Docstoc

Psy 1306 Lecture 7

Document Sample
Psy 1306 Lecture 7 Powered By Docstoc
					Psy1306: Language and
       Thought
   Lecture 7: Spatial Frames of
            Reference
Spatial Frames of Reference (FoR)
Spatial Frames of Reference (FoR)
Talking about locations & directions
  Figure
  (Thing to be located)
                                             Ground
                                             (Reference Object)


   ►   “Where is the girl?”

        The girl is to the south of the umbrella.

        The girl is to the tilted side of the umbrella.

        The girl is to the left of the umbrella.
              Terminologies 101
► Egocentric   vs. Allocentric
   Ego (self) vs. Non-Ego FoR


► Object-centered    vs. Geocentric
   Moving entities vs. Earth-anchored Entities


► Relative,   Intrinsic, Absolute


                                          Where is the girl?
    Relative, Intrinsic, Absolute
► Figure,   Ground, Coordinate System
   Relative = 3 place relations
   Intrinsic = 2 place relations
     ►Ground   = Coordinate System
   Absolute = 3 place relations




                                        Where is the girl?
    Crosslinguistic Variations for
         small scale arrays
► Intrinsic
► Relative:   always have intrinsic
► Absolute
Majid et al.
Experimental Paradigm – ANIMALS-IN-A-ROW Task
(Bird’s Eye View)                                                  Rotation Experiment




               Step 1: Ss memorize items                                                          Step 2: Ss rotated
 north side)
(right side,




                            Subject
                Table 1                  Table 2                                                         Table 1          Table 2


                  Step 3: Ss recreate “same” as Table 1. At least 2 possible solutions.
                Step 3 geocentric tendency                                                              Step 3 egocentric tendency
                          (north side)




                                                    (north side)




                                                                                         (right side)




                                                                                                                                    (right side)
                Table 1                   Table 2                                                        Table 1          Table 2
                 Brown & Levinson (1993)
                             100
                                            Dutch N = 38
                                 90
                                            Tenejapans N = 27
                                 80
                 % of Subjects   70
                                 60
                                 50
                                 40
                                 30
                                 20
                                 10
                                 0
                                       0   1   2   3    4   5
                                     Number of Geocentric Trials

* Also reported in Pederson, Danziger, Wilkes, Levinson, Kita, Senft (1998).
                                 Other Languages reported in
                     Pederson, Danziger, Wilkes, Levinson, Kita, Senft (1998).

           100                                             100
                       Japanese N=16                                  Locality of Tamil Ss:
                90     Longgu N=13 (-3)
                                                                90
                                                                           Ego Ss N=20 -- urban
                80     Arandic N=11 (-5)                        80         Geo Ss N=41 -- rural




                                                % of Subjects
% of Subjects




                70                                              70
                60                                              60
                50                                              50
                40                                              40
                30                                              30
                20                                              20
                10                                              10
                0                                               0
                   0 1 2      3   4 5                                 0 1    2   3 4 5
                 Number of Geocentric Trials                        Number of Geocentric Trials
Chips Task
Maze Task
Inference Task
         Pederson et al. (1998)’s
          Summary of their Data
Pederson, Danziger, Wilkes, Levinson, Kita, Senft
(1998).

Far more than developing simple habituation, use of the
linguistic system, we suggest, actually forces the speaker to
make computations he or she might otherwise not
make...The linguistic system is far more than an available
pattern for creating internal representations; to learn to
speak a language successfully requires speakers to develop
an appropriate mental representation which is then available
for nonlinguistic purposes. (p. 586).
            Levinson et al. 2004
► Proposed    mechanisms of Whorfian Effects
     Perceptual ‘tuning’ and attention
     Re-representation
     Structure-mapping
     Costs of Computation
Perceptual Tuning
Re-representation
Structure-Mapping
Cost of Computation
               Taking Stock
Data show:
► Experimental finding show ADULT speakers’
  “nonlinguistic” behavior correlates with their
  linguistic behavior.

Alternative explanations?
      Alternative Explanations?
► Otherthird factor (e.g. culture experience)
 might be able to explain the language-
 thought effect?

            problems: is Make it the “same”
► Translation
 same across languages?
    Eskimos and Snow Words
► Eskimos  have many snow words.
► Eskimos make fine snow
  discriminations

► Many  snow words  fine
  snow discriminations
► Experience  Fine snow
  discriminations
   Many snow words
   Possible Confounding Factors
Abs/Geo: preliterate                          rural   stable
  Tenejapan, Arrente, Longgu, Tamil (rural)


Rel/Ego: literate                             urban   transient
  Dutch, Japanese, Tamil (urban)
    Possible Confounding Factors

Abs/Geo: preliterate                 rural   stable
Rel/Ego: literate                    urban   transient


               Sending up the
             patient for his north
               eye surgery…
                 Confounding Factors
 Abs/Geo: preliterate                           rural                  stable

 Rel/Ego: literate                              urban                  transient
One Potential Problem: Literacy possibly linked to response preference
e.g. Totonac (Intrinsic language):
 8 relative responders (men, literate) vs. 8 absolute responders (women, illiterate).

In other studies:
     Danziger & Pederson (1998). Acceptance of mirror-images as being the “same”
     is correlated with literacy.         d different than b

    DeLoache et al. (2000). Reading reinforces orienting pictures and words with
    respect to self.                       d different than p
       Environmental Circumstances
       Influencing Spatial Reasoning
 Abs/Geo: preliterate              rural               stable
 Rel/Ego: literate                 urban               transient

Circumstances help support and maintain these spatial frames
of reference.
   Landmarks and Derivation of Geocentric Directional Terms
          Tzeltal “tree standing downhill of man”
          Hai||om “man stands in ‘land of soft sand”
          Longgu “tree standing on side towards sea”
Maze Learning in Rats                                     Rat Setup




“Place-vs.-Response” -- which one is dominant?
  Experimental Setup
   1. Training.          2. Test turn preference when maze rotated 180°
                                   Place                               Response
                                (geocentric)                          (egocentric)
                                                    or



  Experimental Variation
     Teach groups PLACE or RESPONSE . Which is easier to learn?
            Place                              Response
          (geocentric)                          (egocentric)
                                     or
                                 Restle Quote




               Summary of Rat Literature
Restle (1957). Discrimination of cues in mazes: A resolution of the
place vs. response question., Psychological Review, 64, p. 226.

There is nothing in the nature of a rat which makes it a
“place” learner or a “response” learner. A rat in a maze will
use all relevant cues, and the importance of any class of
cues depends on the amount of relevant stimulation
provided as well as the sensory capacities of the animal. In
place-response experiments, the importance of place cues
depends on the amount of differential extra-maze
stimulation.
 Egocentric vs. Allocentric Debate
                    Acredolo & Evans (1980)
 Varied testing context (to be explained on next slide) using the
 basic paradigm below:
Step 1: Train infant to
turn to look at one           Step 2: Rotate infant 180°.
window.                       Which way does infant turn to look?




                                 (allocentric)      (egocentric)
         Infant &
          Mother
 Results from 6-11 months-olds
Test Conditions:
1. Both Windows Plain
2. Fancy Trained Window, Plain Untrained (Direct Landmark)
3. Plain Trained Window, Fancy Untrained Window (Indirect
   Landmark)

Found:
    6 months, prefer egocentric in any environment
    9 months, use direct landmark when available
    11 months, additionally to 9 months, use indirect
   landmark when available
Bottom-line:
Cues in the environment affect spatial behavior
    Egocentric vs. Allocentric Debate      Acredelo Setup




                                One-shot, No Training
                                   Acredolo (1979)
                                                            Step 3: Where does infant
                                Step 2: Move infant 180° to search?
Step 1: Hide object in one of   other side of the table
two locations.




                  Infant &
                  Mother


                      Vary Setting Unfamiliar vs. Familiar
                         • bare laboratory
                         • laboratory with clutter
                         • home
Result on Familiarity of Environment
                 Laboratory            Home
                 UNFAMILIAR            FAMILIAR
Bare             Egocentric


Cluttered        Egocentric            Allocentric


    Bottom-line: Environment affects spatial behavior.
                  Summary
► Experiments  with prelinguistic infants and animals
 (like the previous studies with Penn undergrads)
 show that environmental contexts affect how the
 subjects choose to represent the spatial arrays.

► Environment  affects how one chooses to think
 about spatial relationships. As a result,
 environment could come to influence the kind of
 language that develops.
                 Confounding Factors    Testing Location




Tenejapans tested outdoors on their hill and Dutchmen tested indoors in
laboratory.


                                 S (Uphill)
         House                                             (Tenejapan table setup
                                                           Outdoor, porch next to house)
                                N (Downhill)
            1        2




Shouldn’t spatial performance be influenced by
spatial environment?
Question: Perhaps the Tenejapans’ response NOT result of LANGUAGE, BUT
result of surrounding environment and available landmarks?
Asking experimenters of
Pederson, Danziger, Wilkes, Levinson, Kita, Senft (1998)…
Where were subjects tested?

                      Indoor vs. Outdoor

                                Animals-In-A-Row Task
        Test
    Environment
                   geo resp. ego resp. mixed     mono.      Total # Ss.

       Indoor*        20        58        3         8           89
     Outdoor**        46        6         10        0           62


  * Dutch, Japanese, Arrente, Totonac
  ** Tzeltal, Kgalagadi, Hai||om
                 Eskimo Problem
► Language is tied to the circumstance
► Disassociate language from circumstance
   Move Americans to Vail or Aspen
     ► Start to think more about snow...
     ► Start to get words like “sugar”, “granule”, “powder” for snow
   Move Eskimos to Bermuda
     ► Fewer   uses of snow words, perhaps?

   LIKEWISE – We Move Americans to “Tenejapa”
    (i.e., outdoors to a gridded city where streets run from
    east to west in increasing cardinality)
     ► Start to get more geocentric responses…
     ► Start to get more words like “north”, “south”...
Turning Americans into Tenejapans
       Li & Gleitman (2002)
                             Small Animal Setup




                            Tenejapans

                                                                    S (Uphill)
                                                  House

                                                   1       2        N (Downhill)




                                                       Library
                                                       Walnut St            S
                                                                   Window
                                                       1       2            N
                                                       (room)
                       Condition 1: IRCS Room BLINDS DOWN
                       Placing Americans in a setting like the Dutch.

                       Condition 2: IRCS Room BLINDS UP
                       Placing Americans in a setting like the Tenejapans.
(testing location: IRCS -- indoor)
                           Small Animal Data




                                                                           English Speakers
                     Brown & Levinson (1993)                          Blinds-Down and Blinds-Up
          100                                                       100
                                      Dutch N = 38                                  IRCS Blind Down N =10
                90                    Tenejapans N = 27                   90        IRCS Blind Up N =10




                                                          % of Subjects
                80
% of Subjects




                                                                          80
                70                                                        70
                60                                                        60
                50                                                        50
                40                                                        40
                30                                                        30
                20                                                        20
                10                                                        10
                0                                                         0
                      0  1 2 3 4 5                                              0 1 2 3 4 5
                     Number of Absolute Trials                                 Number of Absolute Trials
   Big Animal Setup




                          Tenejapans

                                           S (Uphill)
                               House

                                1      2   N (Downhill)




Outdoor Location 1                         Outdoor Location 2
                                                        Locust Walk

                           S
                                                                        S


                                               HRN
  1                   2
                                                            1       2
                           N
  Library                                                               N
                                                            Field
                         Big Animal Data




                     Brown & Levinson (1993)                                          Outdoor Conditions
          100                                                            100
                                           Dutch N = 38                                    By HRN N=20
                90                         Tenejapans N = 27                   90          By Library N = 10




                                                               % of Subjects
                80
% of Subjects




                                                                               80
                70                                                             70
                60                                                             60
                50                                                             50
                40                                                             40
                30                                                             30
                20                                                             20
                10                                                             10
                0                                                              0
                      0  1 2 3 4 5                                                   0 1 2 3 4 5
                     Number of Absolute Trials                                     Number of Absolute Trials
                                                          Ducks Setup




                      Increasing Saliency of Landmarks
            Unmentioned Duck Ponds on the sides of tables as landmark.




         Condition 1: Absolute Biasing         Condition 2: Relative Biasing

                Library                                Library
                                     S                 Walnut St                     S
                Walnut St
                            Window                                          Window
                1     2                               1                 2            N
                                     N
                 (room)                                (room)




(testing location: IRCS -- indoor)
                            Ducks Data




                                                                       Ducks on the Tables
                     Brown & Levinson (1993)                          Landmark Manipulation
          100                                              100
                            Dutch N = 38                                     Relative Bias N=20
                90          Tenejapans N = 27                    90          Absolute Bias N=20




                                                 % of Subjects
                80
% of Subjects




                                                                 80
                70                                               70
                60                                               60
                50                                               50
                40                                               40
                30                                               30
                20                                               20
                10                                               10
                0                                                0
                      0  1 2 3 4 5                                  0  1 2 3 4 5
                     Number of Absolute Trials                    Number of Absolute Trials
                   Summary
► Pre-linguistic infants & rats make use of
  both frames of reference
► Cultural difference or other third factor
  might be able to explain some of the effect.
   Eskimo Analogy
   Circumstance  language & behavior
            Task Ambiguity
► Does  the “same” mean the same across
  languages?
► Interpretation of “same” within a language
  requires guessing intent of experimenter.
    Interpretations of “the same”

                   180º



N
            Task Ambiguity
► Does  the “same” mean the same across
  languages?
► Interpretation of “same” within a language
  requires guessing intent of experimenter.
► Ambiguous tasks are subjected to
  contextual influence:
   Environment change (indoor-outdoor)
   Triggers s.a. holding animals
   One table manipulation
Reducing Ambiguity – 1 Table vs. 2 Tables
              One Table 180° Rotation – Same Table, Same Frame of Ref.
                                                        One-Table Setup




         Step 1: Ss memorize items                                              Step 2: Ss rotated




                                         (left side,
            Subject
                                          north side)
                                 Table                                                Table


        Step 3: Ss recreate “same” as Table 1. At least 2 possible solutions.
       Step 3 absolute tendency                                           Step 3 relative tendency
                  (north side)




                                                                                              (left side)
                                 Table                                                Table

(testing location: IRCS -- indoor, blinds-down)
Reducing Ambiguity – 1 Table vs. 2 Tables

   (left side,                  Two Tables With Outer 180° Turn – Control Condition
                                                                   Two-Tables Outer Setup
    north side)

                          Step 1: Ss memorize items                                                            Step 2: Ss rotated




 Subject
                          Table 1                        Table2                                            Table 1             Table 2



                         Step 3: Ss recreate “same” as Table 1. At least 2 possible solutions.
                        Step 3 absolute tendency                                                          Step 3 relative tendency
         (north side)




                                          (north side)




                                                                                            (left side)




                                                                                                                                         (left side)
                          Table 1                        Table 2                                           Table 1             Table2

(testing location: IRCS -- indoor, blinds-down)
                                                                                                  English Speakers
                     Brown & Levinson (1993)     One vs. Two Table Data




                                                                                               One Table vs. Two Tables
          100                                                                       100
                            Dutch N = 38                                                            One Table N=10
                90          Tenejapans N = 27                                             90        Two Tables N=10
                80
% of Subjects




                                                                          % of Subjects
                                                                                          80
                70                                                                        70
                60                                                                        60
                50                                                                        50
                40                                                                        40
                30                                                                        30
                20                                                                        20
                10                                                                        10
                0                                                                         0
                      0  1 2 3 4 5                                                              0  1 2 3 4 5
                     Number of Absolute Trials                                                Number of Absolute Trials
            Task Ambiguity
► Does  the “same” mean the same across
  languages?
► How one’s linguistic community customarily
  speaks about or responds to inquiries about
  locations and directions might come to
  influence what appropriately counts as the
  “same” spatial array.
              Task Ambiguity
► Experimental   Q to be answered next time.
   Li et al. (under review)
  What about Haun et al (2006)
            PNAS?

► Common    phylogenetic inheritance of a
  preference of allocentric strategy
► Such preference can be overwritten by
  cultural preference for egocentric strategy
                Levinson (2003)

Cited evidence of linguistic relativity
► Gesture Data
► Pointing to unseen location
      Pigeons        Tseltal Speakers   Dutch Speakers
                          ‘HOME’




                           P
Slide from Peggy

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:1
posted:5/18/2012
language:
pages:60
fanzhongqing fanzhongqing http://
About