 | |  |
In an important article on "The Text of Dryden's Mac Flecknoe" (Harvard Library Bulletin, VII (1953), 32-54), Professor G. B. Evans has called the attention of future editors of the poem to the textual evidence provided by seven early manuscripts and the printed texts of 1682 and 1684, which he has assembled in several pages of detailed collation. He is unable to establish the relationship of these texts, except that he believes that one of the manuscripts is simply a copy of the 1684 edition, but he argues that when all or almost all of the other texts oppose the text of 1684 they present genuine early readings which are sometimes preferable to the readings of 1684. It is the purpose of the present paper to suggest a relationship for these texts, and on that basis to attempt to distinguish early readings from late and to investigate further the likelihood that the 1684 text is in some places corrupt. The concepts and methods employed are in part, I believe, new, and will, I hope, prove generally useful.
Professor Evans designates his texts by the sigla 82, 84, L, H, B, C, F, M, and I. In his list of variations he had recorded a number of accidentals-- differences in spelling ('upward(s),' line 215), hyphenation ('Royal(-) Barge,' line 39), and the like--which, as Greg has warned, only obscure the relationship of the texts in question.[1] The list should be confined to substantive variations, and to variations in punctuation that markedly affect the sense and so may be called substantive. Substantive punctuational variations are subject to a further restriction, because H and M have almost no punctuation. Therefore, so far as their punctuation is concerned, H and M are the equivalent of imperfect manuscripts, and their punctuation
An examination of this list discloses that there are type-1 variations for each text, but it must be noted that there are none for 84 when I is present. It follows, according to the usual reasoning (see Greg, Calculus, p. 55), that I derives from 84; and accepting this relationship for the present we may ignore I for the time being, and turn our attention to the other texts. An analysis of the type-2 variations and of such of the complex variations as can be reduced to the simple types discloses no consistent pattern, except that of inconsistency. To put it another way, no group of texts occurs (I being now out of the question) from which all the others are excluded or almost always excluded. It is for this reason that Professor Evans, though he persists in thinking of his texts as falling into groups (pp. 43-45), devotes his attention almost exclusively to the instances where the readings in 84 are opposed by all or the majority of the other texts. It does not follow, however, as Professor Evans suggests that it does (pp. 43-44), that the edition of 1682 has no more authority than the manuscripts. Distributional study can indicate no more than something of the transmission of a literary text through or to its various exemplars. Decisions as to the authority of these exemplars must rest upon external evidence, in this case the long
Hill would interpret the variations listed above as indicating that the texts are collaterals independently derived from a common ancestor in the relationship called radiation.[2] Greg's logic allows him to assume simple radiation only when all the texts have type-I variations alone or almost alone; if other variations occur with any frequency he can see no course other than to assume some sort of conflation somewhere (Calculus, pp. 21, 56; 43).[3] Greg does not deal explicitly with chance coincidences, but Hill points out that they occur constantly and indicates something of how often any one may recur and still be accepted as chance (p. 77). The present set of variations, however, is not explicable as the result of chance coincidence or conflation or both. The proportion of chance or the amount of conflation or both would be unbelievably large, for only B agrees, in different variations, with as few as four of the other texts, and 84, L, and H agree in one variation or another with all the others. The only alternative, it would seem, is to accept the relationship of the texts as radiation and to explain the conflicting variations as resulting from authorial revisions, with perhaps some admixture of chance coincidence or conflation or both.
Under the circumstances, it appears likely that I is a collateral instead of a descendant of 84, for in 177, 2 where 84 agrees with B, I agrees with the other texts, and Professor Evans mentions other details in which similar relationships may be observed (pp. 34; 41, variation 211 1). These details seem too many to explain as coincidence, and yet too minute to explain as conflation, and coincidence and conflation are last-ditch explanations in any event. Once it is seen, however, that Dryden allowed copies of his poem to be taken before it was published, and that he revised the text between these copies, then the differences between I and 84 are simply explicable as resulting from small final changes made before he sent his manuscript to the press in 1684.
| | |