Variation in Word Duration and Planning
Alexandros Christodoulou · Carolina Digital Repository (University of North Carolina at Chapel Hill) · 2019
Word duration varies as a function of predictability (e.g., Bell et al., 2009; Lieberman, 1963; Fowler & Housum, 1987; Gahl & Garnsey, 2004; Watson, Arnold, & Tanenhaus, 2008) and planning difficulty (e.g., Bell et al., 2003; Clark & Fox Tree, 2002; Fox Tree & Clark, 1997). Both findings suggest that planning is related to word duration variation. However, the role of planning in word duration variation is debated (e.g., Arnold & Watson, under review; Ferreira, 2007). In four experiments I tested the hypotheses that planning an upcoming word and variation in the timing of planning an upcoming word leads to word duration variation of an utterance-initial word. Speakers named two pictures while trying not to pause. The utterance-initial word was long in duration when followed by a low frequency word. Furthermore, looks to the right object modulated the effect of frequency on word duration. Early looks led to word reduction when the second word was high frequency. Late looks led to long word duration regardless of frequency. The results support the role of planning on word duration. Theories of utterance planning should consider evidence of word duration variation apart from the traditional speech onset time and utterance duration measures.