Sentence Comprehension 2

Question 1

What are the context effects in the garden path model and constraint based model?

Answer 1

GPM = context is initially ignored, but has effect on later stages of processing.
CBM = context has immediate effect (but is not the only factor).

Question 2

What type of context do we deal with in this course?

Answer 2

Discourse context (particularly referential context) that are provided outside the target sentence.

Question 3

How can contact affect syntactic processing?

Answer 3

If presupposition is not satisfied.

Question 4

What is presupposition?

Answer 4

Information about modifiers and number of objects of the same kind in the discourse.

Question 5

What is the referential theory?

Answer 5

It explains how context may affect syntactic ambiguity resolution. It has later been incorporated into constraint based theories.
It predicts that the interpretation that contain the fewest unsatisfied presuppositions is the preferred one.

Question 6

Altmann and steedman (1988) made a 2x2 (referential context attachment) experiment. What did they find?

Answer 6

They found a referential context effect which is compatible with referential theory.

Question 7

Clifton and Ferreira (1989) argue against Altmann and Steedman (1988)’s findings. Why?

Answer 7

They argue that the results can also be explained with the garden path model.
Problems: Their self paced reading task did not distinguish between first analysis and reanalysis. Eye tracking. The GPM argues that the parser has access to the referential information at a later stage

Question 8

What does constraint based models say about context effects?

Answer 8

It’s just one of the many sources of information, it does not always have a strong effect.

Question 9

What does referential theory and garden path model say about context effects?

Answer 9

Context do occur. Context effect is not always as strong or or as early as referential theory claims.
Context makes the non-minimal attachment analysis easier, but not always easier than the minimal attachment analysis.

Question 10

What does garden path model say about frequency effect?

Answer 10

It is initially ignored during syntactic analysis

Question 11

What does constraint based models say about frequency information?

Answer 11

It is used immediately and relative frequencies of alternative usage s could affect ambiguity resolution.

Question 12

What types of frequencies are there?

Answer 12

Argument structure, syntactic categories and tense form. Some verb can take more than one structure, some words can have more than one syntactic structure and some words can take the same form for different tenses.

Question 13

In a self-paced reading experiment. What did Mitchell (1987) find? “After the child had visited/sneezed the doctor prescribed injections”

Answer 13

For “visited “ reading times for doctor was less than for sneezed(incompatible with constraint based models).
For “sneezed” reading time for prescribed was longer than visited (so reanalysis took place for doctor).

Question 14

What was the results from Mitchell (1987) about frequency information?

Answer 14

Frequency information does not provide a strong enough constraint to prevent processor from considering direct object analysis. It’s consistent with garden path model.

Question 15

Tests of architectures cannot test which two things?

Answer 15

Serialism in garden path model and parallelism in constraint based models.

Question 16

When testing architectures GPM and CBM argues what?

Answer 16

GPM argues reanalysis but it can causes processing difficulty. It occurs when initially adopted structure is incorrect.
CBM argues competition but that can also cause difficulty because two syntactic analyses are about equally activated.

Question 17

Who made the race model?

Answer 17

Van Gompel et al 2001

Question 18

Read Van gompel (2001) about race model

Answer 18

Understand what they concluded

Question 19

“While Anna dressed the baby that was small and cute spit up on the bed.” What do people think happened?

Answer 19

They think that Anna dressed the baby, that the baby spit on the bed. Even in control condition (with comma) people thought Anna dressed the baby (12%).

Question 20

What is the good enough representations theory?

Answer 20

Ferreira (2003) argues that people don’t process syntactic structure but construct a good enough in complete representation.
People fail to reanalyze.

Question 21

What does the good enough representations theory become less likely?

Answer 21

In active sentences.

Question 22

Write down the sentence used for the race model

Answer 22

The hunter killed the poacher/leopard with the rifle/scars

Question 23

What are the sentences for the good enough representations theory?

Answer 23

The house ate the baker.

The dog was bitten by the man.

Question 24

Are the findings of the good enough representations theory consistent or inconsistent with the GPM and CBM?

Answer 24

Inconsistent because the two theories assume that people construct complete syntactic representations.