This paper shows that the Aeolic inf. νηφέμεν used in Archil. 4.9 W instead of expectable νήφειν in the beginning of a pentameter is problematic: it cannot be a metrical variant, since Ionic νήφειν would be unimpeachable in the first part of a pentameter colon (where a contracted biceps is entirely admissible), and it cannot simply be put down as an epicism, since νήφω, a very colloquial verb, is never used in the heroic epic that came down to us and is very unlikely to have ever been used in this genre. The rare instances of Aeolic infinitives in -έμεν in the language of elegy can be shown to be dependent on epic models of one kind or another (e.g. γηρᾱσέμεν Simon. 20.7 W.2). Assuming that the form νηφέμεν is Archilochus’ own coinage, therefore, the question is whether the poet intended the final distich of fr. 4 to resonate with the epic tradition; and if so, whether it was a specific allusion to a fixed text or a reference to a broader tradition. It is argued that νηφέμεν should be viewed as modeled on πινέμεν from the famous midsummer picnic scene in Hesiod (Op. 592, construed with αἴθοπα οἶνον): the combination of the high-flown ending, best known from the epic dialect, with a lowly root νηφ- must have both produced a comic effect in the context of an invitation to get drunk on duty and serve as an allusion to the well-known passage in Hesiod.