A Wandering Hyphen, The Reader and the Work

This is a reply to commentary by Matt Cohen, Ian Cornelius, and Alan Galey occasioned by the publication of Paul Eggert’s The Work and the Reader in Literary Studies: Scholarly Editing and Book History (Cambridge University Press, 2019) and to a review of the book by John K. Young. A theory of the work based on the negative dialectic of document and text grounds the work as a regulative idea rather than an ideal entity and finds the role of the reader to be constitutive of it. The relationship (envisaged in the book as a slider) of archival and editorial digital projects, the potential cross-fertilization of philology and textual criticism, and an expanded role for textual studies inspired by D. F. McKenzie’s writings are discussed. “I have so often seen poor Chas stamping mad about such calamities”.

She knew what it was like to have a prickly husband whose temper could occasionally get the better of him, especially when his prized poetry was treated with casual disregard when it was being readied for publication in newspapers, the predominant publication venue for poetry in colonial New South Wales. In the extant draft of a letter that Harpur wrote to an unnamed newspaper editor in ca. 1867-1868, this time relating to some essays on various poets he had written, he complained: Not only am I made to write nonsense by the most stupid errors of the press, such as "a suffering (for sufficing) and final repose &c" but all the niceties of the criticisms depending on italicised lines & passages, scanned metres and so forth, are quite set at naught-separate passages all run and blurred together-&c &c &c! 1 A review of my book The Work and the Reader in Literary Studies made me wonder whether I must count it amongst my personal &c &c &c's -until, that is, I recalled my own arguments about the work-concept and the role of the reader in the work, and until I saw the variety of responses that the book had called out by the scholars participating in this forum, who had, in their distinctive ways, made it their own and part of their ongoing intellectual journeys.
But first to my &c &c &c's. In the review John Young -who devised the round table for which these papers were written -does an excellent job of summarizing some of the principal arguments of the book. I'll quote at some length, finishing at the wandering hyphen that made me temporarily identify with Chas.
Eggert makes the case throughout for a new orientation toward the work-concept, grounded in the role of the reader and also in the ineluctable realities of the digital realm. "What we need now", he concludes, "is not so much a new taxonomy of editing and archiving as a sliding scale of document and audience orientations [. . .] one that can selfadjust as new approaches in the digital arena are projected and put into operation" (86). This "sliding scale" is itself a central metaphor here, as 1. See charles-harpur.org/Letters/Twin/?docid=english/harpur/letters/ML/A87-1/ 1867-HARPUR-UNKNOWN: "repose" was presumably a mistake for "response". The letter in the epigraph is at charles-harpur.org/Letters/Twin/?docid=english/ harpur/letters/ML/A887/28-MAY-1881-MARY_HARPUR-PARKES . Accessed 14 May 2021.
Eggert envisages a slider model for digital editions, which are also almost always digital archives, that would allow editors and users to adjust along an archival-editorial spectrum. The digital edition is therefore "better described as the editorial layer of the complete project" (90), coming into being as the result of particular choices about how to represent historical documents. Eggert clarifies this distinction: "While an archival transcription is an attempt to capture the text of a historical document (representation), an edition claims to make present the text of the thing that has been subject to the editorial analysis (pre-sentation)" (91). (Young 2020, 44). 2 This last sentence would be a perfectly accurate quotation except that, in Young's text, there is a hyphen between "pre" and "sentation". But the hyphen is not mine!, exclaims Eggert, this present Eggert, me. The hyphenated "'presentation'" is in fact hyphenated in the book over the end of a line on page 91 as "pre-/sentation". Ordinarily, we might well pass over this as a typo in the review except that it is reported within inverted commas as part of Young's quotation of me. Having finished the quotation he then immediately quotes from my book's page 7 that "the hyphen makes all the difference", thus making it look as if the difference-making hyphen is the one in "'pre-sentation'". The oddity, drawn attention to in this way, makes it look like something oh-so-cleverly intended by this Eggert character, some nuance or subtlety that, if pondered, reveals all. Alas, no, it doesn't. Eggert, I can report from long experience, is not as clever as that.
In some playful esprit, Finnegans Wake may well contain the nonce-word "sentation" (I haven't checked); but the Oxford English Dictionary abjures all knowledge of it. Nor is it a case of some forgotten medieval rhetorical trope for a deconstructionist suddenly to pull out of the hat. It's a nothing, a non-sense. The Introduction to my book -where the quotation that "the hyphen makes all the difference" comes from -first states the representation vs presentation distinction before Chapter 5 expands on it: the new "edition makes the work present. It does this by resting on a documented past. It is re-presentation but not a representation: the hyphen makes all the difference" (Eggert 2019, 7).
So the real hyphen (from which, in the book, a lot swings) as opposed to the wandering hyphen of Young's report from page 91, is in the word "re-presentation" not in the word he has invented "pre-sentation". My hyphen is meaning-bearing, whereas his is not. My hyphen inserts a distinction between archival representation and editorial re-presentation, with the archival more oriented towards the documents (which can be more or less satisfactorily represented by facsimile image or by transcription), and with the editorial more oriented towards the reader (for whom the editor's argued understanding of the text is made present in the edition). It's the making present that the edition performs. It rests on a historical archive, filtered by editorial assessment of the authority of the documents. The editorial impulse pushes the editor to anticipate the reader's needs at every turn, as each emendation shows. As they play their role in the textual performance, as readers realize the reading text, they absorb the emendations, one after the other, and often without recognizing them as departures from the historical documents on which the reading text is based.
Pondering that errant hyphen -if so little a thing can be ponderedforces our attention back to our own processing of meaning as we read. It makes us reflect on those features of text we deliberately overlook. Now, as Young might well claim, and claim correctly, the hyphen in his nonce word "pre-sentation" is actually there on the page he quotes from. The hyphen is there after "pre", right at the end of the line. So he hasn't misquoted in the sense of making something up or through inattention. Rather, the opposite. He has paid it too much attention. He has assumed to be semiotic and intentional an aspect of text that is merely bibliographical. The endof-line hyphen in a non-compound word such as presentation is a marking on the page that is rendered necessary only as a result of the bibliographic tradition of right-justifying lines of printed prose text. Had the word been a compound, such as re-presentation, Young's wings of fancy might have got him aloft; but, as is, the hyphen means nothing more than: This is a marking necessitated by its documentary carrier and inserted automatically by a typesetting program. It is one of those bookish signals that we train ourselves to ignore. Spaces between words and between lines of type are there too, but we are trained only to read around them or through them. We don't even know how to quote them, only describe them. The ingenious John Young has found a way of quoting the meaningless simply by raising a halfelement of it (the hyphen) and ignoring its partner, the justified righthand margin. 3 The book's author has nothing to do with it. Eggert disclaims all responsibility. Even as he professes himself grateful for the excellent, clarifying distillation of his argument that Young has just offered, Eggert grinds his teeth in chagrin at the wandering hyphen that, in Young's version, threatens to carry away the argument at the very moment when, at Young's hands, it reaches its nirvana, its quintessence. Eggert throws himself into Chas's corner amongst the &c &c's, moaning and whimpering.
Yet has he any right to complain? Editorial theory has been teaching us to see bibliographical nothings ever since the 1980s. Nowadays we all agree that the printed or manuscript page or the screen is the site of meaning; and that we as readers can't but be affected in ways large or small by that location as we raise the continuous flow of meaning that, for simplicity, we call the text. Young's misprision reminds us of what reading, at its heart, is and how it is affected by where it happens and to whom it happens.
Take, for instance, the phrase "Attic vase" in Keats's poem "Ode on a Grecian Urn". It may well mean, for one of Jerry McGann's students, 4 the vase encountered in their grandmother's attic a previous summer, rather than referencing, as Wikipedia tells us, the "sculptural style, beginning in Hellenistic sculpture and vase-painting of the 2nd century bce and climaxing in Roman art of the 2nd century ce, copying, adapting or closely following the style shown in reliefs and statues of the Classical [. . .] and Archaic [. . .] periods". 5 Nevertheless, for that student encountering the phrase for the first time, the "Attic vase" must be one of those half-forgotten vases discarded in an attic. It is all very well to learn that, before writing the poem, Keats had himself "traced an engraving of the Sosibios Vase (which dates to ca. 50 bce) after seeing it in Henry Moses's A Collection of Antique Vases, Altars, Paterae". 6 But you have got to know that, or at least be aware in a general way of what the Attic style was, before it can affect your reading of the poem. The teacher and the student have equal rights 2021): "I enjoyed reading your panel response and more fully understanding the hyphen. I can now see why [the proofreader] and I would not have removed or queried the hyphen. I've attached the MS Word version from which I typeset. It has the reading as printed, and John and [the Reviews Editor] also both proofed the essay, so I think John thought you did mean that the hyphen mattered, but you'll have to ask him". I did, and he did think so; to the poem, don't they? Both are readers; both actualize meanings as they read, meanings they identify with the poem's text?
The answer to this equalitarian dilemma is not to become angry and insist on the earned authority of the teacher over that of the student but to recognize that textual meaning raised in the act of reading is not free floating, as McGann's anecdote teasingly seems to suggest it is. Rather it is always intimately related to the documentary site upon which the raising of meaning occurs. In the book, I describe this relationship as a negative dialectic: that is, the textual and the documentary are dialectical in relation to one another. They need one other to establish their own different identities: a document is only ink on paper until some fragment of text is raised from it. Text, as it were, lifts the status of the ink and paper into that of a document. Once you accept that, there is a consequence: there is no ideal meaning to appeal to that transcends a negative dialectic. Now, ideal texts purport to transcend. They give the pedant in us a stick to beat the student with. Great result. Adhering to the idealist concept (the ideal text of the work) has been a dragging chain for editors for decades. In my workmodel it is dispensable.
But if you decide to dispense with it you have to take the negative dialectic itself seriously: if the textual and documentary dimensions are understood as interlocked in their very identity then our adjudication of possible textual meanings is forced to take cognizance of the carrying documents. Unlike the subjective vagaries of purely textual meaning-making in the present, documents have a definite history that can potentially be traced, and agents responsible for their inscription. We have a method for such tracing. It is called bibliography.
There is another consequence of the model. As well as operating synchronically in the moment of reading, the negative dialectic operates across time, diachronically -as reception histories remind us. Because textual meaning is experiential, readers must be considered to be a constitutive part of the work. This is why we can say works have lives: those agented experiences alter over time and usually in relation to different printed documents.
And there is one last consequence, a little sad: the theory allows Eggert's chagrin to be seen as the fate of all writers, Chas included. "My kingdom for a hyphen" is henceforth to be his motto.
This series of consequences leads to the overarching question: what then is the work? My answer in the book is straightforward: the work is the container for and the name that we give to this unfolding interlocked process of reading and re-reading. The stable thing in all of this is not so much the realized textual meaning (which is inherently shifting) as the biblio-graphic site of it: the material document, which scarcely changes at all. The documentary dimension keeps the work hooked into a history that is material and agented, and into history more generally; the textual dimension releases it into repeated presents and therefore into discursive shifts. We need the work-concept to discuss our experiences of reading. That is, it is a regulative idea not an ideal. Works exist only in our dealings with them and in the dealings of historical others with them. Works also unfold over time. They are never finalized, unless their documentary embodiments are abandoned entirely or utterly destroyed. This work-model is at the center of the book. The rest of the book pursues its implications, especially for literary study. John Young concentrates on how we should understand the relationship of archival and editorial expressions in digital projects. He is interested in the slider, I suspect, because it locates and thereby validates editorial intervention. Michelangelo Zaccarello, who also spoke at the round table, adverted to the ongoing need for critical intervention more than once; and he is a medievalist for whom the need is unignorable. So I have to assume, putting the two together, that the book has hit on what has been an exposed nerve for scholarly editors in recent years. The book is saying that digital archival projects of literary works do not, in fact or in prospect, spell the end of scholarly editing.
The theory on offer in the book is a theory of the work not a theory of communication. It sets out the aims that editions are capable of achieving and the role that the editor plays in them. Materiality -and therefore agency and the intention of the document's mark-makers -have their role to play here. Textual agency remains an interpretative vector, activated by the textual critic on bibliographical ground.
Put another way, what is on offer in the book is a phenomenology of the work rather than an ontology. Pursuing the latter has, in the past, led us into idealist confusions: the text of final authorial intention, the definitive text and related ideals. That said, however, it is equally true that the text of a document is not, or not necessarily, the text of an intended version: establishing the latter, in the slider model, remains an editorial operation. § As I read his review I realized that John Young had gone down this line of reasoning with me. But just at the moment when he sees the representation-presentation divide face-to-face rather than through a glass darkly, his act of reception becomes an act of composition. As author, I had to shake myself to realize this obvious fact: that the review is a shaped performance of his reception of my book, and that he was now on his own. He reads into the wandering hyphen an elusive meaning that he imputes to me but that is in fact his. The hyphen somehow releases him; the baton is handed over; and Young is away. He foreshadows some fertile thinking, yet to come, about how the work-concept might be applied to other media including television, music, and even dance. Now I know something new: that, a little further down the track, I want to be his reader.
Much the same can be said of the way reading the book has stimulated agreement, disagreement, and suggestive tangents from Matt Cohen, Ian Cornelius, and Alan Galey as they wrote their pre-circulated contributions to the forum. I thank them and Michelangelo Zaccarello for taking part, and I thank John Young for originally proposing the idea for the 2020 STS conference and Marta Werner for chairing the forum, which finally took place online, at the STS conference in May 2021. They are all of them strong readers and, even if I may disagree with them here and there, what gives me joy is that their encounter with the book has obviously made them push their own thinking further. If there is misprision it is creative misprision. That's as it should be; it's why members of STS come together as a society of scholars with allied interests. It is to bounce off one another. We stimulate one another to go further in our thinking, even if this generally entails a few side-skirmishes along the way.
Encouragingly, the book offers, in Chapter 9, a general model of this process. Just as authors, along with amanuenses, typists, typesetters, bowdlerizing publishers, and censors must read the text before they alter it for their particular purpose, so readers after publication participate in the work by virtue of their reception of its text. Adaptation requires the taking of one extra step. Reviews and parodies are adaptations that stick close enough to the subject matter of the original for it to be recognizable before launching out on their own. Changes in genre (say, from novel into film) require different techniques and refocusings of the shared subject matter for new audiences. One adaptation may give rise to the next, and so on with increasing variation, so that eventually a new adaptation of its predecessor may share nothing at all with the original work. In each case, reception engenders production; and the new work will be subject to the different cultural valencies of its time and place when and where it must establish its own readership or audience.
Scholarly responses are a special sort of adaptation. They cast the original argument, or, as in this case, the set of related arguments, under a sharp new light that casts unexpected shadows. A reweighting of factors, or the specification of other relevant ones originally overlooked, will be argued for. So, for instance, Ian Cornelius starts by observing that, in the book's "general program for the study of literature centered on the question 'What is the thing read?'[, c]oncepts of document, text, and work are parsed with care, generating many valuable insights and clarifications, but there is need for more thinking about the linguistic medium of literature. To textual studies, bibliography, and book history -the trio of foundational disciplines advocated by Eggert -one should add philology, or the study of literary language" (2021, 39).
The work's capacity to sustain translation and circulations of various kinds is relevant here. Translation can be treated as a deliberately constrained form of adaptation; but more thinking than the book offers is probably needed. 7 Cornelius's comments on world literature make me wonder whether incorporating a geographic or social-geographic vector might usefully particularize or make concrete the book-historical methodology around reception and the life of the work, for which the book argues and then exemplifies in case studies.
Cornelius is also right to note a scanting of linguistic interest in the book. For me, the language medium is the thing that goes without saying, the sine qua non. As mentioned above, the book offers a theory of the work and the reader to undergird literary study, not a general theory of communication in language. Of course, the workings of the work depend on the operations of language, but I have been reluctant to fully engage with this linguistic dimension because of its familiar synchronic or structuralist model for language as semiotic system. My recommended model for the study of literature is fundamentally diachronic, historical. I cannot see how to marry the two approaches conceptually unless it would be based on a social principle that linked back to the diachronic: that is, on language seen as socialized and socializing, and consequently changing over time.
That is not at all to say that my theory of the work ignores the operation of language in the present; rather, it operates downstream of it. For instance, it offers theoretical backing for the teaching of close reading in the present by positioning new readers as part of the reception history of the work and, more fundamentally, by casting readers as being always constitutive of the work through their realization of meanings recumbent in the document.
It is not that the work is implied by the document or that it hovers behind the document. That easy assumption leaves us with an unsustainable ideal-7. See Eggert 2019, 155-59. In FRBR library cataloguing (see 216-17 n 28), the translation is understood as an "expression" of the work whereas the adaptation is seen as a new work linked at the level of subject matter.
ism. Misleadingly, the formulation grants agency to the document when it is in practice the reader who is the agent. The reader activates the work's textual dimension as the negative dialectic unfolds in a present that, soon enough, will become historical. 8 So, for now at least, I am unapologetic about not dealing with the thing that goes without saying; but then I'm not a linguist, and I cannot rule out a future and so-far unconceived Chapter 11 of my ten-chapter book. More promisingly, and now more willingly accepting Cornelius's lead, I can begin to glimpse a redefined philology as that unwritten Chapter 11; and I very much hope he goes on to write it. René Wellek and Austin Warren famously argued in their 1948 Theory of Literature -the textbook that sustained a generation of New Critics -that the term "philology" be abandoned. Literary theorists from the 1980s took them at their word and pronounced the funeral rites. 9 Given the emergence of the new field of material philology in the 1990s, however, a meeting of the two domains can perhaps be envisaged as taking place in ways that would complement the program for literary study that my book puts forward. For instance, Cornelius mentions Sheldon Pollock. My memory of the stimulating keynote that this renowned philologist gave to the 2014 meeting of STS in Seattle at the University of Washington, laden as it was with suggestive parallels with editorial practice and theory, both ancient and modern, suggests that it could. 10 8. For the relevance of this to the work of Bruno Latour and Rita Felski, see Eggert 2019, 15-18. 9. Wellek and Warren said they preferred the term "historical linguistics" (1956,38). 10. The inspiring essay by Pollock (2014), to which Cornelius points, is exemplary in its attempt to incorporate into a single model: (1) historicist readings of works; (2) the tradition of interpretation over time; and (3) readings in the present. The historicist tradition of interpretation was, Pollock explains, initiated by Spinoza and aimed at elucidating the original moment of authorship of the Scriptures in defiance of the dominant Dutch tradition perpetuated by theologians and clerics. As for (3), Pollock brings Gadamer's hermeneutics to bear: "What the text is 'really about,' for Gadamer 'can be experienced only when one is addressed by [it]" (qtd. Pollock 2014, 401). However, by substituting "text" for "work" in many of his formulations Pollock is forced to concentrate on meaning-creation by the reader, so that the identity of the "text" becomes its historical "assemblage" (2014, 410) of meanings. But this is to ignore or downplay the documentary vehicle of the work and thus to minimise the significance of textual variation as a historical thing. It is not just a case of "bad textual § Matt Cohen's response to my book is a meditation on issues it raises as he reflects on how they apply -or don't -to his considerable experience of working on the Walt Whitman Archive. His essay is studded with shrewd observations. I sense that my book has been both a salve and an irritant for him. This is welcome to me because it has evidently brought out his thinking in a way that a book in praise of social editing, say, or on digital editions written from inside the TEI community might not have done. Cohen is mainly concerned with my account of what editions are and how they operate, aspects of which he finds welcome. His disagreement focuses on Chapter 5, which is entitled "Digital Editions: The Archival Impulse and the Editorial Impulse". Clearly, Cohen has fallen out of love with the printed scholarly edition: preparing one is "a truly heroic, longue-durée, or foolish exercise", he says (2021, 34). "Is it time to give up the moniker 'scholarly edition'", he asks (2021,29). My answer to his question is no, since the term denominates a central concern with the transmission of the text that other forms of criticism and theory simply take for granted, to their and to our cost. Cohen is resistant to my position that each scholarly edition embodies an argument about the relevant materials in the archive. He wants to leave room, I think in view of his Whitman work, for "the idea that an edition exhibits heterogeneous arguments" (2021, 31). He also sees problems with my modelling the relationship of the archive and the edition in terms of a slider between the archival and the editorial, whereby one's current position on it reflects one's present task. That task may range from transcribing, with one's eye intently on the document in question, aiming for close fidelity to it (so, more archival) to, say, an editorial reconsideration of the textual relationships of the extant documents. This might lead to emendations aimed at capturing the text of a versional stage in the life of the work, one that is not identical with any of the documents.
variants" (Pollock 2014, 406). With the document sidelined, "the text and the properties the text possesses" (as if this isn't historical too) (2014, 406) start to assume, as a necessary counterbalance, a pseudo-objectivity that undermines Pollock's model. So he is tempted to go further: "every reading is evidence of human consciousness activated by the text in its search to make sense of it" (2014, 407). The problem is in the "it", for there can be no such singular thing if his model is to hold. The idealist ghost re-emerges as the essential thing in the work that resists closure of interpretation.
Cohen objects that, in the TEI world, this slider model pays regard only to the body of the transcription -the text-wordings wrapped in their codings that bend them towards TEI validation -and thereby ignores the TEI header where relationships among documents witnessing a particular work may be specified and where it is possible to record many other interpretations besides. This metadata may guide the transcribers. Furthermore, the well-constructed header may be used to generate different displays for different purposes. Of course, if one is using TEI, he is right. Yet the two activities (provision of metadata and transcription) are different (indeed, the slider implies as much); and the two are generally in a feedback loop until things settle down during the long transcription process. But where the text-critical and editorial interpretations are recorded is not the point at issue in the book. Indeed, leaving a theory dependent upon current technical approaches that will themselves pass with time is a folly the book counsels against. That is why, at a higher level of generality, the book differentiates archival and editorial impulses, the one more intently document-facing, the other turned more towards readers and their anticipated needs.
I do have one bone to pick with Cohen: I don't understand why he thinks my "vision [is] of unitary archives" (2021, 33). My experience suggests that they will usually be growing and changing; and an ideal archive would permit many editions making contrary arguments and serving different readerships. Without Contraries is no progression, said Blake -and it applies here too. There are technical challenges implicit in Cohen's observation that "[t]he permeability of the editorial and archival activities when they have developed as a function of determining the editorial argument means reduced flexibility for those who do not agree with that argument. [. . .] [I]f another editor cannot adjust the metadata in your archive, she cannot make an editorial argument whose consequences will reach all the way down to the basic infrastructure of the document set" (2021, 33). The consequence -that "[s]ubsequent editions, if not performed in collaboration with the generators of the archive, will remain superstructural with relation to the primary data" (2021, 33-34) -is a recipe neither for Contraries nor for progression. Yet this is the corner that TEI pushes you into.
Good editions (in whatever medium) benefit from fresh bibliographical or codicological analysis. As the analysis often exposes gaps in or misunderstandings of the textual transmission, editors sometimes have to establish the texts of intervening documents that no longer exist, meaning that the need to revise digital archives will continue. A unitary archive-edition relationship is not the way to stimulate fresh text-critical thinking. Actu-ally, I suspect that Cohen and I are in agreement about our ideal outcome here: our disagreement may only be around how to achieve the ideal technically. In general, I think it's best to leave things as flexible and simple as is feasible. If that's the aim, then TEI-XML requirements and unfriendly databases may themselves be the problem. 11 The book argues that there can be, in actual practice, neither a purely archival nor a purely editorial position in relation to the relevant documents: that is one reason for the slider. At the start, the project's parameters will have dictated a particular set of documents for a reason, and that reason will to a certain extent set the terms for the initial inquiry into document relationships: for example, the manuscript, typescripts, proofs, and printings of a particular work or oeuvre. Existing bibliographies may help, but understandings of the document set will mature as the transcription stage proceeds.
When the time comes for editions of multi-witness works (or, equally, of multi-witness versions) -rather than only editions of single documents -the needs of readers must figure more prominently in the planning. An edition that, as Cohen puts it, "exhibits heterogeneous arguments" (2021, 31) sounds attractive until you realize it would be in danger of becoming illegible or incoherent to the reader. But editions (plural) that draw upon the same or much the same set of archival materials to embody different arguments will avoid that fate, provided the terms and implications of each argument are made plain to the reader. A good digital archive will provide the conditions for competing editions, argued on different grounds, to flourish. So Cohen's raising the spectre of "the grand-scale development of an argument" (2021, 32) -that is, a single editorial argument as an expression of a unitary archive -is, I believe, a bogeyman that needs to be put to rest.
I think it will, and soon: he says that the Whitman Archive has "a plan to link all known versions of a work using an array of identifiers" (2021, 31). When that is achieved, automatic collation of competing document-texts, 11. In a TEI-XML file, "metadata" means anything that comments on the document as a whole. It is normally added to the source transcription as a header -whereas, ideally for the sake of flexibility, it should be separate from and external to it. Such metadata may include datings and connections to other documents. Hence if other editors want to supply their own interpretations of those connections and datings they would be compelled to add them on top of, and in conflict with, those already produced by the original editors, if their work is to be preserved. This is an intrinsic flaw of the all-in-one TEI approach.
indispensable to the scholarly editor, should be just around the corner. It is good to hear that the Whitman Archive is now heading in that direction: further to the right on the archival-editorial slider. § In Alan Galey's broadened textual-studies response, the ambition behind the project he foreshadows is to be warmly welcomed. In principle, it would include the study of all forms of recorded cultural expression, well beyond literary works and historical documents, that, by virtue of their recorded condition, give rise to variant forms. It would serve as a substantiation of D. F. McKenzie's vision of reconceptualizing textual studies, most famously in his Panizzi lectures of 1985 published the following year as Bibliography and the Sociology of Texts. I like, too, Galey's description of that "strange disciplinary space that textual studies occupies with respect to other fields" and his rightly pointing to "the trans-disciplinary nature of concepts like work, text, and document" that textual studies thrives off (2021,(52)(53). This is surely because of the fundamentals that textual studies engages with and insists upon, but that successive poststructuralist positions scanted. Yet it should not be an either/or here: to make it that way is to put your head in the sand. Literary studies needs richer soil than that, a new conjunction. But how might that drawing-together happen? Well, book history and textual studies offer different modes of reinserting the multiple testimonies of the document into literary-critical and cultural discourse. My work-concept, with its incorporation of the reader, offers a model for putting them into operation together. For instance, circulating discourses are instantiated in both composition and reception. The recorded results or upshots, which textual studies has methods of retrieving, are open to critique. Reception is also performed in close-reading exercises in the classroom, indeed whenever reading is done, and over very long periods of time.
The work model I put forward in the book names the document as the ground upon which meaning is experienced. This way of laying stress on the textual medium is an alternative to McGann's nomination of what in 1991 he called bibliographical codes. 12 Unlike bibliographical descriptions, such codes (if we give the term its literal meaning) are unspecifiable; yet McGann is undoubtedly right that the raising of textual meaning is affected by describable documentary sites. McKenzie opened that phenomenon up as a subject for an expanded textual studies, and in doing so 12. McGann used the term codes metaphorically in 1991 but more literally in 2001.
invoked the shades of A. W. Pollard and, to a lesser extent, W. W. Greg (as opposed to the more strictly bibliographical tradition represented by the proto-scientist in Greg, and by Fredson Bowers and G. Thomas Tanselle). 13 In other words, McKenzie tilled the broader ground, and others (including me) have harvested the crop. Many of us have reason to be grateful, though whether he would have approved of our adaptations we will never know.
In The Work and the Reader I took that shared and personal background as a given; but perhaps spelling it out, as I have done in other writing in the past, would have avoided misconstruction. 14 Describing it as "liberating at the time" (Eggert 2019, 3) was not enough. Nevertheless, I fear that Galey has got the wrong end of the stick in thinking that, because I point out that McKenzie's proposed sociology of texts had, for him at least, no editorial consequences, I am against the larger widening of textual scholarship that he initiated: "[R]ather than rehash editorial theory debates from past decades", writes Galey, "I would instead step back and ask why, today, we should treat scholarly editing as the yardstick by which we judge a concept like McKenzie's sociology of texts" (2021,55). Except, of course, I don't. In the general field of textual studies I see edition-making as indispensable, not as central. Scholarly editing is only one application of textual criticism. Although not the only show in town, it is a durable one. 15 The long-distance view from the shoulders of the giant is, as they say, remarkably clear; you're lucky to be there; but what you can now see may be different from what the giant sees. Accordingly, when the need arose in my book to define the implications for scholarly editing of McKenzie's position I called it as I saw it, drawing on previous arguments advanced by Peter Shillingsburg. I first had a close look at McKenzie's Congreve edition, as Shillingsburg had done, and confirmed Shillingsburg's conclusion that while a "sociology of the text" probably assisted McKenzie's assessment of the early textual transmission of Congreve's works it had no effect on 13. See McKenzie 1993, 13. 14. See Eggert 2002. In Eggert 2019, I describe the digital edition as "only one considered application of the data" (88). This position corresponds to my earlier observation: "As an application of textual criticism, scholarly editing is therefore, I have always believed, done for a purpose. Its function is to enable, in an enduring way, literary critical and historical understanding, to provide them with ways and means. For me, the two kinds of understanding are deeply interconnected. Works are not aesthetic objects pure and simple. They are an interplay between physical documents and meaningful texts, and the interplay occurs in the minds of readers" (2007,69).
the eclectic editorial method he actually adopted, one very traditionally designed to respect authorial intention. 16 This conclusion counteracts a consistent stream in editorial theory and practice since the early 1990s, during which time preference has swung back to basing reading texts of scholarly editions on (usually) first published editions -on the grounds that this was the text encountered by the work's first readership, thus accessing social-historical authority. My book discusses the implications of such decisions and tries to clarify the confused appeals to sources of textual authority that have, as a result, been creeping into editorial discourse. It is time for a reassessment, not least of the terms "social text" and "social-text editing". 17 In the book, I point to McGann, not McKenzie, as the source of the application of textual sociology to scholarly editing (Eggert 2019, 3-5). The consequent withdrawal from critical editing using eclectic methods has been most noticeable in digital projects. They routinely concentrate on providing what their tools allow them to provide: transcriptions of documents rather than critical editions of works or versions. However, I believe this withdrawal will eventually prove to have been temporary, since the needs of readers ultimately have to be met and in as informed a way as possible.
In some of his more memorable formulations, McKenzie brilliantly drew attention to the fact of human agency in texts and books. Galey quoted one in his conference presentation on the day ("all recorded texts are of course multiply-authored . . . the product of social acts involving the intervention of human agency on the material forms" [McKenzie 1991, 163-64]). And I conclude my book with another resonant one ("The book as physical object put together by craftsmen [. . .] is in fact alive with the human judgements of its makers" [McKenzie 1984, 335]). Yet, in order to edit at all, scholarly editors typically narrow down the list of agents for primary attention (the author, a lazy typesetter perhaps, often an interfering publisher or publisher's editor). Of the remaining presence of their coworkers in the text -itself still "alive" with their judgements -there is no doubt. But, for the job in hand, editors don't normally need to concern themselves with the paper maker or the ink maker or the travelling sales 16 staff, unless an effect on the text is detectable and needs explaining and perhaps emending. This editorial exclusion is a reminder that there is no way of duplicating in the new edition the set of social circumstances that gave rise to the original print publication. As I pointed out in 1994 in a response to McGann's The Textual Condition, even facsimile editions -which hew closest of any to the original production -can reproduce nothing but the visual image of the page. 18 There is no way to reproduce the past. Only critical attitudes towards it are possible, and we cannot do otherwise than look at it from where we are. Editors who seek to evade this logic generally end up tying themselves in knots.
Over time, scholarly editors have developed ways of dealing practically with textual agency of this deliberately selective kind. Galey himself refers to "the most precise and well-honed format we presently have for representing textual variation" (2021, 57). But there was, all along, a potential and growing cost to this strategic exclusion, and, in 1985, McKenzie applied the blow torch. What he was in effect arguing was that bibliography's narrowing of attention to text and its transmission had caused a functional blindness: first, to those printed objects that persons in the book trade brought into the world and, by their professional activities, ensured would continue there; and, second, to the meaning-making activities of readers who, over time, took up the printed matter and made sense of it. The solution my book puts forward, absorbing McKenzie's lessons, is to redefine the work so that readership is constitutive of it. Editions make the work present for them.
Galey comments that "[T]he golden age of traditional scholarly editing, enabled by vanishing postwar affluence, has come to an end in North America" (2021, 55). I think that may be true in relation to printed editions, partly because of the finalizing of many of the decades-long complete-works projects initiated from the 1960s to the 1980s. 19 But the commissioning by university presses of new complete-works series, or series conceptualized along alternative lines, depends as much on the expiration of copyright and likely sales projections as on shifts in editorial or literary theory. Changes in institutional arrangements and grant funding, as well as the now-fading effect of Fredson Bowers's opportunistic but effective championship of scholarly editing that inspired a generation of editors from the 18. Eggert 1994: first given as a paper in 1992. 19. The golden age of scholarly printed editions has ended not ended in Britain: see Eggert 2019, 70. 1960s onwards, may be another part of the explanation of the lull Galey sees (and that I discuss in Chapter 4 and in its endnotes). On the other hand, digital-humanities graduate students continue, with some excitement, to be attracted to the activity as it offers research questions that the new digital tools at their disposal promise to help them answer. "Making it new" is a siren call. I hope that my conceptualization of the archivaleditorial relationship in digital projects will help such students make more theoretical and practical sense of what they are doing. Finally, Galey's irritation with scholars who fail to realize that archival work has generated its own theory is very much the professional speaking, lamenting what is unfairly overlooked. I know the frustration and sympathize. He argues that, if an edition is the embodiment of an argument (as I claim it is), an archive should be considered likewise. If this is true (and, though not a professional archivist, I suspect it is true), the argument will nevertheless be of a different kind, or different in scope, to that of the edition. 20 This divide is, in effect, what the slider-model I propose in Chapter 5 tries to bridge. But, as I point out, "the looser metaphorical definition" of archive has attained a currency in literary and digital-humanities circles because it has become useful -not because it respects the principles-based definition shared in the archive community. 21 One implication of Galey's plea for textual scholars to pay more attention to archival theory and practice concerns the book's slider metaphor. He provides an intriguing musical example of two competing recordings of Duke Ellington's famous concert at the Newport Jazz Festival in 1956. The Columbia LP of 1956 released, not the performance that the Columbia microphones had captured, but a studio re-recording in which Ellington and his orchestra willingly participated, produced by George Avakian. Then, in 1999 Phil Schaap produced a new recording sourced from another set of microphones that had also captured the concert in 1956. Galey comments that this, for Schaap, "moved Eggert's slider nearly all the way to the archive setting" (2021, 61). I don't think it does: the archive here has itself become a metaphor used to justify a particular sourcing of sonic files based on closeness to the his-20. See MacNeil 2005. She argues that, traditionally at least, both the editor and the archivist appeal to an absent authority (the author's intended text; the fonds-creator's original arrangement of the contents of the archive). One is at the level of text; the other is at the level of document. 21. Eggert 2019, 80. In two long notes on pages 195-6, I draw out different usages of the term in the two communities.
torical event, together with corresponding production techniques of 1999. That release offers, in editorial terms, an argument that has to stand up against the implicit argument of the 1956 release (justified by aestheticauthorial intention), since Ellington was unhappy with what the Columbia microphones had captured and wanted to improve the music. His career depended on such decisions. What we have here are, in effect, two "editions" based on competing arguments. Neither is at the far left of the slider (where the original sets of sonic files are located). Both releases are to the right on the slider since both are oriented towards the listener, as the promotional material in the CD liner notes for the 1999 release apparently makes clear. The case is a fascinating one, and Galey is perhaps right to imply that the 1956 release should be located further to the right on the slider than the 1999. And Galey is definitely right to observe that the slider is never purely under the control of the participant, in this case, the editor-producer and publicist. To be located on the slider at all is, I would argue, to have to accept the consequences of taking action, whether in one direction or the other: for this is a collaborative enterprise in which everyone has a stake. Listeners and readers and scholars have a right in the work too (in this and in every other one) and will make, if they so choose, their own judgements on its presentation (or should that be its re-presentation?, with its wandering hyphen bobbing up, yet again).