ABSTRACT
Shveitser's judgement on Yakob Retsker was that his taxonomy of translation techniques was based too much on the lower linguistic levels of word and phrase, a criticism echoed by Delisle in relation to all other such taxonomies. The linguistic level at which the translation unit is set has a long and disputatious history behind it and is related, as Albrecht amongst many others points out, to literal translation and free translation. One attempt to extend the analysis of translation to this level has been to invoke generative-transformational grammar. Noam Chomsky proposed a new theory, called 'generative' or 'transformational grammar', which claimed that speakers generate the more or less complicated surface structures of sentences through a series of transformations of much more basic structures called 'kernels' or 'deep structures'. He also suggested that the elements that make up the kernels are universal, existing in all languages.
