Medieval authors continually reworked their own manuscripts, producing a variety of competing texts during their own lifetimes. Later editors often merged, reorganized, supplemented, or suppressed parts of these manuscripts when they were having them copied by scribes, who were themselves sometimes unreliable. Thus, the proper ordering of tales within The Canterbury Tales and the proper sequential reconstruction of the framing narrative that surrounds them remain topics of dispute in Chaucer studies. In recent years, the usefulness of the Ellesmere manuscript as a guide to Chaucer^,s own intentions has been severely criticized, especially in relation to another, very early (though incomplete) manuscript, Hengwrt. In asserting that Hengwrt is closer to Chaucer’s original than Ellesmere is, scholars deploy a number of arguments: that the text of some of the Tales is more comprehensible in Hengwrt than in Ellesmere; that although both manuscripts were written within a decade after Chaucer^’s death, Hengwrt is, by paleographic means demonstrably the earlier; that the very fact that Hengwrt presents the Tales in a jumbled sequence more accurately reflects the probably chaotic state of Chaucer’s papers at the time of his death; and that the same scribe wrote both the Ellesmere and the Hengwrt manuscripts, and since a scribe was unlikely to undo a sensible established sequence, the coherent arrangement of Ellesmere is later and more representative of editorial intervention than authorial intention.

Not all of these assertions can be unequivocally accepted, and even those that can be serve as much to support as to undermine the authority of Ellesmere^’s presentation of the Tales. The chronological priority of Hengwrt, for example, is by no means certain: some expert paleographers find themselves unable to pronounce on the question. And even if one were to accept the outmoded assumption that the medieval manuscript created earliest is automatically the least “corrupt” and could demonstrate that Hengwrt was earlier, its chronological priority would prove little. Ellesmere could after all have been a copy of an even earlier manuscript one unrelated to Hengwrt but now lost to us. Nor is it certain that Hengwrt and Ellesmere were written by the same scribe: certain divergences in spelling strongly suggest they were not. More importantly while Ellesmere certainly does present the Tales in a more artistic and satisfying sequence than does Hengwrt, this hardly proves that the jumbled ordering of Hengwrt is closer either to what was available after Chaucer’s death or to his intention. If Chaucer left no finished manuscript of the Tales, Hengwrt too is the product of editorial intervention: someone gathered the materials for it and put them in a set order for a scribe to copy. Editors cannot put such posthumous materials together without making some hypothesis about authorial intentions the accuracy of such hypotheses has always depended on the information available to them and on the intelligence with which they used it.