Spoiler alert: we don't think (most) people are. If this is the case, how should we think about what people are learning? How does this impact our understanding of current theories of L2 speech sound learning, which are predicated on categories being the target of learning?