1
When Seven Types of Ambiguity was published in the United States in 1931 it was hardly noticed, purchased by just eight people; it sold steadily from the moment it was published the year before in Britain, however, where it was also widely reviewed.1 It was certainly noticed by Riding. Empson’s Dedication (or Preface or Acknowledgments; it is difficult to say how he regarded these sentences) reads: “Mr.I.A. Richards, then my supervisor for the first part of the English Tripos, told me to write this essay, and various things to put in it; my indebtedness to him is as great as such a thing should ever be. And I derive the method I am using from Mr. Robert Graves’ analysis of a Shakespeare Sonnet, ‘The expense of spirit in a waste of shame,’ in A Survey of Modernist Poetry.” This immediately gave rise to a dispute with Laura Riding, co-author of A Survey of Modernist Poetry, about who deserved credit for inspiring this method of reading literature.2 The dispute lasted more than forty years and became more intense each time they renewed it, perhaps a sign of the growing importance of New Criticism in literary history.
From the beginning, Riding regarded the method as developed at least as much by her as by Graves, and she ultimately insisted that the analysis of the sonnet in question was mostly hers. Increasingly frustrated by her determination to claim credit for a method of criticism that she seemed to disdain, Empson professed to find her behaviour “puzzling”: “if you despise it so much as your letters imply, whyever are you so keen to have priority in it?”3 Forty years later, after a year of acrimonious and unavailing correspondence with Empson from 1970 to 1971, Riding folded her tent, taking comfort from what she saw as the tide of literary history: “It seems to me appropriate to record that, without public statement of mine, recognition of my intellectually and verbally sensitive hand within the glove of the Survey method has been mounting, with perception of its connection, via Mr Empson’s hobby-horse use of it, with the ‘New Criticism,’ which tried to make real horse-flesh of it.”4
A silly quarrel comes thereby to obscure for eighty years what ought to have been acknowledged quite unambiguously in 1930: first, that Empson owed his method of reading poetry more to Robert Graves than to any other person; second, that although Laura Riding’s contribution to the development of this method was not nearly as important as Graves’s, neither was it negligible. In the chapters that follow, I will show the sources in Graves’s early works of the method that Empson develops: the method by which the close reading of ambiguity in poetry serves to identify the fullest possible range of meaning within it that is able to be reconciled into a coherent psychological and aesthetic whole. I begin by addressing the dispute between Riding and Empson concerning this question of influence: a question that ideologically progressive literary history has subsequently figured more as a question about who deserves blame for the origin of New Criticism than as a question about who deserves credit for it. It turns out that Empson owes a good deal to Riding, but not as much as she thinks he does.
•
In the fall of 1928, while I.A. Richards was writing Practical Criticism, his new pupil Empson discussed with him the method of reading literature suggested by A Survey of Modernist Poetry. Having completed his Cambridge degree in mathematics in the spring of that year, Empson had enrolled in English as a member of Magdalene College, which led to Richards’s becoming his Director of Studies. With regard to the meetings that followed as part of their effort to define a research topic for Empson, Richards recalled in 1940 that “At about his third visit he brought up the games of interpretation which Laura Riding and Robert Graves were playing with the unpunctuated form of ‘The expense of spirit in a waste of shame.’ Taking the sonnet as a conjuror takes his hat, he produced an endless swarm of lively rabbits from it, and ended by ‘You could do that with any poetry, couldn’t you?’”5
Richards recalls this conversation more than a decade after the fact. Looking back, he seems to have supplied Riding’s name only in retrospect to this account of his recollections of that conversation, for Empson – like so many others at the time, many of whom Riding and Graves chastised in their next book A Pamphlet Against Anthologies – treated A Survey of Modernist Poetry as another book of literary theory by Robert Graves, one that followed logically from the theory of poetry that he was developing in On English Poetry (1922), The Meaning of Dreams (1924), Poetic Unreason (1925), and Impenetrability, or the Proper Habit of English (1926).6 Either Empson had not noticed or he had not taken seriously the fact that there was another name before Graves’s on the title page of A Survey of Modernist Poetry. Prompted by Richards at the end of the meeting described above to explore this way of reading poetry more thoroughly and more systematically than Riding and Graves had done, Empson immediately went off to write Seven Types of Ambiguity – as we have seen, indicating in his introductory note that he thought the book’s reading of Shakespeare’s sonnet was the work of Graves alone.
When Riding learned in 1931 that Empson had made this mistake, she wrote to inform him of the error and to ask for a copy of the book. Empson duly arranged for a copy to be sent to her, and promptly apologized: “I am sorry not to have mentioned your name with his in the preface … I had not the book by me and forgot that it was a collaboration.”7 Riding and Graves had been determined from the beginning that their collaboration be acknowledged, prefacing their second book together, A Pamphlet Against Anthologies (1928), with a list of the newspapers, magazines, and journals whose reviewers had not properly acknowledged Riding’s co-authorship of A Survey of Modernist Poetry, many of them referring only to “Mr. Graves’s book”: “At the beginning of a previous work, A Survey of Modernist Poetry, we carefully described it as a word-by-word collaboration. We did this because it was obvious to us that the vulgarity of a certain type of English reviewer would be encouraged by the combined circumstances that the first author was a woman and that the second was a man whose name was perhaps better known to him than that of the first; and because we were interested to see how far this vulgarity would persist in spite of our statement … We therefore take a statistical pleasure in listing the … papers which succumbed, through their reviewers, to this vulgarity.”8 So it is not surprising that Riding and Graves did not let the matter drop with what Riding characterised as Empson’s “meaningless and useless” personal apology to her: she regarded what he had done as another instance of the “wilful omission of the first author’s name.”9 Publicly exposing this implicitly misogynist vulgarity was important to both authors. Graves wrote a long letter to Empson’s publishers the same day as Riding (an “enormous letter,” Empson recalled, denouncing him “for a plot against women”), endorsing Riding’s request that an erratum slip be added to all further copies sold, indicating the order of authorship as Riding first, Graves second.10 Riding actually wrote again a month later, asking that she be sent a copy of the erratum slip, for she was not willing to accept as fact a mere account from either the author or the publishers of what they had done.
Yet there was more to the story, although it would not emerge for many years – in further exchanges of correspondence between Empson and Riding. In 1966, when James Jensen reported in the Modern Language Quarterly the results of his investigation into the influence of A Survey of Modernist Poetry upon Seven Types of Ambiguity, the editor of the journal published replies by Empson, Richards, and Graves to his queries about their responses to Jensen’s essay – the editor having neglected to consult Laura (Riding) Jackson (as she had since become known), believing that she was dead.11 She was anything but. With his response to the editors, Graves put the cat amongst the pigeons and set up a second round of correspondence between Empson and Riding: “I was, I believe, responsible for most of the detailed examination of poems in A Survey of Modernist Poetry – for example showing the complex implications of Sonnet 129 before its eighteenth-century repunctuations.”12 When Empson read Jensen’s essay and Graves’s letter several years after they were published, he cited this material in his correspondence with Riding to justify once more his omission of her name from his original preface to the first edition of Seven Types of Ambiguity.
Empson wrote to Riding because Chatto and Windus, the publisher of Seven Types of Ambiguity, had asked him to take over their correspondence that year with Riding, who had just learned that there had been a revised edition of Empson’s book, published in 1947, and that in the Preface to this revised edition Empson had omitted all reference to A Survey of Modernist Poetry and thereby retracted his belated acknowledgment via the erratum slip of the dual influence of Riding and Graves upon him: “I ought to say in passing,” he writes in the Preface to the second edition, that “Mr. Robert Graves” “is, so far as I know, the inventor of the method of analysis I was using here.”13 Riding was determined that there be no further misattribution of the origin of the method to Graves, emphasizing in her letters to Chatto and Windus that she was “the actual originator of the technique”: “The method, as anyone familiar with my work, my thinking, my laborings with other poets for this better attention to the requirements of linguistic responsibility, is of my formation.”14
Perhaps still nursing wounded pride at the public revelation of his mistake that the erratum slip had marked, perhaps angered by demands for further acknowledgment, Empson tries to justify himself forty years after the fact by two new tactics. First, he attributes the main influence on Seven Types of Ambiguity to work by Graves that preceded the latter’s collaboration with Riding. He identifies in Graves’s On English Poetry (1922), which he describes as “mainly concerned” with a “Conflict theory of poetry,” “the necessary background for a theory of poetical ambiguity,” and he suggests that Graves had “reached it by 1926, with Impenetrability, or the Proper Habit of English.”15 Citing passages from these early works, Empson explains in 1970 the logic of his introductory note in 1930, insofar as he is able to reconstruct it in retrospect: “these passages, I thought, though they were really very decisive looked a bit scrappy, and when I got round to reading A Survey of Modernist Poetry (1927) I felt that the treatment of the ‘lust in action’ sonnet would be the right thing to mention in my acknowledgment. It dealt with a complete poem, as I was by this time trying to do, and it had a cumulative weight and impressiveness. What I thought about the collaborator I do not remember, but I suppose these few pages, so very unlike the rest of the book, seemed to me such an evident further step by the mind of Robert Graves that no collaborator could disagree.”16
Second, since Graves had recently “confessed” that the reading of the Shakespeare sonnet was his own, and not Riding’s, Empson notes that, as it turns out, his original acknowledgment of just Graves’s influence in the first edition of Seven Types of Ambiguity was actually quite accurate after all.17 Certainly “The analysis uses the idea of ambiguity of syntax, which may need to be made plain by unusual punctuation; I used this in my book a good deal, and it seems quite possible that I derived it from the analysis of ‘lust in action.’ If you assure me that you invented it, and not Robert Graves, I grant that I may be in your debt so far; I don’t remember any case of Robert Graves using ambiguity of syntax in his previous writing.”18 In the end, though, Empson concedes the possibility of Riding’s influence on his awareness of the significance of ambiguity of syntax in 1930 only because he regards it as not worth fighting about, for he has now decided, it seems, that it is not an important ambiguity: “I did use quite a bit of ‘ambiguity of syntax,’ which I now think a very dubious thing, and perhaps one which cannot occur in the sharp form needed to express a conflict. Anyway it is always a temptation to the analyst, because it gives him a big extra chance of forcing in his own ideas against the surface intention of the poem.”19
Finally, Riding and Empson ended their discussion of the question of her influence without resolving it. Riding certainly did not buy what she called Empson’s “early-Graves alibi.”20 The correspondence eventuated by the Modern Language Quarterly article left Riding convinced that Graves was a “liar” and that Empson was “an improviser of fabrications.”21 Empson was nasty too, declaring it impossible that he should ever have been influenced by anything that Riding wrote since “none of your work ever seemed capable of retaining my eye on the page.”22
•
When they were no longer collaborators and lovers, Riding and Graves each claimed the most celebrated parts of A Survey of Modernist Poetry – whether as the one who physically wrote the words, interpreted the literature, or theorized the critical method applied. And even if the analysis of Shakespeare’s Sonnet 129 was written wholly or largely by Graves (and it is certainly consistent with, although considerably more sophisticated than, his interpretation of poetic ambiguities in his earlier works), there is still the question of how much of his analysis was directed by Riding’s principles and ideas, if not actually dictated by her. Since Graves claims in 1966 responsibility “for most of the detailed examination of poems,” but attributes to Riding responsibility for certain of “the general principles,” the question of her influence upon Empson clearly remains unresolved.23 Empson was certainly inspired as much by the theoretical principles that Riding and Graves enunciated as by the practical criticism of the Shakespeare sonnet that they offered.
Similarly, Empson’s Pithian arrow about Riding’s unreadability begs the question of just how much of the “word-by-word collaboration” in A Survey of Modernist Poetry was written by Riding, for even if Graves was entirely responsible for the section on Shakespeare, Empson was certainly also influenced by other sections of the book.24 Besides the analysis of the significance of ambiguity of syntax in Shakespeare’s sonnet in A Survey of Modernist Poetry, for instance, there is a discussion of ambiguity of syntax in an analysis of the poetry of E.E. Cummings that was almost certainly written by Riding. So there are a number of other possible sources of Riding’s influence that Empson does not consider, let alone acknowledge.
An important feature of the reading of Shakespeare’s Sonnet 129 was the close study by Riding and Graves of changes made to the punctuation of the sonnet since 1609. Their attention to the history of punctuation in this sonnet is designed to suggest that it is the intolerable difficulty of reading Shakespeare “that provoked his editors to meddle with his texts as being too incomprehensible as they were written.”25 Their purpose in invoking Shakespeare’s poetry in their discussion of modern poetry is to show that despite the fact that “understanding of … poetry … like Shakespeare’s … is taken for granted,” Shakespeare’s poetry is actually more difficult to understand than that of an apparently difficult contemporary poet like E.E. Cummings.26 In fact, although one would not know it from the correspondence between Empson and Riding, it is in this analysis of Cummings’s work that our attention is first directed to the meaningfulness of ambiguity of syntax.
Riding and Graves do not use either the word “ambiguity” or the word “syntax” in their study of Shakespeare’s sonnet, let alone the phrase “ambiguity of syntax.” Yet their explanation of how “changes in punctuation do the most damage … to the meaning” of the poem certainly amounts to both an analysis and a defence of the poetic meaningfulness of what Empson calls “ambiguity of syntax,” for it is precisely in the suppression of ambiguities in the syntax of the poem that they say editors have done the most to damage a reader’s understanding of the poem.27 They note that in the sonnet’s first two lines, the replacement of the first comma with a semicolon ( “Th’ expence of Spirit in a waste of shame/Is lust in action, [;] and till action, lust”) prevents a syntactical ambiguity: “In the second line a semicolon after the first action instead of a comma … cuts off the idea at action instead of keeping in action and till action together.”28 The semicolon, that is, prevents the reader’s perceiving an instance of parallel construction that the 1609 version allows: “The expense of spirit in a waste of shame is lust in action and is lust till action.” In lines six and seven ( “and no sooner had [, ] Past reason hated”), “particularly serious is the interpolation of a comma after no sooner had; for this confines the phrase to a special meaning, i.e. ‘lust no sooner had is hated past reason,’ whereas it also means ‘lust no sooner had past reason is hated past reason.’”29 Again, emendation prevents the reader from perceiving the ambiguous parallel construction in the 1609 version. Similarly, they note significant changes and restrictions of meaning for the adverb “well” in the last two lines of the sonnet ( “All this the world well knows [;] yet none knows well, /To shun the heaven that leads men to this hell”). This line, “if unpunctuated except for the comma Shakespeare put at the end,” is a general statement of a central theme, “that lust as lust is satisfiable but that satisfied lust is in conflict with itself”: “The man in lust is torn between lust as he well-knows it with the world and lust in his personal experience, which crazes him to hope for more than lust from lust. The force of the second well is to deny the first well: no one really knows anything of lust except in personal experience, and only through personal experience can lust be known well rather than ‘well-known.’ But separate the second well from the first, as in the revised version, and the direct opposition between world and none, well knows and knows well is destroyed, as well as the whole point of the word-play between well knows and knows well; for by the removal of the comma after the second well, this is made merely an adverb to modify To shun in the following line – well here means merely successfully with To shun, not well enough with knowes.”30 In each case of interchangeable adjectival or adverbial modification, and in each case of ambiguous parallel construction, Riding and Graves find ambiguous syntax to be meaningful.
That is, the double readings that are possible in the 1609 version of the sonnet produce a pattern of distinctions which is significant precisely in terms of the doubleness it patterns: “All the distinctions in the poem between lust in action and lust till action, between lust In pursuit and lust in possession are made to show in the end that there are no real distinctions.”31 Depriving the poem of ambiguous parallel constructions deprives it of this theme that a lust is a lust is a lust. Similarly, the ambiguity as to which words certain adjectives and adverbs modify is systematically patterned, and meaningful in this patterning: “It must be kept in mind throughout that words qualifying the lust-business refer interchangeably to the taker (the man who lusts), the bait (the object of lust), and lust in the abstract.” So “Had may mean the swallowing of the bait by the taker, or the catching of the taker by the bait, or ‘lust had,’ or ‘had by lust.’”32 And so their argument that emendation to change punctuation removes meaning from the poem is indeed, as Empson notes, an observation about the significance of “ambiguity of syntax.”
For anyone interested to affirm Riding’s influence upon Empson, it is important to note that A Survey of Modernist Poetry does not confine discussion of the ambiguity of syntax to Shakespeare’s sonnet. In fact, the analysis of Shakespeare is intended to demonstrate that the ambiguity of his syntax is even greater than the ambiguity of syntax already identified in Cummings’s poem “Sunset.” Well before their famous analysis of Shakespeare’s sonnet, Riding and Graves demonstrate that one of the virtues of “Sunset” is that it is “able to stave off death by continually revealing, under examination, an unexpected reserve of new riddles” – the most interesting and fruitful source of which is its riddling syntax: “Did we not accept the poem as a non-grammatic construction and make sense of it nevertheless? Could we not show it to be potentially or even actually grammatic and make sense of it because it was grammatic? By reading swarms and chants, which we have probably been reading as nominative plural nouns, as third person singular verbs, and by reading silver and gold not as adjectives but as nouns?”33 Here is where the analysis of ambiguous syntax that Empson once found so interesting in A Survey of Modernist Poetry actually begins.
This analysis is quite likely by Riding. Graves writes to T.S. Eliot several times in 1926 about the work that his new collaborator Laura Gottschalk (as Riding was then known) is contributing to the book that will become A Survey of Modernist Poetry. Apparently, she was responsible for the book’s analysis of American poets: “She is far more in touch with the American side than I am.”34 Graves indicates that she has written an essay on “The H.D. Legend” ( “it might do for a part of this prospective book devoted to ‘Legends,’ e.g. the other Imagists, Sandburg, D.H. Lawrence”) and that she has written “on John Ransom and Marianne Moore; principally a preliminary essay on Regionalism as a critical clue in American poetry (i.e. a false clue).”35 Such was the division of labour that Riding treats even the American poets that Graves does know well: she writes on Ransom, though, by this point, Ransom has published Graves, corresponded with Graves, and even introduced Graves to Riding, and though, at this time, Graves has just passed along a manuscript of Ransom’s poems first to Eliot and then to the Woolfs, thereby arranging for Ransom’s first publication in Britain.
It is important to note that ambiguity of syntax is presented in both A Survey of Modernist Poetry and Seven Types of Ambiguity not only as a feature of literary language, but also as a feature of the language of the very best literature. For example, Riding and Graves note that “Shakespeare’s punctuation allows the variety of meanings he actually intends; if we must choose any one meaning, then we owe it to Shakespeare to choose at least one he intended and one embracing as many meanings as possible, that is, the most difficult meaning”; they would make the same point about E.E. Cummings in particular and modernist poets in general: “It is always the most difficult meaning that is most final.”36 Analysis of ambiguity of syntax in both the poem by Shakespeare and the poem by Cummings is offered in support of this principle, which will become a fundamental axiom of both Empson’s method in particular and the method of New Criticism in general.
In the conclusion to Chapter Two of Seven Types of Ambiguity Empson states: “I shall now return to Shakespeare and allow myself a couple of digressions; about the emendations of his text and his use of a particular grammatical form.”37 His survey of the history of emendations to Shakespeare’s texts suggests to him that a “conservative attitude to ambiguity” is operative: “it is assumed … that Shakespeare can only have meant one thing.”38 His phrasing follows that of Riding and Graves here. He follows them further in disavowing the practice of restricting meanings to special interpretations of special words – reversing it, in fact, so as to embrace as many meanings as possible – possibility in each case being determined by plausibility.
When reading Shakespeare for ambiguity, Empson seems to be converting the either/or strategy of “three centuries of scholars and critics” into the both/and strategy of poststructural theorists (he confesses that he has often behaved like the Arden editor: “I have myself usually said ‘either … or’ when meaning ‘both … and’”39) but he presents his method as a way of recovering what Shakespeare might have meant and what his original audiences might have thought he meant, thereby affirming an intentionality and a referentiality that are anathema to poststructuralism. The Arden editor, he notes, negates a whole host of associations for the word “rooky” in Macbeth before declaring the meaning he believes that Shakespeare intended: “This somewhat obscure epithet, however spelt (and it should be spelt rouky), does NOT mean ‘murky’ or ‘dusky’ …; NOR ‘damp,’ ‘misty,’ ‘steamy with exhalations’ …; NOR ‘misty,’ ‘gloomy’ …; NOR ‘where its fellows are already assembled’ … , and has NOTHING to do with the dialect word ‘roke’ meaning ‘mist,’ ‘steam,’ etc … the meaning here … I THINK, is simply ‘rouking’ or perching wood, i.e., where the rook (or crow) perches for the night.”40 Embracing all of these meanings as plausible – “these meanings … might, for all we know … have seemed plausible to anybody in the first-night audience; might have seemed plausible to Shakespeare himself, since he was no less sensitive to words than they” – Empson argues that “such a note … makes you bear in mind all the meanings it puts forward” and that “this is the normal experience of readers” of Shakespeare.41 And so “the reader must hold in mind a variety of things he may have meant, and weight them, in appreciating the poetry, according to their probabilities.”42
Similarly, fearing that his analysis of ambiguity in two lines from Macbeth ( “Come what may, /Time, and the Houre, runs through the roughest day”) “seems too elaborate,” he adds a footnote in 1947 to the second edition of his book, defending himself in terms that again recall Riding and Graves: “I cannot see what else (what less) the line means if it is taken seriously as meaning anything.”43 He has accepted the axiom of Riding and Graves that “if we must choose any one meaning, then we owe it to Shakespeare to choose at least one he intended and one embracing as many meanings as possible.”44 Otherwise, one risks stopping at “less” than he means.
Empson argues that it is simply not the case “that a great deal has been added to Shakespeare by the mere concentration upon him of wrong-headed literary attention.”45 The ambiguity revealed by scholarship and criticism is a function not of our interpretive blindness or wilful perversity, but of the text itself: “Here as in recent atomic physics there is a shift in progress, which tends to attach the notion of a probability to the natural object rather than to the fallibility of the human mind.”46 So Empson’s conclusion about the history of Shakespearean emendation is precisely the same as that of Riding and Graves: Shakespeare’s “original meaning was of a complexity to which we must now work our way back.”47
•
This principle that one owes it to the poet to perceive as much meaning as possible in a poem, of course, has practical implications, which Riding and Graves work out for Empson. In fact, his recommendation that one bear in mind, as one reads along, all that words and phrases and lines might mean is a development of the practical approach to reading that Riding and Graves recommend. In the hands of New Critics, this kind of attention to the manifold of meanings presented in a poem will come to be known as close reading.
For “getting out of the prose and into the poetic state of mind,” Riding and Graves suggest “developing a capacity for minuteness, for seeing all there is to see at a given point and for taking it all with one as one goes along.”48 Empson locates in Shakespeare a regular syntactical strategy for stimulating precisely this poetic state of mind: “a linguistic form common in Shakespeare’s verse, and typical of his method; ‘the (noun) and (noun) of (noun)’; in which two, often apparently quite different, words are flung together, followed by a word which seems to be intended to qualify both of them”; since “this form demands that the reader should find a highest common factor of its first two nouns, it implies that he must open his mind to all their associations, so that the common factor may be as high as possible. That is, it is a powerful means of forcing him to adopt a poetical attitude to words.”49
Riding and Graves’s definition of “the poetic state of mind” as a capacity for minute attention to all that words can mean clearly anticipates Empson’s definition of the “poetical attitude to words.” Empson finds this poetic state of mind described not in their analysis of Shakespeare’s sonnet, however, but rather in their analysis of a poem called “The Rugged Black of Anger,” which they present without identifying its author. It is, of course, a poem by Riding. Ironically, then, Empson is influenced by a section of A Survey of Modernist Poetry that requires him not only to read Riding, but also to read what seems to be Riding’s own analysis of her poem, which concludes the book’s central chapters on the American poets Marianne Moore, H.D., and E.E. Cummings (American poets having been identified by Graves as Riding’s bailiwick).50 Although he might not have recognized this as Riding’s poem and would later tell her that “none of your work has ever seemed to me capable of retaining my eye on the page,” Empson seems to have read her so closely as to have accepted that the best way to read all poetry is the way Riding recommends as the best way to read hers!
On the one hand, Empson echoes her in his explanation of why modernist poetry is characterized by ambiguity of the fifth type: “it is in modern poetry, where the range of ideas is great and the difficulty of holding the right ones in the mind becomes acute, that we discover examples of the most advanced numbers of this series.”51 On the other hand, he takes the Riding-focused description of the proper relationship between modernist poem and plain reader and pretends to locate an actual historical model for this relationship in the relationship between Shakespeare and Elizabethan readers. In response to the literary critic’s complaint that Riding’s poetry suffers from “so-called obscurity,” Riding and Graves recommend “increasing the time-length of reading.”52 “The Rugged Black of Anger,” a poem whose “‘obscurity’ … would probably cause it to be put aside by the critic after he had allowed it the customary two-minute reading (for if the poet has obeyed all the rules, this is long enough to give a rough idea of what the poem is all about – and that is all that is generally wanted),”53 is the poem they read slowly, repeating lines, inverting lines, making up transitional lines from the poem’s own words and phrases until, “as a sufficient illustration of the method of letting the poem interpret itself,” “the poem interpreted is practically itself repeated to three times its own length.”54 The time invested in reading is the key to better reading: “The important thing that would be revealed by a wide application of this method to the reading of poems … would be that much of the so-called obscurity of poems was created by the laziness of the plain reader, who wishes to hurry through poetry as quickly as he does through prose, not realizing that he is dealing with a kind of thought which, though it may have the speed of prose to the poet, he must follow with a slowness proportionate to how much he is not a poet.”55 And so “increasing the time-length of reading is one way of getting out of the prose and into the poetic state of mind, of developing a capacity for minuteness, for seeing all there is to see at a given point and for taking it all with one as one goes along.”56
According to Empson, Shakespeare could count on just such slow reading:
One must consider … that the Elizabethans did not mind about spelling and punctuation; that this must have given them an attitude to the printed page entirely different from ours (so that readers must continually have been left to grope for the right word); that from the comparative slowness, of reading as of speaking, that this entailed, he was prepared to assimilate words with a completeness which is now lost; that it is only our snobbish oddity of spelling [that] imposes on us the notion that one mechanical word, to be snapped up by the eye, must have been intended; and that it is Shakespeare’s normal method to use a newish, apparently irrelevant word, which spreads the attention thus attracted over a wide map of the ways in which it may be justified.57
Similarly, by a “form of ambiguity … prominent in early Elizabethan writings,” “Herbert and the devotional poets” require slow reading: they “use a conceit to diffuse the interest back on to a whole body of experience, whose parts are supposed eventually reconcilable with one another; and the reader must pause after each display of wit to allow the various moods in which it could be read, the various situations to which it could refer, to sink into his mind.”58
So Empson cannot deny Riding’s influence by claiming that the only influence that A Survey of Modernist Poetry exerted on him was by means of its initially exciting but apparently wrong-headed attention to ambiguity of syntax in Shakespeare’s sonnet, for the book’s influence on him was not confined to this single discussion. Empson confesses that he wrote Chapter Two of Seven Types of Ambiguity with a certain “excitement” and acknowledges that he was tempted to claim that “this chapter … casts a new light on the very nature of language, and must either be all nonsense or very startling and new.” The principles he discovered in the work of Riding and Graves constituted both a motivating force for beginning his book and an informing presence throughout it, allowing him to convert the critical tradition of editorial either/or into readerly and writerly both / and.59 “Most of what I find to say about Shakespeare has been copied out of the Arden text,” says Empson.60 Be that as it may, it is also the case that very much of what he has to say not just about Shakespeare, but also about the nature of language, the nature of writing, and the nature of reading has been copied out of Riding and Graves.
•
As Haffenden notes, although “the influence of the ‘Graves-Riding’ exegesis of Sonnet 129 is everywhere apparent in Seven Types of Ambiguity,” it is “perhaps nowhere more so than in the discussion of the second type.”61 Haffenden argues that Chapter Two of Seven Types of Ambiguity, about ambiguities of the second type, represents the earliest part of the book to be written:
it seems likely that in the Michaelmas Term of 1928, when Richards told him to go ahead with his task of gathering together his happy “heap” of ambiguities, Empson first looked at the way in which alternative meanings often manage to become reconciled … Not only is Chapter 2 the longest in Ambiguity, it includes an example that is little different from the first piece he ever printed on the subject, “Ambiguity in Shakespeare: Sonnet XVI” (Experiment, February 1929) – which is obviously modelled on the Graves-Riding analysis of Sonnet 129 … Likewise, the second chapter of Ambiguity incorporates another trial piece that came out in 1929, “Some Notes on Mr Eliot,” an analysis of the double meanings created by the confusion of past participles and active verbs in passages from The Waste Land and “Whispers of Immortality.” Since the only other extract Empson printed in advance of the book was a version of the climactic discussion of Herbert’s “The Sacrifice,” it is reasonable to deduce that he started out by looking at the very extremes of ambiguity – that is, type 2, which manifests reconciliation, and type 7, sheer conflict.62
I find Haffenden’s argument here persuasive.
Also suggesting that Chapter Two is the earliest part of the book to have been written is the impression that “the excitement with which it was written” seems less appropriate to the particular type of ambiguity discussed (it is “more common than any of the later types”) than to the sense of discovery with which Empson set out on his project – the sense of discovery, in fact, with which he arrived in Richards’s office.63 His feeling then (recall that when demonstrating to Richards the way Riding and Graves had dealt with the multiple meanings in Sonnet 129, he declared, “You could do that with any poetry, couldn’t you?”) seems to be continuous with the feeling of excitement that he confesses in Chapter Two: he has begun to suspect that the method inspired (as he thought) by Graves “casts a new light on the very nature of language”! Similarly, as Haffenden notes of Empson’s February 1929 essay, “Ambiguity in Shakespeare,” it “opens so briskly that it seems less a poised introduction than an impatient declaration of intent: ‘This is taken out of an essay on the Seven Types of Ambiguity. It is an example of the second type: “two or more meanings which all combine to a single mood and intention of the writer.”’”64 The state of excitement that Empson mentions in Chapter Two and that he shows in his first published work of literary criticism at the beginning of 1929, that is, matches Richards’s description of a student writing obsessively for two weeks after their discussion of the work of Riding and Graves in the fall of 1928 and then returning with “a thick wad of very illegible typescript under his arm – the central 30,000 words or so of the book.”65
More directly focussed on the ambiguity of syntax than any other chapter in Seven Types of Ambiguity, Chapter Two is implicitly the site of the influence by A Survey of Modernist Poetry that Empson concedes in his 1970 letter to Laura Riding. Chapter Two is the one most focussed on Shakespeare, devoting most of the first twelve pages to his sonnets and the last twenty-eight pages to his plays (with twenty-eight pages in between of analysis of type two ambiguities in Chaucer, certain eighteenth-century poets, and T.S. Eliot), deploying across these readings of Shakespeare’s poetry and plays as a whole the particular strategies deployed by Riding and Graves in their treatment of Sonnet 129. We find in Chapter Two, it seems, the first work produced by a student who has convinced himself that Graves was on to something perhaps bigger than he knew ( “Taking the sonnet as a conjuror takes his hat, he produced an endless swarm of lively rabbits from it, and ended by ‘You could do that with any poetry, couldn’t you?’”) and who has sufficiently persuaded his supervisor of this possibility as to have been sent off to test his hypothesis ( “You’d better go off and do it, hadn’t you?”). Empson is excited because he can see that the method he is developing “casts a new light on the very nature of language.”
Empson obviously has Riding and Graves in mind from the beginning of Chapter Two, taking up from where they leave off by extending their reading strategy regarding Sonnet 129 both to other of Shakespeare’s sonnets (more than twelve of them) and to his plays. He explains that in Shakespeare’s sonnets “ambiguity, not of word, but of grammar … is mainly used … to give an interpenetrating and, as it were, fluid unity, in which phrases will go either with the sentence before or after and there is no break in the movement of the thought.”66 In fact, he suggests, “In managing a Sonnet, so as to give it at once variety of argumentation and the close-knit rhythmical unity of a single thought, these devices are more important than they appear.”67
The analysis of Sonnet 129 by Riding and Graves is seldom far from Empson’s thoughts as he writes about the sonnets in general, as one can see by the observation that follows immediately after the sentences above: Empson notes in Sonnet 32 what Riding and Graves note throughout Sonnet 129: “one of those important and frequent subtleties of punctuation, which in general only convey rhythm, but here … amounts to a point of grammar.”68 A few pages later, while carefully unfolding the “variety of meaning … rooted” in the ambiguities of syntax in Sonnet 16, he makes the same complaint against emendating editors as Riding and Graves: “Punctuations designed to simplify the passage all spoil the antithesis.”69
Chapter Two thus shows how Empson went about working out a general methodology from the particular moves that Riding and Graves make in A Survey of Modernist Poetry. For instance, Empson’s observation about Shakespeare’s “use of a particular grammatical form” is inspired by Riding and Graves, for what Empson says about “the way Shakespeare uses a combination of ‘and’ and ‘of’” is suggested by what Riding and Graves reveal about the occlusion of the syntactical ambiguity of “and” by eighteenth-century emendation of the first two lines of Sonnet 129.70 Their complaint that in the second line of the sonnet, the emendator’s placement of “a semicolon after the first action instead of a comma … cuts off the idea at action instead of keeping in action and till action together” is their first example of the general problem they highlight: “The effect of this revised punctuation has been to restrict meanings to special interpretations of special words.”71 Here, that is, the meaning of “and” has been restricted to its meaning “but” or “so that,” whereas in what they call the 1609 version it also has its much vaguer function of simply connecting two clauses (to speak in terms of grammar) or two ideas (to speak in terms of theme) that are to be taken together. Combined with their observation that placement of “a semicolon after the first action instead of a comma gives a longer rest than Shakespeare gave,” this point about the value of “and” in Sonnet 129 looks very much to be the basis of Empson’s point about the value of “and” in Shakespeare’s work generally: “In so far as it is valuable for a poet to include several rhythms, grammatical forms, or shades of meaning in a single phrase, those linguistic forms are likely to be most convenient which insist on no definite form of connection between words and allow you simply to pass on from one to the other. Thus the word ‘and’ will be convenient if you are bringing forward two elements of a situation, conceived as of the same logical type.”72 Just as their treatment of one of Shakespeare’s sonnets allowed Empson to see how they might all be treated, so the ambiguity of “and” that Riding and Graves recover here alerts Empson to the systematic ambiguity of Shakespeare’s “ands” elsewhere.
Riding and Graves even point to the fruitful ambiguity of “of” in the first line, noting “the double meaning of of shame as ‘shameful,’ i.e. ‘deplorable,’ and as ashamed, i.e. ‘self-deploring.’”73 Whereas “the word ‘and’ will be convenient if you are bringing forward two elements of a situation, conceived as of the same logical type,” according to Empson, “the word ‘of’ will be convenient if the two elements are related to the situation differently, and stand in some asymmetrical relation to one another.”74 His example from King Lear ( “The untented woundings of a father’s curse”) demonstrates the same sort of ambiguity in this double genitive: “The wounds may be cause or effect of the curse uttered by a father; independently of this, they may reside in the father or his child. The curse, indeed, might be uttered against the father by the child … All the meanings arrived at by permuting these versions make up one single-minded curse.”75 As with the single-mindedness of the many permutations of Shakespeare’s wounding curses, as spied out by Empson, so with “the double meaning of of shame” and “the double meaning of shame itself as ‘modesty’ and ‘disgrace’” spied out by Riding and Graves: “All these alternate meanings acting on each other, and even other possible interpretations, make as it were a furiously dynamic cross-word puzzle which can be read in many directions at once, none of the senses being incompatible with any others.”76
Of course Riding and Graves demonstrate far more than alertness to “ambiguity of syntax” in their reading of sonnet 129. Having dealt with the changes in meaning produced by changes in punctuation, they take up the interpretation of the sonnet once more, this time concentrating on “a few points … left uncovered in our typographical survey of the poem … principally in the first few lines; for these suffer less from emendations than the rest of the poem.”77 They emphasize the double meanings of words themselves, apart from ambiguities of syntax:
The very delicate interrelation of the words of the first two lines should not be overlooked … the double meaning of waste as “expense” and as “wilderness,” the waste place in which the Spirit is wasted; the double meaning of expense as “pouring out” and as the “price paid”; the double meaning of of shame as “shameful,” i.e. “deplorable,” and as ashamed, i.e. “self-deploring”; the double meaning of shame itself as “modesty” and “disgrace”; again the double meaning of lust in action as “lust” unsuspected by man “in his actions” because disguised as “shame” (in either sense of the word) and condemned by him because he does not recognize it in himself, and as “lust in progress” as opposed to “lust contemplated.”78
So much significance attends these ambiguities that the first line alone hints at much of what is to follow, “the strong parallelism between expense and waste and Spirit and shame expressing in the very first line the terrible quick-change from lust as lust-enjoyed to lust as lust-despised.”79 Riding and Graves, that is, find through their close reading of all the meanings implied by the words of the poem that the ambiguities form a complex whole – the meaningfulness of the ambiguities in the first line suffusing the whole, and the meaningfulness of the ambiguities of the poem as a whole infusing the first line.
Although the complex, systematic whole of multiple meanings that comes of what Riding and Graves describe as “this intensified inbreeding of words,” “all these alternate meanings acting on each other,” “none of the senses being incompatible with any others,” will be examined and explained by Empson as what he calls “the second type of ambiguity” in which “in word or syntax” “two or more meanings all add to the single meaning of the author,” it is not the case that Empson was first alerted to the poetic meaningfulness of words with such double meanings by A Survey of Modernist Poetry.80 I will show in the chapters that follow that Empson owes this awareness to his reading of Graves’s earlier works.
Yet when Empson explains that the multiple “meanings to be extracted” from a poem may emerge from the multiple meanings of constituent words, as in the case of the various meanings of “change” and “earth” in the verse, “Cupid is winged and doth range;/Her country so my love doth change./But change she earth, or change she sky, /Yet I will love her till I die,” he spells things out just as Riding and Graves do in A Survey of Modernist Poetry:
I will love her though she moves from this part of the earth to one out of my reach; I will love her though she goes to live under different skies; I will love her though she moves from this earth and sky to another planet; I will love her though she moves into a social or intellectual sphere where I cannot follow; I will love her though she alters the earth and sky I have got now, though she destroys the bubble of worship in which I am now living by showing herself unworthy to be its object; I will love her though, being yet worthy of it, by going away she changes my earth into desire and unrest, and my heaven into despair … she may change my earth by killing me, but till it comes I will go on loving.81
A perfectly suitable, economical way of meeting the demands of an argument that must show that various meanings of a sentence or phrase are both possible and plausible (paraphrase by direct discourse generally requires far fewer words than paraphrase by indirect discourse), such a style of explication is by no means the only one that Empson might have used, yet it is the one that he uses to the virtual exclusion of any other, and one that Riding and Graves modelled for him.
They deploy this strategy in explaining the effects of ambiguous words on interpretation of line twelve of Sonnet 129 ( “Before a joy proposed behind a dream”). Although they identify “the final meaning of the line” – “Even when consummated, lust still stands before an unconsummated joy, a proposed joy, and proposed not as a joy possible of consummation but one only to be desired through the dream by which lust leads itself on, the dream behind which this proposed joy, this love, seems to lie” – they emphasize that the line “is inlaid with other meanings” that can be extracted from the various meanings of virtually every word in it. Riding and Graves go on to paraphrase successively within quotation marks each possible reading of such a line in just the way that Empson will: “For example the line may also be read: ‘Before a joy (lust) can be proposed, there must be a dream behind, a joy lost by waking’ (‘So that I wake and cry to dream again’); or: ‘Before a joy can be proposed, it must first be renounced as a joy, it must be put behind as a dream; you know in the pursuit that possession is impossible’; or: ‘Before the man in lust is a prospect of joy, yet he knows by experience that this is only a dream’; or: ‘Beforehand he says that he definitely proposed lust to be a joy, afterwards he says that it came as a dream’; or: ‘Before (in face of) a joy proposed only as a consequence of a dream, with the dream pushing him from behind.’”82 According to Riding and Graves, “all these and even more readings of the line are possible and legitimate”; the same point that Empson will make at chapter’s end about the readings of “rooky” as surveyed by the editor of the Arden Shakespeare.83
And just as there are aspects of A Survey of Modernist Poetry other than its analysis of ambiguity of syntax that influence Empson, so its influence extends well beyond Chapter Two. Just as concerning “ambiguity, not of word, but of grammar,” Empson suggests in the second chapter that “where there is a single main meaning” the device of ambiguity “is mainly used, as in the following examples from Shakespeare’s Sonnets, to give an interpenetrating and as it were, fluid unity, in which phrases will go either with the sentence before or after and there is no break in the movement of the thought,” so also he observes the same Janus-like functioning of phrases in Chapter Four concerning ambiguity of the fourth type in Shakespeare’s Sonnet 83: “Line 2 … goes both with line 1 and line 3,” and “the first line may also stand alone, as an introduction … so that line 2 goes with line 3; for this version one would put a comma after therefore.”84
The complaint that Riding and Graves make about emendation of sonnet 129 is that it has destroyed this unity by breaking up the continuity of ideas: “In the second line a semicolon after the first action instead of a comma … cuts off the idea at action … A comma after blouddy separates it from full with which it really forms a single word … Next come several semicolons for commas; these introduce pauses which break up the continuous flow of ideas treading on one another’s heels.”85 Just as their phrase “continuous flow of ideas” anticipates Empson’s “no break in the movement of thought,” so they emphasize – as Empson soon will – the way that, but for emendation, phrases in the sonnet can go either with one idea or another: between lines seven and eight ( “Past reason hated as a swallowed bayt, /On purpose layd to make the taker mad”), “a comma is omitted where Shakespeare actually put one, after bayt. With the comma, On purpose layd – though it refers to bayt – also takes us back to the original idea of lust; without the comma it merely carries out the figure of bayt.”86 Their analysis of other lines highlights again and again the way emendation has robbed the sonnet of similarly meaningful syntactical ambiguities.
•
As is clear from Chapter Two of Seven Types of Ambiguity, Empson found in the readings of Sonnet 129 in A Survey of Modernist Poetry not “nonsense,” but something “very startling and new.”87 Empson was wrong to remove from the second edition of Seven Types of Ambiguity his original acknowledgment of the importance for his own work of this reading of Shakespeare’s sonnet, and it is unfortunate that he left no reference at all to A Survey of Modernist Poetry in his new 1947 Preface. Yet for all the obviousness of the influence of A Survey of Modernist Poetry upon Seven Types of Ambiguity, and for all the controversy as to whose influence is thereby exerted most upon Empson, A Survey of Modernist Poetry is certainly not the book that Empson ought to have acknowledged as either the first, or the main, or the most important source of the method that he developed. And despite his lack of generosity in withholding acknowledgment of the influence of Riding upon him, Empson is not wrong in refusing to award her pride of place.
However much his thesis supervisor I.A. Richards might have determined in his own work to bury the influence of Graves because it represented “current modes of analysis” fraught with “the gravest dangers,” Empson was clearly determined in both the first and second editions of Seven Types of Ambiguity to praise it.88 As we shall see, the person that Empson originally sought to credit for the invention of the method that he developed in Seven Types of Ambiguity not only deserves such credit as Empson offered, but a good deal more. It turns out that Empson’s “early-Graves alibi” is a good one.
1 William Empson, letter to Laura Riding (25 August 1970), Selected Letters of William Empson, ed. John Haffenden (Oxford: Oxford University Press, 2006), 430.
2 William Empson, Seven Types of Ambiguity: A Study of Its Effects in English Verse, 1st ed. (London: Chatto and Windus, 1930; New York: Harcourt, Brace and Company, 1931), vii.
3 Empson, letter to Riding (25 August 1970), Selected Letters, 431.
4 Laura (Riding) Jackson, letter to the editor (14 November 1971), Modern Language Quarterly, 32 (1971): 447–8.
5 I.A. Richards, “William Empson,” Furioso, 1.3 (1940), supplement following p. 44, quoted in John Haffenden, William Empson: Among the Mandarins (Oxford: Oxford University Press, 2005), 207.
6 Robert Graves, Impenetrability; or the Proper Habit of English (London: The Hogarth Press, 1926).
7 Empson, letter to Riding (5 February 1931), quoted in Elizabeth Friedmann, A Mannered Grace: The Life of Laura (Riding) Jackson (New York: Persea Books, 2005), 97.
8 Laura Riding and Robert Graves, A Pamphlet Against Anthologies (1928; repr. New York: AMS Press, 1970), 7.
9 Riding, quoted by Haffenden, Selected Letters, 425.
10 Empson, letter to Riding (25 August 1970), Selected Letters, 430. The length of the enormous letter by Graves grew in Empson’s memory from the nineteen pages he reported to William H. Matchett, the editor of Modern Language Quarterly in 1966, to the forty pages that he reported to Riding in 1971. See Haffenden, Selected Letters, 430n8, and Empson, letter to Riding (29 April 1971), Selected Letters, 433.
11 James Jensen, “The Construction of Seven Types of Ambiguity,” Modern Language Quarterly, 27.3 (1966): 243–59.
12 Robert Graves, letter to the editor, Modern Language Quarterly, 27.3 (1966): 256.
13 William Empson, “Preface to the Second Edition,” Seven Types of Ambiguity, 3rd ed. (1953; repr. London: Hogarth Press, 1984), xiv.
14 Riding, letter to Susan Daniell (12 June 1970), quoted by Haffenden, Selected Letters, 428, and Riding, letter to Susan Daniell (8 August 1970), quoted by Haffenden, Selected Letters, 428.
15 Empson, letter to Riding (25 August 1970), Selected Letters, 429.
16 Ibid., 430.
17 Empson, letter to Riding (29 April 1970), Selected Letters, 433.
18 Empson, letter to Riding (25 August 1970), Selected Letters, 431.
19 Empson, letter to Riding (29 April 1970), Selected Letters, 435.
20 Riding, letter to Empson (13 December 1970), quoted by Haffenden, Selected Letters, 432.
21 Riding, letter to Empson (7 May 1971), quoted by Haffenden, Selected Letters, 436.
22 Empson, letter to Riding (29 April 1971), Selected Letters, 433.
23 Robert Graves, letter to the editor, Modern Language Quarterly, 27.3 (1966): 256.
24 Laura Riding and Robert Graves, A Survey of Modernist Poetry (1927; repr. New York: Haskell House, 1969), 5.
25 Ibid., 63.
26 Ibid., 11.
27 Ibid., 66–7.
28 Ibid., 67.
29 Ibid.
30 Ibid., 73–4.
31 Ibid., 69.
32 Ibid.
33 Ibid., 25, 27.
34 Robert Graves, letter to T.S. Eliot (16 February 1926), In Broken Images: Selected Letters of Robert Graves, ed. Paul O’Prey (London: Hutchinson, 1982), 164. Eliot suggested American poets to be discussed in the book, and strategies for classifying them, in letters of 27 October 1925 and 2 November 1925. See The Letters of T.S. Eliot: Volume 2: 1923–1925, ed. Valerie Eliot and Hugh Haughton (London: Faber and Faber, 2009), 764, 768–9.
35 Graves, letter to Eliot (24 June 1926), In Broken Images, 166.
36 Riding and Graves, A Survey of Modernist Poetry, 74.
37 Empson, Seven Types of Ambiguity, 1st ed., 102.
38 Ibid., 102, 102–3.
39 Ibid., 103.
40 Ibid.
41 Ibid., 103, 103, 104.
42 Ibid., 103.
43 William Empson, Seven Types of Ambiguity, 3rd ed. (1953; repr. London: Hogarth Press, 1984), 202, note.
44 Riding and Graves, A Survey of Modernist Poetry, 74.
45 Empson, Seven Types of Ambiguity, 1st ed., 104.
46 Ibid., 103.
47 Ibid., 104.
48 Riding and Graves, A Survey of Modernist Poetry, 149.
49 Empson, Seven Types of Ambiguity, 1st ed., 114, 115.
50 Graves, letter to Eliot (24 June 1926), In Broken Images, 166.
51 Empson, Seven Types of Ambiguity, 1st ed., 202.
52 Riding and Graves, A Survey of Modernist Poetry, 149.
53 Ibid., 138.
54 Ibid., 148.
55 Ibid., 149.
56 Ibid.
57 Empson, Seven Types of Ambiguity, 1st ed., 105–6.
58 Ibid., 157.
59 Ibid., 102.
60 Ibid.
61 Haffenden, William Empson, 217.
62 Haffenden, William Empson, 217–18. The essays by Empson to which Haffenden refers are the following: “Ambiguity in Shakespeare: Sonnet XVI,” Experiment 2 (February 1929): 33, and “Some Notes on Mr. Eliot,” Experiment 4 (November 1929): 6–8.
63 Empson, Seven Types of Ambiguity, 1st ed., 102, 62.
64 See Haffenden, William Empson, 218.
65 Richards, “William Empson,” in Haffenden, William Empson, 207.
66 Empson, Seven Types of Ambiguity, 1st ed., 65.
67 Ibid., 66.
68 Ibid.
69 Ibid., 71.
70 Ibid., 102, 112.
71 Riding and Graves, A Survey of Modernist Poetry, 67, 74.
72 Empson, Seven Types of Ambiguity, 1st ed., 112.
73 Riding and Graves, A Survey of Modernist Poetry, 79–80.
74 Empson, Seven Types of Ambiguity, 1st ed., 112.
75 Ibid., 113.
76 Riding and Graves, A Survey of Modernist Poetry, 79, 80, 80.
77 Ibid., 79.
78 Ibid., 79–80.
79 Ibid., 79.
80 Ibid., 80; Empson, Seven Types of Ambiguity, 1st ed., 62.
81 Empson, Seven Types of Ambiguity, 1st ed., 63.
82 Riding and Graves, A Survey of Modernist Poetry, 72.
83 Ibid.
84 Empson, Seven Types of Ambiguity, 1st ed., 65, 65, 65, 169–70.
85 Riding and Graves, A Survey of Modernist Poetry, 67.
86 Ibid.
87 Empson, Seven Types of Ambiguity, 1st ed., 102.
88 I.A. Richards, Principles of Literary Criticism (London: Kegan Paul, Trench, Trubner and Co., 1924; New York: Harcourt, Brace and Co., 1925), 29.