3 thoughts on “Weeks 12 and 13 comments and questions

  1. David Erschler Post author

    I am afraid I do not understand how a (the?) correspondence is established between Shepard types and specific artificial language patterns.

    This might well be completely trivial, but without this correspondence being made explicit, the entire narrative seems to become incomprehensible.

    Reply
  2. Joe Pater Post author

    This is an excellent question. It’s only in the simplest cases that it is trivial. For an example, see the translation of Shepard types to phonological inventories in (8) of the Pater and Moreton paper. But in cases like those in the Moreton and Pater reading, the translation often involves some assumptions. For example, the Pycha et al. case in (2d) is categorized as a three feature problem by taking into account the features of the suffix. But you might also say that the task is to use the suffix as a label, and that it’s features are irrelevant. It would then be Type I vs. II.

    Reply
  3. Ivy

    I’m wondering about the order of parameters given in (16) of the Dresher paper. It is unclear to me where this order comes from. The process of acquisition “must” happen this way – is that to mean it always happens this way? Or just that it has to go this way for learning to take place? If it’s the latter then how does the order happen?

    Reply

Leave a Reply

Your email address will not be published. Required fields are marked *