Jump to ContentJump to Main Navigation
A History of Ambiguity$

Anthony Ossa-Richardson

Print publication date: 2019

Print ISBN-13: 9780691167954

Published to Princeton Scholarship Online: January 2020

DOI: 10.23943/princeton/9780691167954.001.0001

Show Summary Details
Page of

PRINTED FROM PRINCETON SCHOLARSHIP ONLINE (www.princeton.universitypressscholarship.com). (c) Copyright Princeton University Press, 2022. All Rights Reserved. An individual user may print out a PDF of a single chapter of a monograph in PRSO for personal use. Subscriber: null; date: 28 June 2022

Introduction A Company of Two Armies

Introduction A Company of Two Armies

(p.1) Introduction A Company of Two Armies
A History of Ambiguity

Anthony Ossa-Richardson

Princeton University Press

Abstract and Keywords

This introductory chapter provides an overview of ambiguity. Doubt and plurality, or plenty, are the twin poles of ambiguity. In language, doubt and plenty are intimately joined in the act of interpretation. The perception of plenty in a word, in a line, in a poem, makes one doubt which meaning is the right one; conversely, it is when one doubts the meaning of a text that one might assert the existence of plenty in it. Such a reciprocity is prominent in the book now most closely associated with the topic, William Empson's Seven Types of Ambiguity (1930), with its claim, supported by a litany of ingenious close readings, that ambiguity is intrinsic to poetry and not a fault but a virtue. What Empson meant by ambiguity should not be taken for granted. His infamous definition is ‘any verbal nuance, however slight, which gives room for alternative reactions to the same piece of language’. The chapter then briefly sketches the fortunes of Empson's ambiguity as a critical concept, so as to define and illustrate the broader questions and problems explored over the rest of the book.

Keywords:   ambiguity, doubt, plurality, plenty, language, interpretation, William Empson, poetry, verbal nuance

Eine gute Vorrede muß zugleich die Wurzel und das Quadrat ihres Buchs sein.

— Friedrich Schlegel, Lyceum Fragmente 8, in KSA II, p. 148.

In the beginning was the Word. No, wait—in the beginning was ho logos, the ‘word’, ‘account’, ‘reason’, ‘plan’, ‘discourse’, ‘message’, ‘rational principle’—or something. And ho logos, whatever that was, was with God, or rather, with ho theos, the god. And ho logos was theos—not ho theos, the god, but only theos, god, or a god, a divinity perhaps, or a divine spirit. Was ho theos the same as theos? Does that ho, ‘the’, matter? That is, is the god that the logos was the same as the god that the logos was with? If not, was it inferior? The text does not tell; it only speaks. Was St John writing for the unlearned, who might naturally assume the identity of theos and ho theos, or for the mice-eyed exegetes, who knew that any distinction counted, no matter how small? Was it St John writing at all, or the spirit of God, or a god, through him? And just when was this ‘beginning’ anyway?

We can barely get started in the world without being ambiguous; the six-yarned samite of Creation is shot through with doubt, verbal, substantial. I put it this way so as to present together the two faces of the term ambiguity, which has always denoted the subjective state of doubt as well as its objective correlative in the world, or in a text, a painting, a sonata. Thus Faustus (I.1.80–2): ‘Shall I make spirits fetch me what I please / Resolve me of all ambiguities, / Perform what desperate enterprise I will?’ Those ambiguities threatened damnation, and not only on the stage. The royalist divine Richard Holdsworth, lecturing at Gresham College in the 1630s, warned that religious ambiguity, which took nothing in Scripture or Creation for certain, was the first step towards faithlessness, just as credulity was towards presumptuousness; true faith offered the golden mean between the two.1


More recent theologians, by contrast, have asked their readers to embrace the world’s ambiguity, to come to terms with their own uncertainty. This seems appropriate to our modernity, which has revelled in hesitation as it has unfastened all certainties—in physics, in warfare, in art, in philosophy—at first conceiving new certainties from its own hesitation, and finally disowning even those. But in adopting ambiguity, our theologians have had to redefine it as plurality, that is, the surfeit of human perspectives, or the ‘strange mixture of great good and frightening evil that our history reveals’.2 Over the past decade or two that plurality of perspectives has come to justify widespread political nihilism, total doubt: the truth of nothing and the permission of all, to paraphrase a line made famous by Nietzsche.3 Every action, every decision, every law, every televised utterance has seemed parsable in two ways or more, depending on one’s ideological commitments. Uncertainty appears all-encompassing.

Doubt and plurality, or plenty, are the twin poles of ambiguity as it is studied in this book. Our subject is the ambiguity not of Creation but of language, of texts—the ways it has been posited, denied, conceptualised, and argued over since Aristotle. In language, doubt and plenty are intimately joined in the act of interpretation. The perception of plenty in a word, in a line, in a poem, makes us doubt which meaning is the right one; conversely, it is when we doubt the meaning of a text that we might assert the existence of plenty in it, and not simply in us.4 Such a reciprocity is prominent in the book now most closely associated with the topic, William Empson’s Seven Types of Ambiguity (1930), with its claim, supported by a litany of ingenious close readings, that ambiguity is intrinsic to poetry and not a fault but a virtue. The book is an extraordinary achievement, wise

and learned, full of a wit recognised even by its detractors, and blessed with compelling powers of observation: under Empson’s microscope, poems come to look just as rough and complex as the seeds and needles in the images of Robert Hooke. The critic’s business is analytical: he is like the dog who relieves himself against the ‘flower of beauty’ and then scratches it up afterwards (STA, p. 9). But his manner is unlike the quasi-scientific mode promoted by his mentor I. A. Richards, and he insists that poetry be treated with sympathy, not merely as an ‘external object for examination’ (248).5 The book’s method, despite its title, turns out to be tactical rather than strategic, arriving at insight not by systematic theorisation but haphazard, as if on the way to something else, in the course of a chat over sherry in the combination room.6

The seven types are ‘intended as advancing stages of logical disorder’ (48), but they keep bleeding into one another. In the first, most general type, ‘a word or a grammatical structure is effective in several ways at once’ (2)—Empson’s first example, ‘Bare ruined choirs, where late the sweet birds sang’, soon attracted astonishment for the number of associations he was able to draw out between trees and ruined monastery choirs. In the ‘most ambiguous’ seventh type the duality of meaning in a text shows ‘a fundamental division in the writer’s mind’ (192), and the book culminates in a reading of George Herbert’s poem ‘The Sacrifice’ as the charged expression of a Christian ambivalence. But between these two extremes lies a wealth of glittering detail. For a flavour of Empson’s typical approach, consider a stanza from the Andrew Marvell lyric ‘Eyes and Tears’:

  • Wha rs,
  • Yea, even Laughter, turns to tears;
  • And all the jewels which we prize
  • Melt in these pendants of the Eyes.

Empson comments:

Melt in may mean ‘become of no account beside tears’, or ‘are made of no account by tears,’ or ‘dissolve so that they themselves become teares,’ or ‘are dissolved by tears so that the value which was before genuinely their own has now been assumed by and resides in tears.’ Tears from this become valuable in two ways, as containing (p.4) the value of the jewels (as belonging to the world of Cleopatra and hectic luxury) and as being one of those regal solvents that are competent to melt jewels (as belonging to the world of alchemists and magical power). (172)

Here doubt (‘may mean . . . or . . .’) insensibly becomes plenty (‘valuable in two ways, as . . . and as . . .’), in such a way that it is hard to know where one stops and the other begins. But the reader, whether or not she accepts the argument, is likely to come away from it thinking only of Marvell’s fulness, having forgotten Empson’s uncertainty. Empson helps her along in this regard, having already asserted that ‘I have myself usually said “either . . . or” when meaning “both . . . and” ’ (81). He confesses that the ambiguities he finds in Shakespeare are mostly copied out of Arden editions, where, in the manner of traditional philology, possible readings and interpretations are considered and dismissed, or else listed as alternatives. But, suggests Empson mischievously, ‘the nineteenth-century editor secretly believed in a great many of his alternatives at once’ (82). How could one see all those wonderful meanings and not think they had occurred to Shakespeare? Better to see the Bard’s ‘original meaning’ as ‘of a complexity to which we must work our way back’. A writer’s intention was of great interest to Empson, unlike many of his successors; in the preface to the second edition he warns that ‘[i]f critics are not to put up some pretence of understanding the feelings of the author in hand they must condemn themselves to contempt’ (xiii–xiv). And so with ‘Eyes and Tears’ above, his discussion concludes with the insistence, forestalling objections, that he has not been making up his own poem but only ‘quoting’ Marvell, on the basis that the poet assumed in his readers a wide acquaintance with ‘conceits about tears’.7

What Empson meant by ambiguity should not be taken for granted. His infamous definition is ‘any verbal nuance, however slight, which gives room for alternative reactions to the same piece of language’ (1).8 But this is not really a definition, as he clarifies in a footnote: it is ‘not meant to be decisive’, and ‘the question of what would be the best definition of “ambiguity” . . . crops up all through the book’.9 A few pages later he specifies both the subjective and the objective, doubt and plenty:

‘Ambiguity’ itself can mean an indecision as to what you mean, an intention to mean several things, a probability that one or other or (p.5) both of two things has been meant, and the fact that a statement has several meanings. (5–6)

Later critics have deplored the imprecision of Empson’s terminology, and particularly his failure to distinguish ambiguity from mere multiple meaning—one recent primer dismisses Seven Types as ‘a very confused book’10—but the foundation of his argument is his own experience of poetic language rather than any desire to clarify and classify concepts; ambiguity denotes, as we have seen above, textual items that have made Empson hesitate.11 Expressions of doubt appear throughout: ‘Not a clear example, and I am not sure that what I said is true’ (5n), ‘I am not sure how far people would be willing to accept this double meaning’ (229). This is something like a negative capability, a ‘being in uncertainties, Mysteries, doubts’, although paradoxically it is also an ‘irritable reaching after fact and reason’.12 ‘Ambiguity’, then, is precisely the correct word. Our estimation of the book’s value must turn not on its theoretical rigour but on Empson’s capacities as a reader, and here we are repaid by his seismographical sensitivity to words, to culture, to society. If it is a confused book, he might rejoin that it reflects a confused subject, and that, like Socrates, he has preferred aporia to false certainties.

Seven Types has long been canonised as a watershed in the history of thinking about ambiguity, starting with Richards’s remark in 1936: ‘Where the old Rhetoric treated ambiguity as a fault in language, and hoped to confine or eliminate it, the new Rhetoric sees it as an inevitable consequence of the powers of language and as the indispensable means of most of our most important utterances—especially in Poetry and Religion.’13 Empson’s criticism more generally has enjoyed a revival of interest in recent decades.14 But beyond pointing to two of the book’s sources—Richards (p.6) and Robert Graves—few have seriously considered what preceded it, and so the nature of its achievement remains unclear. One major aim of this History is to remedy that omission, although I ought to state explicitly that the narratives traced below are not mere back-story, and only in the final chapter do I specifically address Empson’s immediate intellectual genealogies. It is true, as Richards said, that the ‘old Rhetoric’, from Aristotle onwards, treated ambiguity as a fault, but even so, readers had praised the ambiguities of poetry for centuries before Seven Types, and this is to say nothing of the tangled histories of ambiguity in law, biblical exegesis, and philosophical hermeneutics. Each narrative will help qualify our assumptions about the profile of modernity with regards to ambiguity: again and again we find familiar questions raised in past and alien settings. Perhaps most of all, the interpretation of Scripture will acquire a special resonance with Empson’s project. It was not for nothing that Richards paired ‘Poetry and Religion’ in his line on the new rhetoric, and as Empson himself would insist in a later preface, ‘Critics have long been allowed to say that a poem may be something inspired which meant more than the poet knew’ (STA xiv, my emphasis). As we shall see, the idea of divine inspiration, which undergirded so much analysis of multiple meanings in the Bible, was a key precursor to Empson’s argument.

A more detailed synopsis of the book’s chapters will be found at the end of this Introduction. Before that I want briefly to sketch the fortunes of Empson’s ambiguity as a critical concept, so as to define and illustrate the broader questions and problems explored over the rest of the monograph.

Minimisers and Maximisers

(p.7) It will be convenient to begin with Empson’s early readers. A few admired the book for its charm and talent at practical criticism; F. R. Leavis, for one, called its author ‘a mind that is fully alive in this age’, marvelling at his ‘sensitivity’, ‘erudition’, and ‘mastery’ of resources:15 Others expressed surprise at the idea that ambiguity might be a good thing: ‘I have regarded ambiguity in literature, and even in daily speech, as a sin or at least a great mischief. I had yet to realize that it may be almost the foundation of poetic art.’16 Probably the majority were disapproving, or downright hostile. A common complaint was that Empson had taken his passages out of context, and that his readings therefore owed more to his own ingenuity than to the actual texts themselves, as if he had lost sight of ‘the object as in itself it really is’ and thereby abandoned the office of criticism.17 John Middleton Murry concluded that ‘poetry has no particular importance to him save as an opportunity for a free exercise of his abilities’.18 The ageing gourmand Thomas Earle Welby, meanwhile, offered a caustic reductio of Empson’s approach, deploying Swinburne, one of his own specialisms:

‘Swallow, my sister, O sister swallow’ might be paraphrased, “Be a good little girl, and take your medicine.” But it could be so paraphrased only by a reader who was, to be scientific and tabular, (1) ignorant of one of the most familiar as well as poetic of ancient legends, (2) not only unacquainted with the temper of the author from previous reading of him, but utterly incapable of divining it as patent in the very first lines of the poem; (3) deaf and blind without the extenuating circumstance of being dumb.19

This criticism goes right to the heart of Empson’s technique, and represents a central concern of the present book. The assumption is that words or lines are ambiguous only when taken out of historical and textual context, and deprived of their author’s intention; Empson, for Welby, has (p.8) pressed phrases in isolation and so permitted them to mean whatever they can mean.20 The emphasis on context to resolve apparent ambiguity is almost universal in works of classical hermeneutics, for instance in theology and law.21 Unsurprisingly, then, zealous readers both before and after Empson have been accused of inattention to context, a charge encapsulated in a Russian word invoked by Vladimir Nabokov to disparage faithless translations: otsebyatina, ‘what one contributes oneself’, the meaning improperly loaded onto a text.22 Even those in favour of Seven Types saw this threat; one journalist, who called the book ‘brilliant’, also found himself ‘faced with a bewildering variety of meanings’, and noted the apparent irrelevance of the ‘intention or conscious meaning of the author’.23 That reaction is understandable but inaccurate, on both intention and context. Empson did not ignore the former, and saw clearly the dangers of formalistic exegesis, acknowledging that as we acquire a language, or come to understand a particular writer, we ‘learn to cut out the alternative meanings which are logically possible’.24 He even accepted afterwards that in the wrong hands his own method could produce a ‘shocking amount of nonsense’ (xii). But intention is not always straightforward, and in this book we will see Seven Types as the heir to a range of concepts that serve to challenge the primacy and the disambiguating simplicity of intention— above all, dramatic irony and the unconscious.

Context, too, is slippery, because although in ordinary discourse it narrows down interpretive possibilities, there are certain special linguistic situations in which it actually promotes ambiguity, such as witticisms and literature, as I. A. Richards and, more recently, Erving Goffman have observed. (p.9) 25 As I suggest below, this might be seen as an implicit argument of Seven Types in relation to contemporary semantics.26 Context can lure us away from the usual meaning of a word in such a way as to bring both meanings into focus, or it can simply be arranged in such a way as to leave both meanings possible. Empson appreciated this fully. In fact, two of his types, the third (STA 103) and the seventh (192), are defined in terms of context, and it appears in his readings throughout the book (e.g. 55, 167, 203). The same problem had been confronted by earlier readers who posited deliberate ambiguity in poetry and in Scripture, though they did not explore the methodological question in depth.27

Empson’s taste for ambiguity suited an era of literature characterised by a shift away from realism towards other language games—distortion, obscurity, the stream of consciousness, the oblique, the difficult, the ambiguous.28 This was noted by one of his earliest acolytes, the Dublin classicist William Bedell Stanford, who, at the end of his 1939 monograph on ambiguity in ancient Greek literature, pointed to Joyce’s Work in Progress, shortly to be published in full and final form as Finnegans Wake, a book notorious for its polyglot puns and density of allusion, ambiguous in much the same way the sea is wet.29 Roland McHugh, whose volume of Annotations is known to all Wake enthusiasts, once recollected his experiences with a reading group of Joyceans including James Atherton, Clive Hart, and Fritz Senn, held in 1970–1971 at Amsterdam and Brighton. One member of the group, whose name is now less familiar in the field, was Matthew Hodgart, a historian of satire who had, among other things, composed a fifth book of Gulliver’s Travels as an allegory for the ’68 student protests he witnessed at Columbia and Berkeley. McHugh recalls:

At an early stage Matthew Hodgart underlined a distinction: the maximizers, such as himself, were delighted at every additional level that could be envisaged (‘Yes, great. Hamlet, sure. Also King David’s sling.’). On the other hand minimizers such as myself tried to cut the (p.10) allusions to the smallest number which would account for all the letters in the word.30

A good Wake reading group brings to every word the struggle between the free play of the imagination and the sober attention to textual genetics, in search of Joyce’s intentions; in this respect it offers a microcosm of the controversies of literary criticism since the 1920s. But in this book we will see the same conflicts played out among traditions of readers going back to the Renaissance—between Protestants and Catholics on the Bible, between anciens and modernes on Homer, between Alexander Pope and his dunces, among readers of Horace in the eighteenth century and of Sophocles in the nineteenth.

But the greatest maximising statements belong to the twentieth century. In the late 1930s, Seven Types came to the attention of the American critic Cleanth Brooks and his colleagues, who had been weaned on the work of Richards since 1929, and it was an immediate revelation. When the poet John Crowe Ransom came to define that group as the New Critics in his 1941 book of that name, he included a chapter on Richards and Empson, describing Seven Types as ‘the most imaginative account of readings ever printed’, while hesitant about some of its wilder reaches.31 Over the next ten years, Americans responded to Empson’s book in two ways: they critiqued and clarified its terminology, and they domesticated its technique, draining it of doubt and idiosyncrasy to make it more generally applicable in the classroom.32 Both Brooks and Philip Wheelwright complained that the word ambiguity was inappropriate because it denoted an either / or, rather than, as was wanted, a both / and.33

In 1948, the philosopher Abraham Kaplan and the psychoanalyst Ernst Kris collaborated on an article aiming to refine the terminology, distinguishing five types of ambiguity: (1) the disjunctive, in which we are forced to make a choice between incompatible alternatives; (2) the additive, whose various meanings closely overlap with one another, such as ‘rich’ to mean wealthy and abundant; (3) the conjunctive, whose meanings are distinct but both relevant, as in a pun; (4) the integrative, whose meanings relate to and connect with one another in complex ways; and (5) the projective, (p.11) in which each reader must bring his own meaning to the phrase or text—this last was less an ambiguity than a vagueness or indeterminacy. Empson, they concluded, had been interested chiefly in (4), which was the true domain of poetry, and to a lesser extent in (3).34 Meanwhile, the New Critics emphasised that their concern was the plenty in Empsonian ambiguity, not the doubt. Brooks and Robert Penn Warren had ignored the topic in their seminal 1938 textbook and anthology, Understanding Poetry, but when they came to revise it in 1950, they added an entire chapter on it, containing the definition: ‘ “ambiguity” is seen to be depth and richness’.35 Wheelwright’s preferred term in 1940 had been plurisign, capturing multiplicity (pluri-) without uncertainty.36 The strongest assertion was that of Monroe Beardsley in 1958, elevating plenty into a ‘Principle of Plenitude’: ‘All the connotations that can be found to fit are to be attributed to the poem. It means all it can mean, so to speak. . . . The very notion of critical explication seems to involve getting as much meaning out of the poem as possible, subject only to some broad control that will preserve a distinction between “getting out of” and “reading into”.’37 The dynamic modality in that ‘can mean’ and ‘as possible’ is, as we shall see, a grammatical leitmotiv of the maximiser.38

A different kind of maximising is that of the French tradition. The philosopher Richard Gaskin has recently called Empson’s criticism ‘the fons et origo of the doctrine of deconstruction’.39 This is a seriously misleading judgement, for the Francophone avant garde took little notice of Empson.40 (p.12) The exception was the Belgian critic Paul de Man, who in 1956, at work on his doctoral thesis at Harvard, published an article in French against the ‘dead end’ of New Critical formalism, which had aspired to science but finished up as a lifeless traffic with mere words, with paradoxes and ambiguities that could always be neatly tied up by the critic’s business.41 From this he sought to rescue Empson, who had pointed to something deeper—to real doubt.42 Whereas Empson’s second to sixth types dealt only with ‘controlled pseudo-ambiguity’ like that of the New Critics, the first and the last were different. The first revealed that ‘[a]ny poetic sentence, even one devoid of artifice or baroque subtlety, must, by virtue of being poetic, constitute an infinite plurality of significations all melded into a single linguistic unit’.43 The seventh, meanwhile, showed that ‘true poetic ambiguity proceeds from the deep division of Being itself, and poetry does no more than state and repeat this division’.44 De Man’s language, with its smack of infinity and its capitalised ‘Being’, reeling from Sartre and Heidegger, gestures towards the future rhetoric of deconstruction, miles from Empson’s tone of good common sense. It shares something, then, with the excitable, pseudo-theological manner of the early Barthes, who had pronounced three years before that in modern poetry, unlike in classical poetry, the word (or Word)

shines with an infinite liberty and is ready to radiate towards a thousand uncertain and possible connections. . . . Beneath each Word in modern poetry lies a sort of existential geology, in which is gathered the total content of the Name, and no longer the selected content, as in prose and classical poetry. . . . The Word is here encyclopaedic, (p.13) containing simultaneously all the senses from among which a relational discourse would have made it choose . . . a Pandora’s box from which fly out all the potentialities of language.45

All this went beyond Beardsley—an infinite plenitude, with no controlling distinction between ‘getting out of’ and ‘reading into’. Maximising had different modes, then, and different means of expression, but by the 1960s it had become dominant everywhere. The Anglo-American critics who absorbed deconstruction in the 1970s and 1980s—J. Hillis Miller, Geoffrey Hartman, Christopher Norris, among others—had been brought up on New Critical ambiguity, and were liable to reinterpret Empson through the smeared lens of the new French theory, as already proposed by Paul de Man.46

At the same time, Empson’s work came under fire from two brilliant young women. One was the poet and theorist Veronica Forrest-Thomson, who, in her monograph Poetic Artifice—published in 1978, three years after her tragic early death—rejected what she saw as the excessive rationalism of his approach, manifested in his focus on ambiguity in poetry to the exclusion of its extrasemantic effects.47 The other critique, more relevant here, was the diametric opposite, since it came from a structuralist who found Empson’s concept too vague to be scientific; this was Frank Kermode’s doctoral student Shlomith Rimmon, who completed her thesis on ambiguity in Henry James in 1974, publishing it as a monograph three (p.14) years later.48 The question of ambiguity in James had by this stage a long pedigree, entirely independent of Empson. A 1934 article by Edmund Wilson argued that ambiguity—at the narrative rather than the verbal level— was a recurrent feature of the novelist’s stories; thus ‘The Turn of the Screw’, to take only the most obvious example, was predicated on an ambiguity as to whether the ghosts are real or not. The theme was picked up again in the 1960s, beginning with an article by Roger Gard (who would later examine Rimmon’s thesis), and continuing in a series of monographs on James with the word Ambiguity in the title.49 Rimmon took a theoretical approach, devoting her first chapter to the nature of ambiguity, which she treats as an objective, scientifically analysable property of a text, and which she identifies as strictly disjunctive between mutually exclusive options, that is, as Kris and Kaplan’s first type, notated in symbolic logic as AΛ‎B (‘A or B but not both’). It has a close visual analogue, as she notes, in the ‘duck-rabbit’ figure made famous by Wittgenstein. Rimmon carefully distinguishes it from, on the one hand, ‘the multiplicity of subjective interpretations’, that is, ‘the ultimate subjectivity or the “unescapable ambiguity” of all art’, and, on the other, Empson’s ‘use of “ambiguity” as a blanket term for all kinds of secondary meanings’.50 In other words, she seeks to strip ambiguity of plenty and reattach it to doubt, expressed here in terms of choice:

[A]n ambiguous expression calls for choice between its alternative meanings, but at the same time provides no ground for making the choice. The mutually exclusive meanings therefore coexist in spite of the either/or conflict between them. (17)

This is a minimiser’s ambiguity, severely attenuating even the doubt: as with Brooks, there is no doubt about doubt, and the uncertainty itself has become rigorously confined and articulated. Rimmon also insists on the principle of context gestured at by Empson’s first critics; only now she can articulate the point more formally by drawing on the concept of isotopy or semantic redundancy, introduced a decade before by the Lithuanian (p.15) semiotician A. J. Greimas.51 Her example is scriptural, from Deut. 16:20, צ דֶֶק צ דֶֶק תּרְִדּ ֹ ף‎ zedek, zedek tirdof, which might in isolation mean either ‘Justice, justice, shall you pursue’ or ‘Justice, justice, shall you drive away’. It is easy for her—too easy—to argue that our understanding of biblical ethics, as well as of the immediate scriptural context, points us to the correct meaning. For other readers, the lexical ambiguities of the Hebrew Bible had led to quite the reverse conclusions.52

It is worth briefly considering here the shift in academic climate between Empson and Rimmon. He had been the product of élitist centres of education where the reading and writing of poetry was part of the culture, expected even in a mathematics student like Empson. Poetry was still central to the human experience, and the purpose of writing about it was to get at that experience: what mattered was not theoretical rigour but wit, verve, sensitivity, unpredictable insight. Rimmon’s framework was much closer to that of Richards, who, unlike Empson, sought to offer something replicable: method and checkable results.53 Her world, by her own admission at odds with Kermode, was the professionalising, increasingly globalised and inclusive humanities of the 1970s, keen to sacrifice idiosyncrasy for a common, scientific standard of communicable research and theory. Such an ideal glows off the page of her natural academic home, Poetics and Theory of Literature (PTL), the flagship journal of the Porter Institute for Poetics and Semiotics at Tel Aviv University, one of many Israeli institutions funded by the Tesco heiress Shirley Porter, then still in the salad days of her controversial career in politics.54 The journal, edited by Benjamin Harshav (then Hrushovski), was full of work like Rimmon’s, as young scholars from around the world, inspired above all by Barthes and Jakobson, offered technical accounts of literary language and narratology, crisp with diagrams and quasi-mathematical terminology. The first volume contained ‘The Squirm of the True’, a two-part article on ambiguity in James by Christine Brooke-Rose. In 1979 PTL was relaunched as Poetics Today, and its first issue carried Brooke-Rose’s review of Rimmon’s book, cheerfully admitting its superiority to her own work.

(p.16) But it was here, at home, and from a friend, that Rimmon’s neutral, logical ‘ambiguity’ encountered resistance. The third issue (1980) of Poetics Today bore an article on James’s story ‘The Figure in the Carpet’ by Hillis Miller, then at work on his monograph Fiction and Repetition (1982). Miller had read the manuscript of Rimmon’s book for Chicago University Press, and ‘disagreed with almost everything’ but nonetheless recommended publication; the mood between them was warm, and they traded drafts and offprints of their work.55 In his 1980 article he made his reservations public. Later that year Rimmon—now Rimmon-Kenan—was given right of reply, and Miller briefly in return. Their courtly exchange reads as a sort of seduction of her ambiguity by a darker critical force. If Empson was the dog scratching up the flower of beauty, Miller saw himself as one of the little foxes that spoil the structuralist vines. In his view, her concept

is too rational, too ‘canny,’ too much an attempt to reduce the mise-en-abyme of any literary work, for example the novels and stories of James, to a logical scheme. The multiple ambiguous readings of James’s fictions are not merely alternative possibilities. They are intertwined with one another in a system of unreadability, each possibility generating the others in an unstilled oscillation. Rimmon’s concept of ambiguity, in spite of its linguistic sophistication, is a misleadingly logical schematization of the alogical in literature, that uncanny blind alley of unreadability encountered ultimately in the interpretation of any work.56

Here ambiguity, shorn of the doubt that defined its usage in Seven Types, has come to seem a lifeless tool. Miller’s preferred term, unreadability, sought to restore that doubt—it ‘names the presence in a text of two or more incompatible or contradictory meanings which imply one another or are intertwined with one another, but which may by no means be felt or named a unified totality . . . [it] names the discomfort of this perpetual lack of closure’ (my emphasis). It thus belongs to a group of terms like ‘indeterminacy’ which, in the critical discourse of the period, indicate that ambiguities will not be resolved by the act of interpretation, instead remaining open to unsettle the reader even after being identified.57 In practical terms this was remarkably similar to what Rimmon had said of her ambiguity, that it ‘calls for choice between its alternative meanings, but at (p.17) the same time provides no ground for making the choice’, only the emphasis was now on experience, on ‘discomfort’. Just like Rimmon, Miller distinguished his idea from both ‘ambiguity in literature as plurisignificance or richness of meaning’ and ‘the perspectivism which holds that each reader bring something different to a text’. The chief difference is that for Miller all works are, in his sense of the word, unreadable.

Rimmon-Kenan noticed the similarities between the two concepts, as between their respective accounts of James. Her tongue-in-cheek response, ‘Deconstructive Reflections on Deconstruction’, pointed out that unreadability needed ambiguity as much as ambiguity needed unreadability, for each led to the other: Miller, by implying that his account of James was better than hers, had offered only ‘another form of decidable closure’.58 When Miller came to write his own rejoinder, he identified Rimmon-Kenan’s co-opting of his deconstruction as typical of her ‘schematizing rationality devoted to intellectual mastery’; it was her scientific pretensions, and those of the Porter Institute more broadly, that he distrusted. By contrast, he himself had attempted only to express ‘the failure of an attempt at mastery’.59 He confessed his own failure, and yet that was the point: ‘Had I succeeded I would have failed.’ Miller’s tone has the arch self-consciousness we associate with postmodern irony: Rimmon-Kenan in her response was ‘doing it again’, just as he imagines she will conclude of his reply that ‘He’s doing it again.’60 But, as in any good comedy, his conclusion is conjugal. Miller may have been a guest in her house, Poetics Today, but now he will ‘welcome her as a guest’ in his mother’s house—the home of deconstruction—which would instruct her, or she him. It now seemed they had been doing the same thing all along, and doing it together: ‘It is now no longer, “She is doing it again,” but that she starts doing, to some degree at least, what I was doing: “She [He] is [are] doing it again.” ’61

This was all a bit silly: two friends teasing each other about their critical arsenals. If we can get anything useful out of the exchange, it is that criticism in 1980 had reached an impasse on how far doubt was to be legislated: for the Rimmons, it was the critic’s job to be certain, even when she was specifying the doubtfulness of a text, whereas for the Millers, certainty was too much to ask of the critic, whose job was not to clarify but to act out the doubt. This was the continuation of a dialogue about ambiguity given fresh impetus by Empson, but long predating him.

(p.18) Both sides, however, had lost sight of something central to Seven Types— the poet’s intention. They had become formalists. This was no surprise, in the wake of the anti-intentionalist manifestos of the 1950s (Wimsatt and Beardsley) and ’60s (Barthes); to legitimise the intention, it was thought, was to limit the meaning of a text to a unity achieved at the moment of production. But Empson had showed, by contrast, that intentionality was rather more complex than this suggested: it could produce ambiguity because it was divided and ambivalent. Nor was he the first to do so: several other discourses, as we shall see in this book, had discovered ambiguity in language by reconstructing its author’s intention as in some way fractured or multiple. In some instances the meaning was even allowed to change over time, or to be more fully expressed, such as the law that must continually be applied to new cases, or the Hebrew prophet whose words could finally be explained only under the dispensation of Christ. By returning to these models we will see how the supposition of ambiguity in a text, and the acceptance of uncertainty about it, could anticipate the charge of formalism—of otsebyatina.

Ratio Studiorum

Before I explain in more detail what this book is, I ought to say what it is not. First, it is not a history of the word ambiguity or its cognates.62 Nor is it concerned, at least until the final chapter, with non-verbal or moral ambiguity, let alone the sort enshrined in book titles of the twentieth century, from Simone de Beauvoir onwards.63 Third, and perhaps most surprisingly, it is not a history of ambiguities, of ambiguous texts, and certainly not of canonically ambiguous texts like The Praise of Folly or Sartor Resartus. It is tempting to pit literature against literary criticism; for instance, Jacques Dürrenmatt’s 2001 study Le vertige du vague presents French (p.19) Romantic literature, such as the novels of Stendhal and Nodier, as a rebuttal to the hostility to ambiguity prevailing in neoclassical criticism and rhetoric. But if we want to write a coherent history we ought to resist that temptation, partly because such work tends to rely on an imprecise category of ambiguity, lumping it in with obscurity, irony, wit, and so on, but more pressingly because it compares past critical statements with present-day aesthetic judgements of past works, which threatens a serious anachronism. Likewise, a history of ambiguous works would run the risk of being merely a history of works in which we now perceive ambiguity, when that perception is precisely what we want to historicise. For the intellectual historian, ambiguity, as James Porter has said of ‘being classical’, is most usefully understood not as ‘a property an object can have’ (or is now judged to have had), but as ‘the suggestion that a given object has this kind of property, which is why one needs to determine just where in any given case the suggestion originates’.64

For that reason, the basic unit of this history is not the ambiguous phrase but the act of seeing ambiguity in a phrase. In other words, it is a history first and foremost of readers, most of whom are, and have always been, extremely obscure—anonymous ancient scholiasts, late mediaeval scholastics and glossators, Indian poeticians, Counter-Reformation exegetes, editors, canonists, casuists, immortels, Grub Street hacks, schoolmasters, paraphrasts, cloudy German philosophers, Victorian barristers, dantisti, doctoral students and their professors, clerics, psychoanalysts, music journalists.

But what to do with them all? We are faced with a problem of method. Intellectual historians have usually been tasked with capturing the ideas of past writers, such as scientists or political philosophers, and their method for the past fifty-odd years has centred on an attempt to reconstruct the social, political, biographical, and especially the discursive contexts of the works in question.65 The same basic frame has served for reception studies that analyse the reinterpretations of older works in and for new contexts.66 As with the latter studies, we are here concerned with readers more than writers—though of course the readers must also be writers in order to transmit their ideas—but unlike those studies we are interested in reading mostly at the level of words, phrases, and sentences. And at this level we (p.20) find an ongoing tension between, on the one hand, discursive contexts, in the form of hermeneutic rules and shared semantic, aesthetic, or theological precepts, and on the other, the brute act by which meaning, and occasionally multiple meaning, is discovered in another person’s words. The latter act is often impossible to understand by a process of historical contextualisation; it operates at the limits of reason, and in extreme cases its results appear unsayable, even unspeakable. Its eroticism should not be denied: texts seduce us to vulgar interpretive gestures, making us foolish in the eyes of others, who look on with puritan disapproval. This is surely why naughty Empson, banished from Cambridge for a condom, continues even now to receive the headmaster’s slipper for his critical transgressions.67

In this history I have privileged the individual specificity of those acts. But practice must be set off against theory, and so I have explored, alongside those acts, the aforementioned rules and shared precepts, as well as the concepts devised to make sense of ambiguity, such as the classical virtue of elegantia, the Catholic notion of the multiple literal sense of Scripture, and the post-Romantic idea of dramatic irony. Nonetheless, since readers sometimes struggled to articulate the dissonance between their expectation of one meaning and their discovery of several, we need to recognise the limits of our reliance on their conceptual terms or ‘actors’ categories’.68 That is, it must be allowed that we can, now and then, understand past thinkers better than they did themselves, a principle of interpretation formulated by Kant and later insisted upon by Empson.69 Nor will it be improper to observe, where relevant, that readers in diverse eras and disciplines have faced similar problems and sometimes arrived at similar solutions—that certain patterns transcend individual contexts. The point is likely to appeal to those historians who resent the tyranny of contextualism.70 But the narrative of this book has another relevance to their concerns, because it offers, among other things, a genealogy of those (p.21) concerns: to call a text ambiguous is another way of saying that it has been variously interpreted, and to say that its ambiguity is deliberate is therefore to suggest that more than one of those interpretations might be valid. As we shall see, this strategy is not limited to saving an eirenic attitude to Scripture.

A central contention of this book is that Richards’s temporal distinction between Old and New Rhetoric should be replaced by a generic one. On one side, theorists of language in all periods, after Empson as before—writers on grammar, rhetoric, semantics, poetics, general hermeneutics—have understood ambiguity as a pernicious fault. On the other, isolated traditions have acknowledged the deliberate and beautiful ambiguity of certain privileged, exceptional texts, which before Empson fell largely into one of two groups: (a) classical poetry, above all Roman satire and Greek tragedy, and (b) the Hebrew Bible. The strategies of explaining multiple meanings in these two categories differed. Those in classical poetry were evaluated with a concept I label artificial ambiguity, emphasising a speaker’s mastery of words as a means to manage and control other persons, whether benignly as social wit or malignly in acts of deceit. Those in Scripture, by contrast, were justified by later Christian scholars as what I call inspired ambiguity, relying on the figure of the prophet exalted by God to express divine and mystical truths. Where the first reinforces the classical model of the unified subject who deploys language to express his will and exercise agency in the world, the second serves to undermine that model by positing a divided subject whose language is his own and not his own, simultaneously the product of two distinct impulses. These two terms are not actors’ categories, and cannot be found in historical sources; rather, they represent an effort to get a handle on two ways of thinking about multiplicity that can, I think, be abducted from those sources. I have used the word ambiguity here, when some the figures I am discussing, such as St Augustine, would have thought rather in terms of multiple meanings. But the two ideas are only phases of the same moon, and where one reader asserted plurality in a text, another could always denounce it as ambiguity. Empson’s term captures the threat of doubt inherent in all plenty.

It should go without saying that the history of ambiguity is not only complex but extremely non-linear, and therefore that there is no straightforward way to tell it within the confines of a linear prose narrative. Some may be surprised by the method with which I approach their disciplines, such as the history of law, or of literary criticism, which have their own etiquettes, rhythms, styles of citation, and so on. But it has been my assumption that a worm’s-eye intellectual history of the obscure as well as (p.22) the famous, the scholarly communities as well as the lone geniuses, will find the common ground and narrative interconnections between apparently disparate fields and eras. I have thus tried to strike a middle path between the remorseless thick description characteristic of much history of scholarship, and what Germans call the Gipfelwanderung, ‘wandering from peak to peak’ (for instance, from Descartes to Locke to Hume to Kant) of most long-range surveys. Instead I have pursued rivers as they roll down from peaks—from Aristotle, from Justinian, from Augustine, from Eustathius, from Bacon, from Schlegel, from Freud—into trackless valleys, into other rivers, underground. Chronological coherence has been preserved within each narrative, at the expense of a tidy sweep forward overall.

The first half of the book offers a series of disciplinary parameters for thinking about ambiguity, from rhetoric and legal hermeneutics to biblical exegesis and early modern literary criticism. Chapter One is devoted to what Richards called the Old Rhetoric, sketching the long persistence in the West, from Aristotle to the early twentieth century, of a ‘single meaning model’ of language, one that takes ambiguity for granted as an obstacle to persuasive speech and clear philosophical analysis. Within this chapter I also touch on a recurrent fantasy that words ‘really’ (etymologically, or in a speaker’s mind) have only one meaning, which can be recovered by philosophical procedure. This chapter stands apart from the rest of the book, in that it is, to use Saussure’s terms, about ambiguity in langue—that is, in the structure of language, in its lexicon and syntax, not yet realised in use. The later chapters, by contrast, are about ambiguity in parole: in specific utterances, and especially in texts. Langue offers the rules within which parole operates; its ‘ambiguity’ represents the plenty from which doubt may arise on particular occasions.

Chapter Two examines the rôle of ambiguity in a hermeneutic setting that sees it only as doubt and never as plenty, namely, the English common law, where discussions about the nature of ambiguity serve as a proxy for a deeper controversy about what it means to interpret a text—a will, a contract, or a statute. Chapter Three introduces the notion of artificial ambiguity, understood at the level of speech-acts, which classical and early modern scholars usually conceived of either as puns, that is, ambiguities that are not really ambiguous, or as equivocations, ambiguities engineered to deceive; the latter category was the basis of the infamous sixteenth-century debate about Jesuitical equivocation. Chapter Four turns to Scripture, whose ambiguity is seen, following Augustine, both as a difficulty to shake us out of exegetical complacency, and as an inspired involution of multiple meanings on the page; these meanings are not only allegorical, mystical, or typological, but also literal, according to a widespread Catholic idea neglected by previous historians of biblical scholarship. In Chapter Five I return to artificial ambiguity, teasing out its implications for the (p.23) early modern study of classical poetry. This encompasses commentaries on the ‘elegant’ ambiguities in particular lines as well as theoretical treatises and dialogues struggling to make sense of ambiguity as a poetic and political virtue.

The second half of the book, which is more neatly chronological, contains a series of interlinked variations on the themes and ideas of the first. These might be thought of as attempts to reconcile the artificial and the inspired types of ambiguity, or as varying critical responses to the usual hermeneutic focus on the author’s intention. Against those who insist that the intention is single and so must disambiguate the text, it may be argued either that the intention is irrelevant and should be discarded, or, with greater subtlety, that intention is more complex than it looks, and can itself generate ambiguity. The major breakthrough in this respect, as also in reconciling artificial and inspired ambiguity, was the nineteenth-century theory of dramatic irony, which is central to the narrative of the book.

Chapter Six starts in the early eighteenth century with readers arguing over multiple meanings in Homer, and then turns to readers of Alexander Pope, via Pope’s own translation of the Iliad. This marks a surprising episode in the prehistory of ‘close reading’, where the poet’s imputed ambiguities become a counter of hermeneutic authority for which he vied with his hostile contemporaries. Chapter Seven, which centres on the mid-eighteenth-century figures William Warburton and George Benson, considers the way in which the reading of secular poets like Homer and Vergil came to chime with an ongoing debate about the possibility of double senses (and therefore ambiguity) in Old Testament prophecy. It ends in the 1760s, when German scholars, keen readers of Benson and other English theologians, began to reach a rationalist consensus on the unitary sense of prophecy. Chapter Eight examines the reaction against this consensus and the unexpected recrudescence of an older, mystical attitude to interpretation in the work of Johann Georg Hamann, whose writings, whatever their philosophical value, had a seismic impact on the Romantic thinkers of the next generation. A key product of that impact was Friedrich Schlegel’s new notion of irony, and Chapter Nine traces the flattening of this notion into a useful philological tool—dramatic irony—by German and English scholars in the nineteenth century, a process made possible by a new attention to double meanings in Greek tragedy. The result is a kind of ambiguity that is both artificial for the playwright and inspired for the characters onstage.

My final chapter returns to Empson’s Seven Types, a book about ambiguity in lyric poetry, but one that rejects the dominant concept in previous analyses of that subject, namely, artificial ambiguity. Its innovation was to adopt instead a form of inspired ambiguity, one made possible by the earlier invention of dramatic irony, and also, on another front, by the (p.24) Freudian unconscious. To this end, Chapter Ten offers a conceptual archaeology of three keywords in Seven Typesambivalence, primitive, and hypocrisy—an investigation that will lead us outwards, via Empson’s own ambages, to the realm of moral doubt and human understanding, in which lay his book’s greatest originality.


(1) Richard Holdsworth, Praelectiones theologicae habitae in Collegio Greshamensi, ed. Richard Pearson (London, 1661), p. 379 (Lectio 43). (p.2)

(2) David Tracy, Plurality and Ambiguity (Chicago, 1988), whence the quotation (p. 70); Donald Crosby, Living with Ambiguity: Religious Naturalism and the Menace of Evil (Albany, 2008). See also Adam Seligman and Robert Weller, Rethinking Pluralism: Ritual, Experience, and Ambiguity (Oxford, 2012).

(3) Friedrich Nietzsche, Zur Genealogie der Moral, III.24, in his Philosophische Werke, ed. Claus-Artur Scheier, 6 vols (Hamburg, 2013), VI, p. 154: ‘Nichts ist wahr, Alles ist erlaubt’, alluding to Joseph von Hammer, Die Geschichte der Assassinen aus morgenländischen Quellen (Stuttgart and Tübingen, 1818), p. 84, where the tag is attributed to the Ismaili revolutionary Hassan i-Sabbah. The line has since entered popular culture via William Burroughs, Hakim Bey, and even video games; it recently furnished the title of Peter Pomerantsev’s account of postmodern Russian politics, Nothing Is True and Everything Is Possible.

(4) This very ambiguity, which underwrites much of the present book, calls into question the usefulness (and perhaps even the validity) of the distinction between ambiguity in the production of meaning and in its reception, outlined in Susanne Winkler, ‘Exploring Ambiguity and the Ambiguity Model from a Transdisciplinary Perspective’, in Ambiguity: Language and Communication, ed. Winkler (Berlin, 2015), pp. 1–14. (p.3)

(5) Contrast the observation of one admirer, Barbara Hardy, ‘William Empson and Seven Types of Ambiguity’, in The Critics Who Made Us, ed. George Core (Columbia, MI, 1993), pp. 47–58, at p. 49: ‘When one read Empson, one was reading a critic for whom the poems and plays were warmly alive, all there.’

(6) Cleanth Brooks, ‘Empson’s Criticism’ (as in n. 22 below), p. 126, notes that the book ‘reads for the most part like uncommonly good talk’; compare Denis Donoghue, Ferocious Alphabets (London, 1981), p. 72.

(7) For instance, those detailed at STA, pp. 139–45.

(8) This definition, now usually quoted, is actually the wording of the second edition; the first draft, preserved subsequently in a footnote, finishes: ‘which adds some nuance to the direct statement of prose’.

(9) Compare STA, p. 128n: ‘the problem of definition is again secondary’.

(10) Tom Furniss and Michael Bath, Reading Poetry: An Introduction (London etc., 1996), p. 210.

(11) SCW, p. 103n: ‘the term Ambiguity, which I used in a book-title and as a kind of slogan, implying that the reader is left in doubt between two readings, is more or less superseded by the idea of a double meaning which is intended to be fitted into a definite structure. You can still have a doubt as the [sic] whether one or other of two structures is meant but this is much less common and belongs rather to peculiar states of dramatic self-conflict. However the term still seems to me the natural one to use as long as a reader is uncertain. . . .’

(12) I have taken the original definition of ‘negative capability’ from John Keats’s letter to George and Tom Keats, 21–27 Dec. 1817, in Keats, Selected Letters, ed. Grant Scott (Cambridge, MA, 2002), p. 60.

(13) I. A. Richards, The Philosophy of Rhetoric (1936: New York, 1965), p. 40. For a few subsequent examples, see M. H. Abrams, A Glossary of Literary Terms, 7th ed. (Orlando, FL, 1999), p. 11; John Sutherland, How Literature Works: 50 Key Concepts (Oxford, 2011), p. 10; Virginia Ramos, ‘Ambiguity’, in The Princeton Encyclopedia of Poetry and Poetics, 4th ed. (Princeton, 2012), pp. 43–5.

(14) Gary Wihl, ‘Empson’s Generalized Ambiguities’, in Literature and Ethics: Essays Presented to A. E. Malloch, eds Wihl and David Williams (Kingston, Ont., 1988), pp. 3–17; Paul Fry, William Empson: Prophet Against Sacrifice (London, 1991); William Empson: The Critical Achievement, eds Christopher Norris and Nigel Mapp (Cambridge, 1993); Martin Dodsworth, ‘Empson and the Trick of It’; Essays in Criticism 51 (2001), 101–118; Paul Dean, ‘Empson’s Contradictions’, The New Criterion 20 (2001), 23–30; Lisa Rodensky, ‘Empson’s Seven Types of Ambiguity’, Essays in Criticism 53 (2003), 54–67; John Haffenden’s masterful biography, William Empson, 2 vols (Oxford, 2005–2006), esp. I, pp. 176–229; Some Versions of Empson, ed. Matthew Bevis (Oxford, 2007); David Reid, Ambiguities: Conflict and Union of Opposites in the Robert Graves, Laura Riding, William Empson and Yvor Winters (Bethesda, MD, 2012); Donald Childs, The Birth of New Criticism: Conflict and Conciliation in the Early Work of William Empson, I. A. Richards, Robert Graves and Laura Riding (Montréal, 2013); and Michael Wood, On Empson (Princeton, 2017). There is also older literature: J. H. Willis, Jr., William Empson (New York, 1969); William Empson: The Man and His Work, ed. Roma Gill (London, 1974); and especially Christopher Norris, William Empson and the Philosophy of Literary Criticism (London, 1978), based on his doctoral thesis and including a postscript by Empson himself; and, on the Continent, Giovanni Cianci, La scuola di Cambridge: La critica letteraria di I. A. Richards, W. Empson, F. R. Leavis (Bari, 1970); and Horst Meller, Das Gedicht als Einübung: zum Dichtungsverständnis William Empsons (Heidelberg, 1974). A valuable collection of reviews is Critical Essays on William Empson, ed. John Constable (Aldershot, 1993).

(15) F. R. Leavis, ‘Intelligence and Sensitivity’, Cambridge Review, 16/1/1931, repr. in CEE, pp. 37–39. For a broad survey of the book’s immediate reception, see Haffenden, William Empson, I, pp. 274–86.

(16) ‘John O’London’ [= Wilfred Whitten], ‘Letters to Gog and Magog: Ambiguity as Art’, John O’London’s Weekly, 20/12/1930, repr. in CEE, pp. 32–34.

(17) Matthew Arnold, ‘The Function of Criticism at the Present Time’, in his Lectures and Essays in Criticism, ed. R. H. Super (Ann Arbor, MI, 1990), p. 261. Empson had a remarkably Arnoldian view of criticism, STA, p. 245: ‘It is the business of the critic to extract for his public what it wants; to organise, what he may indeed create, the taste of his period.’

(18) John Middleton Murry, ‘Analytical Criticism’, TLS, repr. in CEE, p. 31.

(19) T. Earle Welby, ‘Time to Make a Stand’, The Weekend Review, 3/1/1931, repr. in CEE, pp. 35b–36a, here quoting Swinburne’s ‘Itylus’.

(20) Ibid., p. 35b: ‘to write is not merely to work the desired association of words but to exclude the undesired’. The same point would be made by Isabel Hungerland, Poetic Discourse (1958: Westford, CT, 1977), pp. 26–27.

(21) See, for instance, Chapter Three below, p. 86 on law (Bacon), and Chapter Four, p. 138 on theology (Augustine). On the history of ‘context’, and esp. on its gradual historicisation as a category, see Peter Burke, ‘Context in Context’, Common Knowledge 8 (2002), 153–77.

(22) Vladimir Nabokov, Ada or Ardor: A Family Chronicle (1969: London, 2012), p. 628. He had already offered a fuller discussion in an essay on translation, ‘On Translating Pushkin: Pounding the Clavichord’, NYRB, 30 April 1964, 14–16: ‘This convenient cant word consists of the words ot, meaning «from», and sebya, meaning «oneself», with a pejorative suffix, yatina, tagged on. . . . Lexically translated, it can be rendered as «come-from-one-selfer» or «from one-selfity» ’. Compare John Crowe Ransom, The New Criticism (Norfolk, CT, [1941]), p. 121, afraid that Empson’s readings might be ‘overreadings’, Cleanth Brooks, ‘Empson’s Criticism’, Accent (1944), 208–16, repr. in CEE, pp. 123–35, at pp. 128–29, and W. K. Wimsatt, The Verbal Icon: Studies in the Meaning of Poetry (Lexington, KY, 1954), p. 46, noting the ‘excessive ingenuities of Empson’.

(23) J.D.C., ‘Seven Types of Ambiguity’, Revolt 8 (1930), p. 20, repr. in CEE, pp. 23–25.

(24) On Empson and intention, see Wood, On Empson, pp. 1–25, and Chapter Ten below, pp. 377–80.

(25) I. A. Richards, Philosophy of Rhetoric, New York, 1965, pp. 38–41; Erving Goffman, Frame Analysis: An Essay on the Organization of Experience (Harmondsworth, 1974), pp. 440–44.

(26) See Chapter One below, p. 71.

(27) See esp. Chapter Five below, pp. 186, on Albert Schultens, and 225 on William Baxter.

(28) See R. P. Blackmur’s description of modernism, ‘Contemplation’, in his A Primer of Ignorance, ed. Joseph Frank (New York, [1967]), pp. 59–80, at p. 69: ‘the reassertion of the secret tongue, and the intense declaration of the absolute power of the word as a thing having life of its own and apart from its meaning’. For a rich analysis of ambiguity in modernist literature, see Christoph Bode, Ästhetik der Ambiguität: zu Funktion und Bedeutung von Mehrdeutigkeit in der Literature der Moderne (Tübingen, 1988).

(29) William Bedell Stanford, Ambiguity in Greek Literature: Studies in Theory and Practice (Oxford, 1939), p. 95, labelling Joyce’s work ‘the ne plus ultra of verbal dexterity’ and ‘the ultima Thule of obscurity’. On Stanford see further Chapter Nine below, p. 363.

(30) Roland McHugh, The Finnegans Wake Experience (Berkeley, 1981), p. 67. The phrase referred to is ‘Sling Stranaslang’, FW 338.22; Leo Knuth first mentioned the ‘slings and arrows of outrageous fortune’.

(31) Ransom, The New Criticism, pp. 101–131, with the quotation at 102.

(32) Norris, William Empson, pp. 23–24, reads this as a falsification of Empson’s method.

(33) Philip Wheelwright, ‘The Semantics of Poetry’ (1940), repr. in Essays on the Language of Literature, ed. Seymour Chatman and Samuel Levin (Boston, 1967), pp. 250–63, at p. 252; the idea of ‘plurisignation’ is developed further in Wheelwright’s The Burning Fountain: A Study in the Language of Symbolism (Bloomington, IN, 1954), pp. 61–62, 112–117. Brooks, ‘Empson’s Criticism’, p. 126.

(34) Ernst Kris and Abraham Kaplan, ‘Esthetic Ambiguity’, Philosophy and Phenomenological Research 8 (1948), 415–35. See also Kaplan’s interesting article, ‘An Experimental Study of Ambiguity and Context’, Mechanical Translation 2 (1955), 39–46.

(35) Cleanth Brooks and Robert Penn Warren, Understanding Poetry: An Anthology for College Students, 2nd ed. (1950: New York, 1956), pp. 571–89, with the definition at 573. My thanks to Margaret Meserve for checking the first edition at short notice. The chapter disappeared in the third (1960) and all subsequent editions.

(36) Wheelwright, ‘The Semantics’. Against Wheelwright see Josephine Miles, ‘More Semantics of Poetry’ (1940), in Essays, ed. Chatman and Levin, pp. 264–68, at p. 267, stressing that ‘general language . . . being tentative, fluid, formal, unfixed, is the most plurisignative’. On the heritage of this idea, see Chapter One below, pp. 69–71.

(37) Monroe Beardsley, Aesthetics, 2nd ed. (Indianapolis, 1981), p. 147. The classicist David West, Cast Out Theory: Horace Odes 1.4 and 1.7 ([Alresford], 1995), p. 13, has condemned this notion as ‘the Pansemantic Fallacy’, giving succour to legions of fogeyish critics, such as Richard Gaskin.

(38) See below Chapter Four, pp. 183–84; Five, pp. 196–97; and Six, p. 253.

(39) Richard Gaskin, Language, Truth, and Literature: A Defence of Literary Humanism (Oxford, 2013), p. 187.

(40) Jacques Dürrenmatt, Le vertige du vague: les romantiques face à l’ambiguïté (Paris, 2001), p. 17, n. 3, laments that Seven Types is untranslatable, given its source material. Curiously, a volume entitled Sémantique de la poésie (Paris, 1979) contains a translated excerpt from The Structure of Complex Words (‘Assertions dans les mots’, pp. 27–83) alongside sections of Tzvetan Todorov, Geoffrey Hartman, and François Rigolot.

(41) Paul de Man, ‘The Dead-End of Formalist Criticism’ (1956), in his Blindness and Insight: Essays in the Rhetoric of Contemporary Criticism (1971: New York, 1983), pp. 229–45. For the general point, compare R. P. Blackmur, ‘A Burden for Critics’, Hudson Review 1 (1948), 170–86, at pp. 179–180.

(42) This essay has been criticised—rightly, I think—for its misrepresentation of Empson: see especially Terry Eagleton, ‘The Critic as Clown’, in Marxism and the Interpretation of Culture, eds Cary Nelson and Lawrence Grossberg (Urbana, IL, 1988), pp. 619–31; and Denis Donoghue, Speaking of Beauty (New Haven, 2003), pp. 128–133. Christopher Norris, ‘Some Versions of Rhetoric: Empson and de Man’ (1984), repr. in his The Contest of Faculties: Philosophy and Theory after Deconstruction (London, 1985), pp. 70–96, at p. 90, sees that ‘de Man is constrained to read Empson according to a certain predisposed rhetoric of crisis that is by no means self-evident in Empson’s text’, but allows, p. 95, that comparisons between the two critics are ‘to the point’.

(43) De Man, ‘The Dead-End’, p. 236.

(44) Compare the later formulation of Paul Ricoeur, ‘The Problem of the Double-Sense as Hermeneutic Problem and as Semantic Problem’, in Myths and Symbols: Studies in Honour of Mircea Eliade, ed. M. M. Kitagawa and C. H. Long (Chicago and London, 1969), pp. 63–79, at p. 68: ‘the sole philosophic interest in symbolism is that it reveals, by its structure of double-sense, the ambiguity of being: “Being speaks in many ways.” ’

(45) Roland Barthes, Le degré zéro de l’écriture (1953), in his Oeuvres complètes, ed. Éric Marty, 3 vols (Paris, 1993–95), I, p. 164: ‘le Mot . . . brille d’une liberté infinie et s’apprête à rayonner vers mille rapports incertains et possibles. . . . Ainsi sous chaque Mot de la poésie moderne gît une sorte de géologie existentielle, où se rassemble le contenu total du Nom, et non plus son contenu électif comme dans la prose et dans la poésie classique. . . . Le Mot est ici encyclopédique, il contient simultanément toutes les acceptions parmi lesquelles un discours relationnel lui aurait imposé de choisir . . . une boîte de Pandore d’où s’envolent toutes les virtualités du langage’’. Contrast Barthes’ more restrained language about ambiguity in Critique et vérité (1966), in Oeuvres complètes, vol. 2, pp. 37–44, now responding to Roman Jakobson, ‘Linguistics and Poetics’ (1960), repr. in his Language in Literature, ed. Krystyna Pomorska and Stephen Rudy (Cambridge, MA, 1987), 62–94, at p. 85.

(46) J. Hillis Miller, Versions of Pygmalion (Cambridge, MA, 1990), p. 95, noted of Seven Types that ‘the record of an extraordinarily innovative series of acts of reading . . . has now been assimilated into our culture’. Earlier, in Fiction and Repetition: Seven English Novels (Cambridge, MA, 1982), p. 234, he remarked that he had found Empson useful ‘even if sometimes only to help me work out my own different conclusions’. Compare Geoffrey Hartman, Saving the Text: Literature / Derrida / Philosophy (Baltimore, 1981), p. 23, and idem, A Scholar’s Tale: Intellectual Journey of a Displaced Child of Europe (New York, 2007), p. 137. For an example of ambiguity in use, see J. Hillis Miller, ‘Topography and Tropography in Thomas Hardy’s “In Front of the Landscape” ’ (1985), repr. in his Tropes, Parables, Performatives: Essays on Twentieth-Century Literature (Durham, NC, 1991), pp. 195–212, at p. 206.

(47) Veronica Forrrest-Thomson, Poetic Artifice: A Theory of Twentieth-Century Poetry (New York, 1978), esp. pp. 2–15, wrestling with his interpretation of Shakespeare’s Sonnet 94.

(48) On Rimmon and Kermode’s ‘London Graduate Seminar’, see Bernard Bergonzi, Exploding English: Criticism, Theory, Culture (Oxford, 1990), p. 99.

(49) Edmund Wilson, ‘The Ambiguity of Henry James’, Hound and Horn 7 (April–June 1934), 385–406, repr. in his Triple Thinkers (1938), and now in A Casebook on Henry James’s “The Turn of the Screw”, ed. Gerald Willen (Binghamton, NY, 1960); A. R. Gard, ‘Critics of The Golden Bowl’, Melbourne Critical Review 6 (1963), 102–109; Jean Frantz Blackall, Jamesian Ambiguity and The Sacred Fount (Ithaca, NY, 1965); Charles Thomas Samuels, The Ambiguity of Henry James (Urbana, IL, 1971); Ralf Norrman, Techniques of Ambiguity in the Fiction of Henry James (Abo, Finland, 1977); see also Dorothea Krook, The Ordeal of Consciousness in Henry James (Cambridge, 1967), pp. 228–234; and J. A. Ward, ‘The Ambiguities of Henry James’, The Sewanee Review 83 (1975), 39–60.

(50) Shlomith Rimmon, The Concept of Ambiguity: The Example of James (Chicago, 1977), pp. 12 (citing Stanford, Ambiguity, p. 87) and 17.

(51) On isotopy, see A. J. Greimas, Structural Semantics: An Attempt at a Method, tr. Daniele McDowell et al. (Lincoln, NE, 1983), pp. 78–115, esp. p. 112: a text seems ambiguous when we cannot identify its isotopy.

(52) See Chapter Four below, pp. 149–51.

(53) On Empson’s inimitability, see Wood, On Empson, p. 5: ‘I see the Macbeth passage [in Seven Types] not as a model—who could follow it?’ Compare R. P. Blackmur’s incisive remarks on Joyce in ‘Critical Prefaces of Henry James’ (1934), repr. in his Primer of Ignorance, pp. 244–245.

(54) On the Tel Aviv School, see Brian McHale and Eyal Segal, ‘Small World: The Tel Aviv School of Poetics and Semiotics’, in Theoretical Schools and Circles in the Twentieth-Century Humanities: Literary Theory, History, Philosophy, eds Marina Grishakova and Silvi Salupere (New York, 2015), pp. 196–215.

(55) See the items in UC Irvine Special Collections, J. Hillis Miller Papers (MS.C.013), Box 60, Folder 2. My thanks to Professor Rimmon-Kenan for her personal communication on this matter.

(56) J. Hillis Miller, ‘The Figure in the Carpet’, Poetics Today 1 (1980), 107–118, at p. 112.

(57) This distinction—which does not strike me as a coherent one—is laid out as plainly as possible in Timothy Bahti, ‘Ambiguity and Indeterminacy: The Juncture’, Comparative Literature 38 (1986), 209–223.

(58) Shlomith Rimmon-Kenan, ‘Deconstructive Reflections on Deconstruction: In Reply to Hillis Miller’, Poetics Today 2 (1980–1), 185–88, at p. 188.

(59) J. Hillis Miller, ‘A Guest in the House’, Poetics Today 2 (1980–81), 189–91, at p. 189.

(60) Compare Umberto Eco’s notorious formulation of the ‘postmodern attitude’, Postscript to the Name of the Rose (New York, 1984), pp. 47–48.

(62) A note on word history. It is commonly assumed that the prefix ambi- denotes duality, as in, for instance, ambivalent. But the prefix in ambiguity means ‘around, about’, and the Latin root verb ambigere (< ambi-agere) meant ‘to wander around’, hence ‘to waver, hesitate’; compare Gk. ἀμφι- βάλλειν‎ (Liddell and Scott, sense III), whence the adjective ἀμφίβολος,‎, ‘ambiguous’; and Eng. vague, from Lat. vagus, ‘wandering’. Lat. ambiguus and ambiguitas, and their descendants, have not in usage been limited to connoting mere duality, any more than doubt (< dubitare < duo habere) or combine. However, certain scholars, reinterpreting ambi-, have sought to impose twoness: see, e.g., the jurist Andrea Alciato, in Chapter Two below, p. 83. In English, a few neologisers have coined multiguity, meaning ‘having more than two senses’; there are many recorded examples, most of which were probably formed independently, but see for instance Philip Howard, Winged Words (London, 1988), p. 62. Sadly, or perhaps happily, multiguity has evaded the latest recension of the OED.

(63) Simone de Beauvoir, Pour une morale de l’ambiguïté (Paris, 1947), where the word is a correlate of Jean-Paul Sartre’s être-néant, man’s dual status as free subject and factic object.

(64) James L. Porter, ‘What Is “Classical” about Classical Antiquity? Eight Propositions’, Arion, 3rd ser., 13 (2005), 27–62, at p. 30.

(65) The totemic article is Quentin Skinner, ‘Meaning and Understanding in the History of Ideas’, History and Theory 8 (1969), 3–53.

(66) I am thinking primarily of studies like James Hankins, Plato in the Italian Renaissance, 2 vols (Leiden, 1991); and Sydney Anglo, Machiavelli: The First Century. Studies in Enthusiasm, Hostility, and Irrelevance (Oxford, 2009). But one might also include innovative studies of reading such as Lisa Jardine and Anthony Grafton, ‘  “Studied for Action”: How Gabriel Harvey Read his Livy’, Past and Present, 129 (1990), 30–78.

(67) Most recently in Gaskin, Language, Truth, pp. 186–94.

(68) The latter expression seems to have derived from Irwin Deutscher, What We Say / What We Do: Sentiments and Acts (Glenview, IL, 1973), p. 355.

(69) Immanuel Kant, Kritik der reinen Vernunft, ed. Raymond Schmidt (Leipzig, [1924]), p. 396 (A314): ‘Ich merke nur an: daß es gar nichts Ungewöhnliches sei, sowohl im gemeinen Gespräche, als in Schriften, durch die Vergleichung der Gedanken, welche ein Verfasser über seinen Gegenstand äußert, ihn sogar besser zu verstehen, als er sich selbst verstand, indem er seinen Begriff nicht genugsam bestimmte, und dadurch bisweilen seiner eigenen Absicht entgegen redete oder auch dachte.’ On the subsequent history see Otto Friedrich Bollnow, ‘What Does It Mean to Understand a Writer Better than He Understood Himself?’, Philosophy Today 22 (1979), 10–22. Michael Forster, ‘Friedrich Schlegel’s Hermeneutics, in his German Philosophy of Language: From Schlegel to Hegel and Beyond (Oxford, 2011), pp. 45–80, at 57–59, considers the principle in relation to the unconscious in Schlegel, Nietzsche, and Freud.

(70) See recently Peter Gordon, ‘Contextualism and Criticism in the History of Ideas’, in Rethinking Modern European Intellectual History, eds Darrin McMahon and Samuel Moyn (Oxford, 2014), pp. 32–55.