Word, 1986, № 37:3, p. 141-151.
The purpose of this note is to serve as a word of caution to general linguists concerning the use of Welsh data. Not that there is anything uniquely peculiar about Welsh; it is clear the same situation holds for a number of other attested languages (for instance Classical Arabic; cf. Rabin 1955). But the lack of caution referred to is evident in studies using Welsh data and so Welsh may serve as an example for other languages where the same considerations are necessary.