Affix priming with variable ING in English: Implications for unique vs. dual representation

Variation in the pronunciation of spoken words constitutes one of the primary challenges to theories of Spoken Word Recognition (SWR). In this paper we examine the processing and representation of a type of variation that is connected to morphology: variation in ING, which is found in words that vary between an -ing and an -in’ form. This variation, which is found in monomorphemes like awning in addition to affixed words, has been extensively studied, and has well-known social effects. Crucially, there is no consensus in the field as to whether the variation is morphological – involving distinct -ing and -in’ morphemes – or phonological in nature, with -in’ produced from an underlying -ing form. We connect the morphological and phonological analyses from the sociolinguistic literature to what have been called dual representation and unique representation in the SWR literature. We report the results of a series of experiments that use an auditory priming paradigm to explore the competing predictions of the dual and unique representation approaches. Priming provides insight into what types of representations are shared between the variants, which in turn informs the theoretical opposition at the center of the discussion about the locus of ING’s variation. The first of these experiments reveals priming both within and across ING variants, with significantly more priming found when both variants are -in’. Follow-up experiments manipulating the distance between prime and target, as well as introducing monomorphemes like awning, provide evidence that we interpret as favoring the unique representation view, with the -ing/-in’ alternation being phonological in nature. Alternative explanations are explored as well, with an eye towards the directions that future work on variation might take.