1. What is the question being asked?
What is the right characterization of the comprehension abilities of agrammatic patients?
———————————————————————————————————————————————
2. What prediction(s) does the hypothesis being tested make?
Linear assignment. Hypothesis: the assignment of thematic roles is over linear sequences of nouns and verbs. Predictions: no effect expected because of embedded, only as a result of the order of NPs around the verb. 1 and 3 should show normal performance.
Trace deletion.Hypothesis: the S-structure in agrammatism does not contain traces and their associated indices. Prediction: assignment of thematic roles to NPs is done normally then the NP Is in a thematic position, and by heuristic otherwise. Prediction: change performanc in object gap relatives (2 and 4).
Complexity.Hypothesis: the more complex the constructions, the more likely it is that agrammatics will fail. Prediction: coupled with complexity metric, it can predict which stimuli will be harder, but it cannot predict the actual performance level. Order of difficult from less to more difficult: 2, 1, 4, 3.
———————————————————————————————————————————————
3. What are the experimental conditions (independent variable(s)) being tested (both linguistic and neurological)?
Four types of relative clauses with a 2x2 design: embedding type (center vs right) and location of gap (subject vs object position).
1. The boy who is pushing the girls is tall – center embedded, subject-gap relatives
2. The boy who the girl is pushing is tall – center embedded, object gap relatives
3. Show me the boy who is pushing the girl –right embedded, subject-gap relatives
4. Show me the boy who the girl is pushing – right embedded, object gap relatives.
Neurological: agrammatic patients with lesion in Broca’s area and controls as defined by Mohr’s 1976.
———————————————————————————————————————————————
4. What are the linguistic constructions being tested? What does each contribute to the exploration of the question? (Why do they use these constructions?
See 3 above. They use these constructions because it helps dissociate the results among the three different hypotheses states in 4 below.
———————————————————————————————————————————————
5. What is the kind of measurement (dependent variable(s)) being used?
Proportion of errors
———————————————————————————————————————————————
6. In two sentences, what is the outcome reported?
The patients perform above chance on both types of subject gap relatives and at chance in the object gaps.
———————————————————————————————————————————————