The answer to Hamlet's question was given by Flaubert: Emma Bovary, c'est moi.
The magic tie is identification. Without it, why indeed should our tear-glands become active on Hecuba's behalf? Goethe's early novel, The Sorrows of Young Werther, unleashed an epidemic of suicide in Germany; every romantic young man felt that he was Werther.
The extent to which a character in a novel 'lives' depends on the intensity of the reader's participatory ties with him. To know what Hamlet feels while listening to the ghost, is the same thing as to know how it reds to be Hamlet. I must project part of myself into Hamlet, or Hamlet into myself -- 'projection' and 'introjection' are metaphors referring to the partial breakdown of the crust of personal identity. This remains true, regardless whether the reader admires, despises, hates, or loves the fictional character. In order to love or hate something which exists only as a series of signs made with printer's ink, the reader must endow it with a phantom life, an emanation from his conscious or unconscious self. The major contribution will probably come from the unconscious, which takes phantoms for granted and is apt to confuse personal identities.
Thus the figments of Bovary, little Lord Fauntleroy, and Alyosha Karamazov which float around us in the air, are projections which body forth from our intimate selves, like the medium's ectoplasm. The author has created the prototype-phantoms, and the reader creates out of himself a copy, which he assumes to be like the original, though this is not necessarily the case. Whether the Elizabethans saw Shylock in a tragic or grotesque light, my own Shylock is a tragic figure -- he has a great hook nose like mine, not a snub nose like to thine.
Some novelists give meticulous descriptions of the visual appearances of their characters; others give little or none. Here again the general trend is away from the over-explicit statement towards the suggestive hint which entices the reader to build up his own image of the character. I am always annoyed when the author informs me that Sally Anne has auburn hair and green eyes. I don't particularly like the combination, and would have gone along more willingly with the author's intention that I should fall in love with Sally Anne if he had left the colour-scheme to me. There is a misplaced concreteness which gets in the way of the imagination. It is chiefly due to the misconception that 'imagination' means literally seeing images in the mind's eye; and consequently that, for a character to come alive, I must carry a complete picture of it in my mind. Now this is an old fallacy which affects the subject we are discussing only indirectly, but has a direct bearing on certain basic assumptions about the nature of perception and memory, on which the present theory rests. These are discussed in Book Two, which also contains the detailed evidence for the rather summary remarks which follow.
Phantoms and Images
In the first place, the evidence shows that there are people endowed with the faculty of so-called eidetic imagery -- that is, of really seeing mental images with dream-like, hallucinatory vividness; but this faculty, though relatively frequent in children, is rare in adults. The average adult does not really see anything approaching a complete and sharp image when he recalls a memory -- for instance, the face of a friend -- though he may deceive himself into believing that he does. How do we know that he is deceiving himself? Here is one way of proving it -- among many others. The experimenter lets the subject look at a square of, say, four rows of four letters (which do not form any meaningful sequences) until the subject thinks he can see them in his mind's eye. He can, in fact, fluently 'read' them out after the square has been taken away -- or so he believes. For when he is asked to read the square backwards, or diagonally, his fluency is gone. He has, in fact, learned the sequence by rote without realizing it -- which is quite a different matter from forming a visual image. If he could really see the square, he could read it in all directions with the same ease and speed.
The ordinary citizen, who does not happen to be a painter, or a policeman, or of a particularly observant type, would be at a loss to give an exact visual description even of people whom he knows quite well. What we do remember of a person is a combination of (a) certain vivid details, and (b) what we call 'general impressions'. The 'vivid detail' may be a gesture, an intonation, an outstanding visual feature -- the mole on Granny's chin -- which, for one reason or another, has stuck in one's memory, like a fragment from a picture-strip, and which functions pars pro toto -- as a part, or sign deputizing for the whole.
The 'general impression' on the other hand, is based on the opposite method of memory-formation: it is a schematized, sketchy, quasi 'skeletonized' outline of a whole configuration, regardless of detail. A woman may say to a man, 'I haven't seen you wearing that tie before' -- though she has not the faintest recollection of any of the ties he has worn in the past. She recognizes a deviation from memories which she is unable to recall. The explanation of the paradox is that although she cannot remember the colour or pattern of any single tie which that man wore in the past, she does remember that they were generally subdued and discreet, which the new tie is not. It deviates not from any particular past experience, but from the general code, from an abstracted visual quality that these past experiences had in common. Such perceptual codes function as selective filters, as it were; the filter rejects as 'wrong' anything which does not fit its 'mesh'; and accepts or 'recognizes' anything that fits it, i.e. which gives the same 'general impression'. The gentleman with the new tie, for instance, can get his own back with the remark, 'You have done something to your hair, haven't you?' He has never noticed her previous hair-dos at all, but be does notice that the present one just doesn't go with her mousy appearance. Here the code is 'mousiness' which, like all visual schematizations, is difficult to describe in words, but instantly recognized by the eye. We talk of an 'innocent' or 'lascivious' expression, of 'sensitive' or 'brutal' features -- characteristics which defy verbal description, but which can be sketched with a few lines -- as emotions can be indicated by a few basic strokes indicating the slant of mouth and eyes. Other codes of recognition may combine form and motion, or vision and hearing: a characteristic gait, the timbre of a voice.
Thus recognizing a person does not mean matching a retinal image against a memory image of photographic-likeness. My memory of John Brown is not a photographic record; it consists of several, simplified and schematized 'general impressions' whose combination, plus a few 'vivid details', enable me to recognize him when we meet, or to remember him in his absence. But that remembrance is only partly of a pictorial nature, and much less so than I believe it to be -- see the experiment with the letter-square. The reason for this self-deception is that the process of combining those simplified visual and other schemata and adorning them with a few genuine 'photographic' fragments, is unconscious and instantaneous. The perceptual codes function below the level of awareness; we are playing a game without being aware of the rules. We overestimate the precision of our imagery, as we overestimate the precision of our verbal thinking (quite often we think 'that we have understood the meaning of a difficult text and discover later that we haven't really) because we are unaware of the gaps between the words and between the sketchy contours of the schemata. All introspective 'visual' thinkers, from Einstein downward, emphasized the vagueness, haziness, and abstract character of their conscious visual imagery. True picture-strip thinking is confined to the dream, and other manifestations of the subconscious.
The point of this apparent digression was to show that if the above is true regarding our mental images of real people whom we know, it must be all the more true regarding our images of fictional characters which lack any sensory basis. A character may indeed be 'alive' with the utmost vividness in the reader's mind, but this vividness need not be of a visual nature. The reader may fall in love with Karenina, despair when she throws herself under the train, mourn her death -- and yet be unable to visualize her in his mind's eye or give a detailed description of her appearance. Her 'living image' in the reader is not a photographic image, but a multi-dimensional construct of a variety of aspects
of her general appearance, her gestures and voice, her patterns of thinking and behaving. It is a combination of various 'general impressions' and 'vivid details' -- that is, constructed on much the same principles as images of real people.
In fact, there is no sharp dividing line between our images of people whom we have met in the flesh, and those whom we know only from descriptions -- whether factual or fictional (or a combination of both). The dream knows no distinction between factual and fictitious characters, and children as well as primitives are apt to confuse the two.
Thus the phantoms of Bovary and Karenina which float around us are not so very different from our apparently solid memories of Joe Smith and Peter Brown; both varieties are made of the same stuff. In one of Muriel Spark's novels, a wise old bird asks his woman friend: 'Do you think, Jean, that other people exist? . . . I mean, do you consider that people -- the people around us -- are real or illusory? Surely you see that here is a respectable question. Given that you believe in your own existence as self-evident, do you believe in that of others? Do you believe that I for instance, at this moment exist?' [1]
The only certainty that other people exist, not merely as physical shapes, but as sentient beings, is derived from partly conscious, but mostly unconscious, inference, i.e. empathy. We automatically infer from minute pointers in a person's face or gestures -- which we mostly do not even register consciously -- his character, mood, how he will behave in an emergency, and a lot of other things. Without this faculty of projecting part of one's own sentient personality into the other person's shell, which enables us to say 'I know how you feel', the pointers would be meaningless. Lorenz has shown that the various postures and flexions of the wolf's tail are indicative of at least ten different moods. As we have lost our tails we cannot empathize with these moods -- but since our labial muscles are not very different, we feel at once the significance of bared teeth.
The semi-abstract schematizations which we call 'general impressions' of appearance, character, and personality, are intuitive pointer-readings based on empathy. It is by this means that we assign reality and sentience to other people. Once more, the process differs from bringing a fictional character alive in our minds mainly by the nature of the pointers. A bland face at a cocktail party uttering the conventional type of remark may provide less pointers for empathy and imagination than the cunningly planted hints of the novelist, specially designed to produce positive or negative identifications. Some phantoms can be more real to the mind than many a bore made of solid flesh. The distinction between fact and fiction is a late acquisition of rational thought -- unknown to the unconscious, and largely ignored by the emotions.
Conflict
Drama strives on conflict, and so does the novel. The nature of the conflict may be explicitly stated or merely implied; but an element of it must be present, otherwise the characters would be gliding through a frictionless universe.
The conflict may be fought in the divided heart of a single character; or between two or more persons; or between man and his destiny. The conflict between personalities may be due to a clash of ideas or temperaments, to incompatible codes of behaviour or scales of value. But whatever its motif, a quarrel will assume the dignity of drama only if the audience is led to accept the attitude of both sides as valid, each within its own frame of reference. If the author succeeds in this, the conflict will be projected into the spectator's or reader's mind, and experienced as a clash between two simultaneous and incompatible identifications. 'We make out of our quarrels with others rhetoric, but of our quarrels with ourselves poetry,' said Yeats.
Dramatic conflict thus always reveals some paradox which is latent in the mind. It reflects both sides of the medal whereas in our practical pursuits we see only one at a time. The paradox may be seemingly superficial, as when our sympathies are divided between Hamlet and Laertes, two equally worthy contestants, with the resulting desire to help both, that is to harm both. But at least the double complicity in the double slaughter is prompted not by hate but love, and we are made to realize that it was destiny, not their own volition, which made them destroy each other; the paradox is 'earthed' in the human condition.
Thus the artist compels his audience to live on several planes at once. He identifies himself with several characters in turn -- Caesar, Brutus, Antony, projecting some aspect of himself into each of them, and speaking through their mouths; or introjecting them, if you like, and lending them his voice. He presents Brutus and Caesar alternately in situations where they command sympathy and impose their patterns of reasoning, their scales of value, until each has established his own independent matrix in the spectator's mind. Having acquired these multiple identities, the spectator is led to a powerful climax, where he is both murderer and victim; and thence to catharsis. In the Bhagavid Gita the Lord Krishna appears on the battlefield in the role of charioteer to his disciple Arjuna, whom he cures of his pacifist scruples by explaining that the slayer and the slain are One, because both are embodiments of the indestructible Atma; therefore 'the truly wise mourn neither for the living nor for the dead.' I doubt whether this doctrine, taken literally, had a beneficial effect on the ethics of Hinduism,* but to be both Caesar and Brutus in one's imagination has a profound cathartic effect, and is one way of approaching Nirvana.
Brutus is an honourable man; so is Caesar; but what about Iago? Through pitying Desdemona, and sharing Othello's despair, we are compelled to hate Iago; but we can hate Iago only if he has come to life for us and in us; and he has come to life in us because he too commands our understanding and, at moments, our sympathy -- the resonance of our own frustrated ambitions and jealousies. Without this unavowed feeling of complicity, he would be a mere stage-prop, and we could hate him no more than a piece of cardboard. Iago, Richard II, Stavrogin, the great villains of literature, have an irresistible appeal to some common, repressed villainousness in ourselves, and give us a wonderfully purifying opportunity to discover what it feels like to be frankly a villain.
But true-black villains are limit cases; the more evenly our sympathies are distributed among the antagonists, the more successfully the work will actualize latent aspects of our personalities. Caliban and Prospero, Faust and Mephisto, Don Quixote and Sancho Panza, Christ and the Great Inquisitor -- each pair is locked in an everlasting duel in which we act as seconds for both. In each of these conflicts two self-contained frames of reference, two sets of values, two universes of discourse collide. All great works of literature contain variations and combinations, overt or implied, of such archetypal conflicts inherent in the condition of man, which first occur in the symbols of mythology, and are restated in the particular idiom of each culture and period. All literature, wrote Gerhart Hauptmann, is 'the distant echo of the primitive word behind the veil of words'; and the action of a drama or novel is always the distant echo of some ancestral action behind the veil of the period's costumes and conventions. There are no new themes in literature, as there are no new human instincts; but every age provides new variations and sublimations, new settings and a different set of rules for fighting the old battles all over again. To quote G. W. Brandt: 'There is basically only a limited number of plots; they can be seen, in different guises, recurring down the ages. The reason is in life itself. Human relationships, whilst infinitely varied in detail, reveal -- stripped down to fundamentals -- a number of repetitive patterns. Writers straining to invent a plot entirely fresh have known this for a long time. Goethe quoted Gozzi's opinion that there were only thirty-six tragic situations -- and he added that Schiller, who believed that there were more, had not even succeeded in finding as many as that.' [2]
Integrations and Confrontations
If the individual act of discovery displays essentially the same psychological pattern in science and in art, their collective progress differs in one important respect. We have seen (Chapter X) that the evolution of science is neither continuous nor cumulative in a strict sense; but it is nevertheless more so than the evolution of art.
In the discoveries of science, the bisociated matrices merge in a new synthesis, which in turn merges with others on a higher level of the hierarchy; it is a process of successive confluences towards unitary, universal laws (at least, this applies to a given province of science in a given period or cycle). The progress of art does not display this overall 'river-delta' pattern. The matrices with which the artist operates are chosen for their sensory qualities and emotive potential; his bisociative act is a juxtaposition of these planes or aspects of experience, not their fusion in an intellectual synthesis -- to which, by their very nature, they do not lend themselves. This difference is reflected in the quasi-linear progression of science, compared with the quasi-timeless character of art, its continual re-statements of basic patterns of experience in changing idioms. If the explanations of science are like streams joining rivers, rivers moving towards the unifying ocean, the explanations of art may be compared to the tracing back of a ripple in the stream to its source in a distant mountain-spring.
But I must once more repeat, at the risk of being tedious, that in all domains of creative activity intellectual and aesthetic experience are both present in various mixtures; that 'science' and 'art' form a continuum; that changes of fashion are common in the zig-zag course of science, while on the other hand, development of a given art-form over a period often shows a distinct 'river-delta' pattern.* The modern atom-physicist knows more than Democritus, but then Joyce's Ulysses also knows more than Homer's Odysseus; and in some respects this progress in knowledge, too, is of a cumulative order.