University of Virginia Library

The Early Editions of Dryden's State of Innocence
Marion H. Hamilton

Nine quarto editions of Dryden's State of Innocence were printed before publication of the 1701 Folio of the Works, shortly after Dryden's death, brought to a close the early history of the text with a tenth edition. Although the chronological order of five of these editions is established by the title-page dates, which there is no reason to suspect, the same is not true for the remaining five, three of which bear the date 1684 and two the date of 1695.

In Hugh Macdonald's Bibliography of John Dryden the order of listing of these editions is presumably intended to be chronological, although no statement is made concerning the evidence on which the order is based. The Woodward and McManaway Check List of English Plays 1641-1700 follows the Macdonald listing. However, textual collation of the nine quartos and of the Folio makes evident that this commonly accepted sequence is not correct, nor is the derivation of the various editions a simple matter of successive reprints as has been generally, though implicitly, believed. The following table summarizes what investigation establishes as the true order of the printed quarto texts.[1]

                   

Edition	Date	Macdonald	W & McM
Q1	1677	81a	465
Q2	1678	81b	466
Q3	1684	81e	469
Q4	1684	81d	468
Q5	1690	81f	470
Q6	1692	81g	471
Q7	1695	81i	473
Q8	1695	81h	472
Q9	1684[*]	81c	467

Obviously Dryden was concerned with the printing of the first edition of the play: the prefatory essay "The Author's Apology for Heroique Poetry and Poetique Licence" was written, he states, partly to explain why he chose to publish an opera that was never acted. Since textual collation shows no indication of fresh authority having been added to any of the later editions, and instead discloses only the usual steady deterioration of the text, it is clear that Q1 is the only substantive printed text of the play.

In brief, the derivation of the texts is as follows: Q2 was printed from Q1, Q3 from Q2, and Q4 from Q3. Q4 is a terminal edition from which no other derived. Although Q3 and Q4 were both printed in 1684, the quartos are readily differentiated. Q3 collates A-G4, with advertisements on sig. G4r-v. Q4 collates A-E4 F2 G4 (G1 missigned F1). [See below for the differentiating points between Q4 and its piracy, Q9.] Q5 derives from Q3, Q6 from Q5, and Q7 from Q6. The 1701 Folio was printed from Q7. Q8 was printed from Q7. Since both Q7 and Q8 are dated 1695 and have the collation A-F4 G2 they could be confused unless one notices that in Q7 there is a semi-colon after Innocence in the running-title on $2v4v(—G2v) and a semi-colon after and in the running-title on C3v; whereas in Q8 the semi-colon after Innocence is found only on C2v4v, and a comma appears after and on C3v.

The piracy Q9 imitates the title-page and the collation A-E4 F2 G4 of Q4, but its text derives immediately from Q8. However, there is no difficulty in distinguishing it from Q4 since Q9 uses running-titles in the headlines but Q4, instead, has only page numbers within parentheses.

The total evidence on which this statement of derivation is based is too extensive for presentation here but may be briefly summarized.

The lack of significant variation in the text of Q2, and the fact that its preface and the text for Act I is a page-for-page reprint of Q1, establishes that Q2 derives from Q1 and not from an independent manuscript. There are 22 substantive variants in Q2, eight of which are corrections of obvious errors in its copy. Of the remaining 14, 2 are simple misprints corrected in Q3, 3 are obvious errors but were retained until Q5, and the rest appear, with occasional modification, in all subsequent editions. Of these the most serious was the omission of the second half of the line in Act V (Q1, sig. G2r)

In substantives, Q3 has 18 variants from Q2 in addition to the Q2 variants from Q1 which it follows. In turn, all but three obvious errors in the Q3 total of variants appear in Q4, which—being carelessly printed—adds 49 unique substantive variants of its own. With only one exception, which may just possibly be coincidental, none of these unique variants appears in any later edition, this fact establishing Q4 as a terminal text.

Q5, accepting all but the most flagrant of the Q3 variants, adds 19 to the list, all of which appear in Q6 together with 27 new readings. In turn, Q7, repeating all but four of the most obvious Q6 errors, adds 18 new substantive readings, of which 16 appear in Q8 together with 19 new variants.

Q9, the last of the early quartos, presents a most peculiar case. Dated 1684 and having in its register of signatures and title-page typography a definite relationship with Q4, also dated 1684, it nevertheless shows (except for four pages) a page-for-page correspondence with Q8 printed in 1695. Superficially, the simplest view would be that this edition (Macdonald 81c) was a corrupt version of Q4 and that, eleven years later, it served as copy for Q8. Such an assumption is impossible, however. In the first place, Q4 (M 81d) could not have served as copy for any other edition because

of its extraordinary number of unique and non-reversible readings which prove it to be a terminal text. Secondly, Q8 (M 81h) could not have derived from Q9 (M 81c) since Q9 with 38 unique readings, over half of which are non-reversible, is also shown to be a terminal text. Moreover, there is definite evidence that Q9 textually derives from Q8, a fact which re-enforces the otherwise ambiguous evidence of its general page-for-page correspondence with Q8. Of the 20 substantive variants which first appeared in Q8, 11 are simple errors which Q9 corrects; but Q9 follows the other 9 cases of Q8's hitherto unique variants and adds no less than 39 of its own. Thus Q8 could not have derived from Q9; and instead Q8 must have served as copy for Q9. This can be demonstrated, also, for sheet C and half-sheet F.

We have, then, the peculiar situation that Q9, printed in imitation of Q4, and falsely dated, nevertheless did not choose Q4 as copy but instead was set up from the text of Q8. Whatever the reasons for this oddity,[3] the facts rule out inadvertence, the more especially since Knight and Saunders—listed in the Q9 imprint as sellers, in imitation of Q4—were no longer in business together by 1695. Piracy is the only answer, but the exact date cannot be determined except that it was in 1695 or after.

There is one further peculiarity associated with Q9. Although there can be no question that Q8 served as printer's copy, Q9 corrects six Q8 readings by reversion to the forms found in some earlier edition. Since various of these in the Q8 forms were plausible enough to have escaped suspicion, simple compositorial ingenuity is insufficient to explain their revision, the more especially since they appear in clusters.

             

Q8 reading	Q9 revision	possible source
Shut from his day and that contented skie	contended (C1v)	Q1-4
Thy . . . voice the sleeping Gods we'll reach	will (C1v)	Q1-6
With wings expanded wide, ourselves will reach	we'll (C1v)	Q1-6
What raised thy Beings, ours will take away	Being (F1r)	Q1-6
Such beauty may rise factions in his Heav'n	raise (F1v)	Q1-6
Imprudence was your fault, but Love was mine	is (G1r)	Q1

If for the moment we disregard the variant on sig. G1r, the evidence points only to some edition between Q1 and Q4 being collated to produce these revised readings. Since it is clear that the printer had had access to Q4, it would seem a natural assumption that Q8 for these pages had been collated against Q4. This is probably the answer, although there are difficulties. First, the return to a unique Q1 reading on line 5 of the adjacent page would need to be fortuitous. It is true that in context the change of was to is makes better sense; nevertheless, the was reading had passed muster with the compositors of Q2-8. Hence it might be possible to conjecture that Q1 was the collating copy and that this collation extended over to the top of sig. G1r. Second, it may seem suspicious that in no case does Q9 follow a unique Q4 variant,[4] although two such unique readings would have been present for Q9 sig. F1, one for F1v, one for F2, and three for F2v. The text for Q9's sig. C1v could have contained no unique Q4 reading, however.

Yet we may be the more prepared to accept as chance the collator's invariable rejection of Q4's unique variants (which have nothing much to recommend them at best), since the alternative is the hypothesis that a copy of Q1 was used as the collating edition. In fact, the collation of Q1 (making three editions in the possession of the Q9 printer) could be advanced only if we assumed that Q4, although available for consultation, was not available to be kept on the premises.[5] This seems to invoke such special and peculiar circumstances that we are better off with the simpler hypothesis that Q4, which served as the specific model for the piracy, was also used for the small amount of collation actually engaged in.

That any textual collation whatever with an earlier edition was contemplated, and in fact performed even to a minor degree, is a fact sufficiently astonishing in a late reprint and a piracy without our conjecturing such extraordinary textual conscience in the pirate as to secure a copy of Q1 for collation, when he was also in the possession of Q4. In truth, the only plausible explanation why this collation was ever engaged in was the availability of Q4 in the shop as a model for the piracy.

Notes