University of Virginia Library

New Texts of Marvell's Satires
by
A. S. G. Edwards and R. M. Schuler

Debate on the text of Andrew Marvell's political poems seems to have been stilled by the combined labours of Lord and Margoliouth/Legouis,[1] to whom all students owe a fundamental debt. Yet, indispensable though their work is, it falls short of definitiveness in at least one respect. Neither editor seems to have been aware of an important manuscript authority in Princeton University Library, containing hitherto unexamined, indeed unidentified, texts of some of Marvell's political satires from the early 1670's. The following examination, both of the manuscript and of the Marvell texts it contains, suggests that the Princeton manuscript may be of significance in establishing the text of certain of these verse satires.

Princeton MS. AM 14401 (henceforward P) is a small (6” x 8”) paper manuscript of 417 pages. It contains (among others) texts of poems by Oldham, Rochester, Crashaw and Sedley as well as the two earliest substantial erotic works in English, the verse play Sodom (in two versions) and a prose translation from the Latin of part of Choirier's Satyra Sotadica.[2] It also includes texts of five of Marvell's political poems: "Britannia and Rawleigh," "[Further] Advice to a Painter," the first part of "Nostradamus' Prophecy," "Bludius et Corona," and part of "The Loyall Scot."

There are virtually no clues as to the provenance of this manuscript. The hands of only two scribes appear in it, the second of whom wrote all the Marvell poems. Neither the hands nor the general appearance of the manuscript suggests that it was the product of a professional scriptorium. It seems rather to have been a private compilation assembled piecemeal

over a period of some years. From internal evidence it seems probable that the period of compilation was the late 1670's-early 1680's.[3] Nothing further is known about its history until its acquisition by Princeton University Library in April, 1951, from Rosenbach for $525.00.[4]

The obvious interest of any new manuscript of this kind lies in the nature of the new readings it reveals. In the notes that follow all substantive variants from Margoliouth/Legouis's standard text of Marvell have been noted and then compared in turn with the full manuscript collations in Lord.

i) The Loyall Scott (pp. 362-363): text in Margoliouth/Legouis pp. 184-185, full collations in Lord, p. 458. This extract from The Loyall Scott, comprising ll. 178-185, also appears separately in MSS. Douce 357, B.L. Add. 23722, and Sloane 3413. The full list of variant readings from the standard text follows:

None of these variants is recorded in Margoliouth/Legouis or Lord.

ii) Bludius et Corona (pp. 361-362): text in Margoliouth/Legouis, p. 178; not included in Lord. This is a Latin version of (i) above. The only variants are

iii) [Further] Advice to a Painter (pp. 325-332): text in Margoliouth/Legouis, pp. 176-177, full collations in Lord, pp. 456-457. The full list of variant readings is as follows:

The variants of P diverge significantly from those of the recorded manuscripts. The omission of ll. 25-26 does occur in a number of other copies (All Souls Codrington 116; Bodleian Don.b.8; B.L. Add. 23722; and Osborn Chest II, nos. 3 and 13). In addition, B.L. Add. 23722 includes the same variants at ll. 29 and 48 (the latter also occurring in Osborn Chest II, no. 13). The only other variant P shares with the recorded manuscripts is at l. 35. There is, therefore, insufficient evidence on the basis of the variant readings to suggest a close relationship between P and any of the other manuscripts.

iv) Nostradamus' Prophecy, ll. 1-34 (pp. 367-371): text in Margoliouth/Legouis, pp. 178-179, collations in Lord, p. 459. No other manuscript or printed text appears to contain only the first part of this poem. The full list of variants follows:

Once again the variants show little significant connection with other manuscripts. Two variants, those at ll. 3 and 17, occur in Osborn Chest II, no. 13. But there is no other agreement with extant manuscripts.

v) Britannia and Rawleigh (pp. 279-299): text in Margoliouth/Legouis, pp. 194-199, collations in Lord, pp. 464-465. The full list of variant readings is as follows:

More so than with the other Marvell texts under consideration, the P copy of Britannia and Rawleigh shows extensive variation from the standard text, comprising seventy-six substantive variants in the 194 lines. Of these, eight show some agreement with the nine recorded manuscripts, but even here no coherent pattern of relation emerges: in two cases (ll. 68, 92), P's variant appears in all texts, and in two more (ll. 24 'martyrd', 39) P shares a reading with Bodleian Don.b.8, although l. 39 also agrees with Taylor 1, B.L. Harl. 7315, and B.L. Harl. 7317. The latter manuscript then joins two early printed versions (Lord's e and m, printed 1689 and 1697, respectively) in a single agreement with P at l. 113. Other single agreements are found with B.L. Add. 23722 (l. 19) and with Bodleian Eng. Poet. d.49 (l. 76). Finally, in a single instance (l. 10), P agrees with all manuscripts except Douce 357. Even in the few cases in which there is agreement with other texts, then, P shows no consistent connection with any of them.

An examination of all seventy-five variants supplied by P makes it possible to draw certain conclusions about the P text, the most important being that in a number of respects, P seems superior to the standard text. At various points, e.g., it improves grammar and rhythm.[5] Other readings seem superior in that they seem more likely to have engendered scribal corruption than to be the product of it.[6] In broader terms, P offers a text of greater satiric force, and one that shows less expurgation or bowdlerization than the standard text.[7] An exhaustive analysis of the variant readings in P

would not be appropriate here, but there is sufficient evidence to suggest that any attempt to construct a critical text of Britannia and Rawleigh would have to take serious account of the Princeton version.

The interest of the Princeton manuscript extends beyond its variant readings, and other factors increase its potential significance to the student of Marvell's political poetry. It seems probable, as we have noted, that P is a fairly early manuscript, dating from the late 1670's or early 1680's. The likelihood therefore exists that its readings may be, in certain respects, significantly less corrupt than those of other extant manuscripts. This seems especially true of the text of Britannia and Rawleigh, the latest of the political satires included in P; for the poem itself was completed by early 1675 and is therefore chronologically the closest of the Marvell texts in P to the transcription of the manuscript. Moreover, the P text of Britannia and Rawleigh stands out in its extensive variations from the received texts, and some of these variations are, as our selective analysis has shown, superior to readings in the standard text.

The textual superiority of P, in certain respects at least, is further supported, in that no other single extant manuscript contains all the Marvell poems in P. This fact, in conjunction with the lack of any significant textual agreement between P and other manuscripts, seems to imply that P is an independently derived collection of Marvell's satires, perhaps in some respects closer in time and relationship to Marvell's original text than many of the other surviving manuscripts. These factors combine to suggest that the Princeton manuscript has a potential importance which will invite the scrutiny of future editors of Marvell's political poems.

Notes