Lecture 7: Pragmatics

Question 1

mental model (6)

Answer 1

• Our conceptual representation of the real-world event evoked by a sentence.
• Every sentence contains references;
  
  • Referent: real-world object that’s being referred to (e.g. a box);
  • Referring expression: what you’re using to describe it (e.g. “the box”).
• Speakers can also use pronouns as referring expressions (e.g. “pick it up,” while pointing to a box).
  
  • With pronouns, the same referring expression can be used for different objects (i.e. words like “it” have no inherent meaning).
  • Pronouns are used to refer to things that are accessible in the mental model.

Question 2

factors that influence salience (6)

Answer 2

• Givenness: whether an entity was mentioned.
• Recency: when it was mentioned; recent entities → more accessible.
• Syntactic position in a sentence; subjects/1st entities/more prominent characters → more accessible.
• Gender: when gender info makes pronoun use ambiguous, a pronoun is less likely to be used for the less accessible character.
  
  • e.g. Tom invited Jim/Jennifer to go on a bike ride. He/She was planning to bring some snacks.
  • He is more likely to refer to Tom in the first sentence, but it’s obvious that She refers to Jennifer.
• Pronouns: What is the likelihood of using a pronoun in a particular setting?

Question 3

Arnold et al. (2000a) (5)

Answer 3

• Looked at how available entities are in the mental model, and how gender of S vs. O can create ambiguity, leading to different ways of resolving the ambiguity.
• Participants given sentences containing one or two genders, and had to continue the sentence.
• Measured whether participants would make the subject of the second sentence the 1st or 2nd entity of the first sentence, and if they would use a pronoun.
• Results: 1st entity is referred to with a pronoun more often than the 2nd entity regardless of number of genders.
  
  • Same gender → less likely to refer to the 2nd entity with a pronoun.

Question 4

Arnold et al. (2000b) (6)

Answer 4

• Subjects listened to miniature stories; tracked their eye movements to pictures that depicted the characters and objects involved.
• Results: we resolve the pronoun according to gender matching, and that the order doesn’t matter as much.
• Two different genders:
  
  • If gender info is used immediately (i.e. pronoun is subject) → look to the referent of the subject most.
  • If gender info isn’t used immediately (i.e. pronoun is object) → look to the referent of the object most.
• Same gender:
  
  • Upon hearing the critical pronoun that refers to the subject, listeners will look more to the subject because it’s more accessible.
  • The most interesting condition: the critical pronoun refers to the object. Listeners will actually look to the object most, however, it took longer to identify it when it was less accessible.

Question 5

MacDonald & Just (1989) (8)

Answer 5

• Do our mental models represent propositions or real-world situations (or both, to different degrees)?
• Participants completed a reading-based memory task.
• Sentences contained two nouns; three conditions: noun 1 negated, noun 2 negated, no negation (baseline).
  
  • e.g. Almost every weekend, Elizabeth bakes [no bread but only cookies / some bread but no cookies / bread and cookies] for the children.
• Sentences disappeared → word was presented; participants answer whether or not they saw the word in the sentence (e.g. cookies).
• Measured the latency of word recognition.
• Results: Noun 1 was negated → response times to noun 1 were greater; noun 2 was negated → response times to noun 2 were greater.
  
  • However, compared to the baseline, response times to both noun 1 and noun 2 were still greater even when noun 2 and noun 1 were negated, respectively. This could have something to do with the quantifying words used (some vs. only, and vs. but).
• Conclusion: A concept is activated more strongly when it exists in the real-world situation described by the sentence, as opposed to just in the sentence itself.

Question 6

maxims of cooperative conversation (1)

Answer 6

• A set of communicate expectations that are shared by speakers and hearers regarding how speakers typically behave in order to be understood by hearers.

Question 7

maxim of quality (1)

Answer 7

• If a speaker makes an assertion, they have some evidence that it’s true. Blatantly false statements are typically understood to be intended as metaphorical or sarcastic.

Question 8

maxim of relation (1)

Answer 8

• Speakers’ utterances are relevant in the context of a specific communicative goal, or in relation to other utterances the speaker has made.

Question 9

maxim of quantity (1)

Answer 9

• Speakers aim to use language that provides enough info to satisfy a communicative goal, but avoid providing too much unnecessary info.

Question 10

maxim of manner (2)

Answer 10

• Speakers try to express themselves in ways that reflect some orderly thought, and that avoid ambiguity or obscurity.
• If a speaker uses a convoluted way to describe a simple situation, they’re probably trying to communicate that the situation was unusual in some way.

Question 11

conversational implicature (2)

Answer 11

• A deliberate violation of the maxims.
• An aspect of the speaker’s intended meaning that can’t be derived directly from the linguistic code, but must be inferred by the hearer on the basis of expectations about the speaker’s probable communicative goals and behaviour.

Question 12

scalar implicature (2)

Answer 12

• A type of conversational implicature that occurs when a speaker chooses a relatively vague expression rather than a stronger, more specific one.
• In many contexts, the speaker’s choice of linguistic expression leads the hearer to infer that the speaker has used the weaker, vaguer expression because the stronger one would be inaccurate under the circumstances.

Question 13

Sedivy et al. (1999) (9)

Answer 13

• Demonstrated the maxim of quantity.
• Eye tracking study with two objects that contrasted and a baseline condition (no contrast).
  
  • e.g. “Pick up the tall glass,” with just one glass or two glasses of different height.
• When we’re processing scalar implicatures: when are contrastive inferences computed? Do we need the whole message to compute the implicature? Or do we do it immediately when we hear the scalar implicature?
• Measured where participants looked when they heard tall.
• If we process the scalar implicature immediately, we’d look at the entity (the taller glass) with contrast, as we assume that the maxim of quantity isn’t being violated.
  
  • If we look at the jug, then we’re not paying attention to the implicatures, because the jug is the tallest thing, so we’re interpreting the adjective literally.
• Results demonstrated that we process the scalar implicature immediately.
  
  • Upon hearing the adjective, most attention is to the object with contrast.
  • There are fewer looks to the competitor (jug) in the contrast (critical) condition), with most looks to the target.