No effect of social priming on speech perception in Norwegian

The possibility of influencing speech perception through the priming of a social variable has been shown in studies with subtle and explicit priming methods. However, failures in replicating the effects in newer experiments have cast doubt over the phenomenon. This paper reports on the attempt of replicating subtle and explicit social priming in a Norwegian context, by attempting to prime two Norwegian linguistic variants. Eighty-four participants carried out a vowel matching task in which the vowel /ʉː/ was expected to be perceived differently depending on which condition the participant was randomly assigned to (Urban East Norwegian or Oslo-East). For explicit social priming, participants were told the origin of the speaker who uttered the sentences. For subtle social priming, different newspapers were used with the intention of subliminally priming the participant. The results were analyzed using Bayesian multi-level ordinal probit models. The models showed anecdotal to moderate evidence for the null hypothesis in the case of explicit priming and none to anecdotal evidence for the null for subtle priming. The results from our study cast doubt over whether social priming can affect speech perception. Claims on social priming regarding speech perception thus need careful empirical evaluation.