How To (Learn To) Scan A Poem

First, type out the lines as a prose paragraph.. This is a crucial step if you are a beginner, because most people start off by trying to fit the language into what they think the meter should be, instead of actually hearing the language as it is. All of a sudden they start putting stresses in strange, unnatural places.

Now read the paragraph aloud a few times in as natural a way as you can. Make a recording if you want.

Now circle or highlight all the content words. Nouns, verbs, adjectives, adverbs. Make sure you know which syllable is accented in each of them, if any are unfamiliar. Every content word will have stress on one of its syllables. Listen to your recording and see if you have stressed any of the other words, like pronouns, prepositions. Suppose one like is "He is hiding under the table." There is no question about hiding and table. Did you stress is, under, the? Mark the other words you might have stressed.

By now you will have a good sense of the natural linguistic prosody of the poem. In other words, the way a naturally-speaking native speaker of the language (or competent 2nd-language speaker) would pronounce those sentences. You are almost done and you are miles ahead of almost anyone else.

Now get out the original poem, before you wrote it out in prose. Read it again outloud, naturally. Don't pause at the end of lines (very much) unless there is punctuation, but note where the line breaks are.

Now start to listen to see whether you hear any patterns, and observe whether these pattern happen to coincide with your vague memories of meters. At what point does the pattern you perceive not line up exactly with the stresses you've already determined? Does that make you want to speak the line in a less natural way ? Or does it make you want to ignore the pattern? Find a way of saying the line that makes it sound good without either emphasizing or de-emphasizing the meter.

There, now you are done. If you go any further than you will get severe headaches.

Cadence

Cadence (from Latin cadere, to fall) is a resolution to a musical or poetic phrase. A musical cadence usually falls in descending notes and seems to imply ending, conclusiveness, the way a rising sequence of notes implies initiation.

The Spanish 11-syllable line has a distinctive cadence caused by the "feminine" ending and the accent on the sixth syllable. The cadence is what is most constant, whereas the beginning of the line is what is most variable.

There will typically be three substantive words (nouns, adjectives, verbs, adverbs) in each line, and the 2nd and 3rd of these words will be lined up so that their accents fall on the 6th and 10th syllables. Like this:

Yo quiero ser LLORANDO el HORTELANO
de la tierra que OCUPAS y ESTERCOLAS
compañero del ALMA, tan TEMPRANO.

Alimentando LLUVIAS, CARACOLAS
y órganos mi DOLOR sin INSTRUMENTO,
a las DESAMPARADAS AMAPOLAS

daré tu CORAZÓN por ALIMENTO...

The cadence really dominates this particular poem by Miguel Hernández. I've hypothesized that one reason why blank verse doesn't work in Spanish as well as it does in English is because of the strength of the cadence, which tends to emphasize line-endings. Interestingly, Góngora's more fluid line-endings avoid the cadence by putting accents on 4 and 8 (the so-called Sapphic line) with more frequency, producing lines of 5 + 6 rather than of 4 + 7.

Unamuno tried to write blank verse, but it didn't quite work. If the cadence is strong, enjambment gets awkward. If the cadence is weak, then the line lacks a spine.

Levels of Rhythmic Perception

Imagine a listener, listening to some rather conventional jazz. Imagine that it is me, so I will call him "he."

He perceives the quarter-note pulse. 1 2 3 4 1 2 3 4... These notes are grouped into measures, and those groups are easily perceived too. He has no problem grouping these measures into four- and eight-bar phrases, and these phrases into larger song structures like AABA or the twelve-bar blues.

So he counts four levels of perception: he can easily keep track of the pulse, the measure, the phrase, and the organization of phrases into song-structures. Of course, the actual notes he hears are not just quarter-notes, but whatever notes are actually played, so the fifth level is that of rhythmic detail, the actual swung eight-notes, triplets, etc... that are played, and their relation to the pulse.

These levels can all be perceived easily at the same time because they are hierarchical (smaller units contained within larger) and interconnected. It is not like trying to keep track of five things at the same time, since keeping track of one thing (the pulse say) helps him keep track of everything else.

He can perceive these structures self-consciously, by counting to himself, or simply listen to the music and perceive them unselfconsciously. He could easily teach someone else to hear the music this way as well, if this other person did not already know how. In listening to less familiar genres he does not keep track of things quite as well, but still finds structures more or less "intelligible."

The harmony and melody are also rhythmically, structurally relevant. The listener understands some phrases as answers to others, for example, a falling melodic line as completing a rising one.

***

Suppose there is an eleven-syllable line that is transparently 11 syllables, like Lorca's "Tu cuerpo fugitivo para siempre." The rhythm is immanent, not concealed. I don't count the number of syllables, but simply fit it into a pattern I have heard many times before. Now take the line "No está en mí, está en el mundo, está ahí enfrente." I accept it as an 11-syllable line in its metrical context, but it is not quite as transparent, because it contains seven elided syllables, with elision crossing over phrasal boundaries. "No es / tá en / mi es / tá en / el / mun / do es / tá ah / i en / fren / te." These elisions of "sinalefas" create stress clashes, with heavy syllables falling on positions 123468910. The metrical accents are on 3, 6, and 10, which makes it a nicely "melodic" hendacasyllable on paper.

So in the prosodic example what is the listener keeping track of? The basic meter, the instantiation of the meter in its actual syllables, and the larger structures. Jack DeJohnette can perform as many metric modulations as he wants, as long as I can still keep track of the pulse, there is no problem.

***

If I can grasp the basic nature of the problem, then I will already understand it at an advanced level. If I can state the obvious, maybe I can see what is obvious and what is not.

Update: Areas in red show "rhotic" areas of rural England before 1950, iin answer to Sarang's excellent question.

•••

"Of hand, of foot, of lip, of eye, of brow"

There is another metrical joke of a sort. That's like the stupid high school teacher's idea of what iambic pentameter is: five feet that are also five phrases. The infinitesimal minority of lines are really like that. ("Devouring time, blunt thou the lion's paw.." That's more typical.) The joke is on the Petrarchan "blason of sweet beauty's best..." The implication is that it is monotonous and predictable.

***

The other thing I've noticed recently is Thomas Hardy:

"But that I thus found lacking in your make
That high compassion that could overbear
Reluctance for pure loving kindness' sake"

That briefest, awkward and poignant pause before the word reluctance gets me every time. You have to stop one r sound and begin another.

***

Frost's "Silken Tent" describes a tent swaying in the wind, controlled by "countless silken ties of love and thought," but bound by no "single cord." There is tension and looseness, then, a swaying or swinging motion. The tent is the poem itself. This is a kind of higher onomatopoeia, where the prosody doesn't mimic sounds, but the structure of psychic events. Williams' cat climbing down into the empty flowerpot is another example. The cat is noiseless, but the poem moves with it. Campion's "Rose cheeked Laura" does it with sound, so there's that too.

***

None of this is news, but I have to keep reminding myself of the basic thing that makes poetry valuable to me, which is that it is an art form. Even the students who hate literature (they say) respond to things like this.

Rhyme and Line Endings

Romance languages are really rhyme happy. Some day I'd like to look at rhyme, which has several functions, in terms of one of its main uses: marking the ends of lines. For today I'll just think out loud about it:

Line endings are significant for verse, because verse is defined as a series of verse-units. The end can simply be obvious, in unenjambed verse of the "single moulded" variety. In other words, in some kinds of verse knowing where the end of the line is is unproblematic.

Blank verse does without rhyme but also tends toward increased enjambment. Those two developments are practically one and the same. Heavily enjambed verse with rhyme, then, is an interesting hybrid, since the verse marks line endings while the sense is carried over to the next line. It would be possible to have rhyme that is not all that perceptible to the ear, if you enjambed enough and didn't pause at the end of the line or overemphasize those rhyming words. Rhyme would almost be "internal."

In Claudio Rodríguez, we have a kind of "rhymed blank verse," where there is rhyme, but the movement from line to line suggests blank-verse verse paragraphs. When his verse gets freer, he rhymes more randomly. Rhyme is no longer structural. It occurs where it wants to, and is often internal rhyme. Carlos Piera pointed out to me that the silva, with its combination of seven and eleven syllable words, is really the Spanish equivalent of blank verse. He didn't say that in so many words, but that's the upshot.

Classical verse is rhymeless (Greek, Latin) and sometimes constructed in paragraphs rather in sequences of end-stopped lines (Virgil.) Why is European neo-classicism so rhyme happy?

I've pointed out before that the Johns Hopkins Guide to Literary Theory and Criticism has no entry for prosody, versification, music, or poetics. It has drama theory and narratology, and Murray Krieger, and Roberto Schwarz.

To be fair, the word prosody does come up in 4 articles: Thomas Kuhn, Arabic Theory, Chinese Theory (2); versification comes up 6 times, for Poe, Jakobson, Medieval, Late 18th Century British, Spanish, and Russian Formalism, but some of these references are rather fleeting and insubstantial.

There are 8 references to rhyme.

So what exactly is "Literary Theory and Criticism"? I'd really like to know. Playing gotcha is fun. Looking at what's not visible is highly revealing.

My own termination policy would be for translators who claim to be using "loose" blank verse when they are really not using any meter at all.

Here's a really dumb question. Why is free verse unrhymed, typically? It is dumb, but important. In your answer, an essay of 500 words or less, you may consider the following factors:

The Ogden Nash effect. Lines of unequal length, if rhymed, become doggerel.

The enjambment effect. Rhyme marks line endings strongly. Free verse is associated with two traditions: the Whitmanic one of syntactic parallelism, and the blank verse of Milton with its powerful, heavily enjambed verse paragraphs.

The association of meter with rhyme. Poets throwing out meter will also throw out rhyme, because rhyme doesn't occur unless regular meter does also? Rhyme without meter is comic?

What about Creeley? He is the main free verse poet who rhymes.

The prize-winning essayist will receive a free life-time subscription to BS and SMT.

Reading Fabb and Halle, Meter in Poetry: A New Theory (Cambridge 2008). I vowed I would understand each page before I went on to the next one.

Obviously, tratándose de Morris Halle, it's a variation on generative metrics and a grid theory. The grids are composed of asterisks for syllables and parentheses facing either way for groupings of syllables. If the parentheses face one way, ))))) it's what you would call a rising rhythm, like iambs and anapests. If they face the other way, (((((( then it's a falling rhythms like dactyls or trochees. So you have (***(**( for dactyls, for example. It's basically a foot-based system then, as well as a hierarchical grid. It would have been helpful if they explained the relation between their system and more traditional metrics rather than making me figure it out.

I've long been interested in generative metrics so it's exciting to see whether or not this new theory offers just a new notation or some actually new insight. The authors are not always that good at explaining exactly why their approach is needed or superior, what it gives that other systems did not. One novelty is that it promises to be applicable to the meter of many languages. It's not just a theory of English meter, but a system that can be applied to any metrical system. That's the Chomskyan universalism in their approach.

That's as far as I've gotten. I really checked the book out to get Carlos Piera's take on Spanish meter (he contributes a chapter on Southern Romance prosody, Spanish, Italian, etc...) but since he uses Fabb and Halle's system I have to understand that first, beginning at the beginning.

The central mystery is why and how verse exists. In other words, what is the relation between linguistic prosody and literary prosody. How much of meter is derived directly and without problems from the linguistic prosody of the language in question.

Stay tuned.

Let's take another example:

Siempre la clarldad viene del cielo;
es un don; // no se / halla entre las cosas
sino muy por encima, y las ocupa,
haciendo de ello vida y labor propias

I've bolded the clashing stresses. 6 and 7 in the first line; 6 and 7 again in the 2nd, then 9 and 10 in the fourth. It's true that the preposition entre doesn't have a strong stress, but it's stlll a mouthful. Odd-numbered stresses are just normal variations when they occur on 1 and 3, but on 5, 7, and 9 they create a much more irregular effect, because they clash with the stresses on 6 and 10 (or sometimes 8).

Metrical tension increases from two causes: the variation of pauses (enjambment / heavy pauses within lines), and the variation of the metrical structure of the line itself (radical synaloepha, frequent hiatus, extrametrical accents, avoidance of regularity). Let's illustrate that with "Canto del caminar":

.... Y qué lugares

más sobrios que éstos para ir esperando.

Es Castilla. // ¡Sufridlo! // En otros tiempos,

cuando se me nombraba como a hijo,

no podía pensar que la de / ella

fuera la única voz que me quedara,

la única intimidad bien sosegada

que dejara en mis ojos fe de cepa.

De cepa madre. // Y tú, corazón, uva

roja, // la más ebria, // la que menos

vendimiaron los hombres, // ¿Cómo / ibas

a saber que no estabas en racimo,

que no te sostenía tallo alguno?

I've marked some of the stress clashes. // means a heavy internal pause. / is a hiatus, where elision does not occur. There are a few lines here that are both self-contained and more or less easy to read iambically (que no te sostenía tallo alguno). Most are not like that.

Saintsbury is the greatest historian of prosody, the one who best captured the shape and development of English verse. I don't think he's the best theoretician of prosody or scanner of lines. He was firmly in the foot-sub school.

What I really want to do is to teach the English-speaking reader to read Spanish verse, like Barzun did in his treatise on French verse for readers of English poetry. I don't know if I can do it within manageable bounds.

One way to write a prosodical history of Spanish would be to start with squibs like this (Saintsbury on Browning):

Often describe as a loose and rugged metrist and a licentious, if not criminal, rhymester. Nothing of the sort. Extraordinarily bold in both capacities, and sometimes, perhaps, as usually happens in these cases, a little too bold; but in metre practically never, in rhyme very seldom (and then only for purposes of designed contrast, like the farce in tragedy), overstepping actual bounds. A great master of broken metres, internal rhyme, heavily equivalenced lines, and all the tours de force of English prosody.

Then you could expand each entry into a short essay.

Enjambment is very different in free and metrical verse. In a metrically literate society, could have the convention that verse be written out as prose. It would make no difference, because everyone would be metrically literate enough to know where the line endings were. In free verse, the line endings make the verse: a reader could not reconstruct the lines by applying a set of internalized metrical rules.

For example, Spanish ballads are written either in lines of 8 or of 16 syllables. This creates very little difference. It's the same form no matter what. Anyone could take a poem printed as 8s and write it out as 16s with no problem. My worst student in intro to lit might make a few mistakes, but then you'd call that student not metrically informed.

Anyone should be able to write "fourteeners" out in ballad stanza typography.

I've often been fascinated by the asymmetry of Spanish meter. Lines of 11 accented on 4 break into two syllabic groups of 5 + 6.

"Mil Guadarramas / y mil soles vienen" (Antonio Machado)

Lines of 7 breaks into 3 + 4 or 4 +3.

"La sierra / gris y blanca" (AM).

The traditional combination of lines of 7 and 11 sylables, the silva, breaks into subgroups of 3s, 4s, 5s, 6s. It's a marvelous thing. A hendacasyllable with accent on 6 tends to break up into 7 + 4, the counterpoint to the 5+6. This knowledge is almost secret.

Of course they'll tell you the 11th syllable is extra=metrical. It is, for counting purposes, but I think these extra syllables have a rhythmic function. Imagine if 90 per cent of English iambics had "feminine" endings? Wouldn't that change a lot?

If someone could fully integrate linguistic prosody with poetic prosody they would win the Nobel prize or something.

Gamoneda establishes a unique relation between verse and prose in his work by lengthening the line to such and extent that it becomes a paragraph. I'm exploring his relation to Juan Ramón Jiménez, who increasingly turned to prose in his old age, converting free verse poems to rhythmic prose. So maybe Gamoneda is not so unique after all: he represents a culmination of a long process.

Gamoneda takes some more or less narrative prose and turns it into prose poetry, then takes those same sentences and weaves them into his autobiography many years later. The same phrase or sentence can shift prosodical contexts.

We don't experience Shakesperian blank verse in the theater as "lines" of verse, but as a rhythmic flow. Oral poets, even those composing in isometirc units, probably don't conceive of their poetry in terms of visual lines on the page. Even if they do, that visual representation is not primary.

As a final gesture, one of his final gestures, Juan Ramón Jiménez decided to produce a final edition of his poetic works, what he wanted to save, about 1300 poems, that suppressed line-endings in all unrhymed poems. In other words, any poem without rhyme would be printed as a prose poem. This is a radical gesture. As Howard Young put it in a review of the book--which didn't appear until 1978, years after the poet's death:

The poet who introduced verso libre / desnudo and who eloquently justified its use resorted in his last years to undermining at least part of his theory and permitting the undulating lines he had compared to bird flights or dancers' movements to flatten out in a paragraph”

I've always been shocked and fascinated by this decision. Could you imagine Williams are Creeley doing this?

The page has its own prosody, by which i mean the material support of the text: margins, paper, typeface. It isn't the same to read the same book on different texture paper.

I don't want to fetishize this, arguing that if you don't read an original edition of something you aren't reading it correctly. What I am saying is that if makes a difference to me. A type face has a voice the same way a reader of an audio-book has a voice.

I actually wish I weren't so sensitive to this. It must be nice to not see the book you're reading, to have that experience of unmediated reading.

My students in an undergraduate class, for example, could not pick out assonant rhyme in a Lorca romance when listening to Margarita Xirgú--the great actress associated with Lorca himself--read them out loud. I'm hoping that Graduate Students could do that. In this case the rhymes correspond pretty much with ends of syntactic units, sentences and phrases, and the lines are shorter.

i can hear blank verse, in the sense of hearing where the lines begin and end, even if the speaker doesn't pause at the end of lines. Or if I were reading Milton or Wordsworth written out as prose I think I could simply automatically know where the lines are just by reading it out loud. The idea that we need line-endings to perceive verse is an effect born from over-dependence on the written text.

So the visual effect of line-endings really comes into play where the line endings are not determined by inherent metrical structure. You could write out regular verse any which way and there would be no problem. The weaker or the more irregular the metrical organization, the more that visual aspect comes into play.