately unreal and unbelievable ('supernatural') events, such as those we find in Walpole's Castle of Otranto which is also decorated with historical elements (cf. Lukács, 1956: 11). A realistic explanation may be eventually given, but it is a manifestly unsatisfactory one in comparison to the fantastic explanation based on supernatural forces. Even so, the supernatural is created from the natural by reorganizing commonplace principles of reality, such as cause versus effect, or relaxing the oppositions between existence versus non-existence, and life versus death. 'Magic' is a means for rearranging causes and effects from what would be natural in the real world. For instance, the common magic trick of making something materialize out of nothing defies the basic law of the conservation of matter. All the same, a magic spell doesn't have random effects; it simply introduces a new version of causality. The principle of causality as such remains in force, because it is needed for the processing of the narrative pattern (cf. Beaugrande & Colby, 1979). In general, fantastic events would be meaningless if they could not be understood in terms of their analogy to realistic events. 4.4 Along the parameter between documentation and fantasy, there are many gradations and forms of HISTORICAL LITERATURE, in which certain elements of actual historical periods are used for literary purposes. Lukács (1956) contends that prior to Walter Scott, those elements were deployed in an egregiously non-historical way, namely as superficial props for a culture and cast of characters taken from the author's own times. But Lukács seems determined to equate historicity with a particular kind of realism he favors the critical middle-class 19th century novel). The fact remains that historical elements, however inaccurate, appear very frequently in literary worlds, ranging from ancient times (e. g. the Trojan War in the Iliad) to the present (e.g. the American anti-communist hysteria of the 1950's in Kurt Vonnegut's novel Jailbird). These elements always had an important function. They provided a set of schemas for plot and characters, without committing the author to following any sources faithfully. The fact that these schemas can be shown inaccurate via modern historiography (cf. 4.6 ff) doesn't disqualify the works as literature in the slightest. There is no 'factual' way to represent the Trojan Wars in any detail, because we simply have no eyewitness records of scenes or dialogues that took place there. Literature cannot be 'historical' in the same sense that 'history' is, no matter how much research an author like Walter Scott or Peter Weiss may conduct. Important differences remain: 4.4.1 The two domains differ in focus. The literary representation is by no means obliged to select or focus on the same types of elements that stand out in a primarily historical account. For instance, a history of England in the times of Henry IV wouldn't focus on Falstaff and his followers in any proportion like the one we have in Shakespeare's dramatization. 4.4.2 The two domains use diverse methods of validation. A historian is expected to make use of all the best sources and to adjudicate their merits in cases of discrepancy. A new discovery can make an established work suddenly

62

Robert de Beaugrande

obsolete. A literary author has no such obligations to be thorough and adjudicative. The principle of alternativity gives the literary author the license to abridge available facts and to invent fresh 'facts' that better serve the literary purpose. Also, a new historical discovery rarely makes a literary work obsolete, though it may alter our perception of how the work originated or was first received (historicity in the sense of 4.1.3). 4.4.3 The two domains have different values. A historical document is normally rated by the accuracy of its record with respect to the 'facts' as we know them from other sources. A 'bad' text is an inaccurate one, according to our own modern standards of historiography. But 'bad' literature is hardly the works that fail on this account (cf. 4.9). 4.5 As a Marxist critic, Lukacs was especially concerned with the historical and social preconditions that determined history. He claims that this concern didn't appear in literature until after the French Revolution, and even then, it was absent from many works, such as the novels of Adalbert Stifter. In the psychological definition I am propounding here, all literature is necessarily concerned with the pre-conditions for reality, though 'reality' need not be represented by borrowing from historical sources. In the 19th century, historiography was finally moving outside the control of the church and the aristocracy and entering the sphere of mass education. Hence, it was only natural when literary authors adopted historicity more and more as a SUBSIDIARY SCHEMA for literary communication. Still, that schema was able to coexist with competing schemas such as fantasy precisely because literature is a distinct communicative domain from historiography. Literature added new sources and methods without giving up the old ones. Historiography, on the other hand, cannot keep its old methods once it has created new ones. King Alfonso el Sabio of Spain commissioned a history of the world that included Greek myths from Ovid's Metamorphoses. Today, we draw the borders of history differently and cannot go back to that, though we still take Ovid's myths seriously as literature. 4.6 Whether or not Elizabethan audiences notices or minded it. Shakespeare's historical dramas are typically inaccurate as history. For instance, the second part of King Henry the Sixth has young Richard of York distinguishing himself in the battle of St. Albans, though in fact Richard was only two years old at that time (May 1455). In King Henry the Eighth, Queen Catherine dies before Elizabeth is born, though actually that birth came three years before the Queen's death - a fact people might have remembered when the play was presented some eighty years later (June 1613). But these inaccuracies are motivated by literary considerations. Richard's prowess as a soldier foreshadows his later violent rise to power. The birth of Elizabeth makes a dramatically effective 'happy ending', reinforced by Cranmer's eloquent prophesy of her greatness as future queen (V, v, 18 - 55). The death of the pitiable cast-off Queen Catherine (whose cause the play supports rather strongly) simply wouldn't have done as bright ending.

Schemas for Literary Communication

63

4.7 The convention of prophecy, though a powerful dramatical device, is of course a facile violation of historicity. Writing in hindsight makes it easy to foreshadow things that, in the chronology of the literary world, have not yet happened. For example, Shakespeare devised prophetic speeches in the third part of King Henry the Sixth, which, like the business at St. Albans, foreshadow, Richard's bloody career on the way to becoming King Richard III. Henry himself says (V, vi, 3 7 - 4 2 ) : (1)

And thus I prophesy, that many a thousand Which now mistrust no parcel of my fear, And many a old man's sigh and many a widow's, And many an orphan's water-standing eye Men for their sons, wives for their husbands, And orphans for their parents' timeless death Shall rue the hour that ever thou wast born.

And after murdering Henry, Richard takes up a similar theme: (2)

The midwife wonder'd and the women cried ' O Jesus bless us! He is born with teeth!' And so I was; which plainly signified That I should snarl and bite and play the dog. Then, since the heavens have shap'd my body so, Let hell make crook'd my mind to answer is.

Such speeches build Richard up to be the convincing and formidable assassin that Shakespeare himself would later portray in the drama Richard III. 4.8 Elizabethan audiences must have wanted to have it both ways. The historical dramas were decked out with people and events of the factual past, but also with calculatedly improbable elements, such as prophecies, ghosts, and a baby being born with teeth. On the one hand, Henry the Eighth was presented in 1613 with the title All is True, though it was manifestly 'untrue' in its chronology, and at least some people still living then probably remembered Henry's reign. And Henry the ft/i/?, Canterbury's speech on Henry's claim to the French throne (I, ii, 35-100) is a versification of the same speech in the historiographic source, Holinshed's Chronicle - an unusually direct adoption of a historical source. On the other hand, the audience is expressly enlisted in the same play as co-creators of an imaginary world with a radically displaced chronology: (3)

Let us, ciphers to this great accompt, On your imaginary forces work. Suppose with the girdle of these walls Are now confin'd two mighty monarchies, Whose high-upreared and abutting fronts The perilous narrow ocean parts asunder. Piece out our imperfections with your thoughts [ . . . ] For 'tis your thoughts that now must deck our kings, Carry them here and there, jumping o'er time, Turning th'accomplishment of many years Into an hourglass.

64

Robert de Beaugrande

Evidently, the literary juxtaposition of the documentary and the fantastic appeared to the Elizabethan mind not discrepant, but dramatic. And as a literary strategy, it fits the principle of alternativity very well. The tension between disparate domains encourages an audience to constitute multiple versions of reality. 4.9 To be sure, readers of a quasi-historical work are prone to wonder if things 'really happened that way'. The end purpose and motive of any alternative world is its relation to what is construed to be the 'real world'. The question of factuality naturally enters here, but is not the decisive question, and a literary experience often concludes without attempting to answer it. I think we are dealing with two divergent notions of 'truth'. The 'truth' of literature is never the 'truth' of discourse types such as eyewitness testimony in a court of law. Literature offers us not just veridical details of what actually happened at a specific time and place, but encompassing insights into what is generally true of the human condition. 4.10 In consequence, the 'literary fiction' differs decisively from the 'ordinary lie'. Both cases involve statements that depart from what would be judged 'true' by an actual eyewitness. But the 'literary fiction' is intended to bring us round to a higher-order 'truth' that is fairly immune to variations of circumstance, such as time and place. This outcome compensates us for suspending our insistence on accuracy. The 'ordinary lie', on the other hand, is intended to supplant and hence conceal the truth; it varies circumstances expressly to falsify them, and it is hence of the same order as its denied counterpart, not of a higher order. There is no compensation for accepting the ordinary lie; we reject the lie altogether when we recognize it as such, because it impoverishes, not enriches, our awareness. Thus, we clearly distinguish between the author's act of composing a mythical play like King Lear or Cymbeline versus the characters' (Edmund's and Iachimo's) acts of lying within those plays. 4.11 In recent decades, D O C U M E N T A R Y LITERATURE has undertaken a radical treatment of the problem of fictionality. The author gets the most authentic sources possible and presents them with minimal alterations. For example, Die Ermittlung (The Investigation) by Peter Weiss and DerHund des Generals (The General's Dog) by Heinar Kipphardt were both produced from court proceedings against war criminals from Nazi Germany. The sources made it possible to use the exact words of the historical personages. But even this degree of historicity is not being presented as the reality itself. The end or purpose is not to give out facts, but to make us contemplate the facts and understand how and why they happened, and how they could have been altered by other human conditions, for example, less blind obedience toward unscrupulous authorities. Simply reporting the reality of war crimes in Nazi Germany is not a literary enterprise; gaining insights into that reality and envisioning more humane alternatives is. We must realize that our procedures for constituting and understanding the world (any world) are ultimately negotiable. Hence, even extreme historicity can be readily assimilated as a subsidiary schema for literary communication (4.5).

Schemas for Literary Communication

65

4.12 The second sense of historicity in literature (4.1.2) confronts us with the extraordinary persistence of literary texts over historical time. Often, the best-known documents of a vanished culture are its literary artifacts. Of all the texts preserved from Ancient Greece, none have been read and respected more than the plays of Aeschylus and Sophocles, or the epic poems of Homer. We know the verses of the ancients far better than their histories, war chronicles, and philosophies - and this even though the writers of the latter often cultivatted a style quite close to the literary, witness Lucretius's natural history De rerum natura that expounds atomic theory in hexameters, with original metaphors coined to convey the abstract Greek terms of Epicurus. 4.13 Similarly, no non-literary documents of their time can compare in popularity with Beowulf, Das Nibelungenlied, The Divine Comedy, The Rubaiyat, The Canterbury Tales, and The Faerie Queen. Indeed, many historical personages and events are remembered at all only because they figured in literary texts such as these. Their 'factual existence' proved insignificant alongside their 'literary existence' at least in respect to their long-range presence in public awareness. 4.14 Psychologically, the historical persistence of literature and literary communication is a special challenge to schema theory. This theory assumes that text processing routinely falls back on prior expectations based on shared cultural knowledge and experience. But we can process a work such as The Iliad even though the entire culture which first received it has passed away. Our own culture has more in common with Tennyson's England than with Spenser's; but we can read the latter's Faerie Queen with no less insight than the former's Idylls of the King, and we don't necessarily think the 19th-century work is better than the 16th-century one. Tennyson's Victorian view of courtly conduct radically updates Spencer's Elizabethan view. By being closer to us, however, Tennyson is much further from the age of courtly society. And we don't in fact require this kind of mediation. We can somehow process the older epic even though we are not situated in Spenser's culture, let alone in mediaeval courtly culture itself. 4.15 Here, we have further supporting evidence for the notion of alternativity. We are participating in literary communication by constituting a world which differs from our own model of reality not only via the opposition of fact versus fiction, buy also via the opposition of our own present versus the past of the literary work. Once alternativity is accepted as a constitutive schema for processing, historical remoteness is no longer a significant impediment. We postulate, however simplistically, certain general human truths impervious to the passage of time (cf. 4.9). 4.16 The alternativity of poetic communication is also crucial. The language may have undergone historical changes that disturb comprehension. Yet we can resolve such cases by fitting the obsolete phrase to the world we are reconstructing. Consider these lines from Shakespeare's Measure for Measure (III, i, 94-97):

66 (4)

Robert de Beaugrande Claudio: The prenzie Angelo? Isabella: O , 'tis the cunning livery of hell The damned'st body to invest and cover In prenzie guards!

As far as I know, the word 'prenzie' is not recorded in any other extant sources; it certainly is not a word any 20th century reader would know. But given Angelo's situation in the drama, we can understand the word to mean 'strict', 'prudish' - just the most 'cunning livery' to conceal a licentious character. Having created a schema for the dramatic confrontations in the play, we can oversome historical language barriers. We can resolve the problem of linguistic remoteness and change via the same poetic function that we apply to comprehending the metaphors of a famous passage like this one from the same scene (III, i, 118-128): (5)

But to die and go we know not where; To lie in cold obstruction and to rot; This sensible warm motion to become A kneaded clod; and the delighted spirit To bathe in fiery floods, or to reside In thrilling region of thick-ribbed ice; To be imprisoned in the viewless winds, And blown with restless violence round about The pendant world; or to be worse than worst O f those that lawless and incertain thought Imagines howling!

From the standpoint of modern usage, the rarest words are 'pendant' (i.e., 'suspended in space') and 'viewless' (i. e., 'unseen'). It would indeed be hard to use the word 'viewless' in that sense without reminding people of this same passage, e. g. in Alexander Pope's phrase 'light as the viewless air' (Odyssey, VI, 25). But several words we still commonly use evidently require a different interpretation here than what we would normally mean by them today: 'obstruction' ('stagnation'), 'sensible' ('endowed with the senses'), and 'thrilling' ('sharp', 'painful').4 Since we agree to reorganize discourse as soon as we enter poetic communication, and shift processing toward higher attention levels (3.13,15) we are well prepared to impose coherence on little-known words or fit familiar words to unfamiliar meanings. In other discourse domains, we would expect more standardized word and meanings. 4.17 In the third sense (4.1.3), literature is historical in the act of authorship. Considerable acumen was expended on trying to establish which parts of a play were or were not authored by Shakespeare, or whether ostensibly ancient works such as Ossian's Fingal and Tamora were authentic relics of early periods like the third century after Christ. But here too, literariness brings different standards than does historicity. When I watch an actual performance of Henry VIII, 4

Compare the contemporary usage in Joshua Sylvester's tales of Du Bartas (1605): 'This shaft shall thrill the foes that him assail'.

Scheraas for Literary Communication

67

for instance, my response to specific scenes or passages is not to ponder whether they might be the work of Fletcher, Beaumont, or Massinger, or of some collaboration among them. Instead, I am absorbed in reconstituting the world of Henry VIII and his followers as a unified whole. Or, if it occurs to me to think that the prologue to Pericles, Prince of Tyre was not Shakespeare's, that is because I find the verses forced and ungraceful, and the content trite: (6)

Bad child; worse father! To entice his own To evil should be done by none.

But I am using the issue of historical authorship as a pretext to define my purely literary response. Perhaps I have simply created a author-persona as a literary construct to reflect my own values. If I were a historian, I would probably have to concede the possibility that Shakespeare might have written such doggerel for some reason, e. g. to imitate the archaic diction of John Gower, who supplied the story and occasionally some specific phrases for the play. 4.18 A similar line of reasoning applies to the notion of the author's intention, traditionally construed as the ultimate and conclusive arbiter of literary meanings (e.g. by Hirsch, 1967). The alternatives created during literary and poetic communication need not be the same alternatives entertained by a historical author at a particular time. I might like to imagine that I am uncovering special meanings that Shakespeare thought of; but since the work itself is the most important basis for my hypotheses - sometimes the only basis available - 1 can justify my hypotheses solely by assuming that my own activities as a literary participant belong to the same domain and follow the same general schemas as the author's. For all I can prove to the contrary, I am taking his Shakespeare's cues and doing very different things with them than what he and his contemporaries did. I can simply hope I am in the right domain and using appropriate procedures. Like authorship, authorial intention readily becomes our own construct to account for our responses and values. 4.19 I would submit that these three aspects of the historicity of literature (4.1.1-3) all force us to recognize the central role played by alternativity in literary communication. This mind-set equips readers to rearrange chronologies, overcome shifts in language, and apply cues to probe an author's meanings and purposes. In each case, these activities are specifically literary in their consequences, not historiographic, as compared to the ways a historian would read non-literary samples, e. g. legal and administrative discourse preserved from the times of Dante, Marlowe, or Goethe. The ostensible factuality of those discourses discourages people from expecting to derive from them general truths about humanity. Still, nothing actually prevents us from reading samples originating in non-literary discourse with our literary schemas. By doing so, we transpose the text as artifact over to another communicative domain. In the next section, I shall consider a literary type in which precisely that transposition is made into a powerful subsidiary schema.

68

5.

Robert de Beaugrande

'Foundpoetry'

5.1 A significant and difficult issue is the phenomenon of so-called 'found poetry'. These are texts which the original producer did not intend to be poetic, but which are presented and accepted as poems. For example, found poems may be taken from newspaper items or public signs. The following one was published by Ronald Gross in 1967: (7)

Yield. N o Parking Unlawful to Pass. Wait for Green Light. Yield. Stop. Narrow Bridge. Merging Traffic Ahead. Yield. Yield.

All these phrases are found on traffic signs in the United States. Put together in this way, the effect - according to my students who discussed the poem - is a movement though a narrow and limiting world where you are enjoined to 'yield', i. e. 'surrender'. These signs are all negative, telling you things that slow you down and give the right of way to others. The repetition of the word 'Yield' is on the one hand modeled on the common poetic device of the 'refrain', and on the other hand suggests the linear progress along a road from one intersection to another. Plainly, this mode of presenting the signs is not simply a statement that roads have these signs; nor is it the same experience as seeing the signs in succession along a real road, where they have non-literary consequences, such as stepping on the brakes. Ronald Gross's own contribution was to display at one time two otherwise independent conventions of linear ordering in seemingly unrelated discourse types, poems and road signs. The 'poetic' experience is precisely to confront and resolve this tension between apparent incompatibilities — to fuse together disparate schemas for linearity. 5.2 If successful, the found poem offers insights on two sides. On one side, the possibility of aesthetic perception is introduced into domains of everyday language. Developing a keen awareness of that possibility can heighten the intensity and diversity of a person's overall communicative processes. But on the other side, specialized 'poetic' conventions can be rediscovered and hence re-motivated. In an interview with The New York Times Book Review (June 11, 1967), Gross observed: As I worked with labels, tax forms, commercials, contracts, pin-up captions, obituaries, and the like, I soon myself rediscovering all the traditional verse forms in found materials: ode, sonnet, epigram, haiku, free verse. Such shapes made me realize that these forms are not mere artifices, but shapes that language naturally takes when carrying powerful thoughts or feelings.

Schemas for Literary Communication

69

Gross suggests that the poetic function which originally engendered 'traditional verse forms' grew directly out of utilizing language tendencies already inherent in the 'shapes' of everyday discourse, at least when 'powerful thoughts or feelings' are being expressed. 5.3 Only if we define 'literary' and 'poetic' in terms of processing schemas can we resolve the problem of how a text can be read as a poem when it was not written as a poem. The historical act of text production and the psychology of its intention lose their dominance as determining factors (cf. 4.17f). William Whewell's (1819) textbook Elementary Treatise of Mechanics contains a sentence which, if broken up into lines, reads as follows: (8)

And so no force, however great, Can stretch a cord, however fine, Into a horizontal line Which shall be absolutely straight.

Whether the author deliberately planted this quatrain, or whether he did so accidentally because he was familiar with verse forms, is not the essential issue. If we are prepared to process the text as a poem, it is, in that event, ipso facto poetic in the sense I have defined. 5.4 The found poem is a decisive counter-example for any theory defining poetry by reference to its surface features. The point of the found poem is precisely that the same text with the same features has been moved from nonpoetic over to poetic. At most, we have altered the spacing and the line divisions; and it would be very naive indeed to define poetry as 'a text divided into short lines'. Spacing and line divisions can only be cues we use to identify a poem; they are neither sufficient nor necessary conditions for poetic communication — just the conclusion stated in 3.9. 5.5. This contention can be supported with the findings of a study I did. I expressly took a rather unpromising passage from a biology textbook and divided it up with the typical spacing and line divisions of a poem. In this case, my own intention was, so to speak, 'poetic', because I myself was prepared to regard the text as a poem. Here is the result: (9)

T H E STICKLEBACK THIS is certainly the most sagacoius of the Lilliputian vertebrates: Scarcelly more than an inch in length when full grown It gazes at you With large keen shining-rimmed eyes Takes your measure And dans off with a flirt of the tail That says plainly: Catch me If you can. The sticklebacks are delightful aquarium pets because Their natural home is in still water Sufficiently stagnant

70

Robert de Beaugrande For algae to grow luxuriously Thus we seldom need To change the water in the aquarium Which however should be Well-stocked with water plants And have gravel at the bottom.

The passage appears to have been written for adolescents. Its mode of presentation is not stiffly 'scientific': there are some metaphors, such as 'a flirt of the tail that says plainly'; and the fish is claimed to be sending us a message in English. 5.6. A group of students that were near the end of an introductory poetry course were asked to 'interpret this poem'. I found that neither the line spacing nor the occasional metaphors sufficed to make the text a credible poem for most test persons. One student said the poem was about children and their teasing; another account envisioned a metaphor for a competitive society that worships high-speed car racing. But nearly all the students decided the text meant no more than it would in everyday discourse - in my terms, that there was no alternativity worth pursuing. Here's a typical response; note that the respondent does cite one text passage having more 'personality' (cf. 8.14; 9.1): (10)

I do not like this because it looks like something out of an encyclopedia that has been put into verse, with very little personality added ('Catch Me If you Can'). It could also be a pamphlet that the pet store gives you when you buy a stickleback. It is not poetry.

These findings prove at least that line formatting alone doesn't suffice to convince people to accept a text as a poem. The content overruled the surface format in this case. Most decisive I think was that competing text types immediately suggested themselves as more plausible, witness our sample response: 'encyclopedia' and 'pet store pamphlet'. Neither of these text types is usually processed according to the mode I have called 'alternativity'. That competition probably inhibited students from pursuing alternative meanings, for instance, that the fish is acting as an allegory for territorial instincts in nature and human beings - a 'general truth' of the literary kind (4.9). 5.7 More than any other case, the 'found poem' shows how fuzzy the borderline can be between the literary and the non-literary. In theory at least, any text might be subjected to this schematic transfer. In practice however, what is likely to be successfully transferred should depend on a range of contributing factors. A 'found poem' should be more likely to be processed according to literary and poetic schemas if: 5.7.1 it is included in a collection together with 'non-found' poems by wellknown authors like Wordsworth; 5.7.2 it coincidentally allows for subsidiary features of typical poetry, such as the refrain in (7) or the rime in (8); 5.7.3 its content is not easy to relate to competing text types; 5.7.4 the content is easy to assign an ambiguous or polyvalent interpretation.

Schemas for Literary Communication

71

5.8 So far, we haven't tried to sort these factors out experimentally, but that could readily be done in the future. The sample texts would be 'found' poems systematically varied in respect to the factors listed in 5.7.1-4. Experienced literary scholars could be asked to rate the texts and their consensus would be the basis for rank-ordering the samples and making predictions about the degree of acceptability as poems. Test groups of various kinds could be given tasks such as interpreting or evaluating the samples. The responses could show whether the factors I listed are indeed the relevant variables. 6. Super-coherence as a goal for literary schema application 6.1 One reason why most people do not often feel motivated to read old texts they consider non-literary may be the lesser degree of coherence obtained from the experience. Schema theorists concur that the application of a schema provides a necessary framework for imposing coherence on what is actually said in communication. For instance, if we are reading about a 'rocket', we expect to learn that it will 'take o f f ; if such things as 'a roar and a burst of flame' are mentioned, we assume the 'take-off' must be the cause (Beaugrande, 1980). Hence, even if the text itself seems incomplete or badly organized, people make sense out of it by applying relevant schemas (Beaugrande, 1979a). 6.2 Alternativity may offer unlimited freedom in theory, but it never does in practice. The text is always a SYSTEM, and what you can do at one point is constrained by what happened (or will happen) at another point (Beaugrande, 1980). Disturbances of coherence elicit compensatory processing actions. The same top-level processes whereby people constitute a literary 'text-world' 5 also make sure that such a 'world' is coherent. When the author has designed the text to offer substantial resistance - as in T. S. Eliot's Wasteland, and Joyce's Finnegan's Wake - readers probably won't just abandon their intention to impose coherence. If they feel the trouble is worthwhile, they will to look very hard for means of establishing it; if they feel otherwise, they just won't read any further. They will not merely read on and on without any regard for coherence. 6.3 Now, the historical persistence of literary texts over non-literary texts suggests that this imposition of coherence must be exceptionally intense in literary communication. Somehow literary texts appear to encourage the construction of appropriate contexts (cf. Halliday, Mcintosh, & Strevens, 1964). My explanation would be that to read a text not just for specific events and situations, but for general truths about the human condition, requires us to concentrate on a higher order of coherence (cf. 4.9f). In other words, we make the text super-coherent by finding general truths that subsume and motivate its detailed contents, even when the latter seem to oppose coherence. For example, 5

The 'text-world' is the totality of conceptual meaning assigned to a text; most notions of text meaning in research are 'models' of this 'text-world' (Beaugrande, 1980).

72

Robert de Beaugrande

Hamlet undertakes to appear mad by uttering incoherent speeches (III, ii, 391-400): (11)

Polonius: Hamlet: Polonius: Hamlet: Polonius: Hamlet: Polonius: Hamlet:

My lord, the Queen would speak with you, and presently. Do you see yonder cloud that's almost in the shape of a camel? By the mass, and 'tis like a camel indeed. Methinks it is like a weasel. It is back'd like a weasel. O r like a whale. Very like a whale. Then I will come to my mother by-and-by.

Hamlet violates the principle of consistency by saying that one object is three different things at once; and the principle of relevance by introducing cloud shapes as motives to go visit his mother. The other characters in the play entreat Hamlet to speak more coherently, as in (III, ii, 320): (12)

Guildenstern:

Good my lord, put your discourse into some frame, and start not so wildly from my affair.

We too would probably be disoriented if we met in real life a person whose speeches were disjointed like Hamlet's - because that would not be a literary situation. But when we see Hamlet performed, we can impose coherence on Hamlet's speeches as being parts of his plan to appear mad. Besides, we know when Hamlet is inventing obscure metaphorical allusions to the suspected treachery of his mother and his uncle, as in: (13)

Hamlet: King: Hamlet:

Farewell, dear mother. Thy loving father, Hamlet, My mother! Father and mother is man and wife; man and wife is one flesh; and so, my mother.

In the world of the play, we see Hamlet as coherent, whereas the other persons in the play cannot (except perhaps Horatio). We can apply the 'general truth' that people with arcane or forbidden knowledge appear mad to their contemporaries. We thereby institute a higher-order coherence to resolve a lower-order incoherence. The same is true on a more spectacular scale for Ken Kesey's novel, One Flew Over the Cuckoo's Nest, where the narrator's schizophrenic perception reveals startling allegorical truths amid its distortions. 6.4 The contrast between social reality and the world of the novel might account for people's motives in reading novels to begin with. Most people's lives are replete with isolated incidents, such as events whose causes and effects you never learn, or people you see a few times by accident and then never again. Sometimes, the only connection between your different experiences is the chance factor that they all happened to you. It is not normal to construe all your experiences as part of a strongly coherent design; people who do so, such as August Strindberg in his later years, tend to appear abnormal - fatalistic, superstitious, paranoid, and so on. 6.5 In the world of the classical novels, such as those of Fielding, Smollett, Dickens, or Balzac, we neither expect nor tolerate a comparably weak mode of

Schemas for Literary Communication

73

coherence. If some character is introduced as an orphan, we expect to find out in the course of the novel who the parents were and what happened to them. If a mysterious person enters the scene, we expect to learn who it is; the person can't just disappear and never be mentioned again. We accept startling coincidences and revelations that we may never have encountered in our own lives, such as the return of a long-lost relative, the payment of an unknown inheritance, and so on. In short, we want the world of the classical novel to be far more coherent than our everyday world. And we believe this non-natural coherence quite readily, once we've agreed to construct alternative worlds when we enter into literary communication. 6.6 Value judgements may be derived from the extent to which this supercoherence is attained. If a person in a play is represented in very different ways, and no explanation is given for the transformation, literary scholars tend to consider the inconsistency a weakness. Objections have been raised for instance against Cordelia in King Lear and Casca in Julius Caesar on the grounds that they are suddenly eloquent and profuse, although they had been introduced before as shy of words. Mr. Pickwick has been decried for having a deeper and more sensible character in the second volume of The Pickwick Papers than the one he had in the first. We hardly demand such coherence from people we have only just met in real life, because we don't expect real people to be prototypes of general human truths. 6.7 Thus, the authors of the nineteenth century novels had to respect this consideration. In Dickens' book Our Mutual Friend, we see the dustman burying a bottle and later exhuming it again. These events literally obligate Dickens to give an explanation before the end of the novel; anything less would be counted a gross oversight and literary flaw. The construction of the fragmentary novel The Mystery of Edwin Drood, which Dickens didn't live to finish, is so carefully crafted in its coherence that we can reconstruct the whole mystery from the small part we have - who killed Drood, how, why, and all about it (cf. Baker, 1951). The world of Sherlock Holmes is a supreme example of supercoherence: every detail mentioned, no matter how minute, is taken as a clue for one possible solution of the case at hand, though not necessarily the solution Holmes will find to be correct. Even false clues must appear to be coherent. 6.8 Some novelists have experimented with departures from this degree of super-coherence. Kafka is among the most famous innovators; he recasts the usual causalities and allows his protagonists, such as Gregor Samsa, K., and Karl Rossmann, to traverse apparently senseless and impossible event sequences. But our literary response is merely to transpose our demands for coherence to higher level. We construe the superficially absurd worlds of Der Prozess and Amerika as allegories of an aleatory real-world, of a human existence deprived of discoverable meaning and order. This way, we construct a super-coherent generality to incorporate the ostensibly incoherent specifics of the literary text-world. 6.9 Psychologically, the imposition of tightly organized coherence is evidently highly satisfying. Probably, the AESTHETIC character of literary and

74

Robert de Beaugrande

poetic communication is due to the attainment of super-coherence. For many years, aestheticians have been suggesting that 'beauty' arises from the interaction between 'variety' versus 'order' (Leibniz, 1720), or between 'variety' versus 'unity' (Hegel, 1835), or between the 'manifold' versus the 'unitary connection' (Fechner, 1876). There must be a two-step process here. First, the alternativity principle opens up the freedom to create a world with unfamiliar elements or patterns. Second, we process that world by creating a system in which those elements or patterns are assigned a function. We do not just leave the text-world as a random juxtaposition of unrelated elements; we want to find motives for why these elements were selected and organized this particular way. Thus, the first stage brings an increase in the COMPLEXITY and DIVERSITY of the experience; the second stage decreases those factors again. N o doubt the accomplishment of reconstructing a literary world whose ordering principles we have supplied ourselves is felt to be rewarding. We have in some sense rebuilt the world that, in everyday life, often lacks this degree of coherence (6.4). 6.10 The same effect should obtain for reading literary texts whose poetic use of language creates a noticeable obstacle to comprehension. Rare words or word-combinations at first resist the imposition of coherence (cf. 4.16). However, 'deviant' language, I argued already, is not the actual goal of poetry; deviance is merely a cue to retard our normally 'automatic' processing and to force us to deal with the specific occurrences in the text at hand (3.6ff, 13). We take the words of the poem and assign them as many functions as we can, so that in the end, their specific uses in the poem are seen to be strongly motivated (cf. Beaugrande, 1978b, 1979b): We need merely identify a text as poetic, and the presumption applies that all discoverable means of order have a meaning. (Lotman, 1972: 161).

In terms of systems theory, the text is a system in which linguistic elements are assigned contributive functions (6.2). In a poetic text, this assignment is enriched by alternative functions, again with a super-coherent result. This phenomenon has been variously described as 'polyfunctionality' (Schmidt 1971; Kloepfer 1975), 'secondary modeling systems' (Lotman, 1976) and 'overcoding' (Eco, 1979). Whether language items appear 'deviant' is not decisive, since processing can find new functions for otherwise ordinary items. Coseriu (1971) argues accordingly that poetry makes the fullest possible use of language. 6.11 If authors countermand the schemas of ordinary language use, how can they know what readers will do? How can authors control the outcome when there is no detailed precedent for the processes to be performed? Will readers agree on the best or most rewarding way to deal with deviant word-forms, syntactic displacements, or bizarre combinations of concepts? My suggestion would be that authors at least attempt to pre-figure the reader responses they hope to elicit. The selection of a genre with characteristic conventions is one means of control, but the genre seldom completely dictates the specifics of the individual work (cf. Frye, 1967). Still, the author can try to design each text so as

Schemas for Literary Communication

75

to increase the probability of a particular response, namely by making that response be one of the most satisfying ways to account for the experience of the text. The author can speculate with the readers' preference for super-coherence and pose problems whose probable resolution leads toward the authors' own version of the text-world. For example, it is very hard to make sense of Rilke's Duino Elegies unless we accept some of the odd beliefs upon which they are based: that people who died young are specially privileged, not denied; that suffering is more worthy and precious than joy or pleasure; and so on (Beaugrande, 1978a). The Tenth Elegy in particular is replete with elaborate metaphors that lead toward these beliefs, such as (X, 56-61) 6 : (14)

'We were,' she says, a great race once, we Laments. Our forefathers worked the mines there in the great mountain range; among humans you find sometimes a fragment of polished aboriginal sufferance, or, from an old volcano, cindery petrified wrath. Yes, that came from there. Long ago we were wealthy.'

This speech requires us to at least entertain the thesis that suffering is something of value to make poeple 'wealthy'. The 'Laments' figure as miners deliving up precious stones of great age and hardness. This metaphor resists resolution if we stay within the ordinary view of suffering as a negative accident of no merit, something to be overcome or avoided, not sought out and prized. 6.12 All the same, authors can't enforce a particular response to a text. The literary work is necessarily 'open' (Eco, 1968), due to the very nature of alter nativity. Not all readers could harbor the same demands for super-coherence. Many readers may stop short of the result an author intended. Other readers to beyond the author and assign relations that had not been intended (cf. Beaugrande, 1983b). For instance, our modern understanding of personality psychology enables us to elaborate the motives of characters in literary works, such as Crime and Punishment, or Madame Bovary, and to resolve seeming incoherencies in their actions (Paris, 1974). In so doing, our response reaches beyond the author's intention, because we are applying a recently developed systematics. Bernard Paris (1981) has shown that the Duke's oddly inconsistent actions in Measure for Measure suddenly fall into a coherent pattern if we view him as caught in conflict between self-effacing and perfectionistic tendencies. He thus is too self-effacing to enforce the draconian laws which his perfectionism led him to proclaim. He lives up to high own high standards, but is afraid of judging people who don't, and appoints the radical, single-minded perfectionist Angelo as the enforcer of harsh statutes. This modern theory of personality overreaches the response Shakespeare presumably intended, though it is a powerful resolution of the problem structure we can see in the drama.7 6

7

This passage, like several other quotes from German sources, I have translated myself. On the intricate problems of translating poetry, see Beaugrande (1978a). In Ken Kesey's modern novel One Flew Over the Cuckoo's Nest, clinical psychology is an explicit frame of reference, though filtered through the schizophrenic perception of the narrator, Chief Broom (cf. 6.3).

76

Robert de Beaugrande

6.13 Christian Enzensberger (1977) argues that the super-coherence of literature is used as an antidote for the 'meaning deficit' people experience in daily life (cf. 6.4). He asserts that literary worlds provide a counter-image of reality without changing reality: Literary meanings function as sedatives, dealing with praxis as something imaginary, absorbing social contradictions, and stabilizing the current distribution of power. People read and write literature in order not to act. (Review in Schmidt, 1982: 70)

Enzensberger concludes that social reality is immune to change through literature because the two domains follow incommensurate principles of coherence. Literary super-coherence is a symbol for a social Utopia that has been forgotten and made inaccessible by the praxis of reality. 6.14 This estimation looks too pessimistic for several reasons. Literary communication is found in many societies whose labor conditions and distribution of power are quite unlike the societies Enzensberger explores, such as the England of Oliver Twist. Some literary works have directly affected historical trends, such as RichardII and Uncle Tom's Cabin, though we might argue that these political effects were not a proper part of the literary interaction. But more essentially, I would respond that literary alternativity is neutral and encompassing by nature. A work may reflect a reactionary ideology or a revolutionary one; an intent to change reality, or an intent to replace it with fantasy; a refusal of Utopia or a affirmation of it; and so forth. And of course, any point in between those extremes is attainable. Hence, literature need not be a denial or a deterrent for Utopia; it might equally well be a means for preserving the hope for Utopia, as Bloch (1959) has suggested. Super-coherence is always an ideal (itself a kind of Utopia, if you will), but that just means that we can always move closer to it than we are. 7. Schemas for literary

communication?

7.1 A central dilemma for the psychology of literature might be stated as follows. On the one hand, schema theorists agree that text processing is guided by schemas drawn from a person's expectations and world knowledge; the schemas are abstracted from one's prior experiences. On the other hand, the alternativity principle opens up a wide potential for constituting unfamiliar worlds and innovative experiences. How then can people process literature with prior schemas? This dilemma has significant implications for the whole notion of schema as well. 7.2 One solution might be that the schemas themselves are sufficiently P O W E R F U L and G E N E R A L that they can in principle apply to comprehension independently of what is judged 'factual' versus 'fictional' or 'old' versus 'new'. If schemas are to be flexible enough to handle novel experiences, they must not equate one's prior experiences with the total range of the possible. This way, the potential for processing fictionality and novelty could be built right

Schemas for Literary Communication

77

into one's basic everyday schemas. Still, literary experiences would have to be systematically related (or relatable) to one's real-world experiences, namely at the degree of abstraction that makes schemas functional at all. It follows that the fictional, however fantastic, is necessarily processed in analogy to the factual (my contention in Section 4). 7.3 This account suggests that schemas (at least the dominant or 'top-level' ones) are like outlines with many unbound variables (cf. 2.3). How many of these variables get bound in a context of application depends on how much information is supplied, and what motives the text processor has. In a lowcontext task, many variables are left unbound or are bound with defaults. In a high-context task, many more variables are bound with the specific values being encountered. The. binding terminates when the processor is satisfied that enough variables have been bound for that occasion. The same schema might be bound to a much lesser or greater degree on some other occasion. 7.4 This account would imply that we might not need to postulate a set of specifically 'literary' schemas separated from the person's 'real-world' schemas. The top-level 'constitutive schema' (in the sense of 3.13) is a community of processes that determine which variables get bound, and which sources get used. The experiential 'literariness' of the text is a function of this processing mode. Alternativity would result as the person recognizes the negotiability and range of these variables vis-a-vis one's prior experiences. For documentary literature, we borrow from real life in rich detail; for fantastic literature, we only select isolated points and merge them into much less familiar constructs. For example, readers of John Steinbeck's The Red Pony can directly apply what they know to be true of farm animals. For Charlotte's Web, by E. B. White, we must accept the notion that farm animals can talk, though the farm itself is still run by human beings in the usual way. For George Orwell's Animal Farm, we must allow that animals can not only talk, but take over and run the farm without humans. And for strongly fantastic literature, such as Lewis Carroll's Alice in Wonderland, we must create talking, self-sufficient animals that are built from parts of different real animals, such as the Gryphon (half eagle, half lion) compare the legendary Sphinx and Minotaur. At each step in this progression, we are impelled to discompose our real-world schemas for animals to steadily greater degrees, and to inject elements from other schemas, e. g. the language and actions of human beings. But we remain in a continuum; we do not abruptly cross some borderline into a totally unknown realm unrelated to our real-world schemas. 7.5 The question then is when we can legitimately say that a schema used during literary communication is called from storage and applied, and when we ought to say rather that the schema is created on demand to meet the need of a given context. Probably, the answer depends on the FAMILIARITY of the context (cf. 1.4.3). People are likely to have ready-made - fairly complete and detailed - schemas for contexts they often encounter. But if the context is unfamiliar, schemas are more likely to be manufactured specially for the occa-

78

Robert de Beaugrande

sion. Thus, documentary literature is more liable to draw on ready-made schemas than is fantastic literature. However, a person who participates very frequently in literary communication presumably develops schemas for recurring literary domains, e. g. the 'Elizabethan revenge tragedy' ending with the violent deaths of most of the main characters. Bauer et al. (1972) found that literature students do tend to interpret abstruse contemporary poetry by relying on a restrictive range of literary commonplaces and interpretation strategies. Responses were less uniform among people without such training. 7.6 Hence, the question of whether people have ready-made schemas specially for literary communication would not be answered the same way for everyone. At the top level, the 'alternativity' schema must be admitted as a constitutive set of processes applied to the selection or creation of lower-level schemas and the binding of the latter's variables (3.13; 7.4). Below that, experienced participants in literary communication have steadily more detailed (lower-level) schemas that are influential, though not indispensible. The classical notions of 'genre' belong here. Despite many attempts at uniform definition, there is no good evidence so far that genres are all of the same order. Some genres are very elaborated and uniform, such as the sestina, which stipulates exactly where the end words of the various lines must recur. Other genres are vague and diverse, such as 'prose poem', where each author must define and resolve how 'prose' actually can navigate the borderline of the 'poetic'. Such diversity reflects the variable conditions under which genres developed, and the varying degrees to which innovation was or was not considered a high value at the time. A genre with a very long tradition, such as the sonnet, is likely to assume various forms, simply for the sake of poetic alternativity (cf. Beaugrande, 1978b). 7.7 Even so, it is hardly plausible that ready-made 'literary' schemas could be on a sufficiently low level to stipulate the exact events and patterns assigned to every work, or to most works (6.11). These schemas too must have at least some unbound variables, no matter how experienced the reader. Moreover, the same person can enjoy reading the same work over and over; evidently, the activity of applying schemas and building a super-coherent text world has a rewarding dynamics beyond any removal of uncertainty in the ordinary sense (cf. Beaugrande & Colby, 1979). Accordingly, we need to explore empirical methods for uncovering the natural applications of processing schemas during literary communication. 7.8 The so-called 'cloze procedure' has been tried: every fifth word or so is deleted for test subjects to guess. For example, Faulstich (1976) used a sample (15) by Stephen Crane, which appeared with deletions as (15a). (15)

I stood upon a high place, A n d saw, below, many devils, Running, leaping, A n d carousing in sin O n e looked up, grinning A n d said: ' C o m r a d e ! Brother!'

(15a)

I stood 1 a high 2 , And saw, 3 , many devils, 4 , leaping, And 5 in sin. 6 looked up, 7 , A n d said: '8 !'

Schemas for Literary Communication

79

Faulstich gave the sample first to two English classes at the University of Tuebingen and then to a sample population from various West German universities. The most common responses were then assembled in a new version: (16)

I stood on a high mountain, And saw, below, many devils, Dancing, leaping, And wallowing in sin. They looked up, laughing, And said: 'Join us!'

The enduring problem with the cloze procedure is that, depending on surface structure, topic, and context, it may be tapping very diverse kinds of knowledge. Syntactic structures require for blanks 1 and 6 simple function words, preposition and pronoun, respectively, even with no special concern for the topic. The differences between originals and replacements is trivial, the more so as each blank allowed only one word; there are for instance hardly any oneword subjects available here other than pronouns. However, the content words are also well delimited by the available clues. Given 'high' and 'looked up', the correct recovery of 'below' is to be strongly predicted, whether or not the test persons were envisioning the traditional schema with devils in a low place. Given 'leaping', the other two blanks 4 and 5 in the same modifier will tend to take verbal participles in '-ing', naming typical activities of devils. Probably, blank 7 was filled by analogy via the same considerations - devils are usually pictured as gleeful. Also, the completion of blank 8 is likely to fit the conventional vision of devils trying to lure human souls into their domain; the subjects' guess is a more normal utterance than Crane used for the same purpose (cf. 8.16). A concurrent version (from the first test run) had these lines at the end: (17)

I looked up, sadly, And said, 'Go away!'

anticipating the speaker's probable response to the devils' invitations. This ending is more optimistic than those of the original and of the most common replacement, but hardly very abstruse or creative. 7.9 It's fair to say that cloze tests alone need not uncover the specifically literary nature of processing schemas. These test persons did have to apply their schemas from a fictional domain, namely 'the activities of devils', but the original source is religious discourse more than literary. Other blanks could be filled with little regard for the specifics of the text-world — a general difficulty with the cloze procedure (Groeben, 1982: 70). 7.10 Still, it would be helpful to investigate the tendencies that influence people's responses on cloze tests with a wider range of texts. To explore classes or types of expectations, Piontkowski and Groeben (1975) combined cloze tests with semantic differential tests. They deleted words from both classical and modern poems. They used a semantic differential test to get the subjects (secondary school students at a German Gymnasium) to judge the abstractness or

80

Robert de Beaugrande

concreteness of both original words and the replacement words given most often on the cloze tests. 'Abstract' was defined as 'general, theoretical, vage, not grasped, not measured', etc; 'concrete' was defined as 'measurable, corporeal, visible, real, clear', etc. (Function words were excluded as not relevant to the parameter in question.) The findings were that classical poetry (e. g. Goethe) tends to be considered more abstract, whereas modern poetry (e. g. Krolow) tends to be considered more concrete. The experimenters suggested that readers expect the classical author to include interpretations and generalities, and the modern author to be content with just presenting objects and characteristics for us to interpret and generalize. 7.11 Another route of evidence would be to inquire whether 'literary schemas' are in the main derived from general functions of arousal, curiosity, and complexity control (cf. Munsinger & Kessen, 1964; Berlyne, 1971). For example, the findings obtained by Kammann (1966) suggested that readers prefer a certain degree of complexity as a basis for intensely experiencing a text and judging its quality. Yet Berlyne (1966) did not find that curiosity is necessarily raised by increasing the complexity. Berlyne (1974) postulates an 'inverse Ufunction' in which the rising arousal curve abruptly falls again. Presumably, undue complexity causes this inversion because the experiencer loses control, and pleasure changes to displeasure. 7.12 The nature of associative reasoning, as surveyed in 2.7ff on the basis of list-learning experiments, may be crucial. Those experiments indicated that people tend to agree in the associations they most readily recover for particular words. Presumably, if a person is given unlimited time or strong motivation, many more associates can be found, but steadily less common and probable ones. In text processing, the use of schemas ought to help decide which associates are relevant. For literature, a wider range of associations would be admitted than in other domains. 7.13 One recent unpublished finding from Walter Kintsch's laboratory relates to this account. Kintsch found that priming effects triggered by words in a text being read apply to certain classes of associates according to the timing of cues and responses. When his test subjects had a text about 'water' and were immediately cued after reading the word 'bank', consistent priming effects were obtained for both relevant associates like 'river' and for irrelevant associates like 'money'. But if the cue was delayed by 200 milleseconds or more, only the context-relevant associate was primed. Conversely, a less proximate relevant associate that had no immediate priming effects gained those affects after a delay of about 400 milleseconds. If the text word was 'airplane' and the associates were 'fly' and 'gate', the latter showed no priming on immediate cuing, but did on delayed cuing. 7.14 Apparently, contextualizing is a very rapid and selective process. Initial perception contacts irrelevant associates whose activation is quickly lost in favor of relevant associates. The shift can only be due to the swift and skillful selection and application of schemas, e. g. a 'flight' schema with the knowledge that

Schemas for Literary Communication

81

airplanes are boarded at an airport 'gate'. In literary communication, this normal sifting of potential associates appears to be displaced. In theory at least, readers might postpone schema activation and prolong the active status of all associates contacted during immediate perception. But my guess would be that this mechanism is not practicable, for at least two reasons. One reason would be that the initial schema selection process is heavily automatic and hard to interfere with at that speed (cf. Shiffrin & Schneider, 1977), given the strong need for immediate orientation. The other reason is that a text-world replete with irrelevant associates could mediate against the super-coherence that should be imposed on the literary world. Hence, it is more plausible that the added associates whose range constitutes 'alternativity' are activated in a later stage: after the schema application that delimits context in favor of relevant associates. This process would be more deliberate and more conducive to super-coherence. Whereas non-literary reading would automatically terminate soon after the shift of associative patterns from irrelevant to relevant, literary reading would continue in a more attentional (less automatized) stage where the relevance criteria derived from everyday life are relaxed again, but still controlled by the top-level goal of intensifying coherence. 7.15 This account concurs with the view that 'literary schemas' provide not just structured variables, but also theresholds for one's disposition to admit or address variables (cf. 3.13; 7.4, 6). Yet we still need to clarify why, if general human factors are the basis of these processes, people can differ so widely in their responses. Some readers are evidently motivated by prior training or arousal potential to accept large, complex patterns of variables. The binding of the variables would call for rearranging one's applicable schemas, or even creating new schemas (extreme 'accommodation' in the sense of 7.16). If the text is not felt to elicit sufficiently complex variables, these readers tend to devalue it and exclude it from intense processing. 7.16 Other readers want to work with simpler, less numerous variables, such as those strongly determined by context, or those readily supplied from commonplace schemas (e. g. about what 'devils' do). These readers value a text that they feel stays inside this moderate range without being totally predictable. If that range is exceeded, they soon find themselves on the negative downward slope of the U-function, and their experience is disorienting and unpleasant (cf. 7.11). Groeben (1982: 154) appeals to Piagetian theory and classifies readers according to whether they prefer 'assimilation' or 'accommodation'. Literary communication emphasizes 'accommodative' processing, but not everyone is inclined or trained this way to the same degree. 7.17 An account so derived from general human factors gives PROCESSI N G C O M P L E X I T Y the status of a primary category. Whatever participants in literary communication actually experience in respect to any parameter degree of fictionality, degree of linguistic deviance, degree of uniqueness vis-avis the genre, etc. - is a function of the processing complexity occurring during the encounter. The standard definition of 'complexity', I would think, is: the

82

Robert de Beaugrande

number and diversity of part-whole relations in a presentation (Beaugrande, 1982b). A schema could act as a 'macro-structure' of such relations, whether the constituent elements be unbound variables or standardized defaults. As long as the schema reduces complexity, it need not be required to supply the details of content, e. g. whether the devils are 'carousing' or 'wallowing'. It should follow that literary alternativity would constitute the text-world of a literary work via structural analogy to the real world; but considerable interchanging of constituents within those structures would be readily feasible. For example, each world in Gulliver's Travels by Jonathan Swift is revised according to a few powerful principles: everything is preternaturally small in Lilliput, preternaturally large in Brobdingnag, and so on. As long as the principles are made clear with salient examples, readers are able to adjust their schemas accordingly and process the remaining text without becoming disoriented. The power of the principles allows for numerous alterations (fictionalizations) of detail without a notable rise in processing complexity. 7.18 In sum, literary schemas should be structurally analogous to real-world schemas, but with a top-level constitutive schema that enacts a dominant opening toward unbound variables. Structural analogs reduce complexity without fixing the range of the variables. Readers prefer the degree of complexity they can process at a workable level of arousal and uncertainty. But due to the wide variations in what will be judged (or experienced) as the proper degree of complexity, there will always be a range of responses and evaluations elicited by any given literary work in any specified context. Uniform schemas in literary communication are precisely so designed as to lower the probability of uniform results (cf. 2.9, 14).

8. The ordinary reader: A pilot study of verbalized responses 8.1 Literary studies has devoted most of its resources to displaying or analysing the actions of expert readers - analogous to Riffaterre's (1971) 'superreader'. The ordinary reader, though far more enumerous, has received relatively slight consideration. Any theory of literary communication ought to account for all kinds of readers, not just for professional critics and literary scholars. 8.2 In general, two complementary findings have begun to emerge. As I noted in 7.5, ordinary readers tend to fall back less on ready-made 'literary schemas' for interpretation than do highly trained students of literature (Bauer et al., 1972). I have considerable data indicating that naivety can have a liberating effect, apparently because naive readers feel less pressured to to find 'right' answers and are thus inclined to search and speculate more widely. On the other hand, naive readers tend more to normalize their readings, i. e. bring them closer to their ordinary comprehension of the world (cf. 7.8; 8.16). Eggert et. al (1975) observed that literature students at a Gymnasium in Berlin even reversed

Schemas for Literary Communication

83

the meaning. They read Kleist's disputation where he contemplates giving his son the advice to act first and think about it afterwards, on the grounds that thinking first would confuse and slow down the action; he compares life to a swift and strenuous wrestling match where we must act instantly, though we can reflect back later on to see what we did right or wrong. Most of the students believed to have read just the opposite message: think first, then act. They couldn't imagine a father giving any other advice, and they had heard it from their own fathers very often. 8.3 Though both advantages and disadvantages are involved in naivety, I think the causalities are much the same. Being less inhibited and trained, naive readers feel less responsible for giving a polished, definitive account. Their increased range may both enrich and distort their constitution of meaning. The same freedom enable both discovery and dispersal. This issue deserves much more research. I shall outline one pilot study of mine as an ongoing attempt to define some classes of response among ordinary readers, at least as far as those responses can be verbalized. 8.4 At the end of one term, I ran an unannounced test. During individual interviews, each student was unexpectedly given am unfamiliar poem by Richard Eberhart to read and explain: (18)

H A R D E N I N G INTO PRINT 1 To catch the meaning out of the air 2 Yet have it inviolably there, 3 Life I mean, the glimpse of power, incomparable times 4 Of total splendor, the sudden exaltations, 5 Flash of a thrush, a rush of golden insight, 6 To be caught up in titanic light 7 As if one saw into the depths of things, 8 Yet averts the eye, to try yet further mysteries, 9 It is into this rich reservoir 10 Of knowing and unknowing I flash, 11 And shake high lightning spears of life 12 In the long combats of mortal strife, 13 Thrush song piercing human ills 14 With rigor and wrench so deep, 15 This glimpse is of an immaculate joy 16 Heart suffers for, and wishes to keep.

Intuitively, I rated this poem as challenging, but not unmanageable. 8.5 Its vocabulary ranges from fairly common words (e. g. 'flash', 'glimpse') over to fairly uncommon words (e. g. 'inviolably', 'titanic'). The vocabulary affords a complex mixing of the concrete (e. g. visual or auditory effects) vs. the abstract (e. g. interpretations) (cf. 7.10), notably for associations like 'glimpse of power' (3), 'rush of golden insight' (5), or 'high lightning spears of life' (11). This mixing of opposed conceptual domains can be resolved by constructing an integrative schema of concepts whose content overlaps by virtue of shared categories the reader supplies. The categories of brightness and clearness can draw together 'splendor' (4), 'golden insight' (5), 'light' (6), 'lightning' (11),

84

Robert de Beaugrande

and, less obviously, 'flash' (5,10) (which I consider to be a motion more than an illumination here). Being hidden or momentary are categories associated with: 'out of the air' (1), 'glimpse' (3, 15), 'sudden' (4), 'flash' (5, 10), 'rush' (5), 'depths/deep' (7, 14), 'averts the eye' (8), and 'mysteries' (8). The categories of being strong and uplifting can draw together: 'power' (3), 'incomparable' (3), 'total splendor' (4), 'exaltations' (4), 'titanic' (6), 'high' (11), and 'joy' (15). Categories contrasting against these can also be integrative. Being manifest or long-lasting is suggested by 'hardening' (title), 'inviolably there' (2), 'long' (12), 'keep' (16), and, less reliably, by 'reservoir' (9). Being vulnerable or afflicted is suggested by 'combats', (12), 'mortal strife' (12), 'piercing' (13), 'ills' (13), 'rigor' (14), 'wrench' (14), and 'suffers' (16). For my reading, the subsuming categories work well if grouped into sets of two, e. g. 'hidden' + 'momentary', 'strong' + 'uplifting', and so on. 8.6 My conjecture is that the poet used the tension between opposing categories to enact the main theme of how experience resists 'hardening into print', including a printed poem. The paradox of 'suffering' for an experience and yet 'wishing to keep' it is ambivalent like the experience of processing the poem itself. The 'print' of the poem is a vehicle that is being deployed to mediate against its own stability. The 'hardening' is less decisive than the dynamics ('flash', 'rush') of the original experience. The poet himself judged this interpretation of mine to be appropriate. 8.7 Each student participated in an individual session that was recorded on tape for later transcription. The student read the poem aloud and was then asked to describe or summarize it from memory; however, in most cases, the single oral reading had provided a very slight basis, and responses were negligible. Next, the student get the poem back to look at, and again tried to describe or summarize 'what it's about'. From time to time, I posed questions, such as: 'what images stick in your mind?' or 'how do you think the title relates to the rest of the poem?' I had often asked these questions in class discussions during that semester. 8.8 I predicted that the complexity and ambivalence of this conceptual structure would be a challenge for the naive students. The question was how far they would concur in the ways they constituted and accounted for the textworld, at least as far as verbal reports can afford evidence of these processes. I shall offer a broad classification of the typical operations students performed in reporting their responses to the poem. These operations are by no means mutually exclusive; several may be performed during a single utterance. For instance, when stating the text's topic, students would typically mix key-word association, generalizing, and normalizing. Moreover, I do not claim that my classification is exhaustive or universal. I don't have enough evidence to form any judgement on those factors. 8.9 I illustrate my classification with excepts from the transcripts. My own utterances during the sessions are enclosed in pointed brackets. To avoid interpretive punctuation, I mark a short pause w i t h ' / ' , and a long pause w i t h ' / / ' .

Schemas for Literary Communication

85

Hyphens indicate where a word was broken or interrupted. Actually, a more sensitive representation giving intonation curves would be desirable for a finetuned analysis (cf. Brazil, 1975). 8.10 F R A M I N G includes any elements used for announcing or introducing comments. The transcripts contained examples like these: (19) (20) (21)

the first thing that I got from it was when you look at the title now / you think it makes me think whewww // trying to catch the rest of it // well like from what I can see in these other lines

Framing is especially typical of spontaneous explanations. Speakers gain time to organize their ideas, and can signal where they are getting the evidence for their remarks. 8.11 H E D G I N G is quite common in everyday spoken discourse: signalling that one's statements are provisional, inexact, or personal (cf. Beaugrande, 1983a). The proportion of hedges in these sessions was quite high. Some hedges were also framings, such as 'I think', 'it seems', and 'it looks like'. Some were simple adverbials, such as 'maybe', 'perhaps', or modifiers depending on the preposition/conjunction 'like'. Still others were vague place-holder words like 'thing', 'something', or 'it'. These place-holders came up frequently when students tried to state the main topic, as in: (22) (23) (24)

uh / I don't know it / it seems that uh / someone's contemplating something II and he's // and he can't really make up his mind when you write something / you / you just you think of it / like / up in the air / and then you put it down on paper um / by the things that he's seeing / the things that he sees around him / he uh / he's feels he's feeling joy

Probably, the associations among disparate concepts of concrete versus abstract quality (8.5) increased the students' feeling of uncertainty. Thus, the hedges could be behavioral indicators of having experienced a predominantly open and polyvalent meaning. 8.12 C I T I N G happens when the respondent repeats parts of the original text verbatim. Students typically cited a line or so from the poem and then offered and explanation, as in: (25)

(26)

to catch the meaning out of the air / yet have it inviolably there / it means like / since it's talking about hardening into print / when you write something / you / you just you think of it / like / up in the air / and then you put it down on paper / which is where they get yet have it inviolably there as if one saw into the depths of things / yet averts the eye / to try yet further mysteries / there's so much there's so much out there / that you can't get it all / and uh / and then it's even still more mysterious / I mean there's so

86

Robert de Beaugrande

much up there and it like say averts the eye / you can't see it all you can't get it all / so then it gets that much more mysterious Rehearsing stretches from the original text could help the respondent by reactivating or re-instating the construction of meaning in working memory, so that the respondent can consider and describe it more easily. 8.13 In KEY-WORD ASSOCIATION, the respondent cites a word or two from the original and associates it with other materials. To get a notion of what would most often be treated as key words, I would ask students questions like: 'what things particularly stick in your mind?', or 'do you remember any images from it?' Here are some replies: (27) he kept talking about things flashing / ummm / light / flashes of light (28) the / the lightning / uh the sharp cry of the thrush / um / and I think / light in general (29) did he say something about thrushes? / (mm-hmm / anything else?) / thrushes / and lightning Sometimes students tried to explain why certain key words attracted attention: (30) the golden insight II flash of a thrush / arush ofgolden insight/um/ (what are those things doing in there?) / / y o u mean as / uh the flow of the / okay / they attract your attention because it was a flash and a thrush / the way it was worded / the sounds of the words / annnd // golden insight / umm / / something you can't / that something you can't really pick up in a book / that you have to find out for yourself or see it / see it coming / by something you've done Key words were plausible candidates for being explained, or defined with associated words: (31) an immaculate joy / um / the cleanness / the the purity / the perfection / of of knowledge / of knowing / um / the / meaning of life A common use of key words was in the students' attempts to state the topic. Presumably, the key words were thought to be the best evidence for justifying how the topic was being construed, as is clearly the case in (32) and (33). However, the associations were often quite extensive and mediated in regard to the key words, as in (34) and (35). (32) I think it's about a / a / the ultimate knowledge / because it says the glimpse is of an immaculate joy / heart suffers for (33) sounds like he's trying to decipher the meaning of life / that uh / like he thinks he may either have it or be on the verge of getting it // or it says like as if one saw into the depths of things / yet averts the eye I to try I yet further mysteries / it seems like / you know / he thinks that one can never really um / know the meaning of life / that there is always something more mysterious

Schemas for Literary Communication

87

(34)

by the things that he's seeing / the things that he sees around him / he uh / he's / he feels he's feeling joy and uh just by experiencing and looking around him he's feeling uh / different feelings of joy (35) I notice / like he uses the flash a couple times / um / of golden insight / and a light somewhere / and I get the impression that / he he's I don't know somehow he's / done things in front of people or people have appreciated him a lot and he's wondering what it was all for / if it was for him or if it was for them maybe

Similarly some students brought together key-words from various sections of the text, so that one interpretation would account for evidence from diverse textual sources, sometimes proximate, as in (36) and (37), sometimes more remote, as in (38) and (39). This merging of evidence seemed to make the interpretation more satisfying (cf. Beaugrande, 1983b). These statements also impressed the poet himself most profoundly (cf. also 9.5). Here are some significant examples: (36)

the thrush is something / well what it made me think of at first / was like / carrying a message / like the bird was supposed to carry the message or whatever / because it says the flash of a thrush in a rush of golden insight / like the bird is giving him the secrets of life or something / giving him insight on what he doesn't know // and it says as if one saw into the depths of things / you know / after the bird came then he could see / uh into the deep meaning of things (37) well it's like / the bird is bringing happiness or something / because it says piercing the human ills/ that means getting through / and it says um with rigor and wrench so deep / this glimpse of immaculate joy / and it seems like the bird is bringing joy / into a sick world (38) I think some of the spears of life he described when he described life in the first stanza / glimpse of power incomparable times / total splendor sudden exaltations (39) heart suffers for / and wishes to keep / uh / meaning that your heart wants to know / to to retain this knowledge / and to hold onto it / um and yet // um / once again / you know / from the second stanza / yet averts the eye / heart suffers for and wishes to keep / meaning that / you cannot keep this knowledge / it's brief and passing The words in the title, in this case 'hardening' (or just 'hard') and 'print', were particularly prone to being selected during key-word association. Here again, the associations covered a wide range of elaboration and interpretation: (40) this poem like everything else that anybody writes or says // all it does is really harden / harden into print / it doesn't / go any further (what does it harden?) // the / just what he's / hardening into print / just what what he said / his ideas / are diminished to just / print (41) you get things from around you / and then you put them down // like into

88

Robert de Beaugrande

print / like hardening into print / things are all out there / but I could think that to harden it / to harden it is to finally get a grasp of it / which should go with putting in into print / you finally know what you want to do with it know what you want to say about it / and put in into print The key words from the title also got associated with other key words in the poem: (42) hardening into print II well that gradually the meaning of life / the the meaning of life can be / written down? (what happens to it when you write it down?) // it hardens? joy you mean? (mm-hmm / so what happens when it hardens?) then you can preserve those moments of joy thing? / if life can be recorded into print then that means that even those little moments of joy that you had are part of life (43) hardening into print sounds like / hardening of your heart or something The associated text-word could trigger a further chain of associations, as happened with 'heart' in: (44)

(see anything in the poem that relates to that idea of hardening into print? / the kind of things that might get hardened?) / uh / heart? [...] a heart can get hard from / well from abuse / not just physical abuse but / you think of the heart as / the uh / object of / everything you do in your life / and it could get hardened from well / like this / saying / you know / don't give your heart away / because it is / it is what keeps you going / if you if you don't take it from the physical aspect if you don't look at the heart as the / not only the mainstream of life / but the thing that keeps you going / physically / that is the thing that keeps you going mentally // and it could be hardened from abuse

Some associations appeared to tie in with senses I would be reluctant to assign to the key words in this poem. 'Hard' in the sense of 'difficult' is a case in point. That sense is obvious in (45), and probably encouraged the associations in (46), where the word itself isn't used. In (47), a still different sense of 'hard' namely, 'with sharp impact', is activated. (45) hardening into print / that sounds like um / things you know / things that he's thinking about you know / he's trying to decipher you know what life means when you try to you know / like / get it together and put it down solidly it's just like / it's very hard (46) hardening into print uh maybe it could be / a process of / ideas to a written / um / work like writing maybe a poem / or a / a novel or something like that // you have to go through a lot of work / to d- to obtain satisfaction at / and there's trial and error / uh disappointment (47) hardening into print / uh / perhaps that / he's realizing only after he put it down on paper the hard- / the hardness of the truth / it it hits him really hard

Schemas for Literary Communication

89

8.14 Still, all these results could be due to the same basic processes of taking an activated word sense and associating it with other concepts in memory. This processes are essential to imposing coherence upon discourse in general, as a means to supply contexts for what is actually uttered. For word lists, associations appear to be used as organizational supports for recall (cf. 2.6). Perhaps some similar use is made in poetic discourse to support the retention of concepts while a subsuming framework is being found or constructed. Also, it's noteworthy that people seemed to agree about which were they key words. When Frey (1970) asked students to underline all 'noticeable text passages', they tended to agree on which ones should be marked (cf. 9.1). 8.15 P A R A P H R A S I N G happens when the respondent undertakes to express the same concepts, but in different words. Even in everyday discourse, most paraphrasing entails some degree of approximation and alteration of meaning. Since poetry draws special attention to the organization of discourse elements, these two factors (approximation and alteration) are particularly prominent when a part of a poem gets paraphrased, as in samples (25) and (26). I shall distinguish two subtypes paraphrasing, namely 'normalizing' and 'generalizing'. 8.16 N O R M A L I Z I N G would be the sub-type in which the paraphrase brings the content of a passage closer to the respondent's own everyday discourse. This process should make comprehension and integration easier. N o r malization may be done (a) by rearranging the syntax but retaining key words or word-stems, as in (48) and (49); (b) by supplying commonly associated words, such as 'diving' in (50); (c) by supplying everyday associates for key words, as in (51); or (d) by making commonsense analogies and paraphrases, as in (52), (53), and (54). (55) and (56) are good examples of mixing key-word association with paraphrase. (48) to try yet further mysteries —» it's even still more mysterious (49) yet averts the eye / to try yet further mysteries —» the eye turns away to find / more mysteries to understand other things (50) into this rich reservoir [...] I flash —diving into the reservoir (51) to be caught up in a titanic light —» to be caught up in a in a / in a light of / a large light or a large glow (52) as if one saw into the depths of things —» as if you really understood (53) lightning spears of life —» flashbacks and things you know / like a sudden burst of light you know / knowledge or something (54) caught up in the light —> your mind's clear and open and you can see things and I think that's what he's talking about not just the sunlight / but like / you like I you know / you saw the light or something the light / you think of a lightbulb coming on (55) of knowing and unknowing I flash —» it comes and it goes and he doesn't always know everything (56) knowing and unknowing —* he doesn't know the whole meaning but he find he can figure out different things and some things are unexplainable

90

Robert de Beaugrande

Normalization was sometimes tried by expressly comparing text versions against more normal versions: (57) thrush song piercing human ills with rigor and wrench so deep / / 1 guess / thrush is kinda like a bird singing it's not piercing your ears it's piercing human ills with rigor / and wrench so deep // um // like it's getting into the real meaning behind what's wrong with humans Of course, we could dispute whether these normalizations dismantle the poetic experience by rendering the content in more familiar expressions: the specific alternatives for discourse get interchanged. Still, seeing the relationship between the poetic and the more normal is plainly a major factor in poetic communication (cf. 3.12). 8.17 GENERALIZING is the sub-type of paraphrase that widens or removes specifications. Like associative activation, generalizing is a common memory process. It conserves on processing effort and memory storage by dumping details and replacing concepts with more subsuming ones (cf. De Villiers, 1974). Generalizing is usually accompanied by normalizing, because most general terms are common expressions in the language (e. g. 'life' and 'thing' in the examples below). Also, the degree of abstractness (in the sense of 7.10) tends to be increased. The students were notably prone to generalize when attempting to derive a topic statement from the materials in the poem: (58)

(now tell me what / what you think / it's about) // one word answer / about life (59) it's about the different / things that you encounter / during / during your lifetime / that come up or all the things you pull out / that you do in your life / that will /come in together / in this / in a reservoir // come together as one / to make you a person / the person that you are Students also generalized when expounding a short passage: (60) to catch the meaning out of the air /1 thought / okay / that seems to mean / that um // that he's looking for the meaning for something / that / it's almost like a hit and miss situation (61) the glimpse of power incomparable times of total splendor and sudden exaltations / those are all the nice things of life (62) in the long combats of mortal strife / 1 guess he's talking about the day to day things / and shake high lightning spears of life // that's / that's / that's like / maybe he's really enjoying life (63) in the long combats of mortal strife // um / uh / here's he's describing he's describing / the // how life can be hard / long combats / how life can be hard / mortal strife / um / you know / the problems of / of mortality / of of / the problems of being human Again like normalizing, generalizing tends to dilute the experience of the text as such, but still affords significant evidence.

Schemas for Literary Communication

91

8.18 The affects of normalizing and generalizing should be most negative when processes become irreversible, that is, when readers can't get back from their normalized/generalized version. The student who generalized very drastically: (22)

it seems that uh / someone's contemplating something // and he's // and he can't really make up his mind

remained, throughout all his responses, atypically insensitive to the details reported by other students. 8.19 Combined with associating and normalizing, generalizing can have the negative affect of sidetracking the experience along a tangential associative chain: (64)

(65)

it could be / about a relationship too / that would fit where it says / um / to be caught up in a titanic light as if one saw into the depths of things / yet averts the eye / to try yet further mysteries that's like / you know / you're understanding this person and you're growing to love them or something / and / yet averts the eye and you're left like / you're looking to somebody else to try further mysteries / to see if you can understand something or somebody else hardening into print / putting / a scene into print / or putting something down on paper? / (what happens to it?) / i t changes?/ (how?) changes in meaning? / um / well / a movie's a good example / you read a book and see the movie or vice versa read / see the movie and then read the book / lot of different interpretations / um / dialects // can change / the // appeals to different senses / you can see it whereas you read it and you have to imagine it / and if it's in print / um // you know / whereas one or the other could do more for / the actual story / whereas the movie could do more for it / as / if poe- / it depends on the person too there / if the person / sees and understands better than he reads and understa- and imagines / he would like the movie better / and it depends on / wheth- / see it depends on the person whether he gets more out of the / uh / the print or whether he gets more out of the movie

We need much more research on what participants in literary communication regard as relevant, and how far a perceived lack of relevance is evaluated negatively. 8.20 A case in point is readers' disposition to treat alternative metaphors as relevant despite incommensurate levels of abstractness (again in the sense of 7.10). The students sometimes resolved a concrete, unfamiliar metaphor by recourse to a more abstract and familiar one (cf. 8.17): (25) (36)

to catch the meaning out of the air / when you write something / you / you just you think of it / like / up in the air the bird was supposed to carry the message or whatever / because it says

92

Robert de Beaugrande

the flash of a thrush is a rush of golden insight / like the bird is giving him the secrets of life or something / Or, we might inquire whether one can borrow the structure of an abstraction without the concrete content that conveyed it via the original text. For example, the poem itself seems to me to depend on a tension between opposed categories, e. g. momentary versus enduring, or being uplifted versus being afflicted. Some protocols give evidence of oppositions, but not with the same content: (66) maybe the / uh highness of it and the lowness of it / it's uh / either good or bad/ The question of whether 'the structure of content' can be an independent notion in theories of text processing merits further inquiry (cf. Beaugrande, 1980: 9, 93-102 and 234-240). Literature, governed by the dominance of alternative constructs, should be a useful domain for such research.

9. Future

directions

9.1 On the most basic level, we have the straightforward observation that the students consistently picked out certain elements and passages, such as 'flash' 'golden insight', 'light', 'averts the eye', etc. This finding agrees with those of Frey (1970), who found that readers agreed about which text constituents attracted attention (8.14). Presumably, the readers made their judgements by estimating how far a text occurrence was a more or less common instantiation for a schema. For instance, in this passage: (67)

In der Morgendämmerung war ein Gewitter niedergegangen und hatte die Nacht mit weissen Fackeln und dröhnenden Posaunen verabschiedet. [At dawn, a thunderstorm descended and took leave of the night with white torches and uproarious trombones.]

people singled out the 'white torches' and 'uproarious trombones'8 as the most salient elements, presumably because they do not fit the usual schema for a 'thunderstorm'. In such cases, we are reasonably safe in concluding that people can pick consistent text locations as triggers for the admission of alternatives. However, we have no good reason to expect that the outcomes of such triggerings will agree from one reader to another. The very status of a passage as a trigger suggests that alternative responses are quite probable (cf. 2.9; 7.15). 9.2 Reading and explaining a challenging poem is doubtless a high-context task. Its complexity for naive readers results from its unfamiliarity (cf. 1.4.3). These readers presumably used their routine strategies for explaining texts, augmented with a schematic, attentional opening toward more diffuse variables and associations (cf. 7.3,14). To be sure, reading processes are not identical with * 'Posaune' in older usage, e. g. the Bible, is often found for any kind of brass heraldic instrument, including what we would call a 'trumpet' rather than a 'trombone'.

Schemas for Literary Communication

93

verbalizations of or about such processes. We would still need a theory that relates a typology of discourse operations, such as that sketched in 8 . 1 0 - 2 0 , to the general processes of literary communication and cognition, such as the assimilation and accommodation mentioned in 7.16. Framing and hedging are ways to contextualize one's responses in a domain of relatively high complexity and uncertainty. Citing and key-word associating suggest the processes of selecting a text element or pattern and associating it with one's prior knowledge. Probably, the set of associations actually made during reading corresponds only fuzzily to the set that gets reported. Some of the original set are forgotten, or resist verbalization. Moreover, new associations can be formed during the act of reporting. At least, the processes of text production impose ad additional framework of organization upon whatever was experienced (cf. Beaugrande, 1983a). 9.3 If, as was suggested in 7.14ff, a literary response includes operations of perceiving complexity and then reducing it with a super-coherent schematization, we should predict that some discourse operations reflect or enact the students' attempts to reduce complexity. This function seems fairly certain for the processes of normalizing and generalizing. The mind is trained to respect its limits on memory capacity and processing load by replacing some aspects of a presentation with more normal and general versions, i.e. by increasing the parameters of familiarity and abstractness (cf. 8.17). However, becoming a skilled reader of literature probably entails an effort to resist these tendencies and to pay greater attention to the more unfamiliar and concrete aspects of the individual literary work - to contemplate these particular alternatives for organizing worlds and "discourses. 9.4 Hence, we could use verbal reports as a means to pursue the development of literary schemas during the readers' continuing encounters with literature. My students were still in an early phase, though they had gotten some previous experience in that same semester. Longitudinal studies over several years could probe a sequence of phases moving toward the eventual emergence of the expert. Tests such as the cloze procedure and semantic differential could be used to control for such factors as predictability and abstractness - factors that probably undergo significant evolutions as readers gain expertise. Such research should in turn clarify the question of how far specifically literary schemas are constructed (cf. Section 7). Though I surmise that different readers would tend to adopt comparable schemas, the observed results on a given task might be quite diverse (cf. 2.9; 7.15; 9.1). 9.5 All the same, not every response can be equally satisfying. Though I have no solid empirical evidence for the claim, I am confident that responses are preferred when they account for several problematic elements at once. That is, responses which reduce complexity and also support super-coherence are more satisfying than those which do not. This account fits the notion of 'aesthetic unity' (cf. 6.9) and suggests why literary critics work so hard to put together as many details as they can (cf. Beaugrande, 1983b). An author could try to design

94

Robert de Beaugrande

the text such that the intended response would have this unifying function, even if other responses with the same traits could not be ruled out (cf. 6.11). For example, Richard Eberhart himself strongly approved of student responses that associated several key words from various locations in the poem (8.13). 9.6 To be sure, there is a complex of processes involved in all this - what satisfies which reader for what motives in a given context and culture, and so on. In this paper, I have reviewed some factors that may be generally entailed in all kinds of processes: contextualizing, complexity, familiarity, abstractness, and so on. My contention is that the generality of these processing factors is responsible for the reasonably reliable and orderly nature of communication. Thus, despite inevitable differences in personal experiences and specific responses, people agree on some major literary issues: what passages are salient, what works deserve to be read, what problems and solutions are worth considering, and so on. Accordingly, the psychology of literature can legitimately undertake to uncover the nature of the processing schemas that somehow manage all this complexity and promote all this agreement. This enterprise will also bear upon the fundamental concerns of psychology in the widest sense as an exploration of human processing capacities.

References Adams, M. and A. Collins 1979 "A schema-theoretical view of reading" in Freedle (ed.) (1979: 1 - 2 2 ) . Anderson, J. R. 1976 Language, Memory, and Thought (Hillsdale, N.J.: Erlbaum). Anderson, J. R., and G. H. Bower 1978 "Recognition and retrieval processes in free recall", Psychological Review 79, 97-123. Anderson, R., and J. Pichert 1978 "Recall of previously unrecallable information following a shift in perspective", Journal of Verbal Learning and Verbal Behavior 79, 1 - 1 2 . Anderson, J. R., R. Spiro and W. Montague (eds.) 1977 Schooling and the acquisition of knowledge (Hillsdale: Lawrence Erlbaum) Anisfeld, M., and M. Knapp 1968 "Association, synonymity, and directionality in false recognition", Journal of Experimental Psychology 77, 171-179. Ausubel, D. 1960 "The use of advance organizers and retention of meaningful verbal material", Journal of Educational Psychology 15, 2 6 7 - 272. Baker, R. 1961 The Drood Murder Case (Los Angeles: U C L A Press). Bartlett, E 1932 Remembering (Cambridge, England: Cambridge University Press). Bauer, W., et al. 1972 Text und Rezeption (Frankfurt: Athenaeum). Bauer, M., et al. 1972 Toposforschung (Darmstadt: Wissenschaftliche Buchgesellschaft).

Schemas for Literary Communication

95

Beaugrande, R. de 1978a Factors in a Theory of Poetic Translating (Amsterdam: Rodopi). Beaugrande, R. de 1978b "Information, expectations, and processing", Poetics 7, 3 - 4 4 . Beaugrande, R. de 1979a "Psychology and composition", College Composition and Communication 30,50 - 57. Beaugrande, R. de 1979b "Toward a general theory of creativity," Poetics 8, 269-306. Beaugrande, R. de 1980 Text, Discourse and Process (Norwood, N.J.: Ablex). Beaugrande, R. de 1982a "The story of grammars and the grammar of stories", Journal of Pragmatics 6, 383-422. Beaugrande, R. de 1982b "General constraints on process models of language" in Ny and Kintsch (eds.) (1982: 179-192). Beaugrande, R. de 1983a Text Production (Norwood, N . J . : Ablex). Beaugrande, R. de 1983b "Surprised by syncretism: Cognitive processes in literary criticism", Poetics 12, 83 -137. Beaugrande, R. de, and B. N. Colby 1979 "Narrative models of action and interaction", Cognitive Science 3, 43 - 66. Berlyne, D. 1966 "Notes on intrinsic motivation and intrinsic reward in relation to instruction" in Bruner (ed.) (1966). Berlyne, D. 1971 Aesthetics and Psychobiology (New York: Appleton-Century-Crofts). Berlyne, D. 1974 Studies in Experimental Aesthetics (Washington, D. C. : Hemisphere). Bloch,E. 1959 Das Prinzip Hoffnung (Frankfurt: Suhrkamp) Bobrow, D., and A. Collins (eds.) 1975 Representation and Understanding: Studies in Cognitive Science (New York: Academic). Bransford, J., and M. Johnson 1973 "Considerations of some problems of comprehension", In Chase (ed.) (1973: 383-438). Brazil, D. 1975 Discourse Intonation (Birmingham: English Language Research). Bruner, J . S. (ed.) 1966 Learning about Learning (Washington, D. C. : Hemisphere). Chase, W. (ed.) 1973 Visual Information Processing (New York: Academic). Cofer, C . N . (ed.) 1961 Verbal Learning and Verbal Behavior (NY: McGraw-Hill). Coseriu, E. 1971 "Thesen zum Thema Sprache und Dichtung" in Stempel (ed.) (1968: 183 -188). Cullingford, R. 1978 Script Application (New Haven: Yale Dissertation). Deese, J. 1961 "From the isolated verbal unit to connected discourse" in Cofer (ed.) (1961: 11-31).

96

Robert de Beaugrande

DeVillers, P. 1974 "Imagery and theme in recall of connected discourse", Journal of Experimental Psychology 103, 263 -268. Eco, U. 1968 L'opera aperta (Milan: Bompiani). Eco, U. 1979 A Theory of Semiotics (Bloomington: Indiana University Press). Eggert, H . et al. 1975 Schueler im Literaturunterricht (Cologne: Kiepenheuer and Witsch). Enkvist, N . E. 1973 Linguistic Stylistics (The Hague: Mouton). Enzensberger, C. 1977 Literatur und Interesse (Munich: Hanser). Faulstich, W. 1976 "Die Relevanz der cloze-procedure als Methode wissenschaftlicher Textuntersuchung", Linguistik und Literaturwissenschaft 21, 81-95. Fechner, G. 1876 Vorschule der Aesthetic (Leipzig: Breitkopf and Haertel). Freedle, R. (ed.) 1979 Advances in Discourse Processing (Norwood, N.J.: Ablex). Freedle, R. (ed.) 1979 New Directions in Discourse Processing (Norwood, N.J.: Ablex). Frey, E. 1970 Franz Kafkas Erzaehlstil (Bern: Franke). Frye, N. 1957 The Anatomy of Criticism (Princeton: Princeton University Press). Garvin, P. (ed.) 1964 A Prague School Reader on Esthetics, Literary Structure, and Style (Washington, D. C.: Georgetown University Press). Gibson, E. 1971 "Perceptual learning and the theory of word perception", Cognitive Psychology 2, 351-368. Gomulicki, B. 1956 "Recall as an abstractive process", Acta Psychologica 12, 77-94. Groeben, N . 1977 Rezeptionsforschung als empirische Literaturwissenschaft (Kronberg: Scriptor). Groeben, N . 1977 Leserpsychologie - Textverständnis - Textverständlichkeit (Münster: Aschendorff). Gross, R. 1967 Pop Poems (New York: Simon and Schuster). Grunzenshaeuser, R. 1962 Ästhetisches Maß und ästhetische Information (Quickborn: Schnelle). Halliday, M. A. K., A. Mcintosh and P. Strevens 1964 The Linguistic Sciences and Language Teaching (London: Longmans). Hartmann, P. 1963 Theorie der Grammatik (The Hague: Mouton). Hegel, G. 1835 Ästhetik (Berlin: Duncker and Humbolt). Hernadi, P. (ed.) 1978 What Is Literature? (Bloomington: Indiana University Press). Heuermann, H., P. Huehn and B. Roettger (eds.) 1975 Literarische Rezeption (Paderborn: Schoeningh).

Schemas for Literary Communication

97

Hirsch, E. D. 1967 Validity in Interpretation (New Haven: Yale University Press). Hirsch, E.D. 1978 "What isn't literature?" in Hernadi (ed.) (1967: 24-34). Iser, W. 1978 The Act of Reading (Baltimore: Johns Hopkins Press). Jakobson, R. 1960 "Linguistics and poetics" in Sebeok (ed.) (1960: 350 -377). Jakobson, R. 1968 "Poetry of grammar and grammar of poetry", Lingua 21, 597-609. Jauss, H. R. 1970 Literaturgeschichte als Provokation der Literaturwissenschaft (Frankfurt: Suhrkamp). Jenkins, J. J., W. D. Mink and W. A. Rüssel 1958 "Associative clustering as a measure of associative strength", Psychological Reports 4, 127-136. Johnson, N., and J. Mandler 1980 "A tale of two structures: Underlying and surface forms in stories", Poetics 9, 51 - 86. Kammann, R. 1966 "Verbal complexity and preferences in poetry", Journal of Verbal Learning and Verbal Behavior 5, 536-540. Keele, S. 1973 Attention and Human Performance (Pacific Palisades: Goodyear). Kent, G. H., and A.J. Rosanoff 1910 "A study of association in insanity", American Journal of Insanity 67, 37-96. Kintsch, W. 1974 The Representation of Meaning in Memory (Hillsdale, N.J.: Erlbaum). Kintsch, W. 1977 Memory and Cognition (New York: Wiley). Kintsch, W., and T. van Dijk 1978 "Toward a model of text comprehension and production", Psychological Review 85, 363-394. Kintsch, W., T. Mandel and E. Kozminsky 1977 "Summarizing scrambled stories", Memory & Cognition 5: 547-552. Kintsch, W., and D. Vipond 1979 "Reading comprehension and readability." In Nilsson (ed.) (1979: 329 -365). Kloepfer, R. 1975 Poetik und Linguistik. Munich: Fink. Koch, W. A. 1978 "Poetizität zwischen Metaphysik und Metasprache", Poetica 10, 275 -341. Levin, S. 1962 Linguistic Structures in Poetry (The Hague: Mouton). Levin, S. 1963 "Deviation - statistical and determinate - in poetic language", Lingua 21, 276-290. Levin, S. 1965 "Internal and external deviation in poetry", Word 21, 225 -237. Lotman, J. 1972 Die Struktur literarischer Texte (Munich: Fink). Lotman, J. 1976 The Analysis of the Poetic Text (Ann Arbor: University of Michigan Press). Lukäcs, G. 1956 Der historische Roman (Berlin: Aufbau).

98

Robert de Beaugrande

Mandler, J . 1979 "A code in the node", Discourse Processes 1 , 1 4 - 3 5 . Mandler, J . , and N. Johnson 1977 "Remembrance of things parsed: Story structure and recall", Cognitive Psychology 9, 111-151. Meyer, B . J . F . 1975 The Organization of Prose and its Effects on Memory (Amsterdam: North Holland). Mukarovsky, J . 1964 "Standard language and poetic language" in Garvin (ed.) (1964: 17-30). Munsinger, H., and W. Kessen 1964 "Uncertainty, structure, and preference", Psychological Monographs 79/9. Neisser, U. 1976 Cognition and Reality (San Francisco: Freeman). Nilsson, L. G. (ed.) 1979 Perspectives on Memory Research (Hillsdale, N . J . : Erlbaum). Ny, J.-F. le, W. Kintsch (eds.) 1982 Language and Comprehension (Amsterdam: North Holland). Paris, B . J . 1974 A Psychological Approach to Fiction (Bloomington: Indiana University Press). Paris, B . J . 1981 "The inner conflicts of Measure for Measure", Centennial Review 25, 266 - 2 7 6 . Pavlov, I. 1927 Conditioned Reflexes (London: Oxford). Piontkowski, U., and N . Groeben 1975 "Konkretheit und Abstraktheit in der Rezeption deutscher Lyrik" in Heuermann, Huehn, and Roettger (eds.) (1975: 2 1 5 - 2 3 3 ) . Riffaterre, M. 1959 "Criteria for style analysis," Word 15, 154-174. Riffaterre, M. 1960 "Stylistic context", Word 16, 2 0 7 - 2 1 8 . Riffaterre, M. 1971 Essais de stylistique structurale (Paris: Flammarion). Rosch, E. 1977 "Human categorization" in Warren (ed.) (1977: 3 - 4 9 ) . Rothkopf. E., and E. Coke 1961 "The prediction offree recall from word association", Journal of Experimental Psychology 62, 4 3 3 - 4 3 8 . Rumelhart, D. 1975 "Notes on a schema for stories" in Bobrow and Collins (eds.) (1975: 2 1 1 - 2 3 6 ) . Schank, R., and R. Abelson 1977 Scripts, Plans, Goals, and Understanding (Hillsdale, N . J . : Erlbaum). Schank, R., N . Goldman, C. Rieger and C. Riesbeck 1975 Conceptual Information Processing (Amsterdam: North Holland). Schank, R., M. Lebowitz and L. Birnbaum 1978 Integrated Partial Parsing (New Haven: Yale Computer Sciences Technical Report 143). Schmidt, S. J . 1971 Aesthetizitaet (Munich: Bayrischer Schulbuchverlag). Schmidt, S . J . 1982 Foundations for the Empirical Study of Literature (Hamburg: Buske). Searle, J . 1969 Speech Acts (London: Cambridge).

Schemas for Literary Communication

99

Sebeok, T. A. (ed.) 1960 Style in Language (Cambridge: MIT Press). Shiffrin, R., and W. Schneider 1977 "Controlled and automatic human information processing II: Perceptual learning, automatic attending, and a general theory", Psychological Review 84, 127-190. Skinner, B. F. 1938 The Behavior of Organisms (NY: Appleton-Century-Crofts). Spiro, R. 1977 "Remembering information from text: The 'state of schema' approach" in Anderson, Spiro, and Montague (eds.) (1977:137-165). Spiro, R., B. Bruce and W. Brewer (eds.) 1980 Theoretical Issues in Reading Comprehension (Hillsdale, N . J . : Erlbaum). Stein, N., and C. Glenn 1979 "An analysis of story comprehension in elementary school children", in Freedle (ed.) (1979: 53-120). Stempel, W.D. (ed.) 1968 Beiträge zur Textlinguistik (Munich: Fink). Todorov, T. 1970 Introduction à la littérature fantastique (Paris: Seuil). Thorndyke, P. 1977 "Cognitive structures in comprehension and memory of narrative discourse", Cognitive Psychology 9, 77-110. Tulving, E., and D. Thompson 1973 "Encoding specificity and retrieval processes in episodic memory", Psychological Review 80, 352-373. Underwood, B., and J. Freund 1968 "Errors in recognition, learning, and recall", Journal of Experimental Psychology 78, 55-63. Warren, N. (ed.) 1977 Advances in Cross-Cultural Psychology (London: Academic). Watkins, M.J., and E. Tulving 1975 "Episodic memory: When recognition fails", Journal of Experimental Psychology: General 104, 5 - 2 0 . Weinrich, H. 1973 Literatur für Leser (Stuttgart: Kohlhammer). Whewell, W. 1819 Elementary Treatise on Mechanics (Cambridge, England: Cambridge University Press).

CSABA PLÉH

On Formal- and Content-Based Models of Story Memory Rationale for a comparative

study of different models of story

memory

During the last decade, many attempts were made to characterize the role of story structure in the selective recall of story components. From the point view on the classical Bartlettian (1932) tradition of the psychology of memory these are all attempts towards a more objective description of story schemas than the ones provided by Bartlett. In the Bartlettian view on memory, the very concept of schematization has at least two possible interpretations. In one of them - in the static conception there is a temptation to identify schemata with the structure of the stimulus material, while in the other interpretation, the schematizing activity of the subject, in the form of the elaborations, transformations of the stimulus material is understood under the heading of schemata. This latent dichotomy also exists in connection with models of story memory. Part of these models, especially the so called story grammars, are tempted to make the stimulus side, i.e. the structure "projected into" the story itself responsible for the shematization, while some other, openly content based models are more or less clearly emphasizing the schematizing activity of the subject. In spite of this crucial divergency between different models of story memory, very few attempts were made for a direct comparison of them. The prototypical research strategy in the last decade has been either a preoccupation with one's pet model on the empirical side, where any sings of structural effects on the recall of different parts of a story have been treated as proofs to the specific model tested without due regard to competing models. (The most vivid examples are the empirical studies testing different versions of story grammar all elaborating on Rumelhart's original model put forward in 1975; see Mandler and Johnson, 1977, Thorndyke, 1977, Stein and Glenn, 1979.) The other much exploited possibility has been the theoretical comparison of the different models usually showing that the formal models are only quasi formal (see e. g. the critic over story grammars by Black and Wilensky, 1979), they do not provide for the real choices made during analysis and supposedly during parsing of the stories (Beaugrande and Miller, 1980), and if they are conceptually analyzed, they also lead to an action based model, where, the "formal structure" of the story is organized according to the problem solving activity of the protagonist (see

On Formal- and Content-Based Models

101

Beaugrande and Colby, 1979, Siklaki, 1981, and even an admission of this by Rumelhart, 1980 himself). At the same time there were very few attempts to coordinate the empirical and the theoretical endeavours by allowing the different models to compete in the explanation of the same empirical material. (One of the rare exceptions is the work of Black and Bower, 1980 to which we will return later.) The main purpose of the study reported here was to remedy the lack of predictive power and testability of the story memory models as most clearly exposed by Thorndyke and Yekovich (1980) by following the treatment suggested by Broadbent (1973) for the entire body of cognitive psychology: by putting the different models into direct competition. This was coloured by a further consideration: in view of the debates concerning the extent of schematization and its suggested task and situational dependeny (see Kintsch, 1978 for a program entailing this consideration), we also wanted to see whether the relative explanatory power of the different models depended on the text processing task itself.

METHODS The text and its analysis A Hungarian legendary tale from the collection edited by Hoppal (1973) was selected as the basic material for our experiments. A story with a very clear point was selected, which has familiar actors and familiar topics, while its specific plot was assumed to be unknown to our subjects. The purpose of selecting a culturally familiar frame, but an unfamiliar plot was that in this way the mobilization of all cultural patterns and "top down" effects can be expected without exaggerated distortions. The text was segmented in a quasi-propositional way, as shown in the Appendix. The main deviation from a strictly propositional representation had been the inclusion of the beginning of dialogues in one unit together with the matrix sentence. This was necessitated by the nature of one of the text processing tasks, where the subjects had to construct summaries for the story by using the segmentation units as entire units. Altogether, five different tasks were used as orienting tasks for the processing of the text, and with every task two protocols were obtained from each subject. There was an intentional learning group, where the subjects were instructed to read the story in order to be able to recall it later; in the Scaling and Moral groups, Ss had to rate the story on semantic-differential like scales (interestingboring, conservative-enlightened etc.) and to write a short moral of the story, respectively. In all these groups, written recalls were collected after half an hour and two and a half months. Two other groups had to construct a summary of the story by omitting half of the units, or by selecting half of the units as important

102

Csaba Pleh

ones using a modified version of the procedure of Johnson (1970). (Thus the orienting task here was directed on either the OMISSION of unessential things, or the INCLUSION of essential things.) An unexpected immediate recall followed the construction of summaries in these groups. Ss in all these groups had been highly motivated college students or entrance candidates, who were tested in groups of 20 to 50 people in classroom situations. Altogether, the results are based on the protocols of over 200 Ss. Protocols were coded for the presence of the units of segmentation according to content, and all the analyses reported below are based on this coding. (Except in the case of the summaries, where, the summary itself was the "coded material"). The text was analyzed according to several different models of text structure, and predictions for the recall pattern were obtained on the basis of these analyses. Four different versions of story grammar were considered. According to the models of Rumelhart (1975), Thorndyke (1977) and Mandler and Johnson (1977) structural trees were constructed for the story, and basically the hierarchy prediction a la Thorndyke was tested; i.e. the expectation that the more nodes dominate over a given textual unit, the less its probability of recall. In the analysis of Stein and Glenn (1979) there is no attempt to give a hierarchical structure to the stories, since, in contrast to the other models, EPISODES are assumed to be coordinated here, rather then subordinate to each other. From this model, only predictions as to the differentiated recall of different story grammar categories (e. g. Attempts, Setting, Resolutions) were tested. A narrative production model was also adapted to the purposes of analysis. Labov and Waletzky's (1967. Labov, 1972) approach treats the narrative core of a story as a set of clauses which cannot be moved with respect to each other without changing the meaning of the story. Formally, these narrative clauses which reflect the actual temporal sequence of real life events, are arrived at by determining the displacement set - the moveability limits - of each clause, and then analyzing the overlaps of these sets. This method was adapted to our purposes by determining the displacement set of each segmentation unit, and then taking the assumption that the non-moveable units would be better recalled than the moveable ones. From the general model of text structure and processing put forward by Kintsch and van Dijk (1975, 1978) the folloving aspects were used to obtain testable predictions. In their model, the macrostructure of a text is constructed from the "explicit text base" by the application of rules of deletion, generalization, selection and construction. We tried to apply to each unit deletion rules which are characteristic to simple stories. Mental reasons, goals and consequences; alternatives; usual subsidiary actions; state descriptions which are not conditions for actions; metadescriptions; and dialogues can be deleted according to this model from action discourses like the one in question. By applying these rules, each unit was characterized by how many deletion rules apply to it, and the different types of units were compared on this basis.

O n Formal- and Content-Based Models

103

From the model of story structure and recall proposed by Schank (1975; Schank and Abelson, 1977), similary to Black and Bower (1980) the following aspects were utilized. During story processing, the reader attempts to construct a "causal chain" relating the different events (conceptualizations) of a story into a model expressing the purposes of the protagonist and the consequences of his actions. The principal chain would direct recall, with the consequence that actions leading away from the principle chain would be more likley to be forgotten. Each unit was characterized by its distance from the main causal chain (0, 1 , 2 or 3), and the probablity of the recall of units categorized in this way was compared.

RESULTS Several different statistical methods were tried for the comparison of the different predictions. First, ANOVAs were performed for the average percentage of recall (or inclusion in the summary) over different unit types in each situation, according to each of the predictions. Table 1. summarizes the results of these comparisons. As the distribution of the plusses (significant effects) indicates it, the story grammar hierarchy had clear effects only in the case of subjects who had been working with the stories in a piecemeal conscientious way. Story grammar categories - which are "content based" - however, had a much clearer effect. The differences found roughly suggested an Attempt > Beginning > Solution > State description > Inner reactions ordering. The lack of a significant difference in the first recall of the intentional group according to this analysis can be attributed to a ceiling effect, their first recall being highly accurate in general. The "fixed place" units, (Labov-Waletzky) the non-deletable ones, (Kintschvan Dijk), and the ones belonging to the main causal chain (Schank) were all better recalled, as the last three lines show it on Table 1. However, the effect of these latter variables was weaker in situations where the formal models worked better. Although this type of analysis already clearly showed the superiority of the content based models over the (quasi)-formal ones, it still had to be seen what is the relationship between the predictive power of the different non-formalistic models. For the purposes of this comparison, stepwise regression was used in each situation, basically following the method of Black and Bower (1980). The dependent variable was in each situation the recall percentage of a given unit, while the predictor variables were the numeric characteristics of each unit in every conceivable model. There was only one important difference from the method of Black and Bower: they used the average recall percentage of a given unit to construct the predictor variable for each individual unit concerning story grammar categories. In order to avoid the danger of circularity here, we have included only those predictors in the model, where a meaningful ordinal predic-

104

Csaba Pléh o u Pü s o se

2.00). As may be seen, several of the texts not exhibiting the predicted night-journey quadratic trend for primary process do have a number of significant autocorrelations. Most notable is Novalis' Hymns to the Night. Thus, this text does seem to symbolize alteration in consciousness but not in terms of the simple night-journey plot. By the same token, a large number of autoregressive parameters for Conrad's Heart of Darkness are significant. This suggests that the book does deal with the problem of good and evil but not in a simple fashion. The lags for which autoregressive parameters are significant does not seem to be important. The point is that if autoregressive parameters are significant, this implies that there is some pattern to the type of content being studied. If a pattern exists, then we may infer that

Narrative Pattern Analysis

-o c

> w

e

-a c

C S » • s C rt c wi Ü O wi ü "O (j 4» M 4> ( j 13 . 3 C « 3 c " bùM 4J bßjO 4>äß C ' C ' .SS C 'C , u -S3

« u

• Ì JS

•>

o

Ì C ' .2 i s c v 3 O . >Q

S

£

§

1 1 0 O «

l 1 .a

1 2 rt P on U

c 3 E

S w

%

to c/i « w>

< £ > pä w cn PQ O

=

a

o U

§a 'S§

1-1

c 2

>H Q 2 SS H > U ei w W ö S Q O

6

o U

a3

rt

(J '3 I 3 a U

5

-D £

>

e " J •3

X

o •a rt

c o

"2

c

= 3 1

S s

i l JB C a 3 55 cj « C O

2

r

z

t

-a s o

v> u

« o

c * W

g £

-O T i O

«Ì>• -a

I

P

«2

Q ' W w 2

^

s

3 o •9 b c o u .ü .S c 3 ya 5 «h u

PL, -d

Life vs. Art: the Interpretation

221

filmed are 'unobtrusive mediations'. Events filmed 'live' as they unfolded in front of, and to a large degree for the sake of the camera are 'media-events'. Events scripted, staged and directed by an 'author' working with actors are termed 'scripted events'. If we settle on the first alternative, we are most likely to feel justified in making attributions about the persons in the film (their characteristics, their feelings, their relationships, and so forth). We may feel that we are essentially in the position of an unobtrusive, possibly invisible observer of the actual events, and possessed of most of the information that would be available to such an observer (this feeling would be greatest when viewing sound/color film), watching people "caught in the act of being themselves." There are of course relatively few occasions on which 'invisible' cameras record the behavior of 'actors' who are unaware that they are being filmed. Among the most typical examples of such recordings are the films taken by ceiling-mounted cameras in banks, department stores and gambling casinos. Although many of those who enter these establishments presumably are aware that they are being filmed, it is probably safe to say that almost no one who does not have a 'professional' interest in the practice (from either side of the law) would take the presence of the cameras into account and modify his or her behavior accordingly. In other words, the camera - and the viewer of the film is in the position of an observer of events which can be assumed not to have been shaped by any intention of communicating to the observer: natural events which can be informative about the attributes of the 'actors.' But the viewer of such a film is not in exactly the same position as the camera which recorded the events; there are other factors in the interpretive equation. If we are shown a length of film taken with an 'invisible' camera, while we might reasonably view the behavior of the 'actors' as unintentionally informative, we can have no such confidence about the behavior of the mediator. That is, we have reasons to wonder why (1) the film was shot, (2) these particular segments were selected, and (3) they are being shown to us. Even taking an example such as the camera in a bank, where we easily understand why the film was shot, and do not assume that anyone intended to 'tell' us anything through the film (other than the identity of a bank robber, should one appear), we are still likely to wonder why this particular segment was edited out, and why it is being shown to us. In other words, while we may believe that people may act completely 'naturally' in front of an 'invisible' camera, we also know that intentional choices govern the making, editing and exhibiting of any film. Documentary styles We do not often encounter films which fit the exacting criteria of 'invisible' mediation and 'actors' who are unaware of being filmed. Much more common are films made by visible but unobtrusive filmmakers (camera operators often

222

Larry Gross

accompanied by sound recorders) who attempt to fade into the background and thus minimize the impact of their presence. In such situations we will be less confident in making attributions about the 'actors' than in the case of 'invisible' mediation, since we will feel that the behavior we observe was somewhat constrained by the subjects' knowledge that they were being filmed. That is, their behavior may be less informative because we know it is at least in part 'messageful'. The 'actors' must be presumed to be controlling their 'presentation' to the filmmakers (and thus to the viewer of the film) through choices of commission and omission which reflect communicative intent. The typical practice of documentary filmmakers, particularly in the 'direct cinema'Iánema verite tradition (see Barnouw, 1983, for a discussion of these related schools of filmmaking), rests on the belief that 'actors' who have become accustomed to the presence of the camera will revert to a state of relatively natural behavior and simply 'be themselves.' In fact, it has even been suggested that, "the presence of the camera [makes] people act in ways truer to their nature than might otherwise be the case" (Barnouw, 1983: 253). A pioneer of direct cinema, Robert Drew, thus described the arrangements for his 1960 film, Primary: I settled on a young senator, John F. Kennedy, running for President in a Wisconsin primary against another senator, Hubert Humphrey. I told both Senators that for this new form of reporting to work we would have to live with them from morning to night, shooting anything we wanted to shoot, day after day. They could not know or care when we were shooting, and that was the only way we could capture a true picture of the story. When Kennedy raised an eyebrow I said, "Trust us or it cannot be done." Kennedy agreed. Humphrey agreed." (1982: 19)

This account illuminates several important components of documentary film practice which are described as follows by video documentarían Julie Gustafson: . . . (W)e spend as much time as possible with our subjects so that they will feel as comfortable as possible with us during the taping period. Although we recognize the influence of our presence, we never interfere in the natural flow of events by, for example, asking the subjects to do things over again for the c a m e r a . . . Finally, to create a powerful and attractive story, we use conventional dramatic techniques of filmmakers and playwrights to finally structure our program to draw our audience in. We structure our shooting around strong individuals and clear action. We also, in editing, use the classic methods of dramatic introduction, exposition of the problem, and denouement. (1982: 63 f)

This approach to documentary demonstrates the mixture of presumably natural behavior on the part of the 'actors' and the obviously communicative intentions of the filmmakers in choosing what to film and how to edit and structure the resulting narrative. Clearly, it will be appropriate to regard the final product as a story we are being told by the filmmakers, based to an indeterminate degree upon the 'reality' of the participants' dispositions, motivations and circumstances. However unobtrusive and sincere the filmmakers might be in their

Life vs. Art: the Interpretation

223

attempt to capture the way things were, we are faced with a narrative constructed within the conventions of dramatic realism. Some documentary filmmakers respond to the inevitability of participants' 'reactance' and their own use of narrative conventions by including 'reflexive' devices in order to alert the audience to the fact that they are watching events which are to some degree 'unnatural.' Gustafson again: A ... distancing device is to include glimpses of the mechanics of the shooting process in the final work, a practice drived from the techniques of both cinema verite and American direct cinema. We may tape ourselves setting up the lights and thereby see the effects of our presence on the subject's life. We also include shots of the microphone or other equipment so that the viewer occasionally sees the complete reality of the scene. (1982: 64)

Other styles of documentary filmmaking avoid the inclusion of explicit reflexive signals and even encourage the participants to collude in maintaining the illusion of invisible mediation. The film Salesman by the Maysles brothers can serve as an illustration. Barnouw describes the making of the film: The main characters of Salesman were four door-to-door Bible salesmen. The principal focus came to center on one - Paul. With approval of the company they represented, the Maysles brothers accompanied them on their selling efforts. The salesmen obtained names of potential customers at community churches. At the time of the home calls, the Maysles sometimes started filming as the salesman knocked. They would shoot and record the opening talk at the front door. Then the salesman would introduce the Maysles, and one of them would explain: "We're doing a human interest story about this gentleman and his three colleagues. And we'd like to film his presentation." (The word "sell" was not used.) Usually the reply would be: "Oh, a human interest story. Ok, come in." (Very few people declined to be filmed. Once the Maysles were inside, there was seldom an obstacle to continued filming. Since the salesman was really performing, it was easy for the Maysles to become relatively invisible. Afterwards they would ask the customer to sign a release. Only one refused. (1983: 241)

The film does not, however, include any scenes in which potential customers seem to be aware of the presence of the filmmakers - even when they first answer the door and are confronted with a salesman accompanied by a cameraman and a soundman — and several scenes suggest that the Maysles induced the customers to 're-enact' their behavior, as the invitation to the salesman to enter the customer's home is sometimes shot over the customer's shoulder from inside the home. In other words, participants who have not had an opportunity to become accustomed to the presence of the film crew (as stipulated by practioners of direct cinema), are none the less encouraged to 'role-play' their 'natural' behavior in front of the camera. The basic rule of such filmmaking is to discourage the participants from looking at the camera or speaking to the filmmakers, and to edit out any instances in which they do so. An interesting exception to this 'rule' of documentary filmmaking is the genre of 'personal' documentary (often autobiographical, usually family-centered) in which the filmmaker is a major participant in the real life events being depicted. A good example is Ira Wohl's Best Boy, a film about the filmmaker's retarded cousin in which Wohl's relationship to the participants is a major concern of the film, and in which the 'actors' frequently are shown to be interacting with the

224

Larry Gross

director and the film crew. Such films might be seen as 'professional home movies' in which private lives are presented to public audiences. Conversely, amateur "home movies" might be categorized as a special genre of documentary filmmaking aimed at private audiences, in which participants are often encouraged to role play 'naturally' in front of the camera. However, as Chalfen has shown (1975), home movie 'actors' rarely obey the injunction not to look into the camera. Instructive manuals on home movie making often prescribe strategies much like those adopted by direct cinema documentarians: . . . To capture them un-self-conscious and relatively uninhibited, your best bet is to plan your shooting for occasions when your intended subjects are engrossed in some sort of activity. (Quoted in Chalfen, 1975: 92)

But Chalfen's analysis of actual home movies revealed a different pattern: (1) There is a lot of waving at the c a m e r a . . . (2) Very frequently one sees people, especially children, walking directly toward the camera, sometimes directly into the lens . . . (3) There is an extraordinarily large amount of staring into the lens of the c a m e r a . . . (4) People will strike a pose or present a "camera-face" for an operating movie camera. (1975: 98)

Media events While documentary filmmakers tend to focus on dramatic moments in the lives of their subjects, and home movie-makers are drawn to 'special' occasions (holidays, birthday parties, vacations), none of these are properly seen as having occured in order to be filmed. Media events can be described, conversely, as 'natural performances' which occur in the real world in large part, if not entirely, in order to be seen (simultaneously or subsequently) by viewers who are not present at the event. The most frequent examples of media events are those produced by individuals and groups engaged in political activities. Those who are in established positions which confer on them the power to command media attention will use conventional forms such as ceremonial events, press conferences and speeches to reach audiences far beyond those assembled in front of them. Often they will engage in public activities - touring factories, walking the streets - primarily in order to attract media attention, and audience exposure. Those not blessed with 'official' status are usually required to engage in 'unofficial' but equally conventional means of obtaining such attention: marches, demonstrations, civil disobedience and so forth. In all of these the 'actors' are primarily motivated to convey messages to audiences which will only observe the event if it is transmitted, and will only know it as it is shown to them. Political ads on American television represent a highly evolved genre of seemingly naturalistic media events in which candidates are filmed on location, in live interaction with 'real' people; and many hours of film may yield just the right 30 or 60 seconds of image-enhancing 'character revelation.'

Life vs. Art: the Interpretation

225

Fictional film It may only be a short step, in many respects, from media events to staged, scripted, truly fictional narratives, but the distance between these levels is significant. For, when we are dealing with the interpretation of events which have been 'authored' as well as mediated, we have to recognize a different set of opportunities, constraints and conventions from those we take into account in the case of more or less natural events. The interpretation of fictional narratives calls upon our knowledge of dramatic conventions. These conventions may be nearly the same as those we use in interpreting documentary narratives (and consequently might lead us to similar interpretations), and this is not surprising. We have seen that documentary filmmakers utilize dramatic conventions drawn from fictional genres. It is equally true that naturalistic conventions in fiction aim precisely at evoking attributional stereotypes in order to convey 'lifelikeness' to characters and situations. Culler describes the operation of naturalistic fiction which grounds its conventions in such attributional stereotypes (although he speaks of novels, the point applies at least as well to film): Citing this general social discourse is a way of grounding a work in reality, of establishing a relationship between words and world which serves as a guarantee of intelligibility; but more important are the interpretive operations which it permits. When a character in a novel performs an action, the reader can give it a meaning by drawing upon this fund of human knowledge which establishes connections between action and motive, behavior and personality. Naturalization proceeds on the assumption that action is intelligible, and cultural codes specify the forms of intelligibility. (1975: 142f)

However, the interpretations need not be the same; fiction is not invariably faithful to the naturalism of "general social discourse." We may 'know' that the cowboy in the black hat is the villain without also believing that anyone we see in real life wearing a black hat is a bad guy. The function of genre conventions is essentially to establish a contract between writer and reader so as to make certain relevant expectations operative and thus to permit both compliance with and deviation from accepted modes of intelligibility. (Culler, 1975: 147)

The parallel between written and visual narratives should not be extended too far; there are significant ways in which film and video must be distinguished from stories we read in books and even from those we see acted on stage. Although there are many important differences between these media, I am here concerned with the fact that films invariably present us with a wealth of potential sign-events which we are more likely to see as natural than as symbolic. Image/sound recording presents us with 'segments of reality' which include many elements which are not reasonably taken as having been created or articulated by the author(s) for the purpose of conveying meanings to us, but which are taken to be, simply, real. Although there are striking examples of films which utilize unmistakably artificial sets - I am thinking particularly of early, 'stagey' films like The Cabinet of Dr. Caligari and Metropolis - the vast

226

Larry Gross

majority of fiction films anchor their verisimilitude in the appearance of real world settings. This apparently inescapable aspect of film - the camera throws a net which indiscriminately gathers in whatever is before it (or so many viewers believe) can be discerned in a filmic form of the "general social discourse" Culler described. In a revealingly naive example of inappropriately attributional interpretation, film critic Roger Angell congratulates the makers of Kramer vs. Kramer for their attention to naturalistic detail: There are a great many New York vistas and street corners and storefronts in "Kramer vs. Kramer", but instead of our getting that intrusive little thrill of participation we usually feel when we spot something familiar in a movie, we are more subtly and deeply satisfied here, because the places help us know the people in the movie - people we have come to care about. When Billy Kramer cuts himself badly in an accident in a Central Park playground, I instantly thought, What's the nearest hospital - Lenox Hill? Take him to Lenox Hill! And Ted Kramer sweeps the boy into his arms and runs to Lenox Hill Hospital. (1979: 81)

Although I presume Angell is sophisticated enough about film not to have called out advice to the characters on the screen, neither he, nor presumably his editor at The New Yorker, realized that the filmmakers would hardly have been directing their work at an audience of Upper East Side Manhattan residents who could properly appreciate their skill at "Getting Things Right", as Angell's review is titled. It is quite possible, of course, that the authors of Kramer vs. Kramer were attempting to achieve the sort of detailed verisimilitude Angell so admired. But whether or not they consciously intended to do this, the response to their film evokes the echoes of debates over the nature and scope of realism and naturalism that raged throughout much of the 19th century. Although the authors of realistic and naturalistic novels were frequently accused of attempting to unselectively mirror the world, the accusations were bitterly denied. . . . A contemptible reproach which they heap upon us naturalistic writers is the desire to be simply photographic. We have in vain declared that we admit the necessity of an artist's possessing an individual temperament and a personal expression; they continue to reply to us with these imbecile arguments about the impossibility of being strictly true, about the necessity of arranging facts to produce a work of art of any k i n d . . . We start, indeed, from the true facts, which are our indestructible basis; but to show the mechanism of these facts it is necessary for us to produce and direct the phenomena; this is our share of invention, here is the genius in the book. (Zola, 1880)

In his book, What Is Art?, Tolstoy criticizes the sort of imitation Angell admires: The essence of this method consists in rendering the details which accompany that which is described or represented. In the literary art this method consists in describing, down to the minutest details, the appearance, faces, garments, gestures, sounds, apartments of the acting persons, with all those incidents which occur in life. . . . It is just as strange to value the production of art by the degree of its realism and truthfulness of details communicated, as it is to judge of the nutritive value of food by its appearance. (Tolstoy, 1897)

Life vs. Art: the Interpretation

227

In their defense against the charge of indiscriminate imitation Zola and other novelists were concerned to distinguish the craft of realistic writing from the 'mechanical' naturalism of photography. The writer, de Maupassant argues, recognizes the necessity of selection and composition and thus, The realist, if he is an artist, will seek to give us not a banal photographic representation of life, but a vision of it that is more vivid and more compellingly truthful and even reality itself. (1888)

Photography itself, however, was less easily exonerated from the accusation of ecxessively exhaustive realism. In his famous review of the Salon of 1859, in which photographs were first admitted to the exalted company of officially sanctioned art, Baudelaire denounced the public for their acceptance of photography: . . . (T)he current credo of the sophisticated public, . . . is this: "I believe in nature and I believe only in nature. . . . I believe that art is and cannot be other than the exact reproduction of n a t u r e . . . Thus the industry that could give us a result identical with nature would be the absolute form of art." A vengeful God has granted the wishes of this multitude. Daguerre was his messiah. And now the public says to itself: "Since photography gives us all the guarantee of exactitude that we could wish (they believe that, the idiots!), then photography and art are the same thing." (1859)

Interpretation and aesthetic evaluation As this discussion begins to indicate, the 'mechanical naturalism' of film influences not only our interpretation of the narrative, but it also has consequences for aesthetic appreciation and evaluation. Aesthetic evaluation is greatly influenced by the audience's assessment of the creator(s)'s skill and control over the medium. That is, we appreciate those elements of a work which manifest the artist's successful realization of intentions (cf. Gross, 1973, for a fuller discussion of the relation between competence and aesthetic appreciation). When we encounter elements of a work which deviate from the conventional, we often try to determine whether the deviation results from a successful intention to do something unconventional (and we then will ask whether the innovation is aesthetically pleasing), or whether it reflects the artist's failure to achieve the conventional (in which case we are unlikely to be favorably disposed towards the artist and/or the particular element). This process of assessment and evaluation parallels the attributional analysis Heider ascribed to the "naive psychology of action:" . . . (C)an and try are the conditions of action. Thus, our reactions will be different according to whether we think a person failed primarily because he lacked adequate ability or primarily because he did not want to carry out the action. (1958: 123)

The interpretation and the aesthetic evaluation of film narratives thus involves attributions about the authors which consist of our assessment of their intentions and their ability to realize those intentions. But, as I have suggested, the nature of the medium creates particular complexities and ambiguities for such

228

Larry Gross

assessments. Because the 'raw material' of film includes real people and, often, segments of the real world captured on film, we are less able to determine the boundaries of the artists' control and thus the degree of their aesthetic accountability. This point can be illustrated in the case of actors as well as 'authors.' Again, a somewhat naive film critic gives us an example; this time it concerns Peter Bogdanovich's film version of Daisy Miller: I knew that I'd been moved by the film, disarmed by its surprising wit and, particularly, that I'd admired Cybill Shepherd's Daisy. But f o r all that, I could not honestly tell whether or not Cybill Shepherd could act. W h a t she had done in the film to embody Daisy's ingenuousness and enthusiasm had seemed so unconscious, even accidental, that I could with no confidence decide whether I had been responding to Cybill Shepherd's characterization of Daisy's gaucheness or simply to Cybill Shepherd's o w n gaucheness as an actress. Was Miss Shepherd a thoughtful performer or merely a shrewdly-selected actress caught, held and defined b y a director's c a m e r a - e y e ? . . . We are left not knowing whether she is in control of her performance, whether she knows what she is doing. (Kareda, 1974)

When viewing documentaries, we often are worried that the participants' 'naturalness' has been contaminated by their desire to present a particular persona to the camera. In fictional films, the opposite dilemma faces the viewer who fears that the persona depicted may reflect the actor's 'real' self rather than a crafted characterization. This dilemma is confounded by the commercial practice of 'type casting', in which actors are repeatedly cast in roles which call for similar attributes, thus depriving audiences of the benefit of 'control comparisons' across films. The evaluation of film directors' skill also involves our notions of which elements in the work are attributable to the filmmaker and of the extent of the filmmaker's accountability for all of the elements in the film. As we have seen, Peter Bogdanovich may receive credit for choosing an actress who 'naturally' embodies the personality required for the character of Daisy Miller; but only, presumably, if we do not give Cybill Shepherd credit for skillful acting. Because films are so often shot on location they tend to include parts of the real world as background; does the director deserve credit for the scenery? Does John Ford deserve congratulations for the beauty of Monument Valley, the backdrop for many of his Westerns, or merely for having chosen the location and photographed it so dramatically? Reality, in other words, may become an aesthetic component of film art beyond the conventional level of "general social discourse." The issue of authorial accountability can be briefly illustrated with the example of color. As color came to replace black and white as the standard for films, audiences learned to expect and accept it as 'truer to life' and thus as part of the 'natural background.' In this sense color is rarely a 'sign element' of film, and directors are not given credit for the successful achievement of transparently realistic color. There are instances, however, in which color has been used as a semiotic component in film. One of the earliest, and still best examples is the Wizard of Oz (Victor Fleming, 1939), in which the opening and closing scenes

Life vs. A r t : the Interpretation

229

in Kansas are in drab black and white and the central portions of the film - the Land of O z , with its Yellow Brick Road and Emerald City - are in Glorious Color. The dramatic impact of this symbolic device must have been particularly impressive for audiences as yet unfamiliar with color in films. Although the manipulation of film color has become more common as a framing device (e. g. an opening in sepia under the titles turning to color as the film begins, which tells us we are going from past to present), few directors have attempted to use color as a semiotic component throughout a film. A rare exception is Antonioni, who attempted, unsuccessfully I feel, to use color as a sign vehicle in Red Desert and Blow Up. In Blow Up, Antonioni let it be known, he had the grass painted a brighter shade of green for a crucial scene in a park, thus taking credit, but also becoming accountable for, aspects of the real world not normally included within the scope of authorial control. Conversely, a filmmaker may be given credit inappropriately when the viewer mistakes the author's skill at invention for one at finding appropriate elements (for a discussion of 'making' vs. 'finding' as aesthetic criteria, see Shiff, 1982). F o r the last time, a naive film critic will provide us with an example of attributional error; in this instance the critic is praising Fellini's Amarcord: . . . I can not resist describing the scene that has climbs to the top of a tree in the early afternoon want a w o m a n ! " After an entire day of this, the arrives? A midget nun (where does Fellini find briskly climbs the ladder to the tree and returns

stayed with me longest: when the insane son and starts screaming at the top of his lungs, " I family sends to the asylum for help, and who these people?) whose face we never see, who with the maniac in tow. (Ledeen, 1975: 102)

Ledeen's mistaken belief that Fellini 'found' something he undoubtedly 'made' (perhaps it should be noted that we have no evidence that the person we see is a nun or even a midget; it could easily have been a young boy wearing a nun's habit), illustrates the point that, although events encountered in 'life' and in 'art' may look the same, we need to make different assumptions about the forces that determine their occurence and configuration. Because the conventions of art may be drawn from the "general social discourse" of cultural stereotypes, the conclusions reached may be the same. In order to decide whether an interpretation is attributional (the observer is assessing the event as 'life' - a natural event), or inferential (the observer is assessing the event as 'art' - a symbolic event), one needs to know the grounds on which the conclusions would be justified. If asked how we know something that we have concluded about an event we have observed, we might say that we have based our conclusions on what we know about the way such things happen (attributional interpretation); or we might say that we know it because we are assuming the event was made to happen that way in order to tell us something (communicational inference). The tendency to see documentary films and, often, fictional ones as well, as objective records of events rather than as a filmmaker's statement about events derives from a confusion of interpretive strategies; the unsophisticated assumption that filmed events can be uncritically interpreted as 'natural.' What such viewers fail to understand is that all mediated events are to some degree sym-

230

Larry Gross

bolic. The mediating agent always makes decisions - about what to shoot (and consequently, what not to shoot), and how; and, having shot, about how to edit the footage; and, finally, about when, where, and how to exhibit the edited film. Moreover, the participants in a film are rarely, if ever, just "caught in the act of being themselves." A sophisticated viewer will recognize that the persons, objects, and events in a film are there, at least in part, because the filmmaker included them intentionally; that the sequence of events in the film has been ordered by the filmmakers's intention to say something by putting them in that order (which may not be the order they actually occurred in); and that the overall structure of the film reflects the filmmaker's intention, and ability, to use implicational conventions in order to communicate to viewers competent to draw the appropriate inferences.

References Allott, M. 1959 Novelists on the Novel (New York: Columbia University Press). Angell, R. 1979 "Getting Things Right", The New Yorker, December 24. Barnouw, E. (ed.) 1983 Documentary (New York: Oxford University Press). Baudelaire, Ch. 1859 "Salon Review" quoted in Scharf (1968: 145). Chalfen, R. 1975 "Cinema Naivete: A Study of Home Moviemaking as Visual Communication", Studies in the Anthropology of Visual Communication 2, 2. Culler, J. 1975 Structuralist Poetics (Ithaca: Cornell University Press). Drew, R. 1982 "An Independent with the Networks", Studies in Visual Communication 8, 1 , 1 5 - 2 3 . Gross, L. 1973 "Art as the Communication of Competence", Social Science Information 12, 5. 115-141. Gustafson, J. 1982 "Toward a Cinema of Ideas", Studies in Visual Communication 8, 61-70. Heider, F. 1958 The Psychology of Interpersonal Relations (New York: Wiley). Kareda, U. 1974 "The Signals Movie Actors Give", The New York Times, August 11. Kelley, H. 1967 "Attribution Theory in Social Psychology" in Levine (ed.) (1967: 192-240). Ledeen, M. 1975 "Amarcord", Society 12, 2. Levine, D. (ed.) 1967 Nebraska Symposium on Motivation (University of Nebraska Press), de Maupassant, G. 1888 "Le Roman" quoted in Allott (1959: 71).

Life vs. Art: the Interpretation Scharf, A. 1968

231

"The Technique of Originality: 'Innocence and Artifice in the Painting of Corot, Monet, and Cézanne'", Studies in Visual Communication 8, 4.

Tolstoy, L. 1897 "What is Art?" quoted in Allott (1959: 74). Worth, S. and L. Gross 1974 "Symbolic Strategies", Journal of Communication 24, 4. Zola, E. 1880 "Le Roman Experimental" quoted in Allott (1959: 70).

List of contributors Robert Abelson, Department of Psychology, Yale University, New Haven, CT 06520, USA Robert de Beaugrande, Department of English, University of Florida, Gainesville, Florida 32601, USA Antal Bókay, Faculty of Humanities, Sciences and Education, Janus Pannonius University, Ifjüsäg ütja 6, 7604 Pecs, Hungary Larry Gross, Annenberg School of Communications, University of Pennsylvania, 3620 Walnut St. C5, Philadelphia, PA 19104, USA Laszló Haläsz, Institute for Psychology of The Hungarian Academy of Sciences, 1394 Budapest 62, Pf. 398, Hungary Ravenna Helson, Department of Psychology, University of California, Berkeley, CA 94720, USA Janos Laszló, Institute for Psychology of The Hungarian Academy of Sciences, 1394 Budapest 62, Pf. 398, Hungary Martin Lindauer, Department of Psychology, SUNY College at Brockport, Brockport, NY 14420, USA Colin Martindale, Department of Psychology, University of Maine, Orono, Maine 04469, USA Csaba Pléh, Institute for Psychology, Eötvös Lórand University, Izabella u. 46, 1064 Budapest, Hungary Dean Simonton, Department of Psychology, University of California, Davis, CA 95616, USA

Author Index Abelson, R. P. 2 7 , 2 9 , 3 2 , 3 3 , 3 5 , 3 8 - 4 8 , 38, 41,45,47,48,55,98,103,109,112,115, 116,124 Adams, M . 52,94 Adorno, Th., 16,32 Aeschylus, 194 Aesop, 8 Alba, J. W., 136,137 Albee, E., 44 Allot:, M „ 230 Alloway, Th., 32, 33 Allwood, C. M . , 153,165 Anderson, N . H . , 10,32 Anderson, R. C., 27, 32, 35, 51, 54, 94, 99,

122 Angeli, R., 226,230 Angus, D., 128,137 Anisfeld, M., 53,94 Aristophanes, 194 Arnheim, R., 132,137 Aronson, E., 4,33,35 Artson, T., 188 Asch, S., 4,10,33 Ashby, W. R., 18,33 Auden, W. H . , 59 Ausubel, D., 52,94 Backman, C. W., 6,36 Baddeley, A., 122,123 Baker, R., 73,94 Balzac, H . , 4,72,140 Barnouw, E., 222,230 Barron, F., 16,33,184,191 Barthes, R., 4,23,33 Bartlett, E C., 24,33,52,94,100,111,136, 137 Bauer, M., 78, 82,94 Bauer, W., 94 Baudelaire, Ch., 227,230 Beaugrande, R., 18,20,29,33,49 - 99, 53, 54,55,61,71,74,75,78,82,85,87,92,94, 95,100,101,108,111,115,122 Beaumont, F., 67 Becher, J. R., 14,33 Beckett, S., 44 Bellak, L., 182,191

Berlyne, D . E., 14, 19,33, 80,95,125,134, 136,137 Berkowitz, L., 122 Bien, J., 45,48 Bieri, H „ 16,33 Birnbaum, L., 55, 98 Black, J. B., 41,42,47,48,100,101,103, 105, 111, 115,122,136,137,143,165,183, 191 Bloch, E., 76,95 Blumensath, H . , 34 Boccaccio, 128 Bobrow, D . G., 33,36,95,98, 111, 112, 122,124,143,165 Bodman, P. L., 4,33 Bogdanovich, P., 228 Bokay, A., 3 2 , 2 0 1 - 2 1 4 Bornstein, M . H „ 137,138 Bowen, E „ 190 Bower, G. H . , 27,33, 54,95,101,103,105, 111,115,122,136,137,143,165 Bransford, J. D., 114,122 Brewer, W. F., 26,27,33,35,41,47,48,99, 116,122 Broadbent, D . E., 19,33,101,111 Bruce, B. C „ 45,46,47,48,99 Bruner, J. S„ 13,33,95,180,181 Bungard, 113 Bunyan.J., 172 Burgess, A., 125,137 Canby.H.S., 128,132,135,137 Cantor, N „ 118,119,122,144 Carbonell, J. G., 39,47 Carlsmith, J. M., 4,33 Carroll, J. S., 32,33, 172 Carroll, L., 77 Cermark, M., 114,122 Chalfen, R., 224,230 Chapman, A. J., 4,33,35 Chase, W., 33,35,95 Chatman, S., 1,3,4,5,33 Chatman, J., 140,165 Chaucer, G., 128 Chekhov, A. P., 128,163,190 Cherry, S., 33, 35

234

Author Index

Child, J. L., 4,33 Chomsky, N., 23,33 Cofer, C., 95,165 Coke, E., 53,98 Colby, K.M., 44,47 Colby, B. N „ 55,61, 78,95,101,108, 111 Coleridge, S. T., 172,173 Collier, 137 Collins, A., 33,36,47,48,52,94,95,98, 111,112,122,124 Conrad, J., 172,173,174 Cornelius, D. K., 126,137 Coseriu, E., 74,95 Covello, E., 171,181 Craik, F. I. M., 114,119,122,123 Craik, K. H., 184,185,191 Crane, S., 78 Crothers., E. J., 108,111 Crozier, W. R., 4,33,35 Culler, J., 225,226,230 Cullingford, R. E., 39,47,55,95 Current-Garcia, E„ 128,129,130,132,133, 137 Cybill, S., 228 Cysarz, H., 23,33 Dahl, H., 171,181 Dante, A., 67,168,172 Das, J. P., 4,33 Das, R. S., 4,33 Dashiell, A., 128,135,137 Day, H. J., 137,138 DeCamp, M. A., 4,35 DeCampWilson, T„ 20,35 Deese,J„ 53,95 Dejong, G., 39,47 Derlega, V. J., 192 Deutsch, M., 188,192 DeVillers, P., 90,96 Dery, T., 190 Dickens, Ch., 72,73 Dillon, G.L., 27,33 Donne, J., 58 Dostoevsky, F. M., 128 Drayton, M., 58 Dreiser, Th., 13 Dreyfus, J. H., 40,47 Drew, R., 222,230 Dunphy, D. C., 18 Dyer, M. G., 42,47 Eagly, D. H., 34 Eberhart, R„ 83, 94 Eco, U., 74, 75,96 Eggen, H., 82,96

Ehrenzweig, A., 168,180 Eichenbaum, B. M. 3,33,116,123 Eliot, T. S., 71 Enkvist, N. E., 58,96 Enzensberger, C., 76,96 Erikson, E., 185 Erlich, V., 115,123 Estes, W., 111,112 Euripides, 194 Faulkner, W., 55,128 Faulstich, W., 78,96 Fechner, G., 74,96,115,123 Fellini, F., 229 Fernandez, F., 1,2,33 Festinger, L., 16,33 Fielding, H., 72 Fischer, R., 171,181 Fish, S., 33 Flemins, V., 228 Fletcher.J., 67 Fontane, T., 57 Forgas, P„ 109,111 Forrester, J., 201,204,214 Frances, R., 4,33 Franks, J.J., 114,122 Freedle, R., 33,36,94,96,111, 112, 123,124 Frenkel-Brunswik, E., 32 Freund, J., 53,99 Freud, S., 2,6,15,34,201,202,203,204,205, 206,207,208,209,210,211,212,213,214 Frey, E„ 89,92,96 Frye, N., 74,96 Fukazawa, H., 175,179 Garfinkel, H., 108,112 Garvin, P., 96,98,123 Genette, G„ 23,34 Gergen, K„ 180,181 Gergen, M„ 180,181 Gide, A., 117 Gibson, E., 51,96 Glenn, C. G„ 26,36,54,55,99,100,102, 104,112,116,124 Goethe, J.W., 67,80 Goffman, E., 217 Goldman, N., 98 Goldstein, K., 169,180 Gombrich, E. H., 165 Gomulicki, B., 51,96 Gordimer, N., 190 Gottman, J. M., 174,176,180 Gough, H. G., 183,192 Gower.J., 67 Göpfert, H. G„ 34

Author Index Graesser, A. C., 116,123 Greene, D., 8,36 Greene, E„ 25,34,109,112 Greenberg, M. H., 2, 34 Grimaud, M., 3,33 Groeben, N., 23, 34, 79, 81, 96, 98 Gross, L„ 32,143,166,21}-231,227,230 Gross, R„ 68,69, 96 Grunzenshaeuser, R., 96 Grzelak, J., 192 Gustafson, J., 222,223, 230

Holmes, D.J., 7,34 Holyoak, K.J., 44,47 Homer, 172,173 Hoppäl, M., 101,112 Hospers, J., 128,138 House, P., 8,36 Hovland, C. J., 10,34,35,36 Huehn, P., 96,98 Humphrey, H., 222 Hunyady, Gy., 34 Hupcey, J. A., 27,35,125,136,138

Haberlandt, K., 114,123 Haggard, H. R., 172 Haimowitz, M. L., 1, 34 Haimowitz, N. R., 1,34 Haläsz, L„ 1-37, 5,19, 30, 34,114,123, 140-166,146,148,149,165,183,200 Halliday, M. A. K., 71,96 Hamilton, D. L., 118,123 Hansen, A. J., 128,129,138 Harley, S., 47 Harte, B„ 135,190 Hartmann, N „ 18,22,34 Hartmann, P., 49, 96 Harvey, O. J., 16,34 Hasher, L., 136,137 Hastie, R., 118 Hatakeyama, K., 206,214 Hauptmeier, H., 125,126,132,138 Hauser, A., 148,149,165 Hawthorne, N., 128,130 Hegel, G., 14, 74,96 Heider, F., 8,16, 32, 34,216, 217, 230 Heilbrun, A. B., 183 Heine, H., 206,207 Heise, D.R., 171,180 Helm, J., 112 Helson, R., 13, 31,34,182-192, 187,188, 192 Hemingway, E„ 43, 57,190 Hemstenhuis, F., 115 Herbert, G., 58 Hermann, C. P., 123 Hermann, P., 165,160 Hernadi, P., 96 Heuermann, H., 96,98 Hevner, K., 132,138 Higgins, T., 123,165,166 Himmelfarb, S., 34 Hines, D., 171,180,181 Hirsch, E.D., 33,57,67,96,97,113 Hitler, A., 184,185 Holland, N., 8,34,201,214

Ickes, W. J., 34 Ingarden, R., 18, 24, 34,115, 205,210 Iser, W., 18,34,97 Jacobson, R., 15,18, 34,206, 214 Jacoby, L. L., 119,123 Jakobson, R., 57,97 James, M., 183 Jauss, H. R., 60,97 Jenkins, J. J., 53,97 Jensen, W., 2, 33 Johnson, M. S., 57,95,109,112 Johnson, N. S., 26,35,41,47,54,55,98, 100,102,104,112,116,123 Johnson, P. N., 47 Johnson, R. E„ 102,112 Jones, E. E., 8,34 Jones, R. A., 13,36 Joos, H. N., 122,123 Joyce, J„ 44, 55,57, 71,128,172,173,190 Jung, C . G . , 2,168,180 Kafka, F., 73,190 Kahneman, D., 19,34 Kammann, R„ 80,97,125,138 Kamondy, L., 190 Kaplan, S.J., 134,138 Kardos, L., 34 Kareda, U., 228,230 Katz, H. A., 234 Keele, S., 54,97 Kelly, H., 216,230 Kent, G. H., 53,97 Kennedy, J. F., 222 Kesey, K„ 72, 75 Kessen, W., 80,98 Kihlstrom, J. F., 118,122 Kidd, R. F., 34 Kiell, N., 1,3,34,125,130,138 Kingston, A. J., 7, 34 Kintsch, W., 24,25,26,27,28,35,51,54, 57, 80,95,97,98,101,102,103,104,105, 106,107,108,109,112,114,116,122,123

235

236

Author Index

Kipphardt, H., 64 Kleist, H., 83 Kloepfer, R., 74,97 Knapp, M., 53, 94 Koch, W. A., 57,97 Koestler, A., .13,35 Koivumaki, J. H., 8,36 Konecni, V. J., 15,20,35 Kosztolänyi, D„ 190 Kovel, J., 201,204,214 Kozminsky, E., 25,26,34,97,116,123 Kramers, L., 32 Kreitler, H., 7,8,9,13,35 Kreitler, S.. 7.8.9,13,35 Kris, E., 15,35,115,168,180 Krolow, K., 80 Kumar, P. A., 118,123 LaBerge, D., 35,47,48,122 Labov, W„ 102,103,104,105,112 Lachman, R., 116,124 Langer, C.W., 185 Langer, S., 116,123 Langland, W„ 172 Lardner, R., 190 Lawrence, D. H., 190 Lazursky, 114 Läszlo, J., 30,113-124,118,123,136,138 Lebowitz, M„ 55,98 Ledeen, M„ 229,230 Lehnet, W. G., 42,47,48 Leibniz, G. W., 74 L e N y J . F . , 95,98,122,124 Levin, R., 198 Levin, S., 57,97 Levine, D., 230 Levinson, M., 32 Levy-Bruhl, L„ 171,180 Lichtenstein, E. H„ 26,27,33,35,116,122 Lindauer, M., 4,30,33,35,125 -139,125, 126,130,138,185 Lindzey, G., 33, 35 Lockhart, R., 114,119,122 Long,J., 122,123 Lotman, J., 22,35,74,97 Loyd, B. B., 124 Luchins, A. S., 10,35 Lukacs, G„ 61,62,97 Lüthi, M., 24,35 Mackinnon, D., 182 Madeja, S. S., 138 Magyar, I., 162,165 Mandel, T. S„ 25,26,34,97,116,123 Mandelsohn, G., 198

Mandler, J. M., 26,27,28,35,41,47, 54,55, 98,100,102,104,106,107,109,112,116,123 Mann. H. B„ 153 Mansfield, K„ 190 Marcus, S., 18,35 Marlowe, Ch., 67 Martindale, C„ 15,31,35,125,138, 167-181,169,171,173,174,181,188,200 Massinger, Ph., 67 Maugham, S., 189,190 Maupassant, G., 128,190,227,230 May, C. E., 128,129,138 McGuire, W. J., 16,32,35 Mcintosh, A„ 71,96 McKoon, G., 48 McLuhan, M., 135 Mehrabian, A., 136,138 Melville, H., 172 Meyer, B. J. E, 98 Michael, W. B„ 4,35 Micher, M., 33 Milgram, S., 113 Miller, G.W., 100,108,111 Miller, J.H., 33 Mink, W. D., 53,79 Mischeil, W„ 118,122,144 Mitchell, L., 180,181 Moffet, L. A., 13, 35 Mohrmann, G. P., 134,138 Moles, A. A., 20,35 Montague, W., 94,99 Móricz, Zs„ 119,142, 175, 178,190 Moynihan, C., 136,138 Mukarovsky, J., 59, 98,115,123 Munsinger, H., 80, 98 Murray, H.A., 182,185,188, 192 Natoli, J. P., 138 Neisser, U„ 21,27,35, 51,98,113,115,124, 125,136,138 Neumann, E., 168,181 Nezworski, M. T., 26,36 Nicholas, D. W., 108,112 Nilsson, L. G., 98 Nisbett, E. E., 8,20, 34 Norman, D. A., 143,165 Novalis, F., 172 O'Connor, E, 119,129,133,139,142,175, 177,190 Ogilvie, D. M„ 181 O'Hare, D., 33,35 O'Henry, 130,134 Olimick, S. L., 201,214 Orwell, G., 77

Author Index Osgood, Ch., 16,35 Örkeny, I., 148 Pap, K., 190 Paris, B. J., 75,98 Patrick, W. P., 128,129,130,132,133,137 Pavlov, J., 49,98 Payne, J. W., 32,33 Pecktram, M., 125,139 Perry, M., 10,36,141,157,165 Petöfi, J. S., 165,206,214 Phylyshyn, Z. W„ 114,124 Piaget.J., 81 Pichen, J., 27,32,36,54,94,122 Pinter, H., 44 Piontkowski, U „ 79,98 Pleh, Cs., 30,34,100-112 Pliner, P., 32 Poe, E. A., 128,129,130,173,190 Pompi, K. F., 116,124 Porter, K. A., 190 Posner, M., 24,36 Pratt, M. L., 4,23, 36 Propp, A., 108,112 Propp.V.J., 24,36 Pyron, B., 16,36 Racine, J., 8 Rabkin, L.J., 1,36 Rapaport, D., 168,181 Ratcliff, R., 48 Rath, R., 4,33 Reder, L., 165 Redinger, R. V., 128,132,139 Reid, J., 128,139 Reiser, B. J., 42,48 Reitman, W., 18,36 Reynes, R.C., 171,181 Richards, J. A., 16,36 Ricoeur, P., 201,204,212,214 Rieger, C., 98 Riesbeck, C. K„ 38,48,98 Riffaterre, M., 82,98, 124 Rilke, R. M., 75 Ringle, M. H „ 47 Rittelmeyer, S., 16,36 Roettger, B., 96,98 Rokeach, M., 16,36 Rosanoff, A. J., 53,97 Rosch.E., 54,98,118,124 Rosenberg, M. J., 13,16,34,36 Rosenblatt, L., 141,165 Rosenthal, B. G., 4,35 Ross, L., 8,36 Rossmann, K., 73

237

Roth, W., 171,181 Rothkopf, E., 53,98 Rotschild, S., 206 Rumelhart, D. E., 26,36,41,48,55,98,100, 101,102,104,106,107,112,116,124 Russell, W. A., 53,97 Sackville, C., 172,173 Samsa, G. K., 73 Samuels, J., 35 Sandburg, C., 59 Sanford, N „ 32 Sarbin, Th., 22,36 Sarraute, N., 23,36 Scafer, R., 201,214 Schank, R. C., 38,41,48,52,55,98,103, 104,105,109,112,115,116,124 Scharf, A., 230,231 Schasler, M., 115,124 Scheele, B., 23,34 Schiller, F., 116 Schmidt, S. J., 18,23,36, 74,76,98,126,139 Schneider, W., 52,81,98 Schwartz, J. C„ 118,122 Scott, W., 61 Searle, J., 40,45,48,59,98 Sebeok, T. A., 34,36,97,98,214 Secord, P. F., 6, 36 Seifert, C. M., 42,48 Shakespeare, W., 31, 60, 61, 62, 63,65, 66, 67,75,116,193 -200 Shapiro, Th., 201,214 Sheenan, P. W., 36 Sherif, M., 4,36 Sherrington, C. S., 17,36 Shiffrin, R., 52,81,99 Siklaki, I., 101,112 Silberer, H., 172,173,181 Simonov, P. V., 17,36 Simonton, D. K., 13,31,36,125,139, 193 -200,194,200 Skinner, B. F., 49,99 Sklovsky, V., 117,124 Smith, Y. H., 214 Smith, M.S., 181 Smollett, T., 72 Snyder, M., 141,166 Sophocles, 194 Sözer, E., 206,214 Spenser, E., 65 Spiro, R., 27,36,43,47,48,51,94,99 Stein, G., 57 Stein, N. L„ 26,36,54,55,99,100,102, 104,112,116,124

238

Author Index

Steinbeck, J., 77 Stempel, W. D., 95,99 Stephan, W. G., 8,36 Stevenson, R. L., 129 Stifter, A., 62 Stone, A. A., 1,36 Stone, P. J., 181 Stone, S.S., 1,36 Stotland, E„ 8,37 Strelka, J. P., 139 Strevens, P., 71,96 Strindberg, A., 72 Suci, G. J., 35 Sutton-Smith, B., 171,181 Swift, J., 82 Sylvester, J., 66 Tagiuri, R., 13,33 Tannenbaum, P. H., 35 Taylor, S. E., 8, 37,141,166 Tennyson, A., 55 Thackeray, W. M., 6 Thompson, P., 99 Thomson, J., 54 Thorndyke, P. W., 26,37,54,55,99,100, 101,102,104,106,107,112,116,124,136, 139 Todorov, T., 3,61,99 Tolstoy, L., 226,231 Trabasso, T., 108,112 Treisman, A. M., 19,37 Tulving, E., 54,99 Turner, J. A., 4,33 Turvey, M., 24, 37 Underwood, B., 53,99 Updike, I., 20 Uspensky, B., 12, 37 van Dijk, T„ 2 4 , 2 5 , 2 6 , 3 4 , 3 7 , 5 7 , 9 7 , 102, 103,104,105,107,108,112,114,122,123, 146,165 Vergil, 172,173 Vikstrom, P., 165 Vincent, E. S., 126,137

Vipond, D„ 57,97 Vonnegut, K., 61 Vygotsky, L. S„ 3 , 1 6 , 1 7 , 3 7 Wagner, R„ 172 Waletzky, J., 102,103,104,105,112 Walker, W., 129,139 Walpole, H., 61 Warr, B. P., 34,37 Warren, A., 116,124 Warren, N „ 99 Warren, W. H „ 108,112 Warrick, P., 2,34 Watkins, M. J., 54,99 Weber, J. P., 3,37 Wehrli, M., 23,33 Weinrich, H., 60,99 Weiss, P., 61,64 Wellek, R., 116,124 Werner, H., 4, 37,169,171,181 West, A., 171,181 Whewell, W., 69,99 White, E. B., 77 White, W. E, 7,34 Whitney, B. R., 153 Wilcoxon, F., 160,161,162 Wilensky, R., 41,44,47,48,100, 111, 116, 124,136,137 Wilks, Y., 45,48 William, Ch„ 33 Wimsatt, 210 Winner, E., 27, 37,136,139 Wohl, J., 223 Wordsworth, W., 70 Worth, S., 143,166,215,231 Yang, C. J., 47 Yekovich, F. R., 101,112 Zajonc, R. B., 134,139 Zana, M., 123,165,166 Zender, K., 198 Zimbardo, Ph. G., 113 Zola, E., 219,226,227,231

Subject Index Aboutness 44,46 Abstractness 79,90 A Clean, Well-lighted Place 190 Action 1 , 3 - 4 , 6 - 1 0 , 1 2 , 1 4 , 2 1 , 2 4 - 26,30, 41,43 - 4 4 , 5 2 - 5 6 , 75, 77, 82-83,100, 102-103,109-110,117,127,162,195, 199-200,202,209,212-213,215,227 Aesthetic evolution 32, 227 Aesthetic impect 115 Aesthetic reading 30,122 Alice in Wonderland 77,185 All is True 63 All's Well That Ends Well 197 Alternativity 9,18, 27, 58 - 60, 62, 64-65, 67,70-71,74-78,81-82 Amarcord 229 Ambition and Hilarity 190 Amerika 73 Analogical recognition 11-12,14 Analysis of plots 29,42,46 Aneid 168 Animal Farm 77 Anthony and Cleopatra 197 Anticipation 6,9, 79,115,193,195 Arabian Nights 55 Art 3,16-17,117,119,127,132,168,215, 226,229 Arousal 7,113,161 Artificial intelligence 29, 38, 40 Association 17,53,56, 80, 84, 87- 89,93, 114-115,117,184 As You Like It 197 Attention 7,15,19-21,24, 56, 59, 86, 89, 92,113,115,125,127,130,134-135,145, 159,161,185,196,198,224 Attitude 2,13,19,23,49,207-208,210 Attribution 7,17,21, 31-32,109,117-119, 132,193-195,198,216,218,222,225-227, 229 Authoritarianism 16 Best Boy 223 Bliss 190 Blow Up 229 Bottom-Up processing

146

Brutes / Barbarians 120,142,147,149, 150-152,163,190 Catharsis 14,16-17 Castle of Otranto 61 Causal chain 103,105,109 -110 Channel capacity 18 -19,22 Character 1, 3 - 4 , 6-10,12,14, 22,27, 30, 42 - 46,60 - 61,64,72 - 75, 78,114-115, 117-121,129,131-134,140-142,144, 146-147,183-185,190-191,195,204-206, 209,224,226,228 Charlotte's Web 77 Chronicle 63 Closed-mindedness 16 Cloze procedure 78, 79,93 Cognition 3,17,93,117,169,174 Cognitive process 1,9,21 Cognitive psychology 51,101,113-115,122 Cognitive style 17, 30,121 Coherence 31,66,71-76,81,89,93, 148-149,154,157-158,163,184 Cohesion 24,36 Collative Variable 14,140 Comedy of Errors 197 Communication 1,2, 29, 31, 49-54, 5 6 - 6 0 , 62,64 - 67,69, 73 - 74,76 - 78, 81-82,90, 93-94,200,202,209 Comprehension 1, 7,10, 24-25,28, 30-31, 51,57,66, 74, 76, 82,116-117,126,136 Complexity 1 4 - 1 5 , 2 4 , 2 9 - 3 0 , 5 0 - 5 1 , 5 7 , 74,80-82,92-93,128,171,182,186-188, 198,227 Concept 3,45,53,89,100,207,218 Consciousness 21,23,43,115, 130, 167-169,174,202 Consistency 53, 72, 92,198 Content analysis 31,119,167,169 Context 18,25,43,46,49-56, 71, 77, 79, 82, 89,94,126,171,183,191,194-195, 198-199,203,208,210,212,216 Contextualizing 29,50,51,93,94 Coriolanus 197 Crime and Punishment 75 Critical evolution 29,45 - 46

240

Subject Index

Cross-cultural research 29 Culture 5 7 , 6 0 - 6 1 , 6 5 , 9 4 , 1 0 9 - 1 1 0 , 1 3 5 Cymbeline 64,197 Das Nibelungenlied 65 Decameron 55 Deletion rules 24,102 Deautomatization 104,121 Defense 8,227 Decoding 12,210-211 De rerum natura 65 Der Hund des Generals / The General's Dog 64 Der Prozeß 73 Die Ermittlung / The Investigation 64 Discourse 1, 8 - 9 , 1 1 , 1 3 , 1 7 - 1 9 , 2 1 - 2 5 , 2 7 , 2 9 - 3 0 , 57,59,64,66 - 67,79, 85, 89,90,93, 102,115,117,201 Distortion 7,17,19,24,72,101,159,203 Differential greatness 194,196 Documentary literature 59,77 Dream 13,21,133,202-204,207 Duino Elegies 75 Duke of Milan 60 Eighty Days Around the World 185 Elaboration 100,116 Elementary Treatise of Mechanics Empathy 7,10,12,14,17 Emotion 7,17,127,144,203 Enrichment 20 Enjoying 90,146 Encoding 26,54,118-119 Epic of Gilgames 128 Episode 12 Everything that rises must Converge 43, 119-120,142,147,150,152,164,190 Expectation 4,6,14,27, 79,102, 140-141 Faerie Queen 65 Familiarity 5 1 , 7 7 , 9 3 - 9 4 Film forms 219 Fingal and Tamora 66 Finnegan's Wake 71 First impression 9,146 Frame 93,185 Function 9,49,57,59,61,74,94,159,201, '209,212 Generalizing 84, 89 - 91,93 Gooseberries 190-191 Great Books of the Western World Gulliver's Travels 82

194

Hamlet 3,72,116,193-194 Haircut 190 Heart of Darkness J 74 Henry IV. 61,197-198 Henry V 63,197 Henry VI. 6 2 - 6 3 , 1 9 7 - 1 9 8 Henry VIII. 62 - 63,66,197 Hermeneutics 201 Historical literature 60 - 61,67 Hypothesis-making 140-141,146,149,154, 157,161 Hymns to the Night 174 Identification 7 - 9 , 1 2 Idylls of the King 65 Iliad 61 Image 115 Imaginary 7 - 8,12,18,22 Impoverishment 20 Inference 39,108,114-115,141 Inferno 168-169 Information 1 , 6 , 9 - 1 1 , 1 3 - 1 5 , 1 7 , 1 9 - 2 2 , 25-26,28-30,115,141,144,151,153,157, 159-160,173,188,191 Interaction 6 , 1 5 , 2 3 , 2 8 , 3 0 - 3 1 , 5 0 - 5 2 , 76, 114,141-142,147-149,162,188,190-193 Interest 18,105,182 Interpretation 7 , 9 , 1 1 , 2 5 , 3 1 - 3 2 , 3 9 , 4 4 , 4 6 , 51,56, 82, 84,87,91,100,115,144, 167-168,173,202 - 208,215 - 218,225 Jailbird 61 Joke 202-213 Julius Caesar 73,97 Kingjohn 197 King Lear 64,73,184,197 King of Naples 60 Knowledge 12,15,24 - 27,30 - 31,39,42,49, 51,54 - 55,65, 72, 76, 79,80, 86 - 87,93, 109,114-118,146-153,162,167,210-211, 216,218 Kramer vs. Kramer 226 Learning 28 Levels of processing 30,108,114,119 Literature 1 - 3 , 7,23 - 2 4 , 2 6 , 2 8 , 3 0 , 3 8 , 4 0 , 43 - 44,46,49,61,62,64, 76 - 77, 80,92, 110,114,118,122,125-128,130-131,142, 162,169,182-185,188,193,210 Literary appreciation 29,38,43 Literary communication 31, 51,56 - 57,60, 67, 7 3 - 7 4 , 7 6 - 7 8 , 8 1 - 8 2 , 9 1 , 9 3 Literary discourse 1,4,6, 8 - 1 0 , 1 2 , 1 9 , 21-25,27-30,57,115,117,201

Subject Index Literary narrative 1 4 , 1 1 3 - 1 1 5 , 1 2 1 Literary text 1,18-19,27,29,30,32,56-58,60, 71,73 - 7 4 , 1 1 3 , 1 1 6 , 1 1 8 - 1 2 1 , 1 4 1 , 1 5 7 , 2 0 5 Love's Labours Lost 197 Madame Bovary 75 Macbeth 197 Maria 190 Meaning 15,18 - 1 9 , 2 1 , 2 4 , 2 9 , 4 0 , 4 4 , 6 6 - 67, 70-73,83,85-91,102,119,140-142,154, 160-163,168,174,203,209,212,215-218 Media events 32,224 Measure for Measure 60,65,75,197 Memory 8 6 , 8 9 - 1 0 1 , 1 0 8 , 1 2 6 Metamorphosis 62,190 Merchant of Venice 197 Merry Wives of Windsor 197 Midsummer Night's Dream 197 Mr. Know-It-All 190 Moods 46 Much Ado About Nothing 197 Multilevelledness 15,22,27 Naiv psychology 32 Nasis 190 Narrative production model 102 Night-journey 167-169,173 Normalizing 82,84,89 - 93 Novel 1 0 , 2 2 - 2 3 , 3 9 , 5 0 , 5 5 , 5 8 - 59,62, 72-73,76,88,116-117,125-126,128-130 Narrative pattern analysis 30,200 Paraphrasing 89 Peter Pan 185 Perception of literary characters and works 4, 6,14,28 - 3 0 , 6 2 , 6 8 , 7 2 , 8 0 - 81,117,127, 140-141,144,154-155,159,162 Person 6 - 7 , 9 , 2 0 , 2 5 , 5 1 - 5 2 , 5 4 , 6 8 , 7 2 - 7 3 , 77-79,91,118-119,130,141,143,165, 182-183,211-212,216 Personality 2 - 4 , 3 0 - 3 1 , 4 4 , 7 5 , 1 4 7 , 1 6 7 , 182-185 Personality assessment 182,184-185,188,191 Pinnochio 185 Pleasure 4 , 8 0 , 1 4 3 , 1 6 4 , 2 1 1 , 2 1 3 Poem 14,59,65,68-71,74,79,83-85, 87-90,92,94 Point of view 1,9-10,12,19,23,26-27, 29-31,100,105,108,115-116,118, 121-122,125,149,151,155,159,160-161, 183,186,201 Polyvalence 1 7 - 1 8 , 2 7 , 8 5 Prestige effect 6,17 Primary effect 149

241

Primary process 21,31,115,167,171, 173-174,176 Problem solving 108 Processing 1 , 4 , 1 1 , 1 4 , 1 8 - 3 0 , 4 9 - 61,65,74, 76-82,84,92-94,101,103,108-109, 113-122,140-147,149,153,158-159,161, 163 Progress 174 Protagonist 2 - 3 , 6 , 7 , 9 - 1 4 , 1 7 , 2 2 , 2 7 , 7 3 , 100,103,105,108,141-142,145-146, 148-149,153-154,158,160-161,163 Prototype 1 0 9 , 1 1 7 - 1 1 8 , 1 2 0 , 1 4 4 Psycholinguistic 28 Psychopathology 171 Psychology 21,29 - 3 0 , 4 9 , 5 1 - 5 2 , 5 6 , 6 9 , 7 6 , 94,100,115-116,119,125-126,130,158, 162,185,191 Reader 1 , 2 , 4 , 6 - 7 , 9 , 1 2 - 1 5 , 1 7 , 1 9 - 2 0 , 22 - 2 5 , 2 8 - 3 1 , 3 8 - 3 9 , 4 2 - 4 6 , 6 4 , 6 6 , 7 1 , 74-75,77-78,80-83,91-94,103,108, 114-118,125-127,129-131,140,151,153, 157,158,174,185,188,200,211 Readingtime 10,43,113-114,119 Recall 12,27,54,103,105,108,114 Recipient 7 - 1 3 , 1 7 - 27,115,117,140 - 1 4 2 , 146-147,154,161 Reception 9 - 1 1 , 1 7 , 2 2 , 2 4 , 2 9 , 1 1 5 - 1 1 6 , 121,141-146 Red Desert 229 Reduced 190 Regression 105,168,169,173 Remembering 24 Representation 3 , 7 , 2 6 , 4 3 , 4 5 , 4 6 , 6 1 , 85, 114-115,118,207,212 Rhinoceros 44 Richard II 197 Richard III 63,197 Romeo andJuliet 197 Schema 24,26 - 2 9 , 3 9 , 5 1 - 56,58 - 62,64 - 69, 71,74,76 - 83,92 - 94,100,108 - 1 0 9 , 115-117,141,143,144,149 Script 3 9 - 4 0 Semantic reduction 24 Sensory-motor experience 2 9 , 4 0 , 4 2 - 4 3 , 4 6 Situation 2 , 9 - 1 3 , 1 9 - 2 1 , 3 9 , 5 4 , 6 6 , 7 1 , 9 0 , 101-103,108,116-121,131,140,143,145, 153,188,190-191,202,204,210-212,217 Snobbery 6 Social psychology 2 8 , 5 1 , 1 0 8 - 1 0 9 Stereotype 3 9 - 4 0 , 1 0 9 , 2 1 3 Storygrammar 2 6 - 2 7 , 4 1 , 5 4 - 5 6 , 1 0 0 , 102-103,105,108,116

242

Subject Index

Structure 17,23 - 2 8 , 4 1 , 4 5 - 4 6 , 4 9 , 5 2 - 56, 79,82,84,92,100,102-103,108-110, 114-116,119,130,185,188,193,199-200 Subjectivity of text models 108 Syntopicon 194 Taming of the Shrew 197 Tempest 60,197 Tennysom 65 Text 1,4,6,9-11,15,17,20,22-32, 38 - 4 2 , 4 4 , 4 6 , 5 5 - 6 0 , 6 2 , 6 5 , 6 8 - 71, 7 4 - 7 5 , 7 7 - 81,84 - 8 8 , 9 0 , 9 2 - 94,101 - 1 0 2 , 110,114,117,119,121,126,140-150, 157-159,171,173-174,182,201-203, 209-212 128 The Arabian Nights 190 The Big Blond The Cabinet of Dr. Caligari and Metropolis 225 The Canterbury Tales 65,173 The Dead 190 65,173 The Divine Comedy The Dubliners 44 65 The Faerie Queen The General's Dog 64 The Gift 190 190 The Gold Bug TheHobbit 185 The Jewels 190 The Iliad 61,65 The Investigation 64

73 The Mystery of Edivin Drood The Old Man and the Sea 43 The Organ 190 The Outcasts of Flat 190 190 The Outcasts of Poker The Pickwick Papers 73 The Red Pony 77 The Rocking Horse Winner 190 The Student and the Woman 190 The Wall 190 185 The Wizard of Oz Tibetan Book of the Dead 168,173 Timon of Athens 193,197-198 Thought 14,16,18,21,24,31,57,59,63,65, 67-69,114,126-127,140,143-146,150, 156,161,164,168-169,171,193,202, 204-205,208,211 Top down processing 28 TroilusandCressida 197 Twelfth Night 197 Two Gentlemen of Verona 197 Uncle Tom's Cabin

76

Visual medication 2 1 8 - 2 2 0 Visual narrative 2 1 5 - 2 2 5 Wasteland 71 Winter's Tale 197 Zero degree of Writing

23

RESEARCH IN TEXT THEORY UNTERSUCHUNGEN ZUR TEXTTHEORIE ISSN 0179-4167

Words, Worlds, and Contexts New Approaches in World Semantics Edited by Hans-Jurgen Eikmeyer and Hannes Rieser 1981. Large-octavo. VIII, 515 pages. Bound DM 1 7 8 , ISBN 3 11 008504 6 • (Volume 6)

Psycholinguistic Studies in Language Processing Edited by Gert Rickheit and Michael Bock 1983. Large-octavo. VIII, 305 pages. Bound DM 124,ISBN 3 11 008994 7 • (Volume 7)

Intonation, Accent and Rhythm Studies in Discourse Phonology Edited by Dafydd Gibbon and Helmut Richter 1984. Large-octavo. X, 350 pages. Bound DM 1 4 8 , ISBN 3 11 009832 6 • (Volume 8)

Linguistic Dynamics Discourses, Procedures and Evolution Edited by Thomas T Ballmer 1985. Large-octavo. VIII, 366 pages. With 35 figures and 10 tables. DM 1 6 0 , ISBN 3 11 010115 7 • (Volume 9)

Discourse and Communication New Approaches to the Analysis of Mass Media Discourse and Communication Edited by Teun A. van Dijk 1985. Large-octavo. VIII, 367 pages. Bound DM 160,ISBN 3 11 010319 2 • (Volume 10)

Preisänderungen vorbehalten

Walter de Gruyter

w DE

G

Berlin • New York

GRUNDLAGEN DER KOMMUNIKATION FOUNDATIONS OF COMMUNICATION

Perspektiven auf Sprache Interdisziplinare Beiträge z u m Gedenken an Hans H ö r m a n n Herausgegeben von Hans-Georg Bosshardt Groß-Oktav. XIV, 326 Seiten. 1986. Ganzleinen D M 1 3 4 , I S B N 3 11 010068 1 (Bibliotheksausgabe) Eine Sammlung von Aufsätzen, die anläßlich einer Vorlesungsreihe zum Gedenken an Hans Hörmann, den verstorbenen Ordinarius für Psychologie, entstanden sind. Führende Wissenschaftler aus Philosophie, Psychologie und Linguistik stellen solche Ausschnitte aus ihren Forschungsgebieten dar, die für die psychologische Untersuchung von Sprache und Sprachgebrauch besonders relevant sind.

I. Historische Interpretationen

Theo Herrmann: Psycholinguistisches Nach-Chomsky-Paradigma und mitteleuropäische Sprachpsychologie — John C . Marshall: Einige Zusammenhänge zwischen Sprache und Gehirn.

II. Beziehungen zwischen Wirtschafts- und Alltagssprache

Karl-Otto Apel: die Logoauszeichnung der menschlichen Sprache. Die philosophische Tragweite der Sprechakttheorie — Oswald Schwemmer: Die Verständlichkeit unseres Handelns. Überlegungen zur sprachlichen Repräsentation unserer Handlungswirklichkeit.

III. Mentale Repräsentation, Verwendung und Erwerb von Wissen

Johannes Engelkamp: Sprache, Wahrnehmen und Denken — Giovanni B. Flores d'Arcais: Konzeptionelle Strukturen und mentales Lexikon — Walter Kintsch: Psychologische Studien zum Verstehen von Texten — Hannelore Grimm: Die Ontogenese der Sprache als Fortsetzung nichtsprachlichen Handelns.

IV. Sprachliche und außersprachliche Strukturen

Willem J . M. Levelt: Zur sprachlichen Abbildung des Raumes: Deiktische und intrinsische Perspektive — Dieter Wunderlich: Raum und Struktur des Lexikons — Werner Deutsch: Sprechen und Verstehen: Zwei Seiten einer Medaille?

V. Formen des Zeichengebrauchs

Roland Posner: Zur Systematik der Beschreibung verbaler und nonverbaler Kommunikation. Semiotik als Propädeutik der Medienanalyse. Namensverzeichnis — Sachverzeichnis.

Preisänderung vorbehalten

Walter de Gruyter

DE

Berlin • New York