2 thoughts on “New neighbours make bad fences: Form-based semantic shifts in word learning

  1. David, concerning your speculations about how this effect happens over time, I’m wondering about the relative contributions of comprehension vs. production processes? Your speculations seemed very comprehension-focused, but we do see form based production errors like saying “cavity” for “clavicle” and so on, which seems to suggest a production role.

  2. That’s a great question, Maryellen; thanks. You’re right that the experiments emphasize comprehension, as would be implied by a mechanism involving phonological priming / co-activation of newly learned and existing words. Off the top of my head, I’d suppose that production errors would more likely result in formal shifts (e.g., associating cavity with carving and so pronouncing it “carvity”), given that a substitution error would be more akin to total confusion than a shift. However, a combination of semantic and phonological overlap can and does result in speech errors, and that might tell us something about the co-activation and interference of new and existing words. I’ll certainly give some more thought to the production side of things!

Leave a Comment or Question Below