Verse, Prose, Speech, Counting, and the Problem of Graphic Order

by David Rothman
21 March 1997, Vol. 1, No. 1

© David J. Rothman 1997

My point of departure is a quotation from John Hollander's Vision and Resonance:

By an examination of contemporary poetry in which a purely graphic scheme of line arrangement can operate in open conflict with equally prominent phonemic ones, almost any reader can come to understand how aural and visual entities merge in status when they operate as metrical segments. (138)
Later in the book Hollander comments that we still lack "a theory of graphic prosody" (277). In recent years, several scholars have been turning to this problem and there is a growing interest in what E. A. Levenston, in the title of a recent book, calls The Stuff of Literature, by which he means its physical and graphic aspects. Levenston, Hollander, Eleanor Berry, Derek Attridge, Henry Sayre, Stephen Cushman, Richard Cureton, Richard Bradford, Marjorie Perloff and others include a wide range of phenomena under the heading of such"stuff": typography, spelling, punctuation, iconic representation, "layout" (which includes lineation), and so on. Levenston discusses the ways that graphic organization affects meaning, but still points out, like Hollander, and accurately, that the theoretical questions have never been "systematically investigated" (2). Indeed, most of those who have worked in this area either address themselves to concrete and pattern poetry; or offer catalogues of particular effects (Levenston; W. K. Wimsatt's "In Search of Verbal Mimesis"); or read individual poets (Bradford on Milton; Cureton on Cummings; Sayre, Cushman and Berry on Williams). It becomes easier to understand why the graphic dimension of verse is so difficult to describe—and has been so much avoided and maligned—if we look at the debates over verseform that have set the stage for debate since its beginning. In general, critics who investigate the role of graphic organization in poetry do so on one of two continua: between the aural and the visual; or between the spoken and the written. Those who write in the first vein rarely give versification more than cursory treatment. They descend from Lessing, and interest themselves primarily in questions of representation, the tension between the visual figure and the aural experience. Their concerns are the definition of boundaries between figural art and linguistic artifacts. Therefore, the poetry which they most often address is either patterned, or concrete, or thematically involved with visual representation, as in emblems (Dana Gioia's "The Burning Ladder," the first poem in his first book, Daily Horoscope, is a pattern poem, for example). For such "interart" critics, the graphic qualities of a poem are most interesting in terms of visuality, not language. Interart criticism with a literary bent predictably focuses on The Greek Anthology, Renaissance pattern poems, Blake's illuminated manuscripts, Apollinaire's calligrammes, and modern pattern and concrete poetry. The second group of critics, working on a continuum between the spoken and the written, is less concerned with figural art, and more with poetry as language. They therefore give conventional versification much more attention than the first group. Yet in general these writers have been uninterested in, or hostile to explorations of graphic play in verse, even when considering unconventional typography. They almost always consider writing in verse as only a way of conveying, framing, or orchestrating speech. The fundamental principle is that "[e]very work of literary art is, first of all, a series of sounds out of which arises the meaning" (Wellek and Warren 158). The source of this ancient idea—which still informs all prosodical study—lies in Aristotle's definition of language in The Poetics, which he offers in the context of a verse taxonomy:
Language in general includes the following parts: letter, syllable, connecting word, noun, verb, inflection or case, sentence or phrase. 2. A letter is an indivisible sound, yet not every such sound, but only one which can form part of a group of sounds. For even brutes utter indivisible sounds, none of which I call a letter. (Adams 59-60)
Aristotle, inaugurating western verse theory, defines language as a stream of sound that can be broken into phonetic parts for analysis. Writing is the imitation of language as sound. Verse is not a way of writing, but an imitation of a way of speaking, by definition limited to what it is possible to do with the speech stream. It only records utterance, which is where all meaning resides. Despite the differing contexts, this definition accords well with Socrates' argument in the Phaedrus about the inferiority of the written to the spoken word. Although Socrates' question is which medium is better for the transmission of philosophical truth, Plato's explicit denigration of writing lays the groundwork for Aristotle's categories:
...when [a man with real knowledge of right and beauty and good] is in earnest he will not take a pen and write in water or sow his seed in the black fluid called ink, to produce discourses which cannot defend themselves viva voce or give any adequate account of the truth. (98-99)
Rather, he will engage in dialogue, where meaning can be fixed, rather than silently dispersed. There is an extraordinary range of theories of versification, but they all derive from this ancient premise, however much transformed by criticism since the Renaissance: verse is an aural phenomenon that writing only transcribes, and, if anything, deforms. Elements of writing, such as capitalization, punctuation, margins and blank spaces (between words, at the ends of lines, between stanzas, and so on), which do not represent sounds, but which organize the page, are virtually ignored. Graphic marks are not part of the truth of poetry. Accordingly, most verse theorists transform graphic elements into diacritical marks, removing them from the substance of the poem, relegating them to the status of guides to pronunciation or enunciation. Every element of verse that we can perceive as artifice is subsumed into a general theory of spoken language, or other aural phenomena that language can imitate as sound. A modern version of Aristotle's theory appears in the seventh chapter of Rousseau's On the Origin of Language, "On Modern Prosody." Rousseau, harking back to the classical distinctions, argues that the diacritical marks which indicate pronunciation (in French) as well as prosody, are the product of "derivative languages." These are languages that have lost power (identified with actual tonal values) at the cost of a deadening rationalism (27). As evidence, Rousseau argues that the speakers of the purest languages (ancient Greek and Hebrew), had no need for accent marks as guides to pronunciation, even in verse, as we do now when we scan. The older languages had tonal values, were actually music, and all that was needed for prosody was the phonetic signs themselves: "The older and more original a language is, the less arbitrary its pronunciation, and consequently, the less complicated the signs for indicating that pronunciation" (27). Diacritical marks were unnecessary for the "older and more original" languages because their versification was inherent in natural speech. In other words, for the purest poetry the meaning of verse is the beauty of music, perceived through a pronunciation that requires no diacritical (and therefore graphic) artifice. The metaphorical "music" of verse structure was once indeed music—a music the modern languages seek to revive, but which they have now irretrievably lost to creeping intellect. Rousseau thus writes a genealogy for the troubling prosodical marks (a genealogy we now recognize as bogus), that absorbs them into a concept of pure, or unmediated poetry. According to him, a poetic language of clarity and vigor has no need of diacritical marks at all, but is sound uncorrupted by writing. In hindsight, Rousseau's ideas prefigured one rationale for free verse, which many defined as a use of language rhythms that involves no metrical abstraction from the words themselves (and therefore no diacritical marks). As Pound put it in his free-verse dictum of 1913, poets should aim "to compose in the sequence of the musical phrase, not in sequence of a metronome" (Literary Essays of Ezra Pound 3). It is ironic that this theory has been used to justify extraordinary graphic experiments (Pound, Williams, Patchen, Olson, etc.), at the same time that it leads to a denunciation of "writing" as enervated when compared to the spontaneity of music. When the subject is the overt forms of graphic play in a poem, critical hostility has only tended to increase. In Spectator 60 (May 9, 1711), Addison remarks that "[t]he Acrostick was probably invented about the same Time with the Anagram, tho' it is impossible to decide whether the Inventor of the one or the other were the greater Blockhead" (1: 255). In the same spirit, Dryden banishes Shadwell to "acrostic land" in "MacFlecknoe." The lines combine several charges against graphic play in poetry:
Thy genius calls thee not to purchase fame In keen iambics, but mild anagram. Leave writing plays, and choose for thy command Some peaceful province in acrostic land. There thou may'st wings display and altars raise, And torture one poor word ten thousand ways. (203-08)
And yet, poets have used these graphic and visual techniques in all periods. My argument is that they are only special cases of the more general play with graphic order in which all poets engage when they write. All lines of verse are forms of graphic playfulness, in addition to verbal playfulness. The aural premise also undergirds the project of prosodical investigation in linguistics. Many structural linguists denied any textual quality whatsoever to verse, systematically devaluing writing as anything other than a recording system for speech, and an inaccurate one at that:
...the only adequate record of a poem is a class whose members are (a) some one phonemic transcription of a spoken recitation of the poem, and (b) all exact copies (spoken and written) of this phonemic transcription...No record of a poem written in conventional orthography is an adequate record if judged by the phonemic standard, for, at the very least, its recording of the prosody will be inadequate. (Rulon Wells, Sebeok, ed. 198)
As Seymour Chatman stated in A Theory of Meter (1965), "in the general context of meter...the event is always aural in nature" (24). Without going into the complex details of the many linguistic theories of prosody, it is also safe to say that subsequent developments in generative linguistics have avoided the graphic implications of verse just as assiduously. W. K. Wimsatt and Monroe Beardsley responded to the structural linguists in "Meter: An Exercise in Abstraction," as did Wellek and Warren in Theory of Literature, arguing for the verse-structure of a poem as something different from any particular performance, or sum of performances of the poem:
There is, of course, a sense in which the reading of the poem is primary: this is what the poem is for. But there is another and equally important sense in which the poem is not to be identified with any particular performance of it, or any set of such performances. Each performance of the poem is an actualization of it, and no doubt in the end everything we say about the poem ought to be translatable into a statement about an actual or possible performance of it. But not everything which is true of some particular performance will be necessarily true of the poem. There are many performances of the same poem—differing among themselves in many ways. A performance is an event, but the poem itself, if there is any poem, must be some kind of enduring object...When we ask what the meter of a poem is, we are not asking how Robert Frost or Professor X reads the poem, with all the features peculiar to that performance. We are asking about the poem as a public linguistic object, something that can be examined by various persons, studied, disputed—univocally. (Gross 152-53)
Wimsatt and Beardsley accurately point out that if meter is to exist, it can only work as a non-temporal abstraction against which performances occur, much as we must have the concept of number before we can count to "5" when looking at a stack of bricks. At the same time, however, as Wimsatt and Beardsley dispute the notion that meter inheres in particular performance, they maintain that that abstraction is aural: "Milton's line is not only a visual or typographical fact on the page, but a fact of the language" (159). Thus, with the usual denigration of the materiality of writing, we arrive at the contradictory conclusion that verse structure is made out of sound, and must be understood that way, but its relation to performance is never completely utterable. The fact that this critical quandary can only be construed in the first place because of the existence of writing is not addressed. Some critics have always realized that the avoidance and denigration of the graphic is odd in an art form that takes place primarily on paper. Yet, as I indicated earlier, sympathetic criticism focuses on pattern, concrete, and caligrammatic poetry in the tradition of interart studies, and avoids non-representational graphic play. Despite the explicit graphic playfulness of much modern poetry, most students of versification still react violently against the idea of graphic play as significant in and of itself—often following the dicta of the poets. Yet when discussing versification all critics realize that, as William Carlos Williams put it in an essay on free verse, "The crux of the issue is measure" (Princeton Encyclopedia of Poetry and Poetics 289). Verse—even free verse—is measured language, and that is the key to breaking this impasse. The question should not be in what ways writing and utterance trope each other, but how both are involved with number. Without relating the technology of writing to number (as opposed to sound or drawing), it is impossible to discuss it meaningfully as an aspect of versecraft. What is needed is a way to pry apart the polar, mimetic fiction that undergirds discussions (even sympathetic ones) of writing and versification, and see how we can relate writing to measure. Roy Harris' investigations into the origin of writing make this connection possible. He argues that the origin of writing did not lie in the drawing of figures, or attempts to imitate speech, but in the recording of number. According to Harris, the oldest "writing" that we have, like that on the 11,000-year-old Ishango bone, is in "lines." The surface is scored with rows of short, parallel strokes, which probably served a numerical function. We still use such scoring systems today on occasion. Harris speculates about counting by scoring:
What is relevant for our present purposes is the fact that counting is associated in many cultures with primitive forms of recording which have a graphically isomorphic basis...The iconic origin of such recording systems is hardly open to doubt: the notch or stroke corresponds to the human finger...In short, the rows of strokes are graphically isomorphic with just that subpart of the recorder's oral language which comprises the corresponding words used for counting. It makes no difference whether we 'read' the sign pictorially as standing for so many fingers held up, or scriptorially as standing for a certain numeral. (137)
Along with other evidence, this leads him to argue that the invention of writing—or the division of writing and drawing into separate functions—occurred when the graphic representation of number shifted from the token-iterative system that appears on the Ishango bone, to type-slotting. Harris gives the following example of what he means:
The progression from recording sixty sheep by means of one "sheep" sign followed by sixty strokes to recording the same information by means of one "sheep" sign followed by a second sign indicating "sixty" is a progression which has already crossed the boundary between pictorial and scriptorial signs. A token-iterative sign-system is in effect equivalent to a verbal sublanguage which is restricted to messages of the form "sheep, sheep, sheep, sheep...", or "sheep, another, another, another..."; whereas an emblem-slotting system is equivalent to a sublanguage which can handle messages of the form "sheep, sixty". Token-iterative lists are, in principle, lists as long as the number of individual items recorded. With a slot list, on the other hand, we get no information simply by counting the number of marks it contains. (145)
When this change occurred it opened "a gap between the pictorial and the scriptorial function of the emblematic sign" (133), which had been previously inseparable in the counting represented by rows of slashes. This semiological gap made writing possible because it meant that signs could be manipulated to "slot," or identify, anything whatsoever (155). The open-ended quality of the scriptorial sign was a necessary precondition for the development of writing systems. As Harris points out, no writing system is accurately phonetic. Even the alphabet only highlights certain phenomena in the speech stream. The reason for this is that alphabetic writing did not begin as a simpler or more accurate way to record speech than other writing systems, but as an easier way to write. Through a radical reduction in the number of signs, the alphabet simplified the scriptorial system in and of itself. The evolution of writing therefore may look like this: simple forms of counting preceded the complications of pictorial representation, which in turn led to simplification of the writing system in cultures that adopted the alphabet. Discussing the power of inscriptions of number, Harris points out:
Counting is in its very essence magical, if any human practice at all is. For numbers are things no one has ever seen or heard or touched. Yet somehow they exist, and their existence can be confirmed in quite everyday terms by all kinds of humdrum procedures which allow mere mortals to agree beyond any shadow of doubt as to "how many" eggs there are in a basket or "how many" loaves of bread on the table. (135)

Or, one might add, for how many stanzas there are in a poem, or lines in a stanza, or stresses, feet, or syllables in a line, or occurrences of particular syntactical or grammatical patterns, and so on. As every serious student of versification has always understood, versification is about counting language. The reason to write verse is less to score the voice than to imbue words with the magical quality of counting. That is why meter, or measure, is at the heart of debates over all verseforms, including free verse. Number is one of the creative grounds of poetry, and the idea that writing grew out of counting is the missing link in studies of the graphic in versification. It is almost uncanny that lines of verse look exactly like the most primitive ways of counting—parallel scorings that can be numbered. Verses are countable in exactly the way that token-iterative digits are countable, from either end of the sequence. Each one indicates only its singularity, not a number. Every poem in lines effaces, or predates, the distinction between writing and drawing in the same way as the lines on the Ishango Bone; lines of verse combine functions of writing, drawing, and counting.

As Timothy Steele asserts, when citing Aristotle on the uses of meter, "measure is a fundamental need of the human spirit of the purposes of poetry is to answer this need" (Missing Measures 169). In this context we could also refer to Heidegger's "...Poetically Man Dwells...," where he asserts that "Poetry is a measuring" (Poetry, Language, Thought 221), although of course he means this in the largest possible sense. Still, the observation applies to poetic technique: verse is written language distinguished by the ways in which it continuously refers to its parts as countable. This is true even in the case of open forms, where we confront verses that are simultaneously countable (as lines) and yet, in terms of premeditated art, incommensurate in their parts (as language). In the past, the graphic could not be conceived in relation to versification because it could not be related to counting. But verse—whether free or not—achieves the numerability of its parts by being organized into discrete graphic segments that uncannily recapitulate the origin of writing (assuming Harris' ideas hold up). Indeed, any counting beyond the most basic figuring would be inconceivable without writing, as we silently acknowledge every time we resort to pencil and paper to do mathematical procedures as simple as sums, eventually coming up with words that in no way indicate prior speech. Perhaps there are a few people who could do the following procedure in their heads: 4732 x 9681 = .... That, however, is not the point, which is that even in performing basic mathematical functions our writing does not follow our speech—rather the other way around. We may perform the sentence after we do the sum, but the way we discovered that sum is through an activity that is inherently graphic. The writing comes first. This has large implications for how we understand verse. For verse is a way of writing that is as much like this kind of writing, as it is an imitation of speech. Please note this is not an either/or proposition. Both are necessary. This idea helps us to see how the theories of versification based on the Aristotelian notion of mimetic language put the cart before the horse. Alphabetic writing does not imitate utterance, but seizes upon its partial representation as a way of simplifying writing. As a system, the alphabet is a late development in writing, not the perfection of its original intent. Versification does not imitate utterance; it weds utterance to number through inscription. Therefore, although rhythm inheres in speech, it is only in a society with writing that "meter" could be construed. Even the margin, that space that has so much to do with the appearance of verse, involves the idea of marking and therefore counting. The etymology of margin is "mark," a word that still powerfully connotes number in monetary systems: Deutsche mark, ostmark, reichsmark, markka. This usage signifies what were originally lines ("token iteration") on bars of gold that indicated weight and therefore value. Similarly verse, which is immediately recognizable because of its large and irregular margins, proceeds by numberings and similitudes, by measure or the reaction against it. Writing verse construes even the margin as part of a numbered field. Verse is not language turned back into numbers, although it does look like the earliest recording of them. But the numerical patterning of language in verse encourages creative play with gaps among the aural, the graphic, and the numerical. As in the calligramme or pattern poem, the graphically isomorphic lines of notch-counting erase (or, in this case, precede) the distinction between writing and drawing. This helps us to see why the calligramme is just a special case of a more general phenomenon. All verse, whether in meter or open forms, calligrammatic, concrete, or patterned, is defined by being written in lines, which are all "pictures" of number. The hypothetical origin of writing in numerical inscription provides the Rosetta Stone for understanding one of the most ancient conundra of versification: the difference between verse and prose. Relineation is a key strategy for investigations into this difference, yet relineation has almost always been discussed in terms of the alphabetic polarity between writing and utterance— never in terms of writing itself. In a study of free verse, the critical distinction between the two approaches becomes all the more significant, as so many of the poets have been accused of writing prose chopped into lengths. Yet the way words appear on the page is crucial for metrical poetry as well as free verse and the more playful open and concrete forms. Among other things, relineation of metrical poetry into either irregular lines or prose inevitably changes relative stress values. Consider this prose passage: "No more of talk where God or Angel Guest with Man, as with his Friend, familiar us'd to sit indulgent, and with him partake rural repast..." If we rewrite these lines, the opening to Book IX of Paradise Lost, as verse, we can measure the transformation of the sensual impact of the language by looking at the varying stresses of a word that occurs three times in this short passage, "with":

No more of talk where God or Angel Guest
With Man, as with his Friend, familiar us'd
To sit indulgent, and with him partake rural repast...
In the verse passage, the second line places "with" in the first unstressed position of a blank verse line, then in the second stressed position. In the prose passage, this tension—or, at the very least, this description of a hypothetical tension—cannot exist, as there is no metrical frame. Removed from the artifice of lines, Milton's highly calibrated metric collapses; it loses the force of measure and modulation. Without the ground of counting syllables and stresses by line, the delicate commensurate relations among syllables vanishes. In addition to transforming stress values, relineation of accentual-syllabic verse into prose breaks the word limit rule, which is a graphic parameter that depends on counting. Accentual-syllabic verse is not only an organization of continuous sonic features in the speech stream. Other than in cases of hyphenation, each line-ending must coincide with a word-ending, and in many meters that must coexist with a strict syllable-count. This rule excludes many words from line-ends that would otherwise fit the metrical pattern. Blank verse, for example, is not an unbroken stream of syllables. It requires a line and word break under strict rules. Especially in the case of enjambment, the speech stream is hypothetically no more "broken" at a line-end than at any other non-junctured point in the line. Yet in metrical verse, the line has to be constructed not only according to speech stream patterns, but also word divisions, which are arbitrary in that context, and entwined with graphic representation. In free verse the question of rewriting takes on a different cast than in metrical poetry, if anything, more baffling. Consider this poem by Cummings, the first in 95 Poems:




The poem is verse, but why? The lines are obviously not metrical. The lineation breaks not only meter and rhythm, or even words, but even syllables into apparently random pieces. The verse is not only "free," in the sense of avoiding a rhythmical, metrical, or syllabic norm. In terms of any system of versification that describes verse-writing as the recording or imitation of an actual or hypothetical act of speaking, the poem is out of bounds. It cannot be said to imitate or record actual speech, or register another organic function like breath or the heartbeat. Despite the popularity of Cummings' work among composers, the only apparent musical dimension of language that this poem engages, whether in a literal or metaphoric sense, is silence, as it is manifestly unperformable as speech. Even if a speaker ignores the intrasyllabic enjambments, there is the problem of an unpunctuated sentence parenthetically enclosed between the phonemes of a single syllable. When we relineate Cummings' poem as prose, we get various possibilities:
l(a leaf falls)oneliness
l(A lleaf falls.)oneliness
L(A leaf falls.)oneliness.
L(a lleaf falls)oneliness.
And so on. What I hope these example show is that, even though the poem has no metric, a change in its "local texture" that does not change a letter still disturbs its "logical structure," and things only get more confusing when we try to punctuate it appropriately. Relineation does not simplify the problem of how to perform the poem, or how to discuss its versification in traditional ways, as Cummings has cleverly broken up the sub-verbal units of metrical verse, syllables. There is the possibility that Cummings' versification has another purpose than modulation. But this work is not a concrete poem or calligramme, manifesting the shape of an object relevant to what the words of the poem mean. To the organicist hypothesis that the poem represents how a leaf falls, in the way the reading eye twirls down the page through its narrow column (a theory that has been advanced), we can respond that this is a critical tautology, not a principle of versification. It is a willful (and unconvincing) projection of the interpretation of the words onto their unusual appearance on the page, glorified into an imitative principle that supposedly predates and informs the poem's composition. Consider that if we were to rewrite the poem





it would retain an identical shape, but we would have to come up with a different principle to explain how the eye works in reading a narrow verse structure. Does it now drop like a stone straight down the page? Rewriting free, or non-modulating verse as prose leads back to questions about graphic form as surely as rewriting metrical poetry. Admittedly, Cummings' typographical playfulness, like Milton's blank verse, is particularly amenable to raising the question of how writing and measure figure in versification, and the strange effects of relineation. Yet the questions raised by rewriting Cummings apply to other free-verse poems. We need only consider Mallarme's "Un coup de des n'abolira jamais le hasard," or Pound's and Williams' experimentation with stanza shapes, blank space, and punctuation throughout their work, or Charles Olson's playful use of the typewriter to tilt lines and entire stanzas in The Maximus Poems, to see that explicit graphic play— which does not fit with standard prosodical categories—characterizes the work of many modern poets. These examples lead us back to lineation as one fundamental criterion of versecraft. In fact, the Cummings poem foregrounds tensions that characterize all free verse, in which graphic organization into countable lines is at odds with speech that cannot be meaningfully divided into metrical segments. This is the craft of free verse, and it is as true for Whitman as it is for Cummings. Cummings is not really as special a case as he seems to be—all he's done is push things that are present in all verse a bit further, foregrounding the graphic. Those who think he is not particularly aware that what he is doing is measuring should look closely at all the numbers in this poem, from its "title" (I), to the way it splits the word "loneliness" into a small "l" and "oneliness," and so on. He is referring quite knowingly to the way that he draws upon writing and counting to forge meaning with verse. To conclude: poets write in verse partly because it excites what we could call the numerical imagination, which is both rational and superstitious, quotidian and magical. Versification is inherently a way of asserting the relatedness of words and therefore also things to one another, and it makes sense that poets, even if they define what they are doing exclusively in terms of "voice," or in resistance to "meter," would also seek to organize their writing to concentrate and take advantage of the numerological possibilities of writing, the magic of counting as it interacts with language—for writing is the crucible in which language and counting fuse. Poets therefore play with writing as they do with speech, in order to garner and concentrate as much energy for their work as possible, and the ways they do this literally are the forms of versification, both metrical and free. If we wish to understand the ways in which they do this in individual poems we need to think about number before we think about inflection, not afterwards. Given all this, I want to acknowledge some of the obvious limitations of what I have put forward in this essay. Indeed, my meditations on this subject have led me to hypotheses that I find surprising, and a number of scholars, especially Steven Willett and Richard Cureton (both in correspondence), have raised serious questions and objections to them. Among other things, Willett has pointed out that the principles of lineation which I discuss in English poetry are hardly universal, and has demonstrated how radically different the principles of graphic organization are in other languages, from ancient Greek to Japanese. Both he and Cureton also suggest that rhythmic patterning has to have some other source than artifice, in a realm that is perhaps best described as physiological. In response, I would only make a few comments which I hope will lead to further discussion. First, my aim is not to suggest that the ultimate sources of rhythm do not lie in existence itself. I am more concerned—exclusively concerned—with the ways in which artists organize them. My claim in this context is actually rather simple: poets are not only concerned with the sounds and meanings of words, but also with more abstract ways of organizing them, in particular through counting. To state it more directly: counting is an inherent part of prosodic artifice in poetry, and writing is a crucial part of that artifice. This is a different, and much smaller area of interest than the discovery of the sources and meanings of all rhythms per se. Rhythm I take to be a natural fact; but meter I take to be a kind of artifice. Secondly, it is obviously appropriate to argue that different languages will suggest radically different ways of organizing verse. Indeed, as Willett has pointed out in a lengthy letter, poets in many languages do not rely on lineation at all. At the same time, I wonder to what extent counting does figure in poetic composition in other languages, even where lineation may be of less (or no) importance. Again, my concern is with a realm of poetic artifice which I think is often obscured, even in verse study; but scholars other than I will have to analyze my linked hypotheses in languages besides the modern European ones. To close by stating these hypotheses as straightforwardly as possible: Meter has roots in the rhythmic structures of nature, but in the creation of verbal art the poet organizes such rhythms by means of an abstraction; this abstraction is counting; and counting is a technology tied to writing. In the end, poets draw on the abstract, almost magical power of counting in the same way they use figures of speech, syntax, and various prosodical features of a language: to call forth, focus, and increase imaginative power. Even prosodists have for too long primarily examined only what counting does to words, as they tend to focus on words alone; we also need to examine what counting is. In the end, this can only help us better to understand the measuring which poets accomplish. Notes

Works Cited

  • Adams, Hazard, ed. Critical Theory Since Plato. San Diego: Harcourt Brace Jovanovich, 1971.
  • Attridge, Derek. Well-Weighed Syllables: Elizabethan Verse in Classical Metres. Cambridge: Cambridge University Press, 1974.
  • -----. "Poetry Unbound? Observations on Free Verse." Proceedings of the British Academy 73 (1987): 353-74.
  • Berry, Eleanor. "Visual Form in Free Verse." Visible Language. 23 (1989): 89-111.
  • Bradford, Richard, ed. The Printed Poem and the Reader. Spec. issue of Visible Language 23.1 (1989).
  • -----. "'Verse only to the Eye'? Line Endings in Paradise Lost." Essays in Criticism 33 (1983):187-204.
  • Chatman, Seymour. A Theory of Meter. The Hague: Mouton & Co., 1965.
  • Cummings, E. E. Complete Poems: 1913-1962. New York: Harcourt Brace Jovanovich, 1963.
  • Cureton, Richard D. "Visual form in e. e. cummings' No Thanks." Word & Image 2 (1986): 245-77.
  • Cushman, Stephen. Williams Carlos Williams and the Meaning of Measure. New Haven: Yale University Press, 1985.
  • Gioia, Dana. Daily Horoscope. St. Pau, MN: Grawolf Press, 1986.
  • Gross, Harvey, ed. The Structure of Verse: Modern Essays on Prosody. Revised Edition. New York: The Ecco Press, 1979.
  • Harris, Roy. The Origin of Writing. La Salle, IL: Open Court, 1986.
  • Heidegger, Martin. Poetry, Language, Thought. Trans. Albert Hofstadter. 1971 NY: Harper and Row, 1975.
  • Hollander, John. Types of Shape. New, Expanded Edition. New Haven: Yale University Press, 1991.
  • -----. Vision and Resonance: Two Senses of Poetic Form. New York: Oxford University Press, 1975.
  • Levenston, E. A. The Stuff of Literature: Physical Aspects of Texts and Their Relation to Literary Meaning. Albany: State University of New York Press, 1992.
  • Perloff, Marjorie. The Dance of the Intellect: Studies in the Poetry of the Pound Tradition. Cambridge: Cambridge University Press, 1985.
    -----. "The Linear Fallacy." The Georgia Review 35 (1981): 855-69.
  • Plato. Phaedrus and the Seventh and Eighth Letters. Translated and with Introductions by Walter Hamilton. New York: Penguin Books, 1973.
  • Pound, Ezra. Literary Essays of Ezra Pound. Ed. T. S. Eliot. 1935. NY: New Directions, 1968.
  • Preminger, Alex, et al., eds. The Princeton Encyclopedia of Poetry and Poetics. Enlarged Edition. Princteon, NJ: Princeton University Press, 1974.
  • Rousseau, Jean-Jacques. Essay on the Origin of Languages, which Treats of Melody and Musical Imitation. Trans. John H. Moran. Chicago: University of Chicago Press, 1966.
  • Sayre, Henry M. The Visual Text of William Carlos Williams. Urbana: University of Illinois Press, 1983.
  • Sebeok, Thomas A., ed. Style in Language. New York: The Technology Press of Massachusetts Institute of Technology and John Wiley & Sons, 1960.
  • Steele, Timothy. Missing Measures: Modern Poetry and the Revolt Against Meter. Fayetteville, AK: The University of Arkansas Press, 1990.
  • Wellek, Rene, and Austin Warren. Theory of Literature. London: Penguin Books Ltd., 1956.
  • Wimsatt, W. K. Day of the Leopards: Essays in Defense of Poems. New Haven: Yale University.


David J. Rothman
PO Box 1296 / 1 First St. (UPS only)
Crested Butte, CO 81224
(303) 349-7022 (home phone)
(303) 349-2047 (office phone)
(970) 349-2780 (office fax)
73132,2550 CompuServe

Received: 19 March 1997; Published: 21 March 1997

KEYWORDS: poetic meter, versification, numerical imagination, poetic lineation, graphic lineation, poetry and counting.



What is this?