“…That is, larger gender differences in vowel space expansion were observed in low-density words than in high-density words, and regional dialect variants representing changes-in-progress were more advanced in low-density words than in high-density words, in high-predictability words than in low-predictability words, in second mention words than in first mention words, and in plain laboratory speech directed toward an imagined friend or family member than in clear laboratory speech directed toward an imagined hearing-impaired or non-native listener. These patterns of increased use of stable forms, which have social indexical value, and of more advanced forms in a change-in-progress parallel the findings for phonetic reduction, in which more reduced forms are observed in low-density words than in highdensity words (Munson and Solomon, 2004;Scarborough, 2010;Wright, 2004), in high-predictability words than in lowpredictability words (Aylett and Turk, 2004;Bell et al, 2009), in second mention words than in first mention words (Baker and Bradlow, 2009;Fowler and Housum, 1987), and in plain laboratory speech than in clear laboratory speech (Ferguson and Kewley-Port, 2007;Kuo and Weismer, 2016;Picheny et al, 1986;Smiljanic and Bradlow, 2005). Thus, across a range of contexts that are known to elicit phonetic reduction, individual talkers have been shown to produce a parallel increase in their degree of social indexing, including more advanced forms of changes-in-progress.…”