7 thoughts on “Predicting binding domains: Evidence from fronted auxiliaries and wh-predicates

  1. This is a comment for the presented version! I’m very interested in the finding that people sometimes processed the pronouns in your experiment with a little bit of wait and see approach – that seems different than the greedier profile other studies with cataphoric pronouns have found. What do you all think of this contrast?

  2. So you’re referring to the pronoun+was cases (How proud of him was the woman/man…) where we failed to find an effect of gender (compared to the DID cases)?
    If so, yes that’s at the crux of it. Here’s a thought: the expectations that this is just a matrix gap are so strong that that’s enough for the pronoun to say “nope – antecedent is not gonna be here”.

    1. This would mean that the Gap Filling pressure is prioritized over catephoric dependency resolution. And therefore very different from what Ackerman found (as I understand it) when it comes to the relative priorities of cataphoric dependency formation and local coherence/V+object parses, which could be overturned in favour of the pronoun resolution.

      1. Ah – excellent point. This strikes me as a good way to think of it, and consistent with evidence from other types (the one that comes to mind is Keshev and Meltzer-Asscher’s work on timing in dependency formation for syntactic vs pragmatic dependencies)

  3. Super interesting stuff! Thanks for the great talk!

    I was wondering about the sentence completions that participants produced in the first study.
    (i) What kind of ungrammaticalities rendered a completion ‘ungrammatical’? Secondly, it seems that in the “did” condition participants provided ungrammatical conditions more often (if it’s 252 out of 300). Was this significant and if so, what would motivate such an effect, do you think?
    (ii) Were the ‘was’ (grammatical) sentence completions also compatible with readings that would violate Principle B? (e.g. How proud of him was the boy that he hit a homerun?) And if so, how often did this happen?

    1. These are great questions and we’re still trying to grapple with what we’re really seeing with the completion study.
      (i) the ungrammatical examples were typically those where no gap was provided (which is in itself interesting). There were more ungrammatical cases in the did condition and I think that’s because the WAS-cases are so easy to complete – you can just add a DP subject
      (ii) ok this was very interesting. A lot of the pronoun-WAS continuations sure looked like they were Condition B violations – but they could in principle have been introducing a new discourse reference. We can’t tell – but certainly given the gendered nouns provided, it wasn’t overwhelmingly different genders. This makes me wonder if the participants weren’t really engaging Principle B in their continuations. (But the SPR results seem to suggest they are).

Leave a Comment or Question Below