"Paradise Lost: A Poem Written in Ten Books": An Authoritative Text of the 1667 First Edition

Textual Variants

TEXTUAL VARIANTS In the course of printing, changes may be made to a text as a result of recognition of an error by the compositor, disruption of a letter or letters in the mechanical operation of printing or making changes, and sometimes because of a deliberate change by the author. The first two reasons yield some significance for the authorial text, although they basically comment upon the printer’s work and conditions of work. Yet some of these alterations do tell us about the “authorial” copy-text, and these will be discussed below. At times changes necessitate an adjustment of spacing, and often these printing changes occur on conjugate leaves, suggesting the nature of the compositor’s working habits and attentiveness. Stop-press variants, which created different states of a page, corrected errors made in the course of setting type and printing, resulting usually in a stoppage of printing, correction, and then continued printing. Probably many of the textual variants recorded here are of that nature. Fletcher’s assignment of states is based on such stop-press changes. The pages of the 1667 edition of Milton’s poem evidence the following textual and press variants, listed here as state 1, state 2, state 3, and state 4 when such states occur. Fletcher’s citations do not identify in which copies the states occur (with the exception of a few variations, as noted below), a situation that therefore does not allow us to determine whether states of the text reported are related to issues of the text. We do not know, that is, whether there are true press variants for the 16671 text, thus creating state 2 of a page, or whether the resultant so-called states were created when new issues were produced. However, the mixture of states that exists in some copies of later issues but not in all 395 396 DiscussionoftheEditedText copies of those same issues points to the contention which Fletcher posits, that a state did not come into existence because a new issue was being produced but rather that such states existed earlier than the new issue, which mixed such states at random with other states of the text, thus affecting some but not all copies of an issue. Fletcher treats the issues aside from the sixth as being the same printing as the first except for the title page; that is, the only new printing for an issue was the title page (but see textual variants for book 5 below). For the sixth issue, 16692 , signatures Z and Vv were reset; one concludes that this occurred because not enough copies of those signatures existed for the full sixth issue when that issue was produced. Other pages do not evidence resetting even though variants may appear on them. Generally it has been thought that either copies of the 16671 text had the title page removed and a new title page added for the successive five issues (a very common practice), or a new issue was created by taking signatures from a pile of previously printed signatures (printed in 1667 when the first issue was being produced) but that had not yet been used (in the first issue). Probably both possibilities are correct. This would demand the maintenance of such piles of unused pages from August(?) 1667 through perhaps April(?) 1669. Such piles of signatures would, under the conclusion to which the evidence points, include first, second, and third states, thus accounting for the accidental use of “inaccurate” pages in later issues. It is most curious that almost none of the variant pages appears in copies of 16671 . (See the variants below for book 1, however.) Fletcher seems to base his opinion for calling a page state 1 or state 2 on what he believed were or were not Miltonic practices of spelling and other mechanics of writing, and upon variants that are adjudged corrections , thus being labeled state 2. Further, whenever a reading is not the better reading or is a clear error (this happens particularly when line numbering errors occur), he calls it state 1 and the “better” reading or “correction,” state 2. We may thus find Fletcher’s ordering of states questionable and even unacceptable, since neither possibility is consistently true. The term “state” should mean that the text TextualVariants 397 designated state 1 was set first and that changes to that text (particularly changes made while printing is taking place) thus...