Presented By: Department of Linguistics
Linguistics Colloquium
Meredith Tamminga, University of Pennsylvania
Join us for a colloquium presented by Meredith Tamminga (University of Pennsylvania) in NQ2435 on Friday, April 3, 2026, from 4-5:30 PM!
Title: Sociostylistic variation and the recognition of spoken complex words
Abstract: Variation in word form is a major challenge for models of spoken word recognition. In this talk I present a series of primed lexical decision experiments about how listeners recognize English words containing the sociostylistic variable ING (e.g., workin’ ~ working). First, we cross the two variants in an affix priming design (e.g., jumpin’/jumping > thinkin’/thinking). We find that both -ing and -in’ variants robustly facilitate themselves and each other: jumpin’ does indeed facilitate both thinkin’ and thinking, as does jumping. The strongest facilitation arises when both prime and target contain -in’, but we do not detect differences in facilitation strength between the other three conditions, contra the predictions of accounts rooted in prime/target similarity or prime canonicality. The extra boost in the -in’ > -in’ condition is lost when an unrelated trial intervenes between prime and target, whereas general affix facilitation remains robust. We then move away from the affix priming design to show that when the prime is a monomorphemic noun containing ING (e.g., awning) and the target is a progressive verb, then e.g. awnin’ facilitates thinkin’ but awning does not facilitate thinking. This asymmetry is not simply a property of the surface phonological strings: we do not find facilitation from e.g. dolphin to thinkin’. Finally, to disambiguate between phonological derivation and stylistic incongruency as possible sources of the -in’ asymmetries, we repeat the first experiment with Southern-accented stimuli intended to improve the expectedness of the -in’ variant in the experimental context. Under these conditions, we no longer find significant evidence for the -in’ boost, suggesting that listeners’ sociostylistic expectations modulate the priming effects. I discuss the implications for the grammatical locus of ING variation, for how we study sociolinguistic variation in the lab, and for the incorporation of variation into models of spoken word recognition.
Title: Sociostylistic variation and the recognition of spoken complex words
Abstract: Variation in word form is a major challenge for models of spoken word recognition. In this talk I present a series of primed lexical decision experiments about how listeners recognize English words containing the sociostylistic variable ING (e.g., workin’ ~ working). First, we cross the two variants in an affix priming design (e.g., jumpin’/jumping > thinkin’/thinking). We find that both -ing and -in’ variants robustly facilitate themselves and each other: jumpin’ does indeed facilitate both thinkin’ and thinking, as does jumping. The strongest facilitation arises when both prime and target contain -in’, but we do not detect differences in facilitation strength between the other three conditions, contra the predictions of accounts rooted in prime/target similarity or prime canonicality. The extra boost in the -in’ > -in’ condition is lost when an unrelated trial intervenes between prime and target, whereas general affix facilitation remains robust. We then move away from the affix priming design to show that when the prime is a monomorphemic noun containing ING (e.g., awning) and the target is a progressive verb, then e.g. awnin’ facilitates thinkin’ but awning does not facilitate thinking. This asymmetry is not simply a property of the surface phonological strings: we do not find facilitation from e.g. dolphin to thinkin’. Finally, to disambiguate between phonological derivation and stylistic incongruency as possible sources of the -in’ asymmetries, we repeat the first experiment with Southern-accented stimuli intended to improve the expectedness of the -in’ variant in the experimental context. Under these conditions, we no longer find significant evidence for the -in’ boost, suggesting that listeners’ sociostylistic expectations modulate the priming effects. I discuss the implications for the grammatical locus of ING variation, for how we study sociolinguistic variation in the lab, and for the incorporation of variation into models of spoken word recognition.
Explore Similar Events
-
Loading Similar Events...
