It is with some relief that we can now turn to an area only really approachable by anecdotal means. In ‘Questions about Angels’, Billy Collins addresses and solves the old scholastic chestnut, ‘how many angels can dance of on the head of a pin?’, concluding that

perhaps the answer is simply one:

one female angel dancing alone in her stocking feet,

a small jazz combo working in the background.

She sways like a branch in the wind, her beautiful

eyes closed, and the tall thin bassist leans over

to glance at his watch because she has been dancing

forever, and now it is very late, even for musicians.1

As fine a solution as this is, it also reminds us that poems are not to be trusted, and their methods almost rigorously unscientific. This is not to say such conclusions are worthless: it just means that a poem’s value tends to reside largely within its own instantiation, and – its use as ‘wisdom literature’ apart – poems have little ‘general applicability’ or predictive force. But this is art all over, of course. It’s a bit like Pierre Wantzel proving that the trisection of the angle using a pair of compasses and a ruler is impossible, only to have some passing Giotto wander into his study, take up a pencil and execute it perfectly, freehand. While that might solve everything, it proves nothing. It’s another trick, a brilliant, useless, one-off performance; but it’s through such performances that poetry nonetheless allows us to glimpse those truths we might otherwise find impossible to apprehend.

Poetry is both trustworthy and untrustworthy: it is a truth-telling, but it often gets at the truth at the expense of the facts, facts being something most poets tend to regard as an inconvenience. However, this slipperiness means that they can sneak things past the real and psychological censors that otherwise would have to be declared: there is a secret lining in poetry’s pack of lies in which it will smuggle in its truth, and the surreptitious methods by which poetry might declare love in peacetime are the same as those by which might inspire rebellion in a time of political oppression. Poetry often ‘flies under the radar’. As Hardy once remarked, ‘If Galileo had said in verse that the world moved, the Inquisition might have left him well alone.’2

The English version of that line of Antonio Porchia’s I placed at the start of this book, and mentioned in the introduction – ‘I know what I have given you; I do not know what you have received’ – was made by W. S. Merwin, and appears in his translation of Porchia’s Voces.3 In my role as a publisher, I was once sent an alternative translation of Porchia’s classic. The translator, a native Spanish speaker, had complained that Merwin had travestied Porchia’s original. I turned to ‘Qué te he dado, lo sé. Qué has recibido, no lo sé’, and found: ‘What I’ve given you, I know. What you’ve received, I don’t know.’

One can see why the translator might have thought this superior: its syntax more accurately follows the Spanish. However, it’s in poorer, less idiomatic and less rhetorically effective English.4 The underlying dynamic of mediocre translation is more clearly seen when you run the procedure continuously, in a game of what Americans call ‘Telephone’ and the British still unfortunately call ‘Chinese Whispers’. I ran the Porchia quote through the (Douglas Adams-inspired) translation program ‘Babelfish’, or rather through its ‘Improbability Translator’. This runs a text through every language in its database in alphabetical order – translating it into Afrikaans, back into English, into Armenian, back into English, and so on, all the way through to Yiddish. Via the likes of ‘I know that I gave you, I do not know what you got’; ‘I know what I want, I do not know what happened’ – the Porchia lines were finally spat out the other end as ‘I do not know what it is.’ (Beginning instead with the Spanish original produced the delightful ‘I think I know. You know, I do not.’)

The process is fundamentally subtractive, and shrinks text. What’s shed in each step is often little more than nuance, but nuance grows into lexical affix, affix into word, word into phrase, and phrase into everything. (Another pass gives ‘I do not know what I was thinking’, another ‘I don’t know what to say’.) Because it’s automated and not human, the procedure also sheds any binding semantic context that would have been maintained either by a sensitivity to meaning, or a commitment to making sense. And this is where the fun starts. Here are the first eight lines of Shakespeare’s Sonnet 18 run through the Improbability Translator:

It hardly improves on the original, and in semantic terms it has utterly discohered – but it does sound like poetry of a late modernist sort, and indeed there are magazines which would happily publish this stuff, especially with the additional fondement of the right name attached. Codal incompetence, otherwise known as ‘not getting it’ – is a big problem with any kind of speech act that tries to communicate accurate information; but in poetry, we are not trying to communicate accurate information. When you turn up the ‘codal incompetence knob’, as we have through this automated procedure, the result can often just produce more poetry. The reason this increasingly random garbage still has the feel of the poetic is because we are used to meeting poems halfway. Poems are half-said things, full of deliberate elisions no normal conversation would countenance. The poet has held this space open for the reader to enter – and what they enter into is an act of co-authorship, in the course of which, if they find the poem ‘meaningful’, they make it their own. This is both the point of poetry, but also why poetry is often ‘difficult to understand’. It would not be a poem if it did not stimulate our capacity for oversignifying and overinterpretation (a capacity which derives from our evolutionarily advantageous trick of reading ‘meaning’ into a universe that possesses none, if you’ll forgive me wheeling out my little hobbyhorse for one more ride). Poetry readers are primed to ‘read too much in’ as part of the contract. Thus any random, unconnected input can send the reader’s connective faculties into overdrive – and that’s a feeling they associate with, and can mistake for, poetry. The domain of poetry is therefore what we might call a ‘permissive context’, unlike the heavily rule-constrained context of other domains like legalese, knitting patterns, or physics. Poetry will legitimise misreadings: the unexpected is expected of it.

To give an analogous example, take some common mishearings of The Beatles’ lyric ‘Lucy In the Sky with Diamonds’. ‘Lucy in the Sky with Diamonds’ has been variously misheard as ‘Lucy and this guy are dying’ / ‘Lucy’s getting high with Linus’ / ‘Lucy in disguise with lions’ / ‘Lick me in the sky with Brian’, and so on. (Perhaps the best ‘mondegreen’ of this song renders ‘the girl with kaleidoscope eyes’ as ‘the girl with colitis goes by’.) The listener already knows the song is surreal and drug-addled and half-crazy; therefore the context will mean their mishearings will be rather wild and unconstrained, when compared with those of, say, the less-bonkers domain of a country and western ballad.

I’ve heard several people remark on the ‘poetry’ of the mad spam text you sometimes get in your email inbox. Interestingly, this is also just a way of smuggling in hidden freight: not of declared love or the seeds of political foment – but a Viagra or Xanax advert, a Ponzi scheme, or a kind offer from a member of Nigerian royalty to rest £10 million in your bank account over the weekend. Here’s an example from my own in-tray:

A quick bit of judicious lineation creates something not half bad that, again, would be eminently publishable in certain quarterlies. But once again it’s the discontinuity that’s poetic – and the work we do in the gaps to forge original links and surprising ideas. As I’ve mentioned, this material is generated using Markov chains, a stochastic algorithm which spits out a sequence of random variables; these can be used to generate ‘plausible’ fake texts from a series of real texts. (The above example is garbled P. G. Wodehouse.) The ‘Post-modern Essay Generator’ uses just such a procedure, and so does the ‘Surrealist Compliment Generator’:

You mutter such objects

of equine delight

that the mind’s ability

to sew slices of mordant ivory

becomes tamed with visions

of Tamils in Constantinople …

These have enough local coherence to fool most search-engine rankings, plagiarism software and spam filters, including the one in our temporal lobes – and enough local discontinuity to be found ‘poetic’, if you’ve a mind to. Since poetry is as much ‘read in’ by the reader as ‘written out’ by the poet, and since ‘reading in’ requires a gap in which to do so – the mere presence of that gap itself is often taken by the reader as a cue to make poetry.

The aim of much poetry is not to ‘communicate clearly’, but to communicate its ambiguous and original signs clearly. Its signs have to have enough silence, emptiness and connotative blur around them to allow the reader to adapt them to their own reading – but not so much that the reader has the sense that they could make nothing or indeed anything of them, which is a game most readers find not worth playing; it tends to just make them feel stupid or lonely. (For years what I enjoyed most in the poetry of Clark Coolidge was, I now realise, my own company.) However, ‘clear ambiguity’ is less a paradox than a balancing act, albeit one made more difficult by the stylistic tics of our late modernist style, which has made a virtue and sometimes a fetish of ellipsis, and only increased the potential for slippage between the poet’s giving and the reader’s reception; contemporary readers are also far too tolerant of the many errors of omission that poets commit in the name of this style. Nonetheless without the gap between giving and receiving, the poem cannot, of course, be given at all.

The ‘Store of Signs’ illustrated here5 is a version of the kind of communication model originally proposed by Karl Bühler and Roman Jakobson.6

In the case of poetry, the main difference is that – unlike jazz or algebra or pig Latin – poetry’s gappy communication means that the code that carries its symbols and signs can change frequently, from poem to poem, or within the poem itself. When the channel switches, new receiving equipment must be acquired; and within that channel itself, there’s often a poor signal-to-noise ratio. Unlike algebra, it’s not just a matter of understanding the conventions of symbolic representation: the reader has to be alert not just to the signs, but ready to learn the new code by which they are to be interpreted.7 Every new thematic domain means a new code, a new set of rules and a new set of semantic priorities. If we were to design a model to reflect poetic communication, not only would our channel-box look like points control on the London Underground, the words incompetence and interference would also be considerably larger.

Codal incompetence is often just not understanding stuff. You don’t enjoy Bartok, because you’re trying to listen to him the way you do Mozart, or you’re trying to read Eliot as you would Kipling. Codal interference is often just not liking stuff: you know exactly what Finnish death metal or abstract expressionism and postmodern verse are doing, but because your preference is for a different set of cultural rules and norms, you’d much rather they did something else. However, a further delicious complication that we find within poetry’s unstable and noisy code is that norm and sign are confused or transposed; the reader will muddle incompetence and interference, and think they don’t get stuff when they just don’t like it – and think they don’t like stuff when they just don’t get it.8 This can produce a reader constantly worried about not getting it – or in the grip of a deeply misplaced confidence that they do get it. (Just check below the fold of any online poetry article, if you need proof that such readers are legion.) This is close to something like a paranoid condition, and indeed ‘mild paranoia’ is often the resting state of the contemporary poetry reader. They either worry they do not have the code, or the means to make sense of the detail it carries – or else they panic, and import the wrong code entirely.

One part-solution lies in clarity of deixis. I often find myself in the fortunate position of being able to simply ask poets what they meant by this or that line. Sometimes the shortfall between what they think they’re giving and what is actually being traded on the page is remarkable. While ‘eliminating ambiguity’ cannot possibly be our reasonable aim, two conditions should nonetheless ideally obtain, unless there is a strong reason for them not to: firstly, the poet should be deictically explicit; secondly, the reader should feel able to take the shortest route to literal sense, i.e. the one for which the text gives clearest cue. If both sides stuck to this agreement, I swear we’d eliminate about 80 per cent of the accidental, pointless and unproductive confusion and difficulty that poems appear to introduce. By ‘productive difficulty’, I mean places where the reader will be rewarded for their interpretative investment – which is to say the very paradoxical, inarticulable, evanescent, obscure or ambiguous things that the poet is wrestling with. By ‘non-productive’, I mean either detaining the reader in the meta-game of trying to work out the rules by which the poem is to be interpreted, or information the poet could easily have just told them. If the poet is trying to move or enlighten a reader, trying to propose an idea that can actually be engaged with, rather than just laboriously parsed, unpacked and then exhaustedly set aside, if they are determined to use the poem as more than an excuse to have a conversation about poetry … Then the poem’s codal channels must be kept as noise-free as possible. Plenty of noise is already introduced by the procedure of poetic composition alone.

‘The shortest route to literal sense’ is generally indicated by deictic cue. In speech, we tend to think of deixis as a means of ‘siting’ a point of view via temporal, spatial and pronominal information: once we have the when, the where and the who of it all, we have constructed a mental space, an origo in which events and relations can be clearly imagined and described in an economical way. As I’ve mentioned, deixis in literature has an additional component, one of ‘literality’. In our living day-to-day, we are aware of a broadly non-negotiable outer reality, and a configurable inner imaginative reality. These are, to all but the psychotic, self-evidently distinct. In the depicted reality of literature, the distinction is maintained – except both the ‘real’ and the ‘unreal’ have to be carefully signed to avoid confusion. The reader is perfectly aware of the mimetic convention that their disbelief should be suspended, and that events will be ‘reported’ by the text just as a speaker might relate a real event. However, this means that the non-literal – the figurative, imaginary, metatextual or metapoetic – must be very carefully indicated.

Too often, though, the poet has forgotten to say that the poem is (a) a conversation between two elephants (b) just after the battle of Jutland (c) in a raft in the middle of the North Sea and (d) that it’s really about their abandonment issues. They have assumed these matters to be somehow self-evident. (I exaggerate, but only a little.9) Poetry is close-in work, and this often involves a considerable loss of perspective. Hence my second war cry: anticipate the state of publication. If a poet is doing this properly it will involve shudders, cold sweats and weeping jags: nothing sharpens our perspective more than the prospect of imminent public humiliation, and publication is primarily an exercise in shame. This allows the poet the editorial distance to see that they must return the crucial detail they had foolishly omitted, and to tell a creative gap from a simple lacuna; otherwise the reader will expend all their energy in establishing dumb stuff that the poet could have just told them, and will have none left for the poem’s weightier or more interesting propositions. All the reader requests is a poem free from the kind of inadvertent obscurity which arises from a failure to provide clear context. Such omissions bump up the noise-to-signal ratio to deafening levels; and if we now add to the mix a bold reader who is prepared to make their interpretation regardless of barely being able to make out a word – the result will be a text which has vastly increased its accidental polysemy (or its ‘entropy’, in information theory terms). This is not the fun thing it sounds. With even the best poets, much of the poetic ambiguity and obscurity we routinely indulge turns out not to be deliberate literary tactic, not the snappy stylistic elisions of the late modernist style – but incompetence, in the form of poorly established deixis.

Problems arise most frequently when the poet changes from one deictic space to another. The pronominal, temporal, locational and ‘reality’ co-ordinates of any imaginative space must be clearly defined, unless there’s a very decent reason not to.10 We should, by default, be clear about dramatis personae, speakers and addressees, and consistent in our use of their pronouns; we need to be consistent in our tenses, and make the chronological order of events explicit; we need to show where things are, and where they are taking place, which means being clear about both location and prepositional relations; and we need to be clear about whether they are actually taking place or not. I have lost count of the number of workshops where all our energy has been spent in working out that the ‘she’ in stanza two now has a different antecedent from the earlier ‘she’, and now refers to the poet’s grandmother; or that ‘we’ now refers to the married couple, and is no longer being used to address the reader; or where the road accident just described is taking place before the event described in the first stanza; or that we are now no longer behind but under the desk, or indeed outdoors, underground, in twelfth-century Tibet, down among the quarks or in a fish-market in Victorian Hartlepool; or that the tea-shop described in the final stanza is not intended literally, but is, in fact, the tea-shop of the soul.11 All this, the poet believes, is either self-evident, or will be picked up through some mysterious osmosis. They are too close to the poem to see that clear cues – or at least decent clues – have not been provided, or have, and are highly misleading. The muddle is often retrospectively justified by the silly assumption that a periphrastic or elliptical style is merely the neutral poetic mode du jour. Recent bad precedents – see the indulgence of Geoffrey Hill’s later extempore and antic obnubilations – have done little to arrest this trend.

If it aims for concision and originality, poetic composition will produce metaphoric and metonymic difficulty as a matter of course. Any complex subject will request a complex response, but as complexity increases, so should our counterbalancing commitment to clarity of statement. Clarity need not mean laborious exposition, and is often just a matter of providing a single well-placed function word; if it costs the poet nothing, they should slap it in. (Traditionally, metrical exigency lay at the root of most dubious elisions, and this certainly drove the common poeticism of omitting grammatical articles; the result often sounds like ‘content bingo’. Poets who work in metre should still be keenly aware of such temptations, and poets who do not should be aware they have lost their only excuse.) Also – by definition, poets are too close to their material. If you have spent a long afternoon agonising whether that en-dash should stay in or out, or weeping over the long-suppressed memory your sonnet has unearthed, or wrestling with some fine metaphysical nuance, or just firing the principle of equivalence into the syntagm until everything seems like one giant, meaningless word … You can quickly lose perspective. Our focal lengths are necessarily short. Obsession – and the overfamiliarity and exhaustion with one’s material that it eventually brings – can lead to a total loss of overview, and great confusion over what is and isn’t self-evident in the text.12

Denied a guiding deictic context, the reader gets lost. The phenomenon of the Lost Reader presents a little simulacrum of the kind of codal incompetence we see in clinical psychosis, and which we can discuss in almost exactly the same terms. While this taxonomy is rather obscure, the nature of these semiotic pathologies will be alarmingly familiar to most readers.13 These states can arise through incompetence and interference on either one or both sides. ‘Lostness’ is a disease of degree: at its most extreme, the poet is essentially speaking Klingon while the reader nods in deep misunderstanding. More often, though, the rules have not been made clear on the writer’s side, so another set of rules have been too-precipitately applied by the reader, and it all ends with one side playing checkers and the other chess. (The ability of humans to wilfully sustain such reciprocal misapprehension is a thing of wonder. I once had a ten-minute conversation on The Prelude with a composer friend of mine; we talked of its longueurs, its bright passages, its turgid lines, its pointless repetitions, its lyric grandeur, its overblown rhetoric. Had it not been for him whistling a passage by way of illustration, it’s unlikely we’d ever have discovered that I’d been talking about Wordsworth’s Prelude while he’d been talking about Wagner’s.)

Given the central role that misreading has in poetry, it’s perhaps surprising we have no working taxonomy to describe it. The following may serve as some notes towards a pathology of the lost reader.

Cryptosemia: identifying signs not apparent to others. One often encounters readers who believe poets have hidden messages in their poems. They have been taught in school that meaning is something poets deliberately and sadistically withhold, and that the poem is something you therefore must torture a confession out of. To give an extreme but well-known example: a great deal of numerological and Kabbalistic frenzy surrounds Shakespeare’s Sonnets. (I recall one assiduous cryptographer uncovered a message in one of the sonnets, which reveals that KIT MARLOWE WROTE THIS. By more or less identical methods, I managed to derive the phrase PARIS HILTON WROTE THIS.) This utter nonsense is often fuelled and sanctioned by the genuine Elizabethan obsession with numerology and codes. But cryptosemia is also the classic hallmark of scholarly readings, where a text – however bad or indifferent – will be obliged to live up to the high expectations of the often superior intelligence reading it: this is the kind of thing Helen Vendler indulges in her theory of ‘Key Words’ and ‘Couplet Ties’ in her commentary on the Sonnets – phenomena which exist, alas, only in the capacious and restive mind of Helen Vendler. While there is often far more to poems than meets the eye, there is just as often far less. Indeed if you take the trouble to read a poem slowly enough and carefully enough, you will find many things that are not there. This is not to say that some poets do not weave hidden messages and infra-texts into their poems (barely a single poem by the late Michael Donaghy did not also have a secret life, and almost every poem – maybe every poem – contained a buried Fibonacci number); but when the reader finds one they must be certain, and the text must then clearly confirm that certainty. However, the principal danger of cryptosemia – especially in its academic flavours – is that the hunt for buried signs quickly overtakes any engagement with the poem itself, which has become a kind of textual inconvenience.

Parasemia: the perverse reading of signs through the invention of false contexts. If the thematic domain is not well formed or indicated, the reader will often just decide the poem is about something else, usually something they’re more interested in; they will then make every detail fit this thesis. While many readers default to old standbys (‘no – this poem about learning the piano is clearly really about death/love/God/the act of writing poetry’ etc.) such projections are often wilfully neurotic or sexual in nature, which is to say there is little the poet can do to deflect them in advance. I once met an individual who was genuinely convinced that in Robert Frost’s ‘Stopping by Woods on a Snowy Evening’, the poet was confessing to having interfered with his pony: ‘He gives his harness bells a shake / to ask if there is some mistake’, and so on.14 An excessively fond little poem of my own, written to commemorate my then-baby son’s first smile, was described by a Daily Mail journalist as ‘a love poem from one gay man to another’. The trouble is that if one line seems to fit the hypothesis well – in this case ‘how fine, I thought, this waking amongst men’ – a slack reader will often issue themselves exceptional licence to make all the others fit too. (I admit that in their new thematic domain, certain lines in my poem took on a delightful new sense: ‘his four-day-old smile dawned on him again’, for example, seems to speak eloquently of the stamina of the author.15)

Hypersemia: the foregrounding of some signs at the expense of others. Otherwise known as ‘strenuous overinterpretation’. This frequently arises through the unconscious desire to apply a specialism (I have had reviews of my own work where the critic has identified nothing but the influence of poets I had barely read – poets on whom the critic was invariably expert). It’s also often an error committed in an attempt to nail ‘what a poem really means’ as opposed to just ‘means’ or ‘means to me’; the word ‘really’ marks the speaker as an essentialist. Meaning is no more intrinsic to a poem than to a word; poems don’t ‘really’ mean anything but what we make of them, exactly as words are no more or less than the way we use them. Poets who are more interested in paradox, tension, the freer play of meaning, or open-ended questions are especially badly served by this approach.

This is Paul Muldoon’s ‘Ireland’ in its entirety:

This poem asks us to look at the same image – the parked car with the engine still running – in the two ways that we might think of ‘Ireland’ (the title doing a splendidly clear job of eliminating other frames of interpretation). The last two lines are metonymic and evidential, and supremely economical. They tell us there was something urgent and likely terrible to be done, quickly, at very a specific distance from the car; the map-like precision of ‘two fields and a river’ indicates that there had probably been some reconnaissance involved, meaning the men likely took the trouble to enter through the back door of the house rather than the front, for the sake of surprise; and even if that hypothesis is ‘wrong’, that car ticking like a bomb (an inevitable association, given the context of the title, and the historical period confirmed by the last detail) suggests that whatever it was these men were up to – the detonation of a roadside device also fits – it was the opposite of a young couple nipping into the wood or ducking into the back seat to make love. This seems to me the most sensible line to pursue, ‘the most direct route to literal sense’, given the information we have and the context (provided by the firm anchorage of the poem’s title) which gives it its larger meaning. The poem itself forms a single coherent peristasis, from which the abstract qualities ‘irresolvable doubleness’, ‘instability’ and ‘paradox’ rise to indicate a rich animus, leaving the whole poem a symbol of ‘Ireland in all its contradictory terror and romance’. However, encouraged by the title, more than one person has told me that the ‘two fields and a river’ are ‘symbolic’ of Ireland, the Irish sea and the UK mainland, making the whole poem an allegory. But that’s a wretchedly poor and totally unproductive reading: the poem is about two interpretations held in tension, a double-take, and such a reading leaves the poem overwhelmed by merely one – setting aside the fact of its making no sense of the poem’s composite details. Overinterpretation can take many forms. Readerly incompetence can arise through excited, wilful projection, disproportionate to its textual cue; it can also be the product of overweening confidence, exhaustion, and what we are sometimes obliged to call stupidity.17

Hyposemia: diminishing the significance of important signs. In other words, under-interpretation. One classic example: those many readers who take Robert Frost’s ‘The Road Not Taken’ as an inspirational tract, a poem of bold, inspired, against-the-grain individualism. ‘The Road Not Taken’ is quoted by self-help gurus, recited by high school valedictorians and cannibalised by car advertisements; in China, the poem is taught in schools as a political allegory about turning away from the path of Western capitalism. Heavens, do they all misread it. In their defence, Frost deliberately made a poem to trip the unwary. Yes, he says ‘Two roads diverged in a wood, and I – / I took the one less traveled by, / And that has made all the difference.’18 But nowhere does he say that the difference made was a positive one. Significantly, one often hears the title misquoted as the far bolder ‘The Road Less Travelled’; but the correct title indicates the poem is about the path Frost did not take. To my ear that makes the ‘sigh’ far less one of wistful self-congratulation, and (knowing ‘the domain of Frost’ as we do) likely one of exhaustion, nihilistic indifference, or even bleak regret. Another, equally plausible interpretation exists or co-exists, and it’s just as miserable: the choice made no damn difference whatsoever. (The clue is ‘And both that morning equally lay / In leaves no step had trodden black’.) In this reading, what Frost predicts he will be telling ‘… with a sigh / Somewhere ages and ages hence’ is the self-deluded lie of old men: that their life was a noble journey in which they had been masters of their own destiny. Just read out the last stanza in a sarcastic voice, and you’ll hear it. Either way, Frost knew all about the cheerful errors committed by the glass-half-fullers, and he toyed with them mercilessly.

Asemia: the inability to understand any signs at all. J. H. Prynne’s poetry is either ‘notoriously’ or neutrally discontinuous, depending on your taste. Personally speaking, I now enjoy it rather more than I used to, mainly because I realised that it wasn’t quite as hard as I thought. I used to call such poetry ‘difficult’, but these days tend to think of it as merely ‘bracing’. The fault was both his and mine. Prynne has been notoriously silent on the subject of his own work, but recently more forthcoming:

In short: if you give up on trying to find a paraphrasable meaning in Prynne’s poetry, you can give yourself over to the freer play of sense, sound, intertextual association, and so on; it is text written – to use Louise M. Rosenblatt’s distinction – for an ‘aesthetic’ not an ‘efferent’, information-extracting reading. At that point, I moved from a self-diagnosis of codal incompetence (otherwise known as ‘feeling stupid’) to codal interference, where I was in possession of the correct code and therefore could understand what was going on – but was not necessarily obliged to like it, although I found immediately that I rather did. Admittedly, it might have helped if he’d said all this earlier. (More telling, perhaps, is the total absence of hand-wringing from Prynne scholars over what Prynne himself has apparently dismissed as forty years of misguided and pointless exegesis. Perhaps they think he was joking: I don’t.) Then again there are many occasions when we should merely blame the poet for holding out the promise of sense or thematic coherence (a clear title, a clear initial proposition, a handful of propositions that are obviously connected) and then not making even ‘tonal’ sense, which is the last kind of coherence such a poem can claim. (The habitual elisions, ‘jargon collages’ and hyper-intertextuality of late modernism can easily create poetry that makes intelligent people feel like idiots; if that is the intention, this is an aim that can be sincerely pursued only by the intellectually insecure. Either way, the situation would be improved by some in-house critique that possessed something like a language of disapproval.)20

Dysemia and eusemia: reading signs too negatively or positively; for our purposes, the unjustified projection of negative or positive qualities. These represent the two principal types of critical distortion: the negative assessment of work of self-evident literary merit (while you will never meet a poet who has not suffered as a result of this monstrous perversion, there are reasons to believe the disease is not as widespread as is claimed), and the cheerful projection of the quality of literary merit into work which appears to demonstrate no evidence of it. Both are impossible to prove, but no one doubts their existence. Eusemia is the most prevalent kind of contemporary misreading, and is probably less a reflection of the loss of shared literary values than it is of a politically intermediate age. As with dysemia, it can too easily be dismissed as ‘all a matter of taste’. Nonetheless I have lost track of the number of times I have heard judges on literary panels declare that (what seems to me) self-evidently bad book X is intrinsically and yet inexpressibly superior to self-evidently good book Y, when it’s clear that this assessment is the product of an extraliterary agenda, usually with its origins in a political or personal bias. If they are aware of this, the judges are corrupt. If, as is far more often the case, they are honest and unaware, they are merely hosting an opinion they feel they should have, and reading the book on behalf of an imaginary reader whose standards are superior to their own and to whose taste and judgement they unconsciously defer. There may be any number of factors at the root of this, from intellectual insecurity to tribal or sectional allegiance.21 (The only unforgivable mistake is in declaring yourself free from all such bias, and wholly without blind spots; certainly, I could have been – and indeed may well have been – dead wrong in every case.) ‘Literary merit’, it should go without saying, is never intrinsic; we have nothing to work with but critical consensus and such expertise as we’re inclined to grant one another. But any critic who declares a literary work self-evidently good or bad and yet cannot articulate why, beyond the adduction of their own strength of feeling – should never be conceded any expertise, however reliable their taste may later prove.

Often the easiest way for a poet to win a prize or an endorsement is for the poet to have already won a prize or an endorsement, since this provides the authoritative recommendation an unconfident judge needs to sway their vote; there’s every chance that the earlier prize was secured in the same manner. (I can think of at least one occasion where a poetry prize was awarded to a book no one on the panel had read, simply on the basis that its author was ‘a major figure’.) I don’t doubt that I’ve personally been the beneficiary of such a chain of good fortune. And I suspect that many unaccountably fortunate careers can be traced back to a single favour or piece of luck, one which eventually led to the most advantageous situation a writer can enjoy: to have your book thought well of before the spine has been cracked. There are many more poets who never got the break they needed and deserved.

1 Billy Collins, Questions About Angels (Pittsburgh, PA: University of Pittsburgh Press, 1991).

2 Thomas Hardy, quoted in Florence Hardy, The Life of Thomas Hardy (London: Studio Editions, 1994).

3 Antonio Porchia, Voices, trans. W. S. Merwin (Port Townsend, WA: Copper Canyon Press, 1994).

4 Its weakness is due, in part, to the position of the nuclear stress (NS) in each phrase. In English, NS – unless there’s a overdetermining context involved – tends to land on the stressed vowel of the last content word of the phrase unit. In Merwin’s version, this gives us a rise on given and received; this additional stress makes the antithesis even more stark, and the fact that Merwin has made this already parallel structure heavily metrical (it’s almost clean iambic tetrameter) means that the alternate stress rule (ASR) also comes strongly into play. The ASR works hierarchically over larger units too, but there it becomes an intonational phenomenon: this leaves a fall in pitch between the two nuclear stresses, given to received, which also – to my ear anyway – lightly enacts the giving-receiving synaesthetically. In the new translation, the speaker’s know and don’t know are promoted via NS to the salient rhetorical contrast. But the problem is that readers care about themselves, the you, not the poet; they are far more interested in what they’re getting or not getting than they are the speaker’s relative state of ignorance. Thus Merwin’s translation holds our interest because it directly engages us. Rather ironically, the second translator probably knew what he had received, alright; but he didn’t know what he had to give, and therefore ended up keeping it. After the simple error of unidiomatic ‘translationese’, this kind of codal incompetence is the most common hallmark of poor translation.

5 This is adapted from Philip Tagg, Music’s Meanings (Larchmont, NY: The Mass Media Music Scholar’s Press, 2012). Tagg’s book is a fine introduction to the semiotics of music, and should be eagerly sought out by anyone interested in the subject.

6 Roman Jakobson, ‘Linguistics and Poetics’, in Style in Language, ed. T. Sebeok (Cambridge, MA: MIT Press, 1960), 350–77.

7 A singular code is often, rightly or wrongly, taken as the hallmark of a great style or a great talent. See Wordsworth’s ‘Never forget what I believe was observed to you by Coleridge, that every great and original writer, in proportion as he is great and original, must himself create the taste by which he is to be relished.’ (William Wordsworth, The Letters of William Wordsworth, ed. Alan G. Hill [Oxford: Oxford University Press, 1985].) This statement should perhaps be more notorious than it is, given the comfort it has provided those writers who read ‘failure’ as ‘unjust neglect’, and who hurl no larger criticism at themselves than being too louche to have gotten round to cultivating the means by which their greatness and originality can be appreciated. There is a small, horribly tormented cohort of the living and the dead who go by ‘unjustly neglected’, of course – and it is the responsibility of the community of readers, critics and successful writers to make sure it stays as small as possible.

8 It’s possible to derive a little semiotic square here, which might be fun for someone to run with: this would add two terms to incompetence and interference: delusion, or ‘thinking you get it when you don’t’; and denial, or ‘thinking you like it when you don’t’.

9 This is why God gave us the title. Western convention dictates that titles will be read, invariably, as ‘what the thing under this is about’, so we might as well use them that way. This not only offers a lot of opportunity for subversion, but also simple delegation. A poem called ‘Death’ which merely recounts a dull business lunch has no need to ever mention the morbidity it inspired in the poet. A simple, straightforward title often relieves the poem of the responsibility of having to mention the matter again – meaning that, say, six lines of clumsy deictic exposition about our seafaring Dumbos can be lifted clean from the poem ‘The Insecure Pachyderm in Naval History’. Indeed, a well-chosen title can halve the length of some poems, as readers are more than happy to make up the distance between title and poem, and will more often than not treat it as a creative lacuna in which they can develop an inter-domain ground where they see a shared aspect or property, or an intra-domain link where they spot a direct relation. Another effective technique sees the promotion of a detail in the poem to title-status, leading to its immediate symbolicity and ‘totemisation’; see, for example, Roger Mitchell’s ‘The White Cup’ in Lemon Peeled the Moment Before: New and Selected Poems, 1967–2008 (Port Townsend, WA: Ausable Press, 2008).

Incidentally, we do not need to understand the meaning of a symbol to know we’re probably dealing with one. This technique often creates a kind of desire-vector where the sense of the poem is structured around the attractive force of one partially understood, semi-symbolic detail, one which overlaps with the concepts of both McGuffin and objet petit a.

10 It’s poetry, so often there is a good reason. But the next time a workshop poem generates ten mutually exclusive interpretations, and I hear the poet pipe up with ‘Oh – I really don’t mind if this poem/line/detail is read both ways/different ways/any old way’ when it was perfectly clear that they wrote it with one specific intention, it will trigger my version of Cúchulainn’s warp-spasm. One should mind. Judicious ambiguity or deliberate instability is one thing; self-delighted incompetence is another. The modernist aesthetic – whose elisions and discontinuities dramatically increased the expressive range of the poetic art – should never be the flag of convenience for these rudderless, leaky rustbuckets of poems.

11 Billy Collins’s ‘Workshop’ – a self-descriptive poem written in the voice of an exasperated poetry tutor – is a handy lexicon of deictic errors: ‘First, we’re in this big aerodrome / and the speaker is inspecting a row of dirigibles, / which makes me think this could be a dream. / Then he takes us into his garden, / the part with the dahlias and the coiling hose, / though that’s nice, the coiling hose, / but then I’m not sure where we’re supposed to be. / The rain and the mint green light, that makes it feel outdoors, but what about this wallpaper? / Or is it a kind of indoor cemetery? There’s something about death going on here. / In fact, I start to wonder if what we have here is really two poems, or three, or four, / or possibly none.’ Billy Collins, The Art of Drowning (Pittsburgh, PA: University of Pittsburgh Press, 1995).

12 While I’d maintain that ‘anticipating the state of publication’ is a great cure for lack of editorial distance, there are others, too: the community of poets exists not just as support network for a calling whose principal rewards are mental illness, poverty, broken relationships and substance abuse, but as a ready source of honest editorial perspective. Poets dispense with that community – however late they are in their careers – at their own peril. Who knows how many bad poems or bad books we might have been spared if their authors had had one friend prepared to tell them the truth?

The most common mid-to-late career downturns – especially, one grimly notes, those of male poets – seem to involve either a glum segue from poetry into mere ‘verse’, a descent into unwitting self-impersonation, or some knock-on consequence of their total loss of libido. Many poets would be better shutting up until death is visible on the horizon, something which reliably triggers a late lyric efflorescence. But most poets go on repeating themselves, while seeing in every microscopic advance a radical new direction. (The most exquisite put-down I ever heard was from the mouth of the great Gael Sorley Maclean, a man of few words, but also one possessed of a dry, sly wit. I recall he was being interviewed in his garden on Raasay for a TV programme about George Mackay Brown. The interviewer asked him to give his estimation of Mackay Brown’s literary merit. Sorley rested on his hoe a long while, and then said ‘Lovely poem.’ ‘What do you mean?’ said the interviewer. Sorley raised his eyes to the horizon. ‘Oh; that poem he writes.’)

13 Some of the terms I’ll employ are borrowed from a paper on sign malfunction in psychotic reality. I don’t pretend that this is a complete taxonomy, though I believe one could probably be derived from the four variables of transmissive and receptive codal interference, and transmissive and receptive codal incompetence. The result would be something like a contemporary version of I. A. Richard’s list of the ‘chief difficulties’ of reading. (These – while still relevant – are anecdotal, unsystematic and pretty much useless from an analytical perspective.) Stepan Davtian and Tatyana Chernigovskaya, ‘Psychiatry in Free Fall: In Pursuit of a Semiotic Foothold’, Sign Systems Studies 31, no. 2 (2003): 533–46.

14 Robert Frost, ‘Stopping by Woods on a Snowy Evening’, The Collected Poems (London: Vintage Classics, 2013).

15 Don Paterson, ‘Waking with Russell’, Landing Light (London: Faber & Faber, 2003).

16 Paul Muldoon, Poems 1968–2014 (London: Faber & Faber, 2016).

17 I once judged a poetry competition back in the days when no one used sifters. I was totally page-blind after reading 5,000 entries, could not find a clear winner, and had almost given up. At 4 a.m. I came across a poem called ‘To My Dog Benjy Who Died Under a Landrover Aged Three Years*’. The asterisk led me to a footnote which read ‘Benjy was a cocker spaniel’. In the morning I found that I’d written the words ‘harrowing’ and ‘top 3 at least’ in the margin, before I’d fallen into a coma at the desk.

18 Robert Frost, ‘The Road Not Taken’, The Collected Poems (London: Vintage Classics, 2013).

19 J. H. Prynne, ‘Mental Ears and Poetic Work’, Chicago Review 55, no. 1 (2010).

20 Occasionally one hears accusations of a failure to ‘innovate’; though this is a non-neutral criticism only if you feel that ‘innovative verse’ is a reasonable synonym for avant-garde practice, in which case it amounts to the betrayal of the cause. However, with the exception of a very few commentators – Peter Riley significant among them in the UK – neo-modern or postmodern critics tend to restrict themselves to neutral exegesis and extravagant praise. Anyone in the mood for a graceless, badly timed and poorly judged rant on the subject might want to skim my introduction to New British Poetry, which I edited jointly with Charles Simic (Minneapolis, MN: Graywolf Press, 2004). It was a fractious time in UK poetry, and I was much too taken up with what seemed the avant-garde’s constant abuse of ‘the mainstream’. (I was also too influenced by the late Michael Donaghy, who had become morbidly obsessed with the typographically challenged barbarians at the gate – while occasionally offering them poems for publication, by various alter egos. These heteronyms included the immaculately opaque ampersandeuse Helene-Marie Journod, who wrote acrostic sonnets spelling out things like THIS IS BULLSHIT, all composed with a egg-timer. It was no one’s finest hour.) While it will always be fun to bait the humourless and the ideologues – one should, on principle – my own position is much more live-and-let-live these days; I sense a similar tolerance emerging from the left bank, and the general feeling is one of shaky rapprochement. This has been due largely to the wiser influence of a younger generation who had no idea what we were all fighting about, nor could care less.

21 We must never lose sight of the fact that the great ‘eusemic’ problem over the last couple of thousand years has been the assumed natural superiority of a lot of dreadful white male writing. I think we should all be pretty breezy about the pendulum swinging in a different direction for a while; given that some mediocre poetry will always be unjustly overpraised – better, surely, that it’s the poetry of a more diverse constituency; getting things wrong is all part of serious critical attention. While we should not imagine that historical injustices can be addressed through contemporary correctives, we can at least ask that where ‘generosities are extended’ it’s towards those writers previously denied them.