Following up on my earlier post on Nantucket, as well as all the discussion here about Ngrams, I want to offer a few spontaneous speculations based on a new paper that's the talk of at least a couple of neighborhoods in the town. That paper challenges (though is consistent with) what's called Zipf's Law: George Kingsley Zipf's principle that word-length tends to be inversely correlated with frequency of usage.
Steven T. Piantadosi, Harry Tily, and Edward Gibson -- through some very careful considerations of the Ngrams made possible by Google and also by other databases too (NB) -- argue that in fact word lengths are better predicted by their information content. Such a relation between length and information-content ministers to linguistic efficiency: you get a more constant stream of information if word length is thus correlated, that is you get more uniform information density. As they explain: "A constant information rate can make optimal use of the speech channel by maximizing the amount of information conveyed, without exceeding the channel capacity of speech or our cognitive systems. Thus, lexical systems that assign length according to information content can be communicatively more efficient than those that use frequency."
Now they accept that frequency might very well be related with the amount of information conveyed: though they don't mention Markov chains (if you've ever used the Kant or Hegel generator, or if you're a decent Scrabble player, you know how they work: they grade the probability of the next item in a list based on the previous one. Thus in English an initial S is almost never followed by an R, and a Q is almost always followed by a U), they have something similar in mind when they suggest that predictions of a common next word would mean that word would be both very frequent and not very informative.
To come up with an example that I think makes their point: If you hem you very frequently haw. Now no one haws very frequently compared to all the other things we do, but we do haw an awful lot after we've hemmed. So the word haw is a short one despite its low frequency in the wild because of its high frequency after hawing. And if you ask why "haw" in turn is a short word (since we don't actually spend a lot of time hawing), well the answer is actually that you rarely haw without hemming, so that the unit of meaning from the point of view of hawing is actually a trisyllabic one. You may find yourself going to and fro on this, but then consider that to is a very frequent word that contains almost no information, whereas fro is an extremely infrequent word that is nevertheless invariably preceded by to and.
Now naturally my own interest in these matters has to do with what light they might cast on poetic form, in particular on rhyme and meter. Piantadosi et al. suggest that the way words tend to get shortened over time is that they're spoken more quickly and truncated when they don't contain much information. As a fan of za with shrooms I don't think swallowing an idea like that's infra dig either: anyhow I'll take it on spec.
But what I want to try out here is a way that this claim might illuminate some aspects of poetic meter, at least in English. In an English iambic pentameter line, you'll find (I'm back-of-the-enveloping) that the number of syllables appearing in polysyllabic words gravitates around a mode and probably a mean of about four per line (especially if you leave out feminine endings as essentially moments of breathing); put otherwise you'd expect to find about six monosyllabic words per line. I can think of examples to quote easily: "Of man's first disobedience and the fruit...." "How loved, how honored once, avails thee not." "Yet faithful how they stood, their glory withered." "With naked foot stalking in my chamber." "O, there is blessing in this gentle breeze." "A gentle knight was pricking on a plain." "That's my last duchess painted on the wall." "And thee returning with thy silver wheels." Yes, I'm quoting from memory, but the point here is that these lines are memorable. More technically, it seems to be the case that in all languages almost all poetic lines are fewer than ten words long (hence Pope's parodic "And ten low words oft creep in one dull line"), which means that a ten syllable line usually has at least one polysyllable.
Let's stipulate (for argument and in this context) that the bi- and polysyllabics convey more information than the monosyllables, which I think is true: disobedience, honored, avails, faithful, glory, naked, stalking, blessing, gentle, gentle, pricking, duchess, painted, returning, silver. They're not the only important words, but they do look important. Now, the thing about English is that most two-syllable words are trochaic. It's hard to think of an iambic one offhand, if you disallow what are clearly prefixes which almost function as separate words. (When they don't you get words like Almost, since our minds tend to trochaize words, whether this is an overt (!) process or not, over (see?) time. Trochee's a trochee and so is iamb. And dactyl, spondee, and pyrrhic. So we can say or hazard that the more information-bearing words in a line will be trochaic, though each syllable will carry roughly the same amount of information as its fellows.
But in poetry stress matters too. Unstressed syllables tend to be spoken faster than stressed ones (there's some relation between stress and quantity), so that stressed monosyllables should carry more information than unstressed monosyllables. This seems born out by poetic form, since it's almost definitional of what we call rhyme that it begins with and focusses on a stressed syllable. (Wyatt rhymes "appeareth" and "fleeth," as Saintsbury complains, but that's the rule-proving exception, a kind of exception Dickinson will make her stock in trade. But both of them still rely on assonance in the stressed syllables.) So rhyme presents an interesting phenomenon in the context of the information theory at issue here.
I'll return to rhyme in a minute, but pause to say that it's not the only interesting question. Meter is older and more universal, so let's consider meter in English. If most bisyllables are trochaic, why are most lines iambic? Because the monosyllables fill out the lines. They frequently begin lines, as articles or pronouns or conjunctions or prepositions or even stative verbs: "A gentle knight was pricking on a plain"; "A little more than kin and less than kind"; "My mistress' eyes"; "My first thought was, he lied in every word"; "I am so lated in the world that I..."; "The glory and the freshness of a dream"; "That afternoon they came upon a land"; "Is this the region, this the soil, the clime...?" "And frost performs in these what fire in those." A line will naturally tend to start either with an unstressed syllable or with trochaic inversion, which is the most common form of variation in iambic pentameter. (More spaciously, the most common form of the first four syllables of an iambic line is a choriamb, in which a trochee is followed by an iamb: "Hail, holy light, offspring of heav'n first born"; "Whether 'tis nobler in the mind"; "Season of mists and mellow fruitfulness"; "Swift as a spirit, hastening to his task"; "Down to a sunless sea".
All Indo-Eurpoean verse seems to follow the rules of loose onsets-strict endings. "After great pain a formal feeling comes," and after "After great pain" a formal feeling comes with the words "a formal feeling comes." Lines are maximally free in their first feet, and have little freedom (except the choice of a feminine ending) in the last foot or even two feet. Why is this?
I can think of two reasons in the context of poetic information. Poetry (like music) is about the orchestration of different effects with and in counterpoint to each other. Trochaic words overlap iambic feet. But the importance of this overlap increases towards the end of the line as semantic information gives way to metrical information. If we lose the beat, we need it back, and the line has to make sure to give it back to us as it approaches its ending. The end of the line is its most crucial poetic component. (NB: A line ending may be defined as a place where word and foot-ending coincide with near invariable regularity.) At line ending, the lexical and metrical converge, when foot and word-ending correspond. In end-stopped lines, there's grammatical convergence as well, both syntactic and semantical; whereas in enjambment there's a further dimension put into play with and against the others, and we have to wait for the end of the stanza or even of the poem. Free verse, contrariwise, breaks the connection between word- and foot-ending. Anyhow, this convergence allows for the combination of different kinds of information, and produces the same pleasure of sudden economy that Freud saw at work in jokes.
In rhymed poetry the fact that certain rhymes telegraph their resolutions ("chimes" telegraphs the "sure returns of still expected rhymes") would seem to reduce the information carried by what seems to many the defining characteristic of poetry. We know what Nantucket will rhyme with. Why should that rhyme matter?
I have ideas about this, which I hope to explore later, having to do with the kind of willing that a hearer or reader directs towards a literary work. The rhyme is a sort of ratification of that will: it gives a sense that the bowling ball curved true and yielded a strike because of our body English, our active expectation that the meter would settle into a groove and the rhyme would come. In this context, though, all I want to say is that rhyming and meter smooth the information transfer too. We get to rest our interpretive abilities for a moment and coast along on the rhythmical resolution to the line. If the end-words are are expected (whether as actual rhyme words or as rimes, the rhyming part of the word, or even, in blank verse as metrical qualities), we pay all the more attention to the way we get to those words.
Maybe it would be correct to say that the smoothing of information in a poem takes place more or less in the equivalence of lines as bearers of information. Each line describes a little hyperbolic curve, discounted heavily as to meter at the start, discounted heavily as to referential information at the end. The two discount curves (forwards and backwards) produce a fairly stable average across the line, and from line to line. At any rate, this brings me to my second idea, which is that the rhyme (or last word or whatever) acts as a sort of error-checking-bit, you know the 0 through (Roman numeral) X at the end of ISBN numbers that represents a calculation on all the previous numerals and flags any mistake in any particular numeral.
This description, bloodless as it sounds, actually conforms to a lot of our naive experience of reading poetry: rereading the line to give the right metrical value to various syllables, which in turn can redound upon the meanings of words, and vice versa. I think all of what I've said should feel non-controversial, except for my lightly communications-scientific jargon. (Readers of deconstruction, I've maintained for a long time, should learn a little communications theory, which challenges the central deconstructive tenet that meaning is almost infinitely fragile. It's not.) The point is that there are a lot of different and studiously independent parameters of meaning that converge at the end of a poetic period (line or stanza or whole poem) and that convergence governs our understanding of what's come before. Different kinds of information mesh, and they do so not arbitrarily (even if rhymes are arbitrary) but through a kind of declaration that you now have all you need. The rhyme can be monosyllabic because there's really not much left to say except that the rhyme word is nailing the meter and the line, and it's the experience of nailing it that gives stability to the whole fabric of the poem.
I'm really hoping this is going to turn into less of a tangent and more of a response, but it's a bit of a tangent right now as I'm working through translating some Rabearivelo poems from Malagasy to Tagalog (with the help of French), and finding myself thinking about how default word lengths affect poetic form. Malagasy and Tagalog words tend to be much longer than French and English ones (maybe because they have fewer letters and so fewer sounds, though that's no longer the case with Malagasy, perhaps through African and French influence, I have to figure that out). In any case, the classic Tagalog meter is twelve syllables per line, perhaps to accommodate the large number of polysyllabic words. The poems I'm working on are in free verse so that's not as much of an issue.
However, the other thing your post makes me think of is what meaning not only words but specific grammatical features carry. I'm having this huge conundrum right now with an important word in the poem, "mandroso," which is the present form of "to go" or "to proceed," and is orally pronounced as an iamb with two syllables (the final o becomes silent). The problem is that in Tagalog, the present tense (or progressive aspect, if you want to get all linguisty about it) is usually four syllables, so the equivalent word in Tagalog is "tumutuloy" (two iambs). This is something about Tagalog that I find quite elegant, that the present tense is longer than the past tense (which would be "tumuloy" in this case), so there's this onomatopoeic element built into the word, but it does make for many long words. So do I use this form and disrupt Rabearivelo's rhythm, or do I use a different form that would give the word a somewhat different meaning? Oh translation.
That's completely fascinating. I think (and I think Emily Thornbury might agree) that the feel of rhythms is different in different languages. So there isn't just (or necessarily) a cross-language rule for how the length of a word and its rhythm is going to play out in a line.
But your fluency in both means that what you're hearing and regretting really must be there as an effect one language offers and one doesn't. Greek and Latin didn't really offer rhyme, for example, because of their highly inflected natures (the homoeoteleuton that Emily mentions) meant you had to avoid rhyme (what Milton called "the jingling sound of like endings"); so the question is, did the Greek experience of poetry offer something analogous to the experience of rhyme? Analogous, I think, yes, probably, but homologous? probably not.
I've tried my hand at translating a little Dante (and will again), and there the rhyme words are so important that it's just impossible not to regret the fact that their English translations don't rhyme. When I was very small I used to think that languages were isomorphic in this way, and was surprised when I realized, listening to my grandparents speaking Yugoslav, that the language was almost as transparent to me as English, but that the relations of words to each other (length, rhyme, etc.) were completely and arbitrarily different.
Manipulation of the audience's "willing", as you put it, seems like a great way of expressing the affective power of meter, and I wonder how embedded these expectations have to be in the peculiar properties of the language? In Romance-language poetry, there's a vast array of possibilities for rhyme through homoeoteleuton and (I would guess) that would shape the audience's expectations of the verse's grammar as well as its sound. And that's perhaps why poets actively avoided relying on homoeoteleuton: it too easily fulfilled a desire for parallelism as well as rhyme. Iambic meter in a language with (mainly) trochaic words would perhaps be an institutionalized way of troubling the audience's sense of what it can 'will' in a line of verse.
But I'm not so sure about the working hypothesis that polysyllables convey *more* information than monosyllables. Is "naked" really a more complex information packet than "bare" (or "nude")? "Gentle" than "kind"? Words with multiple prefixes and/or suffixes (like "disobedience") do impose layers of meaning upon a root in a way that monosyllables can't, and mentally must take longer to process. But in the context of your argument here, wouldn't that simple duration work just as well for you? Even if two words held the same quantum of information, it would take longer to access the one stored in the polysyllable, and so the tension between labor invested and meaning gained would form another element in the arrangement of the metrical line. (If indeed demand on the hearer is proportionate to word duration, then that makes what Meredith says about Tagalog verb tense all the more problematic for the translator!) [Edited to add:IOW, I'm not sure I believe that uniform information density is a tenable principle of *poetic* language.]
In Old English, by the way, the most common metrical pattern was trochaic, but rhyming passages tended to be iambic. I think this works with your theory of the 'check digit' function of rhyme; the stress pattern meant the sound repetition wasn't accidental, and couldn't be easily disrupted. I understand that Old Norse skaldic poetry took this even further, and that part of the purpose of its incredibly complex metrical schemes (which included rhyme) was to ensure stability in the course of memorial transmission.
I've really enjoyed your posts on affective meter and I hope there'll be more to come.
... and I have to think about it some more. I certainly need to know more about Old English and to learn something about Old Norse skaldic poetry.
Wimsatt (you prolly know) talks about the difference between rhyme and homoeoteleuton, and sees a studied distinction between them; Empson says something similar about Engish, that the way singulars and plurals work for both nouns and verbs ("Either the nerve swerves or the nerves swerve") prevents falling into predictable patterns based on morphemes and lexemes rather than chance phonetic echos.
About "naked" vs. "bare." Piantadosi et al.'s paper suggests that a lexicon organized as they speculate "would not be the most concise one possible as it would not shorten highly informative words, even if shorter distinctive wordforms were available." I think that Nature commentary on this gave "now" vs. "immediately" as an example. "Put your stuff away immediately" lays stress on the seriousness of the parent's command, in a way that "now" might not (unless stress were added to it in some other way, e.g. through intonation). So the longer word is likely to be more unusual and thus to convey the metainformation that the information here mattered.
I think the same is true in Wyatt: "naked foot" means "bare foot" but it also brings out the very Wyattean feel of the words that follow: "with naked foot stalking in my chamber." Naked as in like an animal stalking its prey; bare or nude would not bring that out, and bare foot would be almost a fill in. So the longer word there at least does seem to carry more information, again because there's a perfectly adequate short word which conveys the literal meaning just as well and more efficiently.
Boswell recalls of Johnson that he said of a play, "it hath not wit enough to keep it sweet," but then corrected himself: "it has not vitality enough to preserve it from putrefaction." I think maybe that a lot of lines I love get to a breakthrough simplicity at the end: that's the sudden Freudian economy.
I guess I agree that uniform density is not a characteristic of poetic language, but that the expectation of uniform density is something that we bring to all language as part of our processing skills? And that instead we get the density ebbing and flowing as it doesn't in most language. I think maybe that's where we get the sense sometimes of catching up with the meaning, as though we're timing how long we have to catch up to the end of the line. (Grammatical expectation must do this in prose as well: this is stuff I know so little about that I'm willing to flail wildly anywhere. Fun!)
A former student of mine, now in comp sci grad school, showed me some stuff, in response to my post, that she'd been working on -- I'm hoping that she'll comment here, because it's really interesting. (If not I'll at least try to get her permission to convey her ideas.)
The other thing to factor in here is entropy. There is some work on the Shannon entropy of English, since Shannon entropy has to do with information. Roughly English has a Shannon entropy of 1.6 as I recall. You know what the next letter is going to be 62.5% of the time. Shorter words have higher entropy, as do shorter lines. This would account for the sense of fitting into a groove which I also like along with the other commenters here! "Groove" would perhaps mean emergent background pattern, yes? I'm writing on this for an essay on deconstruction and ecological criticism.
Emergent background pattern: nice way to put it. Yes, though I'd see we're primed for its emergence, contribute to it.
Sravana Reddy, the student who I am still hoping will post talks about entropy some. It's almost inevitable, isn't it, that shorter words have higher entropy? Which is why my iPhone is pretty good at predicting once we get into the second or third syllable, but doesn't do much predicting on one-syllable words.
My test bed for this business is “Kubla Khan” and “This Lime-Tree Bower My Prison,” two very different poems, which nonetheless have deep thematic connections (e.g. the roaring dell in LTB and the tumultuous fountain in KK, the sun in LTB and the sunny-dome in KK). But they are quite different in matters of versification while LTB tells an intelligible narrative whereas KK is, well, just exactly how does it hang together, if at all?
The following was hastily cobbled together from notes and stuff and so is likely to be a bit disjointed, etc. For that I apologize. But nonetheless some information should tease itself out through the noise.
The sound structure of “Kubla Khan” obviously is elaborately patterned. In contrast, “This Lime-Tree Bower My Prison” is without rhyme and it lacks the “music” so evident in”Kubla Khan.” What is less obvious is that that poems have different approaches to the relationship between line boundaries of higher level constituents. In “Lime-Tree Bower” thirteen out of fifteen section breaks happen in mid-line (ll. 2, 5, 16, 20, 26, 32, 37, 43, 45, 47, 51, 56, and 59, but not 9-10 or 67-68), even including the break between the poem’s two movements in line 43. This contrasts quite dramatically with “Kubla Khan”, where all the section breaks save one (at line 49) happen at line ends.
This puts a brief pause, traditionally known as a caesura, somewhere in the affected lines in LTB. The effect is not entirely clear to me, though it would seem to create a sense of continuity across structural division. In any event, its effect in “Lime-Tree Bower” must be considered in conjunction with Coleridge’s pervasive enjambment in LTB, i.e. constructions that run over from one line to the next. In fact, the caesura section breaks often force an enjambment at the beginning of the new section. Caesura and enjambment are the norm in “Lime-Tree Bower” while they are rare in “Kubla Khan.” Is this independent of the fact that one poem is elaborately rhymed while the other is not or are these two phenomena closely related?
The issue is time, for poetry is necessarily a temporal art. We can only speak one word at a time, word after word, time after time, and we can only read by the phrase. To be sure, we can skip around in the text as we please, we are not bound to the order on the page, but we cannot evade time. Whatever their differences in theme and imagery, “This Lime-Tree Bower My Prison” and “Kubla Khan” exhibit, above all else, two different ways of conducting the trajectory of consciousness through time. Here I am specifically concerned about the relationship between verse technique and conceptual trajectory, between sound and sense.
As I’ve noted, “Lime-Tree Bower” is a narrative. It is relatively easy to follow events from one line to the next. In particular, the transition from the first to the second movement is but a return to a familiar place, the bower. “Kubla Khan” is quite different. Here the transition from the first (ending on “A sunny pleasure-dome with caves of ice”) to the second movement (“A damsel with a dulcimer”) moves you from one conceptual domain to an entirely different one. Not until line 48 (“build that dome in air”) do you come upon an obvious connection to the first part of the poem. In general, it is difficult to get a sense of what is happening in “Kubla Khan,” to feel the poem cohere from one section to the next. In this connection it is worth noting a remark that Elisabeth Schneider made about the poem in her edition, Coleridge: Selected Poetry and Prose (1951, p. 534): “Since Kubla Khan has often been read as a piece of pure music or pure magic without any rational meaning, it may not be out of place to call attention here to what the poem actually says.” She then proceeds to do so.
The point is important, so let us take a closer look at the internal continuity of the first part in each of these poems. Recall the opening of “Lime-Tree Bower:”
Well, they are gone, and here must I remain,
This lime-tree bower my prison! I have lost
Beauties and feelings, such as would have been
Most sweet to my remembrance even when age
Had dimm'd mine eyes to blindness!
In the very first line the poet contrasts his situation with that of his friends, “they.” They’re gone, but he has to stick around. He then goes on to elaborate on his situation. Of course, the referent of “they” in the first line is opaque; someone reading the poem for the first time is not going to know who they are. But the pronoun itself establishes a “place” for them, whoever they are, in the mental model that has been initiated by reading the poem. It sets up an expectation these people will be identified at some later point. That point comes at the end of line five: “They, meanwhile, / Friends, whom I never more may meet again. . . “ From there goes on to follow them in their walk, at least as he imagines it to be. Each successive phase is attached to the previous one in an obvious way.
In lines 5-9 we follow the friends along the heath and hills until they arrive at a dell; we follow them into the dell in lines 10-16, and focus on a particular sight in lines 16-20. In the middle of that line we come to a full-stop. The next sentence opens with the word “now,” preparing us for a change of narrative emphasis. We are finished with one topic and are about to consider a new one. We are still following the poet’s friends—“Now, my friends emerge”—and “emerge” tells us we are going to see something new, which we do. In lines 27 and 28 the poet introduces one among those friends, Charles, and talks a bit of him. Up through middle of line 38 each successive section is clearly linked to the previous one and shifts from one phase of the (imagined) walk to the next are clearly signaled. We break from that, just a bit, in the middle of line 38, when the poet begins to address the sun, ordering it to illuminate the scene. This is the first time the sun has been introduced into the poem, though light figured as a theme from line five, with the mention of future blindness. In line 37 the poet returns once again to his friend and the first movement concludes midway through line 43. The rest of the poem is similarly connected, each section intelligibly linked to the previous one until we arrive at the final invocation of the sun and the rook, which link the second movement to the first. (The poem has three three stanzas. The first two are one movement, bower to sun, and the third is one movement, bower to sun.)
Now consider “Kubla Khan.” The opening couplet introduces Xanadu, Kubla, his decree, and the dome. The next three lines are about Alph, the caves, and some underground sea. The only stated connection between these two topics is about place, and that connection vague. The dome and Alph and the rest are in the same place. Lines 6 through 11 would seem to be elaboration on the decree and the pleasure dome and so they are related to the opening couplet, but not to the next three lines.
Line twelve introduces a “deep romantic chasm,” but gives us no hint of how this topic is related to the those in the previous 12 lines. This is something new. Not until line 24 are we told how to integrate this new material with the old. There we learn that the fountain of lines 17 through 23 is the source of “the sacred river.” At this point we start to move back through material presented in the opening of the poem. Lines 24-28 echo lines 3-5 (the caves and sea) and lines 6-11 (five miles, sinuous rills). Kubla himself returns in line 29, but in a very different role from that he played at the beginning. Lines 31 through 36 then weave semantic elements from the opening 11 lines into a tight fabric of association that is quite unlike anything in “Lime-Tree Bower.” Then we “leap” the gap into the second part of the poem at line 37 and have no explicit connection to the first part until the mention of the dome in line 46, though we might take “symphony and song” (l. 43.) as an echo of “mingled measure” (l. 33).
Thus “Lime-Tree Bower” has a readily intelligible narrative thread that “Kubla Khan” lacks. Perhaps the confluence between sound and sense in “Kubla Khan” is, in effect, a surrogate for the lack of a narrative conceptual frame. The key point is that rhyme adds an element of predictability to the verse; one knows that sounds will repeat at regular intervals and so can anticipate them. Thus the predictability that has been “lost” because the poem does not have a narrative flow is “restored” or “compensated for” though elaborate rhyme. The temporal structure of the poem itself becomes the frame for its semantic trajectory. Imagine a version of ‘Kubla Khan” which was like the existing text in every respect except that it had no rhyme. Would that version work, as a poetic experience? Or would the poem fall apart into discontinuous fragments? Conversely, imagine a version of “This Lime-Tree Bower My Prison” which was rhymed and which lacked caesura and enjambment. Would the sound effects distract ones attention from the narrative flow?
I do not know.
* * * * *
I analyze "Kubla Khan" at some length here, and "This Lime-Tree Bower My Prison" here.
Thanks for this great analysis. I'd actually already submitted a follow-up post, which I guess will take a day or two to appear, before I read this. I talk about some similar issues there. I think I somewhere published a curious observation I made, about Coleridge and others (Spenser, Thom Gunn) that sometimes a speech tag (quoth he, said she) can take the place of a rhyme at the end of a line without our noticing it. It's as though their conventional but necessary appearance does the work of rhyme. I'll probably return to that fact in a week or two, but in the meantime there are the issues you broach here and that I take up in a somewhat complementary way in the next post and an absolutely complimentary way in this comment.
Some more or less quick remarks. What my comment amounts to is that “Kubla Khan” and “This Lime-Tree Bower My Prison” are inscribed within two different modes of (poetic) being. And, of course, the great attraction of KK is that Coleridge has framed it with a creation-story that says the poem embodies a different mode of being, that of opium induced vision.
The remark is more general and it has to do with how language works in the nervous system. Neural tissue all over the neocortex is much of a muchness. The local circuitry is much the same. The major difference between, say, areas of visual tissue and haptic tissue, is not in the processing internal to individual areas, but in the patterns of connections those areas have with the rest of the cortex and, ultimately, with sensory preceptors and motor effectors. Thus, I suggest that what’s going on WITHIN those areas subserving sound structure and those areas subserving semantics is pretty much the same. In both cases we’ve got waves of electrochemical activity. And the total effect is one of interaction between two waves of electrochemical activity. In ‘ordinary’ discourse the sound stream is functionally subordinated to the sense stream; in poetry it is not. At the neural level it’s just interactions between streams of electrochemical activity.
But what does it mean to say that one stream is subordinated to the other? That, it seems to me, is where we move into considerations of the Will and willing.
Of course it’s not even that simple. What I’m calling the sound stream is, in fact, multiple streams, and so it is with the sense streams. And I rather suspect that poetry is calling on streams not used in ordinary discourse. Language IS rhythmic through and through, but it is not necessarily metrical. The use of meter probably invokes other sound streams, and so it is with rhyme. So we’re recruiting more neural tissue to the activity, and perhaps the activity in this tissue serves to ‘frame’ or ‘cradle’ all the other activity.
A final issue: To what extent is the actual sound necessary to get the full effect of poetry? That is, to get the full effect we must either hear the poetry read or read it aloud to ourselves. We’re so used to silent reading that this hardly seems to be a question. I think it is. But I don’t yet know where to go with it.
This is very helpful, and I have found from years of questioning and anecdotal report that subvocalizers (like me) experience "silent" reading differently from more visual readers. (IIRC the split is thought to be about 50-50.)
More important, I think, are some of Ray Jackendoff's analyses about musical processing, about the kinds of autonomic or semi-autonomic processing that occurs in systems that don't lay down long term memories and the kind that that occurs in longer term registers of memory. Certain kinds of surprises never get old because we never can remember to expect them, from one thematic exposition to its recapitulation, but this non-expectation interacts differently with the part of our minds that does remember the theme when it's recapitulated. So the pleasure is in the moving target, and the movement of the target can be very elaborate. "As skaters elaborate their distances to a common end."
I seem to remember that consonants (and whispering) are in general very high frequency sounds, vowels much lower frequency, and that instrumental music tends to make use of the parts of the aural system for language processing tuned towards consonantial processing. Opera, and I guess singing in general, would combine the two in one way, poetry in a related but different way.
I may be misremembering this a bit, but I think Jackendoff's crucial point, which could mutatis mutandis apply to poetry, is that instrumental music isolates and exploits components of linguistic processing -- purifies some of its psychoactive agents. I think (metrical) poetry does this in a related way. Of course, as you say, different poems, different forms, will balance these purified components differently. The whole idea of Lyrical Ballads (e.g.) is in the idea that sometimes lyrical expression, sometimes balladic impersonality, will predominate in their commixture. A "conversation poem" like LTB is going to go towards an imitation of speech; KK to an imitation of chant.
Pentameter vs. tetrameter is crucial here: the more asymmetric the form, the more grammar bears the load of redundancy (error-checking), but also of information carrying, and the more discursive that information is. This is what I talk about in the post TK, but I'll say here that there's a relation between prime numbers and poetic forms. Odd prime forms, like sonnets (decomposing into seven paired lines) and pentameter (five paired syllables) don't balance like composite forms, and require some other scaffolding to hold them in place: grammar in the pentameter LTB vs. medial caesura in the tetrameter KK. Medial caesura in pentameter gives you two contrasting half-lines. Pope rarely used it, so that his lines are like slinkies going downstairs, the imbalance always advancing the poem, even though the rhymes are couplets, so that the balance gets re-established every other line.
More soon. And thanks. /speculation
I have a corpus of poetry -- all rhyming, mostly metrical -- at hand, so I thought I'd look at some basic properties of words at different positions within a line.
Here are the average word length and unigram entropies conditioned on different positions in short lines of poetry. (The line lengths are in number of words. The entropy is conditioned on the line position, i.e., it measures how unpredictable words are at a certain place in the line.)
So, both the entropy and word lengths are highest at line-endings. This doesn't really contradict what you said -- while line-endings may be predictable given a word to rhyme with, they do tend to be longer and more diverse than the rest of the words in the poem (like, say, Nantucket).
If we look at longer lines of a fixed syllable length --
There's this nice little wave that peaks somewhere closer to the start of the line. I think this has something to do with how, if the line is iambic, the first word will likely be a filler monosyllable like "the" or "but", followed by a trochaic words. The length and entropy don't peak until a couple of words in, because fillers are going to be immediately followed by somewhat predictable, shorter words like "only". Once you've come to the peak, you're probably saving your resources to build up to the rhyme, which is when the word lengths decrease again.
And with even longer 10-syllable lines:
That wave is still there, but it gets jagged towards the start of the line. You are mostly confined to monosyllables in these cases. As the 10-word figure indicates, there's this relationship between stress and orthography, where monosyllabic stressed words tend to have more letters than their unstressed counterparts ("And who the knights in green and what the train"), giving rise to the alternating pattern. There is also a greater number of stressed monosyllabic words at hand (as opposed to unstressed filler words), which contributes to higher entropy.
So why is the 'wave' in 8- and 9-word lines in the second half of the line rather than the first? It makes sense in the context of anticipation, and not just of rhyme (though I haven't checked to see if the same pattern exists in non-rhyming lines). That is, you start the line wit monosyllabic words fitting the metrical slots, and then begin a descent until the climax ("The joy, the grief, the envy of the plains").
Speaking of which, here are the average number of syllables at different positions, counting backward and forward from the line edges, for various line lengths:
So it looks like the syllabicity of words more or less alternates from both the start and end of the line, with the line-edge words tending to have fewer syllables, and with polysyllabic words more likely to be in the second half of the line.
And: hmmm.....
Thank you for your discussion of our work! You might also be interested in an earlier paper, The Communicative Lexicon Hypothesis which includes a study showing that across five languages, stressed syllables tend to be more informative than unstressed syllables.
Like you, I'm wary of making generalizations across natural communication and poetic form, but I think your idea that final rhymes might serve as some kind of error correction is interesting. In general, later syllables in words tend to be highly --- often completely --- predictable from the earlier syllables, and so tell you nothing more about the word's identity. One way a language might adapt to reduce this redundancy is to reduce the realisation of later syllables in speech. This could be done by moving stress to an earlier syllable, or by reducing later phonological contrasts as in the final devoicing of German. However, another route would be to harness those syllables to convey some other information, such as the identity of a final word on a previous line.
I don't know of any previous work looking at information contour with poem line as a unit, though I would imagine lines converge with sentences often enough that an aggregate measure might show you nothing beyond the information contour of sentences. Genzel & Charniak (2003) do show that later sentences in poetry contain more information than earlier ones when taken out of context, which they argue means that sentences interpreted in context are roughly equally informative throughout a text. But this is a text-level contour rather than a line-level one, and the effect is weaker in poetry than in any genre besides print advertising, popular magazines and tabloid news text!
That's fascinating. Thank you so much for the information. And of course for your stimulating papers (I am about to download the one you linked to).