Ballads and Sonnets (1881), Proofs

The proofing process for DGR's two 1881 books, the Poems. A New Edition and the Ballads and Sonnets, did not carry on for as long as the notorious case of the 1870 Poems. Nonetheless, the process was in certain respects every bit as complex, and the documentary witnesses that record DGR's meticulous attention to the proofs are almost as numerous. As in the case of the 1870 volume, here DGR used the proofing process as a compositional device. Not only did he make numerous textual revisions and additions to specific poems in the proofs, he used the visibilities of the print medium as a stimulus for rethinking the organization of the book as a whole, and especially of the content and sequencing of the poems that come in the last two sections of the book.

When DGR began to consider publishing a new book in 1879 he imagined it would be one volume, an augmentation of the 1870 volume. But as he began to get practically involved in the project, he discovered that he had too much material for a single volume, and too little for two. Throughout 1880 and into early 1881, therefore, he was heavily involved with assessing and revising old work as well as in writing new poems (see commentaries for the two volumes). The printer's copy for the 1881 Poems. A New Edition was a corrected set of pages torn from a copy of the 1873 Tauchnitz edition of the 1870 volume, augmented with manuscript copies of poems that had not appeared in any of the 1870-derived editions (the most significant of these being “The Bride's Prelude”).

The printer's copy for the Ballads and Sonnets was a heterogenous mixture of printed and manuscript materials as well. The opening section featuring the three new ballads, “Rose Mary”, “The White Ship”, and “The King's Tragedy”, was printed from DGR's fair copies of those works. Printer's copy for the 1881 version of “The House of Life” was a composite amalgam of (a) corrected printed texts taken from the torn-up Tauchnitz edition that DGR used for printing Poems. A New Edition (see above); and (b) fair copy manuscript texts for the sonnets not found in the 1870-derived editions of the sonnet sequence. The printed texts for these sonnets do not appear to have survived, but the two bound volumes of manuscripts assembled after DGR's death by Fairfax Murray—the Troxell manuscript and the Fitzwilliam manuscript—both contain many individual sonnet manuscripts that were originally printer's copy for the 1881 volume. Printer's copy for the next section of the book survives: it is the bound volume in the Troxell Collection headed “Lyrics”. An integral manuscript for the final section of “Sonnets” does not appear. In this case the printer's copies are scattered in various places.

The extant proofs are the following: a revise proof of Signature R at Yale; a miscellaneous set of proof signatures at Princeton (Signatures B, E, F, G, and L); at Harvard, at set of proofs comprising one copy each of Signatures A through L; at Delaware, an enormous set of proofs comprising multiple copies of every signature except Signature I (a total of 160 distinct signature units, most complete, some partial, and all representing various stages of the proofing process from the first author's proofs to final proofs); finally, a single set of proofs for the entire volume (now at the British Museum, but once part of the Rosenbach Library in Philadelphia). This set of proofs is gathered in separate signatures within a mockup of the design for the cover. There is one copy of each signature and all the copies are either corrected first author's proofs or corrected first revise proofs, with the exception of Signatures M and P, which are later (printer's) proofs

Analysis of these materials shows that first author's proofs for each signature were pulled on the following schedule: B (5 April); C, D (6 April); E - G (7 April); H - L (13-14 April); M - N (25 April); O -Q (3-4 May); R - X (5-6 May); Y - Z (7 May). Revise proofs proliferated from the point when DGR received these first proofs, which we can see must have been sent to him immediately. The first author's proof for Signature B, for example, is dated 5 April while the first revise is date stamped 7 April, which obviously means that DGR must have received the first author's proof for this signature and returned it with his corrections right away.

It also suggests, what is borne out by the proof witnesses, that DGR would be correcting different proof states at the same time. That situation did not prevail in the case of the long and complex proofing process for the 1870 Poems, which developed through its series of complete and integral proof states. The situation in 1881 meant not only that DGR was always proofing in medias res, so to speak, but that changes he would make in one signature would be affecting signatures that were in process of printing but that he had not yet received. As a consequence, certain of the signatures—Signatures L and N are the most outrageous examples—would pass through multiple states of revision: twenty different examples surive of Signature L, seventeen of Signature N. But several of the signatures have ten or more witnesses. Multiple copies of each signature were pulled, partly because the printer kept at least two copies in house, and partly because DGR himself wanted at least two copies (one of which he regularly had sent to his brother for comment and suggestions). Because DGR's revisions and changes were numerous, and because of the peculiar seriatim way in which the proof signatures were sent out, those four initial copies could and did proliferate into many revises.