The model of the phonological component of a generative grammar that we’ve been developing to this point can be seen to consist of two parts. First, a set of underlying representations (URs) for all the morphemes of the language and, second, a set of phonological rules which determine the surface forms (i.e. the phonetic forms) for these underlying representations (see Section 8.3). So, for a word like ‘pin’ the UR, i.e. the phonological information in the lexicon, might be /pn/. To this form, rules like ‘Voiceless Stop Aspiration’ (see Section 3.1.3) and ‘Vowel Nasalisation’ (see Section 4.3) will apply, giving the phonetic form (PF) [ph˜n]. As in this simple example, underlying forms may be affected by more than one rule; the series of steps from UR to PF is known as a derivation, which is the concern of this chapter.