Notes on Textual Instability and Editorial Idealism

As I continue my readings in bibliography, I've moved on to "Textual Instability and Editorial Idealism" by G. Thomas Tanselle, which sounds like it should be right up my alley given the conundrums I'm dealing with in the preparation of a final text. Merry Devil has textual instability aplenty, and right now if there's one thing I'm clearly lacking, it's enough experience as an editor to have a guiding ideology of how to best proceed with this project. Lets see what I can learn here....

Tanselle begins with the observations that the study of texts as social objects that are newly edited each time they are read has become popular in the last half century, but that this does not necessarily invalidate the Greg-Bowers tradition of establishing authorial intent. The vogue for "anti-foundationalist" literary theory leads to an anti-foundationalist textual theory, but neither can be more correct than the other (1-3).

Bowers preferred dealing with the problems presented by specific situations to discussions of theory (3).

The question of "multiple authorship" can only be answered based on how one define multiple authors. In Multiple Authorship and the Myth of Solitary Genius, Jack Stillinger defines multiple authorship so broadly that a single author must be someone who never incorporates anything outside of themselves into their work, and never revises it at a later time. Otherwise, any contribution made to a text is made by another "author." Tanselle finds this proposition ridiculous (3 - 5).

Stillinger's proposition is further undercut when he makes the distinction between changes that an editor makes to a text that are then authorized by the original writer, and changes that an editor makes to a text that are not; in this case calling them corruptions. These corruptions to the original authors text are, however, still authorized by the editor who makes them, and thus are "legitimate" for that particular new work (6). So, in a nutshell, if I make a change to a text by Shakespeare, I've corrupted Shakespeare's text at the same time that I've authorized by own, but my text has at least two authors (myself and Shakespeare). If an actor asks to change a word in the text and I let them, I am authorizing that change, and thus they become another author of text B, but if I don't and they do it anyway, they create a text C with three authors anyway. You can see how quickly this lies to madness, for if we follow this logic, by choosing to not make the change my actor suggests, they author a new edition because I have had to consider an alternative and choose not to use it, so they force me to re-evaluate my original choice, so in reality we create text C (actor changes word without me authorizing the change) and C prime (me not making the change after the actor has suggested it.  This sounds like a Caucus Race.

Questioning the importance of authorial intention in textual studies is a worthwhile endeavor, but only if it can account for how the socialized, anti-foundationalist understanding of a text can fit together with the foundationalist approach. "Textual idealism" is not a synonym for "textual perfection," and constructing a text as it was conceived of by the author at a certain point in time does not undermine the notion of further editions of that text as products of social and historical change, nor does it invalidate those other texts (12).

It is further important to distinguish that bibliographers who attempt to discern authorial intention are not attempting to discover an original idea behind the work, but are attempting to reconstruct a material work that is not extant in physical form (12).

An author may have simply made a mistake in writing down their ideas, or they may have been limited by available materials or technology, but these limitations or accidents do not invalidate the text they intended to write. An intentionalist editor is therefore not trying to recover the idea of the text, they are simply intending to recover the text that was conceived of, but not necessarily written down properly by the author (12). Honestly, I'm a little shaky on this one. When it comes to typos &c, seeing through to authorial intention, that is the idea behind the text, is almost necessary. When an author calls a character by the wrong name (for example, omitting "Sir" in a speech prefix) and names another character that is clearly not present in the scene, I don't think it wrong to say that we can see the idea behind the text. Reconstructing the material textual object would mean reconstructing the error, but in this case it is more useful to reconstruct the idea behind the text and simply note the error. There's no need to muck this up with poetic notions of trying to see through to the genius or the muse behind the idea, for in this case we are seeing an idea clearly manifested in symbolic form that, for whatever reason, simply did not translate well into a material object. Maybe this is just splitting hair between the meaning of "idea" and "text as conceived."

It is important to remember that an intended text is in itself the product of a historical (and a social) process. Since no author writes in a vacuum, the texts they produce, whether intended or unintended, are distillations of a particular historical/cultural moment, and the texts intended by authors at the moment of conception should have their place in the history of the text (13-14). Thus, understanding of an author's intention means having to understand the author's particular historical/social/cultural context.

Editors attempting to create a text that reflects an author's intention by making emendations do so by attempting to fix the text as intended at a specific time by the author. This can mean incorporating later emendations into an earlier text on the premise that those emendations more accurately reflect the intended earlier text, or vice versa. Still, the editor is also fixing the edited text in a certain moment in time, and as with any historical research, will be required to fill in the gaps with their own critical judgment (17).

In the anthology Editing in Australia, Peter Shillingsburg's "The Autonomous Author, the Sociology of Texts and Polemics of Textual Criticism" advocates an approach to reading, and thus editing, based on context. A text needs to be read in light of all possible contextual factors. This position is not a new one, but Shillingsburg sets it forth very well, and this essay is therefore worth following up with (23).

James McLaverty suggests the two important questions when considering a text are what versions are, and which text editors should present to readers. He goes on to argue that any distinctive "utterance" of a text somehow connected to other utterances constitutes a new version. The text editors should present to readers encapsulates one or more of the author's intended utterances, and also includes the structural apparatus for connecting these utterances, and for connecting this particular utterance with other utterances (25 - 26).

Joseph Grigley builds on McLaverty's work with the observation that individual utterances cannot be repeated because each utterance is its own event. Even if the precise words of an utterance can be repeated, its context cannot. Grigley's conception of the timely nature of utterances may not offer anything new to editors, who have known for generations that they can never recapture the complete context of a work, but it may help clarify what editors have done and will continue to do going forward: capture the interpretation of a textual moment (26).

The binary between "modernist editors" who seek to create a stable and definitive text and "post-modernist editors" who highlight the instability of texts and the value of interpretation is, as most binaries tend to be, too neat to be practical (30).

Morris Eave's entry in Cultural Artifacts and the Production of Meaning: The Page, the Image, and the Body offers a view of editing that delights in the instability of texts. He argues that all editorial decisions are legitimate, and links the decisions of the editor to the same decisions the artist makes as part of "the process of history." Intentionalist editing is one more form of interpretation and creation, and should not be dismissed (32 - 33). A-men I say to that. The fact is that it's very difficult to present Renaissance or earlier, or even early 20th century or earlier plays in their entirety to a modern audience. The theatrical conventions of the time will dictate an audiences expectation and basic level of tolerance for how long they're willing to sit still. Like it or not, in the early 21st century, that means the 10 or 15 minute span of a YouTube video.

Tanselle's got my back on this last point. I'll quote him here: "Thus when Eaves says that 'whatever is may be . . . on its way to becoming right,' he recognizes that change may be necessary to produce rightness for a given audience" (33).

Ann R. Meyers, in "Shakespeare's Art and the Texts of King Lear" notes that even attempts at documentary editing still require a degree of selection and interpretation. In presenting the two versions of King Lear in their Oxford Edition, Wells and Taylor have made emendations to the quarto Lear where Qa has "crulentious," Qb "tempestuous," and F "contentious," Wells and Taylor have used the Folio reading in their quarto version of the text based on paleographic evidence that demonstrates the Folio provides the correct reading to the compositor error of the Quartos. Despite similar paleographic evidence for reading for the correctness of Folio's "Come unbutton" (as opposed to Qa's "Come on be true" or Qb's "Come on") Wells and Taylor have left the reading of "come on be true" because it makes to them "local and contextual sense" (34 - 35). Wells and Taylor are the ones that determine the sense of their inclusion or rejection of certain readings, and thus they still are creating a text that is a conflation; an edition of their own interpretation.

In "Text as Matter, Concept, and Action," Peter L. Shillingsburg offers an ambitious argument for the existence of non-material texts, in which he demonstrates how non-material texts are linked to their material counterparts. He conceives of works as entities that are manifested in material and linguistic form, but that the conceptual version of a text should not be confused with the Platonic ideal of "text" (37). This is definitely one I'll have to read. Where does one draw the line between ideas, texts as conceived but not written, and the text that appears on the page. I can only imagine some future textual scholar trying to make sense of the stew of tracked changes that Google Docs keeps of "Battleground State." Precisely how much weight should we give to the text as conceived when the author might decide that the text as conceived sucks, and needs extensive re-conception before it's fit to be put in front of people?

Texts can only be understood through observation, but every observer has the right to choose how they wish to observe a text (or anything else for that matter). Any theory of textuality must encompass every process by which a text may be observed, which includes a range of methods and tastes as diverse as the number of observers (39-40). Observing a text through an historical scope requires a different apparatus than observing that text with a final performance in mind, and observing a text as a work of art fit to be hung on a wall requires a still different apparatus. Before presenting an edition of Merry Devil, I must determine who that edition is for.

Paul Eggert, in "Document and Text: The 'Life' of the Literary Work and the Capacities of Editing," also argues that all editions of a work should be considered and studied as authentic iterations of the work and should be studied as authentic (41).

Tanselle says this of Grigely's "Textual Criticism and the Arts: The Problem of Textual Space," but I think it bears quoting, as it has general application, and should, I think serve as a framework for understanding any essay arguing for any theory of a text: "Practically everyone understands that the experience of a text, like all other experiences, is colored by one's entire previous life and that it will continue to reverberate in one's mind, affecting all experiences afterward. An essay that says something like this does not mark any advance in thinking" (45).

If modern approaches to editing emphasize the plurality of texts rather than the intentions of a single author, than it is curious that modern textual scholars have a low regard for the work of the intentionalist editors of previous generations. The texts they produced are, after all, a further contribution to the plural text that modern textual scholars celebrate (49 - 50).

New theoretical insights into the nature of editing have the danger of creating a sense of the futility of the editing process. If all versions of a text ought to be considered in their own right, it is impossible to present a modern reader with an accurate distillation. It is likewise impossible to create an edition that encompasses the needs of all possible readers of a text. Through this, it is important to remember that editing a text is a creative activity, akin to writing an essay or giving a performance (50 - 51). So it is more than helpful, it is essential to determine the ultimate goal of a text before it is produced. Is the goal for the text to be read in a classroom or to be performed on a stage. What kind of classroom? What kind of stage? Determining the ideal audience for a work will provide the context in which the work is to be created. The way you advertise your show for a Philly Fringe audience will of necessity be different from the way you advertise your show for a largely academic audience in rural Virginia.

While some more modern philosophies of editing have heaped criticism on previous generations of editors for preferring a "single text" edition of a work, the form of the codex lends itself to this style of presentation. Critical editors of the past might very well have wanted to present multiple text editions of their work, but this option was unavailable given the physical and economic limitations they faced (52-53).

Here's something worth quoting: "The acts of constructing texts and works are social events, as many textual theorists have been telling us; but we are not going as far as we can toward understanding those events if we limit ourselves to surviving objects and exclude from our deliberations the mental events that are a fundamental part of the textual process" (58).

Quoting Virginia Wolfe: "I believe that the main thing in beginning a novel is to feel, not that you can write it, but that it exists on the far side of a gulf, which words can't cross: that its to be pulled through only in a breathless anguish. . . . But a novel, as I say, to be good should seem, before one writes it, something unwriteable but only visible; so that for nine months one lives in despair, and only when
one has forgotten what one meant, does the book seem tolerable. I assure you, all my novels were first rate before they were written" (58-59).

Summary: Texts that are conceived are inherently different from texts that are written, and if we are truly to explore textual objects as cultural artifacts, this requires seeing beyond the words printed or written on them. The form of the printed book has as much as any guiding philosophy led to the development of single text editions, but single text editions will always be a requirement of publishing. Someone with the proper knowledge and training will always be necessary to distill information that a less well informed reader does not have, or does not have access to. In the complete social history of a text, a single text edition of that text is another version of the text.


Wow. This was a thick survey of the development of theories of editorial instability and editorial theory. It's left me with a lot to think about, and some other things to explore.

Interesting Textual Variants

Stern has a lot of good ideas about how "Shakespearean" texts were made, how they got on the stage, and how they were printed, and I have this funny feeling like her work will be extremely handy in bringing The Merry Devil of Edmonton to life, but as part of this project involves the creation of my own new critical-edition of Merry Devil, I am struck by two things: first of all is the overwhelming regularity of the three quartos that I have so far collated; few variations exist between them.

Where these quartos differ most, however, falls in a seemingly trivial moment early in the play, when Fabell and Raymond accuse Frank Jerningham of taking part in the plot for him to steal Milliscent away from Raymond. Frank describes his affections for another lady:

                              ...but thou know'st
That Essex hath the saint that I adore,
Where ere did we meet thee and wanton springs,
That like a wag thou hast not laughed at me,
And with regardless Jesting mocked my love?

That's the text from Q1. Q2 reads thusly:

                               ...but thou know'st
That Essex hath the saint that I adore,
Where ere did'st meet me, but we two were jovial,
But like a wag thou hast not laughed at me,
And with regardless Jesting mocked my love?

Note the changes in bold. Also note that Q2's line is less poetic. One must be cautious about such proclamations in a play like this, where characters of all classes regularly change between verse and prose throughout, but Q2's line change creates a metrically irregular line in an otherwise regular passage. One might also make the argument that "wanton springs" is more idyllic than "we two were jovial."

The delightfully apoplectic W. W. Greg summarizes his analysis of the variant texts thusly:

I think that if a rough author's draft (after having served for the production of a prompt-book) had been severely cut with a view to preparing an abridged version of the play, and had then come into the printer's hands, the resulting edition might have been something like what we find in the extant text (130).

So why the variation? Stern points out that country audiences were generally less sophisticated than their London counterparts, and thus the plays were cut (145-146). Her further suggestion that prompters themselves would be a common source of emendation to might lead one to think that the Q2 printing followed from a prompt copy further edited from the original (144). Of the four extant copies of Q1, none are identical, which is not necessarily odd given the practice of stop-press correction that was common in London printing houses (Greg, 125). Greg, however, traces the lineage of Q2 to a cousin of what he calls Q1B with one exception, and this based on a line that occurs just a little bit later in the passage I have examined above (127). It may serve us to examine that variation. Q1 is as follows:

Before I would give o'er the chase, and wrong the love,

Q2:

Before I would wrong the chase and leave the love,

Q3:

Before I'd wrong the chase and o'ergive love,

 I'm apparently working from Q1B, because Greg gives the reading of Q1A as

Before I would unage the chase and overgive love (127)

The plot thickens, eh? It's difficult to say why such variation would arise. Pretending that "unage" is a misreading for "wrong," there seems to be an almost casual swapping of verbs in the line. If Greg is correct and the Q1B reading represents the text in an intermediary state of printing, that would simply confirm the statistically likely possibility that Q2 was not printed from one of the four extant Q1s (127). Nothing Earth shattering there, but why the other edits?

Lets play the game "everybody is right." If I'm right that there are emendations to the text that represent changes made out on the road between Q1 and Q2, that means that Arthur Johnson needs to have got his hand on those changes. If he wanted to do a simple reprint, he could have either printed from an existing Q1, or from the papers that he had from that printing. The advantage of the latter could derive from dissatisfaction with the first printing; he did change printers from Henry Ballard (Q1) to Thomas Creed for Q2, after all, and his further change to G. Eld for Q3 seems to indicate that he may have had a more habitual dissatisfaction with printers.

Yet it is also worth considering that the book that the King's Men supplied Johnson for printing was the touring prompter's book. No copy of the text would be more dispensable than an edited for touring text when the London theatres were open, after all. If Johnson printed Q2 from this book, which would likely have been edited over years of use, that may account for some of the variation in this passage. It would, of course, be just as likely that the line had been edited back for subsequent printings (if Johnson continued to use the touring book), or perhaps, unable to obtain it, he printed from one of the variants of Q1.

Maybe I even get a little bit of support for the ever-changing touring prompt book theory from the evolution of another line within this same passage:

Q1: I have taught the watchful Nightengale to wake,
Q2: And I have taught the Nightengale to wake,
Q3: I have taught the Nightengale to wake,

Q3 bears markers of both Q1 and Q2 in this case.  Again, Q1 seems the more poetic, and depending on the pronunciation of "have" may read as a regular verse line. Q2 omits the description of the nightingale, but is metrically regular without elision. Q3 is what a music director friend of mine would describe as a hot mess. It just doesn't work as verse. Through these three printing, we see the evolution of a line of verse into a line of prose. 

Anything I've said here needs to be read in light of the disclaimer that this is all mere conjecture. Yet given the evidence that Stern has put forth in Making Shakespeare, it is worth examining the book printing industry and the playing companies as having more of a symbiotic relationship than some would suggest. If Johnson regularly obtained copy for subsequent printings of The Merry Devil from the King's Men, the printed history of The Merry Devil may reflect an increasing simplification of dramatic writing.

Works Cited
Greg, W. W. "The Merry Devil of Edmonton." The Library: Transactions of the Bibliographical Society. 1944. London. Oxford UP. vol. s4-XXV 3-4 p. 122-139.

Stern, Tiffany. Making Shakespeare: From Stage to Page. Routledge. London. 2004.