By the somewhat shaky standards by which we can judge the popularity of any single medieval text, the Canterbury Tales was something of a bestseller. It survives for us in 84 fifteenth-century manuscripts as well as four early prints by William Caxton (1476/7, 1482/3), Richard Pynson (1492?) and Wynkyn de Worde (1498). I’ll discuss some of these witnesses in more detail as I go on, but if you’re interested in the specifics then you can find extremely thorough descriptions in Daniel Mosser’s freely available catalogue. It’s far from the most widely circulated Middle English text – the Prick of Conscience (a long penitential poem) survives in about 130 witnesses, and the two versions of the Wycliffite Bible clock in at over 250 – but its corpus is more than large enough to provide us with all sorts of useful clues as to how it was transmitted and who read it.

Now would be an opportune time to note what we don’t know. We don’t know what Chaucer ultimately intended for the work. When Chaucer died in 1400 he was either still writing the Tales or had given up on the project altogether, as he seems to have done with his Legend of Good Women and perhaps his House of Fame. The two earliest surviving manuscripts, which I’ll discuss in greater detail shortly, are clearly incomplete: there are stories missing, stories unfinished (most famously, albeit debatably, The Cook’s Tale) and the frame narrative makes no geographical sense. The order of the Tales varies wildly across the manuscript corpus, and not all later manuscripts seem to descend from the earliest surviving pair. As such we can’t speak with any certainty about how the Tales as we have them came to be, but there is a general consensus that a person or persons arranged, copied and circulated versions based on loosely arranged ‘fragments’ that were in the possession of Chaucer and/or his acquaintances when he died. In fact, we have no manuscripts of any work by Chaucer which were actually produced in his lifetime; the best we have is a reference to the sale of a lost copy of Troilus and Criseyde in Southwark in 1394 (at the very modest value of 20 shillings).

Whether Chaucer himself had a particular reader in mind is therefore not something that the manuscripts of the Tales themselves can tell us. But we do know enough about who he was circulating poems with during his lifetime to start to construct a picture – a picture in which a court circle with connections to the London urban middle classes looms large. Chaucer himself straddles this social border. He was the son of a successful merchant, a young servant (of the glamorous variety) to an aristocratic household, the husband to a well-connected wife (Phillippa de Roet) and an important public servant for many decades under three consecutive monarchs. He may not have been noble, but he had access to the court (and his son and granddaughter would climb the social ladder still higher).

In the course of all this, he wrote poems with implicit connections to the patronage of John of Gaunt (The Book of the Duchess) and Queen Anne (The Legend of Good Women), as well as a begging poem explicitly for the attention of Henry IV (‘conquerour of Brutes Albyon, / Which that by lyne and free eleccion / Been verray king’, The Complaint of Chaucer to His Purse). He dedicates poems to the poet and royal tutor Henry Scogan (Lenvoy de Chaucer a Scogan), the Yorkshire knight Sir Peter Buckton (Lenvoy de Chaucer a Bukton), the lawyer/poet John Gower and the lawyer/scholar Ralph Strode (‘O moral Gower, this book I directe / To the and to the, philosophical Strode’, Troilus and Criseyde). On top of this all Chaucer had connections to the so-called Lollard Knights like Sir Lewis Clifford, and through them a more nebulous relationship with some French poets including perhaps Deschamps, Froissart and Grandson. The fifteenth-century English monk/poet John Lydgate wastes absolutely no opportunity to insist that Chaucer was his friendly poetic “master”, as to a lesser extent does the bureaucrat/poet Thomas Hoccleve – scholars dispute these connections, but there’s no doubt that both read a good amount of Chaucer within a couple of decades of his death.

This was probably the sort of circle in which Chaucer intended the Canterbury Tales to circulate, a mixed aristocratic/upper-middle-class urban set with a good amount of money and a healthy interest in the patronage and production of literature. In Canterbury Pilgrimage terms, think of the Knight and the Squire, the extravagant Monk and Franklin, the scholarly Clerk, the comfortably well-off Merchant and Wife of Bath. Think of the Host! Harry Bailly was the very real owner of the very real Tabard Inn, as well as a many-time local tax collector, controller, coroner and – like Chaucer himself – sometime member of parliament. Thomas Spencer, who sold that 1394 copy of Troilus I mentioned, was also a Southwark innkeeper (and scribe, which may account for his possession of such a cheap book) and at the more “urban-middle” end of Chaucer’s contemporary audience. This class would only get more important in the fifteenth century (Lydgate was forever churning out commissions for guilds and merchants’ companies alongside his aristocratic requests) and Chaucer seems to have had his finger on that particular literary-mercantile pulse.

But let’s get back to the surviving manuscripts. Although they can’t tell us who Chaucer himself intended to read The Canterbury Tales, their varying quality and occasional medieval readers’ notes attest to a similar picture to that which I’ve already painted. The best place to start is with the early pair from which much of our modern idea of the “authoritative” Tales is drawn: San Marino CA, Huntington Library MS 26 C 9, the “Ellesmere Manuscript” and Aberystwyth, National Library of Wales, Peniarth MS 392D, the “Hengwrt Manuscript”. Both were copied in London in the first decade of the fifteenth century by the same scribe named (probably) Adam Pinkhurst, who may have been in Chaucer’s own employ and may in fact have been partly or wholly responsible for the earliest circulation of the Tales.

Hengwrt is earlier, more modest in size and quality of parchment, and is generally considered to represent an early stage in the attempt to assemble Chaucer’s fragmentary and incomplete poem into a coherent book. It doesn’t have a confirmed named owner until the rich Chester draper and sometime mayor Fulke Dutton in the first half of the sixteenth century. However, Pinkhurst’s copying work is supplemented by five other scribes (sharing work was a common practice for large texts copied in London in this period), one of whom looks to have been Thomas Hoccleve and one of whom copied Henry V’s presentation manuscript of Hoccleve’s Regiment of Princes. Even if Hengwrt was never meant as anything more than a modest early attempt at coherent compilation, it must have had some early owners/readers – and those readers very probably had some sort of personal connection to the Chaucer/Pinkhurst/Hoccleve/wider court set.

Ellesmere is more straightforward – it’s one of the nicest copies of the Tales in existence. It’s all in Pinkhurst’s best hand, it’s expertly decorated, it’s accompanied by the very famous pilgrim portraits and another artist was brought in especially to add the picture of Chaucer himself. This is obviously a deluxe manuscript for a rich urbanite and/or aristocratic owner, and it’s become an almost Platonic ideal of the early fifteenth-century Tales in Chaucer scholarship. The look of the manuscript is borne out by readers’ notes from the first half of the fifteenth century which connect it with the influential London-Norfolk Paston family and the closely related de Vere family (through whom there is perhaps a connection to Chaucer’s son Thomas). I should stress that while we have a uniquely clear picture of Ellesmere’s early production and reception, not least because so much critical interest has been lavished upon it, it’s by no means a singular example of its kind. Plenty of the higher-end copies of the Tales, such as the second-quarter-of-the-fifteenth-century Oxford, Bodleian Library MS Bodley 686 (which carries the most idiosyncratic Chaucer Portrait), would have been a good fit for this class of book owner.

What about the other end of the scale? I’ll illustrate with a manuscript I know well, Oxford, Trinity College MS 49. It dates to around the third quarter of the fifteenth century and is a decidedly modest production (I once heard the eminent manuscript scholar Richard Gameson describe it as the worst-looking copy of the Canterbury Tales that he had ever seen). It’s a fairly large book, as a full copy of the Tales necessarily is, but otherwise cheaply made: it’s on rag paper of mixed stock, it’s unillustrated, headings and capitals are in simple red and blue, there are odd portions of text missing or repeated, it’s patchily corrected, there are some clanging errors both by the scribe and by the painter of the blue capitals, etc. As books of the period go it’s fine and perfectly readable (confused text aside), but clearly an object of more utility than beauty.

Continued below.