University of Virginia Library

During the past decade there has been renewed scholarly interest in the popular culture and literature of nineteenth-century England, including — seemingly for the first time — serious attention to Victorian theatre and popular magazines. Much in evidence among these studies is a small but generally perceptive and carefully executed body of essays and books on W. S. Gilbert and the Gilbert and Sullivan operas — perceptive and carefully executed, that is, as long as the works confine their interest to criticism, biography, and literary and theatrical history. On the other hand, the scholarship seemingly relaxes its standards for scrupulous accuracy, sound theory, and consistent methodology in matters relating to the bibliography and attribution of Gilbert's texts. Mr. James Ellis' "The Unsung W. S. Gilbert" (Harvard Library Bulletin, 18 [1970], 109-140) forms a useful example. It is, as a critical and historical discussion, an insightful and useful introduction to the Gilbert of the Bab Ballads, yet its bibliographical and textual studies need some correction, containing as they do various factual errors and questionable theories. Of more general interest are the stated or tacit approaches to bibliography, attribution, and textual problems that have implications reaching beyond the treatment of a single author and into the whole emerging field of the serious study of nineteenth-century periodical literature.

The problem of fixing the publication date for a book with an undated title-page is seen in the assignment of 1873 to the first Routledge edition of More "Bab" Ballads (pp. 109 and 110, n. 2), the source for this date presumably Townley Searle's bibliography, Reginald Allen's checklist,[1] or the published catalogue of the British Museum. The dubious authority of such secondary materials in matters of dating is demonstrated by the observation that two other generally reliable catalogues, those of the Newberry Library and the Library of Congress, give the date of this book as 1872. The chief problem in accepting evidence from these catalogues, checklists, and bibliographies without further authentication is the absence from them of discussions on how each conjectural date was determined. In the case of More "Bab" Ballads, it is likely that 1873 became popularly established

as the publication date through the appearance of reviews in that year. Reviews are, at best, a kind of secondary evidence for questions of dating. As in all matters bibliographical, the chief source of information on dating should be the physical book itself, but if, as in the present case, no clues to the publication date can be found in the physical make-up or text of the book, the collateral evidence of documents pertaining to the printing and publishing of it may be brought to bear on the problem. For More "Bab" Ballads this kind of evidence exists in the form of a contract between Routledge and Gilbert dated 18 May 1872, agreeing that an edition of two thousand copies of the book will be published in 1872, presumably in time for the Christmas-gift market.[2] This dating is corroborated by the admittedly secondary yet still contemporary evidence of advertisements and entries in the "Publications of the Week" column of the Spectator and in the English Catalogue which fix the latest possible date for the publication of More "Bab" Ballads at 26 October 1872.[3]

Misstatements about the texts occur at three points, the first two having to do with the same poem, "The Advent of Spring," the earliest verse positively attributable to Gilbert. A footnote states that Punch reprinted this poem "with the title 'Sing for the Garish Eye'" (p. 111, n. 8). In fact, when Punch published it (albeit in a version somewhat altered from Gilbert's original verse), it was printed without a title, beginning only with a new first line "Sing for the garish eye," the "S" being a large illustrated initial capital. Similarly, "The Advent of Spring" appears in Mr. Ellis' essay with its original text but without Gilbert's comic by-line "By a Devout Admirer of Mr. T------n," an omission which seriously impairs a full

understanding of the poem. It is quite correct to describe the piece as Gilbert's "only attempt at sustained nonsense writing,"[4] but the poem is more than that. The omitted by-line taken together with the character of the nonce words and syntax in the text demonstrates that this verse is Gilbert's deliberate parody of early Tennyson, whose most salient poetic feature was a plethora of archaic and obscure words and word-forms.[5]

The third has mostly to do with the nature of the material contained in one of Gilbert's magazine contributions and his source for that poem. The author maintains that "an article in the Marseilles Publicité on 'la grève des célibataires' forms the basis of one poem ('The Bachelors' Strike')" (p. 117). The text of this poem as it appears in Fun begins with the following prose quotation enclosed in quotation marks as it is here:

Gilbert lifted this head-note to his poem verbatim and in toto from page 10 of the Times for 10 July 1865. The appearance of this paragraph along with the verse suggests that it and not the French original was Gilbert's source. In his edition of Gilbert's poems Mr. Ellis quotes the French original, but he attributes the Times paragraph to Gilbert as his "note of explanation" to the verse (pp. 317-318).

In the area of enumerative bibliography, at several points the essay mentions the number of poems Gilbert contributed to Fun during stated periods of time; twice these figures are incorrect (pp. 117 and 123). It is stated that "Gilbert definitely contributed twenty-three" poems to the New Series up to May 1866, whereas a count of his verse contributions to that date—even excluding dramatic burlesques wholly or partially in verse—shows the number to be thirty.[6] The number twenty-three corresponds to the number

of poems from this period which Mr. Ellis includes in his edition of "all of Gilbert's poems," (p. 4) combining two individually printed verse contributions under one title and excluding altogether six other verses as "some brief macaronics which he [Gilbert] describes as French translations of poems by Thomas Moore, and a few bits of unworthy doggerel."[7] Again in the essay it is observed that "only five of Gilbert's poems" appeared in Fun "in the twelve months following Archy's story" (June 1866 through May 1867); there are, however, seven.[8]

The footnote citing "To My Absent Husband" as the only poem of Gilbert's in Punch (p. 119, n. 12) reveals another instance of the untested authority of secondary evidence. The note tacitly approves the popular misconception that no records exist to identify the anonymous pieces published in Punch during the nineteenth century. In fact, the contributors' ledgers for every year from 1848 forward are extant and housed in the library of Punch Publications, Ltd. The ledger for 1865 discloses that on 21 October of that year Punch printed "The Return," an unillustrated poem by Gilbert exploiting the same subject matter as "Back Again!" which he had published in Fun two weeks earlier.[9] The ledger also corroborates the identification of "To My Absent Husband" as Gilbert's, an attribution which otherwise could only be conjectural, based as it is on the attendant drawing signed "Bab."

The Punch problem falls between purely bibliographical matters and questions of attribution and evidence. Similarly, to say of Gilbert that

"Down to the Derby" (28 May 1864) is "the last of his poems in the old series of Fun" (p. 113) contradicts the remark that after June 1864 "he presumably wrote more in the old series" (p. 117). The latter statement is a well-considered and plausible conjecture, although the absence of such external evidence as the Proprietor's Copy provides for the New Series precludes positive identification of Gilbert's contributions. The former assertion, however, that a given poem was Gilbert's "last" in the "old series" skirts the entire complex matter of identifying any of Gilbert's "old series" work without the aid of documentary evidence.[10] It would be more accurate to offer "Down to the Derby" as the last of those poems in the "old series" of Fun which may be even tentatively attributed to Gilbert on the tangential evidence of the accompanying illustration signed "W. S. G."[11]

Such occurrences of signed artwork—"closely related, initialed drawings known to be his"[12]—furnish the author his only evidence for ascribing five poems from the "old series" of Fun to Gilbert. Yet one of his footnotes wisely cautions against automatically attributing to Gilbert every item attached to a "W. S. G."—signed illustration, justly citing Gilbert's disclaimer of "The Lie of a Lifetime" in support of this caveat.[13] The grotesque caricatures of Napoleon III which accompany "The Lie," however, as well as the graceful young lady who adorns Tom Hood's "Croquet: An Anticipation,"[14] are as closely related to the texts of their respective

poems as are the illustrations upon which the attributions put forth in the essay are grounded. These drawings are more faithful in detail to the accompanying texts than the invariably toothy representations of the Baron which jar with his description as "a tyrant, a Tartar, a toothless and tough 'un" in the First Fytte of "The Baron Klopfzetterheim," a poem which the essay assigns to Gilbert with no reservations.[15] The one surviving piece of external evidence regarding a verse printed with "W. S. G."—signed illustrations (the letter on "The Lie") argues against Gilbert's authorship. This should not imply the equally extreme position that the presence of a drawing signed "W. S. G." denies to Gilbert any poetry or prose it accompanies. Rather, it should suggest that, since the only external evidence relates a series of such drawings to the verse of another contributor to Fun, Gilbert's practice as an illustrator of texts not his own is documented, thereby raising at least a reasonable doubt as to the authorship of verse or prose found in conjunction with drawings signed "W. S. G." Hence, the absence of external evidence should preclude admission into the canon of the five "old series" poems so illustrated and signed. Indeed, the footnote on "The Lie" argues convincingly for the exclusion of these five poems, whereas the text proper offers no valid justifications for their place among Gilbert's verse writings.

Before directly analyzing further the problems of attribution, some explanation seems desirable for the above insistence on the need for external evidence in identifying Gilbert's, or for that matter any Victorian journalist's, anonymously published magazine contributions. The case for the superior value of external evidence and the pitfalls of casually treating internal evidence in an area as seemingly remote from the present concern as Tudor and Stuart drama has been stated in a now near-classic essay by S. Schoenbaum,[16] which transcends most considerations of period and genre. The conditions of the Victorian comic paper were such as to create what amounts to a species of mass-produced literature (this view is not meant to denigrate artists who triumphed over the strictures imposed by production schedules); and this mass-production was at least one of the causes responsible for phenomena with which the student of authorship must work. Briefly, when numerous comic journalists (many of them hacks) were

busily cranking out countless pieces for openly commercial—and, hence, not altogether "respectable"—enterprises, the net result for attribution studies is the high percentage of similar styles (or lack thereof) among the authors and the equally high percentage of anonymous publication. The investigator of authorship is thus faced with a vast quantity of markedly similar unsigned work.

Professor Schoenbaum's dictum of "no style, no stylistic evidence" (p. 109) has as much validity in assigning authors to Victorian comic periodical verse as Elizabethan drama. Or, perhaps his advice to "beware . . . the formulas of style and expression that are the common currency of an age" (p. 119) is here more to the point. To read through issues of Punch, Fun, or other humorous weeklies of the 1860's is to discover an almost overwhelming, flat similarity of style and tone, not only in the verse and prose, but in the engravings as well. Perhaps this flatness, often characterized in prose by prolixity and in verse by tortured wrenching of syntax to achieve rhyme, is an absence of style, but taken altogether it could be considered a recognizable style of its own—the overall style of mid-Victorian comic journalism, but not to be associated with any specific writers in particular. And much of Gilbert's output in the mid-sixties is no different from, no better than, the rest.[17]

The most eloquent testimony to the dangers of relying on stylistic evidence to identify Gilbert's anonymously published verse is the work of Townley Searle and John Malcom Bulloch in the 1930's. In his Lost Bab Ballads (1932), Searle wrongly included four poems, not counting others of dubious authorship. His brief introduction does not discuss methodology and omits mention of the criteria by which he admitted unsigned poems into the collection. It is fairly evident, however, that the "Bab"-signed drawing for "Croquet: An Anticipation" led him to include that poem, while Searle's own intuition of some "Gilbertian" style in the drawing and text of the unsigned "Jopkins's Ghost: An Irregular Ballad" (Fun, 17 August 1867) and "King Archibald Naso" (Fun, 2 May 1868), and in the text alone of the unillustrated "Crinoline on the Ice" (Fun ["old series"], 16 January 1864) must have been responsible for their temporary intrusion into the canon.[18]

Similarly, Bulloch's alphabetical list of Gilbert's poems includes forty-three verses which Gilbert did not reprint in his own published collections. Just under half of these had previously been discovered by Dark and Grey,

Goldberg, and Searle, and most of them are correct attributions; Bulloch, however, lists twenty-five additional poems (admitting his indebtedness to Searle in locating them), of which six are from the New Series of Fun. The Proprietor's Copy shows that five of these are not by Gilbert, and while the sixth is, it is not a verse at all but an illustrated prose piece.[19] The remainder of Bulloch's new attributions are from the "old series" of Fun and The Comic News, papers for which no contributors' ledgers have yet been located, thus leaving his accuracy in doubt. His poor record, however, with the New Series of Fun raises sufficient question as to the validity of his method, one which apparently relied heavily on internal stylistic evidence. The chief fault with the approach of Searle and Bulloch is summed up by Professor Schoenbaum's point that "intuitions, convictions, and subjective judgments generally, carry no weight as evidence" (p. 112).

Returning to Mr. Ellis' article, one could suggest that the five "old series" poems have been admitted into the Gilbert canon not by internal evidence at all, but on the basis of the "W. S. G." signatures to the illustrations, which might be taken as analogous to the author's name on the page or in the table of contents or index to the periodical. And yet, the facts of "The Lie" and "Croquet" demonstrate that an illustration and a text must be treated as separate entities, and that the signature to one cannot necessarily be taken to apply to the other.[20]

The essay contains direct application of stylistic evidence in dealing with "The Cattle Show" in The Comic News, although this attribution is cautiously qualified with words like "perhaps" and "probably."[21] The process begins with the observation that the poem has the same title and subject as a verse in the 12 December 1863 issue of Fun, a piece earlier ascribed to Gilbert on the strength of a "W. S. G."—signed drawing. This method of attribution by means of comparing things greater than direct verbal parallels—here, specifically, content—Professor Schoenbaum calls "the larger correspondences of thought and theme" and again he points out that such matters are as common to a period or genre as to an individual author (p. 117). Vogues for apt subjects are observable in the London comic weeklies of the 1860's; some, such as lampoons on Napoleon III and jokes about chignons or crinolines, appear in bunches as the object of the humor is currently topical; others, such as pieces on Derby Day, the Christmas

Pantomimes, and the Cattle Show, are cyclical and come in waves as the annual events come around. That the Cattle Show was a favorite target for humorous treatment by the comic papers and other popular journals is documented by the fact that the December issues of Fun, Punch, and London Society for 1861 through 1865 contain no fewer than twenty-two poems, prose pieces, and engravings concerning the event, nine of which use "the Cattle Show" as a part or the whole of their title. Nor is it proof of mutual authorship that in 1863 the alleged Gilbert "Cattle Show" in Fun preceded by only a week the piece of the same name in The Comic News. Without further documentary evidence these events can be taken only as an example of the popularity of the topic and, perhaps, as the conscious imitation of one journal by another; H. J. Byron was the editor of both, and he may have been trying to make as much as possible of a proven annual subject for humor. There is no direct evidence to support the contention that this is an example of Gilbert's known economical re-use of material. A demonstrable instance of this practice was noted earlier with regard to the similar poems on returning from the continent which were published two weeks apart in Punch and Fun. But this observation of a particular case of one of Gilbert's life-long habits of composition could not be made until the two poems were shown to be his through the agency of documents external to the texts. This illustrates Professor Schoenbaum's principle that "internal evidence can only support hypotheses or corroborate external evidence" (p. 105).

The last step in the attribution of the second "Cattle Show" poem employs the device of comparing stanzaic forms. It is noted that the same stanza is used in "Down to the Derby" (another poem identified only by the "W. S. G." illustration), although Mr. Ellis is careful to point out that the stanza was not peculiarly Gilbert's—a frank admission of the weakness of metrics as evidence (pp. 113 and 113, n. 10).

Given the desirability of physical data in bibliographical investigations and the validity of Professor Schoenbaum's arguments for external evidence in attribution problems, the specific difficulties in the study of Gilbert's journalistic verse may be seen as part of larger weaknesses in the overall approach to the subject: a neglect of factual information, the treatment of tentative and conjectural possibilities as proven certainties, and the lack of a consistent and workable method fit for the problems and materials involved.