General principles

The special status of names

Naming, although semantically a specialised function, in other respects forms part of the everyday language. Phonemic material has to be the same, and to follow dialectal and chronological paths that are related, albeit not invariably identical. The morpho-syntactic features of names must fit with general ones. Lexical material and modes of word-formation too must reflect those of the language at large. Indeed, place-names normally start as plain descriptions of the sites concerned: e.g. Kingston < cyninges tūn ‘the king's estate’, Pyrford < (æt) pyrigan forda ‘(the settlement at) the ford by the pear-tree’ (PN Surrey: 59, 132; illustrative examples will usually be given in normalised rather than documentary form). Personal names, although less transparently motivated, likewise ultimately derive from elements of common language.

Before becoming truly a ‘name’, a descriptive formation must, however, be divorced from its etymological meaning in such a way that the sound-sequence, no matter how complex its structure or plain its surface-meaning, becomes a simple pointer; ‘one might claim that unintelligible names fulfil their role more directly’ (Gardiner 1940; Nicolaisen, in Gelling et al. 1970:14). Bath, as a place-name, coincides in form with the common noun, and awareness survives of the Roman baths that it commemorates; but, for all that, the name's everyday ‘meaning’ is independent of etymology. Such independence is clearer still with names which, like London, have, since records began, apparently been opaque to their users (Rivet and Smith 1979:396–8).