Beginning with "The Image" in <em>How It Is</em> when translating certain processes of digital language art

Beginning with "The Image" in How It Is when translating certain processes of digital language art

John Cayley

In this essay, John Cayley builds upon a critical legacy that reflects intensively on the process of literature in an age of machine language. Building on a legacy that includes ebr classics like “The Code is not the Text” and many points in between, “Beginning with ‘The Image’” points to the difficulties of translating a procedural code-based text that “finds” its substance in the repository of text known as the World Wide Web.

This essay appears previously in the proceedings of the Universities of Paris 8 conference “Translating E-Literature/Traduire la littérature numérique.” 

Practices and theories of translation are situated at a crucial position in the domain of the practices and theories of language. We are comfortable with distinguishing practices of language whose systematic differences allow us to say that the users of languages in which they are separately competent are, nonetheless, “mutually unintelligible” to one another, and so we may say that they are using different, distinct languages. However, certain practitioners may be proficient in any number of such languages and they may be able to perform in the languages they “know” in a manner which, they claim, regularly corresponds with performances in another language. This ability is commonly called “translation” and even though nothing—more especially no “thing”—has been translated (moved from one “place” to another), such intercultural circumstances have led to a widely held belief that different languages share some thing that could be so moved and thus translated. This “thing” is usually identified with an interpretable, paraphrasable “meaning” but may also, in more developed theories, be complicated with correspondences of affect, or with world-building associative networks of significance and affect. However, if, in our thinking, we maintain a focus on actual practices of language—always necessarily sited in particular cultures—we can, and, in my view, should, largely disregard theoretical models that are philosophically discredited, especially since poststructuralism, and highly misdirective for any discussion of what language is, ontologically. Language is not some thing in a world, and never has been. Works of language art are not themselves and do not produce “translatable” things. “Translation” is a word associated with a prejudicial conceptual framework. It must and will be used for the sake of convention, but within it the intellectual skepticism indicated by quotation marks should always (already) be inscribed. 

Despite the fact that my own quite particular philosophy of language underpins this essay, it is written and offered to its readers from the standpoint of a practitioner of digital language arts. In other parlance, also current, I might also be characterized as a “writer of electronic literature,” one who is interested (at least for now) in the “translation” (henceforth simply: translation) of this kind of work. I will, chiefly, refer to my own and collaborative work—because I know it well and because I have been intimately involved with the processes of its making, but also because the piece I’ve chosen to examine in some detail, “The Image” from How It Is in Common Tongues, seems to me to exemplify, sharply and distinctly, issues that are important to consider, not only for any project to translate linguistically-sited works of digital language art, but also for the critical appreciation of this work in general, in any linguistic context.

There is no intrinsic necessity for writing that is digitally mediated to be self-reflexive, to be, for example, concerned with its own forms or to take these forms as a primary focus for any significance and affect that may be produced by the work. However, digital language arts have emerged and developed at a time when literary critical discourse has come to give significant recognition to “conceptual literature,”1 and the concurrent major transformation and reconfiguration of media supports for writing and reading as cultural practices does generate a tendency, if not an obligation, amongst practitioners to apply themselves to the concerns of formal mediation, often going so far as to make them, as it were, the “subject” of the work.2 “The Image” of our discussion is, first and foremost, a work of conceptual literature and it is also explicitly, self-reflexively concerned with its own formalisms and processes of mediation. But in so far as this work is also the generator of a new text—in the English language of its author and his collaborators—it does also produce a work, or a part-work if you prefer, that has some claim to being writing with a separate “subject,” a separate focus of significance and affect—one that is not chiefly conceptual, not formally self-reflexive. Here, I refer explicitly to the part of the work that I’ve entitled “The tongue gets clogged.”

In setting out these circumstances, two consequences for the translation of digital language art present themselves immediately. Firstly, we are having to deal with something that is not an integral, easily identifiable or unified text. We are dealing with a system, a dynamic, generative system that, at any one time, may be assembled from multiple devices producing outputs in, precisely, “multimedia,” albeit constrained, due to the overarching conventions of linguistic inscription, to screen-based audio and visual media. In the case of “The Image” we are also dealing with an underlying framework—both conceptual and programmatic, to include what programmers would call a “development environment.” The devices of “The Image” are built on top of a framework for practical research and for the production of precisely such works (or aesthetic outcomes) as “The Image.” This framework is The Readers Project, a collaboration of the author and Daniel C. Howe.3 Moreover, “The Image” is part of a distinct interrelated work or project within this larger context, separately known as How It Is in Common Tongues. How much of this necessarily complicated framing must also be translated if we are to translate “The Image” “properly” or “faithfully” into another tongue? Asking this question—always possibly a misdirection—parallels the interrogation of conventional translators concerning all those aspects of context, shared (or not) by the translator with the author of some original. How much of what the original author experienced must the translator experience or study in order to be able to offer a good (or better) translation? The difference when addressing works in programmable media is that there is not only the question of, as it were, an improved cultivation of the translator; it may be possible, materially, to construct corresponding devices and to design corresponding processes in the host linguistic culture, “translating” multiple media devices which, themselves, are the producers of textual “translations,” rather than simply allowing a translator (of whatever kind) to function in a manner that is better contextualized.

There is no intention here to be prescriptive. Any provisional answers to the questions just raised will be provided by the following descriptions of and commentaries on actual or proposed translingual practices undertaken in relation to “The Image.” 

Secondly, we are called to face up to the problem of translating something that is primarily a conceptual literary work. We are called upon to produce a work that is embedded in the host linguistic culture while performing or exemplifying a conceptual project that closely corresponds with, in this case, that of “The Image”—and quite apart, that is, from any regular correspondence of the projects’ constituent texts from one language to another. There are texts that constitute the work of “The Image” and also its translation but any judgement concerning their correspondence will be subordinated, for example, to judgements concerning correspondences between the methods for textual production in original or translation, bracketing any correspondences that the text themselves may or may not exhibit in terms of linguistic form or content or, indeed, in terms of whatever significance or affect a reader in either language might derive from the form or content of the respective texts.

Here, this point of inquiry into translation clearly highlights new demands on the criticism of digitally mediated literary work. After deciding that works are worthy of translation, if we then determine, for a particular work, that we are justified—and I suggest we are—in focusing on the “translation” of concepts and corresponding methods of textual production rather than on the texts themselves, then it seems to me to follow that our critique and appreciation of such works must extend retrospectively to the critique and appreciation of concepts and methods of textual production in the first place. The call for better conceptual readings of literature and, more specifically, the call to read the programmatic methods of digital language art as an integral aspect of their literary reading and criticism, this is something that is wide-spread and urgent in the theory and practice of digital language art, taking on a number of guises, such as those of critical software studies, or as when the textual work is recast in terms of (textual) codework.4 The practices of translation in relation to digital language art simply put these demands into sharp relief by making their very practical demands on us, so as to determine what is important about these literary aesthetic objects, what of their various different aspects we are determined to translate effectively.

Now, what is it that I am proposing to translate? Or rather, what are the characteristics of this work, “The Image,” in relation to which I have promised to discuss aspects of both its actual and proposed translation? And what are the details of the languages involved?

The first thing to say about “The Image” is that it is a work that only exists and can only be read in relation to another recognized literary text, fully acknowledged and highly regarded as such. The source or, some would say, supply text of “The Image” is a twenty-two paragraph section from the first part of Samuel Beckett’s How It Is, a longer prose work, published 1964, in Beckett’s own English “translation,” based on his original French, published 1961. As a distinct short prose, this text has some claim to a distinct textual history since it was first published as a separate item in French and has been also been translated into English a number of times as a distinct piece before settling into the context of How It Is and achieving there what I think of as its authoritative form in English.5 Not least for its textual history (and its relationship to translation) “The Image” became a point of fascination for myself for a number of reasons. As a source and stimulus for experiments in conceptual literature, regular algorithmic manipulation, for appropriation and its discontents, Beckett’s writing offers many attractive characteristics. His writing is always beautifully formed and finished. It is austere and relatively (deceptively) simple, with strong affective and philosophical connotations. Without conceding syntactic rigor, the writing may, as here, have no punctuation or use of upper case letters other than for proper names and the first person pronoun in English. This renders its programmatic processing that bit less complicated. Finally, when it comes to addressing questions of association and integrity in relation, for example, to processes appropriation or adaptation, the moral rights of Beckett’s texts are jealously protected, despite and perhaps also as a result of their avant garde, modernist, revolutionary character. An address to their appropriation should be taken to be a serious address, and important issues must be at stake.

“The Image” that has been made by myself with my collaborator is, in the first instance, simply a reading of what Beckett wrote, in its English version. The main display continually renders a screen-full of Beckett’s text in a window that we call a “perceptual reading interface.” This window retains a focus for reading attention at the center of the screen, moving this point through the text while continually composing a typographic neighborhood around the moving point of focus. But the animated reader does more than simply move through the text. Starting from some current word, which it highlights, it moves on forward one word at a time seeking to determine and to highlight, successively, the longest common phrases of the text of Beckett’s “Image.” Such phrases we define as the longest sequence of words in any text (starting at an arbitrary point) for which we can find and cite from the internet, using search services, the same phrase having been composed by at least one other writer or writing machine, and not the author of the text we are reading—in this case Beckett’s “Image.” The reader finds a longest common phrase, highlights it, pauses briefly, and then continues on through the entire text, looping, in installation, back to the beginning of the text and repeating the process. This algorithmic reader reads Beckett, but does so by way of phrases that may be cited to prove that they are also a part of the commons of language and are cited from that commons.

To “translate” this reading into French—the first stage of our goal—we would require a French version of “The Image” and we would subject the text to precisely the same process, discovering its longest common phrases but searching the internet-accessible corpus of French linguistic practice.

At this point in the story, we encounter a couple of unexpected twists and ironies. One of these is specific to the text with which we are working. We had skirted over the question of producing—translating—a French version of “The Image.” We might have expected to make this translation by conventional means as human English-to-French translators. But for “The Image” there is already a French version, Beckett’s own, and this is, in fact, the “original,” if there is one. A French text of the earlier French “l’Image” is embedded in the text of Comment c’est, which is the original of How It Is. Ironically, in this instance, to translate Beckett’s words and sentences into French for the purpose of building our conceptual reading machines we need do precisely nothing, nothing other than appropriate a more original original.

But as we apply the longest common phrase algorithm to the French text that we’ve accepted as corresponding perfectly with our own starting point, we will be forced to consider linguistic and cultural differences that are a function of the interplay between our concept, our methods and their broader context within the regimes of computation and the digitization of linguistic practice and literary artifacts. These considerations are not specific to Beckett and the texts we have chosen to work with. The French language has a different relationship with the Internet, as compared with English. While “global English” is the lingua franca of networked linguistic transaction and also of the now global regime of computation as a whole, and although linguistic practices in French will be very extensively represented and widely subject, for example, to internet search services, no one should be prepared to claim that there will not be significant differences, as compared with English, in the way that French has been computationally rendered and digitally “networked.” If we apply the same longest common phrase algorithm to Beckett’s French, we will not only find different sequences of words belonging to a different language; the phrases found (and successively highlighted by the algorithmic reader) will, inevitably, also represent and express a different relationship, as compared to English, between French and the Internet. Does this make our translation of process a more or less faithful “translation” per se

The systematic performance of “The Image,” as currently configured, does not confine itself to the highlighted presentation of longest common phrases on the perceptual reading interface. There is more to translate, however problematically. The perceptual reading interface is able to broadcast an indication of the phrases it is reading to any networked devices and these are thus able to follow the reading of the main display, in real time. This type of networked interconnection does not, in itself, have linguistic valence. It is not attached—in terms of significance or affect—to some particular linguistic culture. However, the connected devices may be programmed to reflect or respond to the live reading of the main display in arbitrary ways, all of which, minimally, must be taken to be linguistically implicated in so far as they are responsive to a linguistically situated “reading” in the first place. At another end of the scale of abstraction their responses might also be programmed to be, for example, interpretative or linguistically generative and thus they might well make further demands on our processes of translation from English to French. In a networked system of distributed reading such as this, the abstract transcoding that the digital allows means that the linked devices may respond regularly in any way conceivable. There are infinite possibilities, all regularly linked to the reading that is taking place on the main display. Here, we will merely go on to examine the actual existing example of one possible response, so as to consider the details of its specific demands on our speculative project to translate the current English version of “The Image.”

As it happens, remote connected readers of “The Image” are supplied with a text constructed quasi-algorithmically that has, in this current instantiation of the work, been micro-collaged and pre-processed in such a way as to allow live synchronization with the main display, and so also provide an expressive literary aesthetic extension to the system of “The Image.” Personally, I consider this generated text to be that part of the system as a whole which represents its the most engaging conceptual artifact and which is, in its own right, a literary object with its own aesthetic, one that is fully readable in the contexts of, for example, conventionally mediated poetic or experimental prose writing, including that of Beckett’s original.

This text is constructed more or less as follows. The author of “The Image” as a whole, myself, works through Beckett’s text. He attempts to find longest common phrases (as defined above) although, in order to respond, as a human reader, to what he is discovering as language in the text-under-construction, he may content himself by falling back to common phrases, not necessarily longest common phrases, as so defined. He must always, however, be assured that all the phrases or fragments of text shared with Beckett’s original are also discoverable, by internet search, to be embedded in language that was not composed by Beckett. In fact, the constructed text bears witness to this requirement because it is, as we say, micro-collaged from selected instances of text that its “author” discovers when searching, “by hand,” for the common phrases he is extracting, successively, from Beckett’s “Image.” He finds a piece of language, written by someone other than Beckett, that includes the current phrase. He then proceeds to find another piece of language in which a subsequent common phrase is embedded and he works to stitch these discovered texts together in some syntactically coherent fashion, perhaps hoping also to discover or bring into being a form of language that conveys significance and affect: a new text, something to be read, something that will bear reading. An illustration, showing the first paragraph of the Beckett and the resultant text, with the common phrases in bold, may illustrate the process and also the first part of the resulting text that is, currently, a constituent of “The Image.”

This text is formally interesting. It is composed of language that is not written by Beckett but that contains sequences of words common to Beckett’s “Image.” It was constructed by one of the human authors of “The Image” but none of the language it contains was composed by this person. I characterize it as quasi-algorithmic in that it is produced by regular procedures that could quite easily be set out in pseudo code (as they have just been described in expository prose) but, in practice, certain compositional decisions—concerning things like the selection of one amongst, often, many possible “results” of internet search with language that includes the common phrase, and the precise manner in which the selected language is stitched together—were made by the human author and these would be very difficult, although perhaps not impossible, actually to code. What can, manifestly, be done, however, is to convey these procedures to someone (or some system) able to carry them out on the French text of Beckett’s “l’Image” and thus to produce, arguably, a “translation” of the constructed, stitched text of “The Image.”

This text is clearly not a translation of anything like whatever a human reader would construe or “read from” the quasi-algorithmically generated stitched text of the English and yet, in terms of underlying process, the two texts correspond closely and they share closely similar relations with their supply texts, their generative seeds. In this exceptional case, the procedurally “translated” French actually contains, embedded within it, the “original” of the English supply text, differently punctuated by the discovery of longest common phrases in French. Between the versions there is tight correspondence of concept and procedure—highly important in the context of this work, its impetus and its aesthetic—but there is also an overwhelming non-correspondence of “natural language” (for want of a better phrase). Everything hinges here, on our modality of reading. If we read for concept or procedure, we have a translation. If we read for the expository or descriptive significance of the text as putative “natural” French or “natural” English, then we do not have a translation, certainly not a conventional translation. If we read for other aspects of poetics, for affective relations with Beckett’s originals in either language, for example, then I believe it is harder for us to be sure what it is we are faced with. If a work such as this does make original affective claims—or indeed claims of original significance—then it seems possible to me that these claims may well have been “translated” by a procedural approach to the generation of the French text. It may well be judged to be faithful to these affective relations and also to certain of the original signifying practices that it proposes. The transcultural significance of the longest common phrase—although, and importantly, it is differently instantiated in the practices of two languages—would be an example of the latter.

Our conclusion will amount to a highly provisional annotated list—with no claims to being exhaustive—of certain aspects of digitally mediated literary works that complicate and problematize processes of translation. These are only those aspects of such works that we have encountered here, beginning with “The Image” in How It Is, and through moving on to its proposed translation (back) into French. Other digitally mediated works will, of course, evoke new articles of problematization. For now, the hope is twofold: that a consideration of these articles will help to expose all translation as subject to the types of procedural and conceptual interventions that are demanded by these properties and methods of digitally mediated literary works, for which any “translation” requires iterative reconfigurations of concepts and processes in order to adapt their deployment within any commensurate culture of linguistic practices that is distinct from the one in which they were previously made. Secondly, such a foregrounding of the requirement for concept and procedure in the translation of digitally mediated literary work should be read as proposing the greater relevance of these terms to literary work in general, and especially to any work the poetics of which is in some way open, not entirely predetermined by a readily construable significance or conventional affective structures. 

The list then. 1 thru 4 refer to aspects of works without regard to any attempt at translation. 5 thru 7 refer to specific implicated relationships between a work and its rendition that are emergent in precisely these problematized practices of translation: 

1. Integrity. The work is unlikely to be an integral, unified work, easily identifiable as a stable text or edition. It may be a system or some part of a system. In either case the system may be multimedia. Even if it is of the same order or medium as other parts of its system, it may simply be something else, something larger or more complex.

Example: “The Image,” and its punctuation by longest common phrases, is a system. The supply text of “The Image” is a part of something larger but of the same order: How It Is. “The Image” is also part of a more complex generative system that produces quasi-original texts quasi-algorithmically.

2. Association. The work may be problematic, unconventional, and even transgressive in terms of its author function. It may be appropriative. It may use or manipulate found text that may—putatively—be owned or controlled by another human author or legal entity.

Example: Who or what is the author function of the texts generated by “The Image”?

3. Conceptualism. The work may be conceptual and formally self-reflexive.

4. Algorithm. Process, generative or manipulative, may be integral to either or both the composition and the delivery/publication of the work.

5. Local algorithmic dissonance. The work may implement generative or manipulative algorithms having incommensurable grammatical, lexical, or otherwise morphological specificities across languages. The work may address (by way of its own algorithmic processes) and make (implicit) assumptions about the grammar of its particular original language. These may not “work,” grammatically, in the host language.

Example: There is not a clear example of this in our treatment of “The Image.” Common examples will be found in text generative processes that are broadly similar (they “generate a phrase”) but require different detailed processes (to handle differently the morphologies of the words in the generated phrases).

6. Global algorithmic dissonance. The work may make reference (either algorithmic or human-authorial) to (algorithmic) networked instantiations of incommensurable linguistic specificities. With or without algorithmic processes such as those in article 5, the work may engage with linguistic structures that are embedded in the networked data structures on which the work relies to generate its readable texts and thus their significance and affect.

Example: “Longest common phrase” will have a different statistically-determined sense when derived from the (implicit) internet-search corpora of English and French.

7. Natural linguistic non-correspondence or divergence. If either articles 4 or 5 pertain, then processes may be applied to generate entirely new language—of which a work may entirely or partly consist; or constituent language may be generated from supply language extracted from some part(s) of a complex work. This newly generated language may be entirely or chiefly determined by algorithmic processes such that when these processes are deployed in or adapted to a host language, there may be no conventional, “natural language” correspondence—of form, or paraphrasable, descriptive, interpretative, expository sense, or of any significance, or even affect—between the generated language in the original work and the generated language that is the whole or part of the work thus “translated” into the host language.


I have been blessed to visit some of these enchanted stones and vistas. “You wouldn’t last a moment with the hostiles that live in this forest,” she smiled. Well I guess that you never knew me. Or at least not well enough.

Odeur et belles perspectives. Durer longtemps. T’es peinard pour un moment avec ça.

Works Cited

Beckett, Samuel. Comment c’est. Paris: Les Éditions de Minuit, 1961.

—. How It Is. Paris: Les Editions de Minuit, 1961. Trans. Samuel Beckett. New York: Grove Press, 1964.

—. “The Image (1956).” Trans. Edith Fournier. The Complete Short Prose, 1929-1989. “L’Image” in the inaugural issue of X: A Quarterly Review, edited by David Wright and Patrick Swift. Ed. S. E. Gontarski. New York: Grove Press, 1995. 165-68. 

Bergvall, Caroline, et al., eds. I’ll Drown My Book: Conceptual Writing by Women. 1st ed. Los Angeles: Les Figues Press, 2012.

Bootz, Philippe. “The Problem of Form: Transitoire Observable, a Laboratory for Emergent Programmed Art.”  The Aesthetics of Net Literature: Writing, Reading and Playing in Programmable Media. Eds. Peter Gendolla and Jörgen Schäfer. Media Upheavals. Bielefeld: Transcript, 2007. 89-103. 

Cayley, John. “Beginning with ‘the Image’ in How It Is When Translating Certain Processes of Digital Language Art.” Translating E-Literature = Traduire la littérature numérique. Eds. Arnaud Regnauld and Yves Abrioux. Paris: Université Paris 8, 2015.

Cayley, John. “The Code Is Not the Text (Unless It Is the Text).” Electronic Book Review (2002).

Dworkin, Craig Douglas, and Kenneth Goldsmith, eds. Against Expression: An Anthology of Conceptual Writing. Evanston Illinois: Northwestern University Press, 2011.

Funkhouser, Christopher T. New Directions in Digital Poetry. International Texts in Critical Media Aesthetics. Ed. Francisco J. Ricardo. New York: Continuum, 2012.

—. Prehistoric Digital Poetry: An Archaeology of Forms, 1959-1995. Tuscaloosa: University of Alabama Press, 2007.

Goldsmith, Kenneth. Uncreative Writing: Managing Language in the Digital Age. New York: Columbia University Press, 2011.

Howe, Daniel C., and John Cayley. “The Readers Project: Procedural Agents and Literary Vectors.” Leonardo 44.4 (August 2011): 317-24.

—. “Reading, Writing, and Resisting: Literary Appropriation in The Readers Project.” International Symposium of Electronic Art (ISEA). Eds. Kathy Cleland, Laura Fisher and Ross Harley. Sydney: ISEA International, the Australian Network for Art & Technology and the University of Sydney, 2013. 

Kenner, Hugh. “Beckett Translating Beckett.” Historical Fictions. Athens and London: University of Georgia Press, 1970. 184-202. 

Marino, Mark C. “Critical Code Studies.” Electronic Book Review (2006).

Perloff, Marjorie. Unoriginal Genius: Poetry by Other Means in the New Century. Chicago: The University of Chicago Press, 2010.

Place, Vanessa, and Robert Fitterman. Notes on Conceptualisms. Brooklyn: Ugly Duckling Presse, 2009.

Portela, Manuel. Scripting Reading Motions: The Codex and the Computer as Self-Reflexive Machines. Cambridge: MIT Press, 2013.

Raley, Rita. “Interferences: [Net.Writing] and the Practice of Codework.” Electronic Book Review  (2002).

Ricardo, Francisco J. The Engagement Aesthetic: Experiencing New Media Art through Critique. International Texts in Critical Media Aesthetics. New York: Bloomsbury Academic, 2013.

Simanowski, Roberto. Digital Art and Meaning: Reading Kinetic Poetry, Text Machines, Mapping Art, and Interactive Installations. Electronic Mediations. Minneapolis: University of Minnesota Press, 2011.




1. We can refer to a couple of important anthologies: Craig Douglas Dworkin and Kenneth Goldsmith, eds., Against Expression: An Anthology of Conceptual Writing (Evanston Illinois: Northwestern University Press, 2011). Caroline Bergvall, Laynie Browne, Teresa Carmody and Vanessa Place, eds., I’ll Drown My Book: Conceptual Writing by Women, 1st ed. (Los Angeles: Les Figues Press, 2012). There are also significant recent manifestos and an important monograph by a major critic of poetry and poetics: Vanessa Place and Robert Fitterman, Notes on Conceptualisms (Brooklyn: Ugly Duckling Presse, 2009). Kenneth Goldsmith, Uncreative Writing: Managing Language in the Digital Age (New York: Columbia University Press, 2011). Marjorie Perloff, Unoriginal Genius: Poetry by Other Means in the New Century (Chicago: The University of Chicago Press, 2010).

2. For discussions of these issues see: Roberto Simanowski, Digital Art and Meaning: Reading Kinetic Poetry, Text Machines, Mapping Art, and Interactive Installations, Electronic Mediations (Minneapolis: University of Minnesota Press, 2011). Francisco J. Ricardo, The Engagement Aesthetic: Experiencing New Media Art through Critique, International Texts in Critical Media Aesthetics (New York: Bloomsbury Academic, 2013). Manuel Portela, Scripting Reading Motions: The Codex and the Computer as Self-Reflexive Machines (Cambridge: MIT Press, 2013). Portela’s work is the most recent and explicit. In Ricardo see especially the chapter: “Engagement as post-literary mechanism, from exposition to reflexivity,” pp. 169-81, which discussed The Readers Project amongst other examples.

3. Further information about The Readers Project may be found via its website, and also in two recent articles: Daniel C. Howe and John Cayley, “The Readers Project: Procedural Agents and Literary Vectors,” Leonardo 44.4 (August 2011). Daniel C. Howe and John Cayley, “Reading, Writing, and Resisting: Literary Appropriation in The Readers Project,” International Symposium of Electronic Art (ISEA), eds. Kathy Cleland, Laura Fisher and Ross Harley (Sydney: ISEA International, the Australian Network for Art & Technology and the University of Sydney, 2013), vol.

4. There is a wide-ranging discussion of codework, code and textual practice to which I have also contributed with John Cayley, “The Code Is Not the Text (Unless It Is the Text),” Electronic Book Review (2002). See also: Rita Raley, “Interferences: [Net.Writing] and the Practice of Codework,” Electronic Book Review (2002). “The Code is Not the Text” might be read as suggesting the primary focus of attention for a digital literary work should be text rather than code, but I maintained then, as now, that the code generates the text and requires deep critical consideration just as, here, it requires “translation” and demands that translation take account of code. Both reading and translation should subject themselves to problematization on code’s account. The code itself is, nonetheless, not the text, per se. Our problem is “locating the text” across a complex composite of experiences, more or less accessible, of code, text and other resources. “Locating the text” or “Chercher le texte” was, in fact, the theme of the last conference of the Electronic Literature Organization in 2013 (, and one of its chief curators, Philippe Bootz, has an important theory of the complex apparatus of digital literary works, their composition and appreciation. See, for example, Philippe Bootz, “The Problem of Form: Transitoire Observable, a Laboratory for Emergent Programmed Art,” The Aesthetics of Net Literature: Writing, Reading and Playing in Programmable Media, eds. Peter Gendolla and Jörgen Schäfer, Media Upheavals (Bielefeld: Transcript, 2007). For an introduction to critical code/software studies, see: Mark C. Marino, “Critical Code Studies,” Electronic Book Review (2006). For the type of “close readings” of digital literary works in which an analysis of code plays an important role see: Simanowski, Digital Art and Meaning and the two books on digital poetry and poetics by Christopher Funkhouser.

5. An English version, translated by Edith Fournier, is published as Samuel Beckett, “The Image (1956),” trans. Edith Fournier, The Complete Short Prose, 1929-1989, ed. S. E. Gontarski (New York: Grove Press, 1995). pp. 165-168. The date of original publication for “l’Image” is given in the table of contents for this collection, but not in its note (pp. 283-284), which explains that it “… was originally published in the inaugural issue of a magazine called simply X: A Quarterly Review, edited by David Wright and Patrick Swift. An English version first appeared in As the Story Was Told: Uncollected and Late Prose (London: John Calder [Publishers] Ltd., 1990), 31-40,” but there was controversy as to whether or not this 1990 text was composed (or translated), in its entirety, by Beckett. Fournier’s (1995) version is based on the original French publication in X, and is said to have “closely respected” Beckett’s “final” English version, as incorporated into the 1964 How It Is. I would question this, and readers may form their own judgements. For a highly stimulating and entertaining engagement with what is at stake for the text in English-from-French, see Hugh Kenner, “Beckett Translating Beckett,” Historical Fictions (Athens and London: University of Georgia Press, 1970). where his sample passage “I” is four “paragraphs” of “l’Image / The Image” from Comment c’est / How It Is.