Subphonemic Variation and Lexical Processing: Social and Stylistic Factors

Abstract

Different pronunciation variants of the same word can facilitate lexical access, but they may be more or less effective primes depending on their phonological form, stylistic appropriateness, familiarity, and social prestige, suggesting that multiple phonological variants are encoded in the lexicon with varying strength. The current study investigated how subphonemic variation is encoded using a lexical decision task with cross-modal form priming. The results revealed that the magnitude of priming was mediated by stylistic and social properties of the auditory primes, including speaking style, talker dialect, and duration. These interactions provide evidence that phonetically reduced forms and forms that are not socially prestigious are not as robustly encoded in the lexicon as canonical forms and forms produced in prestigious varieties.

Relative Contributions of Social, Contextual, and Lexical Factors in Speech Processing

“…Thus, dialect familiarity was not the primary source of the observed processing difference across stimulus dialects. Rather, this result is consistent with previous work on lexical processing of these dialects by these populations (e.g., Clopper & Bradlow, 2008;Clopper et al, 2010) and suggests a mismatch in production and perception, in which a processing benefit for the ideologically more standard Midland dialect is observed even among those who may use the Northern dialect in their own speech (Benson, 2005;Clopper & Bradlow, 2008;Jones & Clopper, 2019).…”

“…For trials with phrases produced by Northern talkers, semantic predictability was found to influence speech processing, such that high-predictability keywords corresponded to 46 ms faster response times than low-predictability keywords (Node 7), as expected. This interaction reflects a sub-additive effect, in which the predictability effect was limited to Northern stimulus materials, similar to the sub-additive interactions that Clopper (2012) and Jones and Clopper (2019) observed for semantic predictability and speaking style, respectively, and regional dialect.…”

“…For trials with low-frequency targets, stimulus dialect was an important predictor of Midland listeners' response times, with 36 ms faster response times associated with keywords that followed Midland primes than keywords that followed Northern primes (Node 9). This pattern suggests that for these Midland participants, the stimulus dialect effect was limited to low-frequency keywords, similar to the sub-additive interactions that Clopper (2012) and Jones and Clopper (2019) observed for semantic predictability and speaking style, respectively, and regional dialect.…”

“…For example, Goldinger et al (1989) observed that the processing advantage for high-frequency words relative to low-frequency words was larger for low-density words compared to high-density words. Focusing on social and contextual factors, Jones and Clopper (2019) found that the effect of speaking style on form priming in a lexical decision task was larger for a familiar dialect than for an unfamiliar dialect. Similarly, Clopper (2012) found a larger semantic context effect for familiar dialects than for unfamiliar dialects.…”

“…A processing benefit for familiar regional dialects over unfamiliar regional dialects has been demonstrated in a number of studies (e.g., Adank et al, 2009;Clopper & Bradlow, 2008;Floccia et al, 2006;Labov & Ash, 1997). In lexical decision experiments, response times are slower by 10-120 ms for unfamiliar regional dialects compared to familiar local or standard dialects, particularly in non-ideal listening conditions (Dufour et al, 2007;Floccia et al, 2006;Jones & Clopper, 2019;Sumner & Samuel, 2009). In speech intelligibility tasks, listeners more accurately identify words and sentences produced in familiar dialects than in unfamiliar dialects (Clopper & Bradlow, 2008;Clopper et al, 2010;Clopper & Wagner, 2019;Jacewicz & Fox, 2012;Labov & Ash, 1997;Smith et al, 2014;Wright & Souza, 2012).…”