As one of the more striking contrapuntal devices of late thirteenth-and fourteenth-century polyphony, the hocket has been studied at great length. Until recently, however, little attention has been paid to the question how this texture accommodates the verbal text with which it often appears: as it breaks up a musical phrase into small segments separated by rests, it often necessitates a breaking up of the words themselves. Already some contemporaries had condemned this as one of the principal defects of the hocket, and most modern editors (implicitly or explictly agreeing with these comments) have sidestepped the issue by moving the text as much as possible to non-hocketed sections of the music. This article attempts to take a positive view of the ways in which words were matched to notes in the hocket. It distinguishes between 'unbroken hockets'-where the words and the notes are devised and deployed in such a fashion that word breaks and rests coincide-and 'broken hockets'. In the latter, it is argued that the way in which words are split up is by design and not happenstance; taking the etymological meaning of 'hocket'='hiccup' as a point of departure, it is shown that the most logical way to bridge the rests is not between syllables, but in the middle of syllables. This solution is corroborated not only by etymology, but also by the underlay in the sources, by the compositions themselves (in which by virtue of this 'hiccup underlay' musical phrases and text phrase match seamlessly, as elsewhere in the repertory) and by some theoretical evidence as well.