During the second half of the 1380s, John Gower, the leading fourteenth-century poet and an acquaintance of Geoffrey Chaucer, was working on his longest English poem, the Confessio Amantis. Chaucer, in turn, is believed to have been writing some of the material that would later form The Canterbury Tales. In addition, Chaucer was probably finishing Troilus and Criseyde, which must have been available before March 1388, at least in part, to Thomas Usk, the poet and undersheriff of London, who names the poem and borrows from Chaucer's Boece in his Testament of Love. 1 It is in the Confessio and in the Troilus, that is, in works written in the second half of the 1380s, that Chaucer and Gower first refer to one another in a literary context. 2 The only other instance that connects the two names is a 1378 legal record, in which Chaucer hands power of attorney to Gower and the lawyer Richard Forster. 3 Rather than reading this document as proof of Gower and Chaucer's supposed personal friendship, I adduce new evidence for Gower's legal training that suggests that the 1378 record was a purely professional arrangement-Chaucer might simply have needed a team of lawyers at the time. There is no reason to read this document through the prism of an instance of poetically embedded praise some ten years later, particularly given Gower's likely career as a lawyer. The new evidence for Gower's legal training changes what we know of the relationship between the two poets and, by virtue of pushing forward their literary acquaintance to the late 1380s, brings into sharp relief their deep ties to Southwark, where Gower may have resided at the time and where Chaucer launched his Canterbury Tales. Furthermore, on closer inspection, this new focus on Gower's and Chaucer's work in Southwark has the potential not only to foreground the role of an emerging literary culture in the area but to challenge existing models of read-I would like to thank the Warden and Fellows of All Souls College, Oxford, for electing me to a visiting fellowship in Hilary Term 2016. The generous support I have received from the college made this work possible. I am indebted to the editor of Speculum, Sarah Spence, and to the journal's two anonymous readers for their guidance and vital improvements to my argument. I am especially grateful to