Listeners rely on prosodic cues to disambiguate syntactic structures. One such ambiguity arises from how nouns are grouped in a sentence. Grouping nouns together as compounds compared to non-compounds should result in temporal adjustment within the word. We investigated how speakers disambiguated the two types using temporal planning, and how these temporal cues were exploited during perception. As expected, compounds showed shorter durations than the noncompounds, with the first word of compounds being shorter than in non-compounds. Compounds were also recognized faster than non-compounds in an eye-tracking task, suggesting a close link between production and perception.