The Stuff of Fiction Digital Editing, Multiple Drafts and the Extended Mind

Since genetic criticism regards modern manuscripts as a research object in and of itself, it objects to an editorial practice that treats manuscript studies as a mere tool towards the making of a scholarly edition. Still, an exchange of ideas between genetic criticism and scholarly editing can be mutually beneficial and may work in two directions. This essay therefore starts from digital scholarly editing, more specifically from recent developments in computer-assisted collation of multiple draft versions, to see how it can contribute to the study of modern manuscripts. The argument is that the combination of textual scholarship and genetic criticism can be an effective instrument for literary critics, enabling them to study the material aspect of the writing process as an inherent part of what cognitive philosophy calls “the extended mind”; and that this extensiveness does not only apply to the writer’s mind, but that an awareness of manuscripts as a crucial part of the “stuff of fiction” can also contribute to a better understanding of literary evocations of the fictional mind.

The title of this essay refers to the word "stuff" in its material sense and to the role of primary objects, artifacts, and archival materials as the basis of, and challenge to, textual scholarship. In the past, there have been quite a few attempts to define the distinction between textual criticism and genetic criticism. Daniel Ferrer has suggested that textual criticism focuses more on "repetition", treating variants as deviations from a copy-text, whereas genetic criticism focuses more on "invention" (Ferrer 2010, 21), treating variants as forms of rewritings. Pierre-Marc de Biasi also advocates the use of the term "rewritings" or "réécritures" (de Biasi 2000, 20) instead of variants, arguing that one cannot speak of a variant if there is no invariant to compare it with. But in some cases, one can use another version (even a rough draft) as a "temporary invariant" to compare. And more generally, the rejection of the term "variants" also has to be seen in the historical context of the early days of genetic criticism, in the 1960s, when it was compelled to establish itself as a separate disci-pline by distinguishing itself clearly from textual criticism and scholarly editing, or what is usually called "philologie" in France. In the meantime, a rapprochement between genetic criticism and scholarly editing is not just desirable, but may actually be mutually beneficial. This essay is an attempt to contribute to such a rapprochement.
From the perspective of textual criticism, scholarly editors may object to de Biasi's rejection of the term "variants". Nonetheless, reconsidering "variants" as forms of "rewriting" ("réécritures") can also be an invitation to scholarly editors to treat variants as forms of creative undoing and inventive revision, not only in terms of deviations from a copy-text. Since genetic criticism has objected to the subservient role of manuscript research in scholarly editing, I would like to propose a reversal of these roles for the purpose of this essay. Instead of employing manuscript research with a view to making an edition, I will start from digital editing and treat it as a tool for manuscript research and literary criticism, starting from a concrete case: the ongoing research into the possibility of computer-supported collation of modern manuscripts, applied to the works of Samuel Beckett. From there, I will work backwards from digital editing to a consideration of multiple drafts and, finally, to a more abstract and philosophical investigation into what some thinkers call the "Extended Mind".

Digital Editing
For the Beckett Digital Manuscript Project (BDMP) -a digital edition that brings together facsimiles of all the manuscripts of Samuel Beckett -the Centre for Manuscript Genetics at the University of Antwerp is working together with Gregor Middell, Ronald Dekker, and Joris van Zundert at the Huygens ING in The Hague to try and make a computer-supported collation tool (CollateX) and implement it in the BDMP to compare multiple versions. A specific problem with the collation of modern manuscripts is that it involves the treatment of cancelled text. One manuscript version can often be subdivided into several writing stages, as they were originally called by the TEI Special Interest Group (SIG) on "Genetic Editions". 1 1. In a draft encoding model for genetic criticism, a stage is defined as "a reconstructable stage in the evolution of a text, represented by a document or by a revision campaign within one or more documents, possibly assigned to a specific point in time" (TEI Special Interest Group 2013). In the meantime, the term "stage" is no longer used for this purpose, for several reasons, including the potential confusion with stage directions.
The classical problem, however, is that it is often hard to identify different stages and the writing sequence within each writing stage. Especially if an author uses the same writing tool for all the text on the document (including cancellations and additions) it is often almost impossible to discern separate stages. As a first step, we therefore started by turning every manuscript into a straightforward textual version, by regarding the manuscript as "a protocol for making a text", according to Daniel Ferrer's definition (Ferrer 1998;2011, 43). A pragmatic application of this protocol model is to work with the uncancelled text of each manuscript: a reading text of a draft, without the deleted passages, generated from the XML transcription by ignoring the passages marked by <del>. . .</del> tags. This pragmatic approach aligns with Ferrer's protocol model since the uncancelled text is usually an author's last instruction to himself when he is on the verge of making a new version of the text, such as a fair copy or a typescript. This system of working with the uncancelled text was used to test the first research results of CollateX integrated in the BDMP. Evidently, researchers working on modern manuscripts are usually not just interested in the uncancelled text of a manuscript, but especially in the cancellations and substitutions. The challenge was therefore to try and find a solution for computer-supported collation of modern manuscripts, including cancellations. One of the difficulties of working with "versions" and "stages", as initially suggested by the TEI SIG working on genetic editions, was that these concepts usually apply to the entirety of the work (for instance a poem, a short story or even a novel). If, say, Samuel Beckett makes corrections in blue-black ink on a typescript, it is easy to discern this writing layer in blue-black ink as a separate stage in the revision process. But Beckett often used the same blue-black ink he used for the first draft, fair copies, and revisions, which makes it hard to distinguish separate stages.
For instance, the following sentence from the manuscript of The Unnamable is written in blue-black ink: how can you reflect think and speak say something at the same time 2 The words "reflect" and "speak" are both cancelled in blue-black ink and the additions "think" and "say something" are also written with the same writing tool. In this particular case, it is likely that Beckett made the two substitutions during the same stage of revision ("campagne de révision'), which could be represented as a two-stage development: S1: how can you reflect and speak at the same time S2: how can you think and say something at the same time But theoretically it cannot be excluded that "think" was added later than "say something", i.e. that these two separate additions constitute two separate stages: S1: how can you reflect and speak at the same time S2: how can you reflect and say something at the same time S3: how can you think and say something at the same time Nor can it be excluded that the order of the two different additions was the reverse: S1: how can you reflect and speak at the same time S2: how can you think and speak at the same time S3: how can you think and say something at the same time The markup of a transcription would have to indicate that, in the first case, "think" was an addition at stage 2 and "say something" an addition at stage 3; and the reverse in the second case. In complex cases, it is sometimes impossible for the editor to make this kind of decision, and researchers may want to check against a facsimile of the original manuscript anyway.
My suggestion is therefore to apply the notion of versions to smaller units of text, and to specify the size of the unit. Apart from text versions (Textfassungen) one could, for instance, work with sentence versions, or, for the purposes of collation, with even smaller units, such as versions of a word or of a "segment" (which can be as small as a space or a punctuation mark). This is another example from the manuscript of The Unnamable: unable to stop seeking why, why the cause of the cause of this need to talk (HRC MS SB 5-9-3, 15r) Instead of dividing this one sentence into several witnesses for collation purposes and obliging the editor to make sometimes impossible decisions with regard to the relative chronology of different substitutions, working with segment versions only requires a chronology of each substitution in isolation. In the example quoted above, the XML encoding could read as follows: unable to stop seeking <subst xml:id="subst1"><del xml:id="del1">why, why</del><add xml:id="add1"><subst xml:id="subst2"><del xml:id= "del2">the cause of</del><add xml:id="add2">the cause of</add></subst> </add></subst> this need to talk This could be visualized synoptically: In this example, "witness 1" consists of the uncancelled text of the manuscript, which serves as the protocol for the writer when he makes a fair copy, but the information about all cancellations and additions is preserved and can be visualized if the user so desires.
In the first typescript, the "réécriture" is rewritten yet again ("the cause of" becomes "the cause, the cause of") and another unit ("seeking") is briefly revised ("searching for"), only to be restored to "seeking": If users do not wish to be bothered by the cancellations and only want to see a collation of the uncancelled text, an option "hide cancellations" (as one of the "Tools" in the menu) could simplify the alignment table, reducing it to a visualization of the different versions' uncancelled text only.
From the point of view of scholarly editing, this would be a way of bringing genetic and textual criticism closer together, by both visualizing the "réécritures" and collating variants between multiple drafts. why, why the cause of w1 unable to stop seeking the cause of this need to talk why, why the cause of w1 unable to stop seeking the cause of this need to talk searching for w2 unable to stop seeking the cause, the cause of this need to talk w1 unable to stop seeking the cause of this need to talk w2 unable to stop seeking the cause, the cause of this need to talk

Multiple Drafts
Multiple drafts can also be approached from other perspectives. In the 1990s, the philosopher Daniel C. Dennett suggested his so-called "Multiple Drafts Model" to describe consciousness. 3 Dennett compared the workings of the conscious mind to a process of editorial revision: "These editorial processes occur over large fractions of a second, during which time various additions, incorporations, emendations, and overwritings of content can occur, in various orders" (Dennett 1991, 112). Observations or feature discriminations are spatially and temporally distributed over various specialized parts of the brain and combine into narrative sequences that are subject to continuous editing. The result is that "at any point in time there are multiple 'drafts' of narrative fragments at various stages of editing in various places in the brain" (113) and there is "no single narrative that counts as the canonical version, the 'first edition' in which are laid down, for all time, the events that happened in the stream of consciousness of the subject, all deviations from which must be corruptions of the text" (136). This is an adequate description, not just of the workings of consciousness, but also of genetic criticism's attitude towards textual versions, and towards "authorial invention". This "invention" consists of a dialectics of composition and decomposition, as Nelson Goodman already suggested in Ways of Worldmaking (1978; see also Ferrer 2011, 180). And it is not only applicable to literary invention. Take for instance the Declaration of Independence. Thomas Jefferson wrote the draft of the Declaration, and in his autobiography he describes the reactions to it when he presented it to Congress. On 28 June 1776, a Friday, the draft was read out to the members of Congress. They had a weekend to think about it and the next week they discussed it. Jefferson includes a transcription of the original draft in his autobiography, with the following justification: "As the sentiments of men are known not only by what they receive, but what they reject also, I will state the form of the Declaration as originally reported. out by Congress shall be distinguished by a black line drawn under them, and those inserted by them shall be placed in the margin, or in a concurrent column" (Jefferson 2003, 337). One of the passages that remained unchanged was "the pursuit of happiness". Only one man stood in the way of this pursuit of happiness, the King of England. As part of the rhetorical strategy of the Declaration, all the King's misdeeds were enumerated, for several pages. One of his cruelest outrages was to allow the slave trade and "to keep open a market where men should be bought and sold" (Jefferson 2003: 340). This important paragraph on slavery was famously cut by Congress. As a collective, the members of Congress weighed the options -should they abolish slavery or not? They eventually decided to omit the paragraph. This omission constitutes one of the most striking "what ifs" of American history. What if this paragraph had not been omitted, could the civil war have been avoided? This draft is therefore a historically valuable, material remnant of democratic decision making, and perhaps even more importantly, a material trace of democratic hesitation, if hesitation can be collective. For decision making often implies hesitation -both public and private decision making. This document is also a material vestige of a more private process of hesitation and decision making. The notorious slavery passage was preceded by a shorter one that was cut in one go. It starts as follows: "He has incited treasonable insurrections of our fellow citizens, with the allurements of forfeiture and confiscation of our property" (Jefferson 2003: 340;emphasis added). With a team of preservation researchers at the Library of Congress, Dr. Fenella France discovered in 2010 that the word "citizens" was written on top of another word, which Jefferson had carefully scraped away. The scraping is significant in itself, because elsewhere he merely cancelled passages by crossing them out. With a technique called "hyperspectral imaging" the team managed to decipher the word. Jefferson had first written "our fellow subjects". The difference between "subjects" and "citizens" is vast. After centuries of being ruled by kings who claimed to have received their mandate directly from God, a people now decided that this was no longer the case: a leader receives his mandate from the people, and therefore has to earn it. And yet, even for Jefferson, it turned out to be hard to put this into words. Even as he corroborated this tyrannical King's incompetence with numerous pieces of evidence, he kept referring to the people as "subjects". The decision to write "subjects" may have been prompted by what neuroscientist Jan Lauwereyns calls "bias". Investigating the neural underpinnings of decision making, he emphasizes the crucial role of "the prior" in the assessment of probabilities and the way "neural circuits weigh the options" (Lauwereyns 2010, 14). But after having written "subjects", Jefferson apparently noticed the "bias". He caught himself lapsing into this prior, default formulation, scraped it away as thoroughly as he could, and replaced the word with "citizens".
This may have been a one-off lapse, but a little further in the document one notices that, even for the most eloquent revolutionary, it took some time to become fully aware of, and come to terms with, the idea of freedom. Jefferson still thought it necessary to present the citizens as "a people who mean to be free [. . .] a people fostered and fixed in principles of freedom" -hesitantly wrapping the idea of freedom in a cloud of words instead of unwrapping it completely: A prince whose character is thus marked by every act which may define a tyrant is unfit to be the ruler of a people who mean to be free. Future ages will scarcely believe that the hardiness of one man adventured, within the short compass of twelve years only, to lay a foundation so broad and so undisguised for tyranny over a people fostered and fixed in principles of freedom.
Only as part of a collective, together with all the members of Congress, did he decide to replace the long, hesitant description by one powerful adjective, "free" -a free people: A prince whose character is thus marked by every act which may define a tyrant is unfit to be the ruler of a free people.
But it was the hesitation itself that served as the catalyst for decision making. And this process of hesitation, creative undoing, and decision making has left its material traces on a document.

The Extended Mind
Perhaps the expression "to leave traces" is not precise enough as a metaphor as it may give the impression that hesitation takes place "inside" people's heads and that the paper traces are merely a record of this cognitive process. According to a current paradigm in cognitive sciences, the mind is not something inside our heads; it is "extended" (Clark and Chalmers 1998;Clark 2008;Menary 2010;Stewart 2011).
The "extended mind" is the interplay between intelligent agents and their cultural as well as material circumstances. These material circumstances in the environment can be anything. In the case of a writer, for instance, this environment can simply be a piece of paper. I suggest that we can regard manuscripts, not just as "traces of" a cognitive process, but as "parts of" a cognitive process, parts of an extended mind.
This brings us to the title of this essay, "The stuff of fiction", a quotation from Virginia Woolf's essay "Modern Fiction". In line with her rhetorical strategy to present her generation of writers as "modern" and completely different from the previous generation, she writes "that the proper stuff of fiction is a little other than custom would have us believe it". This sentence is part of the paragraph that starts with the motto "Look within": Look within and life, it seems, is very far from being "like this". Examine for a moment an ordinary mind on an ordinary day. The mind receives a myriad impressions-trivial, fantastic, evanescent, or engraved with the sharpness of steel. From all sides they come, an incessant shower of innumerable atoms; and as they fall, as they shape themselves into the life of Monday or Tuesday, the accent falls differently from of old; the moment of importance came not here but there; so that, if a writer were a free man and not a slave, if he could write what he chose, not what he must, if he could base his work upon his own feeling and not upon convention, there would be no plot, no comedy, no tragedy, no love interest or catastrophe in the accepted style, and perhaps not a single button sewn on as the Bond Street tailors would have it. Life is not a series of gig lamps symmetrically arranged; life is a luminous halo, a semi-transparent envelope surrounding us from the beginning of consciousness to the end. Is it not the task of the novelist to convey this varying, this unknown and uncircumscribed spirit, whatever aberration or complexity it may display, with as little mixture of the alien and external as possible? We are not pleading merely for courage and sincerity; we are suggesting that the proper stuff of fiction is a little other than custom would have us believe it. (Woolf 1972: 106;emphasis added) Unlike the Realists, Woolf claimed that her generation was going to "Look within" and study the mind. To a large extent, critics of Modernism have taken over this internalist way of presenting the mind, which has been called the "inward turn" of Modernism (Kahler 1973). The term "interior monologue" is a good example of this discourse. But recently this "inward turn" has been questioned: Is this really what literary Modernists were doing, "looking within"? It is clear that they try to focus on the workings of the mind, but more often than not this mind turns out to be an "extended mind". In her literary writings, even Virginia Woolf herself seems to present the workings of the mind as interplay between an intelligent agent and his or her cultural as well as material circumstances. A good example is her story "The Mark on the Wall", which is not just a illuminating exercise in what is usually referred to as "interior monologue", but it evokes a mind that is at work thanks to the interaction with an environment: a mark on the wall.
Against the background of the motto "Look within", it is remarkable that in the opening sentence of "The Mark on the Wall", the protagonist does not look "within" but looks "up": "Perhaps it was the middle of January in the present that I first looked up and saw the mark on the wall". Three times, the first-person narrator says she looked up, before the "innumerable atoms" start "falling": "I first looked up and saw the mark on the wall" [. . .] "I was smoking a cigarette when I looked up and saw the mark on the wall for the first time. I looked up through the smoke of my cigarette" (Woolf 2000, 53) 4 -and in the subsequent paragraphs she comes up with several hypotheses about what the mark might be -the head of a gigantic old nail, a hole made by a nail (but "it's too big, too round, for that"), a roseleaf, a crack in the wood. Again and again, the mind in the story "swarms" in divergent directions, but each time the text returns to the mark on the wall, and a new hypothesis about the mark sets off a new string of thoughts. The structure of the text reflects this interactive way in which an intelligent agent negotiates opportunities for interaction with an environment. In other words, the text demonstrates an "extended mind" at work. There is a divergence between what Virginia Woolf did and what she said she did ("Look within").
Another example of such an "extended mind" at work is a late text by the late modernist, Samuel Beckett. The text, entitled "Ceiling", is about the slow process of gaining consciousness or "coming to", as it is called in the text. The narrative situation is a man lying in a bed, opening his eyes. 4. James Harker analyses the story "The Mark on the Wall" and makes a link with Noë 2004: "Woolf's ordinary mind, inextricable from its moving body, is in line with the 'enactive' approach in contemporary understandings of perception. 'Vision,' cognitive philosopher Alva Noë claims, 'is a mode of exploration of the environment drawing on implicit understanding of sensorimotor regularities' (29-30)" (Harker 2011, 8).
The first thing he sees is the white ceiling: "On coming to the first sight is of white" (Beckett 2009, 129). The "dull white" of the ceiling can be read as a metaphor of the dull white of the paper (Van Hulle 2012, 285). As Richard Menary argues in "Writing as Thinking", writers' interaction with the paper is part and parcel of the cognitive process and that "writing transforms our cognitive abilities" (2007,621). The nexus mind/manuscript is a constant interplay that helps constitute the mind in the first place. In this respect, the first line of the first draft of Beckett's next work, Stirrings Still / Soubresauts, is interesting. It is a sentence that is not finished and that is interrupted by the words "comment dire": "Tout tout le temps Toujours à la même distance comme c'est comment dire?" (Beckett 2011, www.beckettarchive .org). The phrase "comment dire" indicates a failure to utter, but at the same time it serves as a driving force of the cognitive process of invention, because this is the first line of a draft that admittedly led to many aborted sections, but eventually also to the published work Stirrings Still (translated by Beckett as Soubresauts). The same phrase eventually became the title of Beckett's last work: "Comment dire", or in his own English translation: "what is the word". The text was written after an accident. In July 1988, Beckett fell in his kitchen and was discovered unconscious. The diagnosis was inconclusive; he had a neurological illness, but the cause was uncertain. The effect, however, was clear enough: he temporarily suffered from aphasia. While he was recovering, he wrote this text, first in hospital and then in the nursing home Tiers Temps, as he indicated on the first page of his first draft.
The work looks like a poem, but it can be read as the attempt to write one single sentence and the failure to complete it. The text contains more lines of creative undoing than lines that represent the sentence that does not manage to be written. The longest version of this sentence is: folly for to need to seem to glimpse afaint afar away over there what -This is followed by the last few lines, "what -", "what is the word -" and again, after a blank line, "what is the word", this time not followed by a hyphen.
The process of writing or failing to write the sentence is scrutinized meticulously. At regular intervals the text is interrupted by the phrase "what is the word". In the earliest French version this phrase is "quel est le mot" ["what is the word"], not yet the more idiomatic French expression "comment dire". And the first word was "mal", not yet the word "folie", as in the published version. 5 What drives the text is the question of an accurate description of "tout ce ceci-ci". In Beckett's own translation, "all this this here" is presented in the following way: folly seeing all thisthiswhat is the wordthis thisthis this hereall this this here - (Beckett 2009: 133) The word "seeing" is reconsidered and replaced by "given": "folly given all this this here -" and eventually this whole part of the sentence is simply cut. From that moment onwards the word "folly" is immediately followed by "for to -" and by the process of choosing the right verb. The first verb that comes to mind, "see", is immediately replaced by alternatives that imply more ambiguity: folly seeing all this this herefor towhat is the wordseeglimpseseem to glimpseneed to seem to glimpsefolly for to need to seem to glimpse -In the manuscript of the French original, this is the place where the text splits up into two alternatives. The first alternative starts with the verb "to see" (à voir); the alternative in the second column suggests that even the verb "voir" was preceded by a moment of hesitation (à -/ comment dire -/ à voir -). At this juncture in the manuscript, the text (which up to that point looks like one long column of short interrupted lines) splits up into two columns, two alternative versions. This is more than just a moment of hesitation in the manuscript. Beckett may have been in doubt as to how 5. The French published version opens as follows: "folie -/ folie que de -/ que de -/ comment dire -/ folie que de ce -/ depuis -/ folie depuis ce -/ donné -/ folie donné ce que de -/ vu -/ folie vu ce -/ ce -/ comment dire -/ ceci -/ ce ceci -/ ceci-ci -/ tout ce ceci-ci -".
he was going to proceed; at the same time this doubt has left its traces on the manuscript in the form of a draft that splits into two alternatives, and this hesitation is subsequently "performed" in the published text. In his last work, Beckett is giving shape to what, in his first published text, he said about James Joyce's "Work in Progress" (before it was called Finnegans Wake): "Here form is content, content is form. [. . .] His writing is not about something; it is that something itself" (Beckett 1984, 27). To illustrate his point, Beckett chose the word "doubt". The English language, according to Beckett, was "abstracted to death" and Joyce brought it back to life, by developing a new expression of the abstract concept "doubt": "in twosome twiminds" (28). This is an apt description of what happens in the manuscript of "Comment dire": the text is constantly "in twosome twiminds" and the process of thinking and writing is driven by a dialectic of composition and decomposition. What he wrote in his first publication (the essay on Joyce) can be applied to his last work: here, form is content, content is form; the text is not about a cognitive process, it "is" that cognitive process itself. Even though the text never makes explicit what "this this here" is, the insistent deixis draws its readers' attention to the materiality of the process of cognition, the interaction between neural processes and the writing surface.
In conclusion, my suggestion is that "this this here" -the material aspect of the writing process -is something we, scholarly editors and genetic critics, can draw literary critics' attention to, by arguing that it is an inherent part of the "extended mind", and by showing that this material aspect can contribute to a better understanding of the Modernists' literary attempts to evoke the workings of the mind. This implies that our task is not just to produce scholarly editions, with critical apparatuses, but that we might also consider it our task to show that also the cancellations in even the most chaotic drafts matter, that these "multiple drafts" contain relevant information on the "extended mind", and that "all this this here" is the "stuff of fiction".