1. A rule of five

This is not a book which discusses ‘form’ in anything other than a strictly metrical sense, but – at the risk of appearing even more of a crank – I have found the following little table useful in maintaining distinctions of formal procedure and structure. (It’s derived, in part, from Vedantic nyaya traditions.) ‘Structure’ is generally far simpler in poetry than in other literary forms; our sophistications lie elsewhere. While the list below presents idealised categories, real-world poems often switch between these strategies freely, or nest one within another. The categories themselves can be interpreted in a great number of ways: they can be read as analogous to the operations of physical law, or those of the conscious and unconscious mind; or as descriptive of various logical, philosophical and literary procedures. The scale itself might be interpreted as (a) the progressive corruption of atemporal unity by temporal sequence; (b) the progression of an absolute singularity towards absolute dispersal, driven by entropy; (c) the non-sequential, symmetric unconscious moving towards sequential, asymmetric consciousness; (d) a progression from absolute innocence to an absolute skepsis, an absolute interconnectedness to absolute disconnectedness, or an absolute stasis to an absolute Heraclitean flux. There is a very odd (and likely coincidental) scalar correspondence to the perception of sound-events, which I will also include below; since hearing is the sense which measures time most accurately, it seems a relevant inclusion.

  1. No A, no B: unity. There are no separable elements to be compared. Perceived simultaneity of auditory events (below 0.003 seconds apart). Continuous mode. No paradigm or syntagm. No temporal or causal sequence. Via positiva. Positively defined strategies. Pure ‘image’ poems; free-standing and deliberately inconclusive meditations; ‘epiphanic’ forms, such as haiku.
  2. A is B: symmetry. Elements compared for similarity only. Perceived separation but not sequence of auditory events (above 0.003 seconds apart). Litanic mode. Strongly paradigmatic. Sequence without temporal or causal relation. List-poems; prayers; curses; incantation; dithyramb; blazon, etc.
  3. A both is and is not B: correspondence. Elements compared for similarity and dissimilarity. Perceived sequence of auditory events (above 0.03 seconds apart). Argumentative mode. Paradigmatic and syntagmatic. Causal relation and sequence. Argument; conceit; syllogism; intertextuality; dialectic; eclogue, etc.
  4. A is not B: asymmetry. Elements compared for dissimilarity only. Auditory response time to sound-events (i.e. time in which non-passive, asymmetrical reaction can be made): 0.3 seconds. Narrative mode. Strongly syntagmatic. Temporal and causal sequence. Narrative forms; dramatic monologue; epic; anecdote.
  5. A neither is nor is not B: disunity. Elements are not compared. Limit of perceptible rhythm: i.e. sound events repeated 2–3 seconds apart. Discontinuous mode. Neither paradigmatic nor syntagmatic. Indifference to temporal or causal sequence. Via negativa. Negatively defined strategies. Surrealism; collage; cento; cut-up; postmodern or experimental discontinuous practices.

2. Symmetry and generalisation in the unconscious

Ignacio Matte Blanco proposed that the unconscious was structured by two principles. Firstly, the principle of generalisation, which states that ‘Unconscious logic does not take account of individuals as such; it deals with them only as members of classes, and of classes of classes’; and secondly, the principle of symmetry: ‘the Unconscious can treat the converse of any relation as identical to it; that is, it deals with relationships as symmetrical’.1 Unlike the workings of the conscious ego, the unconscious deals with neither instance (it only understands sets) nor asymmetrical relation. If a relation exists, its direction is reversible and its agents interchangeable. If differences between things cannot be acknowledged, there can be neither chronology nor causality, since sequence requires differentiation. In the unconscious, what X did to Y, Y did to X; if I hate you, someone or something also hates me; the cat knocked over the vase both before and after it smashed; what happened yesterday is also happening today, and will happen tomorrow. The whole business will be fully addressed in the second part of the book, but I must make this point here: I feel the principle of generalisation erodes the paradigmatic ‘axis of selection’, which depends on family resemblances between stored items or ‘individuals’, while the principle of symmetry erodes the syntagmatic ‘axis of combination’, which depends upon rulebased category and sequence. Now: if the unconscious functions as a serious influence in the composition of poetry, Matte Blanco may have partly explained the psychological origins of Roman Jakobson’s holy formula, ‘the poetic function projects the principle of equivalence from the axis of selection into the axis of combination’. My position, which I will explain in due course, is that the principle of rule is also projected from the axis of combination into the axis of selection: the poetic function is really a partial collapse and conflation of both axes. While it is probably unfalsifiable, Matte Blanco’s hypothesis would explain this perfectly. This kind of thinking has, however, fallen out of fashion, and recent neuroscience seems to contradict both the clear-cut distinction between conscious and unconscious (consciousness seems more a matter of specific focus, and of our awareness of our awareness), and the idea of a childhood as an atemporal and indivisible Eden. But the fact that we continue to hark back to it as such – that we often see ‘child-like presence’ as an artistic or meditational corrective to the jagged discontinuities of the conceptualising machine – strikes me as almost as important.

3. Noise and meaning

White noise and brown noise can also be thought of as, respectively, the universe’s paradigmatic and syntagmatic forces in their virgin state, i.e. when one is entirely unmitigated by the operation of the other. In the first case, all terms are interchangeable, and meaninglessly correspond; in the second, all terms are arranged by inflexible rule, and meaninglessly relate. ‘Meaning’, or ‘meaningful arrangement’, can be considered a human-projected consequence of the interrelation of both axes; or, better, a bootstrapping effect where a natural organism whose brain is constructed by the interrelation of both axes projects itself into a natural habitat which has evolved from the same principle: ‘meaning’ is a closed-system feedback process. When this stable, complex ‘pink’ system (i.e. ‘us’) projects itself into ‘pink’ features of nature, ‘meaning’ is equivalent to ‘intentionality’.

‘White’ = lexis, ‘brown’ = grammar is another fun line to pursue, and might lead us to find poetry ‘a browner shade of pink’ than regular speech, given its fondness for rule-based structures which correlate its parts (among which I would also include the subservience of its elements to the overdetermined rules of the ‘thematic domain’, explored elsewhere). This appears to be an opposing tendency to that identified by Jakobson, in that it is a strengthening of the axis of combination, but it might merely be a natural way of compensating for the projection of parallel arrangement across the syntagm. In other words, the rules of the entire domain are overdetermined in an attempt to limit the homogenising effect of the paradigmatic.

Incidentally, there are many ‘colours’ of noise described in acoustics and signal processing; my favourite is that unofficial racket known as ‘green noise’, which roughly corresponds to the sound of the world when you open the window: a low-level pink noise with an emphasis around 500Hz (just below C above middle C; the earth blows a little flat).

4. Computers as composers

As a result of the telescoping of the serial process of composition in computer-based music, all the analogue noise and error of us that we call ‘humanity’ has been moved to a different part of the compositional chain, and shifted from its traditionally frontloaded position (i.e. the recording of live performance); the fact that noise, delay, rubato, mistiming, timbral variation, error, slippage and goof are now deliberately inserted as part of a compositional procedure – and not left to performative expression and incompetence – appears to have affected the ‘humanity’ not a jot. Paradoxically, the necessity of ‘programming in’ the humanity has led to a bizarre expertise in the expressive exaggeration of human error, a music often full of glitch and crackle, staggered and staggering beats, over-hit snares and under-hit hi-hats. Computer-based music is already in its Romantic phase. By contrast, very few poets would dare to write ‘badly’, lest the gesture be misread as simple ineptitude. This ‘constraint of demonstrable competence’ is something one finds only in overdemocratised art forms: in poetry’s case, the bar has been set very low by its admitting poets who could not write, poets from whom the rest would like to clearly distinguish themselves. The exceptions are pastiche or satire (see, for example, Peter Reading’s Stet), or the poetry of those too smart, literate and self-confident to care (see Louis MacNeice).

Generative music programming – essentially the art of judiciously modulating ‘white’ aleatoric data with some ‘brown’ grammatical rules to produce something ‘pink’ – is fairly close to providing some very listenable music, and reminds us that Bach’s genius was computational as well as inspirational. Poetry is a vastly more complex business, because its listable parameters are far more numerous, and its signs too quickly destabilised; but not – who would seriously claim this – because it captures any more of the human spirit than does music. Current efforts at generative poetry are still aspiring to the merely daft, alas, and most depend upon stochastic algorithms to produce surreal effects of ‘the fatuous banana spliced my unwarranted windmill’ variety, which appear ‘poetic’ only because nonsensical linguistic input overstimulates our connecting faculties – a feeling we reasonably might associate with poetry, but would be wise not to confuse with it. Poets, needless to say, have had nothing to do with their programming. An algorithm for poetry would be incredibly complex, but not infinitely so; and its detachment from such catastrophically overvalued and sentimental constructs as ‘the individual voice’ (the one all young writers are presently required to ‘find’) could be just the thing to see the cultural return of anon., or even propel us into a new era of Classicism, should we desire or require such things. However, a fact too rarely pointed out is that a computer is going to have to be able to knock out a decent after-dinner speech, a piece of good journalism and a cracking short story before it gets anywhere near a great poem, which will be the very last thing it produces.

5. ’Allo allophones

Foreign allophones are the source of much mildly racist humour, but also the hinge on which many fatal shibboleths have turned. In the South-East Asian theatre in World War II, anyone approaching a US army checkpoint would be asked to say ‘lollapalooza’, which the Japanese l/r allophone makes very difficult to accommodate: anything close to ‘rorraparooza’ would be met with a burst of gunfire. Allophonic and ‘missing phone’ shibboleths have decided the fates of stuttering innocents throughout human history, from the Ephraimites who couldn’t say the word for ‘flood stream’ (the original ‘shibboleth’, though it’s sometimes translated as ‘ear of corn’) to the Dominican Haitians who couldn’t say ‘parsley’ (the perejil of Rafael Trujillo’s horrible ‘Parsley Massacre’), right up to both Catholic and Protestant Northern Irish: one way or the other, your gunpoint recitation of the alphabet would stop at the eighth letter, since either haitch or aitch would declare your schooling, and therefore your sectarian allegiance.

To return to mild racism: I once heard of a jazz promoter who put together a tour of Japan featuring three guitarists chosen for no other reason than requiring that Japanese MCs announce ‘Larry Coryell, Emily Remler and Bireli Lagrene’. A friend recounts an uncomfortable encounter with the p/f allophone in Filipino English: passing a garden party thrown by Imelda Marcos, she was treated to the hostess’s rendition of the well-known Barbra Streisand number, ‘Peelings’. But the problem is that smug Anglophones forget it cuts both ways. English ‘t’, for example, is an allophone of six different sounds, separately differentiated in other languages. English allophones produce much Turkish and Mandarin hilarity, I gather, particularly in our (purely psychological) conflation of aspirated and unaspirated consonants. Even though we use both sounds, we have tuned out their differences so effectively we’re quite unaware of them. Try saying ‘nitrate’ and ‘night rate’: you have to hold a candle-flame to your lips to make yourself aware of the clear difference in the articulation of the central /t/.

6. Present and absent zebras

If we set aside the arguments for its part-iconicity – the word ‘zebra’ is an arbitrary sign for both our real and conceptual zebra. However, let me offend the intelligence of every semiotician in the room and suggest that the word functions indexically when the zebra is visible or present to us, and symbolically when it isn’t, designating the first and invoking the second. The ontological change in status of the referent from present to absent is hardly, I’d propose, a small matter; it entails a change in the function of the name from intra-domain to inter-domain – in this case, from one that works within the domain of a real and present world, to one that works between that real world (if we assume speech itself has a real-world presence) and its memorised simulacrum. I appreciate that this is an argument for the word ‘table’ in ‘put it on that table’ not having exactly the same referent as ‘take it through to the table in the next room’ – but I suspect that, given the opportunity, present-table trumps concept-table in a way that goes well beyond mere deictic convention, and that we simply neither think of nor use the words in quite the same way. (It also proposes that conceptualisation is very much a disease of degree. The table in the other room must be ‘imagined’, since it’s not visible; though not to the extent that we have to ‘imagine’ a table that hasn’t even been manufactured or purchased yet, but which we might still discuss as existing.) Nothing will settle this but an MRI scan, of course.

Our ability to hold the known world in our heads when it isn’t physically present is unbelievably cool, but the facility came at a price: the consequent superimposition of ‘the known’ over the real world deepens the sense of all-pervasive dream, and diminishes our sense of our own real presence within that world. (Hence the parlous state of ‘the’ environment; the definite article stands before a pernicious misconception regarding our relationship to it.) It is precisely this superimposition that poetry alerts us to, undermines, punctures, and ‘gets between’. Language is the main way in which we express, codify and consolidate what we hold to be the case; the poetic function is language’s built-in epistemological tool for challenging the truth of our casual certainties and broad assumptions wherever we sense them to be incorrect. Were language able to function only indexically, there would be no need for such a function; its symbolicity, however, demands it.

7. Overtones

In audio signal processing there is a procedure called ‘additive synthesis’. It has its origins in the nineteenth-century French physicist Joseph Fourier’s work on the heat equation. The Fourier series he subsequently developed decomposes the periodic functions of complex waveforms into an infinite series of simple sines and cosines. This led – much later – to the discovery that any sound, no matter how complex, could be replicated merely by stacking up sine waves (think of the pure tone you get from a pipe organ) at different powers, at frequencies corresponding to the partials of the harmonic overtone series measured from a fundamental or root tone. Nowadays it’s possible to resynthesise even a complex musical sound – a violin note, a human voice, or a whole piece of music – and convert it to an additive series of sine waves. Computers allow us to visualise the whole thing spectrally, and ‘see’ how the sound is composed – as well as cut into it at any point, and change its inner composition as drastically as we like. The business of semantic connotation is central to how poetry works. In poetry, one is very conscious of not just the ‘meanings of words’, but also their ‘overtones’ – those secondary associations in the more peripheral regions of a word’s conceptual domain, which are constrained, shaped and boosted by the context of the poetic line and the thematic domain as a whole. Even though I know the exercise is quite impossible, it strikes me that if you could ‘resynthesise’ a word’s site-specific contextual meaning, you could visualise it as a unique semantic spectrograph, where connotative overtones would be expressed at various strengths over the harmonic fundamental of its denotative sense. Thus in a poem – to use my favourite go-to noun – ‘moon’ might denote the heavenly satellite, but also carry aspects of whiteness, roundness, light-carrying, tide-pulling, coldness, remoteness, barrenness – at various amplified or diminished levels, depending on its various semantic, phrasal, lyric, syntactic and thematic circumstances. (The word’s changing sense over time – which occurs through its subjective rereading and subsequent recontextualisation – maps beautifully to the more complex procedure of inharmonic additive synthesis, an analogy the reader will be relieved to know I do not intend to explore here.)

8. In defence of Lakoff

Patrick Errington, a St Andrews PhD candidate with whom I have argued long and hard over the chapter on ‘The Conceptual Domain’, feels the critical consensus is swinging back towards Lakoff, and writes to me (I have omitted most of his copious references; my italics):

Numerous brain-imaging studies of linguistic and conceptual processing strongly suggest that word- and sentence-understanding recruits sensorimotor brain areas, particularly the motor and pre-motor cortices in action-verb processing, which, to my mind, throws something of a wrench in this reading of Pinker’s theory. A general hypothesis, deeply tied to studies of what has been dubbed the ‘mirror system’, suggests that comprehension involves neural simulation of events and states that the words provoke (and not just as a side-effect), as well as simulation of the action of producing the words themselves in both speaking and writing. A recent survey of such studies by Giovanni Buccino et al. leads them to suggest that noun, adjective, and adverb comprehension recruits motor/premotor simulation as well, potentially due to interactive ‘affordances’ encoded by the hypothesised ‘canonical neurons’ as well as emotional reflexes, which implies that meaning is based in collections of real-world experiences. Additionally, other research has shown fairly consistently that action words in metaphors (e.g. ‘he grasped the situation’) do recruit sensorimotor simulation in the processing of their meaning, though generally seem to show a diminishing level of activation inversely correlated to the subject’s familiarity with the metaphor. This doesn’t undermine Lakoff and Johnson’s theory, but supports one based on a variation of Rachel Giora’s ‘Graded Salience Hypothesis’, wherein I would suggest that novel metaphors (as all metaphors are, at one point, even our commonplace ‘conceptual metaphors’) recruit a wider or deeper spectrum of the sensorimotor systems to process it – though, over time, and with increased familiarity, the extent of the activation is diminished (there is research that may show a shift in activation from mostly right hemisphere to mostly left), with novel metaphors recruiting and eliciting greater emotional responses. Importantly, novel variations of familiar conceptual metaphors can still be understood as metaphors, as Lakoff and Johnson suggest (‘the water went from hot to cold’ can still be understood when it’s changed to ‘the water meandered from hot to cold’), which strongly suggests that those metaphors are not completely dead, and can thus be re-anchored in the wider-spectrum sensorimotor system by novel variations.

This ‘embodied’ approach seems to me a welcome middle way between the two positions, and is also consonant with what one might expect from good common sense. Lakoff has nonetheless exaggerated the power of conceptual metaphor, whose effects inevitably depend not on their verbal formulae but the actual deep referents themselves – which rise, sink, alter in metonymic emphasis or disappear, while the metaphor itself often retains the identical form of words.

9. Linguistic relativism

The idea that metaphors have the ‘power’ to reframe or shape thought is closely related to the linguistic relativism (or determinism, in its ‘strong’ form) of the Sapir-Whorf hypothesis. Only the weak version of the theory – the proposition that language influences, rather than determines, the thinking we do – has much credibility these days. As Lakoff has pointed out from the start, most of our basic metaphors are ‘embodied’, i.e. they emerge directly from our corporeal status, our physical orientation and topology, our spatiotemporal dimension, our proprioception, and so on; and therefore it might seem reasonable to assert that they do indeed ‘shape our thought’ to some degree. However, ‘embodiedness’ is not an arbitrary cultural convention, and I find it hard not to see these ur-metaphors as ‘merely the shape our thought takes’. (I suppose one may consider our human embodiedness a ‘culturally relative state’, but this Olympian perspective is not one we have the luxury of indulging for long.) Even our original, newly coined metaphors are merely those which we are capable of thinking, and we are no more able to conceive of tropes outside that set than we are to imagine colours we cannot see. What looks like a symbiosis between a formal expressible idea and the dynamic thought that takes place within it may just be their consubstantiality. (Interestingly, this is a perceptual error that likely originates in the false propositional metaphor indicated by the words ‘within it’ – the result, if you like, of the ill-advised extrapolation of an otherwise perfectly useful embodied-type metaphor, adequate to some descriptions of reality but not others.)

Elsewhere, local, culturally fixed and unconsciously deployed conceptual metaphors can be explained in much the same way as anomalous linguistic systems of colour, number or tense: not as evidence of a fundamental difference of neural wiring, but mere reactions to cultural or environmental necessity, topographic circumstance, and so on. Doubtless these are self-reinforcing; doubtless they are censorious of certain kinds of thought – but I am sceptical as to whether they ‘place limits on them’. Whenever a genuine exigency presents itself – whether it’s perceptual, sequential, economic, existential, sexual, or whatever – and requires us to think beyond those local conceptual metaphors, we tend to have little trouble in doing so. Our concept-blindness is then revealed as less a conditioning of our thought than merely a sensible and economical adaptation to a fairly stable set of cultural constraints. However, I would concede that – based on what I understand of the current evidence – the constant reinforcement of specific neural pathways can lead to their overwiring, and possibly even the abnormal growth or contraction of parts of the brain’s architecture; certainly there is now proof that neurogenesis occurs in the hippocampus of human adults, and appears to be shaped by personal experience and the patterns of behaviour that result from it, meaning that the brain is consequently ‘individualised’. One would not be surprised to find that the patterns of an idiomatically distinct culture might have the same effect; the idea that a repeated thought should exert some downward causality and restructure its own material basis in a way that would then affect the kind of thought it subsequently produced – isn’t an outrageous one. No thoughts are, after all, ‘unembodied’; but this is a dynamic, not a ‘deterministic’ model. My gripe is with those who seem to insist that language influences or limits the mind through some magical supervenience. Language is merely an aspect of mind.

10. Figures of speech

Below is a reasonably comprehensive list of those rhetorical devices which still crop up in contemporary poetic composition. They have all occurred regularly in the poetry of the last 100 years, although most often go unidentified. However, as much as a few older scholars still like to present rhetoric as a subject which enjoys an unusual rigour, it is really a rag-bag of poorly defined, barely agreed, contradictory and often overlapping concepts. Moreover, the set of concepts that should be comprehensively reflected by these terms is covered patchily at best. I would avoid – like the plague – making any proscriptive or prescriptive use of them: this is a recreational bug-hunt, not a proper discipline. (It used to be, but we should now take it about as seriously as alchemy, i.e. as a historically interesting precursor to more sensibly disciplined approaches.) Their arrangement here is, inevitably, a little arbitrary, and some terms could arguably have been as sensibly placed under different headings. I’ve made three divisions: first, rhetorical figures proper: these are performative devices concerned with ways of saying, of making convincing and manipulatively effective speech; secondly, those figures that describe deviations from and unusual ways of manipulating ‘normal’ syntax; and thirdly those figures that operate at the level of the single word. (These are useful in discussing older poetic dictions.) My rule of five – here used to distinguish figures of addition, repetition, arrangement, substitution and omission – is elsewhere often a rule of four, with addition and repetition conflated. The distinction seemed to me just about worth making, however. (The popular ‘trope and scheme’ distinction is terminally flawed.)

Either way, their names are still darned pretty, and occasionally they are clearly useful in the identification and description of an effect. I still use them frequently, but I appreciate that mine might be the last generation to do so; if that’s the case, no tears should be shed, but something should be found to replace them. They are still a good, handy way of caricaturing a style: most poets can be identified by their excessive fondness for half-a-dozen of these figures, and noting examples of their effective use can be a highly useful exercise for the apprentice poet.

1. Rhetorical

i. Addition

exemplum: citing an illustrative example

metanoia: the correction of a statement immediately previous

apophasis: inclusion by affecting to omit

enumeratio: detailing parts, causes, effects, or consequences to make a point more forcibly

metabasis: a brief summary of what has been said and what will follow

asterismos: adding a word to emphasise the next statement

ii. Repetition

repetitio: irregular repetition of words or phrases

amplification: repetition of a word or expression, adding more detail

accumulatio: repetition in other words

antanaclasis repetition of a word used in a different sense

antithesis: repetition by the denial of the contrary

sententia: concluding or summarising foregoing passage in a pithy statement of general wisdom

iii. Arrangement

auxesis: material presented in ascending importance: can take the form of climax or bathos

iv. Substitution

periphrasis: more words for less

euphemism: pleasant words for unpleasant

dysphemism: unpleasant words for pleasant (or neutral words)

irony: any speech where the literal meaning of the words and their actual intention is opposed

hyperbole: overstatement for effect

meiosis: understatement for effect

litotes: a form of understatement, denying the opposite or contrary of the word which otherwise would be used

v. Omission

(figures of contraction, or ellipsis at the level of clause)

enthymeme: the omission of a clause which is logically implied

anapodoton: the omission of a clause which is not logically implied, but can be inferred by context

aporia: the profession of a doubt over what to say or choose

erotesis: a rhetorical question

aposiopesis: a breaking off, as if one were unable to continue

anacoluthon: the act of breaking off and finishing a sentence with a different grammatical structure from that with which it began

praecisio: the omission of the whole: the conspicuous silence

2. Syntactic

i. Addition

pleonasm: addition of superfluous words

polysyndeton: addition of conjunctions between clauses

ii. Repetition

anaphora: repetitions of beginnings

epistrophe: repetitions of ends

epanalepsis: repetition of the beginning at the end

inclusio: epanalepsis at the level of passage

epanados: repetition in reverse order

chiasmus: epanados at the level of passage

anadiplosis: repetition of the end at the next beginning

gradatio: successive, linked anadiplosis

conduplicatio: similar to anadiplosis but it repeats a key word (not just the last word) from a preceding phrase or sentence, at the beginning of the next.

epizeuxis: immediate repetition or duplication

diacope: repetition, with a word or two between; a kind of syntactic tmesis

isocolon, parallelism: duplication or successive similarity of syntactic forms

paregmenon: the use of several words of a common root together

polyptoton: a paregmenon where the repeated word has a different grammatical function

iii. Arrangement

hyperbaton: displacement of a single element; uncommon syntactical usage

anastrophe: a hyperbaton in which noun and adjective, or two other elements are reversed

hypallage: an anastrophe which changes the sense

hysteron proteron: an anastrophe in which temporal order is reversed

metastasis: an unexpected change of tense

parenthesis: a word, phrase, or whole sentence inserted as an aside in the middle of another sentence

hypotaxis: using a series of subordinate clauses or phrases (the opposite of parataxis)

parataxis: the use of successive independent clauses, with or without conjunctions

iv. Substitution

metonymy: the substitution of a word for a related or aspectual term (such as cause for effect or effect for cause)

synecdoche: locally, the specific metonymy of part-for-whole

metalepsis: a double metonymy, a metonymy substituted for a metonymy

anthimeria: deliberately ‘wrong’ substitution for one part of speech for another

hendiadys: the use of ‘and’ to split a single qualified noun into two parts

antiptosis: using a prepositional phrase for an adjective

catachresis (often in the form of transferred epithet): an apparently ‘inappropriate’ substitution, usually for figurative purposes

enallage: an effective grammatical mistake

paronomasia: a pun, or play on words; a word used in a way that draws out its ambiguous meaning. It is behind some highly suspect ‘suggestion’ techniques in the quack science of Neuro-Linguistic Programming: here the ‘top’ meaning of a word used in a local sense veils a second meaning in a generic sense, or vice versa. (Poetry employs it in much the same way.)

v. Omission

ellipsis: omission generally

asyndeton: the omission of conjunctions

scesis onomaton: the omission of the main verb

zeugma: yoking two words to a single word which has a different meaning for both

3. Lexical

11. What’s not in a name

While my acquaintance with the philosophy of language is that of a complete amateur, I’ve read just enough in Kripke and Frege not to be too breezy in my asseverations regarding ‘names’. But there’s a little infinite regress here: if ‘Don Paterson’ means ‘he-who-goes-by-the-name-of-Don Paterson’, then what does the proper name in ‘he-who-goes-by-the-name-of-Don Paterson’ mean? ‘He-who-goes-by-the-name-of-“he-who-goes-by-the-name-of-Don Paterson’’’? You see the problem. However, I think this nonsense also directs us towards the truth of the matter, which is a version of the ‘use-mention’ distinction. The name ‘Don Paterson’, when interrogated, can be seen to be commonly employed as a metonym for ‘he-who-goes-by-the-name-of-Don Paterson’; however when the name occurs casually within in that sentence, it is being employed as an arbitrary name proper – that is, a pure ‘asymbol’ whose connection is forged by convention alone, the Millian idea of a ‘name’ as a pure denotation. Peirce thought their changing semiotic status was tied to one’s degree of acquaintance:

This passage still seems staggeringly insightful. However, while it seems to me that names are invisibly symbolic in their unconscious conversational use, they still appear indexical when consciously analysed in isolation, as one reacquaints oneself with the surprise of meeting them the first time: they are de-iconised. Again, one to be settled by the MRI scan, I suppose; but I’d point out that if you had substituted your own name in this self-conscious footnote, you would be just as struck by its strange, queasy index as I currently am by mine – enough to scream inwardly ‘but that’s not my name!’ and try to either separate it from yourself, or attempt to flip it quickly into blissfully unthinking symbolicity, so as not to disown it completely. This seems to point to not just a trivial trick of the mind, but evidence of a fundamental change in its function under these conditions. (It also points to something else: the status of the name as both semiotic symbol and index means they are doubly arbitrary, suggesting that ‘names’ might be the wave-particle duality of language; although, as far as proper names go, I am not convinced that we can ever have much perspective on a linguistic phenomenon in which we are all so existentially invested. A conscious banana would probably be appalled to know what name it went by, and then spin off into something identically bewildered.)

I mention all this for one simple reason: I believe – in its fetishised attention to sound-detail, in its careful originality, in its deliberately slowed delivery – that poetry insists on our consciousness of words as strange indices, not mere symbolic tags in a sign-system. It ‘novelises’ language, and in doing so conducts a global shift towards the indexical function: this creates the paradox of moving language one step closer to the reality it seeks to represent, while reawakening us to the very strangeness of the signs by which it does so.

12. Tonality, atonality and expressive range

At the end of Frost’s poem – just when you feel things couldn’t get any more depressing – the poem’s apparently throwaway last line, ‘If design govern in a thing so small’, makes a deictic shift into a domain far more appalling than that of our merely evil demiurge. Frost’s sudden reframing of his material is so radical, I can hardly read the line without feeling my stomach lurch: the poem, I think, is saying … ‘Maybe God doesn’t sweat the small stuff.’ At least with the big stuff, we have the sense of someone taking an interest, even if it was a bitter or malevolent one – but one who nonetheless creates the horses, the forests, the sunsets, our children. Maybe what we’re seeing in this Boschian Lilliput we find in the undergrowth is simply the nature of physical law. Unfortunately, it turns out that the physical law on which this universe is founded is a rebarbative, amoral, murderous enormity. One ends the sonnet nostalgic for the demon-designer of line 13; anything, rather than be left alone in a godless universe founded on such horrific principles.

Note that this line symbolises practically the whole rhetorical range of the traditional aesthetic, and the astonishingly complex and subtle effects it can produce. No experimental poem would ever ‘throw away’ its biggest line – nor arguably ever could, since this kind of nuanced understatement is possible only in the context of a poem which signs its allegiance to and is read within the traditions of consistent argumentative structure and rhetoric. I know that sounds like tedious finger-wagging, but I really do intend it as a neutral point: experimental poetry is not generally concerned with the kind of argument that could make use of such an effect in the first place. Either way, its saliences are rarely achieved by meiosis, something that may account for its attenuated expressive range.

This is a perennial limitation of the discontinuous approach in all the arts; an intrinsic lack of resolution is often culturally interpreted as a sign of tension, a tension which tends to then infect the whole domain. I recall a conversation with the late Michael Donaghy about the film Amadeus (initiated by the fact that Donaghy was a hilarious double for its lead, Tom Hulce). In the film, Mozart’s own music served as the soundtrack to his life, in all its love and loss, its triumph and disaster. Imagine, Donaghy proposed, that the same trick had been attempted in a biopic about Schoenberg: you’d spend the entire ninety minutes – whether our hero was on the tennis court, sharing a candlelit dinner with a lover, hailing a cab, giving a lecture or wheezing on his death-bed – worried that the lunatic from Friday the 13th was about to jump out from the bushes or the cupboard with a machete. The emotional palette of non-tonal music is, for nearly all listeners, constrained and tense. The ear helplessly resolves everything into tonic-dominant patterns of tension and resolution, and while symbolic or ghostly substitutions for tonic resolution are possible – Alban Berg is the go-to name to drop here – we mostly perceive serial or atonal music as an extended and unresolved dominant sequence. (Moses und Aron put me in the same state of psychosis as would a month in the trenches.) The effect can be suspended and diminished through exposure, familiarity, and the slow letting-go of the deeply inculcated expectations of functional harmony, but it is not replaced by some atonal analogue of the tonal language of feeling.

Too much can be – and perhaps is – made of the tale of George Rochberg’s conversion to tonality in the sixties. Rochberg was a very decent composer, but no one is pretending he was Webern. Nonetheless, the story is a poignant and instructive one. After the death of his son, Rochberg found that the icy angles, symmetries and dissonances of his serial technique simply could not express his grief: he needed a tonic as well as a dominant to compose music adequate to his tragedy. For the same reason, I would gently propose that a neo-modern experimentalist writing a ‘heartfelt’ poem for the funeral of a beloved friend or family member might find, say, the technique of switching out every noun for the next one in the dictionary inadequate to their purpose. The constriction of emotional range that discontinuous technique usually implies is another reason poets might be better to think of it as an occasionally useful mode, not a revolutionary solution.

13. Chicken and egg

So much depends, as they say, on what you think about things, and what you believe about reality. As for us poets: William Carlos Williams’s absurd insistence on the integrity of the object and the intrinsicality of its own meaning was nonetheless driven by a laudable distaste for the way writers had rendered chicken, wheelbarrow and everything else unthinkingly subservient to hierarchical, classical, Eurocentric, colonial and patriarchal frames – a situation that, if anything, was made worse than ever by the palaeo-modernists WCW despised. However, his position remains that of a typical neo-modern absolutist: his distaste for all literary comparison (metaphor does indeed narrow our view of a thing, but it simultaneously makes it part of a new compound sign), and his embarrassing Schwärmerei towards Gertrude Stein, who, he famously claimed, ‘has gone systematically to work smashing every connotation that words have ever had, in order to get them back clean’3 reveals someone as clueless about the business of naming as he was about prosody. Mercifully we do not judge poets on the quality of their ideas. (As much for my own sake – I pray we do not.) By contrast, the arch-relativist Eliot’s formula would run something like this: humans produce language and culture, and only culture produces reality; reality is therefore an adaptable construct in which the ‘word’ is so inevitably, so wholly continuous with the thing it denotes, combinations of words can create and destroy reality in a stroke. (Occasionally this omnipresent and invisible rule will burst into popular consciousness – see the furore around the revision of Pluto’s status from major planet to dwarf, for example; its demotion was treated by some as a ‘real thing’, though we can be certain that Pluto itself was broadly indifferent to the matter. The beautiful, lonely little globe certainly looked that way on our recent flyby.) I would prefer to say that the word and/or the concept it denotes circumscribe a phenomenon; without this circumscription it is literally unthinkable, and with it – however absurd or fantastical the thing may be – the thing exists by virtue of its mere adduction, its very ‘thinkability’. My own position is therefore that if there is no intrinsicality, and if ‘definition’ is a human act of semantic circumscription, it follows that the ability to be ‘thought of’ is the only sensible qualification for ‘a thing’s’ existence. (One thing remains arguably intrinsic: that of intent, i.e. that motive, emotion or desire which drives or informs our actions. But there, alas, I am not sure we are any further forward than Berkeley: ‘… we cannot know the existence of other spirits otherwise than by their operations’.4) As I trust may be apparent from these remarks, I line up behind Quine’s anti-essentialism, and the assertion that all properties are ‘accidental’. ‘Things’ may be partly composed of what I refer to as ‘core connotations’, but those are a mere matter of cognitive designation and cultural consensus. There are no primary qualities. An egg is not really an egg, but a set of expectations and assumptions about an egg. Of course the egg has ‘chemical properties’ and through these might initiate a ‘process’ – but ‘process’ is not ‘motive’ (the ‘intentional stance’ is a grand heuristic and predictive tool, but nothing else), and by ‘properties’ we do not mean anything like the ‘properties’ we intend when we ascribe meaning, role, function, and so on. (For a start, one can guarantee that the hen has a very different take on the thing – one that even humans might not dismiss as irrelevant.) Poets are mostly reality-sceptic, and expert in both pointing out these merely consensual frames, and positing the alternative lives of ‘things’. ‘What do you know / Of the revolutionary theories advanced / By turnips, or the sex-life of cutlery?’5

14. A brief defence of jargon

It’s impossible to have a wholly nonsensical abstraction, as they always hold out the possibility of their own hypostatisation. This is why they can be so freely combined; randomly thrown-together terms like ‘polysemic intransigence’ or ‘vectorised incoherence’ meet with no real objection from the mind, and sound like (and indeed are) plausible or potential states or conditions. This is the idea behind the Sokal hoax and the ‘Postmodern Essay Generator’, but also the stylistic raison d’être behind this notorious, prize-winning corker from Judith Butler:

The move from a structuralist account in which capital is understood to structure social relations in relatively homologous ways to a view of hegemony in which power relations are subject to repetition, convergence, and rearticulation brought the question of temporality into the thinking of structure, and marked a shift from a form of Althusserian theory that takes structural totalities as theoretical objects to one in which the insights into the contingent possibility of structure inaugurate a renewed conception of hegemony as bound up with the contingent sites and strategies of the rearticulation of power.

Some might say that formal logic and algebra were developed to relieve language of the task of articulating such ugly specificities, and to deny it the opportunity to pull a fast one. Not that I doubt that Butler has the real world in all its hypostatised richness in mind; however the lapse into pure abstraction has, in the eyes of some, signalled her detachment from it – and, we are told, led her into stylistic disaster. This is not just unreasonable: it betrays a foolish prejudice. A jargon’s usefulness depends entirely on one’s familiarity with it, since it’s merely a form of in-house shorthand. For those who have learned an abstract jargon-word in its professional context, its definition is usefully narrow and precise; within its own discourse group, it can – if it needs to – invoke a synonym, or indeed an entire hypostatic set of concrete exempla along the axis of selection. The fact is that well-understood and narrowly defined abstracta assist enormously in both the rapid juggling of terms and the expression of larger, composite, nuanced ideas. The problem is that many of these words sound familiar to readers outside the circle – but within that circle, they are not used with the ambiguity or fuzziness others might hear, and indeed often have different senses entirely (as anyone who has encountered, say, the word ‘perversity’ in a critical theory context will confirm). Some read Butler as a model of lucidity, and I believe them. Other writers are less sincere, and this mode offers cover and comfort to much intellectual charlatanry and self-delusion – from undergraduate to professorial level. Its main crime is the casual hurling around of abstracta with the confidence of concrete substantives in a way that has been insufficiently jargonised. The test is simple: ask for a rough synonym of the word, or paraphrase of the passage. If none can be produced, you’re either dealing with a chancer, or someone halfway towards a private language. Only poets are allowed say ‘my immortal line can suffer no human paraphrase’, and even then you should raise at least one eyebrow.

15. Music, synaesthesia and the poem

Many musicians have a touch of sound–colour synaesthesia; personally, I see/hear A major as a bit pale green, and A minor as a bit dark green. But the fact that I know another musician hears it as blue renders my association firmly ‘asemic’. Relativists will also tell you that the emotional associations I ascribe here to certain harmonies are culture-specific, and by implication also asemic: they will point to the variation in the emotional interpretation of identical signs to prove the mutually exclusive musico-semantic nature of different native or folk musics. ‘Minor’ Indian ragas, for example, are not necessarily perceived as ‘sad’, any more than the microtonal intervals of Turkish music are experienced by Turkish listeners as out-of-tune. However, I think the point, while accurate, is also seriously overstated. The quality of the feeling may be culturally shaped, but it is very far from asemic. The extent to which musicians from any culture, whether improvisers or composers, quickly ‘master’ dodecaphonic tonal harmony points to it being, if not a universal system, one which can be very quickly and universally understood, as well as being a partly hardwired principle tied to the harmonic series. (There is a link to the anti-Whorfian argument here: cf. the effortless acquisition of Portuguese by the Pirahãn Amazons.) The music of Toru Takemitsu or the improvisations of Djivan Gasparayan are heavily influenced and coloured by their native musics, but the dodecaphonic, triadic harmony on which they are often built seems close to an emotional lingua franca. The argument that all this is merely Western cultural hegemony in action is wishful thinking: the strain would show. These musicians are not speaking in any ‘second language’, and compose with the idiomatic fluency of any ‘native speaker’, while maintaining, effortlessly, their own cultural inflections. (This is setting aside the more mundane observation that they play exactly same mix of algorithmic EDM, stadium belters and what my kids call ‘douche rock’ on, say, Ice FM Nuuk in Greenland as they do everywhere else, with few obvious signs of cultural coercion.) At the end of the day the harmonic series is not negotiable, and has a relatively finite number of interpretations. The dominant, the fifth above the tonic, is the first ‘new’ non-root note in the harmonic series; whatever music we identify as being built on it has a ‘tense’ quality that demands resolution. (Even in the most atonal music, musicians and musically sophisticated listeners will instinctively posit a tonal centre not via a tonic, but the vestigial appearance of what sounds like dominant harmony; in such circumstances, the existence of a tonic is only confirmed by the appearance of dominant tension.) Since the fifth is embedded into almost all musical notes that aren’t pure waveforms, it should be no surprise to find it watermarks all our human musical thinking. Basic music theory shows that once we have a tonic and dominant, some from of triadic harmony is almost a logical inevitability. If there are deep psychological mappings between harmony and ‘meaning’, the (very) vague aesthetic similarity we may sense between, say, Bach and Dickinson on the one hand, and late Schoenberg and J. H. Prynne on the other, may be more accurately defined than merely ‘tonal vs atonal’; we may identify a lack of cadential resolution, for one thing, as well as an absence of the usual hierarchy of detail or scale-tone that triadic harmony provides; conversely in ‘tonal’ verse, one might even identify ‘plagal’ areas of structure which seek resolution by first moving away from a thematic centre (such as those typified by second quatrain of the English sonnet).

Of course, this isn’t intended to suggest that other kinds of tonal organisation aren’t equally legitimate – just that only one is unavoidable. There are theories as to how this situation was encouraged, though one of the simplest states that while monophonic, melodic folk music can take many forms, it is primarily an ‘outdoors’ or ‘small space’ music. It can move into an art music as it acquires complexity, but adds that sophistication horizontally in the form of intricate melodic line, especially when enhanced by improvisatory skill (Indian classical music being the prime example). When music moved into large, echoey spaces with resonant frequencies, long decays and complex reflections, triadic harmony was the inevitable, vertical product of hearing overlapping notes, whose consonance or dissonance was suddenly made unavoidable. (This is not an explanation of polyphony, a far older phenomenon.) Something like a universal grammar of harmony might not explain the consistent connection between the signs it creates and the broad emotional effects they provoke – that mechanism will remain obscure for some time – but it might at least shed some light on the stability of the relationship. By the same argument, one would be surprised not to find interlingual phenomena where identical or close tonal qualia were evoked by similar musico-semantic means; though for now these are almost wholly inscrutable.

16. Paradox and intertextuality

Forgive me for pursuing another logical quibble some small way. There are other reasons we are more conscious of the paradox of the indexical status of the ground under ‘intertextual conditions’. The relationship between the p-text allusion and the a-text mimics the semiotic feedback loop that we find in the symbolic. (If you recall, this is intra-domain on the way out, inter-domain on the way back: an index is propositional of an abstract concern – the ‘mirror-like sea’ is a salient detail which, if we read it as an index, suggests ‘peace’; that abstract concern becomes part of the thematic domain, which in turn appears to have hypostatised the same detail as a totem of its animus. This means that on a second reading of this poem we now know is about ‘peace’, the ‘mirror-like sea’ now seems less evidence proposing a theme than symbolic of a pre-existing one.) Initially the relationship between allusion and a-text appear less like V–T and more like a double subject: i.e. it looks isologic, where, if you like, the indexical content of the ground points in both directions. (In the example given, the Donaghy points to the Marvell, but because the sign is compound, the Marvell now seems to point us straight back to the Donaghy.) However, an isologue with one text-absent component is asymmetrical, and therefore a logical impossibility; hence my again referring to the status of the intertextual as a paradox. In a normal isologue, the indexical function of the ground is partly switched out for the active projection of shared content and family resemblance between two roughly equal domains. (In my imaginary poem ‘The Final Hurdle’, the fast car that crashes in l.3, say, and the sprinter in l.9 are clearly connected by their shared attributes.) The weak, two-way indexical function is largely activated by the reader. In an isologue where one subject is text-absent, the ground must indicate (‘allude to’, here) the text-absent domain for us to establish its identity; but the trade-off is that even though we discover what has been indicated is ‘merely another subject’ which might then form the basis of an isologue, it has by then effectively been upgraded to a ruling tenor. This tenor will to some extent overdetermine the meaning of the entire present-text in a way that inevitably renders it minimally subservient to its domain: the Marvell partly ‘rules’ the Donaghy. This is strongly reminiscent of the operation of symbol. Just as with the symbol, we are unconsciously gripped, I would claim, by a deep and pervasive ontologico-cultural metaphor: absent domains are by definition dominant (possibly because they do their work invisibly, and by unseen, ambient or mysterious means, conditions we may unconsciously associate with ‘the work of God’). Despite ‘intertextuality’ appearing to propose a connection between two equal subjects, the asymmetric relationship created by text-absence means that the power relation between the domains is similarly one-sided, and (as with the symbol, where the animus extends its influence beyond the totemic detail which first proposed it) the influence of the a-text may be considerably more broad than just a local allusion. A-text is always the boss of p-text.

17. The burden of burdens

Repeated burdens are a problem in poetry; these often have their origin in song-forms – but if you’re singing, the same words can be shaped to mean a thousand different things. In the poem, a line repeated is often just a line repeated. Poets will go to great lengths, sometimes ingenious and sometimes contrived, to put a different spin on each repetition. The villanelle is best passed over in silence, I feel; it is the S. T. Coleridge of poetic forms, and not nearly enough decent poems have been written in it to justify the discussion it has received. (Indeed its best-known example is one of the silliest poems I know, Dylan Thomas’s ‘Do not go gentle into that good night’. Carol Ann Duffy’s perfect and regrettably long-suppressed squib ‘Fuckinelle’ really should have been the last we heard from the damn thing: ‘The poet has tried to write a villanelle. / He’s very pleased. The audience can tell.’) The sestina was invented in the twelfth century by the French troubadour Arnaut Daniel. Its rules are even more tortuous than those of the villanelle. The sestina offers no repeated lines, but something even worse to test the ‘inventiveness’ of the poet: a fixed pattern of ‘homoioteleutons’ woven through six stanzas of six lines, followed by a three-line envoi. The six words that occur at the line endings of the one stanza have to be repeated in the next, in the precise order 615243; this procedure is then repeated with each successive stanza until six permutations have been performed. The six words must then reoccur in their original order in the envoi, and no-one has ever explained to me why. The result is something like the tuba solo of poetry, with the difference that in the more sensible world of music the tuba solo is more or less banned for anything but comedic effect. By the sestina’s end, the poet has usually tied themselves in such wheezing, desperate knots that the envoi resembles a drunk game of Twister in a care home, with the final lines either strained beyond credibility, inexplicably surreal, total rubbish, or technically no longer in the host language. I keep a small anthology of sestina envois that I show students to make the point. I don’t like sestinas. No one in the history of poetry ever fell into a swoon and was dictated a sestina by an angel. No brilliant line ever told its author it wanted to be a sestina when it grew up. The odd exceptions tend to succeed by tackling themes which tightly align with the repetitive, claustrophobic or Kafkaesque spaces the form can create. (Elizabeth’s Bishop’s ‘Sestina’ being the most obvious example, but there is a fine narrative sestina in Glyn Maxwell’s long poem The Sugar Mile, and there are handful of blinding experiments by Paul Muldoon.) The explanatory note that accompanies Billy Collins’s priceless ‘Paradelle for Susan’ summarises the deranged constraints of the Provençal forms: ‘… [the Paradelle] is a poem of four six-line stanzas in which the first and second lines, as well as the third and fourth lines of the first three stanzas, must be identical. The fifth and sixth lines, which traditionally resolve these stanzas, must use all the words from the preceding lines and only those words. Similarly, the final stanza must use every word from all the preceding stanzas and only those words.’ The poem itself concludes: ‘I always cross the highest letter, the thinnest bird. / Below the waters of my warm familiar pain, / Another hand to remember your handwriting. / The weather perched for me on the shore. / Quick, your nervous branch flew from love. / Darken the mountain, time and find was my into it was with to to.’ For his trouble, Collins received some genuine, foam-flecked abuse for his failure to honour the paradelle’s noble tradition, despite it not actually existing.

18. Love & marriage

I confess that there are very few hobbies more rewarding than winding up the radical wing of the avant-garde. Never has so little provocation met with such gloriously disproportionate response – at least from that shirty corner of la rive gauche who insist on the meek silence of ‘the mainstream’ as the price it must pay for its modest commercial success, and who will self-combust with righteous apoplexy whenever the cultural value of their own contribution is gently questioned (by adducing something as tawdry and irrelevant as, say, actual book sales). Or, I’ve noticed, when some hapless mainstream-ite attempts any kind of assertion about anything, however anodyne: one opens one’s mouth and is immediately gummed alive by their hipster outriders and keyboard ninjas. But all poetic factions, my own included, are guilty of perpetuating what Michael Donaghy once called ‘a knife-fight in a phone box’. What would have amounted to a little recreational cattiness over a half of mild in an upstairs room forty years ago will now, alas, take the form of a social movement. Because our lop-sided debates are now largely conducted through social media – where dissent can be blocked, and no passionate minority is small enough not to feel like a caucus – shared grievances have been refined to something like weapons-grade, though it often seems that the energy required to sustain them exceeds any being ploughed into the art they supposedly defend. (Our enemies form our characters to some extent, and perhaps that is no bad thing. I find myself at least as grateful for their advice as I am that of my mentors, as it is often the more honest. I was for years the proud owner of an indefatigable troll; behind his various cryptic noms de théâtre lay an otherwise clever and funny academic, a man over whom – cheeringly, if confusingly – I appeared to exercise some kind of terrible psychosexual hold. I fear the poor chap will end his days shivering at my graveside, riding out wind, rain and sleet like some tank-topped Greyfriars Bobby. But I should miss him too: he was often right.)

All the same, the mutual exclusivity of our various sects can serve a real purpose. The avant-garde remain critically important to the poetic biosphere, and contribute an alternative gene-pool that regularly saves the mainstream from death by inbreeding (the British Poetry Revival of the 1960s, for example, rescued UK poetry from a post-Movement, Americanophobe doldrums that might have rivalled the Georgians for mediocrity), as well as providing the new bacterial strains that lightly infect and then immunise the mainstream from the avant-garde. (One symptom of full-blown avant-gardism is the total loss of a lay readership.) They must be read and taken seriously, if not perhaps always as seriously as they take themselves. But as for those naïfs who seek nothing but the harmony of the broad church: without taking up a strong position, there can be no dialectic, and sometimes that position may involve an element of disingenuous self-caricature, as I confess it occasionally does in my own case. One follows the work of les avants far more closely and with far more pleasure that it sometimes suits one to admit – and it is important to: they celebrate their best writers like any other group, but their instinct is to keep them secret. (Witness the justified pride they have taken, say, in the plainly radiant talent of Veronica Forrest Thompson over the years; a pride which nonetheless has managed to do literally nothing to increase her readership, and indeed has had precisely the opposite effect.)

To finish with a more serious point: among the more extreme proponents of the avant-garde, experimental, ‘innovative’ postmodern, ‘Cambridge’, ‘non-conformist’ and weak-hypothesis versions of the L=A=N=G=U=A=G=E schools, there remains a lack of understanding – a code of denial, even – over the practical consequences of discontinuous speech, the superstitious avoidance of the language of grammatical function, and what is lost when sense cannot leap the phrase boundary. Not just sense, but metrical impetus, phrasal rhythm, tone, style, lyric continuity, argument, narrative – all of which require larger parsable syntactic structures to provide their carriage. This is not to say there aren’t other pleasures to be found besides, although ‘pleasure’ is perhaps not the aim. Some avants delighted in Geoffrey Hill’s semi-buccolingual pronouncement ‘Whatever strange relationship we have with the poem, it is not one of enjoyment’. No less celebrated was his perennial claim that ‘genuinely difficult art is truly democratic’; and at this point Hill may be said to have firmly joined the avant-garde, whose war-cries are largely indistinguishable. Alas, this is not ‘a counterintuitive truth’ but a demonstrably false statement. There is nothing democratic about the business of making intelligent men and women feel stupid, decade after decade. Often the fault is the poet’s, and any fool can express themselves inadequately, or provide difficulty without reward (he said, knotting his own noose). But intelligent people also have to accept that they might occasionally feel stupid because the argument or idea that the poem incarnates is intrinsically difficult, and its nuance is indistinguishable from its value. These poems are the products of superior intelligences. (Here we may indeed just think of Hill’s best poetry; who would dare change a word of The Mystery of the Charity of Charles Péguy?)

1 Matte Blanco, The Unconscious as Infinite Sets: An Essay in Bi-Logic (London: Karnac, 1998).

2 C. S. Peirce, The Collected Papers of Charles Sanders Peirce, vol. 2, ed. C. Hartshorne and P. Weiss (Cambridge, MA: Harvard University Press, 1958).

3 William Carlos Williams, Selected Essays (New York: New Directions, 1954), 63.

4 George Berkeley, The Works of George Berkeley (London: Macmillan, 1871), vol. 1, 231.

5 Derek Mahon, ‘The Mute Phenomena (after Gérard de Nerval)’, Raw Materials (Loughcrew: Gallery Press, 2011).