CHAUCER'S "BOOK OF THE DUCHESS": A PROPOSAL

Abstract

The poem is imperfectly extant in three manuscripts in Oxford (Bodleian Library, MSS Fairfax 16, c.1450 and its affiliate Bodley 638 after 1475, and Tanner 346 c. 1440) and in Thynne's edition of Chaucer's works of 1532. The distance between the original poem written c. 1370 and these copies is marked by a noticeable degree of scribal contamination. The first modern editor to address the textual problem was Skeat in 1894 who offered several emendations, mainly to metrically uncertain lines, and explanatory notes which listed some unsolved cruces.1 He was followed in 1928 by Koch, who re-edited the poem with full textual apparatus and advanced the scholarly enquiry.2 Both editors believed that the poem was written in octosyllabic couplets which scribal error and linguistic change had distorted. In 1933 (and in a second edition in 1957) Robinson abandoned such enquiry for a more conservative approach:it appeared unsafe to indulge in wholesale emendation to improve the text of an early poem in a meter of rather rough and free traditions.3This approach was followed, without a critical apparatus of depth, in three collections of Chaucer's verse edited by Donaldson in 1958 (2nd edn 1975), Baugh in 1963, and Fisher in i977.4Each editor explained his policy thus: Donaldson, Ι have usually followed the manuscripts traditionally considered the best, except in a few cases where I have experimented ... in a number of lines which the scholar will recognize I have adopted a reading I consider superior'; Baugh, 'the text is based on MS Fairfax 16. The erratic use of final -e, often ungrammatical, has been corrected where necessary'; and Fisher, 'the text follows MS Fairfax 16 more closely even though the spelling is not always consistent and produces many hypermetric (though not necessarily unrhythmic) lines'. And in 1987 the Riverside Chaucer, which marks a divide between the old and the new, simply copied Robinson's text and slightly updated his apparatus, content to follow the dicta in its introduction (by Norman Davis) that the poem was written in the 'familiar type of verse' in four-stress couplet. This brief apparatus refers the reader to the fuller coverage of an unpublished edition (by D. A. I. Dickerson, Chapel Hill, NC, 1968) and of the more recent edition by Phillips in 1982 (2nd edn 1993).5 The latest edition of the poem appeared in 1999 in a spacious CDROM edition by McGillivray with a comprehensive apparatus which recorded even the punctuation marks of the scribes. These three editors adopt a conservative approach to their copy-texts, Phillips emending 'only where [her copy] is very unsatisfactory' and McGillivray offering even fewer textual changes.6None of these editors therefore has produced a text which aspires to be much more than a slightly polished transcription of their copy-text, and none has attempted even a critical reconstruction of the lost archetype of the four extant copies. The heavy weight of Robinson's example in editing this poem at least has been wholly malignant, leaving unresolved the many cruces and rough phrasings that, despite Skeat and Koch, disturb the extant copies. Thus the reader of the Book of the Duchess who is aware that these copies contain a more rewarding poem slightly out of focus must return to the primary texts and consider their evidence afresh. Immediately such investigation requires three determinations. First, to establish the most suitable (or least problematic) copy among the extant affiliation on which to stand and, as a minor corollary, to establish the value of Thynne's print of 1532. Secondly, to determine the metrical shape of the couplet, either the octosyllabic lines advocated by Skeat and Koch or the 'rough and free traditions' favoured by Robinson. Thirdly, to bring to the reading of the poem the import of modern scholarship on 'Chaucerian' language and usage and scribal practice observable in a poem written c. 1370 but only extant in imperfect copies made at least seventy years later. …