Closure and the impossibility of prosody Chris McCully [email protected] [email protected] Presentation for Metrics symposium, University of Essex,

Closureand the impossibility of prosody

Chris McCully

[email protected]

Presentation for Metrics symposium, University of Essex, June 2015

This paper has a long history, beginning in 1979….

An undergraduate course (University of Newcastle) taught by the late Barbara Strang:

• Sub-module called ‘The continuity of English metre’• Teaching handout (1979 - typed, with diacritics

handwritten into the text) told me that

Metre adds a form of patterning, [specifically] the line as unit defined by line-end marking, to a stress-timed rhythm which characterises all English utterance

Variety of line-end marking

The same handout taught me that metrical lines of [post-Conquest] verse ‘could be identified by the presence of one, two or three of the following line-end markers

(a) rhyme(b) a silent stress(c) monosyllabic foot [‘foot’ here = temporal measure, McC]

Choice of only one marker makes the metrical patterning unobtrusive….; use of all three makes it obtrusive’

Gray’s ‘Elegy Written in a Country Churchyard’ (published 1751) as a test case

Highly obtrusive line-end marking, here scanned as including a salient pause (= silent sixth stress at the line-end, here symbolised by the caret, ^)

Now fades the glimmering landscape on the sight, ^

And all the air a solemn stillness holds, ^

Save where the beetle wheels his droning flight, ^

And drowsy tinklings lull the distant folds ^

(Note: the website http://www.thomasgray.org [Thomas Gray Archive] scans this as iambic and pentametric; Strang read this instance of the ‘pentameter’ as temporally six-footed…and here, at the outset of analysing ‘closure’, we’ve hit our first major snag….)

Problem 1: temporal theories of metre

• There’s a very long tradition of temporal theories of metre, beginning with Joshua Steele (1779), which has been almost wholly ignored in the literature, though in different frameworks, both Attridge (1982, 2013), Hayes and MacEachern (1998) and Kiparsky (date) admit salient pauses into their metrical analyses of folk verse.

• Salient pauses (silent beats) admitted into ‘folk verse’ and ‘song-verse’ (Hayes) but distinction between ‘art-verse’ and ‘song-verse’ unhelpful

• Should metrists look again at temporal theories?

Strang’s thinking on metrical closure extended to pre-Conquest verse

‘In pre-Conquest verse only one type of structure is permitted, though it is of great diversity’ (teaching handout, 1979)

It’s worth remembering that this ‘one type of structure’ lasted for at least 600 years. The structure itself was a cultural sign and also suggests that some durable isomorphism between language form and poetic form had been found (rhythmicity, together with tight mapping between morphology and metrics)

Closure in pre-Conquest verse

‘One of the most important constraints on the highly-wrought, aristocratic verse of the Anglo-Saxons, for example, is precisely a constraint on metrical closure – on what sort of syllable can occur in the fourth position of the half-line. Barbara Strang put this most succinctly in her great History of English (London: Methuen, 1970), p.326: ‘[T]here is between half-lines a break, or line-end marker….The end of a half-line is always determinate….[I]f it is occupied by a lift [stressed syllable: McC] that goes without saying, but if it is occupied by a drop there is the special restriction that the drop must there be monosyllabic.’

(Chris McCully, (to appear) ‘Beginning with endings’. PNReview. My emphasis.)

Closure highly ranked as a formal constraint on OE verse

• Position: each half-line contains four filled metrical positions (restated by Cable)

• Closure: position 4 of the half-line is filled by L (a lift) or with exactly one unstressed syllable (Strang)

• Freedom: ‘beginnings free, endings strict’ (Hayes)• Expansion: only one expansion per half-line (Cable,

Russom)• Specifically metrical twiddly bits: verbal prefixes, the

prefix ge- and the negative particle ne may be extrametrical (Russom)

(adapted from McCully and Hilles (2005:175)

Closure may in fact be the highest-ranked (inviolable) OE verse constraint

(i) Position >> Closure wouldn’t work. This would mean ‘it’s more important to have four positions than to have closure’…but closure is apparently never violated, therefore

(ii) Closure>>Position (and other constraints). This means ‘it’s more important for potential half-lines to have closure than it is for them to have four filled positions’ – and while lines with what most analysts handle as a defective number of positions are vanishingly rare in most metrical accounts of e.g. Beowulf, they do occur, e.g. wæs mīn fæder (Beo.262a).

(iii) The ranking Closure>>Position accommodates Bliss’s light verses, e.g. half-lines with only one stress. Sedgefield (1933) noted that many lines in both Beowulf and e.g. Genesis A pattern as ‘a one-stress B [Type]’ or more commonly ‘a one-stress A [Type]’ (1933: 226 [check])

(iv) Closure in OE suggests that the metrical analysis of e.g. sealfode (and the –ode suffix of all Class II wk. verbs) should be séalfòde (Campbell, Hogg) and not séalfode (Russom, Getty)

Problem 2: synchrony and diachrony

• In all Sieversian D-type cases known to me it can’t be accidental that strings such as –ode (in class II wk. verbs) contain a syllable that is historically long (thus –ōde)

• OE metre seems (to me) to treat such strings metrically as SW, thus séalfòde (secondary metrical stress on –o(de)); note that SWW closures would be debarred under the principles I began to introduce above

• If so, OE metre is inherently archaising; I’m perfectly prepared to grant that sealfode was synchronically produced …SWW

• How does any metrist reconcile the claims of synchronic and diachronic analysis? (And how could one teach anything like that to literary undergraduates, for whom such a question might itself be meaningless?)

The death of OE metre

• There’s universal agreement that whatever the OE half-line was, eventually that precise form of metre dies, though there’s an efflorescence of alliterative poetry in the C13th and 14th (largely in NWMidl., e.g. GGK)

• Classical OE system replaced by accentual-syllabic metre: a different kind of counting (syllables counted, not metrical positions) and a different kind of closure (rhyme)

• Many possible reasons for the loss: imperfect training of poets and scribes (a vulnerability of scribalism), change in composition of potential audiences, cultural pressures > loss of prestige (esp. after the Conquest), but one reason seems to offer compelling linguistic reason for this instance of metrical death: morphology.

Metrical death and (Old) English morphology

• Russom (2012): on the basis of comparing the expanded dips (runs of weak syllables) in Beowulf and Maldon Russom argues that there is ‘a felt need for increased use of function words’ in later OE and that this compromised classical OE metrical principles

• To oversimplify radically: a largely morphologically-based system (OE, where half-lines have the prototypical metrical shapes of well-formed compound words) became a primarily phonological one (ME, accentual-syllabic). Crudely, the original work of inflections (e.g. to indicate possession, location, movement) came to be handled by more systematised use of function words such as prepositions.

• However….

Problem 3A: Literary history and English morphology

If morphology and linguistic change interacts with the death and birth of English poetic forms, and therefore with literary history, then

(a) do students of (English) literary history need to know about (English) morphology? And

(b) what authoritative history of English morphology would be useful in this context?

Problem 3B: What courses?

• Old English is now rarely taught at undergraduate level in the world’s universities

• Courses on ME literature seem largely to begin with Chaucer and may not cover ME alliterative verse. Further, Chaucer and (sometimes) ME alliterative verse seem now, where they’re studied at all, often to be studied in translation (e.g. Armitage 2009)

• What conceivable courses are there now, or could there be in the future, which could begin to explore questions such as that posed in problem 3A? Where syllabuses are driven by criteria of relevance, what would any metrist say to the charge that those problems outlined above are largely or wholly irrelevant?

Chaucer, metrical inventiveness and closure

Duffell (many works, most recently p.c. 2015) makes Chaucer’s metrical inventiveness abundantly clear: in order to (single-handedly) invent the English pentameter, Chaucer needed to engage with those closures he found in his models (Boccaccio, Petrarch)

Chaucer, closure and counting ‘For both Boccaccio and Petrarch it was possible to allow triple-rhythms both before a mid-line caesura and before the tenth syllable of the line. Thus Italian ten-syllable lines could both begin and end ‘(de-)de-de-DUM’ (io de Parnaso le Muse pregare, Filostrato Book 1, example from Duffell, p.86, runs of unstressed material underlined): “Chaucer’s great innovation….was to elevate the iambic rhythm he found in an overwhelming majority of endecasillabi to a structural principle, and he did this by not placing lexically strong syllables in odd-numbered positions, and by placing stresses in most even-numbered ones” (Duffell, p.87). By (almost wholly) ruling out triple-time closures, therefore, Chaucer single-handedly constructed “the first true pentameters in any European language” (Duffell, p.87). In these dynamics a new metre was born.’ (from Chris McCully, ‘Beginning with closure’ (to appear; see also Duffell, Martin A New History of English Metre 2008)

However….

Chaucer’s metrical inventivenes and underlying structure

• For Duffell’s analysis to work, there must be an underlying structure to the relevant line, since in Chaucer’s models triple-time rhythms could be allowed both before a mid-line caesura and before the tenth syllable of the line

• When Chaucer ruled out these kinds of line-internal and line-final closure, his metrical revolution consisted precisely in matching linguistic materials against an underlying metrical template – a structure that must in some way have been ‘mentalistically real’ (and bears comparison with Coleridge’s ‘secondary imagination’)

• However…

Problem 4: underlying structure and dualism

• While I doubt he’d deny that underlying structure exists, Attridge (2013) identifies a ‘danger of laying too much emphasis on the opposition of rhythm and metre…. [A] dualistic model of poetic movement is produced, in which the actual contours of the poem are regarded as being in constant tension with a Platonic…ideal’ (2013:123). cf. here Kiparsky 1977 and theories of metrical matching > complexity (‘The complexity of a line may be deduced from the number of mismatches in it’ [p.ref. insert])

• For Attridge, ‘complexity’ is a product of ‘[T]he event of the poem…involving…shifts in register, allusions to other discourses…., rhythmic patterning, linking rhymes, movements of sounds….all operating in a temporal medium…. I would also include…the meanings of the words and sentences, for once we conceive of a work as an event, meaning becomes an occurrence, not a substance or an abstraction’ (2013: 29)

• How should metrical analysis handle the dangers of dualism?

Closure and metrical clines

It’s possible to use even a simple set of metrical constraints, e.g. COUNT (‘something is counted’) and CLOSE (‘metrical domains must be closed’) to generate a cline of metricality. (Note that I’m using COUNT and CLOSE as headers for what is probably a family of related constraints under each head.) Such a cline might look something like the following:

Least metrical More metrical More metrical Fully metricalMost prose-like Less prose-like Less prose-like Least prose-like

No COUNT COUNT CLOSE COUNTNo CLOSE No CLOSE No COUNT CLOSE

Nonmetrical forms of verse

In such a schema, nonmetrical forms of writing are automatically (and in a principled way) included into those materials it’s the job of the metrist to analyse, even to explain.

Problem 5: describing the rhythm of prose

I don’t know of much if any recent work on the rhythmic analysis of prose, though it seems to me that in certain cases – Joyce, Beckett, some contemporary drama and writing for radio – such analyses might be well worth doing.)

Yet how much agreement would there be among metrists, critics and writers about the prosodic analysis of even the most straightforward piece of contemporary prose? And for whom might such an analysis be useful?

+CLOSE, -COUNTVerse that has metrical closure but where nothing is counted:

Frankly I don’t know how you can stand it – [SW]For months the same broken doors in the kitchen;The relics of weeks-old meals on the bedspread;Vitreous stains that haven’t received due attention.But perhaps after all a life should be measuredBy the capacity for bearing what its critics call squalor,And equally, perhaps it’s merely over-fussyOr prurient to find dirt somehow deficient,Whereas clearly it’s a symptom of long-drawn-out resentmentThat begins in the classroom, ramifies through families,Dispatches its lovers late at night to the wine-shopsAnd will drop its fag-ash even into the open palm [pattern breaks: anticipation]That has begun to beckon towards the final judgement….

(Chris McCully, ‘Letter to Torquatus’, from Polder, 2007)

-CLOSE, +COUNT

• Includes e.g. haiku. Not coincidental that popularisation of haiku in the West came with Imagism (France 1906 (Couchoud); F.S. Flint > Pound/ Amy Lowell). Pound’s ‘In a station of the metro’ appeared in 1916.

• Note that in Japanese the haiku consists of 17 on (morae) set into three phrases; typically, two images are juxtaposed via a kireji (‘cutting word’). When the form was adapted into English, morae weren’t counted but syllables were, and the characteristic lineation (three lines, 5.7.5) mimics the three phrases of Japanese originals, which were set in a single vertical line of characters.

Haiku: counting but no closure

– fu-ru-i-ke ya (5 morae)

– ka-wa-zu to-bi-ko-mu (7)

– mi-zu-no-o-to (5)

– old pond . . .

– a frog leaps in

– water's sound (literal)

– Pond-glaze. Water-light.

– Stretched leatherlegs, frantic sky.

– Summer riddles noise. (English adaptation into syllabics)

-COUNT, -CLOSE

There’s an automatic temptation to say ‘This means prose’ but in fact there are many varieties of modern and contemporary vers libre that can perhaps be distinguished from prose because they make use of anti-closural devices. That is, they deliberately resist closure or indeed association with other structures often found in ‘Poetry’.

Case study in anti-closure: Ashbery

It’s this crazy weather we’ve been having:

Falling forward one minute, lying down the next

Among the loose grasses and soft, white, nameless flowers.

People have been making a garment out of it,

Stitching the white of lilacs together with lightning

At some anonymous crossroads….

(John Ashbery, ‘Crazy weather’ [excerpt], in Houseboat Days (1977: 21))

Comment on ‘Crazy weather’

‘If there’s a principle at work in the rhythm, it’s a principle of avoidance: the regular patterns of metrical verse are kept at bay by the constant changes in the configuration of stresses and the refusal of the syntactic units to match familiar metrical structures…’

(Attridge 2013: 116; my emphasis, CBMcC)

-COUNT, -CLOSE: Prose and poetry

And of course prose can itself contain – usually as special effects for special purposes or occasions – language that isn’t merely rhythmical but also metrical:

The river – with the sunlight flashing from its dancing wavelets, gilding gold the grey-green beech-trunks, glinting through the dark, cool wood paths, chasing shadows o’er the shallows, flinging diamonds from the mill-wheels, throwing kisses to the lilies, wantoning with the weirs’ white water, silvering moss-grown walls and bridges, brightening every tiny townlet, making sweet each lane and meadow, lying tangled in the rushes, peeping, laughing from each inlet, gleaming gay on many a far sail, making soft the air with glory – is a golden fairy stream….

(from Jerome K. Jerome, Three Men in a Boat (1889), chapter XIX)

-COUNT, -CLOSE again: the case of vers libéré (freed verse)

• Pound’s Cantos (composed 1915-1962)

• Themes (often drawn from literary history), juxtapositions, fragments of aural architecture, mixture of registers, insistence on the contemporaneity of myth

• Emphasis of Pound’s earlier work on language

Ezra Pound and language

‘What obfuscated me was not the Italian but the crust of dead English, the sediment present in my own available vocabulary ... You can't go round this sort of thing. It takes six or eight years to get educated in one's art, and another ten to get rid of that education…Neither can anyone learn English, one can only learn a series of Englishes. Rossetti made his own language. I hadn't in 1910 made a language, I don't mean a language to use, but even a language to think in’

Imagism and language

And in the earliest phase of Imagism, Pound claimed that the ‘movement’ was ‘concerned solely with language and presentation’ (1912). The following were stated as desiderata:

1. Direct treatment of the "thing" whether subjective or objective.

2. To use absolutely no word that does not contribute to the presentation.

3. As regarding rhythm: to compose in the sequence of the musical phrase, not in sequence of a metronome

Imagism and its manifesto (there were many; below is c.1916)

1. To use the language of common speech, but to employ the exact word, not the nearly-exact, nor the merely decorative word.2. We believe that the individuality of a poet may often be better expressed in free verse than in conventional forms. In poetry, a new cadence means a new idea.3. Absolute freedom in the choice of subject.4. To present an image. We are not a school of painters, but we believe that poetry should render particulars exactly and not deal in vague generalities, however magnificent and sonorous. It is for this reason that we oppose the cosmic poet, who seems to us to shirk the

real difficulties of his art.5. To produce a poetry that is hard and clear, never blurred norindefinite. 6. Finally, most of us believe that concentration is of the very essence of poetry….

The later Pound: the Cantos and intonation

‘In Canto 34, for instance, Pound lifted phrases, sentences and paragraphs from The Diary of John Quincy Adams 1794-1845, published in a 1928 edition: “As he went through it [Pound] made pencil-marks in the margin against numerous passages…. These if placed together amount to thirty-four pages; in the canto Pound reduced them to eight….”’

(Noel Stock, The Life of Ezra Pound, 1970: 381, qtd. in McCully 2007).

Intonation and lineation in Canto 34 (1)

Oils, beasts, grasses, petrifactions, birds, incrustations,Dr. Mitchell’s conversation was various….And a black manservant, to embark on a voyage to Russia…Consistent with their peace and their separation from Europe….English pretentions, exclusive, auf dem Wasser…. (a.d. 1809)“En fait de commerce ce (Bonaparte) est un étourdi,” said Romanzoff…Freedom of admission for ships, freedom of departure, freedom of

purchase and sale….Are the only members of the corps diplomatique who have any interest

in literature, conversation….

(opening, Canto XXXIV)

Intonation and lineation in Canto 34 (2)

[[Oils]PP[beasts]PP[grasses]PP[petrifactions]PP[birds]PP[incrustations]PP]IP

[[Dr. Mitchell’s conversation]PP[was various]PP]IP

[[And a black manservant]PP[to embark on a voyage to Russia]PP]IP

[[Consistent with their peace]PP[and their separation from Europe]PP]IP

*

PP= Phonological phrase, IP = intonational phrase (on the prosodic hierarchy, see Hayes 1989)

Intonation and lineation in Canto 34 (3)

• Pound here apparently chose to edit (compose) not primarily by lineating syntactic constituents but by lineating intonational ones: many lines are right-edge-bounded by alignment with right edge of IP

• (Very) general tendency in many Western literatures for metrical domains to match up with larger intonational units (PP and/or IP) – dolnik, pentameter, Alexandrine, hexameter

• Was Pound ‘making it new’? Arguably, he was doing something remarkably old.

However….

Problem 6: novelty and inconsistency

• Pound was inconsistent. Many lines in Canto 34 don’t have neat right-edge intonational closures (though enough lines do to be at least suggestive)

• This inconsistency means that for many readers, the lines of free(d) verse could be any length the poet chose

• Pound’s relative scrupulousness and astute historicity – how often taught in today’s creative writing classes?

Nostalgia for closure: Olson and ‘the breath’

• Charles Olson (1910-70) associated with Black Mountain group, also Williams and the New York School

• Links with Dorn, Creeley, Pickard (and with Essex)

• Inspired by Pound: Olson’s Maximus poems are an epic of place (specifically Gloucester, Mass.)

• Interest in intonation, manifested on Olson-Creeley correspondence, much of which centred on ‘the breath’ (lineation of free(d) verse tracked – or was alleged to track - breath groups)

Olson’s practice: lines and breath (PP)…This, is no bare incomingof novel abstract form, thisis no welter or the formsof those events, this,

Greeks, is the stoppingof the battle

It is the imposingof all those antecedent predecessions, the precessions

of me, the generation of those factswhich are my words, it is comingfrom all that I no longer am,yet am, the slow westward motion of

more than I am

There is no strict personal orderfor my inheritance….. (from ‘Maximus to Gloucester: Letter 27 [witheld]’)

Olson, intonation and syllables

• Olson (like Pound) inconsistent.• Difficult to teach his principles in e.g.

writing classes• Yet Olson, for all his inconsistency, clearly

wanted to base his practice on an aesthetic rooted in language. One facet of this nostalgia for language was ‘the breath’. The other was…

• The syllable

Olson, the syllable and projective verse

Olson goes by ear, and his lines are breath-conditioned. The two halves, he says, are: "the HEAD, by way of the EAR, to the SYLLABLE/the HEART, by way of the BREATH, to the LINE." He believes "it is from the union of the mind and the ear that the syllable is born. But the syllable is only the first child of the incest of verse. . . . The other child is the LINE. . . . And the line comes (I swear it) from the breath. . . ."

http://www.poetryfoundation.org/bio/charles-olson, accessed May 2015; note that in the same essay Olson also rejects set feet: ‘all the conventions that "logic has forced on syntax must be broken open as quietly as must the too set feet of the old line”.’

Problem 7: teaching about syllables

• Syllable has a hierarchical structure• My attempt to teach creative writing students about

syllables (Onsets, Nucleus, Coda) in 1990s foundered• The same students (many of whom were deeply

influenced by Ashbery and the New York School) immediately rejected any analysis that involved the notion of ‘hierarchy’

• We departed mutually unenlightened

How can – how should – we teach literary scholars and writers about syllables? And what is our response to be if people simply don’t want to know?

Ten problems – and the impossibility of prosodyProblem 1: temporal theories of metre

Rhythmic salience; inclusion of salient pauses difficult to teach; some writers and students resistant to studying fairy tales and folk verse (which they see as ‘non-literary’); division between ‘art-verse’ and ‘song-verse’ (Hayes and MacEachern) unhelpful

Problem 2: synchrony and diachronyProsody has much to say about literary history and transmission. How teach this?

Problem 3A: literary history and English morphologyLacking an authoritative history of English morphology; also an authoritative history of English stress. Both bear critically on poetic change in English (and other languages)

Problem 3B: what courses?What courses could offer engaging and provocative work on English metrics?

Problem 4: underlying structure and dualismIf poets (and readers, listeners) can and do hear or use underlying structure, how can this be demonstrated?

Problem 5: describing the rhythm of proseWhat model would be useful and compel assent?

Problem 6: novelty and inconsistencyPractice of many C20th and contemporary poets seems inconsistent. How engage with this?

Problem 7: teaching about syllablesHow persuade literature students that attention to syllable structure and weight is important?

*Problem 8: notation and terminology (most notoriously, ‘the foot’) Problem 9: no career path for those early-career researchers interested in these matters

*Problem 10: what, precisely, do we think we’re describing, and why?

Thank you

Selected references

Armitage, Simon (2009) Sir Gawain and the Green Knight. London: Faber and Faber.Attridge, Derek (2013) Moving words: Forms of English poetry. Oxford: Oxford UPDuffell, Martin J. (2008) A new history of English metre (Oxford: Legenda, Studies in Linguistics 5, Modern

Humanities Research Association and Maney Publishing).Getty, Michael (2002) The metre of Beowulf. Berlin: Mouton de Gruyter.Hayes, Bruce (1989) ‘The prosodic hierarchy in meter.’ In Paul Kiparsky and Gilbert Youmans, eds.,

Rhythm and Meter. Orlando: Academic Press, 201-260.Hayes, Bruce and Margaret MacEachern (1998) ‘Quatrain form in English folk verse’. Language 74: 473-

507. With appendices in http://www.humnet.ucla.edu/humnet/ linguistics/people/hayes/metrics.htm.Jerome K. Jerome Society, http://www.jeromekjerome.com/2014/07/jkjs-hiawatha-parody/, accessed 9 May

2015.Kiparksy, Paul (1977) ‘The linguistic structure of English verse’, Linguistic Inquiry 8/2: 189-247. Kiparksy, Paul (n.d.) ‘A modular metrics for folk verse’. http://web.stanford.edu/~kiparsky/Papers/hayes.pdf,

accessed May 2015McCully, Christopher B. (to appear, 2015) ‘Beginning with endings’. PNReview.McCully, Christopher B. (2007) ‘The wreck of the Pound’. PNReview 172: 39-42 McCully, Christopher B. and Sharon Hilles (2005) The Earliest English. London: PearsonRussom, Geoffrey (2012) ‘What explanatory metrics has to say about the history of English function words’.

In ed. Denison et al., Analysing Older English, Cambridge: CUP, pp.15-27Sedgefield, W.J. (1933) ‘Further emendations of the Beowulf text’. Modern Language Review 28/2. 226-230.Steele, Joshua (1775) Prosodia rationalis. Second edn. (1779) ccessible via

https://archive.org/stream/prosodiarationa01steegoog#page/n7/mode/2upStrang, Barbara M.H. (1970) A history of English. London: Methuen.