An Appetite for Poetry: Essays in Literary Interpretation 
by Frank Kermode.
Collins, 242 pp., £15, November 1989, 0 00 215388 2
Show More
Show More

Frank Kermode belongs to no sect of literary criticism, and he has founded no school. Like William Empson, whom he praises as a ‘genius’ of criticism, Kermode has always been more interested in a poetic than in a theoretic approach to the study of literature. He thinks that literature itself – rather than theories about it – is our best guide to how to read critically, and he has devoted the better part of a long career to this conviction. Kermode is a working critic and proud of it. When he does turn to the sustained consideration of questions of theory and method, it is almost always in the interest of identifying what is useful or illuminating, what can help us to read better, and what impedes us from doing so, in a given critical position or practice, rather than in elaborating a comprehensive theoretical perspective of his own. As for what ‘reading better’ consists in, Kermode would be fully justified in pointing to his own vast production of literary critical essays as a case in point: the bibliography of his works published between 1947 and 1988, appended to Poetry, Narrative, History*, lists 308 items. Over a long career, Kermode has been a consistently patient, generous, informed, and above all intelligent reader of literature. We can say of him what he says of Empson: ‘He never loses class’.

This ‘class’ is amply displayed in An Appetite for Poetry, which consists of ten essays originally published between 1975 and 1989 and a Prologue, itself alone worth the price of the volume, which dilates on the current state of affairs in literary criticism and theory. The essays span the period marked by the composition of such books as The Classic (1975), The Genesis of Secrecy (1979), The Art of Telling (1983), Forms of Attention (1985), History and Value (1988) and Poetry, Narrative, History (1989), and they deal with topics and address questions that have concerned Kermode since the appearance of his first book, The Romantic Image, in 1957. Admirers will be pleased to find new appreciations of Eliot, Stevens and Milton and reconsiderations of such theoretical matters as the nature of symbolic language, the relation between fiction and history, and the notion of narrative. These familiar Kermode topics are utilised in this volume to focus attention on and reassess the issues animating current debates over the value of literature, the notion of the classic, the concept of canonicity and the nature of interpretation.

In Kermode’s view, the proper task of the critic is to read and to teach others to read attentively, not only because sustained attentive reading is very difficult, but also because some writing, the kind we used to call ‘literary’, is, pace many of the new theorists of literature, worth all of the care and attention we are capable of spending on it. Not that Kermode is ‘against theory’ per se. On the contrary, he knows that such an anti-theoretical position, when not simply meant to be mischievous, is anti-intellectual. Many of the essays in this collection – ‘Divination’, ‘The Plain Sense of Things’, ‘The Argument about Canons’, ‘The Bible: Story and Plot’, ‘Freud and Interpretation’ – address theoretical matters directly; the Prologue, which reflects on the relation between literary theory and reading, can be called a metatheoretical exercise. But Kermode’s purpose is to take us back to the basics of literary criticism, and he aligns himself with similarly inclined critics (Hugh Kenner, John Hollander, Richard Poirier, Lionel Trilling, and above all Empson) who use theory, when they use it at all, for the most part to clear the ground for attentive reading.

The phrase which serves as the title of the collection is taken from Paul Valéry’s remark about certain ‘men with no great appetite for poetry’ who nonetheless presume to study, judge and cultivate it. Kermode cites it in reference to certain older advocates of the scientific study of literature (Wellek and Frye) and some of the current purveyors of literary theory (Culler and Eagleton) who appear to see no difference between literary and ‘subliterary’ uses of language, or who, when they do see the difference, appear to have little need of and therefore no ‘appetite’ for literature itself. Kermode continues to believe not only in ‘literature’ but in its ‘value’ as well.

The choice of the word ‘appetite’ is significant: it suggests an impulse more rational than ‘hunger’ and less aesthetic than ‘taste’. Latin appetitio connoted both a certain rationality and an element of ethical awareness in the choice of an object of interest. Kermode’s use of the term is intended to indicate a desire for literature that is disciplined by knowledge and a sense of ethical responsibility to the community to which the critic belongs: it also indicates a sense of ethical responsibility to literature itself. Scientists and theoreticians do not have to like, value or respect what they study, but in Kermode’s view, literary critics do.

Kermode shows in the Prologue that he knows his theory and appreciates those who – like Derrida and the late Paul de Man – can theorise about literature with the requisite tact and learning. He recognises that we ought to reflect on such theoretical problems as the relation between literature and other kinds of writing, the connection between the institutions of writing and those of politics or society, the nature of canons, what interpretation consists of, and the like. In his own critical writing, Kermode uses theory in a modest and understated way to discriminate among different ways of reading. The essays are replete with allusions to and reflections on the work of Freud, Schleiermacher, Jakobson, Barthes, Genette, and – surprisingly, to me – Heidegger (in a beautiful meditation on Stevens’s notion of what it means to ‘dwell poetically’ in the midst of ordinary life). But Kermode never allows an interest in theoretical questions to deflect him from the proper work of the critic, which is to teach us to read both literature and other kinds of writing, including theoretical discourse, attentively and to grasp the value of specific literary works. And by ‘attentively’ he means patiently and generously, in the first instance; with proper respect for learning and science, in the second; and with the kind of ‘abnormal’ common sense that only those who have an ‘appetite for poetry’ can be expected to possess, in the third.

All this provides a ground for a deceptively modest theory about the relation between literary classics, canons and the morally responsible interpretation of literary works. In his polemical Prologue, Kermode takes to task current defamers of the canon of literary classics both for their ignorance of the history of the notion of canonicity and the imprecision and inconsistencies of their attempts to conceptualise their own causes. While granting that the notion of a canon presumes some system of inclusions and exclusions, he denies that the Western canon of literary classics is necessarily an instrument of the ‘cultural racism’ of ‘white males’. Nor does he have much patience with the idea that this canon is made up of ‘static monuments’ supportive of ‘political oppression’ and ‘totalising’ in their psycho-social effects. He points out that canons – even religious ones – change over time, are responsive to the body of commentaries produced by both their devotees and their critics, and contain as many self-deconstructing texts as ‘totalising’ ones. And while granting that canons do embody ‘traditions’, he points out that, without some equivalent of the notion of tradition, there would be no need for ‘the special forms of attention elicited by canonical texts’ and so, incidentally, no room for such special forms of attention as deconstruction.’ Kermode notes the contradiction involved in wishing, at one and the same time, to abolish the canon and to reform it by admitting to it the work of the oppressed and marginalised of the world, and he remarks on the fatuity of those who think that either of these goals could be accomplished by ‘doing theory’ rather than by writing literature or practising criticism.

In his reflections on modern Biblical interpretation – considered as a sacred disputation of which the current argument about literary canons can be considered a profane counterpart – Kermode gently but firmly deflates those interpreters, of both sacred and profane texts, who confuse their own ‘foreunderstanding’ of texts with a putative ‘foreknowledge’ of what can or must be found in them. Such is the danger run, not only by all prejudicial criticism, but by most theory-guided criticism as well. In its search for the ‘codes’ in which a discourse is cast, ‘brave Theory puffing by/in silks that whistle’ tends consistently to miss the ‘poetry’ that gives the literary work its unique value. Kermode prefers the older hermeneutical practice of interpretation to both theory-building and historicist contextualisation. Hermeneutics licenses a kind of poetic criticism, which is to say, a criticism both imaginative and responsible to the science and learning of the society for which it is produced.

In four of the essays, Kermode shows us how poetic consciousness works: three are appreciations of poets (Milton, Stevens and Eliot) and one of a critic, Empson, whom Kermode takes to be the very incarnation of a poetic kind of critical intelligence. Kermode reminds us that one error to be avoided is that of the literal-minded reader, who constantly discriminates between what is said explicitly and what is only figuratively connoted, provides a ‘plain’ rendering of the latter and then proceeds to decide the truth or falsity both of what has been said directly and what has only been suggested by figuration. Another error is characteristic of the rhetorical reader, who constantly hunts for the ‘codes’ by which poetic sequence can be sublated into semantic structure. The former fails to see that the literal meaning of literary works is context-bound, that contexts are constantly changing, and that, consequently, the criteria for deciding the difference between the literal and the figurative meanings of any work are also always changing apace, while the latter fails to notice the relation between a work and its context altogether.

Thus, for example, those custodians of justice and truth who find in Eliot’s poetry evidences of attitudes reactionary and even fascistic in nature miss the extent to which his poetry agitates the seeming stability of what he says both literally and figuratively. In his appreciation of Eliot as modern metoikos or poet of urban exile, Kermode shows how the poetry transcended Eliot’s explicitly held political and social prejudices, as well as many of his implicit ones: ‘In consciously holding together ... those diverse ideas of an ideal eternity and a decadence in time, Eliot was unique among modern poets ... an outsider, an exile from easy opinion, banished and banishing, honoured and deplored.’

Poetry, poetic vision and poetic speech take their rise on the boundary between ‘easy opinion’ and the enigmas of human existence which always recede to just beyond the advancing frontiers of our sciences. In his essay on the aged Stevens, Kermode asks what it means to ‘dwell poetically’ in the world – a question inspired by Stevens’s work and an interest the poet expressed in old age in obtaining a copy of Heidegger’s essay on Holderlin’s line:

Full of merit, yet poetically
Man dwells on the earth.

A specifically poetic kind of insight is expressed in Holderlin’s use in this line of the conjunctive adverb ‘yet’ (doch), which suggests both ‘nonetheless’ and ‘additionally’ and therefore makes ambiguous the quality to be ascribed to man’s ‘merit’ and the ‘poetic’ manner of his ‘dwelling’. Kermode affirms the ‘enigma’ expressed in that ‘yet’. He glosses the phrase as suggesting that ‘we dwell poetically on this earth, even in a time of destitution, and that our doing so is somehow gratuitous, independent of our merits, a kind of grace.’ And he goes on to suggest that Stevens’s poetry consists in the effort to descend from the position of ‘merit’ to the apprehension of the value of local places, worn textures, everyday experiences: ‘splendour in poverty, death in life ... the casual boons of the world of poverty’. For the aged Stevens, Kermode remarks, it was not so much a question of confronting a ‘grim reality’ as, rather, coming to terms with ‘plain reality’.

To dwell poetically means, in the case of Milton in his old age, writing Samson Agonistes, ‘a heroic exclamation of ... intimate impulses, and a vindication of the peculiar justice of a God whose arbitrary decisions and devices alone make sense of a hero’s world, and are that of which he must make sense’. Kermode compares the vision expressed in Samson Agonistes to that of Freud’s Judge Schreber, which ‘offered another famous instance of the human imagination at work on the grand questions of election and suffering and the divine plot’.

‘The grand questions of election and suffering and the divine plot’: Kermode makes much of the similarities between religious and poetic vision and between theological and literary interpretation. On his view, the poet shares with the theologian and metaphysician a desire to grasp the ‘divine plot’, even after there is no possibility of imagining any divinity behind it. The notion of a divine plot is not, however, to be confused with the kinds of generic plot-types which both formalists and archetypal theorists identify as the – or at least one – meaning of literary texts. On the formalist view, the meaning of the literary text is exhausted by the literal and figurative (or allegorical) meanings of the words and phrases and the dialectical relations among these that comprise the text. According to Kermode, however, the determination of these various levels of meaning does not touch on – because it cannot apprehend – the relations between the formal features of the work and ‘the human need’ which ‘the poet sees’. The ‘divine plot’ has to do with ‘the grand questions of election and suffering’ with which the aged Milton wrestled in Samson Agonistes. It is not so much the plot of all plots of human life as the effort to grasp some third alternative to the unrealistic option of a perfectly free will, on the one side, and the depressing likelihood of an ironclad determinism, on the other. ‘Divine plot’ is a metafiction – a possibility which makes the distinction between fiction and non-fiction credible.

Kermode is fond of St Augustine’s remark: ‘Not everything we make up is a lie.’ He has always been interested in the possibility of a third alternative to the unalloyed truth, on the one side, and falsehood, on the other. This has been the burden, as I understand it, of his extensive inquiries into the nature of the languages of ‘secrecy’, in which figures of speech and of thought are used not so much to distort truth as, rather, to hide it and close it off to those who are not yet ready for or are deemed unworthy of it. Kermode remarks that, in the reading of any literary work, ‘Coding operations are called for,’ but ‘more difficult books reduce their audience by calling for a higher competence than that.’ This ‘higher competence’ Kermode calls ‘divination’, which is the art or skill of discerning those ‘occult relevances’ of which neither the code can speak nor the author be consciously aware. This skill or art requires a ‘third ear’ because it attends to what we may call a ‘third voice’, which is that of the poet neither representing a world nor figuring it but dwelling in it poetically. The ‘third voice’ is manifested not only in ‘what is explicitly stated and conventionally coded’, but above all in what, after Freud, we can call the ‘condensations and displacements, the puns and parapraxes’ of a text – which is to say, those elements of an utterance in which an ‘occulted’ sense of some ‘divine plot’ of the world are indicated.

Thus, Kermode comes, in his essay on ‘The Bible: Story and Plot’, to notions of story and plot quite different from those which inform the conventional notions of codes. He postulates, over and beyond the recognisable plot-types – of romance, comedy, tragedy, and so on – another kind of plot, ‘an occult plot’, which establishes relations between the parts of the narrative that can be explained by ‘appeal to causality’ and the parts that cannot be so explained. The events that signify the presence in the text of this kind of plot are not those that manifest a ‘law immanently inherent in the world’ so much as those that appear to be nothing more than ‘incident, anecdote, news, in what [Lotman] sometimes calls “scandal” or “excess” ’. The ‘occulted’ plot is related to its conventionally-encoded counterpart as ‘scandal’ might be said to be related to ‘myth’.

All this may seem quite new to Kermode’s readers, but it is perfectly consistent with the main tenor of his thought since at least The Sense of an Ending. He has always held that the literary work – like all speech and all discourse – is at least once removed from reality. It is displaced by virtue of the fact that it is a mediation between a specific aspect of the world and other literary expressions of such experiences. Every writer writes within a tradition or complex of traditions and hews the wood of his or her experience of the world in terms conformable to the traditionally provided matrices thereof – so that every literary work refers experience to the forms of what passes for literature (the canon) of the time and place of the writing. Literature is identifiable by this conformity of the individual work to the canon, which determines what will or can count as literature in a given time, place and cultural condition.

Accordingly, every claim to represent reality by means of literary writing makes of the reality represented a fulfilment of one or another of the structures of meaning available in the literary endowment of the writer’s own culture. Joyce’s Ulysses represents the events of an ordinary day in Dublin, but also replicates a structure of meaning found originally in Homer’s epic. This replication can be said to fulfil a ‘figure’ found in Homer, just as the New Testament replicates and thereby ‘fulfils’ the Old Testament by referring the events of Jesus’s life to the ‘figures’ of them found in the Torah. But these fulfilments are themselves figures, and therefore ‘figures of the truth’ manifested and concealed in the earlier representation. Kermode’s notion of ‘figures of truth’ is a promising new point of departure for the discussion of the relation between fictional and factual discourse, pointing in the direction marked out by Ricoeur’s recent work on the metaphysics of narrative. This notion may contain the key to Kermode’s insight into the language of secrecy as the third alternative to the discourses of truth, on the one side, and of the lie, on the other.

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