The question of whether The Canterbury Tales is a finished work has not been answered to date. There are 84 manuscripts and four incunabula (printed before 1500) editions of the work, dating from the late medieval and early Renaissance periods, more than for any other vernacular English literary text with the exception of The Prick of Conscience. This is taken as evidence of the Tales' popularity during the century after Chaucer's death. Fifty-five of these manuscripts are thought to have been originally complete, while 28 are so fragmentary that it is difficult to ascertain whether they were copied individually or as part of a set. The Tales vary in both minor and major ways from manuscript to manuscript; many of the minor variations are due to copyists' errors, while it is suggested that in other cases Chaucer both added to his work and revised it as it was being copied and possibly as it was being distributed. Determining the text of the work is complicated by the question of the narrator's voice which Chaucer made part of his literary structure.
Even the oldest surviving manuscripts of the Tales are not Chaucer's originals. The very oldest is probably MS Peniarth 392 D (called "Hengwrt"), written by a scribe shortly after Chaucer's death. Another famous example is the Ellesmere Manuscript, a manuscript handwritten by one person with illustrations by several illustrators; the tales are put in an order that many later editors have followed for centuries. The first version of The Canterbury Tales to be published in print was William Caxton's 1476 edition. Only 10 copies of this edition are known to exist, including one held by the British Library and one held by the Folger Shakespeare Library.
In 2004, Linne Mooney claimed that she was able to identify the scrivener who worked for Chaucer as an Adam Pinkhurst. Mooney, then a professor at the University of Maine and a visiting fellow at Corpus Christi College, Cambridge, said she could match Pinkhurst's signature, on an oath he signed, to his handwriting on a copy of The Canterbury Tales that might have been transcribed from Chaucer's working copy. Recent scholarship has cast severe doubt upon that identification.
In the absence of consensus as to whether or not a complete version of the Tales exists, there is also no general agreement regarding the order in which Chaucer intended the stories to be placed.
Textual and manuscript clues have been adduced to support the two most popular modern methods of ordering the tales. Some scholarly editions divide the Tales into ten "Fragments". The tales that make up a Fragment are closely related and contain internal indications of their order of presentation, usually with one character speaking to and then stepping aside for another character. However, between Fragments, the connection is less obvious. Consequently, there are several possible orders; the one most frequently seen in modern editions follows the numbering of the Fragments (ultimately based on the Ellesmere order). Victorians frequently used the nine "Groups", which was the order used by Walter William Skeat whose edition Chaucer: Complete Works was used by Oxford University Press for most of the twentieth century, but this order is now seldom followed.
General Prologue The Knight's Tale The Miller's Tale The Reeve's Tale The Cook's Tale
|02Fragment II||B1||The Man of Law's Tale|
|03Fragment III||D||The Wife of Bath's TaleThe Friar's TaleThe Summoner's Tale|
|04Fragment IV||E||The Clerk's TaleThe Merchant's Tale|
|05Fragment V||F||The Squire's TaleThe Franklin's Tale|
|06Fragment VI||C||The Physician's TaleThe Pardoner's Tale|
|07Fragment VII||B2||The Shipman's TaleThe Prioress's TaleSir Thopas' TaleThe Tale of MelibeeThe Monk's TaleThe Nun's Priest's Tale|
|08Fragment VIII||G||The Second Nun's TaleThe Canon's Yeoman's Tale|
|09Fragment IX||H||The Manciple's Tale|
|10Fragment X||I||The Parson's Tale|
An alternative ordering (seen in an early manuscript containing The Canterbury Tales, the early-fifteenth century Harley MS. 7334) places Fragment VIII before VI. Fragments I and II almost always follow each other, just as VI and VII, IX and X do in the oldest manuscripts. Fragments IV and V, by contrast, vary in location from manuscript to manuscript.