One of the major achievements of Britsh linguistic scholarship before the 19th century was John Wilkins’ (1609–72) Essay towards a Real Character, and a Philosophical Language (1668), which attempted to construct, for scientific purposes, a language in which the elements were isomorphic with the categories of reality (as they were perceived by Wilkins). Immediately after its publication, the Essay was presented to the scientists of the newly-founded Royal Society for their critical appraisal. Since the committee appointed to examine it never reported, it has usually been assumed that they were uninterested or disapproving. It can now be shown, however, that it was certainly not lack of enthusiasm among Wilkins’ contemporaries that led to the absence of a report, and that three members of the original committee took part in a project to revise the Essay after its author’s death. It has long been known that a small group were informally engaged on its revision in 1678, according to a report of the antiquarian John Aubrey (1626–97), F.R.S., but hitherto nothing has been known of the enterprise. Recently, their correspondence has been discovered among Aubrey’s collection of manuscripts in the Bodleian Library, and these letters, besides showing links with the original committee, illustrate the growth of linguistic insight in the would-be improvers, particularly in respect of semantic classification and various problems in the phonetics of English. The course of their discussion is traced here, and the reasons for their eventual rejection of Wilkins’ scheme. Yet the immense undertaking was never wholly forgotten; it aroused the interest of at least one eminent 18th-century scientist, and became one source of inspiration for Peter Mark Roget (1779–1869), creator of the famous Thesaurus.