The Uses of Error 
by Frank Kermode.
Collins, 432 pp., £18, February 1991, 9780002154659
Show More
Show More

Frank Kermode having now become ‘Sir Frank’, it seems a good moment to take a look back over his remarkable career: though by no means because that career is at an end, for he is producing at such a rate just now that it is quite a job to keep up with him. Very broadly, one can think of his career so far as falling into four stages. The first stage, from Romantic Image (1957) to Puzzles and Epiphanies (1962), was very much imbued with Symboliste theory, and Kermode was ready to go along with the notions of the autonomy and organic unity of the work of art (and with the word ‘art’ itself) and with the identity of form and meaning. Merely – though of course it was not a small ‘merely’ – he argued against the hermetic tendencies of Symboliste aesthetics. He praised Yeats for insisting that poetry was made for ordinary human beings and for ignoring the forbidding notice ‘No through road to action’, and he contested the idea, implicit in Eliot’s ‘dissociation of sensibility’, that to embrace Donne you had to give up Milton. It was a stage in which, not for the last time, Kermode offered himself as a reconciler and peacemaker. Romantic Image, moreover, in the way it pursued a certain figure, that of the Dancer, through innumerable avatars, anticipated the long vistas of his later work.

The second stage is represented by The Sense of an Ending (1967) and the collection Continuities (1968), and is deeply concerned with narratology, and especially with the idea that human beings, readers and writers alike, are incorrigible chiliasts and find they can only make sense of their existence by interpreting it through paradigms like ‘crisis’, ‘apocalypse’, ‘decadence’ and ‘renewal’. The habit of resorting to these comforting fictions, so the argument runs, is inescapable, and a writer’s duty lies, not in trying to eschew them, but in constantly checking and regulating them against his or her sense of reality. One notices how much capital and rhetorical effect Kermode gets out of that propitiatory pronoun ‘we’, by which he offers himself as a scapegoat or lightning-conductor on our behalf.

Then, just about the year of The Sense of an Ending, the French invaded our shores, and his new book, according to his own rueful account, ‘became antediluvian almost on publication’. His reaction was wholly admirable. He founded a seminar at University College London to explore ‘Structuralism’ and kindred imported theories. It lasted from 1967 to 1976, attracting what for us now are some very familiar names – Jonathan Culler, Christopher Norris, Annette Lavers, Stephen Heath etc – and under his guidance, we gather, they all got on extremely well, ‘preserving a tone of good humour in the midst of the most serious, even the most fierce, exchanges’.

Kermode remained, as he declared in Continuities, ‘more in favour of continuities than of schisms’, and he became more interested (or interested in a different way) in the long vista of continuity represented by ‘the classics’. Interest in the apocalypse was joined, under the influence of Frances Yates, by an interest in the ‘Imperial theme’, the theme of timeless empire continually renewing itself over huge tracts of history by means of ‘translations’ and ‘renovations’. This imperium sine fine was seen by Kermode as a paradigm of the ‘classic’ – that is to say, of that notion of a literary model or measure binding on all European cultures. In his T.S. Eliot Lectures of 1973 (published as The Classic) he widened Eliot’s own discussion in What is a classic? by relating it to current controversies about ‘intention’, about whether it matters what the author of an ancient work actually meant by it himself. Which should a literary critic model himself or herself on, asked Kermode: the hermeneutically-minded archaeologist or philologist who tries to project himself into dead ways of thinking and feeling, or the Medieval commentator who ‘accommodates’ awkward texts by allegorising them? His answer, it emerged, after some vigorous re-interpretations of modern classics, was a typically Kermodian and irenic one: that ‘hermeneutics’ and ‘accommodation’ come to much the same in the end. ‘Imperial concord’ seemed about to descend, olive-crowned, and everything appeared in danger of fitting almost too well.

These were golden days and were not to last. ‘Deconstruction’ arrived and gave a new and fearful jolt to literary minds. Kermode went to Cambridge, and this was not perhaps altogether a happy move, for soon afterwards the place was ravaged by the McCabe affair. Kermode could have no doubt, he reflected disconsolately in Essays on Fiction (1983), that his own inadequacy as a mediator had been demonstrated. There was a war on, ‘and he who ventures into no-man’s-land brandishing cigarettes and singing carols must expect to be shot at.’ It was altogether a turning-point for him. Like St Jerome in Bethlehem, he retired to his study and applied himself to his Bible (the evenings in Cambridge, he has written recently, seemed very long).

The outcome was The Genesis of Secrecy, which addressed the ‘hermeneutic problem’ and the paradoxes of the ‘hermeneutic circle’ in more depth: a book which offered ‘an interpretation of interpretation’. Of the alternative approaches, hermeneutics and accommodation, he was by temperament more attuned to the second, and he now sought a justification for the activities of literary critics in those rabbinic commentators and New Testament gospel-authors who bridged the gap between the first readers of a sacred text and later ones by rewriting and augmenting the text, thereby explaining it in a more comfortable sense. Kermode is still concerned with narrative, and with two problems about narratives – what kind of thing they are, and how they are produced – which, by an ingenious manoeuvre, he argues to be one and the same problem. It is not merely commentators – rabbinic, Christian or Princeton professor-type – who ‘interpret’, it is what any storyteller does. The Evangelist, for whom what makes a story true is that it interprets and fulfils an Old Testament testimony or prophecy, is in no very different situation from a novelist, with whom the later pages of his novel are the interpretation of its earlier ones. This is how narratives are generated, and it need not surprise us if literary critics continue the same interpretative process after the book has been printed.

Kermode’s Biblical studies continued for some ten years, but when he was invited to deliver the Clarendon and Northcliffe lectures for 1987 he thought it a good occasion to make ‘a clean break’ with them. In these lectures, published together in 1988 as History and Value, and in two succeeding books, An Appetite for Poetry (1989) and Poetry, Narrative, History (1990), he took for his subject the question of value in literature. It is an issue which, he said, had tended to be lazily or discreetly edged aside over the last few decades and had subsequently returned in a new form: concerning ‘canon’-formation and institutional control of the ‘canon’. My periodisation becomes a little shaky here, for his cunning polemic Forms of Attention (1985) was already all about canons: nevertheless we may regard this as the fourth, which is to say the current, stage in his development.

His project in History and Value was at first sight a very promising one. He would have another look at the writers, those of the Thirties, who were producing during his own adolescence – with the thought that, even over so short a space as fifty years, the laws of literary history and hermeneutics might apply and he might find that, to salvage or recuperate admired works of that period, he would need to resort to midrash and drastic re-interpretation. The problematics of history and value would, moreover, be posed with peculiar force by Thirties writing, seeing that the extremer Marxist theorists of this period were liable to say that a new political order would require the overthrowing of all previous notions of literary value, or even that English literature from Caxton to the present day, being composed in ‘an artificial jargon of the ruling class’, might have to be given the boot.

In practice, this experimental revaluing of the Thirties does not strike me as having worked very well, for reasons I shall come back to. The strength of Kermode’s present standpoint is better seen in the lengthy prologue to his succeeding book, An Appetite for Poetry (1989), an impressive, embattled and somewhat gloomy piece in which he defines his current stand on all the issues that most concern him: the ‘contingencies of value’, the take-over of criticism by Theory (and the politics of Theory), the nature of canons (and their inseparability from commentary), and the status and concept of literature itself. His conclusion is that literary criticism (by which we are to understand academic literary study, for, so Kermode holds, there is no longer any such person as the ‘common reader’) is in a mess, as evidenced by the fact (a question of value if ever there was one) that it is possible to make a career in a ‘literature’ department, not only without having, but without pretending to have, any ‘appetite’ for literature. (He retails, with Stoic mildness, a horror-story about the English faculty at Duke University where, according to a report in the Wall Street Journal, one highly-paid lecturer declared they were ‘throwing out the notion of good or bad, or ignoring it’, and another confessed to only ‘using’ Shakespeare to illustrate the mistreatment of women in the 17th century.)

It is no small thing to have behind one so long and coherent an intellectual career, one with so many ‘continuities’. For one’s overriding impression is that under all the ‘translations’ and ‘renovations’ one is watching the evolution of the same pursuit or set of preoccupations. In a panegyric on Frances Yates in the collection The Uses of Error, the book now under review, Kermode speaks of her as ‘one who, in the course of time, remained ingeniously and beautifully faithful to some hint of illumination that was worth a life’. It could be said of Kermode himself. He would, of course, as a peace-maker, have been attracted by her vision of ‘imperial concord’, also maybe envious of her capacity to believe anything as wholeheartedly as she could. He was much and long influenced by this Britomart-like enthusiast, though there are hints that after her death, he has, like many others, succumbed to a sneaking feeling of having – ever so slightly – been taken for a ride. (Banish it as one will, when one reads Yates on the shadowy but ubiquitous ‘Rosicrucian enlightenment’, the word ‘Baconianism’ does keep slipping into the mind.)

It can be fairly said, too, that no one has done more than Kermode, or even as much as him, to hold the literary community together. He has confronted several bloody revolutions in literary studies without losing his nerve or, except very occasionally, his temper; has shown that one can, as one should, take on everything in the way of serious novelty and yet keep one’s own independent course or, as he likes to say, ‘follow one’s nose’. He writes for the dry-as-dust scholar, for the advanced literary theorist, and for the casual reader of literary journals, and all this neither with trendiness nor with pretended indifference to fashion. When in 1970 he launched the ‘Modern Masters’ series, or as he now describes it his ‘Guru series’, it was for the not ignoble reason that fashion called for it just then. ‘People wanted to know about these people rather than to know in detail what they said.’

In the Introduction to his new collection of reviews, The Uses of Error, Kermode speaks up very becomingly for the dignity of reviewing. It is, he says, ‘something you can only do well enough if you are also doing something else well enough’: but, that said, it presents its own opportunities and challenges. The sharp distinction made by Germans between Tages-kritik and Literaturwissenschaft seems to him culturally harmful. As for writers republishing reviews, it is, says Kermode, hard to see why, if they can get away with it, ‘they should not be allowed to enjoy the measure of permanence, and the measure of vanity, proper to their station, especially if they believe, as I admit I do, that some of their best writing has been “buried” in reviews’. It is plain that he writes his own reviews with every expectation of reprinting them, and since they are an essential part of his oeuvre, no one is likely to complain.

There would he plenty to say about this new collection, but I shall not attempt to say it or spend very long on anything in the volume, apart from one piece, the remarkable sermon (it is, he says, the first and last effort he intends to make in that direction) which gives the book its title. More of this later. Meanwhile, just to give the flavour, let me mention an item or two at random. A review of R.B. Martin’s Tennyson: The Unquiet Heart provides some good examples of Kermode’s felicity as a phrase-maker, as when he speaks of Tennyson’s ‘entranced research into the possible sounds of English’ and of Tennyson’s ‘so private a gift, an endless mouthing into shape of sounds, a fumbling at dark phonetic limits’. The article ‘Remembering the Movement’ illustrates his knack of insightful generalisation: ‘the best work of the group was a poetry, and a fiction, of disappointment: work that was aware of what it had given up to be what it was’. Even better, in this line, is his characterisation of H.G. Wells as a writer always, and by destiny, ‘betwixt and between’. ‘Even the means by which he had escaped the oppression of the class system – his writing – took him into a profession where he was neither one thing nor the other, neither journalist nor artist.’ A subtle and sympathetic piece, ‘Fighting Freud’, convicts the revisionists and detractors of Freud, the ones who accuse him of being ‘unscientific’, of measuring him against the wrong kind of science and making a confusion between the physical and the hermeneutical sciences. There might be ways in which Freud’s theories were wrong, but ‘these ways are not to be discovered by indicating differences between his theory and theory of physics, though he himself might have supposed so.’

He includes one or two autobiographical pieces in the book, including an essay on what it meant to be born a Manxman; and he reprints his tough and furious rejoinder (first published nine years ago in the periodical Raritan) to a famous attack on him by Helen Gardner. It was a bizarre episode. Gardner, invited to give the Charles Eliot Norton Lectures for 1979-80, described in them how for ten years she had been preoccupied with professorial duties and close textual editing and had only recently felt free to take a look around at what had been happening in English studies meanwhile; and what she found, she said, had alarmed ‘this kind of Rip Van Winkle’ exceedingly. She met with extraordinary jargon, borrowed from half a dozen other disciplines; grotesquely irresponsible theories about the indeterminacy of meaning and the dispensability of the author; and, worst of all, a deliberately anti-humanistic trend. And the ringleader or one of the ringleaders in this, she found, was her friend Frank Kermode! She pointed out various gross fallacies in his reasoning and even, unwisely, hinted that his Biblical scholarship was not all that it might be: and various reviewers expressed joy at seeing him thus ‘covered with blood’, with Imagination, Humanity and Principle saved from his machinations. This mild controversialist was taken aback and thoroughly stung, and in return he really let fly, accusing Gardner of arrogance, ignorance and, at one point, of ‘baseness’.

Helen Gardner’s was the wrong way to argue with Kermode. Still, there are problems with his ‘system’. It is hard to get over the feeling that he, and other exponents of hermeneutics, simply exaggerate the remoteness and inaccessibility of works of art of the past. I have an especial affection for his book Forms of Attention, which deals with the rediscovery of Botticelli in the mid-19th century and his re-admission to the canon after three centuries of neglect. For this return to the canon to become permanent three things were required, wrote Kermode: ‘opinion’, in the form of a felt need, on the part of Burne-Jones, Swinburne and Walter Pater, for a certain kind of early Renaissance art; ‘knowledge’ (that is to say, solid art-history), since Pater and co got a great deal wrong – for instance, the ‘cadaverous’ colour so admired by Pater in The Birth of Venus was merely the result of deteriorating pigment; and ‘interpretation’ – the continuance of Botticelli’s ‘modernity’ requiring the wild theorising of Aby Warburg about ‘mnemic energies’ and eternal recurrence. In other words, once a work has been admitted, or re-admitted, to the canon, it will receive – and will need, if it is to retain its position – forms of attention quite different from those granted to non-canonical works.

I find this extraordinarily illuminating: yet I am suspicious of the larger conclusions Kermode draws from it. It is certainly true that Botticelli was not ‘available’ to the age of Voltaire, nor was Gothic architecture, and such rediscoveries (Vermeer’s is another case) are an immensely significant part of cultural history. But, so far as I know, Michelangelo and Rembrandt have never needed to be rediscovered; and one might argue, and I think I do argue, that it is the remoteness and alienness of Botticelli and Gothic architecture that was the illusion, needing to be dispelled, and not the other way round. There is nothing immutable about the reputations of Michelangelo and Rembrandt (or of Chaucer, Shakespeare and Pope), nor about the things we shall say about them, but it is unlikely that we shall wake up one day to find they have slipped out of reach and that we stare at their work in blank incomprehension or indifference.

Then, we need to remember that the term ‘canon’, as Kermode uses it, is only a metaphor, and a rather loose one. The canon of Scripture is closed for ever, whereas the secular ‘canon’ is (if we follow Eliot’s view) a living affair, admitting of new inclusions, with a jostling effect on the existing incumbents. Again, the contents of the Scriptural canon are known, which is not at all the case with the secular one. If you asked an English lecturer to enumerate the canonical English writers, he or she would no doubt name Shakespeare and Wordsworth and a few others, but otherwise there is no knowing how long a list he or she would give (twenty names? fifty? a hundred?), or whether, shall we say, Thomson of The Seasons will get in. The term ‘canon’, in Kermode usage, is no more than a broad gesture, much like the term ‘the establishment’. Kermode says in History and Value that what opponents of the canon tend to want is not so much to abolish as to capture it, and this is a true remark, but it needs the rider that the canon is, anyway, a gaseous or elastic entity.

It is of importance that Kermode is nowadays less inclined to use the term ‘art’ in discussing literature, and that other literary critics boycott it on principle. For to hold on to this term goes with accepting the fundamental kinship of the arts, and this will work against some current assumptions about ‘literature’. For instance, it is the hermeneutic assumption that what the reader and the critic is looking for in a work of literature is its ‘meaning’, whereas the first question we ask of a quartet by Haydn or a still-life by Chardin will hardly be ‘What does it mean?’ Which is not to say that, framed in a certain way, this will not at some stage prove a sensible question. The answer to it is likely to be, initially, an ontological one, though leading out from this to human activity in general. (Could there be a more profoundly significant cultural phenomenon than the way that, in the Haydn period, musical form suddenly developed an impetuous forward movement and began to hurl itself towards a goal?) Naturally, the fact that ‘literature’ uses words, the currency of everyday communication, and is thus embedded in ‘meanings’ in the more straightforward sense, must put it in a different situation from music or even from painting: but all the more reason – and this is where the term ‘art’ is beneficent – not to see this difference as more absolute than it is.

Again, the sober and unproblematic sense attached to ‘interpretation’ in music is a useful reminder. One of the objections one feels inclined to make to the hugely increased importance attached to interpretation by hermeneutic theorists, Kermode among them, is that it seems to want to overlook literary genres. Certain genres of poem, in particular allegory, specifically call for interpretation, according to strict regulations laid down by itself. Fairy stories of the Hans Andersen kind thrive on unregulated interpretation, and as it were universal applicability, and would be pointless without it. By contrast, certain kinds of Symboliste poem appear to be intended to defeat all attempts at interpretation, this being again an aspect of their genre. The debate in Essays in Criticism, many years ago, about Eliot’s ‘A Cooking Egg’ brought this out. (Of course we were all less smart at interpretation then.)

To think of literature as one of the ‘arts’ has the further advantage of bringing up the issue of value. For it seems fair to say that music, once it had gained its independence from the church, had no other function than to be valuable. What raison d’ être can a Haydn quartet be said to have apart from this? It certainly isn’t useful; value is its mode of existence. (I am here saying nothing about what kind of value.) But can it be true that what is not merely central to, but the whole justification for the existence of, ‘pure’ music is not a large, even a predominant factor in all the arts?

But if so, is not Kermode actually taking the past neglect of literary value rather coolly? The law seems to be that if one considers ‘value’ to be the concern of the literary critic at all, it is going to have to be his prime concern. Talk of ‘value’ on critics’ lips tends to sound feeble and plaintive if ‘value’ is considered merely as one item of concern among many: it only makes sense if one posits – as I would myself – that, directly or indirectly, valuation, which Paul de Man has called ‘the critical act in its most naive form’, is the only thing that criticism is concerned with. Kermode’s way of reviving the issue of literary value, as a matter of canons and the politics of canons, is valuable and suggestive, but it does not seem to me adequate (of course partly, but by no means only, because the canon is such an uncertain fiction). What would be involved in giving value primacy is a basic conviction that art is, in origin, entirely an empirical matter, a history of discovering certain things to be good. (It helps to defuse the controversy over authorial ‘intention’ to remember that artists, who are involved in value-judgment at every moment of the day over their own work, habitually speak, not of intending certain things, but of discovering them.) It also implies a theory of criticism or aesthetic conversation something like Wittgenstein’s – as being a sophisticated kind of pointing, almost as if by dumb-show.

It may be significant that Kermode’s experimental return visit to Thirties literature does not appear to have worked too well. Reinterpretation did not enable him to find more to say in favour of the ‘proletarian’ novels of Walter Greenwood, Walter Brierley and Lewis Jones etc. At most, it helped him to be more appreciative of the ‘bourgeois’ Edward Upward’s trilogy ‘The Spiral Ascent’ and to make out a convincing case – one I would certainly go along with – for the greatness of Auden’s ‘Spain’, a poem he had earlier described in Romantic Image as ‘once greatly admired but now embarrassing’. It was not much of a harvest for midrash. Indeed he was, on this occasion, depending more on hermeneutics than on ‘accommodation’; and according to Julian Symons,* who is part of the very past that Kermode was patiently reconstructing, he did not get it all quite right – did not know, for instance, what the ‘boiled orange’ was, in a Kenneth Allott poem. Then, what also amazes one – though of course Kermode is not peculiar in this – is the insouciance with which so subtle a student of rhetoric uses those intensely complex and devious terms of rhetoric, ‘bourgeois’, ‘proletarian’, ‘upper-class’, ‘middle-class’ and ‘working-class’. If nothing else, should he not be arrested by the acute dissonance, the ‘judder’, between the binary scheme bourgeois/proletarian and the tripartite scheme upper-class/middle-class/working-class?

One can discern two different identities in Kermode. There is the civic-minded critic who strives, sanguinely at some periods and grimly and near-despairingly at others, to hold things together and induce conflicting factions to speak to one another; and then there is the marvellously-gifted individualist who trusts his interpretative luck and ‘follows his nose’. In The Genesis of Secrecy he says that, though ‘we’ know that any literary interpretation must depend, tacitly, on institutional training, ‘we tend to reserve our highest praise for those interpretations that seem most intuitive, most theory-free, seeming to proceed from some untrammelled divinatory impulse.’ This certainly applies to Kermode himself. He has a natural gift for divination that other critics would sell their birthright for. It was Kermode who, noticing that the names ‘Pyncheon’ and ‘Maule’ in The House of the Seven Gables are derived from type in the printing-house sense, performed a sweeping and most dazzling interpretation of Hawthorne’s novel in terms of ‘types’ of every kind – in photography, natural history and Scriptural allegory. It was he, again, who diagnosed Hemingway as a ‘one-handed writer’, suggesting that what he learned from Gertrude Stein was how to make writing difficult for himself, as if by tying one hand behind his back: with the consequence – this is no mean fiction – that he became interested in the myth of manliness in general and the cult of the crippled hero, a subject-matter which brought out the sentimentalist in him. It was Kermode who, in The Genesis of Secrecy, having invoked Hermes as the god of hermeneutics, conducted a striking interpretation of Henry Green’s Party Going, in which he was able to identify a character unnamed by Green as Hermes in person!

There is, one observes, a kind of joyous superfluity about this last example: his exegesis did not depend on his discovering Hermes among the cast of Henry Green’s characters – it is a charming bonus. This brings me to Kermode’s sermon on ‘The Uses of Error’, delivered to the University of Cambridge in King’s College Chapel on 11 May 1986. It begins with the very complicated history of the passage in the Book of Job in which Job’s wife says to him, as he sits among the ashes: ‘Dost thou still retain thine integrity? Curse God and die’. In the Hebrew text what she says is ‘bless God and die,’ and this is how St Jerome translated it in the Vulgate. But commentators affirm that ‘bless’ is a euphemism for the tabooed word, and that St Jerome expected his readers to realise this. This leads Kermode to meditate on Georges de la Tour’s magnificent painting which was once though to depict ‘The Clothing of the Naked’ or an angel liberating St Peter from prison, but in 1935 was shown to represent ‘Job Visited by his Wife’. He points out how, as soon as this identification had been made, people began – almost as by a gestalt transformation – to read the expression of the female figure quite differently and as representing cruelty. It suggests to him the potential fruitfulness of error: for the possibilities opened up by this case are incalculable and extend in every direction. Was St Jerome right in believing his original readers would understand ‘bless’ correctly? Is it conceivable that La Tour took ‘bless’ literally? What exactly is Job’s wife saying anyway? What conclusions should we draw about the contingencies of our ways of looking at works of art? Later in his sermon he goes on to follow the trail of midrash-like augmentations and ‘wrong’ readings, going back to the Old Testament, which led to the writing of Vaughan’s exquisite lines in his poem ‘The Night’, where the fruitless visit paid by Nicodemus to Jesus by night is transformed into a supreme illumination. Kermode’s sermon is in part, and explicitly, a justification of the errors any interpretative critic is bound to commit: a very up-beat apologia, taking a much more buoyant and consoling view of ‘blindness’ than de Man’s. But further, it exhibits the supererogatory aptness, and unsought-for richness of relevance to his long-term concerns, that characterises his best writing. By a sort of double-take or triple-take, this writer manages not only to practise what he preaches but to practise it at the same time as he is preaching it.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences