Following a flurry of interest in Romano-British placenames after the publication of Rivet and Smith’s book in 1979, there has been little further published work on the subject. This is in many ways surprising given the smallness of the corpus and the comprehensive linguistic notes which form Part Two of their work. The best overall treatment of the subject remains Margaret Gelling’s (1978) Signposts to the Past, which first appeared before the publication of Rivet & Smith’s corpus.
Her analysis divides the names into four basic types:
an adjective (with optional suffix)
a topographical term (with optional suffix)
an adjective plus topographical term (with optional suffix)
a river-name (with optional suffix)
She notes that topographic names predominate, with the order of elements generally being adjective + noun, rather than noun + adjective as in modern Welsh (although it is not strictly correct to treat Welsh placenames as compounds: Jackson 1953, 225).
The corpus is small: Rivet and Smith list 453 names which they consider genuine, although I have argued that they have unnecessarily reduced the number of names given by the Cosmographer; when these are taken into account together with a few names they were not aware of or which have been discovered since 1979, the total reaches 471. In addition, 22 more names can reasonably be inferred from existing place-names (such as *Cunetio, the River Kennet, from Cunetione, Mildenhall), bringing the total to almost five hundred.