History of the manuscript
This was one of the collection of manuscripts amassed at the mansion of Hengwrt, near Dolgellau, Gwynedd, by Welsh antiquary Robert Vaughan (c.1592–1667); the collection later passed to the newly established National Library of Wales as the Peniarth or Hengwrt-Peniarth Manuscripts.
The Hengwrt manuscript's very early ownership is unknown, but by the 16th century it can be identified as belonging to Fouke Dutton, a draper of Chester who died in 1558. It then seems to have passed into the ownership of the Bannester family of Chester and Caernarfon, and through them was in the possession of an Andrew Brereton by 1625; by the middle of the 17th century it had been acquired by Vaughan.
The Hengwrt Chaucer has been in Wales for at least 400 years, and recent research by English scholars suggests that Chaucer himself may have partly supervised the making of the manuscript, before his death in October 1400, according to the Welsh newspaper The Western Mail.
Peniarth MS 392 D contains 250 folios with a page size of around 29 x 20.5 centimetres. It is written on heavily stained and rather damaged parchment. The main textual hand has been identified with one found in several other manuscripts of the period (see below); there are a number of other hands in the manuscript, including one of a person who attempted to fill in several gaps in the text. This has been tentatively identified as the hand of the poet Thomas Hoccleve.
There is some illumination in blue, gold and red, used on the border and on initital letters at the opening of individual tales and prologues, but the manuscript contains no illustrations.
Scribe and relationship to other manuscripts
The lavishly illustrated Ellesmere manuscript, which for many years, following the examples of the editor Frederick Furnivall and W. W. Skeat, was used as the base text for more modern editions of the Canterbury Tales, is now thought to have been written by the same scribe, though the arrangement of the individual tales in the two manuscripts varies widely. Linne Mooney, a literary scholar at the University of York, has claimed to have identified this scribe as Adam Pinkhurst, the same Adam to whom Chaucer wrote a poem, admonishing him for his occasionally inaccurate copying skills.
Since the work of John M. Manly and Edith Rickert in compiling their Text of the Canterbury Tales (1940), the Hengwrt manuscript has had a much higher degree of prominence in attempts to reconstruct Chaucer's text, displacing the previously prominent Ellesmere and Harley MS. 7334. Recent scholarship has shown that the variant spellings given in the Hengwrt manuscript likely reflect Chaucer's own spelling practices in his East Midlands / London dialect of Middle English, while the Ellesmere text shows evidence of a later attempt to regularise spelling; Hengwrt is therefore probably very close to the original authorial holograph.
The manuscript is conventionally referred to as Hg in most editions giving variant readings.