The Verbal Analysis

William Empson

From The Kenyon Review, Autumn 1950, Vol. XII, No. 4

When I was asked by letter to contribute to the symposium from Peking, my first impulse was to say that I didn’t believe in having any Credo; a critic ought to trust his own nose, like the hunting dog, and if he lets any kind of theory or principle distract him from that, he is not doing his work. This does seem to me the deepest truth about the matter; but the bottom, as Mr. T. S. Eliot remarked, is a great way down. There is the same position about a moral or ethical theory; however firm your belief in it, and however definite its ruling on a particular case, you still have to see whether your feelings can be brought to accept the results in that case. If they can’t, well, you may be wrong, but if it gets too bad, you have to give up the theory. All the same, there is clearly a need for such theories; for one thing, without a tolerable supply of handy generalizations you can’t stretch your mind to see all round a particular case. And the theory alters the feelings no less than the feelings alter the theory.

The metaphor to keep in mind, surely, is the standard one of Taste. It has become less obvious in the last generation or two because of a kind of traffic jam in both literature and criticism, one which nobody need be blamed for. We know a good deal more than we did about the human mind, and expect a writer to work from the whole body of his experience, so that references to psychological and sociological and anthropological affairs may very properly be cropping up; and the critic is to expect the same things in earlier writers even if used “unconsciously.” Also we have only recently begun to try and appreciate all the arts of all periods of all countries, or rather to try and pick out from them representative high-standard cases, fix in our minds the whole range of possible achievement; and meanwhile we have had on our hands the results of the first attempts at universal education ever made. People were bound to look round for some System which would give quick answers and break up some local traffic jam, often very usefully. Going back to the metaphor of Taste, however, any such theory needs to be regarded as a salt or what not which is to be dissolved into the blood; in its crystal form, when it is portable or transferable, it is quite useless for the ultimate purpose; it must be digested, but it may none the less really be needed. What the thing is all about, the test of the value of learning these theories, is when a foreign body is masticated and brought up to the taste buds (or palped, if you prefer, by the “tact” of an exploring fingertip); there is a diffusion across a skin, which needs to hold firm, but there is an otherwise direct action between the foreign body and the living blood (for the finger you would say the living nerves).

It often strikes me that students going through the kind of critical course I think good are liable to get into a mood of excessive anxiety about their own capacity for tasting, and that others (probably the ones with less natural taste to be anxious about) tend to complain fretfully that the machine offered them is not a completely reliable machine, such as would be guaranteed to save them from having to risk tasting anything at all. Of course the second feeling is ridiculous; even with the most self-acting modern machine, some real human being has to check whether it is out of order or broken down. The first is not absurd, and any serious type of training has to be watched for the same danger; it begins to give bad results if it is screwed up to the point of excessive strain, excessive, that is, for the particular people concerned. And of course in any training for literary criticism one needs to get quite clear, what I think usually is clear to those concerned, that they are expected to use their own Taste, and that it would be a very bad sign if they never disagreed with Teacher.

Another trouble that seems to crop up is the idea that poetry is good in proportion as it is complicated, or simply hard to construe; it seems quite a common delusion, and always shocks me when expressed. And yet I suppose it is very near my own position; in any case it joins on to I. A. Richards’ Theory of Value as the satisfaction of more impulses rather than less, and T. S. Eliot’s struggle to find a poetic idiom adequate to the complexity of modern life. But, without disagreeing with these figures at all, it is necessary to see the point of the reply of Wordsworth:

                                                The gods approve
        The depth and not the tumult of the soul.

Indeed “depth analysis” is probably the best way out of this limiting critical impasse. If you realize the weight of the latent politics in Wordsworth’s apparently simple descriptions of Nature, both in the metaphors used and in the intense political experiences he had actually gone through before arriving at them, you are no longer likely to complain (of the Prelude, at any rate) that he is only serving up apple-dumplings at his state banquet. It may be, as some critics would object, that this would only mean going off on another side-line (though I do not believe so myself) but it would at least be a reassurance allowing taste to act on the verse without being put off by its absence of surface complexity.

The essential thing is to get the process the right way up. If the reader feels a passage is good, let him by all means direct his attention to considering what the profound complexities are, which his theory leads him to expect there; the trouble is when he presumes it cannot be good if it does not seem complex.

I am thus very little willing to sympathize, in one way or another, with critics who feel objections (or a crusading spirit) against analytical study of literary texts. But I cannot believe they mean to go to the startling lengths they sometimes imply. There is an important distinction to be made here, one that Mr. Ransom was pointing out not long ago. It is all very well to say that the learner may learn to use his own taste, and therefore must be simply “exposed” to the work of art in question; but you have still got to get him exposed to it. As is particularly clear with children from homes where they don’t read poetry, but also sometimes true I am afraid of all of us, it is quite possible to be confronted with a work of art and not see what the point of it is, what it is trying to do, how one part of it is supposed to affect another. There is room for a great deal of exposition, in which the business of the critic is simply to show how the machine is meant to work, and therefore to show all its working parts in turn. This is the kind of criticism I am specially interested in, and I think it is often really needed. Anyone who objects to it because it does not try to give a Final Valuation of the work, in relation to all other work, seems to me merely irrelevant. Where I should heartily agree with him would be if he said that, after all this supplying of the reader or the student with the machine, the vital question is whether he can make it go, whether he gets the experience in question; and also, for that matter, whether it leaves him better equipped to do the same thing on another occasion of his own accord.

Going back to the question of Valuation: I do not mean to say, what would be a very foolish thing to say, that criticism has nothing to do with valuation. It has to do with it all the time, because you cannot even say just how some element works without suggesting how well it works. But to assess the value of the poem as a whole is not the primary purpose of this kind of criticism, or at any rate ought only to emerge from the analysis as a whole. There is a tendency to feel that, if the critic is offering a really efficient machine, it ought to be able to say whether marmalade is better than sausages; but even the most expert cook cannot say that; sometimes you want one, sometimes the other. Especially in our own age, the first to make a serious effort to appreciate the whole variety of good literature, this kind of absolutism seems to me comical. In any case, there is no question of the critics providing a Last Judgment about the works of the past; Mr. T. S. Eliot once remarked that a critic could only hope to illuminate the work of a past period from the point of view of his own. The metaphor deserves pondering, because it is not denied that he gives real light such as may clarify the work for a still later generation, but only that he can claim to look from all possible historical or cultural points of view at once. As to the poems of his own generation, surely there can be no doubt that he ought to judge them, after proper mastication, by his own Taste; any theories he may have, based on the experience of the past, are precisely what the new work ought to be testing. Finally I do not deny that it may be a splendid thing to have a grand synthesis of human experience, a single coherent Theory of Value which could be applied to all works of art and presumably to all human situations; but it seems hardly reasonable to grumble, in the present state of affairs, that nobody has provided one; and if it did exist it would clearly be a philosophical synthesis rather than a literary one.

The kind of criticism that most interests me, verbal analysis or whatever one calls it, is concerned to examine what goes on already in the mind of a fit reader; sometimes bringing it up from levels of unconsciousness deep enough to make it look rather surprising, but even so not expected to make much difference to the feelings of the fit reader after he has got over this surprise. Like all theories about the action of the mind, in short, there is a sense in which it does not need to be expounded; if it is true, we are already acting on it all the time. The only use of it is when something goes wrong; but this is true of a good deal of knowledge, such as the ordinary car-driver’s knowledge of the working of the carburetor. A quite practical problem therefore often arises, as to how much of the analysis needs to be written down; often a very great deal could be written down which though true doesn’t need saying (except indeed to forestall a certain type of objecter, who likes to tell the critic he was ignorant of what was too obvious to need saying). If you are trying to tell your audience what it is missing, what the Elizabethans for example would feel about a passage whose language has been dulled, it is clear that you need to know your audience as well as your topic. I should think indeed that a profound enough criticism could extract an entire cultural history from a simple lyric, rather like Lancelot Andrewes and his fellow preachers, “dividing the Word of God,” who were in the habit of extracting all Protestant theology from a single text. A critic obviously does not need to do this kind of thing often, if at all; it is not really a convenient way to teach cultural history. But it is not my fault, or the fault of any other analytical critic, that our equipment threatens to make us become bores; it is wonderful how many ways there are to be a bore, and almost any line of intellectual effort, however true and useful, presents this threat. I do not know that any of us would deny that we had better try to keep to saying what is worth saying.

The business about digging up the Unconscious, at its various levels, has a separate difficulty of exposition. For example Mr. Elder Olsen , in a recent number of Modern Philology, was objecting to a passage I had written long ago (my Ambiguity, p. 49) about some remarks of Macbeth:

                                                      If th’ Assassination
        Could trammell up the Consequence, and catch
        With his surcease, Success. . . .

He seemed particularly irritated by the sentence, “Tramnmell was a technical term used about netting birds, hobbling horses in some particular way, hooking up pots, levering, and running trolleys on rails.” This and similar remarks about other words in the speech he called “a meaningless and tasteless muddle,” because these meanings are clearly not all meant to appear. But I do not think I have ever heard anyone use the word trammell in ordinary life; it still seems to me sensible to go to the dictionary and find what sort of thing it “meant” to an Elizabethan. These extra meanings are present, not in any deep unconscious, but in the preconscious levels where we handle lexicon and grammar, in our ordinary talk, at the speed we do (surely the various current uses of a word must be in the mind somehow, or how can we pick out the right one so quickly?). These current uses of the word appear to Shakespeare or his audience as a kind of feeling about what you could do with it; and the literary effect here, though simple in its way, is I think very strong. Macbeth is trying to feel that this is only a kind of engineering problem; if only he can get the murder done efficiently, he thinks, all this fog will lift and he will be able to see clearly again. Both the mechanical analogy and its underlying complexity still seem to me very direct parts of the speech. You may say, no doubt, that I ought to have indicated how far these meanings were supposed to be “unconscious,” but if I had done that all through it would have made the book much longer without really making it clearer. Mr. Olsen goes on to argue, from a kind of Aristotelian position I think, that the context of the dramatic character, and the stage reached in the structure of the planned work of art, are both extremely relevant to a literary effect, without being part of the meaning of the words used there. This is merely a question of definition; I should include among what a passage “meant” (to a fit reader) its whole literary effect, from whatever cause, and I claim I was patently considering that here in saying what the words “meant.” What I would be inclined to claim, after this point of definition had been overcome, is that cases arise where we have forgotten how the whole structure would strike a contemporary, and are doubtful between one theory and another (so far from its always being obvious as Mr. Olsen appears to assert) and then we can decide between such theories by examining minor word-pointers which at the time merely “fitted in” and did not seem important.

I hope that this more or less covers the points at issue; but I have not yet approached a passage in the original letter which asked me to discuss the co-operative work of critics, and the social responsibilities which their method of work requires them to assume, as well as the vital philosophy from which it must have proceeded. I do not wish to appear flippant, but cannot at the moment raise the spirits to answer in the high tone which these questions deserve. As to co-operation, I hope I don’t refuse it, but I have noticed that, when you give a party, the best thing is not to rush at everybody and try to force them to talk to each other; in a party that goes unexpectedly well (in the way of making disparate groups talk to each other) you often notice that the host has been stuck away in the corner most of the time, talking to some expert about a technical point he is really interested in. As to responsibility, there was an earnest lady in the last war who took the opportunity of an introduction to President Roosevelt to urge upon him the great weight of the responsibility which she trusted he was properly conscious of, and his only reply (with a rudeness quite unusual to him) was an impatient movement of his hand. Obviously it would have made his judgment fatally bad if he worried all the time about his responsibilities. Even in so humble a walk of life as literary criticism, it seems to me, a man might feel the same; the best thing for him to do as a critic is to do his work as best he can, and he has still plenty of responsibilities as a social being. But of course I do not mean to deny that he really has responsibilities as a critic too. It seems clear that critics have been making a steady effort to act on them.

Back to top ↑

Sign up for Our Email Newsletter