Retroflex variation and methodological issues: A reply to Simonsen, Moen, and Cowen (2008)

We argue that the differences in the articulation of Norwegian retroflex consonants described by Simonsen, Moen, and Cowen (2008) as individual variation may instead be due to factors such as individual and dialectal background, rather than variation across a single variety. Our main argument is based on existing dialect literature and speech corpus data, which show that the phonemes involved in the retroflexion process are not present in the same linguistic contexts in all dialects. SMC's experimental stimuli and conditions include linguistic contexts which do not necessarily induce retroflexion naturally, and therefore cannot be relied upon to provide an accurate picture of retroflexion in natural speech contexts. The peculiar retroflex pronunciation that can be heard in some of SMC's sound files may be due to the invasive intra-oral equipment or to the unnatural retroflexion contexts introduced by their stimuli.