University of Virginia Library

Yale University Library MS Osborn b. 105 is the best known of the large body of manuscript miscellanies of verse from the Restoration period. Even in its present mutilated state, with a total of forty-five leaves excised by some prudish past owner, it has been accepted by both David M. Vieth and Keith Walker as the most important single witness to the text of Rochester. Yet it is also fair to say that, since the manuscript was first brought to the attention of scholars in 1963 through Vieth's monumental study in Attribution in Restoration Poetry, there has been no substantial increment to our understanding of it.[1] The problem is not so much one of lack of evidence as of uncertainty about what questions scholars should be asking. In the feeling that a quarter of a century is too long to have been waiting for a significant advance on Vieth's findings, I would like to present some new hypotheses concerning the origins and production of the manuscript.

First of all, it is necessary to consider what is already known or has been plausibly guessed about Osborn b. 105. It is a collection that before mutilation probably consisted of between seventy and eighty poems in the first or original hand, along with eight added later by a second hand. What immediately follows is concerned only with the poems in the first hand. The most important datum is that the contents of its surviving pages and their order are largely identical with those of the corresponding sections of the 1680 Poems on Several Occasions by the Right Honourable, the E. of R--- (henceforth 1680H), with the exception of eleven poems unique to the manuscript and one unique to the edition.[2] Vieth (pp. 80-83) demonstrates by means of line counts that the correspondence probably continued through the various gaps caused by excision, apart from two instances where the manuscript again had room for additional material. Although the precise relationship of these two texts has yet to be established, and Vieth suspected some measure of conflation from an outside source, it was undoubtedly close.[3] Vieth (p. 74) believed that the lost manuscript (henceforth MS1680H) which served as copy for 1680H was the work of the same scribe and resembled Osborn b. 105 in containing a number of ascriptions to authors, arguing that these were omitted by the printer in order to pass off the edition as being entirely the work of Rochester. This too is simply a hypothesis, and one that should not be accepted uncritically, but is plausible in the light of our present knowledge. The alternative would be that the attributions were added from personal knowledge

by the scribe of Osborn b. 105 or in some manuscript intermediate between it and the nearest common ancestor. Most of the poems also exist in other contemporary manuscript anthologies and sometimes also as single sheets of the kind that were passed from hand to hand at court, sent through the post with newsletters or made available to the frequenters of coffee-houses.[4] Vieth's suggestion, based on the estimated date of the last-composed item, that the manuscript's contents were first assembled in the spring of 1680 will hardly be argued with (p. 65). It is written in a clear but unfussy professional hand which is also found in Princeton MS Taylor 1 (see below) composed of political satires of the time. As mentioned, the surviving sections of the manuscript lack only one poem present in the edition, and that the very last, "On Rome's Pardons." Vieth wonders if this may have been added by the printer in order to fill up space on the last sheet (pp. 79, 355-356).

The consistently good quality of the text of the 1680 compilation (here used to cover both the manuscript and the edition) is confirmed by the practice of recent editors of Rochester who all show a strong preference in their choice of copy texts for one or the other, usually favoring the manuscript.[5] The poems themselves range in date of composition from the mid-166os to late 1679, and some at least had been circulating in manuscript as separate entities for many years prior to the compilation of the collection.[6] These considerations suggest a collection assembled step by step over a considerable period by someone who was consistently able to lay his hands on good texts of poems soon after their appearance, and who therefore presumably moved in the same literary circles as the writers. A useful contrast can be made with the topical verse sporadically included in the commonplace book of the courtier Sir William Haward (Bodleian MS Don b. 8) which includes a number of items which also appear in the 1680 compilation. Haward's contacts were also good and he often obtained copies at very close to the apparent date of composition, but neither was his collection as comprehensive nor his information about authorship nearly as assured as that of the 1680 compiler. His texts, moreover, vary greatly in their quality.

The presence of eleven poems in the surviving leaves of the manuscript which are not found in the edition (with more probably lost from the excised sections) raises the question whether we are dealing with a collection that expanded in the course of recopying (with the edition representing the earlier state) or whether MS1680H had the same contents as Osborn b. 105. Vieth argues for the second alternative, maintaining that MS1680H probably contained all the poems contained in Osborn b. 105 but that some had to be left out of the edition for reasons of space. His assumptions about the nature of MS1680H become evident in the course of speculations about the nature of the nearest common ancestor (called by him "the archetype") from which they derived the readings they share against other sources:

Most of this is plausible, but none of it is conclusive. Surviving families of Restoration miscellanies are not so numerous that we can generalize from them without risk. Moreover, if we widen our horizons by looking at scribal publication as a phenomenon of the earlier as well as the later Stuart period, and as embracing such diverse activities as the distribution of musical part books and reports of parliamentary proceedings, it can be argued that the dominant model is that of the "rolling" archetype from which old material is continually removed and new material added in the course of production.[7]

Aware of this objection, Vieth invokes the evidence of "linked groups" of poems, which appear in the same order in other manuscript sources, arguing that in four cases it is Osborn b. 105 which preserves the "normal" group while the edition has it in reordered or (in one case) truncated form (pp. 76-79). In these cases it is hard not to agree that the "archetype," however conceived of, will have preserved the attested order; yet it would still be possible that the changes originated during the copying of MS1680H rather than when it was being marked up for printing. Vieth's response to this possibility is to argue that it is the change of medium which is more likely to have provoked the alteration:

Here the argument, although logical, is not as strong as it seems. In the first place, the point made in the last sentence only holds if one restricts oneself to the eleven poems that can actually be shown to be missing from the edition (and whose order in Vieth's reconstructed contents table to Osborn b. 105 of 72 items is 3, 47, 53, 55, 58, 61, 62, 67, and 69-71) and ignores his demonstration (pp. 93-100) that additional material totaling more than seven pages had also been present in the fourth and fifth gaps, corresponding to items 17-22 and 34-63. Moreover, even if true, it would not meet the objection that, in a "rolling" collection, new material which was not linked in an obviously thematic way with items already present would also tend to cluster towards the end. A more general objection is that a printer setting from a bound manuscript, as Vieth believes to have been the case, would have no reason to alter the order of items, this being much more likely to happen when a text is being copied from an exemplar in the form of separate leaves or booklets. Viewed in this light the case for a rolling archetype against a static one is stronger than Vieth gives it credit for being, though not strong enough for us to assume without further corroboration that 1680H represents an earlier state of the compilation and Osborn b. 105 a later, enlarged one. Vieth's uncharacteristic failure to think through this question seems to arise from his attention having been directed towards problems of determining attributions and establishing a text for Rochester, in which the issue is not of very great moment. However, it becomes of crucial importance (as will be seen later) as soon as one wishes to analyse the contents of the compilation for evidence of its genesis.

A further argument for accretion is that a process of this kind can be observed at work in the manuscript itself in the form of the group of poems added at the end by the second hand. These are of similar date and may well have been added by an assistant to the original copyist, but are all absent from 1680H and seem unlikely to have been present in MS1680H.

Another issue is whether Osborn b. 105 is to be considered a "factory manuscript," produced for sale, or a private commission from a litterateur to a scrivener or secretary. Vieth's speculation that it "formerly existed in several more copies in the same handwriting" endorses the first of these possibilities and leads him to the further suggestion that the manuscript was one of a set of complementary compilations, another of which is represented by Princeton MS Taylor 1, whose contents are entirely composed of political satires. Noting that the only songs contained in Osborn b. 105 are those by Rochester, Vieth (p. 90) further proposes that there might have been a third volume composed only of songs. Although there is no evidence of a connection it should be noted that exactly such a collection has become incorporated into Edinburgh University Library MS DC. 1. 3 (pp. 108-112), a compendium of six independently circulated miscellanies. However, to accept the likelihood that the manuscript was produced for sale (an assumption strengthened by evidence to be presented later) is not to imply that those responsible for copying it were also the original compilers of the collection.

One last piece of textual evidence that only became available in 1986 is that the text of Rochester's "Seneca's Troas" in Osborn b. 105 is identical in its substantive readings with one printed by Charles Blount in his edition in 1680 of The Two First Books of Philostratus.[8] Blount's source for this was presumably the manuscript of the poem he mentions in a letter to Rochester of 7 February 1679 in terms which have led some writers to assume that it was a gift from Rochester, though in my view with insufficient justification. The simplest and in textual terms the most economical explanation for this close agreement would be that the compiler of the manuscript took his text from Blount's volume; but other explanations are possible and might be worth reviewing should any further evidence emerge concerning the matter.[9]

This discussion does not attempt to deal with the intricate question of

the relationship of the "archetype" to other surviving collections containing a high proportion of the same poems, such as Stockholm Riks-Bibliotheket MS Vu 69 (Vieth's "Gyldenstolpe MS"), Leeds Brotherton MS Lt. 54, Nottingham University Library MS Portland PwV. 40, Victoria and Albert Museum MS Dyce 43, Vienna Nationalbibliothek MS 14090, and Harvard MS Eng. 636F. The most likely possibilities are (1) that the poems common to these manuscripts and the "archetype" are derived from an ancestor of the archetype, or (2) that they and the "archetype" were independently garnered selections of material that was in scribal circulation as single poems and groups of poems. It does not seem to be the case that the texts of either group are descended from those of the other.[10] The matter is one to which I hope to return on another occasion.