Your browser (Internet Explorer 7 or lower) is out of date. It has known security flaws and may not display all features of this and other websites. Learn how to update your browser.


Text as Technology: Rethinking John Bryant’s “Fluid Text”

John Bryant, in his essay “Witness and Access: The Uses of the Fluid Text,” argues that virtually every literary work exists in multiple versions that are both materially and textually distinct. One major source of such variation is revision, whether by authors, collaborators, or editors. Bryant urges the critical community to witness writers’ revisions, by turning away from “a single, non-variant text” (26), to read instead “the text of the history of the book” (21). This is what Bryant calls the fluid text.

I am currently researching S.T. Coleridge’s process of revising and representing the radical poems of his youth, poems that mortified him as his political views developed. I share Bryant’s concerns, and I am only too ready to agree with him that revision is a worthwhile object of study.

However, “fluid text” is a odd term for describing the diachronic and plural existence of literary works. Just as the old term “textual corruption” suggests that textual change is unnatural, even lethal to the literary work, so the new term “textual fluidity” suggests that texts are naturally inconstant. Worse, the term “fluid text” reifies the notion that literary works are ultimate examples of linguistic fluency. The implication is that the text adapts, ineluctably, to each new environment; the text is thus always optimally intelligible.

I am not suggesting that Bryant intended “fluid text” to mean all these things — only that the root of the word fluid associates the “fluid text” with fluency: with so-called standard language, produced coherently and without show of effort. There are many literary texts that are unintelligible by contemporary standards of linguistic fluency, if only because they use archaic or invented words.

Linguistic fluency is an imaginary standard that has no material existence. When we speak we misspeak, stutter, prevaricate, repeat ourselves, repeat others, transpose one phoneme for another, and “repair” our utterances as we go. There are few full sentences. Speech works because it is a social technology that we change by using, and speech is never fluent — except by degrees of fluency, which are always open to question.

Textuality is not a natural phenomenon, but rather an extension of the social technology of speech. Textuality is natural only in the sense that it relies on the complex anatomy that apprehends, comprehends, and articulates speech: Broca’s region, the voice box, the alveolar ridge, and so on.

Text is a technology that varies and repeats speech productions, having transferred them into new media. Text develops speech, extending its productions — whether vernacular or literary, real or virtual — over expanses of space and time, and intensifying rather than erasing its halting characteristics. Text holds speakers accountable for their productions in more durable ways than listeners can. The production originally may have travelled through the air, may have been extant in the subvocal motions of silent reading, or may have done nothing more than activate neurons before its textual debut.

When we claim to read and understand a single, non-variant text, we may create the illusion that text has achieved the linguistic fluency that speech cannot. To collate and read multiple variant texts is to read a protracted textual production that is emphatically dysfluent, or a speaking text.

“Speaking” is not a metaphor for what the text does, if one assumes that text is an extension of speech: every text speaks, imperfectly, on the behalf of those who participated in its production, and the collation of variant texts allows the reader to witness those texts as a palimpsest of interruptions, revisions, plagiarisms. One text may reiterate last year’s speech, speak to the occasion, or retract last year’s speech. Its strategies of representation may do all three of these.

To examine texts in their material diachronicity is both to understand the work as a potentially wide-ranging social entity with multiple producers, and to understand text as technologically developed speech. “Textual fluidity” is a useful shorthand for the complexity Bryant begins to describe in “Witness and Access,” and it certainly sounds more desirable than “textual corruption.” Better terminology would recognize how texts are non-fluent social productions, inseparable from the speech communities that affirm and contest their coherence.

Theorizing textuality by analogy will always be problematic — but to function as a logical analogy, there must be aspects of the entity (the “text”) that resemble aspects of the supposedly analogous entity (any thing that is fluid, i.e. any “fluid”). Since texts and fluids are totally dissimilar, “fluidity” is a metaphorical description of textuality, rather than an analogous description. It is more logical to measure text against a phenomenon it resembles. I propose that since text is the transcription of extratextual human language,  an analogy between text and speech is preferable to an analogy between text and liquid.

Written for “Versioning & Collation in the Digital Environment” at the Digital Humanities Summer Institute, 2013.

Works Cited

Bryant, John. “Witness and Access: The Uses of the Fluid Text.” Textual Cultures: Texts, Contexts, Interpretation 2.1 (2007): 16-42. Project MUSE. Web. 6 July 2013. <>.

Olivia Ferguson

is an MA student in English at the University of Victoria.

Leave a comment


email (not published)