9.012  Cognitive Science

another clarification about Gibson language test

question 7c says the following:

Hasson et al (2004) observed that the activity in many regions in participants’ brains as measured by activation in fMRI are correlated when listening to stories. Would you expect the same regions to be as correlated during a resting state (no task)? Why or why not? Which regions are responsible for the high correlations during story comprehension? Provide evidence for your claim.

Idan didn’t end up talking about this particular issue in his lecture (which I have discussed in previous years). I have uploaded Blank & Fedorenko (2017) and Hasson et al 2004 to make this question more answerable.

Announced on 21 May 2017  11:06  a.m. by Edward A Gibson

clarification about q 9b

There are two hypotheses that one might consider here:

(1) that there are optional orderings in the syntactic rules in the language, which drive the observed effects. For example, many verbs allow two orders of their post-verbal dependents:

(a) I talked to John about Mary
(b) I talked about Mary to John

Maybe the dependency length differences come about because languages minimize dependency lengths in these optional situations. E.g., as in:

(c) I talked to John about the problem that was trying to understand.
(d) I talked about the problem that was trying to understand to John.

(c) would be preferred over (d) under this formulation.

(2) that the reason for dependency length minimization comes about from differences in rigid rule orderings within a language. E.g., that a language has both verbs before complement clauses, and complementizers at the beginning of these clauses, or both at the end. But there is no optionality in any rules.

Announced on 21 May 2017  12:10  a.m. by Edward A Gibson

Test 3 is now available

due Sunday at midnight.

email Ted if you have questions.

Announced on 17 May 2017  11:03  p.m. by Edward A Gibson