Attribution Scholarship in the Curran Index
Whenever possible the Wellesley Index investigators based their attributions on publishers’ or editors’ records, apparently authoritative marked files, and contemporary letters from presumably knowledgeable people. These external markers remain the “gold standard” for attribution scholarship, and, like its predecessor, the Curran Index has relied on them whenever they were available. Thus, as examples, the Bradbury and Evans payment account book for Once a Week has been the primary source for attributions to that periodical; a verifiable marked file for the early issues of the English Review has been a key source for that quarterly review; and comments in letters from the editor Frederick Marryat have supported many 1832-1835 Metropolitan Magazine entries in the Curran Index.
It should also be noted that, although the early- and mid-Victorian norm was anonymous publication, some articles in some periodicals were signed. Sometimes the signatures were in the form of recognizable names or meaningful pseudonyms or initials; sometimes they were in the form of a useful cross reference, such as “the author of . . .”; and sometimes the signatures were themselves puzzling. Without debating the true level of “anonymity” of these articles, the Wellesley Index included them in its listings. The value of offering complete table-of-contents and collected individual bibliographies was paramount. The Curran Index has adopted the same guideline. In particular, we have found that verse was sometimes published with meaningful signatures; for example, most of Felicia Hemans’ poems published in the New Monthly Magazine or in Blackwood’s Edinburgh Magazine either bore her name or the initials “F. H.”
Of course, in most cases articles were truly unsigned, and often surviving publishers' records or relevant letters may either not exist or not be readily identifiable. In those cases the Wellesley investigators considered the subsequent reprinting of a periodical article in a signed collection or document to be reasonable evidence of authorship. Unfortunately, matching various collections to periodical articles in the twentieth century was something of a hit or miss affair. How could one systematically determine what collections were relevant, and what items in a given collection matched what contributions to which periodicals?
Now, with the advent of digital tools such as the British Newspaper Archive, Google, Google Books, and the Hathi Trust archive, that situation has radically changed. Searches can be made using the title of the article and / or a key phrase from an article to see if that article was discussed or claimed, or reprinted, in now-digitized books or periodicals. WorldCat and the on-line catalogues of major libraries such as the British Library or the Bodleian Library at Oxford can assist with the tracing of cross-references, or even establish linkages to reprinted pamphlets or volumes that may not have been digitized.
Many of the attributions in the Curran Index are based on evidence from these internet-age tools.
Some scholars (perhaps most notably Alexis Antonia and Ellen Jordan at the University of Newcastle in Australia) have used the Burrows Method of textual analysis, which seeks patterns in the statistics of word usage, to probe the identity of the authors of anonymous articles. The Curran Index has cautiously integrated some of their findings into its records; however, we note that statistical analyses seem to be inherently more appropriate for larger articles than for smaller ones, and that a greater degree of editorial involvement in the wording of a particular article might undermine the premise of the analysis.
Of course, sometimes an attribution cannot be definitively established, but various sorts of internal evidence -- usually encompassing a mixture of circumstances, subject matter, expressed opinions, article style, and the established attribution of related articles – suggest a likely author. Following Wellesley, we believe that the positives of having a more complete set of attributions outweigh the certainty that some uncertain attributions will be erroneous, as long as the user of the Curran Index is suitably warned. Accordingly, if we somewhat subjectively believe an attribution is at least 50% likely, but not 90%, we identify it as “probable.” Attributions which have a reasonable basis, but are realistically still quite doubtful, are marked as “possible.”
The Curran Index distinguishes between the terms “contributor” and “author.” The original developer of a translated piece of literature remains, of course, its author. When a story by Dumas or Hugo, or a poem by Heine or Schiller, was translated into English and published in a British periodical, the original author should be recognized. Following that principle, the Wellesley Index sometimes (and inconsistently, sometimes not) listed among its contributors many non-English writers who did not intentionally publish items in the British periodical press. The Curran Index, however, is more concerned with identifying those individuals who actively interacted with editors, who submitted manuscripts and deliberately contributed to British periodicals. Translators played an important role, and their contributions should not be obscured. Accordingly, whenever known, the authors of originally non-English items are identified in comments or in article titles, but translators receive primacy as contributors.
It should also be noted that sometimes, even when an article signature or a publisher’s record connects a name with a given article, the article may still be effectively anonymous. Outside of other information, who would know, for example, what to make of a “J. Smith” associated with a particular article? Our position is that the curtain of anonymity has not been fully pierced until some descriptive matter can be attached to a name. A great deal of effort has been made, using tools such as Frederick Boase’s Modern English Biography and digital tools such as the genealogical website FamilySearch, the British Newspaper Archive, and Worldcat, to attach some meat to the bones of otherwise unrecognized names. In essence, we regard attributions as incomplete until information such as life dates, occupation, nationality, education, or other writings can make attributions more meaningful.