EXP3604 SPRING 2008

EXAM #3 SHORT ANSWER (ANY FOUR)

1. State how one of the approaches to concept representation (e.g., definitional, exemplar, etc) explains what makes some things members of the category, and others not, and one finding that suggests that this is what we do.

e.g.: Prototype; we compare the instance to an “average” or ideal member of various categories. One finding is the “typicality” effect, where instances rated as more typical of a category are verified more quickly (a robin is a bird), primed more (ROBIN / bird lexical decision), given as associates more often, etc. Exemplar: compare to various instances or examples in memory. One finding is that we can categories “exceptions” like ostriches, or bats, into appropriate categories.

[Part credit for describing an approach but not giving some evidence for it; or for describing one approach but calling it another (common for prototype/exemplar distinction).]

2. Collins & Quillian (1969) compared the time needed to verify sentences like “a canary is yellow” versus “a canary has wings.” Which of these was faster, and how did they interpret this outcome?

“A canary is yellow” is faster; the claim was that traits that are unique to that level (canary) are directly associated with that concept, but traits that are common to a superordinate category (BIRD) are stored at the most general level possible – “cognitive economy.” So we have to “infer” that canaries have wings by accessing two levels of propositions, and that takes longer. [Some said wings would be faster because it’s available in the image of a canary, or something along those lines; part credit]

3. According to the results of Swinney’s (1979) study of priming of ambiguous words like “bug” by sentence contexts that suggested one of its meanings (e.g., SPY), what happens as we encounter such words?

As the ambiguous word (bug) is heard, both meanings are apparently “automatically” activated, since we get priming for visually presented words related to either meaning (SPY, or ANT). But very quickly (within a second after BUG is heard – 700 ms actually), the contextually appropriate meaning is selected – another example of the “immediacy” principle – and only the appropriate meaning (SPY in this case) is still “primed.” [Similar to the Neely study of the role of expectancy in priming with single words, we have different outcomes at the different timing intervals between the prime word and target. Part credit for giving one of the two outcomes, or for suggesting that the context that follows the ambiguous word determines the selection.]

4. What is the relation between the “thematic importance” of a fact in a text passage, and the probability of its recall, as a function of the coherence of the text passage?

When we build a “propositional model” of a passage, some facts directly concern the “theme” of the passage, others are less directly related details. Recall is better, the closer the fact is to the main theme. But if the passage is “incoherent,” it’s hard to build that model; we don’t understand what’s an important fact and what’s not, and the importance effect can disappear, with overall recall lower as well. [The key phrase here is “as a function of coherence,” which asks about the “interaction” of importance and coherence. Part credit for correctly describing the importance effect only, or for treating importance and coherence as two independent effects, without stating that the effect of one (importance) depends on the level of the other (coherence) – that’s an interaction.]

5. What is one way we’ve demonstrated the “psychological reality” of the syntactic structures described by linguists (where syntactic form or complexity affects performance)?

Several possibilities from lecture or text: garden path sentences, where we’re led to parse one structure but it turns out to be another (The horse raced past the barn fell), causing disruptions in reading; “chunking” studies, where rapid presentation that conserve phrase boundaries (The man / who walked / was old) are perceived better; the “click” studies, where perception of a click seems to “migrate” to phrase boundaries; the contrast between “subject relative” and “object relative” sentences (The senator who married /who the secretary married) in brain activity and reading speed. [Part credit for a description of a semantic factor or effect, as in Kutas’ ERP study of semantic anomaly.]