Philosophy Looks at the Arts
ed. Joseph Margolis
(Third Edition, 1987)
[381]Intention and Interpretation in Criticism
FRANK CIOFFI
If we adapt Wittgenstein's characterization of philosophy: "putting into order our notions as to what can be said about the world," we have a programme for aesthetics: "putting into order our notions as to what can be said about works of art."
One of the tasks of such a programme would be to elucidate the relation in which biographical data about an author, particularly of the kind loosely known as knowledge of his intentions, stand to those issues we call matters of interpretation. I.e., the relation between questions like these:
Whether it is Goethe who is referred to in the first line of the first canto of In Memoriam.
Whether it is the poet who is speaking in the concluding lines of Ode on a Grecian Urn.
Whether Pope is 'screaming with malignant fury' in his character of Sporus.
Whether Hamlet in his famous soliloquy is contemplating suicide or assassination.
Whether Milton's Satan in his speech in Book IV beginning 'League with thee I seek' may be wholly or partly sincere.
Whether the governess who tells the story in James's Turn of the Screw is a neurotic case of sex-repression and the ghosts not real ghosts but hallucinations.
Whether Wordsworth's Ode: Intimations of Immortality is 'a conscious farewell to his art, a dirge sung over his departing powers' or is a "dedication to new powers'; and whether the 'timely utterance' referred to in that poem is My Heart Leaps Up or Resolution and Independence.
[382]
Whether we are meant to reflect that Othello becomes jealous very quickly on very little provocation.
Whether the Moses of Michelangelo is about to hurl the tablets of the law to the ground or has just overcome an impulse to do so.
Whether Shakespeare's Sonnet 73 contains an allusion to despoilt and abandoned monasteries.
Whether the image which floats before Yeats's mind in Among School Children, 'hollow of cheek,' is of an old woman or of one beautiful in a quattrocento way.
Whether the metaphors in Othello's soliloquy which begins "Steep me in poverty to the very lips' are deliberately inappropriate so as to suggest the disorder of Othello's mind.
Whether Ford Maddox Ford's novel Parade's End is a trilogy or tetralogy.
Whether on reading the line 'in spite of that we call this Friday good' from East Coker we are to think of Robinson Cruse's friend.
Whether Gertrude's marriage to Claudius was incestuous.
Whether Othello was black or brown.
Whether Pippit in Eliot's A Cooking Egg is young or old, of the same social status as the speaker or not, and whether the connotations of the expression 'penny-world' in that poem are sordid or tender.
Whether in The Mystery of Edwin Drood Dickens has deepened his analysis of Victorian society to include Imperialism; and whether John Jasper in that novel is a member of the Indian sect of Thugs.
Whether we should identify with Strether in The Ambassadors and whether the Ververs in The Golden Bowl are unqualifiedly admirable.
And statements like these:
That Eliot associates the Hanged Man, a member of the traditional tarot pack, with the Hanged God of Frazer and with the hooded figure in the passage of the disciples to Emmaus.
That Hopkins said: The Sonnet on Purcell means this: 1-4 I hope Purcell is not damned for being a protestant because I love his genius, etc., etc. 'Low lays him' means 'lays him low, 'listed is enlisted' etc., etc.
That Wordsworth wrote Resolution and Independence while engaged on the first part of the Immortality Ode.
That Henry James in 1895 had his faith in himself shaken by the failure of his plays.
That Donne's Valediction: Forbidding Mourning was addressed to his wife.
That James was conversant enough with English ways to know that no headmaster would have expelled a boy belonging to a county family without grave reasons.
That A. E. Housman vehemently repudiated the view that his poem 1887 contained a gibe at the Queen.
[383]
That Swift was philanthropic and well-loved by his friends.
That Maude Gone was an old woman when Yeats wrote Among School Children.
That Keats in his letters uses the word "beauty' to mean something much more subtle than is ordinarily meant by it.
That Eliot meant the lines 'to Carthage then I came, burning, burning, burning.. to evoke the presence of St. Augustine and the Buddha, of Western and Eastern Ascetism.
That Abraham Cowley had had very little to do with women.
That ruined monasteries were a not uncommon sight in 1685,
That Henry James meant his later novels to illustrate his father's metaphysical system.
That Conrad in The Arrow of Gold has an unrealistic and sentimental- ized portrait of a woman.
That Wordsworth nowhere in his work uses the word 'glory' to refer to his creative powers.
That Eliot told someone that Richards in his account of A Cooking Egg was "barking up the wrong tree.
That Tennyson shortly before his death told an American gentleman that he was referring to Goethe when he wrote 'of him who sings to one clear harp in divers tones.
That there were no industrial mills when Blake wrote "Jerusalem."
That in the book of Genesis Moses shattered the tablets of the law.
That Dickens wrote a letter at the time of the Sepoy mutiny advocating the extermination of the Indian people.
In this paper I have assembled and invented examples of arguments which use biographical claims to resolve questions of interpretation and confronted them with a meta-critical dogma to the effect that there exists an operation variously known as analyzing or explicating or appealing to the text and that criticism should confine itself to this, in particular eschewing biographical enquiries.
By now any of you who are at all interested in this topic must have had the phrase "the intentional fallacy' occur to you. This phrase owes its currency to a widely anthologized and often-alluded-to paper of that title by two Americans, Wimsatt and Beardsley. I want now to try to bring what they say in it into relation with the issue I have raised.
I
The first statement of their thesis runs: "The design or intention of the author is neither available nor desirable as a standard for judging the success of a literary work of art." These words don't really mean what they say. They don't mean that an artist may have intended to create a masterpiece but for all that have failed to do so; for the authors go on to
[384]
say of their thesis that it entails "many specific truths about inspiration, authenticity, biography, literary history and scholarship," etc., and none of these specific truths follows from the truism that knowledge of an artist's intentions cannot provide us with criteria for judging of his success. The charitable conviction that they mean more than this is borne out by a later statement of the thesis; this time to the effect that it is a thesis about the meaning of a work of art that they are concerned to advance: that certain ways of establishing this meaning are legitimate whereas others are not. So, presumably, what they intended to say is: "the design or intention of an author is neither available nor desirable as a standard for judging the meaning of a literary work of art." But no argument can profitably be conducted in these terms. For if a discrepancy should come to light between a reader's interpretation of a work and the interpretation of the author or his contemporaries, no way of determining which of these could be properly described as the meaning of the work could be produced.
What an author meant, by a poem, say, what his contemporaries took him to mean, what the common reader makes of it and what makes the best poem of it are usually concomitant and allow us to speak of the meaning without equivocation. If when confronted by instances in which this concomitance breaks down we appeal to only one of the ordinarily coincident features as if we had a settled convention behind us, the question becomes intractable. If the question is expressed instead as' "How should this poem be read?" it at least becomes clearer what the issues are. So the thesis becomes, "The design or intention of an author is neither available nor desirable as a standard for judging how a work of literature should be read." But does any criticism of literature consist of the provision of standards by which you may judge how the work should be read? One of the pieces of criticism which the authors have provided in their paper as an illustration of how it should be done concerns Eliot's The Love Song of J. Alfred Prufrock. They say that when Prufrock asks, "would it have been worth while ... to have squeezed the universe into a ball," "his words take half their sadness and irony from certain energetic and passionate lines of Marvell's To His Coy Mistress." This may be true and it may be helpful but nothing in it answers to the description of providing a standard by which the work may be read. What they have done or have tried to do is to produce in the reader a more adequate response to Eliot's lines by reminding him of Marvell's. If we bring their thesis in line with their practice it becomes: "The design or intention of the author is neither available nor desirable as a means of influencing a reader's response to a literary work." But since they give as an example of what they consider as irrelevant to criticism the fact that Coleridge read Purchas, Bartram, Milton and Bruce and this is not a fact about either
[385]
his design or his intention it is obvious that they mean something rather wider than this, something which the expression "biographical data' would be a closer approximation to. This gives us "biographical data about an author, particularly concerning his artistic intentions, is not desirable [I omit available as probably being just a sign of nervousness] as a means of influencing a reader's response to a literary work."
What any general thesis about the relevance of intention to interpreta- tion overlooks is the heterogeneity of the contexts in which questions of interpretation arise. This heterogeneity makes it impossible to give a general answer to the question of what the relevance of intention to interpretation is. There are cases in which we have an interpretation which satisfies us but which we feel depends on certain facts being the case. It may involve an allusion and we may wish to be reassured that the author was in a position to make the allusion. In this case biographical facts act as a kind of sieve which exclude certain possibilities. Then there is the case where we are puzzled, perhaps by an allusion we don't understand, perhaps by syntax, and reference to the author's intention, though it does not guarantee a favorable response, may at least relieve this perplexity and make one possible. There are cases in which we suspect irony but the text is equivocal, and cases where we aren't sure what view the author wishes us to take of the situation he places before us. Then there are the most interesting cases, those in which the text seems unmistakably to call for a certain interpretation and this is found satisfying, but in which we learn with surprise that it has been explicitly repudiated by the author. Even within the same kind of context the author's intention will vary in relevance depending on the kind of question involved; whether it concerns the meaning of a word or the tone of a passage, the view to be taken of a character or a situation or the general moral of an entire work.
Why did Wimsatt and Beardsley think they had a general answer to the question of deciding what the response to a work of literature should be? This is what they say:
There is a difference between internal and external evidence for the meaning of a poem. And the paradox is only verbal and superficial that what is (1) internal is also public. It is discovered through the semantics and syntax of a poem, through our habitual knowledge of the language, through grammars, dictionaries and all the literature which is a source of dictionaries, in general through all that makes a language and culture; while what is (2) external is private or idiosyncratic; not a part of the work as a linguistic fact; it consists of revelations (in journals, for example, or letters, or reported conversa- tions) about how or why the poet wrote the poem--to what lady, while sitting on what lawn, or at the death of what friend or brother. There[386]
is (3) an intermediate kind of evidence about the character of the author or about private or semi-private meanings attached to words or topics by an author or by a coterie of which he is a member. The meanings of words is the history of words, and the biography of an author, his use of a word, and the associations which the word had for him are part of the word's history and meaning. But the three types of evidence, especially (2) and (3), shade into one another so subtly that it is not always easy to draw a line between examples, and hence arises the difficulty for criticism.It is not clear from this account what the authors mean to exclude as illicit sources of interpretive data. Once the author's character and the private associations a word may have for him are admitted among these, along with all that makes a language and a culture, what is there left to commit fallacies with? Were it not that their illustrations give a much clearer impression of their attitude than their attempts at explicit formula- tion of it do, and show it to be much more restrictive, they could be suspected of advancing one of those enchanted theses which possess the magical power of transforming themselves into truisms at the touch of a counterexample. They say of a line in Eliot's The Love Song of J. Alfred Prufrock: "I have heard the mermaids singing each to each" that it bears some resemblance to a line of Donne's: "Teach me to heare Mermaids singing" so that the question arises whether Eliot's line contains an allusion to Donne's. They go on to say that there are two radically different ways of answering this question. The way of poetic exegesis and the way of biographical enquiry, and the latter would not be a critical enquiry and would have nothing to do with the poem. The method of poetic exegesis consists of asking whether it would make any sense if Donne's mermaids were being alluded to. The biographical approach would be to ask Eliot what he thought at the time he wrote it, whether he had Donne's mermaids in mind. The answer to this question would be critically irrelevant. It is not surprising that their example bears them out since it was hand-chosen, as it were. To expose its tendentiousness we need only take an example in which it was felt that a literary allusion would enhance the value of the lines. Let us take their own example of Marvell's To His Coy Mistress, familiarity with which they maintain enhances the value of certain lines of Eliot's. If we take the case of someone not familiar with Marvell's To His Coy Mistress, then the biographical claim that Eliot alludes to it in Prufrock would enhance its value for them. If on the other hand they merely applied the test of poetic exegesis and incorporated the allusion to Marvell's To His Coy Mistress into the poem without knowing whether Eliot was alluding to it, it is doubtful whether their appreciation would
[387]
survive the discovery that he was not. If a critical remark is one which has the power to modify our apprehension of a work, then biographical remarks can be critical. They can serve the eliminative function of showing that certain interpretations of a work are based on mistaken beliefs about the author's state of knowledge.
II
We can illustrate this eliminative function of biographical data by taking the very case on which Wimsatt and Beardsley based their arguments as to its irrelevance. They quote a quatrain from John Donne's A Valediction: Forbidding Mourning:
Moving of the earth brings harmes and fears,
Men reckon what it did and meant,
But trepidation of the spheares,
Though greater farre, is innocent.They then go on to criticize an interpretation of this quatrain which basing itself on the biographical fact that Donne was intensely interested in the new astronomy and its theological repercussions sees in the phrase 'Moving of the earth' an allusion to the recently discovered rotation of the earth round the sun. Wimsatt and Beardsley show the unlikelihood of this, not by disputing the well-authenticated facts concerning Donne's interest in astronomy, which would be to use a biographical method, but through an analysis of the text. They maintain that whereas the fear which is produced by the rotation of the earth is a metaphysical, intellectual one, the fear which Donne is attempting to discourage is of the emotional kind which an earthquake is more likely to produce and that this accords better with the "tear-floods' and 'sigh-tempests' of the poem's second stanza than the earth's rotation. Let us concede that the authors have made it very plausible that Donne was alluding not to the heliocentric theory of the earth's rotation but to earthquakes. The gratuitousness of the conclusion which they draw from this becomes apparent if we ask the following question: have they established that Donne was not referring to the rotation of the earth as conclusively as the fact that Donne was ignorant of the heliocentric theory that would establish it? Wouldn't this external fact outweigh all their internal ones?
At this point someone who finds my question unrhetorical is thinking to himself dark satanic mills.' It is true that the knowledge that the poem that prefaces Blake's Milton is not an expression of the Fabian sentiments it has been traditionally taken as has not caused the traditional interpreta-
[388]
ton to be abandoned. I suggest that what we have in this case is something in the nature of a spontaneous adaptation of Blake's poem. It is unlike what we ordinarily consider an adaptation in not being conscious (initially at any rate) and not involving any physical change in the work adapted. Does the fact that this was possible in the case of Blake's lyric reflect adversely on it as poem? Does the fact that the melody of God Save the Queen could be fitted with new words and become the national anthem of a republican nation reflect on it? The combination of resolution and exaltation which characterize Blake's poem carries over into its adaptation; it functions like a melody. We should see cases like that of "Jerusalem" as continuous with more obvious cases of adaptation. When Pistol tells French audiences that his "rendezvous is quite cut off," his Doll lies dying of Maladie of Naples. Does anything follow as to the relevance or irrelevance of an author's intentions? Then neither does it in the case of Jerusalem. It would only follow if the discovery that a work was an adaptation made no difference. There is one sort of literature in which adaptation is a matter of indifference: jokes. Wilkes becomes Disraeli and Disraeli becomes Birkenhead. The two Jews become two Irishmen or two Chinese. But then, we speak of the author of a poem but not of the author of a joke. I am saying: we don't stand in the same relation to Blake's lyric after changing our conviction as to what he meant to convey as we did before. If the case were one in which the discrepancy between the author's interpretation and the reader's were one as to the very emotions expressed and not just the accompanying imagery our attitude would be very different. Frank Harris read A. E. Housman's poem 1887 as an anti-im- perialist gibe and the expression God Save the Queen' which recurs in it as a sarcastic jeer until Housman revealed otherwise. Thereafter he naturally found it difficult to do so in spite of his conviction both as to the superiority of his interpretation and its greater consonance with Housman's general outlook. "How was I to know that someone steeped in a savage disgust of life could take pleasure in outcheapening Kipling at his cheapest?")
The following examples should make it clear how inept Wimsatt's and Beardsley's characterization of the role of biographical data in critical discourse is. An example which seems to support their account is Leavis' reaction to John Middleton Murry's attempt to give the word 'beauty' in the concluding couplet of Keat's Ode on a Grecian Urn a less limiting sense based on the use Keats made of the word in his letters. "To show from the letters that "beauty' became for Keats a very subtle and embracing concept and that in his use the term takes on meanings that it could not possibly have for the uninitiated is gratuitous and irrelevant. However, his use of the word may have developed as he matured, 'beauty' is the term
[389]
he used and in calling what seemed to him the supreme thing in life beauty' he expresses a given bent--the bent everywhere manifest in the quality of his verse, in its loveliness… and that beauty in the Ode on a Grecian Urn expresses this bent is plain, that it should as the essence of the poem, and there is nothing in the poem to suggest otherwise."
This may sound as if a general principle akin to Wimsatt's and Beardsley's is being employed, but that this is not so Leavis's practice elsewhere shows. For example, "Hopkins' Henry Purcell is a curious special case, there can be few readers who have not found it strangely expressive and few who could have elucidated it without extraneous help. It is not independent of the explanatory note by Hopkins that Bridges prints. Yet when one approaches it with the note fresh in mind, the intended meaning seems to be sufficiently in the poem to allay at any rate the dissatisfaction caused by baffled understanding." We must not be misled by the expression "a curious special case." Leavis's dealings with The Waste Land make it clear that the only question which arises in connection with notes or other extraneous aids to understanding is not one of their legitimacy but of their efficacy. For example, Leavis says of The Waste Land that it "sometimes depends on external support in ways that can hardly be justified. for instance, the end of the third section The Fire Sermon. .. No amount of reading of the Confessions or Buddhism in Translation will give these few words power to evoke the kind of presence that seems necessary to the poem." Of another passage he writes: 'it leaves too much to Miss Weston; repeated recourse to Ritual and Romance will not invest it with the virtues it would assume." On the other hand, of Eliot's note on Tiresias, Leavis remarks, "if Mr. Eliot's readers have a right to a grievance, it is that he has not given this note more salience." Power to evoke,' Invest with Virtues,' these are not the idioms in which the probative value of statements is weighed.
Wimsatt and Beardsley are aware of the problem posed them by Eliot's notes to The Waste Land and make the suggestion that the notes should be considered as part of the poem. They thus become internal evidence, and may be consulted with a good conscience. Does it follow that since the effectiveness of certain lines in Prufrock depends on familiarity with Marvell's Coy Mistress, Marvell's poem should be considered part of Eliot's, or does this not follow because whereas we are expected to be familiar with Marvell's poem, familiarity with the contents of Eliot's notes is not expected of us? Then is this what the distinction between external and internal evidence comes to; the difference between what we can and can't be expected to know? and how is it decided what we can be expected to know? Leavis has said of Quinton Anderson's book on Henry James, "thanks to the light shed by Mr. Anderson, we can see in the peculiar
[390] impressiveness of Mrs. Lowder of the Wings of the Dove a triumph of morality art." Is Mr. Anderson's book also to be considered part of James's Wings of the Dove then?
No amount of tinkering can save Wimsatt and Beardsley's distinction between internal and external evidence. It isn't just that it's made in the wrong place, but that it is misconceived from the start. A reader's response to a work will vary with what he knows; one of the things which he knows and with which his responses will vary is what the author had in mind, or what he intended. The distinction between what different people know of an author before reading his work or what the same person knows on successive occasions can't be a logical one. When is a remark a critical remark about the poem and when a biographical one about the author? The difficulty in obeying the injunction to ignore the biographical facts and cultivate the critical ones is that you can't know which is which until after you have read the work in the light of them.
The assumption which stultifies their exposition is the conception of critical argument as the production and evaluation of evidence. They say that there are two kinds of evidence: that provided by poetic exegesis and that provided by biographical enquiry. But the examples they give of poetic exegesis seem not to be evidence but conclusions or judgments. For example, that the lines from Prufrock take half their sadness and irony from lines in a poem of Marvell's, or that the mermaids in Prufrock derive no benefit from a reminiscence of the mermaids in Donne. We could construe these statements as evidence, only by taking them as biographical statements about Wimsatt and Beardsley, but so taken they would stand in the same relation to critical judgment as biographical statements about Eliot. If a critical remark fails to confirm or consolidate or transform a reader's interpretation of a work it will then become for him just evidence of something or other, perhaps the critic's obtuseness. Biographical remarks are no more prone to this fate than any others.
III
In the sixth stanza of Yeats's Among School Children there occur the lines:
Plato thought nature but a spume that plays
Upon a ghostly paradigm of things;
Solider Aristotle played the taws
Upon the bottom of a king of kings.Many editions give the first word of the third line 'solider' as "soldier.' This is due to a compositor's error, a transposition of two letters which went
[391]
unnoticed because by a fluke instead of producing gibberish, it produced the English word 'soldier.'
The American critic Delmore Schwartz was thus led to advance his well-known interpretation to the effect that the expression 'soldier Aris- totle' alludes to a legend that Aristotle accompanied Alexander on his military expedition to India. Since there is obviously a contrast intended between the unworldliness of Plato and the down-to-earthness of Aristotle, Schwartz's military interpretation accords well with the rest of the poem. But in spite of this, now that we know of the error wouldn't we insist on the restoration of the lines as Yeats wrote them and regard the view that there is a military allusion in the lines as a mistake? It might be objected that this is not to the point because the case here is one of a discrepancy between what the author wrote and what we made of it and not between what he meant and what we made of it.
But can this distinction be upheld? Can't we imagine cases where the words were homophones? In such a case the only distinction between what an author wrote and a mistaken reading would be what he meant. In fact, we needn't imagine such a case. Hopkins's note on his poem Henry Purcell provides us with one: "One thing disquiets me: I meant 'fair fall' to mean "fair (fortune be) fall?: it has since struck me that perhaps 'fair' is an adjective proper and in the predicate and can only be used in cases like "fair fall the day,' that is 'may the day fall, turn out fair.' My lines will yield a sense that way indeed, but I never meant it so." Is the possible meaning mentioned but rejected by Hopkins any more tenable than 'soldier Aristotle'?
There is thus no doubt that there are cases in which knowledge of an author's avowed intention in respect of his work exercises a coercive influence on our apprehension of it. The question now arises: When doesn't it? My answer is, "When the issue is of a complexity comparable to that which would cause us to discount his avowed intention in respect of something not a work of literature." To put it another way, we tend to think that there are cases where we override the author's intention and persist in an interpretation which he has rejected, but what we are really doing could less misleadingly be described as favoring one criterion of intention as against another. If we establish the existence of a discrepancy between the interpretation we give to a work of art, and that of the author, we haven't shown that the work has a meaning independent of what the author intends because what the author intends will now be the interpreta- tion given to the work by us and his own statement as to its meaning an aberration. The notion of the author's intention is logically tied to the interpretation we give to his work. It's not just that our language works this way; but that our minds do. Confronted with a choice between saying that an effect so complex could have come about by accident and that the
[392]
author was mistaken we would opt for the latter. The work will be considered more conclusive evidence of his intention than his own state- ments. The color flows back.
Edmund Wilson's dealings with Henry James's The Turn of the Screw bring this out clearly. The Turn of the Screw was generally considered a superior ghost story until Wilson popularized the view that the ghosts were figments of the narrator's imagination and the work a study in thwarted Anglo-Saxon spinsterdom. He thought he had discovered that the text was skillfully ambiguous so as never unequivocally to imply the ghosts' objec- tive existence. He was able to interpret some passages in James's preface to the book to similar effect. The publication of James's notebooks some years later, however, made it clear that James's conscious intention was to produce a ghost story. At the same time Wilson came to admit that the text itself was not completely reconcilable with his thesis that the ghosts were hallucinatory. Nevertheless, Wilson continued to insist that it was not a straightforward ghost story, but a study in the neurotic effects of repressed sexuality. His arguments for this provide an excellent example of what I have called the color flowing back. Instead of simply enjoying a gratuitous effect for its own sake, Wilson convinces himself on the basis of certain biographical facts about James that at the time the book was written, his faith in himself had been shaken and that "in The Turn of the Screw, not merely is the governess self-deceived, but that James is self- deceived about her. The doubt that some readers feel as to the soundness of the governess' story are the reflection of doubts communicated uncon- sciously by James himself."
The real interest of this kind of example is that it brings out quite clearly what otherwise is not so apparent; that there is an implicit biographical reference in our response to literature. It is, if you like, part of our concept of literature. It is only when it is missing that we notice that it was always there.
I want now to deal with some notorious ostensible counterexamples. This is the fifth stanza of Yeats's Among School Children:
What young mother a shape upon her lap
Honey of generation had betrayed,
And that must sleep, shriek, struggle to escape
As recollection or the drug decide,
Would think her son, did she but see that shape
With sixty or more winters on its head
Compensation for the pang of his birth,
Or the uncertainty of his setting forth?[393]
There is an accompanying note to this poem which indicates that the phrase "honey of generation" is taken from an essay of Porphyry's and that Yeats has arbitrarily used it to refer to "the drug that destroys the recollection of pre-natal freedom." It is then, the shape upon the mother's lap, the child, which has been betrayed by being born. John Wain has put forward a reading according to which it is the mother who has been betrayed, and "honey of generation" is an allusion to the sexual pleasure which accompanies conception, and the desire for which has betrayed her. Doesn't this example show the irrelevance of intention? Not necessarily. It could be interpreted as a case where we take the poem as better evidence of what the poet intended than his own explicit remarks on the subject. To persist in an interpretation in spite of an author's explicit disavowal of it is not necessarily to show an indifference to the author's intention. For we may feel that he was mistaken as to what his intention was. A case which comes to mind is Goldsmith's withdrawal of the gloss he offered on the word "slow" in the first line of his poem, The Traveller, "Remote un- friended, melancholy, slow." Goldsmith said it meant "tardiness of locomo- tion" until contradicted by Johnson. "No sir. You do not mean tardiness of locomotion. You mean that sluggishness of mind that comes upon a man in solitude."
Though it might be true, as Wimsatt and Beardsley say, that critical enquiries are not settled like bets, neither may questions of intention. I can't resort here to the argument I used in the case of Donne's Valediction and ask you to imagine what your attitude to Wain's interpretation would be if you were convinced that Yeats was ignorant of the fact on which it is based, since this fact, that sexual pleasure is an incentive to procreation, is not such as can be overlooked. Nevertheless, I want to maintain that we don't, if we accept Wain's interpretation, think it an accident that it should be possible to read the text as he does, but we feel that the ambiguity which makes it possible was the result of a connection in Yeats's mind between the expression "honey of generation" and sexual pleasure. (In fact, this can be demonstrated.)
In order to convince you that an implicit biographical influence is at work even in Wain's interpretation, I want you to imagine the case altered in some important respects. Imagine that the reading according to which it is the mother who is betrayed, was also that of Yeats, and that there was no footnote referring to Porphyry's essay, of which Yeats was com- pletely ignorant, but that a reader familiar with it and sharing its views on prenatal existence, insisted on taking the expression "honey of genera- tion" as an allusion to the drug which destroys the recollection of prenatal freedom and, therefore, to the infant and not the mother as betrayed. Wouldn't our attitude to this interpretation be quite different from our
[394]
attitude to Wain's? Wouldn't we feel it perverse? And since it can't be the text which makes it perverse but only the facts about Yeats as we have imagined them, doesn't our implicit biographical or intentionalistic ap. proach to literature emerge quite clearly here? Of course, there are cases where the pleasure we take in literature doesn't depend on this implicit biographical reference. Literature, as Wittgenstein probably said, is a motley. Nursery rhymes come to mind as the most notable example. But in general we do make such a reference. Eliot's attitude toward a line of Cyril Tourneur's illustrates this reluctance to take pleasure in what is accidental and unintended. The line is: "The poor benefit of a bewildering minute, " which is given as "The poor benefit of a bewitching minute" in the texts both of Churton Collins and of Nicoll who mention no alterna- tive reading. Eliot comments; "it is a pity if they be right for "bewildering' is much the richer word here." It has been argued that if the folio text of Henry V was right and Theobald's lovely guess wrong so that Shakespeare made the dying Falstaff allude to a painting rather than babble of green fields most of us would persist in reading the traditional and incorrect version. We probably would but it would worry us; and if "a babbled of green fields" wasn't even Theobald's guess but a transcriber's or printer's unthinking error, it would worry us even more. The suspicion that a poetic effect is an artifact is fatal to the enjoyment which literature characteristically offers. If the faces on Mount Rushmore were the effect of the action of wind and rain, our relation to them would be very different.
IV
In the course of their criticism of the interpretation of Donne's poem which saw in it an allusion to the rotation of the earth round the sun Wimsatt and Beardsley remark, "But the text itself remains to be dealt with ..."!
Where understanding fails, says Goethe, there immediately comes a word to take its place. In the case the word is "text." Let us appeal to the text. But what is the text? These critics talk of the text of a poem as if it had an outline as neat and definite as the page on which it is printed. If you remind yourself of how questions about what is "in the text' are settled you will see that they involve a great deal which is not 'in the text. Though there are many occasions on which we can make the distinction in an immediately intelligible and nontendentious way, where an interpretative issue has already arisen, the use of a distinction between internal, licit considerations, and external, illicit ones is just a form of question-begging.
What are we to say of attempts to support an interpretation by citing other works of the author? For example, Leavis on Conrad's Heart of
[395]
Darkness; "If any reader of that tale felt that the irony permitted a doubt regarding Conrad's attitude towards the Intended, the presentment of Rita (in The Arrow of Gold) should settle it." Isn't this illicit? Isn't the common authorship of several works a biographical fact?
What of the use of previous drafts of a work for critical purposes? Leavis in commenting on Hopkins's Spelt from Sybil's Leaves is able to enforce his point that "Hopkins' positives waver and change places and he is left in terrible doubt by showing that in a previous draft of the poem the word-order in the phrase 'black, white; right, wrong' was convention- ally symmetrical "black, white; wrong, right." The only doubt which might arise is connection with Leavis's point is whether the word order may have been altered to avoid a rhyme, but this is equally intentionalistic.
Marius Bewley supports his interpretation of James's The Turn of the Screw by pointing out that when James collected his stories for the definitive edition he put it in the same volume as one called The Liar.
Even if the anti-intentionalist thesis were qualified to accommodate all these there would still be a fundamental objection to it.
You must all have had the experience while reading of having the words suddenly undergo a radical transformation as you realized you had missed the end of a quotation, say, and mistaken the speaker. The more familiar the speakers the greater the transformation when you realized your mis- take. Doesn't this illustrate the importance of implicit biographical as- sumptions in interpreting what we read? Here's an illustration: In Rudyard Kipling's Loot occur the lines:
An' if you treat a nigger to a dose of cleanin'-rod
'E's like to show you everything he owns.Hugh Kingsmill has quoted these lines as an example of Kipling's brutality, and even Kipling's biographer Edward Shanks is embarrassed by them. Edmund Wilson, on the other hand, in his well-known essay on Kipling, says this about them: "Kipling was interested in the soldier for his own sake, and made some effort to present his life as it seemed to the soldier himself. The poem called Loot, for example, which appears to celebrate a reprehensible practice is in reality perfectly legitimate because it simply describes one of the features of the soldier's experience in India. There is no moral one way or the other." T. S. Eliot takes a similar line in his introduction to his selection of Kipling's verse.
How is this issue to be decided? By an appeal to the text? Isn't it rather our sense of Kipling which will determine the side we come down on? A sense built up not only from the other tales but from his autobiography and other sources as well? Don't these throw a 'field of force' round the
[396]
work? If it had been written by someone else wouldn't this make a difference to our apprehension of it? Isn't this like the case described by Wittgenstein in the Philosophical Investigations? "I see a picture which represents a smiling face. What do I do if I take the smile now as a kind one, now as a malicious one? Don't I often imagine it with a spatial and temporal context which is one either of kindness or of malice? Thus I might supply the picture with the fancy that the smiler was smiling down at a child at play, or again on the suffering of an enemy.
The difference of opinion between F. R. Leavis and Marius Bewley over James's What Maisie Knew is an excellent illustration of an interpretation depending on 'the fiction I surround it with'. Unfortunately it is too long to quote, but the gist of it is that Bewley finds the atmosphere of the book one of horror and its theme the meaning and significance of evil, whereas Leavis can detect no horror and sees it as an extraordinarily high-spirited comedy reminiscent of the early part of David Copperfield. Bewley in attempting to locate the source of their difference says that it has its "origin in areas not readily open to literary-critical persuasion" and "that the way one senses the presence of evil and horror in the novel may be due to one's conception of them outside the novel.
There is one aspect of our response to a work of literature to which biographical data seem to have particular relevance and that is our conviction as to an author's sincerity. It is certain that there are cases where biographical considerations are genuinely relevant and equally cer- tain that there are cases where they are intrusions which we feel we ought not to allow to condition our response. But it is difficult to know where the line should be drawn. I suppose that we would all consider Beethoven's inability to get on with Scott's Kenilworth because "This man writes for money" as eccentric, though the decline in Trollope's reputation which followed his revelation as to his methods of composition and his businesslike attitude toward his writing in his autobiography, show it is not rare. Perhaps these responses should be considered more as moral gestures , like refusing to hear Gieseking perform, rather than as aesthetic responses .
A good example of a response which is genuinely critical but which we would all consider misplaced is Johnson's criticism of Abraham Cowley's The Mistress: "But the basis of all excellence is truth: he that professes love ought to feel its power. Petrarch was a real lover and Laura doubtless deserved his tenderness. Of Cowley we are told by Barnes, who had means enough of information, that whatever he may talk of his own inflam mability, and the variety of characters by which his heart was divided, he in reality was in love but once, and then never had resolution to tell his passion!
[397]
"This consideration cannot but abate in some measure the reader's esteem for the work. .
Another, perhaps slightly less conclusive example is provided by John- son's remarks on Cowley's poem on the death of Hervey . but when he wishes to make us weep he forgets to weep himself, and diverts his sorrow by imagining how his crown of bays, if he had it would crackle in the fire. It is the odd fate of this thought to be the worse for being true. The bay leaf crackles remarkably as it burns, as therefore this property was not assigned to it by chance, the mind must be thought sufficiently at ease that could attend to such minuteness of physiology." "It might be argued that Johnson has indicated a source of dissatisfaction in the poem, the bay leaves image. But it is what this enabled him to infer about something outside the poem concerning Cowley which abated his esteem. If he could have been convinced that Cowley was ignorant of the propen- sity of bay leaves to crackle remarkably and the felicity of his image therefore fortuitous, Johnson would presumably have liked the poem better. But it would be a mistake to think Johnson simply absurd here. Suppose that the poem in question were Bishop King's Exequy and the biographical fact that he was never married and therefore never bereaved. Some of us would decide it didn't matter, some that it did and some would oscillate. This is an example of the more general dilemma which arises when an empirical concomitance on which we habitually depend and so regular that it has influenced the build of our concepts, disintegrates. Van Meegeren's Disciples at Emmaus, the poems of Ern Malley, Macpherson's Ossian, Chatterton's Rowley, all point the same moral.
D. W. Harding raised a related issue in a vivid form some years ago. He wrote:' "We think of it (a work of art) as being a human product, as implicitly sanctioning and developing interests and ideals and attitudes of our own. That being so it does become disconcerting to find that for the author it satisfied certain impulses which we ourselves are glad not to possess, or which if we do possess we think better left unsatisfied. The same thing goes on in social intercourse of a simpler kind than literature. We enjoy the bon mot with which our friend disposes of a charlatan, but if we know that he is incidentally working off irrelevant spite against either the charlatan or the world in general the flavour of the remark is spoilt. The bon mot is as good as ever regarded as something impersonal, but as a human product it no longer gives us pure satisfaction--an element of distaste or regret comes in and makes our state of mind more complex. Many people find this more complex attitude extremely difficult to main- tain... especially because in most actual cases the neurotic flaw can be detected in the work itself." But once in possession of biographical data it is difficult to be sure what is "in the work itself." Leavis has suggested
[398]
it is a pity much is known of Pope's life since the expression of spite, envy, venom, and malice so often found in his work is a consequence of the distorting effect of this knowledge.
What I have called "putting a field of force round a work," surrounding it with a web of associations, may be effective even when it doesn't deserve to be. But this kind of suggestibility is a risk all critical remarks run and not merely biographical ones. Would anyone have found the last few lines of Bishop King's Exequy productive of an effect of terror if Eliot had not said so? And would Eliot himself if he had not first come across them in Poe's The Assignation? It is the fact that we can speak of criticism which is effective but mistaken which makes the analogy with argument so tempting for there too we speak of conclusions seeming to follow but not following; so it seems that we can have specious criticism in the same sense in which we have specious argument. But this is an illusion. You don't show that a response to a work of literature is inadequate or inappropriate in the way that you show that the conclusion of an argument has been wrongly drawn.
Wittgenstein has some remarks in Part Two of the Investigations, which shed light on the nature of the intractability which characterizes so much critical argument and makes its prevalence less surprising. His remarks though concerned with the question of the genuineness of an expression of feeling have a more general application. He contrasts our judgments about sincerity with those about color. "I am sure, sure that he is pretending: but some third person is not. Can I always convince him? And if not, is there some error in his reasoning or observations?" Though there are those whose judgment is better in such matters and rules for determin- ing this, these do not form a system and only experienced people can apply them. There are consequences which distinguish correct from incorrect judgment, but these are of a diffuse kind and like the rules incapable of general formulation. 66 . only in scattered cases can one arrive at a correct and fruitful judgment. » It is not surprising, then, that "the game often ends with one person relishing what another does not."
Conclusion
What I have been saying is this: a conviction that a poet stands in a certain relation to his words conditions our response to them. That this should be so seems to me part of the physiognomy of literature (as Wittgenstein might have put it). We are not ordinarily aware of this as these convictions tend to be held in solution in "the work itself.' It is only in exceptional circumstance that we crystallize them out as explicit beliefs and become aware of the role they play. Why should anyone wish to deny this? Because
[399]
it is then only a step to the production of phantasy-theses like Wimsatt's and Beardsley's, "What is said about a poem is subject to the same scrutiny as any statement in linguistics or in the general science of psychology."
This in its turn has its source in the determination to tidy up the activity of reading and to reduce what it involves to a neat logically homogenous set of considerations such as guarantee a readily communicable rationale. The idea of a work of literature as 'a linguistic fact' or an 'integrated symbol' is comparable to the notions of 'a concept' in philosophy or "behavior' in psychology in being the manifestation of an irresistible demand for discrete, coherent, and enduring objects of investigation. But, "Literature is a motley."