Example 9

Hawthorne, The marble faun, 1860

Nowhere is the debate on editorial principles livelier, nowhere has disagreement between textual bibliographers been more pungently expressed, than were it concerns the editions sponsored by the Center for Editions of American Authors (the CEAA). The CEAA  is a Committee of the Modern Language Association of America set up in 1963 to encourage the responsible editing of American authors, dispensing grants from the National Endowment for the Humanities, and approving the results with a ‘seal’. The Center has been concerned with editions of the novels of Crane, Hawthorne, Howells, Melville, Simms, and Mark Twain, each one a vast and costly enterprise supported by a particular university and university press; and, although the Center is likely to become less effective as a result of the withdrawal of its funds in circumstances of economic recession, the influence of the editions it has sponsored—now numbering more than a hundred— is far-reaching.¹

The central editorial principle of the CEAA is that an eclectic ‘clear text’ should be established according to Greg’s theory of copy-text, adapted to nineteenth-century textual situations, and should be supported by an apparatus of detailed evidence.² Another main principle of the Center is that the investigation of the textual evidence should be so thoroughly carried out and so fully recorded that the apparatus will be ‘definitive’, by which it is meant that it will not be necessary for the investigation to be repeated in the foreseeable future.³

The textual bibliographer at the centre of this debate has been Fredson Bowers, the scholar was more than any other has been responsible for establishing the editorial principles adopted by the CEAA, and who is himself the most distinguished of the CEAA editors. Amongst  other works Bowers edited the Ohio ‘Centenary’ edition of Hawthorne’s The marble faun (1968) which he believed to be a classic case of the use of demonstrable evidence in a scientific manner for the solution of difficult editorial problems.⁴ It is indeed a central CEAA text, and it will serve as a demonstration not so much (as with the other examples) of the various ways in which a work of literature might be edited, as of the actual strengths and weaknesses of a typical CEAA edition.

The marble faun, Hawthorne’s last major work of fiction, was first drafted from July to December 1858 during a family tour of Italy, and was rewritten in a revised version from July to November 1859, after the Hawthornes had returned to England.⁵ The holograph manuscript of the revised version went to the printer to be set in type, Hawthorne corrected the proofs, and the first edition was published in London in three volumes at the end of February 1860 (1860a).⁶ Sheets of this printing, meanwhile, had been sent to America, and the first American edition, set from 1860a with a few deliberate alterations, was published in two volumes in Boston about a week after the English edition (1860b). In England the novel was called Transformation; the American title (which Hawthorne himself preferred) was The marble faun.

The reception of the book was mixed, and almost immediately after  publication Hawthorne wrote an afterword to explain its ending; this was added as a ‘Postscript’ to the second printing of 1860a and—with a slightly altered text—as a ‘Conclusion’ to the fourth printing of 1860b. There was a Tauchnitz edition in 1860 (set from sheets of the second printing of 1860a), and second and third English editions in 1860 and 1861. In America, although there were several further printings  of 1860b from plates in 1860, 1864, and 1865, the book was not reset until 1876, after Hawthorne’s death.

The first draft of The marble faun has not survived, but the revised draft made by Hawthorne in 1859, which was used as copy by the printer, was given by Hawthorne to his English friend H.A. Bright, and was presented to the British Museum by Bright’s daughter in 1936.⁷ The proofs are lost, so the only other early documentary evidence for the text are the various printed editions of 1860 and 1861. Of these 1860a, the first edition set from the surviving manuscript, is much the most important; all the other printed texts derived from it, although there were some deliberate alterations in 1860b, the first American edition.

Hawthorne’s holograph fair copy of The marble faun, which is thus of  crucial importance for the establishment of the text, was legibly written on half sheets of paper folded once to make post quarto bifolia. The compositors, whose names were written on the manuscript at the beginnings of  their stints, had no great difficulty in transcribing it. There are a few verbal differences between the manuscript and 1860a which appear to be copying errors, and a larger number which appear to be the result of deliberate revision in proof. The compositors normalized the text freely, altering a good  deal of the punctuation and making changes of capitalization and spelling.

Here as an example is fo. 125 of the manuscript, followed by the passage in 1860a that was set from it:

There are two verbal variants between the manuscript and printed versions of the extract: ‘could’ in the manuscript becomes ‘would’ in 1860a (a6/b7) and ‘rejoined’ becomes ‘replied’ (a31/b36). The fifteen non-verbal variants between the manuscript and 1860a, which are typical of the book as a whole, comprise the deletion of seven commas and a dash, the addition of three commas and a hyphen, the substitution of two punctuation marks for others, and the alteration on one capital letter to lower case.

There were no changes in the extract between 1860a and 1860b.

Let us look, then, at the Centenary edition of The marble faun. At the heart of the edition is a plain text—it is the laudable aim of the CEAA to lease its edited texts inexpensively to reprint publishers⁸—which is based, in accordance with the principles of the series, on the manuscript, not on the first printed edition. The text is preceded by a short historical ‘Introduction’ by Claude M. Simpson, which describes the genesis and composition of the work, and a long ‘Textual introduction’ by Fredson Bowers, which explains and argues the case for the editorial procedures used. The lengthy textual apparatus which follows the plain text is in seven main parts: (1) 35 textual notes on particular readings; (2) a list of 688 editorial emendations to the copy-text; (3) a list of 190 verbal variants from the manuscript which appeared in the first printed edition but which were not incorporated in the edited version (on the other hand there is no corresponding list of non-verbal variants from the manuscript found in 1860a but not incorporated in the edited version); (4) lists of line-end hyphenation in the edited version and in the manuscript; (5) a historical collation, recording the verbal variation between the early printed editions, and also all variation between 1860a  and 1860b; (6) a list of 1,447 alterations made to the manuscript (mostly by Hawthorne) while it was being written out or soon afterwards; and (7) an analysis of the 145 compositorial stints into which the manuscript was divided for setting, which also serves as a concordance between the pages  of the manuscript, the first printed edition, and the edited version. At the end of the apparatus the ‘Editorial principles’ of the series are reprinted from its earlier volumes; and a list of corrections made to the introduction and apparatus is added to the second printing (1971).

Assessment of the Centenary Marble faun must rest on the answers to three main questions. These are, first, has the textual evidence been thoroughly investigated, and have the results been adequately and accurately recorded? Is the edition in fact ‘definitive’ in the sense that the investigation need not be repeated in the foreseeable future? Secondly, is the edited version well founded: are the principles upon which it is based generally sound, and in this particular case efficacious? And thirdly, does the editor offer a convincing treatment of the text? Are his 688 emendations of the copy-text acceptable?

Collation of the holograph manuscript of The marble faun with the Centenary text and its apparatus, and with the early printed editions, shows a very high standard of accuracy in the edited version and in the records of variants, etc. This is of course what we should hope to be the case, and indeed what we should expect, since the early texts were each checked three times, comprehensively and expensively, by the editorial team. Yet mistakes can be made in collation by the hardest-working editors, and it is no small matter that this great quantity of detail has been noted and used with so little error.⁹

The records of variants, etc., in the textual apparatus are therefore reliable (if somewhat awkwardly divided) as far as they go; but they do not include a record of the non-verbal variation between the manuscript copy-text and the first printed edition, for some of which the author could have been responsible. The omission of this record means that the edition is not ‘definitive’ in the sense intended by the CEAA, for without it a future investigator of the text would have to do this part of the work again.¹⁰  In any case this omission suppresses the evidence of the normalized punctuation, etc., of the first edition—accepted and perhaps refined by Hawthorne but not incorporated in the edited version—which is required for an assessment of the Centenary text. No doubt the inclusion of such a long list in the apparatus would have been difficult and expensive; but a solution to the problem might be to place it, together with the rest of the apparatus, in the form of microfiches in a pocket at the back of the volume.¹¹

Next we consider the CEAA’s central editorial principle, the establishment of an eclectic text according to Greg’s theory of copy-text. Greg offers a procedure for choosing copy-texts for editions of Renaissance works which, he insists, is to be used with flexibility and critical discretion. (Taking it for granted that the author’s manuscript has not survived and that the author himself was not much concerned with the printing and reprinting of his book, Greg recommends that the earliest in an ancestral series of printed texts shall normally be chosen as copy-text because it comes nearest  to the author’s ‘original’.¹²)

The application of Greg’s theory by the CEAA to the editing of nineteenth- and twentieth-century works is both more comprehensive and less flexible than Greg’s pragmatic approach, and it denies the editor much critical discretion in the choice of copy-text. The CEAA’s Statement of editorial principles says that the author’s manuscript should be chosen as copy-text—or, if the manuscript is lost, then the version nearest to it—because the manuscript ‘almost always’ represents the author’s intentions more closely than any later version.¹³ Greg’s theory was based on the reasonable assumption that the earlier printed versions  of Renaissance texts were likely to be more authoritative in detail than the later ones; but the assumption underlying the CEAA’s editorial principle is the very dubious one that, in the absence of evidence to the contrary, a nineteenth- or twentieth-century author ‘intended’ the manuscript rather than the printed version of his work. In fact, of course, such an author may have preferred the printed version, but without actually saying so.

In practice, as we have seen, the authors of nineteenth-century novels did not as a rule specifically approve or disapprove the normalization of the text that accompanied the transformation of their works from manuscript to print; we know that they accepted it, and usually that is all we know. The editor cannot normally say that this version or that comes closest to the author’s intentions. All he can say is that this version is to be preferred to that one other—perhaps critical—grounds.

Sometimes (as other examples have shown) there are critical reasons for choosing an early printed version of a work as copy-text for an edition even though the author’s final manuscript is available; but sometimes it is better to use the manuscript, and The marble faun is such a case. The details of Hawthorne’s printer’s-copy manuscript, and especially its excellent punctuation, are complete and well considered, and they convey the rhythm of the dialogue and the balance of Hawthorne’s precise, parenthetical prose in a way that the mechanical details of the normalized printed texts do not consider, for instance, how Hawthorne’s style is blunted in the 1860a version of the extract by the addition of three and the deletion of seven commas. So, in spite of the fact that Hawthorne accepted, and perhaps refined, the normalized details of the printed text, there is a good critical case for basing an edition on the text of the final manuscript, which is that it offers the most convincing punctuation of the text. There is no need to introduce the assumption, which cannot be supported by evidence, that Hawthorne intended or would have preferred his own punctuation to that of the first edition, although the editorial principles of the series indicate that it was partly because of this assumption that the manuscript was chosen as copy-text for the Centenary edition of The marble faun.

The Centenary editor reproduces his manuscript copy-text with admirable accuracy, and with similar accuracy he reports his 688 editorial emendations. The remaining questions are whether these emendations are both necessary and right, and whether they are adequately supported by annotation.

About half of the 688 emendations of the copy-text incorporated in the Centenary text had first appeared in 1860a. Most of them corrected obvious errors in the manuscript, while a few appear to have resulted from proof-revision by the author. It is plainly right that these 1860a. readings should be included in the edited version (although there is a lack of supporting annotation—which obliges the reader to search for their origins in the lists of the apparatus).

Then there is a group of verbal variants between the manuscript and 1860a which might have resulted either from Hawthorne’s corrections in proof or from mistakes made by the compositors. The editor has approached them by counting the transcription errors made by each compositor in a control section of the text. (This was a part of the book which was sent to the American printers as uncorrected proofs of the English edition, and which was then proof-corrected by Hawthorne before the English edition was published.¹⁴) It was supposed that, if a doubtful variant appearing in another part of the text had been set by a compositor who made many mistakes in the control section, it was more likely to be an error (not an authoritative correction) than if it had been set by one who made few. While this supposition is likely to be true on average, it cannot properly be used to decide particular cases. The danger of doing so is illustrated in our extract, of which the 1860a version was set by a compositor called Mintern who happened to be the man with the highest rate of error (in the control section) of all the compositors who worked on The marble faun. In 1860a two words differed from the manuscript (a6/b7, a31/b36). One of them was written ambiguously in the manuscript, and it is likely enough that Mintern misread ‘could’ (p.186, a6) as ‘would’ and that Hawthorne did not notice the change; the editor reasonably prints ‘could’. But ‘rejoined’ was written reasonably clearly in the manuscript (p.186, a31), and ‘replied’ is more likely to have been a proof-revision by Hawthorne than a mistake by Mintern. Nevertheless the editor prints ‘rejoined’, arguing:

Ordinarily a change like this from MS ‘rejoined’ to E1^a [= 1860a] ‘replied’ would be imputed to the author. But the word is not very legible and could readily be confused by such a careless compositor as Mintern.¹⁵

Thus a general suspicion of Mintern’s competence has led the editor to reject, on the doubtful ground that the manuscript reading is illegible, a particular emendation that he would otherwise have accepted.

Finally there is the considerable group of emendations in the Centenary text that were introduced by the editor. Apart from the officious correction of the prepositions in three Italian proper names, there is only one verbal emendation, and that one is wrong in a way that indicates a critical insensitivity to the meaning of Hawthorne’s text. Hawthorne wrote:

And yet Donatello’s heart was so fresh a fountain, that, had Miriam been more world-worn than she was, she might have found it exquisite to slake her thirst with the feelings that welled up and brimmed over from it. She was far, very far, from the dusty mediæval epoch, when some women have a taste for such refreshment.¹⁶

By ‘the dusty mediæval epoch’, Hawthorne referred of course to Miriam’s middle age, from which she was as yet many years distant. In the Centenary edition ‘have a taste’ (which is the reading of the manuscript and of all the early printed editions) is emended to ‘had a taste’—as if Hawthorne was referring not to Miriam’s middle age but to the historical Middle Ages!¹⁷

Most of the editorial emendations regularize details of the copy-text in an undesirable way. Hawthorne sporadically attempted to use English spellings for the benefit of the English compositors, sometimes spelling ‘-our’ in place of the American ‘-or’, etc. Here the editor has altered a large number of Hawthorne’s ‘-or’ forms to ‘-our’, even when it was not normal English practice to do so, and has for instance spelt ‘horrour’ ten times when Hawthorne himself spelt ‘horror’ nine times out of the ten. (Similarly absurd are the emendations to ‘Emperour’s’ and ‘errour’ in lines 6 and 8 of the Centenary version of the extract.) There is also much regularization of Hawthorne’s capitalization, hyphenation, etc., which is not only unnecessary but also objectionable where such details convey meaning. For instance Hawthorne generally wrote ‘Nature’ where he intended personification and ‘nature’ where he did not, but in the Centenary text the word is  regularized to ‘Nature’ throughout.¹⁸

Most of these editorial emendations, very few of which are supported by textual notes, would have been better left unmade. Too much editing can be as bad as too little, and the Centenary text would have had greater critical value if the editor had done no more than transcribe the manuscript and emend it with all the verbal variants from 1860a.

The CEAA and its editors have been noble in vision, rich in good intentions as well as in funds, indefatigable in industry, and scrupulous in accuracy. These qualities, and the features of the CEAA editions that derive from them, are applauded even by the CEAA’s critics. Yet the whole great  enterprise seems to have gone astray. CEAA editions are not and never can be ‘definitive’; their main editorial principle is unsound yet inflexibly applied; and individual editions sealed by the Center can be grossly imperfect.

It would seem that editing—which is at least as much a part of literary criticism as of bibliography—cannot well be regimented, and that editors should always consider the why and the wherefore and the how of their work according to the circumstances and the needs of each individual case. Books of rules can prove delusive guides.¹

Since this was written the CEAA has been replaced by the Center for Scholarly Editions (CSE), which is likewise a committee  of the MLA. The aims and accomplishments of the CEAA (which was originally called the Center for American Editions—the name was changed in 1967) are surveyed in Bruccoli, M.J., ‘A few missing words’, PMLA, Ixxxvi, 1971, pp. 587–9; Bruccoli, M.J., and others, Statement of editorial principles and procedures: a working manual for editing nineteenth-century American texts, revised edn., New York 1972 (the original edition was prepared by W.M. Gibson and others, [New York] 1967); and Tanselle, G. T., ‘Problems and accomplishments in the editing of the novel’, Studies in the novel, vii, 1975, pp. 323–60.

The publications in the debate on the work of the CEAA up to 1972 are listed in the Statement of editorial principles (pp. 17–25), and perhaps there are further references in Tanselle’s ‘Problems and accomplishments’. The most recent, and perhaps the best, contribution is Tom Davis’s ‘The CEAA and modern textual editing’, The Library, xxxii, 1977, pp. 61–74. See also Freehafer, J., ‘Greg’s theory of copy-text and the textual criticism in the CEAA editions’, Studies in the novel, vii, 1975, pp. 375–88, and the ‘Forum’ on the CEAA which follows Freehafer’s article (pp. 389–406); Tanselle, G. T., ‘Greg’s theory of copy-text and the editing of American literature’, Studies in bibliography, xxviii, 1975, pp. 167–229; and the references given in p. 3 n. 5 (Peckham and Tanselle) and p. 184 n. 5 (Bowers and Freehafer).

² Statement of editorial principles (see n. 1 above), chs. 3 and 4.

³ Bruccoli, ‘A few missing words’ (see p. 183 n.1), p. 587; Bowers, Two lectures (see n. 4 below), pp. 25–7.

⁴ Hinman, C., and Bowers, F., Two lectures on editing, Columbus 1969, pp. 55, 57. Bowers’s lecture, ‘Practical texts and definitive editions’ (Two lectures, pp. 21–70), is the best exposition of his views on editing nineteenth-century American novels.

⁵ The background to the writing of The marble faun is well summarized in the Centenary editioin, Columbus 1968, pp. xix-xliv; see also Hawthorne: the critical heritage, ed. Crowley, J. D., London 1970.

The major articles on the Centenary Marble faun are Bowers’s ‘Practical texts and definitive editions’ (see n. 4 above), and Freehafer, J., ‘The Marble faun and the editing of nineteeth-century texts’, Studies in the novel, ii, 1970, pp. 487=n503.

⁶ For the bibliographical details of the early printed editions, see the Centenary edition, pp. xIv–Iii.

⁷ It is now British Library MS. Add. 44889–90.

⁸ Laudable but not necessarily attainable; for, rather than pay even the modest royalties asked by the CEAA, a publisher may prefer to reprint for nothing the unedited text of a work that is out of copyright.

⁹ Total accuracy is of course impossible, even if annotated photo-reproductions are used. The editor cannot avoid all error himself, nor can he be in complete control of his printers. This Centenary edition is accurate enough for practical purposes; and the suggestion that CEAA collations are likely to be inaccurate because they are  carried out by graduate students and by professors of literature poorly qualified for the task (Morse Peckham in Studies in the novel, vii, 1975, p. 402) is not borne out here.

¹⁰ It is hardly likely that a future editor of Hawthorne would feel that he could rely on the collations, etc., of the Centenary editors, or of anyone other than himself; the ‘investigator’ postulated here would be someone other than an editor who needed detailed textual information for a study of Hawthorne’s work.

¹¹ A similar conclusion was reached independently by Tom Davis (see p. 183 n. 1).

¹² Greg, W.W., Collected papers, Oxford 1966, pp. 374–91. See also the references in p. 183 n. 1; and, on Greg’s ‘substantives’ and ‘accidentals’, p. 5 above.

¹³ ‘When an author’s manuscript is preserved, this has paramount authority, of course’ (Bowers, F. T., ‘Some principles for scholarly editions of nineteenth-century American authors’, Studies in bibliography, xvii, 1964, p. 226).

‘…it is almost always true that the accidentals and of course the vast majority of the words—will be closest to the author’s intent in a finished or fair-copy manuscript, if it exists, or,  if not, in the proofs or first printing. For this reason, one of the earliest forms of the text will normally be chosen as copy-text… .

‘If the sole form of the author ‘s work is manuscript unpublished (or imperfectly published) during his lifetime, the manuscript becomes copy-text. ‘Where both manuscript and printed edition are available, still the manuscript, if it is a finished or printer’s-copy manuscript, normally becomes copy-text… . ‘If the sole surviving forms of the author’s work are printed … . that form which stands closest to the missing final manuscript is normally chosen as copy-text.’ (CEAA Statement of editorial principles and procedures, revised edn., 1972, pp. 4–6.)

¹⁴ Centenary edition, pp. cvff.; Two lectures (see p. 184 n. 4), pp. 56–7.

¹⁵ Centenary edition, p. 474.

¹⁶ MS fo. 92; Centenary edition, p. 80.

¹⁷ This emendation is not the subject of an editorial note.

¹⁸ The Centenary editor could not see any system in Hawthorne’s use of ‘Nature’ and ‘nature’ (Centenary edition, p. 471), but Professor Freehafer demonstrates Hawthorne’s intentional distinction between the two forms (Studies in the novel, ii, 1970, pp. 496–7).