3. Image and imagination
The title ‘Image and Imagination’ is the editor's, not Lewis's. The essay appears untitled in a ruled school-notebook of the sort that Lewis typically used for his drafts. It is possible that the title Lewis intended is one of those mentioned in a letter of 2 June 1931 that he sent to T. S. Eliot, asking about the possibility of publishing some of his work in The Criterion, which Eliot founded and edited. Lewis wrote:
The essay [‘The Personal Heresy in Criticism’] does, as you have divined, form the first of a series of which I have all the materials to hand. The others would be
The whole, when completed, would form a frontal attack on Crocean aesthetics and state a neo-Aristotelian theory of literature (not of Art, about which I say nothing) which inter alia will re-affirm the romantic doctrine of imagination as a truth-bearing faculty, though not quite as the romantics understood it.
A draft of ‘Objective Standards of Literary Merit’, but titled ‘Objective Value of Poetry’, appears in the same notebook as ‘Image and Imagination’, which is probably, therefore, part of the extensive argument re-affirming ‘the romantic doctrine of imagination’ Lewis planned.
Many people think that literature is one species of a genus called Art. This seems to me doubtful. But the doctrine which I am going to develop does not depend on this doubt, and it will be possible for a consistent reader to agree with what follows and still to think that literature and art are to one another as species is to genus. But many people who think this, draw from it a conclusion which I not only doubt but which I claim to refute. It is held that if anything appears in literature other than those characters which are common to all art, then that thing belongs to the book or poem in question per accidens and is foreign to its poetical nature. This would be supported by analogies.
A particular triangle may be blue; but as colour is not included in the definition of all triangles, the blueness of this one is foreign to its triangularity. It is blue not quâ triangle but quâ something else; change its colour and its triangular qualities remain as they were. In the same way, it cannot be denied that a particular poem may tell a story which is in fact historical, or a particular painting may resemble a real man; and if this is so we may, if it so happens, derive knowledge from both. But music cannot thus convey facts, and the enjoyment of music cannot be knowledge. But music is an art. Therefore ‘conveyance of fact’ or ‘knowledge’ cannot be included in the definition of Art. Therefore, wherever it occurs in a given work of art, it occurs per accidens. The poem gives knowledge, or possesses truth, not quâ poem, but quâ something else: and the knowing cannot possibly form part of our appreciation of the poem as a poem.
Fortunately we need not inquire into any of the facts underlying this argument, for it can be refuted on logical grounds. It depends on an exploded view – what may be called ‘the L. C. M. view’ – of universals, on which we may consult any formal treatise, such as Joseph's Logic. To take an example from that work, we need only point out that locomotion cannot be included in the definition of ‘engine’. But in the case of a given locomotive engine, we cannot say that locomotion is foreign to its engine-character, for ‘being a locomotive’ makes it precisely the engine which it is. In the same way, though truth cannot be included in the definition of Art, it might well be (for all the present argument shows) that all literature was true, and true quâ literature, and that the definition of literature was ‘that species of Art which is true’. Thus, again, if we were dealing with Beauty instead of Art, we certainly could not include truth in our definition; but this could not prevent a particular poem from being true, nor could it prevent the truth from being intrinsic to the beauty. For neither could we include in our definition of Beauty such things as ‘a spondee in the fifth foot’ or ‘a sequence of back and front vowels’: yet in a given line the spondee or the sequence might be the beauty. In other words you cannot find which of the characters of a given poem, or animal, or engine, are accidental and which are intrinsically poetic, animalic, and mechanic, by the easy process of looking up your definition and finding which are common to the whole genus. Rule of thumb will not help, and we have to think instead. Hasty deductions from a highly abstract conception of Art, formed by leaving out all the concrete characters of all the poems and pictures we know, are one of the prime sources of critical error. It is not surprising that ‘Art’ should be amoral, alogical, a-this and a-that, when you have arrived at Art by sheer subtraction.
Having got rid of the logical blunder, we are now in a position to devote ourselves to literature and explore the characters it really has – undeterred by the false theory that anything which literature does not share with the other arts must be accidental and unliterary. Now the obvious differentia of literature is that it uses language. And language means words combined so as to have meaning. No doubt we can derive a certain degree of pleasure from hearing a poem read in a language we do not understand – at least, if the language is melodious. But in so far as the reader, by his voice and eye, gives us an inkling of the meaning, the language is language and not mere sound to us. If he does not, then the pleasure is simply a low form of musical pleasure and does not concern us here.
Literature, then, uses language. And language is language to us only when it has a meaning. It is true that we may find it difficult to say what the meaning of a given sentence in a poem is – though Dante, by the way, has a relevant comment on this point. But the demand is perverse. If it means that all significant language can be translated without loss, we can only reply ‘Why should it?’ If it means that all significant language has a meaning restatable in the same language, in different words, we can point to such words as ‘if’, ‘but’, and the like, which are significant but cannot be paraphrased. If it means that ‘the significant’ must coincide with ‘the logical’ – i.e., with that which can be represented by pure conceptual symbols – then this is sheer dogma. I am speaking of meaning in the most elementary sense; and in this sense all literature uses significant language, and to enjoy it involves a knowledge of the meaning – as for example, you could not enjoy the lines in Keats’ ‘Ode’ if you thought that a casement meant a kind of case, or that opening on meant ‘opening on top of’. Whatever difficulties there may be about the ‘meaning’ of poetry in some more ultimate sense, we clearly need to know ‘what it means’ in that lowest sense whereby a schoolboy knows or fails to know ‘the meaning’ of a bit of Caesar. You cannot have even a fairy tale unless the reader and writer understand what a castle, or a stepmother, or a giant is.
Now I have used the word ‘is’ advisedly. To read the fairy-tale we must know what a castle is; and that would seem to be a bit of knowledge about the real world. It will be objected, of course, that you may also need to know what a fairy ‘is’; and this does not seem to be equally a bit of knowledge about the real world. It would, therefore, seem safer to emend ‘know what a castle is’ and say ‘know what the author means by a castle’. That is (apparently), we do not need to know about things: we need only know of what imaginations in the author's mind each word is the name. But this substitution, which seems so reasonable, does not really help us at all; and the exposure of its illusory offers of assistance will occupy us for the next few pages.
Let us suppose that the narrator begins: ‘Once upon a time there was a princess, who lived in a tower.’ Suppose, again, that we interrupt him by asking ‘What is a tower?’. ‘Aha!’, replies he, ‘You are not going to catch me that way. The tower is not any thing in rerum natura, it is only the name of something in my imagination.’ – ‘Very well. Of what thing in your imagination is Tower the name?’ – ‘Oh, a tall kind of building, you know, with five or six stories and battlements on the top.’ – ‘It is one kind of building, then. And what about the word Building? Is that also the name of something in your imagination?’ – ‘Certainly. I mean a thing made out of some hard substance – stones or brick or wood – with empty spaces inside, where people can live.’ – ‘Dear me, this gets more and more difficult. We have now a whole host of things: stones, brick, wood and people, not to mention “substance”. Are these the names of realities, or of things in your imagination?’ – ‘Of things in my imagination. But don't let us have any metaphysical nonsense about substance. I mean matter.’ – ‘Which again is the name of something in your imagination. And this imaginary matter is doubtless not merely a series of coloured shapes, but solid and resisting like the matter in the real world?’ – ‘Certainly.’ – ‘Then it has an “inside” no doubt. Between the surfaces of each bit there is something – whether continuous stuff, or discrete atoms, or forces?’ – ‘I suppose so.’ – ‘And this is all present in your imagination?’ – ‘I am not sure.’ – ‘And again, this matter occupies space, which, no doubt, is also the name of something in your imagination.’ – ‘Go on.’ – ‘And I suppose any particular piece of this imaginary space has other pieces surrounding it; as, for example, there is more space (filled or empty) to the left and right of the space which the tower occupies, and above it and below it?’ – ‘Naturally.’ – ‘Very well. Does this space in your imagination go on for ever or does it come to an end somewhere?’ – ‘It goes on forever.’ – ‘Then the space you imagine is infinite.’ – ‘Yes.’ – ‘Then an infinite is present to your imagination?’ – ‘I don't know.’ – ‘And if you turn to the princess, I suppose you will say that she has parents, though only in your imagination?’ – ‘Well?’ – ‘And grandparents and remoter ancestors, back to the beginnings of human life, and beyond that to the beginnings of organic life? And all these exist in your imagination.’ – ‘I don't know.’ – ‘But surely you do know that there must be some flaw in the original position. For if the words in your story are really the names of things in your imagination, it will follow that in every sentence, and more than once in every sentence, a complete universe in all its detail must be present to your imagination. But that is impossible. We must have made some sort of mistake.’ – ‘I suppose we must.’
And with this the argument collapses. If the words used in literature name only imaginata, then, since any one object implies an infinite series both in time and space, every single imaginatum implies an infinite imagined series. For the imaginata are not in the same position as real objects. A real object, no doubt, demands a context reaching away on every side to infinity: but just because it is an object in the real world that context is provided for. It is there, in fact, whether we think of it or not. But the whole contention about literature was that it referred to objects existing only in imagination; and imagination cannot provide (as reality can) the necessary context. It is clear, therefore, that we shall have to allow some sort of interplay, or overlapping, between the imagined and the real, for only in the real can imaginata find the context they need. Monadism, even if it were true of the other arts, must be untrue of literary art. What we are really doing is to call the monadists’ bluff: they are ready enough to talk of a tale or poem as presenting ‘a world of its own’: but we can show them that the word ‘world’ will have to be taken more seriously than they ever intended. To get a single blade of grass growing in that imaginary world you must make it a world indeed – a real universe, self-sufficing in space and time. But only one author can do that sort of stage-set. On this view only God can tell stories.
It will be objected that I am driving poets and story-tellers into a quite unreal difficulty by forcing on their imaginary worlds just that logical connection and reciprocal implication which it is their privilege to neglect. A real tower, you say, needs stones, matter, gravity, space and time, but an imaginary tower does not. That is the whole point about an imaginary tower. But what does this defence really mean? If it means that we do not imagine the infinite series, I agree that we do not and cannot. But does that mean that we can dispense with it? For, of course, if we need it, and yet do not imagine it, the conclusion will be that we are using the real world to provide a context for our imaginary tower. Let us, then, inquire how we would get on with no context.
Let the narrator take up a new position. ‘I see where I went wrong’, he may begin, ‘I should never have admitted that what I meant by a tower was a building etc. I grant you that once you let that in, you let in the whole universe. All I was really imagining was the gleam of its white pinnacles above the trees, and the cool, dark pit of its doorway, and the moss in the crevices.’ – ‘I see’, returns the tormentor, ‘In fact Tower, in your story, is the name of an image – or rather of a group of images, visual, tactual and what not.’ – ‘Yes, exactly.’ – ‘And that is all you need for your literary (or perhaps you would rather say “aesthetic”) purposes?’ – ‘Yes. I have already admitted that.’ – ‘I suppose, now, that in a certain story, you could have a man wandering in an enchanted country where the mere appearances of buildings were constantly meeting him; things, I mean, that looked just like towers, but which he found he could walk through.’ – ‘I suppose I could if I wanted to.’ – ‘And again, you could have, in a story of a different kind, say a novel, a man encountering a real tower?’ – ‘Of course.’ – ‘Is there any reason why the mere image of the phantasmal tower should be any different from that of the real tower?’ – ‘I can't tell you without writing both stories.’ – ‘Well, at any rate, as far as a priori grounds go, you can't foresee a necessary difference?’ – ‘Perhaps not.’ – ‘Yet they would differ aesthetically wouldn't they? I mean they would be suitable to quite different genres?’ – ‘Well, of course a story of magic would be different from a realistic novel.’ – ‘Yes. But I do not mean merely that the actual walking through the Tower, and thus proving it to be phantasmal, would be different from any episode in the novel. I mean that whenever the phantasmal tower turned up, whether the hero walked through it or not, it would affect the reader differently, strike a different note, from a real tower. And that would be simply because he had learned to take it as a phantom: not from any fresh imaginary experience of its insolidity.’ – ‘I suppose so.’ – ‘And, yet, what he was visualising, the mere images, might be just the same as the images he would have if you had offered him a “real” tower?’ – ‘All right.’ – ‘So that precisely the same image may play different roles – aesthetically different.’ – ‘Well, as I said before, I doubt if they would be the same. It is a very far-fetched hypothesis: quite a marginal case.’ – ‘Perhaps it is. But what would you say to the two Florimels in Spenser. The whole point is that they are to look just the same. That is the poet's hypothesis. But you will admit that the two are aesthetically distinct. They affect us quite differently. And that is because, while the images remain indistinguishable, the one is “taken to be” a mere phantom, and the other is “taken to be” a woman who could have parents and children and could love and so forth. I say “taken to be” because, of course, we don't make the parents and children etc. actually present to the imagination.’ – ‘I don't know why we shouldn't.’ – ‘Because if we do (don't you remember the princess?) we get landed in the whole universe.’ – ‘Oh damn! So we do. All right.’ – ‘It seems then that aesthetic differences are not necessarily differences in images. Or take another example: “Childe Roland to the dark tower came.” Surely you just take it for granted that it is a “real” tower – I mean a solid tower, with an inside, and all that? You are not in any doubt?’ – ‘Well, I don't know. The poet might have gone on to make it a tall ogre in disguise. In a certain kind of poem that might work rather well. The moment of discovery would give an agreeable shock.’ – ‘Yes. But it would give no shock at all if it didn't surprise you: and it would not surprise you if you had not begun by taking it as something else. And “to take it as something else” means to have something more than an image. For, by hypothesis, the mere image is the same, whether it's going to turn out a tower or an ogre.’ – ‘Look here. I am getting very bothered about this everlasting harping on the “image”. I’m not sure that I have these distinct images at all when I am telling or following a story. I think you are assuming that every one is an extreme visualist.’ – ‘No, that is what you are assuming. For on your view, in order to get rid of the infinite series, the naked image is all that you want for literature. The image is what counts aesthetically. But if so, a bad image means imperfect appreciation of the story. On your view only visualists can be poets or readers of poetry.’ – ‘Remember, my tower wasn't only visual. There was coolness: and in fact there would be smells and all sorts of things.’ – ‘That makes no difference. I have used the word “visualist” for brevity; but if you like I will go through images of all the senses in turn.’ – ‘God forbid! I’d rather give in than that.’ – ‘We agree, then, that identical images can differ aesthetically, and, in general, that the aesthetic difference between two situations is not in proportion to the difference of the images? And, perhaps, now that we have raised the question of the visualist, you will admit that the images actually present to a reader matter very little to his appreciation. In fact, the image is not what counts. But this surely means that the words in a story cannot be simply names for images in the teller's imagination.’ – ‘Have it your own way.’
The attempt to do without a context for the imaginata thus collapses. Take away from the imagined tower all its implications and it ceases to be an imagined thing and becomes merely an image. But images are not enough: for the way in which they affect us depends, not on their content as images, but on what they are taken to be. Mention a tower, or a king, or a dog, in a poem or tale, and they come to us not in the nakedness of pictured form and colour, but with all the associations of towerhood, kinghood, and doghood. And there is this further defect in the image theory, that it throws an altogether artificial stress on the function of explicit psychological imagery in literature. What sort of blotch or blur is my ‘image’ of the dark tower to which Roland came? I hardly know. I know only that it is different every time I read the line, and that there is no proportional difference in my poetical experience. On this point I speak with some authority, having been born an extreme visualist, and having learned that this unruly power – in truth not the ally of imagination, but a mere nuisance to it – must be corrected and restrained in dealing with literature. Our imagination uses our images for poetical purposes, much as a child uses material objects for its games. An imaginative man can make of very scanty and crude images all he needs for appreciation of the greatest books, as a child worth its salt can make a liner or a railway station out of the first two or three bits of furniture it finds in the nursery. It is not the children with the costly toys who play best: or if they do, they do it in spite of the toys.
We now know that the words which occur in literature are not the names of mere images; and if they are the names of things, then these things demand a context which imagination cannot supply, and therefore these things are not, in any unequivocal sense, ‘things in the imagination’. They seem, in the most troublesome way, to be half in the imagination and half in the real world; and this, however troublesome, is precisely the view we shall have to come to. When a fairy tale mentions a tower, it refers us to some rudimentary image which is just ‘taken to be’ a tower, and therefore affects us as a tower, though the infinite series involved in towerhood is not, and cannot be, present to imagination. There is always some tacit assumption, some taking for granted, about the words used in literature, whereby a ‘girl’ or ‘storm’ in a poem is enabled to let in on us the whole meaning of girlhood or of rough weather.
It is convenient, at this point, to remind ourselves of the situation as it exists, not in literature, but in ordinary life. In ordinary life we never see a ‘tower’. We see a coloured shape, which we take to be a tower, because we believe that it has another side, and is solid, and hollow, and fulfils all the other conditions of towerhood, which, however, are not given in the experience of seeing it. In other words, we turn a mere sense-datum into a tower by attributing to it a context. We link up the present experience with other actual and inferred experiences in one determinate way, and this fixes the present experience as a ‘tower’: if we had linked it up differently, or failed to link it up at all, it would not be a tower. In the case of sense data, then, in taking any sense datum as a ‘thing’, we assert categorically certain connections between it and the rest of our real world. Our problem is to find the parallel process whereby we take a mere image as an imagined ‘thing’. Or, if there are some who do not visualise at all, how do we make a non-visual idea into an imagined ‘thing’? For, of course, a non-visual idea, even a highly intellectual conception, can no more contain all the necessary conditions of thinghood than an image can. We can no more think the infinite series than we can picture it. In either case, something immediately present has to be treated as, or taken as, a thing, though in itself it cannot give us thinghood. In a sense, no doubt, this has to be done not only when we listen to a story, but whenever we listen to speech of any kind. But no other speech is in the same unhappy position as imaginative literature, for all other speech claims from the outset to deal with a real world. If you mention the Parthenon, I cannot furnish its context in my own mind: but then the real Acropolis and the Balkan Peninsula and the solar system can. But if you speak of Valhalla, and I cannot furnish the context, and without the context it is a mere image, who or what comes to our aid?
In a very rough and ready sense we may reply that this context, though not explicitly present to the imagination, is assumed. But we must find out what this means. We must not, for instance, interpret it by supposing that what is not explicitly present can be ‘implicit’, as unexpressed stages can be implicit in a chain of reasoning. For the character of reasoning is that every judgment rigidly determines and is determined by others: but the character of imagining is that it is free – at least within very wide bounds indeed. If you imagine a tower, even a solid tower, foundations and builders are not necessarily implied: for you are at perfect liberty if you choose to suppose that it was produced by the fiat of a magician and is supported by four buried Jinns. In conceptual thought, if more is meant than meets the eye, you can tell what that ‘more’ must be, because it must be related thus and thus to what does meet the eye. But in imagination, if there is any more, you cannot tell what it is, for it might be related to the explicit imagery in almost any way. It is like the difference between a mirror and a picture. If you ask what the invisible parts of the looking glass room are like, you can give an answer – pace Alice. You know where all the lines go to beyond the frame. But if you ask where a road in a landscape goes to when it passes into the frame, we can only reply either that it does not exist, and therefore goes nowhere, or that it may go anywhere: for it might turn sharply the moment it was out of sight. Similarly, of that which is not consciously present to imagination, we must say that perhaps it does not exist, and that if it does it might be almost anything. That it exists at all appears to me a rash assumption. To say that we imagine that of which we are unconscious, seems to set up with imagination itself a distinction, reality and appearance, which is all but meaningless. Surely here, if nowhere else, esse is percipi?
The ‘assumptions’ or ‘things taken for granted’, then, which supplement the images or ideas actually present and convert them into imagined objects, cannot be, in the strict sense, ‘implicit’. And yet they must exist. There remains only one possible explanation. If we asked the teller what he meant by a tower, he would, after the tortures to which he has been submitted, reply: ‘I was imagining the sort of thing which you could walk all round, in which you could find stones and mortar, which would have foundations etc. – which would in short fit into a whole spatio-temporal system in one way and not in another.’ This reply makes a great advance. The context does not actually exist: it is hypothetical. But all hypothesis rests on some actuality. Every conditional sentence has an ‘if’-clause, concealed or expressed. ‘I could do this’ means ‘I could do this if I liked.’ We must therefore find the hidden protasis in the sentence: ‘I was imagining the sort of thing which you could walk all round.’ Clearly the protasis ‘If you liked’ will not serve. For, short of direct spatio-temporal inconsistency, you could do anything you chose in imagination; and to enumerate one or two of the myriad incompatible possibilities, therefore, will not determine your imagined thinking. It remains, then, to supply the only protasis which is left, namely ‘If it were real’. When I imagine a tower I imagine a sense-datum which, if it were given in sense, would be connected thus and thus with the real world. The same conclusion can be reached in a different way. What we require in order to turn mere images into imagined things is something that can be assumed without being explicit or present. Here, as in all cases where an appearance is to be treated as a thing, more must be meant than meets the eye. But in imagination, either there is no more than meets the eye, or if there is, we do not know what it may be and therefore cannot assume it. Therefore imagination cannot yield the ‘something more’. Therefore reality must yield it. But since reality does not actually give it, it must be given hypothetically. That is, what is really implicit in imagination is hypothetical assertion: and all hypothetical assertion is about reality. The statement ‘I am imagining that, if it were real, would have the properties X Y Z’ is a statement concerned just as much with the real world as with imagination; for it can be easily converted into the form ‘The universals X Y Z, in the circumstances P Q R, would yield the appearance which I am imagining.’ And this is a statement of exactly the same kind as any of the statements we meet in the works of science.
What we do when we imagine, then, is to suppose (with or without the support of explicit images) a reshuffling of universals taken from the actual world. When we imagine Britomart we take our idea of ‘a girl’, which is part of our general knowledge, and our idea of ‘medieval knight’, which is another part of our general knowledge, and put them together. To get into imagination itself what we mean by either of the two terms is impossible. They are not imaginations: they are summarised knowledge of the real. Always the real world is the bank on which the poet draws his cheques; and though a metaphysical lyric may be a fine and private place, all the meanings embraced within it are but passengers who come there from the public, eternal, objective world of reality and haste thither again. Aristotle was right. Poetry presents ο
α 
ν γένοιτο, things that might be – it recombines elements which belong to the real, and to appreciate poetry involves at every moment a knowledge of those elements and therefore of the real.
What obscures this truth is the fact that we may find fairies as well as towers in our story. If all literature were realistic the doctrine which isolates ‘art’ from human experience as a whole would not be even plausible. The sort of monadism which I am rejecting may be regarded as one of the many attempts to explain the marvellous in literature: parallel to the allegorisations with which our ancestors excused their pleasure in Ovid, and equally shallow. In fact, the wildest passages of Italian or Indian epic conform to Aristotle's definition as easily as one of Thackeray's novels. Let us try the experiment. Disturb the realism of Roland's dark tower a little. As Roland approaches, let the bells within the tower begin to ring. But instead of sounds let them give forth great birds: and yet we understand somehow that the birds are the sounds. This seems a far cry from the real world. Yet even these birds affect us only because of their hypothetical connection with reality. For example, ‘bats’ would have a different poetical value from birds. Yet the images will scarcely be distinguishable: and neither their bat nature or their bird nature will be explicitly imagined. It is enough to give them the one or the other name and they will affect us accordingly: and this because we know what bats and birds are in the real world, and therefore what these would be in the real world. And however we vary the fantasy we shall find the same result. Let these bell-born birds be no common birds but the souls of dead men whose blood was used to temper the bells: and let them fly out singing with human voice Justorum animae.
Does any one suppose that the imaginative value of such fantasies can be divided for a moment from our knowledge of death, and blood, and Christianity? You may say, indeed, that it is not our crystallised knowledge of these things that counts, but the emotions which have gathered about their names. But it is easy to answer that where a reader depends merely on the associations roused by the words, he will be right only by accident: as many a young reader has found to his cost. Good reading implies, and good writing demands of its readers, that the emotion should depend not on the name alone, but on the name understood. The name, indeed, rouses emotion, but rouses it through the memory of the thing; that is, through knowledge. And we have seen that no degree of unrealism in the imagination impairs this principle a whit. We can cut the bells off from the sound which they would really have and substitute birds; but then the quality of such a marvel depends on our having expected sounds, and we expected them because of our knowledge of real bell-hood. The birds again affect us by their hypothetical connection with real zoology: cut this off, and the whole process is repeated with the dead men. In other words, we can sever any number, one by one, of those connections which our imagined objects would have in reality, but for every one we sever we set up a new one. The monadist is fighting with a hydra. If A were real it would imply B. We can remove B and substitute C which A could not have in reality. But if C is to have any poetic significance for us, then C in its turn will be determined by its hypothetical connection with the real: C will imply D. You can change D if you choose, and D again will have meaning for us in virtue of its hypothetical E. Remove E and it will be the same again. You must stop somewhere: and wherever you stop you will find reality waiting for you. You may change, as much as you please, the character which your objects would have in reality: but reality furnishes both that which is changed and that by which you change it.
If this is true of a fairy tale, it is more obviously true of a tragedy, a novel, or an epic. A single image, a break in the rhythm, a ‘Prithee undo this button’ becomes the key to a whole situation. But these things can have no meaning to us which is separable, even in idea, from our knowledge of what they would mean in the real world. It is not the words, taken as mere sounds, that enlighten us. They reveal, as we are told, the character's [emotions]. Lear's speech shows what Lear felt. But then there is no Lear, and he never felt anything, if by Lear you mean something other than the words and, at the same time, actually present: the attempt to get Lear – as opposed to what Lear says – inside the imagined, ends in the infinite series. When we know Lear through the play, we know that if such things were said and done and suffered in the real world it would be by a man of such and such a kind. And this is a bit of knowledge about human nature, differing not at all from that knowledge of human nature which we use in our daily life.
Some will say that I am confusing ‘aesthetic experience’ with the cognitive experience which may lead up to it and merely forms its basis. But this is untrue. When one experience is merely the support or external prius of another, it does not survive in it. A comfortable chair may be the prius of appreciating a poem: but the sense of comfort has vanished from consciousness before appreciation begins. A knowledge of Greek grammar is the prius of enjoying Aeschylus: but we are not aware of paradigms when we enjoy – nay, how many of us, if we are not schoolmasters, could write out, in answer to a grammar paper, those very forms which we understood as we read? But knowledge of human nature is not thus externally related to our imaginative apprehension. The character of Lear and the process by which we know it are one. The words, the rhythms, and the architectonics of the play are not separable from some ‘subject’ or ‘content’ which we reach through them. Form and matter – data and conclusion – are abstractions. What we have is matter informed, form incarnate: the tragic situation, or character, embodied in these words and no others.
And so an argument in which I have deliberately chosen the wildest fancies for my illustrations brings us back into line with the grand orthodox tradition of European poetics. Poetry imitates the universal, and when we read poetry we are engaged in knowing. Thus said Aristotle, thus the Neo-Platonists. To this Rymer and Voltaire commented, though in a crude and vicious fashion, misconceiving the ‘general’ as the socially, or politically normal. On this Johnson based his disagreement with Rymer and Voltaire, returning like the saner Aristotelian he was, to ‘just conceptions of general Nature’. This was the same orthodoxy to which Wordsworth returned, for Wordsworth also is in the true line of descent and agitated repeal rather than revolution. He would have us return ‘to eternal nature’ and leave those arbitrary idiosyncrasies which ‘though they may have prevailed hundreds of years, a philosopher will look upon as accidents’. It is not the romantics, but the Croceans, who are the heretics.