 | |  |
Notes
This paper retains the form of public address of its abbreviated version read before the Society for Textual Scholarship in New York City on April 10, 1987.
Critical edition is a more comprehensive term than critical text since to the edited text it adds such important accompaniments as an analytical textual introduction and various series of apparatus. For the moment I am not concerned with these appurtenances to the text and so for convenience may use the term edition and text interchangeably when in fact I am concerned strictly with the text. However, the presence in a scholarly edition of an apparatus that records editorial alterations in depth is an essential factor behind my views of the manner in which texts may be emended.
In modern texts the effect on meaning of the different forms of accidentals varies considerably. It is true that spellings are an essential part of the philological historicity of a text and not just a means for imparting an agreeable flavor of the past. But compared with punctuation, capitalization, emphasis italic, and paragraphing, for example, spelling as such has little effect on the direct transmission of meaning. Punctuation as it illuminates the contextual modification of word, phrase, or clause, may sometimes be as vital to the meaning as the substantives, the words themselves; and insofar as capitalization may distinguish concepts it can be of more than casual significance. Structural paragraphing marks meaningful progression within discourse. Normally, punctuation when not setting off syntactical boundaries in the interest of clarity enables an author to direct the flow of his style as if he were reading aloud, with the pauses and emphases that he felt when writing and that he wants to transmit to the reader. It is not by chance that Nathaniel Hawthorne, for example, utilizes a heavy parenthetical punctuation whereas Stephen Crane, with an entirely different purpose, omits some of what are usually considered to be normal punctuation divisions.
Once an appropriate copy-text has been selected, many single-text modern works
Very strictly speaking, any compositorial departure from copy is an error, a statement that would include all housestyling. But when the setting copy is not preserved, the presence of such "error" can only be inferred, sometimes with near certainty as with the appearance of an alien for an idiosyncratic authorial characteristic, often with mere suspicion, but more often with complete uncertainty whenever neutral details are involved. For our purposes, therefore, it will be convenient to enlarge error from its technical to its popular sense, that is, of something wrong, not just something altered by another hand from the original in detail that has little or no exact application to the sense.
For example, in recent editions of Elizabethan dramatists an attempt has been made in old-spelling editions to reach out to a more general audience by removing unessential barriers to a general reader's, or early-student's, approach. One now finds most editors substituting the short s for the old long s, modernizing from an only slightly later period the old u-v and i-j conventions, normalizing speech-prefixes, making stage-directions more consistent, and slightly adjusting the more ambiguous and potentially confusing examples of freewheeling early punctuation. Even undergraduates, I am told by an experienced teacher, seem to respond to this approach when they are properly introduced to it. The means for dealing in this manner are treated in my "Readability and Regularization in Old-Spelling Texts of Shakespeare," Huntington Library Quarterly, 50 (1987), 199-227.
The problem of regularization and normalization in manuscripts is usually more acute than with prints, which willy-nilly have already gone through some resolution of inconsistency at the hands of compositors or copyreaders, a process only imperfectly to be isolated by an editor when a work is preserved in a single authoritative document. This is no place to expatiate on the differences between the two media in respect to the editorial problems they individually present. In fact, the principles are identical for both in the matter of regularization, and the difference consists mainly in the larger amount of attention ordinarily required for the presentation of a critical reading edition of a manuscript as copy-text in comparison with that needed for a print unless the manuscript is being treated in a diplomatic, or documentary, manner where inconsistency must be preserved as a virtue.
I do not know when copyreading became standard in commercial book publishing. Certainly the preserved manuscripts of books by Nathaniel Hawthorne, Stephen Crane, or William James (my chief acquaintances) show no signs of copyreading. On the other hand, the results suggest that attention was paid to styling by the editors of various magazines that published these authors, but how far this pre-print styling extended to the accidentals is uncertain. On the whole one is probably safe in earlier printing in assessing variation as mainly compositorial. Nevertheless, even when copyreading became the rule and compositors were expected strictly to follow copy, variation could result from oversight in preparing the manuscript. A copyreader may not have been letter-perfect and might miss variation or could inadvertently create variation himself by memorial lapses. Strange things could and did happen which the average author was powerless to prevent or repair, given the present custom in commercial publishing of not resubmitting copyread manuscript to the author for approval before typesetting. When my wife, Nancy Hale, came to read galleys for her biography of Mary Cassatt, published in 1975 by Doubleday, she discovered that about halfway through the typescript a second copyreader with quite different styling had replaced the original reader so that two divergent and inconsistent systems separated the halves of the book's typesetting.
In the event that insufficient holograph material is preserved, it is more likely that
One may cite William James's excentric, canvass (as a noun), or hindoo (for Hindu). In The Marble Faun Hawthorne out-Englished the English in preparing his manuscript specifically for an English publisher, especially in the extension of -our endings beyond custom. The English compositors were sometimes irregular in restyling this copy to normal British usage. Moreover, the copy itself was inconsistent when Hawthorne inadvertently forgot to write according to his self-imposed system. Like the early James neccessary, true misspellings are not to be preserved, of course, no matter how characteristic provided that they are indeed genuine misspellings. But an editor is faced with Hobson's choice about Hawthorne's regular cieling since this is not a misspelling proper but an old-fashioned holdover from an occasional eighteenth-century spelling apparently affected by an earlier seeling spelling. The Centenary editor of Hawthorne was no doubt mistaken to flinch at reproducing cieling from the copy-text manuscripts and to substitute the compositorial normalization ceiling from the prints.
All references are to the separately titled volumes and the page-line numbers of The Works of William James (Harvard University Press, 1975-88). Hawthorne is cited from the Centenary edition (Ohio State University Press, 1968-74) and Stephen Crane from the multi-volumed Works (University Press of Virginia, 1969-75).
It must be regarded as a pedantic distinction mistakenly based on a too scrupulous definition of unit when the textual editor of the James edition retained preëxistent in his reprint of an added preface prefixed to the second edition, presumably on the ground that the printer's copy differed between second-edition preface and that behind the first-edition text.
The following manuscript compounds differ from manuscript in the print: school-room, class-room, turning-down, belly-band, straw-hats, gold-background, earthborn, offhand.
In fact, this statement is truer in theory than in practice since in this part of Lecture III the only non-Jamesian form for which we have evidence in VII-VIII is the book's repetition twice of any one. However, the evidence of Lecture VII is still useful in regularizing these two occurrences to anyone even though for this word the printed lecture of 1898 and a typescript made at that time from James's rough manuscript also read anyone: each is an independent witness to the lost manuscript but one cannot guarantee that each had accurately copied the form of this word (although apparently they did).
The following manuscript compounds differ in Chapters XIX and XX of the book: burnt-offerings, marlin-spike, beggarlike, re-awaken, and re-assert.
In the process of regularization, as here, it may seem legitimate to extend specific evidence to that of analogues. However, an editor must be prepared with evidence that such analogues are legitimate ones. For instance, although James spelled connexion with a x, his spelling for reflection was conventional with the ct. James invariably spelled colour but for similar words he always wrote -or.
It must be admitted that in a posthumous volume of editorially collected essays of different dates from manuscript (possibly mixed with printed copy) an attempt to cross-regularize might well be pushed over into the field of normalization. In such circumstances it might sometimes be better for a regularizing editor to confine his emendations to the evidence of precedents within any individual essay, thus treating each essay as a separate unit. In such a collection different accidentals systems ought to be accepted with equanimity by any reader. It must also be admitted that the examples of Pragmatism and the Varieties are not necessarily exact parallels to a hypothetical authorially collected volume of essays. In both books the sections of lectures were written in sequence and in what were essentially similar manuscripts; hence it was possible to extrapolate the evidence of the preserved part of such an overall manuscript back to the lectures set from lost manuscripts. (The only exceptions to this neat pattern are the three Pragmatism lectures set in the book from annotated journal copy, but in turn these journals were set from the same set of manuscripts and so the end is the same.) However, an author's own collection of unpublished units set from manuscript (perhaps mixed with previously printed units) would very likely derive from less consistent copy which if written over a substantial period of time might
Some linguistic specialist may find it interesting to inquire into the details of Hawthorne's attempt as well as the details of the printer's rejection of various of his putatively anglicized spellings such as liquour. There is also the problem of unique or consistent spellings that would need to be normalized without the documentary evidence within the book's unit available for regularization. Analogy would need to be called on heavily.
An exception was made for neighbor, which was so far in the majority—with only a scattering of neighbour spellings—that it was retained. Whether this regularization on the basis of numbers was wise may be put in doubt: even the few neighbour spellings attest to Hawthorne's general intention in a case where otherwise habit proved too strong for him. It is unlikely that he felt that neighbor should be excluded from the -our spellings that he was trying to adopt.
The details of this investigation are elaborated in the Centenary Textual Introduction, pp. 319-329.
I notice at 128.14-15 that in the Faun manuscript a comma comes before the first parenthetical dash but none appears before the second. It is a commonplace in William James's manuscripts to see him interline parenthetical dashes where he had written commas in the text. But almost invariably he did not delete the commas at the time he was interlining the dashes. Whether this was simple negligence or whether he intended the comma-dash combination in these circumstances (although simple dashes are common in the same manuscript) is moot. The odds would seem to favor negligence.
In 1962 the textual editor of The Scarlet Letter grasped the significant point that regularization should be confined to evidence within the unit (p. xxxv), and he was prepared to alter to Hawthorne's preferences three spellings that were variant in the print. But he was not prepared to regularize punctuation except in error and felt that compositorial pointing otherwise had to be preserved faut de mieux (p. lxiv). At that time, however, the differentiation had not formed in his mind between certain formal aspects of punctuation that can be regularized to what would almost certainly have been in the manuscript as distinct from ordinary syntactical pointing so subject to authorial mood or indifference that editorial regularizing could seldom guarantee a plausible reconstruction of the manuscript in any given case. Since the evidence of the variable comma-dash in the Letter may be assigned as a general compositorial styling and Hawthorne's habitual practice without commas was known, The Scarlet Letter text might safely have been regularized in this formal feature by removing the commas. In the matter of spelling, the housestyling was consistent for certain words in -or (Hawthorne's spelling) but also consistent for others in -our (compositors' spelling). Since there was no deviation in any of these words, the lack of specific inconsistency did not seem to allow regularization: the alteration of the consistent -our forms to -or seemed to be normalization and thus highly debatable. Consequently, in 1962 the editor felt there was no choice but to follow copy, even though this meant reproducing what was demonstrably a compositorial and not an authorial divergence. I am inclined to agree that for better or worse the rules work so logically that usually they ought to be kept, within flexible limits. But the results can be illogical, as in The Scarlet Letter. Thus the temptation is strong to apply normalization to these -our spellings since we can be certain the compositors in them departed from the manuscript setting copy. Unless it is weaseling, in the end we may come down to a definition of what is meant by variation within a unit. The Centenary editor applied the definition narrowly to each individual word. But if one were to argue that in this case the general classification of compositorial -our versus authorial -or should be the basis and not specific words, then it might just possibly—and with some stretching—be considered as regularization (and not normalization) to alter the various -our words to -or on a more broadly viewed concept of what constitutes variation within a unit. Since this distinction of the two kinds of variation within regularization is highly debatable, an editor need not be faulted if he chooses one or the other: there is a logical principle behind each. I sympathize with the broader view— even if it is close to normalization—but must confess I have not experimentally worked out all of its ramifications if generally applied to other texts. Certainly the kinds of differences
It is no doubt a distinction without a difference to note that regularization in printed texts endeavors, on some documentary evidence of compositorial variance, to restore the hypothetical authorial reading in the lost manuscript from compositorial interference, but that regularization within a holograph manuscript—no matter how necessary—cannot restore demonstrable authority. Strictly speaking, the vagrant comma in The Marble Faun mentioned in footnote 20 above is authoritative since it comes in holograph and is the only evidence we have. Thus an emending editor cannot assert restoration of an authorial reading here in the Faun but only restoration of a firm authorial characteristic from the author's own thoughtless lapse—in this case an actual error on his part. Since the comma-dash combination is practically unknown in Hawthorne's manuscripts one might speculare that in writing out the Faun fair copy he made an incomplete alteration of what could have been parenthetical commas in the draft.
In Chapter XXI the youth and Wilson have one speech together jumbled without differentiation of speaker (120.37-39), all of this in dialect. But which part is the youth's and which Wilson's (if Crane himself ever differentiated them) is moot. The youth's last line in Chapter XXIV (132.17) is probably to be taken as unrevised dialect.
It is difficult to understand the basis for recent statements that it was Crane's editor Hitchcock and not Crane himself who deleted the endings of certain chapters scored through in the manuscript and absent from the book. (1) Most tellingly, certain of these deletions were made in the same blue pencil in which Crane in his own hand wrote some further alterations. (2) Plenty of textual evidence of coincidences in typist's errors as well as in indifferent readings between newspaper and the book against the manuscript demonstrate the same basic copy—a typescript and its carbon—was used to set both newspaper and book. (3) Hitchcock accepted the Red Badge after reading the clippings of the previously printed newspaper version that Crane brought him. (4) The newspapers—which admittedly contain a vastly cut version—do not print any of the material deleted in the manuscript. (5) All facts indicate that the deletions in question were made before the final typescript was ordered which in turn was in back of the newspaper version printed before Hitchcock came into the picture. (6) No scrap of evidence is preserved that Hitchcock ever saw the manuscript. (7) The deleted chapter endings thus were initiated solely by Crane and represent his last intention before he ordered the pre-Hitchcock final typescript and carbon. (8) Every piece of evidence indicates that the manuscript was not touched in any way after the typescript and its carbon were made from it except to abstract a few pages of a battle scene for separate magazine publication. A recent edition of the Red Badge trumpeting the reprinting of these repetitious chapter endings as the restoration of Crane's true intention from Hitchcock's requested deletion is just plain wrong. The evidence of the manuscript clearly shows that it was Crane who on his own initiative deleted them, and wisely so.
Comparison against the originals in quotations in James's manuscripts used as setting copy, or journal articles used as copy for books, and then against the final printed accidentals, shows that compositors had no especial respect for reproducing copy in quotations as against their treatment of regular text but that they imposed housestyling on quotations as freely as on the quoting author.
| | |