Learning Wh Gaps

[[Classnotes MOC]]

[[Psycholing Classnotes]]

[[Learning Wh-Gaps, Utku Turk]]

Date:: 2022-02-06

Class:: Psycholinguistics II

Tags:: #umd #classnotes #umd/psylx

Were the results surprising?

• Locality matters
• At Issue, Not At Issue information apparently does not matter
• Usually people talk about locality in terms of structural locality, and why it matters. It is not the case here. Linear locality mattered more.

Put the frog on the napkin in the box. Children process on the napkin as a part of the verb “put”, rather than DP modifier. Verbs like “put” are very hungry when it comes to the location. With different verbs that require location less ambitiously, child do better.

!Paper! Developing incrementality in filler-gap dependency parsing, by Atkinson et al.

!Paper! Locality and Word Order in Active Dependency Formation in Bangla by Chacon et al.

Maybe there are some information structure related issues with Children? The structure used in English is so weird that participants might have a really weird time with an uncooperative speaker, which might not be the case in Japanese.

Also, the moment participants realized it is not monoclausal was really late in Japanese.

Input vs uptake

Do kids really hear what they are heard? Maybe they distort the input and have their own “uptake”?

But still, maybe kids still use the corpus, even though they distort it.

They do not allow “subcategorically” weird combinations, but they do allow re-statement of certain elements, like double locative?