Not everything that competes means something: evidence for competition among word-forms in a novel-word learning paradigm

The aim of the present study was to investigate whether learning a meaningless phonological word-form, can affect its ability to compete with other words shortly after it was learned. According to previous experimental work we expected that a semantic referent (Leach & Samuel, 2007), and/or consolidation over a significant amount of time (Gaskell& Dumay, 2003) are necessary for a novel word-form to be able to engage in lateral inhibition with other words. In order to examine this we used the experimental design that was used by Dahan, Magnuson, Tanenhaus and Hogan (2001). Experiment 1 was a replication of the Dahan et al (2001) study. In Experiment 2 we added a condition in which a novel word was now assigned the role of the competitor, by inserting a nonword learning task (that was performed right before the Dahan task). The goal was to see whether any differences would arise between this new novel-word condition and the nonword condition. The results from Experiment 2 were inconclusive due to the stimulus set and this is why we conducted Experiment 3, which was similar to Experiment 2, but had a different stimulus set. The results of Experiment 3 showed that, in contrast to the predictions, a novel word can compete with other words, even if it does not have meaning and, moreover, this happens immediately after training. These findings indicate that 1) a word does not have to be complete (i.e. include semantic information) in order to compete with other words and 2) connections between novel and known words can form faster than what has been suggested.