I am reading a first book by A., a poet who began as, and may remain, a performance poet - someone who writes for recitation, who evidently wants immediate reaction to oral performance from a live audience - along with a midcareer book by B., who is also a novelist, whose poems tell stories; as I admire the characters they reveal, the wayward insights and the comic self-description, I find myself thinking If these are good poems, then there is no difference between poetry and any other sort of writing.

I am reading a book by C., a rather good, if derivative, poet, the sort of writer whom Pound placed among 'good writers without salient qualities... men who wrote sonnets in Dante's time, men who wrote short lyrics in Shakespeare's time or for several decades thereafter', a writer who has taken over and as it were domesticated already existing ways to write. I find myself thinking, these are good poems but I already know what kind of poems they are: they are the kind already invented by D.

I am reading a late-career book by D., and I like it; I find myself thinking, I already know what kind of poems they are: they are the kind he invented twenty years ago, and has simply continued to write. Were I to come upon this book before I had come upon any of D.'s other books, though, I would have been bowled over, blown away.

I am reading the second or third poem I have ever encountered by E., who blows me away: she seems to have invented a new kind of poem, a new subgenre. When do my expectations settle enough to allow me to recognise that kind of poem when I encounter it again?

When does (when should) E. stop trying to write that kind of poem? After the second attempt? The second book? After E. sees other poets who have tried to write it? After she sees them succeed?



When students, when audiences, ask me what makes a poem good, I say that I call a poem good when I want to reread it several times with sustained attention to how it sounds. But that description also applies to works of linguistic artfulness that are not commonly called poems, e.g.Mrs. Dalloway. I conclude that if I am to describe my own usage (rather than the larger, sometimes contradictory corpus of usage for 'poetry' and 'poem') there is no ahistorical criterion that will separate all the works I call 'good poems' from all the works I do not commonly call poems.

On the other hand, very few of the works I do not call poems have the sort of compression, the invitation to be read not two but twenty times, that I find in the works I am likely to call good poems.



How many styles, how many developments, in contemporary poetry might be construed as reactions to the appearance, or allegation, that nobody reads it!

The hysterical boast, the style apparently meant to attract attention by any means; the strenuously topical headline-hammering culture critique, insistent (like so much journalism in prose) on speaking of matters in which we are already interested; the first book whose tone, whose 'voice' gives for all the world (and for all its patina of theoretical justification) the impression of a destructive and wilful child; the book of appalled satire indistinguishable from self-satire, supposedly immune from outside insult because the poet insults himself first; the book of in-jokes, designed to attract a great deal of attention from an in-group; the book of poems that double as reportage, and would be of interest as reportage (the grim history of Guam, the travails of coal miners, the life of an airline stewardess) whether or not the poems were labelled as poems; the book whose compressed poems seem to seek the solidity and the humility we associate with solid objects, real things, natural or physical (as opposed to social) facts, which do not require our attention, our consciousness, in order to remain real.

And every one of these suspect modern strategies (which poets may not adopt consciously) gave rise to one or more books that I really enjoy.



Asked to contribute a 'critical credo' (a well-established mini-genre, I discover, dating back to the old Kenyon Review) I realise that I don't know what I believe. Or rather I know what I believe about particular works (poems, books of poetry) and only by generalising, with some trepidation, from those beliefs can I make any guess about what I believe, or do, or know with respect to poetry, or to literature, in general.

Which is to say that either I walk around in an untheorised muddle, or else (as I prefer to say) literary reading, and hence literary criticism, are a habitual, accretive, impressionistic matter, that my habits of expectation and response (like those of earlier generations of readers, from whom mine differ in many particulars) have been built up over many small encounters, many cues and many chances to see how other people (teachers, friends, students, family members) read: these habits make up, for me (as they seem to have made up for earlier generations of readers), a more flexible and more interesting set of ways to respond, consciously and unconsciously, to a new text than do any ways that follow from explicit rules about how to read.

If this account seems 'conservative', or Burkean - being a defence of intricate, partly unconscious, partly un-self-conscious, pre-theoretical traditions and habits - it also permits an account of how those habits change; and it should sound 'conservative' only in the sense that environmentalists and ecologists are also 'conservative', wanting to understand complicated and beautiful systems, and also to keep them around.

I discover that I behave as if I believe:

That there are people, individuals, who have ideas and experiences; these ideas and experiences are subjects of immediate and sustained interest in literary writing. Sometimes they have causal force; sometimes they are epiphenomena, comprehensible only when we see large social or even biological factors that make us whatever we are, that make us do whatever we do.

That there are kinds of poetry, and in particular a kind or super-kind (a set of kinds) called lyric, in which poems represent individuals (so that 'lyric' in general represents 'the person' in general).

That some of the most fascinating poetry of recent decades - especially but not only in the US - attacks, or appears to attack, the two propositions above, claiming that there may be no individuals (that the individual, the person, is always the wrong unit of analysis), or that lyric fails to represent them: this poetry asserts, or seems to assert, that its characteristic forms and modes reflect our uneasy dependence on, even our emergence from, systems much larger and more interesting than we are. And yet we still say that somebody wrote these poems.

That poems should have something personal, and something impersonal, about them, at the same time. 'All that is personal soon rots: it must be packed in ice or salt', wrote W.B. Yeats, implying (correctly) that poets have something personal which they, or we, want to preserve.



Poetry criticism should be easier to write well than other sorts of criticism because the poetry critic can always give the evidence: we can insert a bit of the art itself in the midst of discussions about it, so that our readers can simply compare, as they go, a piece of the art work to our description of it. (Music critics can do the same thing, albeit not in print: on the Internet, or on the radio.)

But poetry criticism should also be impossible: if a poem is any good it should exceed and complicate any statement that you want to make about it - the trick is to say things that are true nevertheless. (If you do not feel that your task is impossible to execute completely then you are doing it wrong, or else you are discussing a very minor poem.)



I am less often afraid of dying than I am afraid that I have already let somebody down; that I have neglected a real responsibility, that somebody I respect and trust will come to see me as a bad person, a slacker, a selfish type, and that this portrait of me will be my fault.

No wonder I choose to devote so much of my energy not to the making of original poetry (that most self-indulgent and supposedly autotelic of linguistic acts) but to the applied and even ancillary arts of literary criticism - describing, giving explanations to, proposing just deserts for, things that someone else has made; I give my time and attention to poems that represent another human being, that can never represent only me.



When I write poetry I reveal (as Kipling, teasingly, put it) 'something of myself': when I write about poetry by G. I reveal something of myself, too, but only what the writings of G. permit me (by contrast, or by echo) to reveal. I may then signal, or disguise, that revelation, as I try to write with constant reference to just what I have found in G.



When you are reading a poem as a critic, try to read it without thinking about who else might read it, who else would like it, who else may have already read it.

Do attend, however, to what sort of people, with what sort of knowledge and what sort of expectations, the poet expects her readers to be.



People who ask what distinctive role poetry plays for us, what distinctive moral, cultural or intellectual gifts poetry alone can bring, are asking impossible questions. 'Poetry', the word, at its most useful, denotes a set of techniques that writers can use and a set of frames that readers can adopt, frames and techniques that enabled Alexander Pope and Louise Glück and Marianne Moore and Pablo Neruda to play their roles, to bring their gifts: these gifts, these roles, share less with one another than each writer shares with other writers who are not poets, and with artists who are not primarily writers.

Poetry critics, as such, are people especially useful, or eager, or talented, in their attention to those frames, those techniques.



When I see the damage that mistaken notions in other fields (in politics, in technology) have done, the harm that apparently innocent ideas (say, lead in petrol, for a smoother ride) have caused, I take comfort in thinking that poetry is at least (as the Hitchhiker's Guide said of our Earth) mostly harmless: to say with Auden that poetry 'makes nothing happen' is to say that in our present state of civilisation it is unlikely to hurt anyone.



J. came to understand contemporary poetry - her own poetry as she wrote it, the poetry of her own contemporaries - through an enthusiastic coterie of American West Coast avant-gardistes: poems had to speak to the systematic bad energies of late capitalism, to undermine the prison of conventional language, and 'self-expression' was for bourgeois losers: under no circumstances should the contemporary poem project a consistent fictive 'I'.

J. has now written an enormous - and brightly readable - memoir, based on her journals and tracking her adolescence, more than 600 pages long; her new book of verse, moreover, addresses and mythologises, with an unironised verve, her life partner and friends, almost in the manner of Shelley or Coleridge.

When you try hard and seriously to expel from your poetry the figuratively personal, the fictive 'I', sooner or later it may return as the literal 'I': as an attempt to present and preserve real people, with real names.



On the other hand, you can try to convey the personal, the fictive 'I', by only apparently impersonal means: through description - I am the sort of person who would notice this and this and this and this and this and this, think them worthy of record (but who would probably overlook that). Elizabeth Bishop made this way of proceeding unmistakable, and obviously useful; R.F. Langley seems to have taken it as far, in verse, as it can go.



Frost/Williams: Whose Era? Critics older than I am grew up asking, or arguing, whether Pound or Stevens ought to dominate whatever passel of tastes and techniques modernism has given us: hence Marjorie Perloff's title 'Pound/Stevens: Whose Era?' But if we want to ask what the modern history of poetic technique can tell us about history in any larger sense the choice lies instead between Robert Frost and William Carlos Williams.

When we read Williams, even at his most pessimistic ('These', for example; 'Impromptu: The Suckers'; 'The Yachts'), the style alone tells us that societies do change, that they can become more (or less) just, and more (or less) fun; that new styles can come into being to reflect those advances (or retreats). His style says, always, that we can solve old problems, or discover new things, together if not collectively, through the shared medium of our changeable language. (The same style, with its rough edges and its reversals, says that Williams himself does not necessarily know what we will discover next.)

When we read Frost the form and tone say the reverse: though we may be Americans in New Hampshire, we live, at best, in the universe of Virgil and of Horace; nothing can change, and nothing improve, from generation to generation. Whatever language we speak, whenever we speak it, the crops grow and die, and we die, alone, and the poet at his most ambitious has to lead us, in code ('so can't get saved, as St. Mark says they mustn't') or in plaintext, to that one truth. We would know as much already, alas, were we (as he is) 'versed in country things': and Frost writes, usually, as the man who already knows - so his finish, his flawlessness, even his sometimes jocular aura, confirm.

When I say that I would like to side with Williams, I do not mean that the weight of historical, nor of literary-historical, evidence has already convinced me to do so.



When we judge works of art we speak, quite rightly, about the value of surprise, of novelty. But there is a delight in finding difficult familiar patterns performed very well: perhaps the art of poetry has as much (or as little) in common with the sport of ice skating as with the work of scientific discovery. And - as with ice skating - the same difficult move may seem new, may hold our attention, may differ enough from the last performance of the same move, simply because we are seeing or hearing a different person perform it.



We hear, still, that one or another artistic move deserves praise and attention because it 'makes us aware of our role in the construction of meaning', or some similar locution; but really we are already aware of that role: beginners are all too aware of it, since they are often asked to find meanings, or effects, that they do not (yet) know how to construct. Sometimes construction is harder than disassembly; it is harder to bake somebody a cake than to give them a recipe, harder to make for them a working bicycle than to hand them a bag of bicycle parts.

We should not ask that all works of art be 'subversive', any more than we should ask of gardeners or farmers that they always plough up and turn over (sub-vert) new ground. (But there would be no crops, and no new flowers, if the people who grow them never turned over the ground.)



I don't want to read arguments about poetry when I can just read poems - but I cannot always read poems. I can read literary criticism, and other sorts of expository and argumentative prose, at the breakfast table, where I cannot read fiction or poems: they would be too absorbing, too distracting - I would neglect the people who eat breakfast with me.



Niedecker contra Zukofsky. I dislike, in principle, though I may admire in the execution, any work of art that attempts to create a small band of devoted and specialised readers, readers who will devote their lives to its untangling. (See Bob Perelman, The Trouble With Genius.) It seems to me arrogant, if not misguided, for an artist (if he has a choice) to ask us to devote so much of our time to himself (and it is usually him-self).

I am, however, happy to advocate work that in practice, for whatever reason - from the difficulty of its sentences, to the obscurity of local references, to the fact of being published by a small or badly distributed press - has only 100 readers at present. Such work may seek attention (don't we all, sometimes?) but it does not aim to monopolise its readers, to take them away from other art, from other ways to see the world, which is after all bigger than even the most ambitious art that attempts to contain it.



I prefer to make claims about single authors, single works of art, by giving evidence and weaving arguments around it: in such cases, from a review-essay to an academic book, sufficient evidence can be given. But I prefer this generalised and fragmentary form when making claims about reading and writing and art in general, in part because I do not know, for such claims, what sort of evidence would suffice.