Realization of just: Speech reduction in a high-frequency word

Speech communication in conversational contexts often contains word forms distinct from careful speech, as speakers modify or omit segments in high-frequency words, producing reduced pronunciations (Aylett and Turk, 2006; Bybee, 2001; Fidelholz, 1975; Fosler-Lussier and Morgan, 1999; Hooper, 1976; Munson, 2007; Pluymaekers et al., 2005). The present study explores how the high-frequency word “just” in American English is realized with regard to its acoustic features, how the acoustic features of “just” and its reduction differ based on adjacent segment types, and how they vary depending on the predictability of word sequences. Results reveal that “just” is more reduced in duration and vowel articulation when followed by sibilants compared to vowels and other consonants. Preceding sibilants cause less fronting of the vowel’s backness but do not significantly affect duration or vowel height. Additionally, higher word predictability correlates with increased reduction, although specific acoustic cues are not affected by the word predictability. These findings emphasize the systematic, context-dependent nature of speech reduction and support its characterization as a gradient phenomenon (Ernestus and Warner, 2011) with the observed variability in the reduced form of “just.”