Psycholinguistics part 4 Flashcards
Pragmatics
Mental models
- Detailed conceptual representations of real-world situations the sentence evokes
- We situate the linguistic form in mental models, and we retain more information from mental models than the exact linguistic form
- Not fixed or static: can incorporate spatial and temporal information
- Can affect how active/accessible words are in our memory
Negation in mental models, MacDonald & Just (1989)
Participants read a sentence, the sentence is hidden, and they must judge whether a word displayed on the screen appeared in the sentence or not. Some words in the sentence are negated and some are not, i.e. ‘Elizabeth baked bread but no cookies’
Results: Reaction times were slower for words which were negated, indicating that those words are less active in memory
Mental models vs. linguistic content, Bransford et al., (1972)
Findings: variation in the linguistic form of a sentence is more likely to be recognized if the mental model the sentence produces is also different
Factors affecting pronoun resolution
Referents for pronouns are determined by:
* Real-world knowledge / context
* Syntactic prominence (subjecthood)
* Gender
* Givenness
* Recency
* Syntactic constructions
Arnold et al. (2000)
Visual world eyetracking
Major findings:
* When at least one of the cues (syntactic prominence or gender)
matched the pronoun referent, they were able to correctly identify
the referent
* People can simultaneously and rapidly use multiple cues (i.e.,
syntactic prominence and gender) in real-time to resolve ambiguity,
UNLESS there was pronoun ambiguity (no gender distinction) AND
the referent is the less prominent entity (the non-subject entity)
Repeated name penalty
The difficulty in reading a sentence where a highly salient referent is
referred to by a full noun phrase than by a pronoun
Focus constructions
These constructions can change the accessibility of a referent during pronoun resolution.
Ex. It-cleft sentence: ‘It was the bird that ate the fruit. It was hungry’ -> the bird is more accessible
Ex. Wh-cleft sentence: ‘What the bird ate was the fruit. It was rotten’ -> the fruit is more accessible
Grice’s conversational maxims
- Quality
- Quantity
- Relevance
- Manner
Contrastive inferences, Sedivy and colleagues (1999)
Participants’ gazes are tracked as they hear ‘pick up the tall… (glass)’ and view 1 of 2 displays: 1. a display with a glass, a pitcher, and two distractor items, or 2. a display with a glass, a shorter glass, a pitcher, and a distractor item
Results: Upon hearing the adjective (tall):
* No contrast: higher incidences of looks to the competitor (pitcher) than to target; looks to target occurred considerably later
* Contrast: people are faster to attend to the target (tall glass) than to the competitor (pitcher)
Major findings: Information about the objects used with knowledge of the contrastive function of the adjective; incremental use of contrastive inferences based on the maxim of quantity
Referential communication
task, Clark & Wilkes-Gibbs (1986)
An experimental task in which speakers refer to a specific target object in the context of a number of other objects
Findings: The mean number of words used by the director and the turn-taking between the director and the matcher declined from trial 1 to trial 6 -> Joint effort between the director and the matcher (the conversational partners) to establish common ground (information that is accepted by both interlocutors)
Aspects of the common ground, Clark (1996)
- Linguistic co-presence: what both interlocutors have heard or spoken previously
- Physical co-presence: what both interlocutors can see or hear
- Community membership: what interlocutors know because they
belong to certain groups
Isaacs & Clark (1986) on community membership
Referring expressions used will depend on community membership and mutual awareness of community membership (i.e. expert versus novice)
Conceptual pact
Tacit agreement on a linguistic expression for a particular referent between interlocutors over the course of a communicative exchange