The There There and Not There in the Writing of Writing: Textuality and Modern American Poetry

In Black Riders: The Visible Language of Modernism (1993) Jerome McGann proposes that “the history of modernist writing could be written as a history of the modernist book.”[1] This claim reflects his sense that the material conditions of writing and publishing—including print itself, the various processes of printing and book making, and the mechanisms for distribution that mediate between writer and reader—not only change over time but become, as they change, factors that help shape what might be termed the cognitive dynamic of writing.  If so, the technical means and cultural conventions for the production, reproduction, and reception of texts can contribute to our understanding of literary practice and meaning, because they help shape the rhetoric of a period and the ways these historically grounded rhetorics can mean.

In Black Riders, McGann focuses on how the modernist poets, especially Ezra Pound, responded to the technology and publishing institutions of print in the early decades of the twentieth century as they developed their poems and deployed their theories about poetry.  He suggests that modern print created “a social environment dominated by typographical media and publishing institutions,” and this domination meant that poets could “no longer” imagine that they stood “in the same immediate relation to their work as the dancer might be imagined to stand [to the dance]” (45).  This separation inclined the modernists, as McGann sees it, to experience the relationship “between the poet and the execution of the poem” as a “physical gap” (45), which combined with the practices of the period to shape a “culture” where print became the primary embodiment or realization of “linguistic meaning” for these poets.  And this, in turn, “foreground[ed] textuality as such” and encouraged the modernists to understand “words” (and by implication poems) as “ends-in-themselves” rather than “means” (74).  For McGann, that is, the culture of modern print “encouraged” poets to think of writing as the constructing of an aesthetic object (the poem) that readers and poets might regard and appreciate for the meaning the poem embodied more than it encouraged them to think of writing as the action of addressing the reader through the mediation of the poem in an imagined exchange or implicit dialogue.[2]

McGann’s approach clarifies features of the modernist agenda and grounds their development historically.  He is able, for instance, to demonstrate the relevance of the “late nineteenth-century’s Renaissance of printing” to modernist experiment.  William Morris at the Kelmscott Press and others who approached printing as he did emphasized “the physical presentation of texts” as “a fundamental feature of their expressiveness” (77).  McGann suggests this emphasis on poetry as writing expressed, even embodied as print (not just represented in print) contributed to the modernist tendency to “foreground textuality as such,” and he shows how the emphasis in the work of Morris and his circle on producing elegantly bound books of hand printed pages (often intricately designed) in small runs for a coterie of elite readers offered the modernists a counter-example to the book as a mass media commodity marketed to a general audience and, thus, contributed to their sense of audience and their sense of the poem’s relationship to the page.

For McGann textual media and practices are shaped by material culture and are also aspects of it.  Material culture is always historically and sociologically conditioned, and as a result the functions (and to some extent the meaning) of writing texts, circulating them, and reading them in a given period always derive in part from the terms of the material culture as it plays out through social practice.  In Black Riders McGann offers an account of what might be termed the cognitive dynamic of writing in the poetry of the Anglo American modernists and how this dynamic clarifies their poetry and criticism.  The question is whether the cognitive dynamic of writing in the modernists is true primarily for the modernists themselves or true more generally for poets of the period.  Do the modernists exemplify the period’s conditions of writing in a particularly intense, significant, and realized manner, or do they exemplify one possible approach to, or realization of, the period’s conditions of writing?  Or to put it more bluntly, are the modernists, for the purposes of criticism and literary history, so fully and comprehensively “modern poetry” that nothing else matters?  Or are the modernists a chapter among other chapters when it comes to the poetry of the period?


In Black Riders McGann leans, I think, toward the first possibility: that the approach the modernists took, in their determination to make poetry culturally significant and powerful in the modern context, reveals the cognitive dynamic of writing in the period and gives these particular cognitive, social, and technologically material conditions their fullest expression.  However, the cognitive dynamic of writing is not only a matter of how writers conceptualize writing under particular conditions but also how they conceptualize speaking (the other major modality of language) and writing’s relationship to it.  McGann points to this in claiming that the culture of modern print encouraged modernist poets to “foreground textuality as such” but leaves this point undeveloped (perhaps because it seems both temporally and technically prior to the historical developments of writing within the institutions of print that are his primary focus).  But the story of Pound’s efforts to fashion a model for poetry and its circulation that could recast the terms of modern print into high art is not only a story of how nineteenth century traditions of printing as the craft of producing beautiful pages for a coterie of aesthetically sophisticated and committed aficionados offered Pound and others a counter-example to the book as mass produced commodity for an amorphous generalized audience; it is also a matter of how his commitment to poetry as written construction (the poem as image, equation, vortex, ideogram) rather than represented speech involved not only an allegiance to specific practices of print and printing but also a rejection of the alternative possibility: conceptualizing the poem as a mode of speaking to readers and casting writing as an imitation of or representation of or version of speaking.  To appreciate Pound’s work and his influence on both poets and theoreticians of poetry we must, as McGann so powerfully demonstrates, attend to his attempts to privilege a particularly rigorous sense of the written over the merely spoken.  But it is also necessary to consider this privileging as one possible response to the period’s cultural, aesthetic, and textual conditions and to consider as well that other possible responses might also be possible and aesthetically viable.

Speaking is inherently a social process where what is being accomplished behaviorally (bonding, negotiation for position, whatever) is often more immediate and important than what is said explicitly (“Come here often?” “I saw you and George at the mall,” “Please pass the peas,” “Since we have evolved from carnivores, eating meat is morally superior to being a vegetarian”; and the like).  The give and take of speaking can unfold as a kind of dance, because speaking is interactive; it is speaking and listening by turns.  In speaking, words function more as “means” than as “ends-in-themselves.”  However, the page—both as the site of writing to an absent (and often abstract) other and as the device that preserves and circulates writing—disrupts this interactivity.  While language as writing remains a social behavior, it does so in a deferred, attenuated sense.  The page, paradoxically, is both a “physical gap” (as McGann puts it) between writer and reader and also a mediation that links them.  The “gap…between the poet and the execution of the poem” that McGann identifies as a particular condition of the modern poet is, then, actually a factor for all poets who write (as opposed to those who compose orally in actual performance for an audience of immediately present listeners).  It may well have been that the dynamics of modern print (both economic and technological) made this gap harder to ignore for modern poets, but poets who write (as opposed to those who perform improvisationally without writing) never “stand in the same immediate relation to their work as the dancer might be imagined to stand” (and even for contemporary slam poets, there is still an awareness of a gap between slam as a mode and writing for the page).

Yet McGann is right, I think, to see this more general condition of writing as having become particularly acute and problematic in the early decades of the 20th century.  While writing in general lacks the interactivity and immediacy of speaking, some modes and occasions (letters, for instance, even journals and diaries) can continue to operate as if language is a “means,” because the rhetorical situation still involves something of the sense of a socially real reader—an actual other who figures for the writer as a “you” rather than as the abstract position of “reader.”  In writing a letter, one can still imagine writing as an interaction, as a kind of speech, in spite of the page.  Similarly, some circumstances make it more possible to imagine or project writing as speech-like interactivity that can bridge the gap of writing and page that stands between poet and reader (the Renaissance tradition of writing poems to share in manuscript with a circle of intimates would be an example of this; in a quite different way, Melville’s sense that he was in part writing for and to Hawthorne as he composed Moby-Dick might be another).  This suggests that McGann’s point might be rephrased along these lines: The conditions of modern print culture—where the poem is a manufactured commodity produced for a market rather than for a specific reader or circle of known readers—intensified the “physical gap…between the poet and the execution of the poem,” which made it more likely (but not inevitable) that poets in this period would approach the page more as a space for constructing written objects that readers would later possess, while making it less likely (but far from impossible) that other poets would respond, in spite of this, by seeking ways to make the page continue to support imagined exchanges between poet (or speaker) and reader.

What I would like to suggest, partly in support of McGann and partly in opposition, is that the material conditions and social practices of the early twentieth century (in particular publishing as a burgeoning mass medium and how this was changing the relationship of writer, publisher, and audience) underscored for would-be-poets the problematic relationship of speaking and writing as modes of language.  The challenge of revitalizing poetry and making it sufficiently modern and powerful (even as poetry lost ground in the literary marketplace to fiction) meant, among other things, working out some response to the dialectic of poetry as speech and writing—both in practice on the page and in theory.  For the modernists this encouraged or conditioned or caused or became a factor in a shift toward understanding the poem as crafted visual object (with its corollary emphasis on writing as a visual system) rather than approaches that would cast the poem as the activity (interactivity) of speech given written form.  While Pound’s allegiance to the Black Riders of the printed character on the page is one particularly powerful and influential resolution to the dialectic of speaking and writing, it was not the only possible response in the modern period nor the only significant one.


The word writing has several overlapping meanings.  Perhaps most fundamentally writing is a system of notation that we can use to place and represent language on a surface such as a page (sometimes for ourselves, more typically for some absent, future reader).  By a kind of metonymical extension writing also becomes a term for things we make out of writing, i.e. texts (poems, novels, philosophical treatises, and the like).  And by a further extension—more metaphorical than metonymic—writing becomes synonymous with language, and this metaphor gathers and expresses writing’s power as a system, as an activity, and as a set of products or texts.  Used metaphorically, writing not only names a mechanism for representing language and the practices this mechanism supports, but it also comes to stand as the epitome of language, its paradigm, its fullest realization.  In this sense, to understand writing is to understand language, and it’s as if to understand language we must, primarily, understand writing.

It makes some sense that those of us in literary studies should focus on the activity of composing language for the page, on written works, and on the process of reading them.  It makes sense that we come to see writing—as the set of these three elements—as a full, transparent, even natural embodiment of language.  But like all metaphors, this metaphor distorts and conceals as well as reveals.  Through a kind of linguistic imperialism it reduces plain or ordinary speaking to a colonized territory that we value only for its potential to be more fully realized as writing.  It brings writing to the fore by pushing the possible relevance of speech and speaking into the alley behind the luxury hotel of our various critical practices.  We operate as if understanding writing in and of itself is sufficient for literary study, because we have cast speaking/listening either as something that can be ignored (once a text is written what relevance could speech possibly have?) or as something that is preliminary to writing, lesser than writing, and therefore inconsequential.

If speaking/listening were merely the frog of language awaiting the prince’s kiss to awaken redeemed as writing/reading, we could, perhaps, learn whatever speaking/listening might teach us about language by focusing only on writing.  But what if speech is an alternative mode of language with functions, qualities, and norms that differ from writing?  What if speech is language in a different sense and in different ways than writing?

While our habit may be to think that language is language, whether spoken or written (and to think of the written as more fully realized than the spoken), some of our typical usages of each show that speaking and writing can have quite different characteristics and operate quite differently.  The move and counter move of a perhaps-couple-to-be in a restaurant trading glances through the candle while chit and chatting with seeming randomness is one thing; a technical manual is something quite different.  For the restaurant couple, content counts for little; the gaps of silence (which are actually part of language in this context) can communicate as much as or more than a comment; the fragments and phrases function less as propositions or claims or representations and instead function more as gestures; the dialogue is a process, even more an interactive behavior.  In the technical manual, the emphasis is on precise, concise representation; the exposition is a linear structure of enumerated elements, steps, or relationships; indirection or nuance or innuendo is avoided or carefully and fully subordinated; the overt content is all that matters; the constructed text is not a behavior but instead a carefully manufactured product; an object.  Even more telling are those practices or occasions when we mix speaking/listening and writing/reading in various hybrids: a conference paper that has been written to be read (voiced) to colleagues knocking back lattés to stay awake; a “transcription” of a contentious but lively panel discussion where the elliptical, at times ungrammatical and chaotic give and take has had to be translated onto the page so that the written representation can convey not just what the original audience actually heard but what an engaged member of that audience might actually have understood (transcribing the tape, i.e. making it literal, written, visual, reveals that what seemed the dynamic wholeness of the actual oral performance was something in part constructed in the act listening).

Speaking/listening and writing/reading are, it seems, each modes of language that at times operate independently of each other (as speaking/listening operated prior to the development of writing) and at other times enact various linguistic tangos, where for one pass across the floor the particular dynamics of speaking (what works for the ear and how it works) take the lead and during a second pass writing (the compression it supports, the intricate subordinations it allows) defines the dance.


Historically, language as speech preceded language as writing.  We have no trouble accepting that people talked for many thousands of years before the invention of systems of marks (hieroglyphs, ideograms, alphabetic characters) that could be used to represent language visually on a surface (stone, clay tablet, papyrus, paper, computer screen) that could be preserved and even transported.  But the introduction of writing did more than provide a way to store language in visual characters.  In speech, language is an action, and each act of speaking is unique.  We can remember and imitate what others said; we can re-perform and thereby imitate something we have said; but these secondary actions approximate rather than duplicate what was said originally.  They are not the actual and the specific interaction of the original saying.  The second act of behaving in language can refer to the original act but cannot be the original act.  Writing not only records language; it also alters it.[3] Language can still be understood as an action in writing, but it is implicitly a doubled action: the act of writing and the act of reading, and these are usually separated by time and often by space.  The objects produced in acts of writing can, also, be studied in detail, word by word and repeatedly.  And there is little need to re-perform or approximate or imitate an act of writing, since its results can (unlike an act of speaking) be reproduced as well as reread (especially once printing and later various other systems of reproduction such as xerography come into play).  The process of writing and using writing as readers inclines us to see language more as a medium to formulate acts of reflection, conception, and communication than as itself (as in speech) a mode of action, a category of behavior.  Language becomes something we make with more than something we do.

The classicist Millman Parry (in the 1920s) was one of the first to recognize that the difference between speaking and writing was a factor in how cultures that had not yet developed or acquired writing and those that had differed in their strategies and techniques for producing verbal works.  Cultures where speaking/listening has been the dominant or exclusive mode of language have had to employ—as a matter of material necessity—different procedures for generating, distributing, and conserving verbal works than have cultures that use writing to preserve legal codes, religious doctrines, important practical and technical information about the environment, and (yes) the stories and poems we think of as literature.  To Parry, the prevalence of repeated, apparently conventional and formulaic epithets in the Homeric epics suggested that these poems had developed initially as an oral practice in an oral culture without the aid of writing and then later been recorded in or adapted by writing.  The poems were originally, his work suggests, not so much texts as they were unwritten, but learned, performance traditions that were refined, preserved, and circulated through repeated oral performance.  He argued that the repeated use and variation of stock phrases in these epics reflected the way oral poets drew on their command of a stock of set plots and characters, thematic motifs, and learned, formulaic phrases to recreate in the real-time of actual performance extended narratives that were too long to memorize verbatim.

Parry and those who took up his work after his death, particularly Albert Lord, tested and refined this position through their study of contemporary oral poets in the Balkans, who could neither read nor write but who could, like ancient oral poets, perform and recreate versions of epic poems from the communal stock of forms, phrases, episodes, and themes.[4] Taking his cue from Parry, the classicist Eric Havelock has considered both how the transition from the orality of the Homeric period to the literacy of Plato’s epoch occurred and its possible role in a gradual shift away from a kind of thinking in and through narrative (which he sees as a feature of the oral epic tradition) to the emergence of the kind of sustained, abstract reflection of classical Greek philosophy (and the privileging of such thinking), which he argues is both enabled by the introduction of writing and reading and in part caused by it.  Alexander Luria’s work in cognitive psychology, the work of the anthropologist Jack Goody, the developmental psychology of David Olson, John Foley’s investigations of how shifts in literary genres might relate to shifts in the role of writing and its modes of circulation all point to the conclusion that the oral creator is a performer (whether of something as sustained as epic or brief as an African American folk blues) and that those who generate oral works relate to their materials and shape their productions in a fundamentally different manner than those who deploy systems of visual characters to compose texts (whether a John Milton, a Charles Ives, or a Barry Manilow).[5]

For literary studies, Walter Ong’s Orality and Literacy (1982) has been the central synthesis of the strands of research into the nature and extent of the differences in cultures and traditions where speaking is central and those where writing is central.  Drawing on his own investigations of rhetoric, intellectual history, and the impact of writing and print on consciousness as well Luria, Goody, Havelock, and others, Ong suggests that the development of writing not only makes it possible to preserve and thereby build a larger repository of verbal works (one could not remember and recite the collected works of Henry James much less the works of the Victorian poets or the scientific treatises of Newton) but that the dynamic of speaking/listening and the dynamic writing/reading are different enough that they encourage different cognitive styles and that oral cultures and literate cultures thereby tend to develop different cultural patterns and mechanisms.  For Ong, most simply, orality and literacy are different ecologies of language, and they operate within different linguistic horizons.

This claim that the different media used for language (speaking, writing, print, word processing) might not only enable but in part encourage different styles of thinking and different cultural dynamics has been controversial.  For some it is overly deterministic.  For some it exaggerates the differences between consciousness and imaginative discourse in non-writing cultures and cultures where writing is the norm.  Both responses are, I think, at least somewhat problematic.  That the materiality of a mode of discourse (speaking, print, and so on) might in part condition the dynamic, structure, and function of that discourse and could, thus, be an element in the experience of it and contribute to its meaning is no more deterministic than the now common postmodern view that subjects do not speak and write as idealized, fully realized subjects but are instead woven into a social and cultural nexus that in part speaks and writes through them.  (Is it the seemingly greater materialism of the research Ong synthesizes that makes his position seem more deterministic and more threatening?).  Also, what some have seen as Ong’s over-emphasis on the differences between oral consciousness and culture and writerly consciousness and culture may reflect in part his decision to cast Orality and Literacy not just as a treatise on the spoken and written but as an advocacy of why we need to consider how the dynamics of language might vary culturally, historically, and for different media if we are to avoid imposing the values and logics of our own print literacy on other cultural productions, traditions, and systems.  If Ong’s rhetoric at times casts orality and literacy in opposition as a dichotomy, his actual argument is that they are (since the development of writing) a dialectic and that they have interacted in varying ratios as our practices of writing, our systems for reproducing and distributing it, and our cultural uses for it have continued to change.


For the most part studies that have explored the possible implications of the differences between the oral and the literate have tended to focus on earlier literary periods where the cultural context is still largely oral or on periods when a culture is in transition from orality to literacy (a situation that Ong has termed “residual orality”).  This reflects, I think, the sense of those who have pioneered this approach that our assumptions about language are so strongly conditioned by the dynamic of literacy (especially literacy as shaped by modern print) that we can only perceive the nature of orality and its significance in periods where orality is still distinct from, even opposed to, literacy.  Yet for Ong and others who have theorized orality, its significance is not just that it calls attention to what is not literacy but that it problematizes our understanding of literacy as well.  It is not just that orality might differ from literacy but also that the existence of this difference and its nature suggests that the cognitive dynamic of language is not fixed but shifts over time as the material media for language shift and as the institutions for its practice change.  If so, the recognition that orality and literacy differ and have combined differently in different periods and contexts points to a need to use this perspective to consider not just earlier, more oral periods but also recent periods where writers are already deeply, fully acculturated in writing and where writers seem to have no direct participation in an oral tradition or any particular investment in oral forms.

I’m suggesting, that is, that the issue is not so much whether the modality (or modalities) of language in thoroughly literate periods is of critical significance but rather how to get at that significance.  For one, while the oral may still be an element in more recent periods, “orality” as such (in Ong’s sense of it) has either disappeared or survives only in culturally marginalized groups or isolated communities within the society.[6] Writers may at times draw on the productions, the “folk” practices, of these communities and subgroups (as Ralph Ellison draws on—and riffs on—folk blues in Invisible Man), but this is more a matter of writers using these performative practices as raw material for literate compositions than an attempt to write as a folk artist or to incorporate the dynamics of folk performance directly into the act of writing.  Modern literate writers may at times represent folk material in their work, but they do not attempt to enact folk practice or processes in writing (the obvious, substantial, and significant exception to this is probably jazz, a performance tradition that enacts a synthesis of folk practice and composed art music).[7] In earlier periods the interplay of orality and literacy as differing modes and tendencies seems to make both strands and their interplay more available to analysis than with modern and contemporary literature where the dynamic of literacy seems to have become the whole show.  The work of Ong and others suggests that orality and literacy operate, intersect, and diverge in modern and contemporary practice, but these discussions seem not to include the critical and theoretical saws and hammers needed to frame the critical discussions for periods where literacy dominates.

This problem may be more apparent than real, though, if we shift our terms slightly.  Orality itself, what Ong describes as primary orality, may no longer be a significant factor for most writers of the industrial and post-industrial west, but even so, language is more than the practices, economies, institutions, and dynamics of print-based literacy for these writers.  For modern writers the issue is only occasionally how primary orality bears on literacy; more, it is a matter of how a writer’s approach to writing relates to the practices of speaking and listening as they have evolved (in a dialectical relationship with writing and reading) from the orality that preceded writing.  Even more, it is a matter of whether a writer understands the production of visual language (i.e. the procedures and visual marks of writing) as the act of representing speaking and giving it a fixed form; or understands it as a process that intensifies or distills or extends speaking; or understands it as something distinctly different from speaking.

The work of the Prague linguist Josef Vachek helps explain why our linguistic identities as writers of language remain, always, to some extent entangled with our linguistic identities as speakers of language.  Moreover, his model for how and why speaking continues to be a presence in writing systems and the use of writing systems (i.e. writing) suggests how the doubleness of language as writing and language as speech can be used as a critical tool for the study of modern and contemporary literature.[8] Ironically, part of Vachek’s agenda is to counter what he sees as the over-emphasis on speaking in modern linguistics with its corollary devaluation of writing as a distinct and significant modality of language.  In the tradition Vachek critiques, the purpose of linguistics is to understand the nature and process of language in its most primary form; language is most fundamentally the dialogue of native speakers; and speech should (thus) be the focus of inquiry and analysis.  From this perspective, writing figures primarily as a mechanism for transcribing what people say and thus figures as a lesser, secondary system to speech.  Writing can be used to represent language, to store it, to extend its circulation, and to fix it for the purpose of reflection and analysis.  But writing is not language; rather, it is a representation of language that can capture speaking with sufficient transparency that writing can be analyzed as if it were speech.  This approach, somewhat paradoxically, defines writing as a system for representing language (but not language itself) but then, by insisting on the transparency of representation, collapses writing into speech and casts the differences in media (sight and sound) and production (the direct interaction of speakers, the writer’s interaction with the page to create an indirect interaction with a future reader) as insignificant.

To Vachek, this account of writing’s relationship to speaking is fundamentally flawed.  While he concedes that alphabetic writing may well have been developed initially as a system to represent language (i.e. speech) rather than initially itself having been language (in the sense that speech is), he contends that over time writing itself becomes language in an equally primary sense (functionally) as speech and speaking.  Once writing becomes well-established in a society (especially when it begins to be used for legal and bureaucratic purposes, scientific investigation, and literature), those who write and read fluently come to use writing and understand writing as in itself language rather than as a secondary system—a representation of speech.  Once this happens, writing is no longer bound to speech but can, and does, develop its own distinct stylistic traits and norms, and neither the writer nor the reader are obligated to process the visual marks of the code as sound in order to comprehend the marks.  Written words can, that is, be read solely by the eye.  Once this happens, writing shifts from being a representation of language to being itself language, or rather it is a mode of language that operates alongside speech as a mode of language.

For Vachek, the conditions of modern print literacy drive a situation where writing and speaking are each modes of language, each with its own particular social and cultural functions, specific strengths and weaknesses, distinct norms, and differing cognitive dynamics.  Writing neither serves speech (as it does for those linguists who conceptualize speech as the paradigm for language) nor is writing privileged over speech (as if writing were the fuller realization of language and its capacities).  Rather, for some contexts and functions the medium of speaking is the more appropriate, productive, and efficient fit; for other contexts and functions the medium of writing is the better fit.  Modern literacy, as Vachek models it, involves a kind of bi-lingualism, in which people use both speaking and writing fluently, shift between the two without reflecting on their different functions and norms (at least outside the classroom), and remain mostly unaware of their differences as media (the transitory sound of voiced linguistic gestures, the stabilized visual marks of writing).[9]

Several factors condition this modern doubleness of language.  One is how we learn to read.  We are taught, first, to speak the words on the page as if they represent speech (who can forget the public spotlight when one’s turn came to stammer out “See Dick run.  See Jane run”).  But once we can voice the words from the page, we are pushed to master what we are told is true reading, silent reading.  To be sufficiently efficient readers we must skip the step of converting the visual code into its voiced equivalent and treat the word as a self-sufficient visual symbol.  This transition shifts the emphasis from expressing words derived from the page to instead comprehending the words we see on the page, and the measure of reading becomes “reading speed” and “comprehension” (and comprehension figures primarily as a matter of tracking content rather than interpretive subtlety or power).

Another factor is how and when we use the two modes.  Speaking remains the more useful approach when we need language as social behavior, as interaction, which is why, for Vachek, the norms of speech are relatively informal and allow for considerable variation.  In writing, the audience is usually not present (unlike listeners, readers are often not only absent but are often an abstract position or function reflecting particular conventions for how certain kinds of texts are to be composed and read).  In writing, the emphasis is more on representation, exposition, analysis, and the like; and the norms are more formal and stringent in order to support speed of reading and efficiency of comprehension.  In speaking, slurring, even eliding, a word is fine as long as the verbal gesture remains clear; indeed slurring and eliding can function expressively and add force to the gesture.  In writing, the equivalent tactic of misspelling or deliberately respelling a word is usually “read” as an error, not an expressive device, because it disrupts the reader’s processing of the marks on the page that are, visually, the word.

One might, here, invoke Twain’s use alternate spellings to project the illusion of Huck’s vernacular voice as a “Yes, but” to this point.  However, the humor and expressive nuance in Twain’s style is possible precisely because the conventions for writing/reading that are part of circulating texts in print establish the expectation that spelling will be consistent, correct, and thereby what amounts to neutral.  We typically “see” the letters of prose as if they evoke words that are either silent or are pronounced as we are used to hearing them (which is, stylistically, akin to being silent).  Twain’s distortions of the code trick us into hearing the words and into hearing them in particular ways, and in the process his writing destabilizes the relationship between speaking and writing.  It is precisely Twain’s presentation of Huck’s speech encourages us to treat the visual code of the page as something more like the speech from which it originates that the written norms can be partly set aside.  Chatty letters and dialect in novels reflect not only how easily we move back and forth between speaking and writing as modes of language but also how these two modes (though distinct in their tendencies) can interact and blend as part of the condition of modern literacy to support such hybrid practices as the conference paper (to which I could add “noted above” if I wanted to invoke the model of writing or to which I could add “as mentioned earlier” if I preferred invoking the model of speaking).

The example of Twain points to a second factor that contributes to the modern doubleness of language—at least when the writing system is alphabetic.[10] The set of letters that are a written word can convey something of the sound of the word; the letters can, that is, function as an image of the spoken word.  But even when words are spelled correctly (i.e. per current convention), if we imagine them as spoken or voice them to ourselves as we read, we probably hear them through our own habits of pronunciation (I grew up hearing “wash” pronounced “warsh” and tend to hear the word on the page that way, unless I remind myself that most people would hear it “w-ah-sh”).  To convey dialect we manipulate the alphabetic code to evoke these non-normative sounds (“I’m going to warsh the dawg,” he said, as he kicked at the loops of hose in the wheel high grass, slipped in the mud, fell on his butt, and the dog squirted out the gate.).  But most writing does not involve manipulating spelling to convey non-standard pronunciations: “wash” is “wash.”  That two readers might hear it differently from the page is generally of little importance, in part because such pronunciation differences have little impact on the function of the word as a unit of visual language, however much the pronunciation of the word when spoken might convey information about social class and education.  Most reading is silent, and differences in pronunciation aren’t factors in silent reading.  In writing, that is, the set of letters that make a word can function either as an image of sound, in which case the letters point to the spoken word (and the spoken unit is the real word), or the unvoiced letters together become a visual sign, and this unvoiced visual sign is the real word.

This distinction is clearer if we consider the conventions for writing numbers.[11] In English “two” and “2” can be swapped about without changing meaning, but “two” has a different relation to sound than “2.”  We could, for instance, imagine the comment: “My old math teacher loved the sound of numbers.  I can still remember his ooing sigh of ‘two’ in ‘one plus one is twooo.’”  But we could not evoke the sound of the final word using numerals (“222” would not be “twooo”).  Similarly, “2” has the same meaning in English, French, German, and so on but the words in these languages that convey “2” alphabetically (“two,” “deux,” and so on) are pronounced differently (“two” has meaning in English, “deux” does not, unless it occurs in a technical term borrowed from French, as “pas de deux” in dance).  For the linguist Roy Harris this capacity of writing to function as systems of visual marks that do not have to be translated back to or through the sound of speech is a fundamental quality of all writing systems.  It isn’t just that “2” functions primarily as a visual unit of meaning but that “two” in English and “deux” in French also come to function primarily as visual units of meaning in the writing and reading of the two languages.

The recent fad of custom license plates illustrates Harris’s point.  One could, for example, express “I am before you” in license plate as “I M B4 U.”  Here, the characters convey the phrase, but the usage is eccentric enough that the characters function neither as immediate visual symbols of the intended words nor as direct transcription of sound.  Instead, they are a kind of low order puzzle, where we use the visual marks to imagine the sounded words, then jump back to the intended words as visual signs.  The common bumper sticker, “I ♥ NY” involves a similar process, though one we “read” more directly because the ad campaign that popularized the visual slogan has established that “♥” is “love” and not “heart” and the “NY” for “New York” is a common substitution.  In the case of “I ♥ NY” the visual code is transparent enough that the phrase can function immediately as visual language without the support of the sound of the words.  (This visual code does not, though, function equally well for all states.  “I ♥ ME” can be read either as “I love Maine” or as “I shor dew think I’m pretty hot stuff” depending on context, and “I ♥ ME” can also be read as parodying “I ♥ NY” and thereby as a humorous comment on this particular visual system.  My recent misreading of the plate on a large blue van, in which I initially took BLUBRY to be “blubbery” when “blueberry” was probably intended also shows the kind of slippage that can occur between the visual code and what is being represented.)  The process of decoding license plates, whatever pleasure it adds to rush hour traffic, points to the value of conventional spelling: it enables the reader to perceive and process the written marks directly as visual units (words) without having to process the marks into an image of sound and then judge from that aural image what word is actually intended.  Spelling conventions, as Vachek has stressed, make reading more efficient.[12] They contribute to our being able to process written words without bringing them to mind as sound and speech, and this contributes to writing being able to function itself as a mode of language that operates directly and fully as a visual system and not just as a representation of language as speech.

Vachek notes another dynamic of conventional spelling that reflects how it enables writing to function as visual language and not just as a visual representation of speech.  In writing, the need for efficient, consistent visual cues supports not just the regularizing of spelling but also promotes the use of different spellings for words that have the same pronunciation but different meanings or functions.  Thus, the words “two,” “to,” and “too” have different spellings precisely because they sound the same, and the numeral “2” can assume the role only of the first of these (unless, that is, one is writing in license plate).  One might think also of the visual distinction between “it’s” and “its,” which has no auditory equivalent, (which is probably why students so often confuse them in their writing).  And then there is Gertrude Stein’s “There is no there there,” which turns on the several functions of “there” and how the spelling, in this instance, fails to differentiate between them.  For readers who move easily back and forth between writing as representation of speech and writing as visual language, the ambiguity of “there” in Stein’s phrase functions as a kind of visual pun, one that requires us to hear the inflection that the words would have as a gesture in speech to distinguish “there” from “there,” even as the abstraction of the phrase is more typical of writing than speaking.  Stein’s phrase, in effect, stages the doubleness of language as speaking (aural action) and writing (visual code), and this has something to do with the way it is simultaneously simple and complex, banal and resonant.  It is also, I think, why inexperienced readers often have trouble with it.  Inexperienced readers are more used to reading silently and aren’t sure how to adjust to the kind of hybrid processing that this phrase encourages (requires?).  This, at least, would explain why these same “readers” can usually distinguish “there” from “there” quite easily if someone reads the phrase aloud with the appropriate inflection, and it suggests that these same readers would not be thrown by “There is no there there” if “there” had two spellings for the two functions (say “ther” and “there”).  Stein’s phrase could then be written “Ther is no ‘there’ there” or even extended as “Ther is no ‘there’ over there in their there” and inexperienced readers would probably still track it precisely because there would be little need to draw on the possible sound of the phrase and how it would be inflected in speech (and how this inflection would substitute for the device of distinct spellings for distinct functions in writing).

There is another aspect of Stein’s play on “there” that should be noted.  In conversation distinguishing words that sound the same from each other is usually simple.  When the speaker and listener are both physically present, the rhetorical situation is typically specific, concrete, and unambiguous; and physical cues (posture and gestures) and auditory cues (tone, inflection, pace) help us distinguish one “there” from another (and either “there” from “their” and “they’re”).  In writing, the rhetorical situation is less immediate, often more abstract, and at times ambiguous; and the shift from sound to visual marks erases or attenuates the physical and auditory cues (though they can, as in Stein’s phrase, be implied).   Spelling distinctions, though, often help differentiate words that sound the same—but not always.  Stein’s phrase works in part because spelling does not distinguish one “there” from the other, and this ambiguity (in writing divorced from speaking) underscores the difference between speaking and writing.  Even more it underscores Vachek’s point that writing is itself a doubled system.  Writing can function, that is, as a representation of speech (and thus as an analogue to speech as language), and it can function as a fully visual system of language where sound is irrelevant.  The play in Stein’s phrase, then, is less a matter of the slippage between speaking and writing than a matter of the slippage that can occur between writing as a representation of language as performed speech and writing as itself a system of language.[13]


The pragmatic nature of most writing means that the doubleness implicit in Stein’s phrase is, and should be, of little importance.  For most occasions (scholarly articles, proposals, technical reports, attempts at persuasion, historical narratives, etc.) we write for the convenience of the eye and for the efficiency of whatever generalized version of the reader we project as the audience.  Or we optimize the text for the ear when the occasion (political speech, conference paper, banquet introduction, etc.) calls for that.  Either way, the functional occasion determines the approach to writing, and Steinian “there’s” are to be avoided.  But even so, the primary doubleness of language as speaking and as writing and the way writing can function both as alternative to speaking (i.e. visual language) and as representation of speech is a factor in literature, and not just for experimental figures like Stein.  For one thing, while the occasions and processes of writing poems, novels, creative non-fiction, and the like are as socially and politically situated as the more immediately functional occasions of writing professional talks and technical reports, literature is less a matter of recording or communicating or persuading and more a matter of engaging, evoking, and expressing.  In literature the need for language to be resonant competes with the more restrictive norms of precision, efficiency, and clarity.  In literature we often push the boundaries to make language do more or do new things; verbal play is allowable, even desirable.

Whatever the precise reason or reasons, in modern and contemporary literature the doubleness of writing as representation of speech and writing as visual language is a fundamental aspect of the linguistic horizon, just as the opposition between the oral culture and literate culture was part of the linguistic horizon for the writers of ancient Greece and medieval Europe.  And this is especially true for poetry, which (unlike the novel with its roots in the silent reading of print) emerged first as epic poesis—a mode of performative speaking.  Even after poetry became a written practice, the use of formal sound patterns (meter, rhyme, alliteration, etc.) remained the norm—in some contexts to enhance the text for oral performance, in some to support memorization, in some as the mark of the poet’s technical skill.  Writing and print changed the ecology of poetry, but traditionally sound has been so fundamental to poetry that even free verse poets who in part use the visual space of the page to measure their lines (and whose readers often read the poems silently) have continued to believe that the words in a poem are at least partly aural and should be voiced rather than treated as primarily visual cues as in expository prose.[14] Robert Creeley’s brief “I Know a Man” illustrates the odd mixture of the aural and visual that can result:

As I sd to my
friend, because I am
always talking,—John, I
sd, which was not his
name, the darkness sur-

rounds us, what
can we do against
it, or else, shall we &
why not, buy a goddamn big car,
drive, he sd, for
christ’s sake, look
out where yr going.[15]

The poem offers a brief conversational exchange, an act of speaking, but it does so in a curious way.  Some details, such as the way the line break stretches and inflects the word “sur-/rounds,” suggest that Creeley is more committed to writing as representation of speaking and voice than writing as visual system.  Yet “&” is a purely visual element (akin to the ideogram that for a while named the artist formerly and now, at least the last I knew, again known as Prince).  It needn’t be pronounced or heard to convey the function “and.”  We “hear” it in Creeley’s poem as “and” because the poem seems to dramatize two acts of speaking: the speaker’s exchange with the John who is not John and his narrating this exchange to the reader who is, seemingly, to play the role of listener.  Similarly, the abbreviations “sd” and “yr” function visually; they do not as fully record the sound of “said” and “you’re” as the full spellings do.  Yet because the poem is so clearly speech and speaking, the abbreviations come to function more as a sign of the informality, the immediacy, even the colloquial edge to the voices of “John” and the speaker, as if “yr” were “yer” or “yur” and not the more fully, carefully enunciated “you’re.”  The abbreviations “sd” “yr” “&” are also the sort of shorthand one might use in writing a quick note to one’s self, and this informality of the visual inscription encourages us to hear the words they point to as being spoken in a similarly impromptu manner.  Creeley is, that is, manipulating writing as visual system to evoke registers of speech rather than using writing as a direct representation or image of the sound of the words the two figures exchange.  Creeley probably didn’t think of “I Know a Man” as an experiment in, or demonstration of, the doubleness of writing as represented speech and visual system when he was writing it.  Rather, the poem shows him drawing on and manipulating the stylistic resources for poetry at the time he wrote the poem (1955), including the example of Pound’s earlier use of such “visual” abbreviations as “sd” to convey spoken words.  Still, the poem’s stylistic features and how and why they work make more sense if we recognize that the poem manipulates this doubleness.

In “I Know a Man” it is unclear whether Creeley’s allegiance is more to the act of speaking in and through poems or to poems as constructed objects of visual language (i.e. writing as itself language rather than writing as represented sound) that might include represented moments of speaking in them.  Or perhaps for Creeley the formal significance of the poem would be precisely its ability to function as both simultaneously or as a hybrid of the two.  For Ezra Pound, though, his campaign to rescue poetry from effete sentimentality and make it a thoroughly modern art seems to have involved a commitment to writing as a visual system of language, not writing as an analogue to speaking.  For Pound, it seems, the sound of writing had poetic value, but the poem was neither a mode of speaking nor speech.  His “Treatise on Metre” from ABC of Reading, points to his sense of poetry as what might be termed written or composed sound distilled in visual form:

In making a line of verse (and thence building the lines into passages) you have certain primal elements:

That is to say, you have the various “articulate sounds” of the language, of its alphabet, that is, and the various groups of letters in syllables.

These syllables have differing weights and durations.

A. original weights and durations

B. weights and durations that seem naturally imposed on them by the other syllable groups around them.[16]

By arranging these “weights and durations” the poet “cuts his design in time” as if “TIME” were a surface (i.e. as if “TIME” did not unfold but were instead static, which requires converting time into space, which in turn is a matter of conceptualizing time as if it is a matter of conceptual and imagistic relationships that exist beyond time and are thereby “in TIME” because cut into space but beyond “TIME” because stabilized, even frozen, as “design”).

For Pound the goal in handling these “weights and durations” is not to present or intensify the speaking voice and its cadences.  Rather, his goal is to use the sounds, patterns of stressed and unstressed syllables, and the relative duration of syllables as encoded in the rules of the writing system and written words as elements in the “design” of the poem.  For Pound, the explicit analogy for meter in poetry is the role and treatment of rhythm in music:

Rhythm is a form cut into time, as design is determined space.

A melody is a rhythm in which the pitch of each element is fixed by the composer. (199)

In the first of these comments, though, Pound seems to be treating rhythm as a visual matter, something akin to sculpture or perhaps a painted surface, more than he is treating it as something that unfolds in time, the way a melody does as it is played or the way the poetic line does as the eye moves across and down the page or as the ear hears either the poem or piece of music in performance.  His comment, that is, treats “time” metaphorically as a physical substance or surface (something that is outside of time as the unceasing unfolding of moments, something that can be cut, something that can reflect that cut to the eye), and this “TIME” as physical substance or surface then enables and preserves the poet’s design.

Pound’s image of rhythm is as much (or more) visual and spatial than it is aural.  One could dismiss this seeming difficulty as nothing more than focusing on the wrong part of the game.  He is stressing here, after all, the process of composition, not the later reading (or playing) of what has been composed and notated by the system of the visual code onto the page, and he is talking (rather writing) metaphorically.  Still, Pound in his Imagist and Vorticist days had little use for the ambling discursiveness of a Wordsworth, shuddered at the chanted oratory of a Whitman, and blushed in exasperation at the rhetoric of Victorian verse.[17] He called for a practice in which each word is necessary to the distilled, compressed, multivalent image that had been incised onto the page and which the page in turn offered to the reader.  And this commitment to writing as a visual system, rather than writing as a recorded act of speaking, and thereby in turn to the poem as composed system rather than the poem as spoken act, is apparent in the key essay “Vorticism,” from spring 1914, which he included as chapter XI of Gaudier-Brzeska: A Memoir.

In “Vorticism,” Pound justifies “Imagisme” in part by explicitly setting it in opposition to “rhetoric” and in part by aligning it with visual art.  He asserts, for instance, “The ‘image’ is the furthest possible remove from rhetoric.  Rhetoric is the art of dressing up some unimportant matter so as to fool the audience for the time being.”[18] Whether poetry can so simply escape the rhetorical dimensions of language, of writing, by simply defining rhetoric in this extremely narrow and reductive manner is open for debate; Pound seems not to admit the possibility that in rejecting one rhetorical mode for poetry (which he sees exemplified in the work of such figures as Tennyson) that he is simply opting for a different rhetorical mode (the rhetoric of the modernist approach to poetry that he is advocating).  One is tempted to observe that Pound’s “rhetoric” in this key move has proven persuasive, even as (or because) it is misleading.  In any case, as Pound notes, in his view “there is another sort of poetry [i.e. a poetry of the image] where painting or sculpture seems as if it were ‘just coming over into speech’” (82).  This second comment could be read as casting poetry as speech or speaking.  Rather, it casts poetry as what is prior to, more fundamental than, and ultimately independent of “speech” and speaking.  The reader may, in effect, speak in response to the experience of the image the poem embodies and operationalizes, but this response by the reader to the poem as crafted, compressed, and realized aesthetic object is necessarily belated and secondary to the poem itself.  In Pound’s imagining of the imagist poem, the poem is not simply timeless, it is beyond time—both the time of its making and the time of the reader’s reading.  To speak from or to the image (poem) “just coming over into speech” is to lapse back into time.

Even if this pushes the implications of Pound’s flight from the rhetorical and from speech and speaking too far, it is still clear, I’d suggest, that the effect of the imagist poem as Pound imagines it in “Vorticism” does not come from it being an act of speech, nor is it a matter of the poem presenting a speaker embodied within the poem who speaks to an auditor within the poem, so that the poem, for the reader, becomes an overheard dramatic episode.  It is even less a matter of the poem representing the poet as speaking to the reader (unless it is in the same sense that a painter might be said to paint to the viewer).  Either of these would, in Pound’s sense of the term, condemn the poem to being merely “rhetoric.”  Instead, the effect of the poem is to inhere in its construction and how the poem—as constructed object—creates energy for (or transfers energy to) the consciousness that contemplates the poem.  In Gaudier-Brzeska, Pound introduces “Vorticism” by quoting himself from the first issue of the journal Blast:

Every concept, every emotion presents itself to the vivid consciousness in some primary form.  It belongs to the art of this form.  If sound, to music; if formed words, to literature; the image, to poetry; form, to design; colour in position, to painting; form or design in three planes, to sculpture; movement, to the dance or to the rhythm of music or verses.

I defined the vortex as “the point of maximum energy,” and said that the vorticist relied on the “primary pigment,” and on that alone. (81)

Sound is a factor, it seems, but a secondary one: rhythm is a quality of “verses” but “literature” is “formed words” and real “poetry” (perhaps as distinguished from “verses”) might be said to be formed of “formed words” arranged into image.  The poem, that is, is an image or set of images that aims to work for and through the eye rather than through the representation—and mediation—of socially or dramatically or rhetorically situated speakers and speaking: “The point of Imagisme is that it does not use images as ornaments.  The image is itself the speech.  The image is the word beyond formulated language” (88).  For Pound, that is, the crafting of the poem is not a process of speaking, and the poem transcends speaking.  For Pound, at least as Imagist and Vorticist, the image is a written construction that operates, finally, in opposition to the process of speech and speaking.

In “Vorticism” Pound also describes the composition of “In a Station of the Metro” to illustrate the project of Imagisme.  His account underscores his commitment to writing as in itself language rather than writing as a code that stores speech.

Three years ago in Paris I got out of a “metro” train at La Concorde, and saw suddenly a beautiful face, and then another and another, and then a beautiful child’s face, and then another beautiful woman, and I tried all that day to find words for what this had meant to me, and I could not find any words that seemed to me worthy, or as lovely as that sudden emotion.  And that evening, as I went home along the Rue Raynouard, I was still trying and I found, suddenly, the expression.  I do not mean that I found words, but there came an equation…not in speech, but in little splotches of colour. (86-87)

For Pound narrating the experience would not yield a poem.  Instead, the poem happens through the discovery of an “equation” that abstracts the experience from its unfolding as an emotional encounter in time and transforms it into a relationship that is visual and has the simultaneity of the visual elements in a painting.  Pound describes the actual writing of the poem (as distinct from the experience that was its occasion but which the poem, he would have us believe, completely supercedes) this way:

Victor Plarr tells me that once, when he was walking over snow with a Japanese naval officer, they came to a place where a cat had crossed the path, and the officer said, “Stop, I am making a poem.”  Which poem was, roughly, as follows:—

“The footsteps of the cat upon the snow:
(are like) plum blossoms.”

The words “are like” would not occur in the original, but I add them for clarity.

The “one image poem” is a form of super-position, that is to say, it is one idea set on top of another.  I found it useful in getting out of the impasse in which I had been left by my metro emotion.  I wrote a thirty-line poem, and destroyed it because it was what we call a work “of second intensity.”  Six months later I made a poem half that length; a year later I made the following hokku-like sentence:—

“The apparition of these faces in the crowd:
Petals, on a wet, black bough.”

I dare say it is meaningless unless one has drifted into a certain vein of thought.  In a poem of this sort one is trying to record the precise instant when a thing outward and objective transforms itself, or darts into a thing inward and subjective. (89)

The “(are like)” that Pound adds to the poem of the Japanese officer indicates that he sees the haiku as a kind of equation—an equation that can (his last comments suggest) fuse perception and consciousness and communicate them directly to the reader’s consciousness without recourse to an embodied consciousness speaking about it.  The poem is “cat steps on snow/plum blossoms” more than it is “I see cat steps on snow, ah, and think of plum blossoms.”  And Pound’s particular reading of the officer’s poem, the way “(are like)” becomes the pivot, points to “In a Station of the Metro” as neither about the Metro nor as the story of his attempts to construct the poem but rather as an image/equation, in which his own emotions as he perceived the beautiful faces are no longer directly relevant or directly present.  As Pound would have it,

The difference between art and analytical geometry is the difference of subject-matter only.  Art is more interesting in proportion as life and the human consciousness are more complex and more interesting than forms and numbers. . . . By the “image” I mean such an equation; not an equation of mathematics, not something about a, b, and c, having something to do with form, but about sea, cliffs, night, having something to do with mood. (91-92)

The experience is the poem’s occasion, but the poem neither refers to nor celebrates nor comments on the experience and its emotions; these merely contingent matters, the chance of circumstance, have been abstracted out, so that the distillation of the poem can more purely enact/express/be the image/equation of the emotions.

Pound’s account of writing “In a Station of the Metro,” his insistence that poetry should be at least as well-written as modern prose, his interest in the ideogram, all reflect his conviction that poetry should privilege the compositional precision and clarity of writing as visual language over the behavioral immediacy of language as speech and spoken interaction.  But it is worth noting how easily “In a Station of the Metro” could be transformed into a poem that would resolve writing’s doubleness as represented speech and visual system quite differently.

The way Pound transforms his experience into an equation that the text enacts as if directly means that the poem has no speaker in the usual sense of the term.  The phrases are not spoken by a figure of the poet or a dramatic character, even though they are carefully cadenced and musical (especially the second line where the percussive first syllable plays out through progressively longer, more open syllables).  The poem is not, that is, speech or represented speech, in spite of its aural dimension.  Rather, it is writing that inscribes the perceptual equation on the page, and the visual space, the indentation, that opens the second line in its appearance in this essay, is as much a visual symbol expressing function (“=” or perhaps “≈” or perhaps something on the order of “→”) as it is an indication of a pause.  It is equivalent to the  “(are like)” of the officer’s poem.  We could, then, as Pound did with the officer’s poem represent the poem this way:

The apparition of these faces in the crowd:
(are like) Petals, on a wet, black bough.

And the equation (as Pound defines it in his essay) would still be intact.  It would still be, functionally, the same poem.  To write the poem this way would, however, alter it fundamentally:

The apparition of these faces in the crowd:
(Ah!)  Petals, on a wet, black bough.

Adding in “Ah,” or for that matter “Ooh” or “Oh” or “Zowie,” transforms the poem from written equation into various gestures of represented speech.  It shifts the poem from a moment of distilled consciousness that can (in turn) elicit heightened consciousness for the reader into a dramatic occasion of apprehension and emotional response to which the reader might respond with recognition, empathy, and the like.  It turns the poem into a poem that more directly, in fact, reflects the story Pound tells about the experience that led to the poem.[19]

Clearly such a recasting (if we believe the argument in “Vorticism”) would not have pleased Pound.  But even so, adding an “Ah” to the second line wouldn’t necessarily make “In a Station of the Metro” a lesser poem.  Rather, it would make it a different kind of poem, one operating as a kind of speech and speaking rather than visual composition, as the example of Edna St. Vincent Millay’s almost equally brief “First Fig” suggests:

My candle burns at both ends;
It will not last the night;
But ah, my foes, and oh, my friends—
It gives a lovely light!

For Pound the page is a visual field for staging writing.  The language of the poem is visual, and even the unwritten (unprinted) surface of the page functions as an element of the crafted object (the two small lines of “In a Station of the Metro” against the empty page are a visual cue that the poem is a particularly distilled and energized moment; the space that opens the second line—typically omitted in reprints of the poem—is the “=” or “≈” or “→” that links the two units of the equation).  And the poem is a crafted object, an aesthetic object, mounted on the gallery wall of the page for our readerly attention: in the explanatory prose “Vorticism” Pound shares the anecdote that cues us to his experience of the presumably actual moment behind the poem—that moment of beauty breaking through the banal and ordinary.  He shares it, though, not as an explanation of the poem or as a further gloss upon it but shares it almost as the curator of an art exhibition might offer context to the amateur art patron so that the patron/reader can more fully recognize and appreciate how much the moment of actual experience has been fully transformed into, replaced by, and transcended by the timeless aesthetic construction that is the poem (in Pound’s paradigm the experience from which the poem is made, the making of the poem, and the study of the poem occur within time, but the equation that is the poem and the grasping of that equation stand as if beyond time, which is why the poet cannot be figured as present (either directly or as if directly) in the poem.  It is also why the reader’s task is not to look through or beyond the surface of the poem for an experience it enacts or an experience it represents.  Instead, the reader is expected to attend to the elements of the system (embodied as writing rather than being a spoken action recorded in writing) to intuit the equation that is the poem.  There is, of course, a representational dimension to “In a Station of the Metro,” but that representational dimension is neither the source of the poem’s aesthetic power nor its significance.  As Pound understands his writing of the poem and the way that mode of writing defines how the poem should be read, the reader is to experience an intensified moment of perception and consciousness by grasping the equation.  That equation is made from, but essentially replaces, the experience from which the poem is made.  In Pound’s imagining of the poem, in his critical framing of it, both mere rhetoric and mere representation have been transcended, and the reader (by engaging the poem correctly) can possess the “beauty” of the faces directly rather than through the mediation of the poet’s witness and testimony (which the tale of the experience in “Vorticism” might be said to offer).

For Millay, though, the page is more a dramatic space, a location, for enacting voice.  And whether we hear her speaker as a character separate from the poet or as a figure for Millay as an individual who happens to be a poet (and given our associations with Millay and her career, we are more likely to hear it as the latter), we recognize that the poem functions as a moment of speaking.  In this sense the writing is aural, not visual, and her piece works quite differently than Pound’s.  If we read “First Fig” as an “equation” (in the sense that Pound uses the term in “Vorticism” or as he operationalizes it in “In a Station of the Metro”), “First Fig” reduces to something like “double-ended candle burning/lovely light” or “life in the fast lane/ooooh.”  The first is insipid; the second a bad rock lyric.  Either way, we affirm or deny this “equation” so quickly and with so little investment that it has little impact.  The equation expresses a cliché, not a discovery.

“First Fig” is not, though, a failed Poundian equation.  Its impact comes from the inflections of the represented voice as that voice addresses to the reader as imagined auditor.  This is especially clear in the third line, where the counterpoint of the two spoken interjections sets up the last line not as a simple observation but as a dramatically resonant assertion that evokes the speaker’s complex experience of the gains and losses of her burning.  “First Fig,” that is, is a poem we read to hear as an act of speaking rather than read to see as written object, and because it involves hearing (and requires imagining the conditions of the figure who speaks it from the page), the response of the reader/auditor can draw on the dynamics of speaking and listening and include such mechanisms as empathy.  With “First Fig,” to abstract the speaker out of the text (to read it, that is, as if it is a Poundian equation) destroys the poem.

“First Fig” and “In a Station of the Metro” illustrate the doubleness of writing that Vachek has noted.  The differences between writing as a representation of the sound and action of speaking and writing as a direct visual system can drive quite different understandings of the mechanism of the page and generate different understandings of what poems are and how they function for the reader.  Pound’s Imagist pieces emphasize words as units of visual action rather than embodied, spoken gestures.  The sound the words create is a supporting element, but sound is important for the musicality it adds to the “equation” and how this enhances the poem’s status as aesthetic object; sound does not imply or evoke the presence of a speaker, a persona, to whom we listen and respond as if in a dramatic exchange.  In a piece like “First Fig,” however, the poem turns on spoken gestures.  Even as the meter and rhyme and its brevity mark it as crafted writing, we hear the words as if spoken, and we track the dramatic inflections that the writing records and implies as if we are to respond to the speaking.  In both “In a Station of the Metro” and “First Fig” language is both aural and visual, but writing as representation of speech and writing as visual system combine differently in the two and reflect the poets’ different understandings of the mechanism of the page, of writing and poem, and the reader.  The poems, that is, enact different models of textuality and thereby demand different kinds of reading.

While the differences between the work and careers of Pound and Millay involve more than one having emphasized writing‟s potential as a visual system and the other drawing more on writing‟s capacity to represent speech and speaking, the differences between “In a Station of the Metro” and “First Fig” do suggest that American poets of the first half of the twentieth century did experience a dichotomy between orality and literacy but experienced it not so much as a primary opposition of speaking to writing but instead as a kind of secondary difference within writing itself (writing’s capacity to mime speech and give it visual form and its capacity to operate independently of the dynamics of speaking and be a direct visual system). If so, the various textual rhetorics that modern poets fashioned as they negotiated writing‟s two potentials offer an additional way to explore the array of voices and experimentation that characterized American poetry in the first decades of the 20th century, as poets attempted to move beyond the poetry of the recent past (which already seemed sentimental and quaint even before the cultural disaster of World War I), as they worried over poetry‟s increasingly secondary place in the literary marketplace as prose fiction consolidated its position as a (even the) major literary form and not just a popular art, and as they faced the possibility that poetry in particular and literature in general might simply be irrelevant or inadequate to the conditions of modernity. As poets sought to renew or recast the poetic idiom and to make poetry culturally relevant (by variously exploiting the potentials of print or instead by resisting or seeking to recast print in ways that could make a poetry of speech that could address modernity within a culture and technology of print as mass media commodity), matters of style were also inevitably matters of substance, and this suggests that at least some of the differences between significant poets in this period stem from their different senses of whether the page should be treated a space for writing (as “In a Station of the Metro”), a space for speech and speaking (as in “First Fig”), or as a space where various hybrids could be deployed (the work of Robert Frost, Langston Hughes, and others).

In considering how the material culture of modern print functioned as a matrix for the development of modern poetry, McGann in Black Riders emphasizes the visual potential of print and printing.  From this perspective, Pound’s version of the terms for constructing modern poetry and his theorizing of the page become both foundational and paradigmatic.  But this highly illuminating and productive move can also shade (or more accurately mute) the work of poets who sought to fashion modern poetry while yet imagining poetry as a kind of speech or speaking.  As such McGann’s study reflects the tendency in accounts of modern American poetry to emphasize the “visible language” of his title over the aural language of other poets (often relegated in the standard anthologies to the limbo of the category of “anti-modernists”) and to focus (as a result) largely on poets who approach poetry as writing constructed on the page rather than spoken action mediated through the page.

McGann’s discussion also reflects one of the reasons for this critical and historical preference: if the sociology of print is (as he suggests) the matrix for the period, then poetry that tests the limits of writing as “end-in-itself” (as Pound’s does) becomes the more significant response to the period’s possibilities than poetry that doesn’t (such as the work of Pound’s imagist fellow traveler and eventual adversary, Amy Lowell).  The visualist bias in Pound’s criticism has been another factor favoring the recognition of modern visualists at the expense of modern auralists, both because he offered a particularly systematic and acute account of what poetry should be and because he argued against the spoken by damning it as mere popular “rhetoric” and positioning it as the antithesis of the poetic.[20]

Pound’s familiarity with, and enthusiasm for, developments in the visual arts may have been a factor in his early championing of the visual and his suspicion of the aural.  His contemporaries, including Gaudier-Brzeska and the early Cubists, had already made their break with 19th century modes and styles and were producing work that seemed fully and distinctly modern.  Lawrence Rainey suggests another factor: Pound’s need to craft an aesthetic that would position his work and the work of those he was supporting to compete in the critical marketplace with the enthusiasm for F.T. Marinetti and Futurism.  In Institutions of Modernism: Literary Elites and Public Culture (1998), Rainey suggests that Pound’s campaign to establish Imagism reflected, on the one hand, a need to counter the continued popularity of the Georgians and, on the other, the enthusiasm in the popular press for Marinetti.  Rainey notes that Harold Monro, editor of Poetry and Drama, devoted the entire September 1913 issue to Futurism,

And in a long editorial, Monro praised Marinetti warmly, hailing him for auguring a dissolution of every distinction between poetry and popular culture, art and life. . . . Here, he [Monro] declared, was poetry that was no longer written for “close and studious scrutiny by the eye,” poetry “no longer…withheld from the people” by “educationalists,” “intellectuals,” or the commercial press, but poetry intended “for the ear,” for immediate and wide circulation,” poetry “regaining some of its popular appeal.”[21]

Here, that is, was a poetry that could claim to be art yet was popular—and popular in part because it appealed to the ear and could be performed for a general listener.

Equally important to the critical tendency to focus more on poetry as visual writing than poetry as speaking recorded in writing (or written speaking) is Eliot’s influence.  His advocacy of aesthetic impersonality has different roots from Pound’s advocacy, but his vision of a poetry purged of the merely personal so that it expresses the tradition more than the poet (and gains its authority from its power to extend the tradition rather than its power to transform and present the poet’s own personal experience) contributed to the New Critical preference for a poetry of well wrought urns that did not refer back to or depend on the implied presence of a speaking poet for authenticity or effect.[22])

But if American poets in the first half of the 20th century did face a crisis of the page (as I believe McGann rightly argues) driven by such factors as the economy of modern print and the dynamics of modern print literacy, then it may be that it was possible for poets to experience this crisis not only as the challenge to reinvent the page as a space for writing that resisted the culture of modern print (even as it circulated within it) but also as the challenge (under these same conditions) to reinvent the page as a space for speaking that similarly cut against the grain of modern print culture.  If so, this suggests the need to consider ways that the story of modern American poetry might be not only one of canonical writers like Pound perfecting the page as visual space and object but also as the story of how other poets worked to recuperate the page as mediating link to the reader and use it as a space for poetic speech.  In the process of this rethinking we might, for instance, come to a better understanding of the nature and implications of the contrasting experiments of Vachel Lindsay and e.e. cummings, and we might discover that attending to the poem as a mode of speaking can add to our understanding of the way Langston Hughes and the Robert Frost of the early narratives enact speaking in their work and how (and something of why) their approaches to this differ from Eliot’s in The Waste Land and Pound’s in various of The Cantos.  But above all, I would suggest, we would come to a fuller understanding of the array of possible responses to the Black Riders of modern print and in the process recover poets and poetries that our literary histories have marginalized.

[1] Jerome McGann, Black Riders: The Visible Language of Modernism (Princeton, NJ: Princeton University Press, 1993), 76.  Subsequent references noted parenthetically in the body of the text by page number.

[2] One might also say that McGann demonstrates that the modernist poets treated “poetry” as an occasion for “composing” with language rather than “performing” in language.  For a discussion of this distinction, see, Tim Hunt, “The Muse Learns to Tape,” in Reimagining Textuality” Textual Studies in the Late Age of Print, ed. Elizabeth Bergmann Loizeaux and Neil Fraistat.

[3] Walter J. Ong’s Orality and Literacy: The Technologizing of the Word (New York: Methuen, Inc., 1982) remains the best introduction to impact of the development of writing, especially the first four chapters.

[4] See Albert Lord’s the Singer of Tales (Cambridge, MA: Harvard University Press, 1960).  Also, Ong offers a brief overview of Parry’s position in Orality and Literacy (see pp. 18-28).  See, also, John Miles Foley’s The Theory of Oral Composition: History and Methodology (Bloomington, IN: Indiana University Press, 1988) for a broader overview of the development of the “Oral-Formulaic Theory” catalyzed by Parry’s work.

[5] See John Miles Foley, Immanent Art: From Structure to Meaning in Traditional Oral Epic (Bloomington, IN: Indiana University Press, 1991); Jack Goody, The Interface Between the Written and the Oral (New York : Cambridge University Press, 1987); Eric A. Havelock, The Muse Learns to Write: Reflections on Orality and Literacy from Antiquity to the Present (New Haven, CT: Yale University Press, 1986); A.R. Luria, Cognitive Development: Its Cultural and Social Foundations (Cambridge, MA: Harvard University Press, 1976); and David R. Olson, The World on Paper: the Conceptual and Cognitive Implications of Writing and Reading (New York : Cambridge University Press, 1994).

[6] See Bruce Jackson’s Get Your Ass in the Water and Swim Like Me: Narrative Poetry from Black Oral Tradition (Cambridge, MA: Harvard University Press, 1974) for an example of field work that illustrates this.

[7] See Tim Hunt, “The Muse Learns to Tape” in Reimagining Textuality: Textual Studies in the Late Age of Print, Ed. Elizabeth Loizeaux and Neil Fraistat (Madison, WI: University of Wisconsin Press, 2002). 189-210.

[8] Josef Vachek’s seminal 1959 essay, “Two Chapters on Written English,” is collected in Selected Writings in English and General Linguistics (Prague : Academia, 1976), 408-41.  Vachek later updated and extended his discussion of these matters in Written Language: General Problems and Problems of English (The Hague, Mouton, 1973).

[9] This might seem to contradict Ong’s claim that literacy and orality are deeply different.  Rather it is simply an aspect of what happens to speaking under the conditions of modern literacy.

[10] See Vachek’s “Characteristic Features of the Written Norm of English” and “Problems of the Orthographic Reform of English” in Written Language, 40-68.

[11] In Rethinking Writing (London: Athlone, 2000), Roy Harris explores, among other things, the implications of numerals for understanding what writing systems are and are not; see, especially, the chapter “Notes on Notation.”

[12] “Characteristic Features of the Written Norm of English” and “Problems of the Orthographic Reform of English”

[13] Stein’s phrase also illustrates that words in speech and words in writing do not always work quite the same way, as two further examples illustrate.  Shel Silverstein (who wrote “A Boy Named Sue” for Johnnie Cash, drew cartoons for Playboy, and earned a dime or three off his children’s book The Giving Tree) has one humorous song, “The Mermaid,” that turns in the final line on the way “tale” can also be heard as “tail.”  The line has been composed to be heard, not read, and in fact would not work on the page where the need to print the word either as “tail” or “tale” would preempt the other meaning.  Conversely, one of Bob Dylan’s puns in “Subterranean Homesick Blues” is clearer in writing than when sung:  For a clip in the documentary film Don’t Look Back Dylan mimes along to the recording, holding up a series of cards with the rhyme words written on them.  For the line “try to be a success” the card reads “suck-cess,” which mimes the pronunciation in the song as performed and visually reveals the point of the pronunciation by emphasizing the pun that the ear might otherwise hear as just another of Dylan’s vocal exaggerations.

[14] In free verse there is no sound pattern to show where one verse line turns and the next begins; the line break has to be seen on the page to be registered.

[15] The Collected Poems of Robert Creeley (Berkeley, CA: University of California Press, 1982), 132.

[16] ABC of Reading (New York: New Directions Books, 1960) 198-99.

[17] Pound does, of course, include spoken passages in The Cantos, but he typically subordinates these units to the space of the page.  They are less moments or acts of speaking (as if to the reader) than blocks of speech sculpted as written units and positioned on the page along with the other written units that function more obviously and directly as distilled, Vorticist “equation.”  The more personal, sustained “voice” of the “Pisan Cantos” is an exception to this but one, I’d suggest, that reinforces the general point.

[18] Gaudier-Brzeska: A Memoir (New York: New Directions Books, 1960), 83; additional references noted parenthetically in the body of the essay.

[19] Willard Bohn notes that “In a Station of the Metro” “was modeled on the Japanese haiku” and that its original published form in the April 1913 issue of Poetry was

The apparition        of these faces        in the crowd     :
Petals          on a wet, black     bough       .

Bohn, who I wish to thank for pointing out this initial form of the poem, notes that “Pound’s experiment with visual effects was extremely short-lived,” but his comments in “Vorticism” suggest that he continued to think of poetic language more as “writing” than as “speech” or “speaking.”  See Bohn, Modern Visual Poetry (Newark: University of Delaware Press, 2001). 33.

[20] In “Vorticism” Pound declares, “The ‘image’ is the furthest possible remove from rhetoric” (83).

[21] Institutions of Modernism: Literary Elites and Public Culture (New Haven, CT: Yale University Press, 1998), 31.

[22] That Eliot later moved toward a voice that can be seen as more directly personal (as with Pound’s move in “The Pisan Cantos”) complicates this claim but doesn’t change the basic logic of his initial practice or the implications of the earlier critical pieces, which have been so influential in shaping the criticism of this period.

Creative Commons License
The There That’s There and Not There in the Writing of Writing: Textuality and Modern American Poetry by Tim Hunt is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.


Back to Critical Work