HOW DO WE USE LANGUAGE?
✏ D+L2 LECTURE 4

Direct and Indirect speech

Searle (1975); Gibbs (1986)

  • Direct speech acts (<10%)
  • Literal meaning intended
  • Indirect speech acts (>90%)
  • Need to move beyond the literal meaning

Non-literal speech acts:

  • Idioms --> "kick the bucket"
  • Metaphor / simile --> "She's a brick"
  • Irony --> "What an exciting lecture, Professor Gaskell"

"British Politeness"

How do we understand indirect speech?

Searle (1979)

  • Two-stage mechanism: literal meaning first and then non-literal interpretation

Gibbs (1989)

  • One-stage: parallel literal and non-literal, or only non-literal
  • Mixed evidence - depends greatly on the context and the type of indirect speech
  • “Kick the bucket” probably gets a lexical entry of its own (Swinney & Cutler, 1979)
  • Subtle irony may take several seconds to detect (if at all)

Scalar Implicature

"Some tuna are fish"

  • (1) Has semantic meaning (at least one tuna is a fish - true)
  • (2) And a pragmatic meaning (some tuna are fish and some are not - false)
  • “Have the cakes been eaten?”
  • “Some of them have”
  • Most people opt for the pragmatic interpretation (i.e. some tuna are fish is FALSE)
  • But is there competition between the two meanings?

De Neys & Schaeken (2007)

  • Asked participants to verify statements while manipulating cognitive load (maintain patterns in memory)
  • Difficult / high load led to fewer pragmatic responses
  • Competition between automatic semantic interpretation and more resource-dependent pragmatic interpretation?

Conversational Structure

Grice (1975)

  • Co-operative principle, 4 maxims:


    (1) Quantity: Make your contribution as informative as required and no more.


    (2) Quality: Make your contribution true.


    (3) Relation: Make your contribution relevant to the aims of the conversation.


    (4) Manner: Be clear.


  • Like grammar, another mutually agreed convention?

  • Co-operation between speaker and listener to maximise information flow.

Evaluating Grice


Linguistic contribution

  • Useful for description of conversation
  • Also identification of violations (e.g. jokes, political interviews)
  • Good example - Jeremy Paxman interviews Michael Howerd
  • Same question asked 12 times (“Did you threaten to overrule him”)
  • Clear violations of many maxims (although perhaps not quality)

Psychological contribution

  • Several offshoots of Grice’s work now research areas in their own right.
  • Issue of collaboration/common ground
  • What does the speaker know about the listeners task/mental representation (and vice versa)

Can speakers avoid ambiguity?

Ferreira et al. (2005)

  • Pairs of participants, one has to name objects on the screen, the other has to mark the objects on paper form.



    Both see the same pictures in three conditions:

  • Non-linguistic ambiguity
  • Linguistic ambiguity

  • No ambiguity



    Speakers are good at avoiding non-linguistic ambiguity.

  • Evidence of common ground



    Less good at avoiding linguistic ambiguity

  • Often unaware of a problem.
  • Number represent the order in which the speaker has to name the object.

Can listeners read speakers' minds?

Keysar et al. (2000)

  • Two participants:
    • Director (confederate)
    • Addressee (naive participant) - eyes tracked.

  • Participants informed that some shelves in the grid are blocked from view of the director.

  • Director performs set of instructions


    • E.g. “put the truck one slot down”
    • Key instruction: “Now put the small candle above it.”

  • Addressees generally picked the appropriate object


    • Suggests a representation of common ground

  • But their initial response was often egocentric


    • Sometimes hand movements towards occluded objects
    • Many fixations toward occluded objects

  • Simpler egocentric interpretation needs to be inhibited?


Cultural differences in Perspective Taking

Wu & Keysar (2007)

  • Are some cultural groups better at seeing the other person’s point of view?
    • Tested two groups
      • Collectivist - Chinese
      • Individualist - American
  • Chinese listeners showed better ability to determine speaker’s point of view - even in initial eye-movements.
    • But are there other ways of explaining the data?

How far does mind-reading go?

  • Speakers and listeners read each others’ minds.
  • But what about larger groups? (e.g. friends, lecture)
    • Do we still read each others’ minds in larger groups?

Social N400

  • We can read co-listeners’ minds
    • At least when directed to
  • Conversation involves tracking common ground in terms of multiple participants.

Is it all about mind-reading?

  • Individual differences:
    • ASD individuals often have problems with pragmatic inference:
      • A: Would you like some chicken soup?
      • B: I am a vegetarian
  • Fits with the “Theory of Mind” deficit interpretation
  • We can assess mind reading explanations using these groups.


Pijnacker et al. (2009)

  • (1) "All sparrows are birds"
  • (2) "Some sparrows are birds"
  • (3) "All birds are sparrows"
  • (4) "Some birds are sparrows"
  • Tested high function ASD adults and controls.
    • Groups generally well matched on performance.
    • Interpretation in this case may not rely on mind reading skills.
      • Might be a more automatized process?

Conclusions

  • Language understanding involves going beyond the literal meaning
    • Particularly difficult in ASD
  • In a conversation, the participants establish common ground.
    • Speakers and listeners attempt to read each others’ minds, with varying success.
    • More automatic egocentric skills may supplement this mind reading.