DD303 Concepts

Intro
1. Concepts
  1. "General ideas that apply to every member of a category"
  2. Internal to mind; categories external
  3. Can map to multiple categories
    1. eg "chest'
Categorisation
1. Bruner
  1. Individual things in terms of group membership
  2. Implies
    1. Conceptual grouping via behaviour
      1. eg petting dogs
    2. Different behaviour = thinking differently about different concepts
2. Sorting tasks
  1. Ross & Murphy
    1. Eggs
      1. Breakfast
      2. Dairy
Concepts & Cognition
1. Makes it easier to remember stuff
2. Semantic classification
  1. Recognition
  2. Units of semantic membership?
    1. Store facts
    2. R'ship between concepts
      1. eg cat & heart
3. Enable inferences
  1. Reasoning
  2. Simplify remembering info
4. Lexical concepts
  1. Category represents what we believe 'cat' to mean
Explaining Categorisation
1. 1) Classical view
  1. Things belonging to category
    1. Common properties
      1. Necessary condition
    2. If item has properties common to category's members, it must be a member too
      1. Sufficient condition
  2. Evidence for:
    1. Bruner et al
      1. People associate common properties with category members
  3. Criticisms
    1. Typicality
      1. Robin more typical bird than penguin
      2. "All or nothing" ≠ reality
        Rosch
        Ps verified sentences 'Robin is a bird' quicker than 'Penguin is a bird'
        Robin = more typical
      3. Categories have rich internal structure
        Not explained by classical view
    2. Borderline cases
      1. If "all or nothing" ≠ borderline
      2. Red/orange?
      3. McCloskey & Glucksberg
        Ps rated items inconsistently & @ different times
        eg bookends as furniture
    3. Intransitiivty
      1. If A is in B and B is in C then A is in C
      2. Hampton
        Ps judgements not fit with rule
        eg car seats
    4. Lack of definitions
      1. Wittgenstein
        All members share common properties?
        Games?
        Olympic
        Card
        Board
        Most categories indefinable
2. 2) Prototype view
  1. Prototype
    1. Most typical examplar of category
    2. Object = member if meets enough properties of category - ie similarity to prototype
      1. ie NO necessary condition
    3. Property can have many values
      1. eg apple colour
    4. Typicality weightings
      1. Highly typical instances match highly-weighted properties
      2. Typicality correlated with how widely members share attributes
  2. Criticisms
    1. Meaning of typicality effects
      1. Armstrong et al
        Concepts not organised around prototype
        BUT existence of typicality effects aren't conclusive evidence membership determined by similarity to prototype
        Suggests: Dual process model (ie combine both theories)
        1) Concept core - judge category membership
        2) Identity procedures - match instances to category
    2. Context sensitivity
      1. Typicality effects alter w/ context
        Medin & Shoben
        Spoons
        Metal more typical than wooden
        Small more typical than large
        BUT large wooden more typical than small wooden!
    3. Complex concepts
      1. not cope with concept combinations
        eg red car
        How combine prototypes?
        eg Stone lion?
3. 3) Common-sense theories
  1. Issues with similarity based approaches (classical & prototype)
    1. Meaning
    2. Items can share infinite things in common
      1. Plumb and Lawnmower
  2. Evidence for:
    1. Rips
      1. Pizza & US 1/4
      2. 3rd object: more like pizza but more similar to 1/4
        Dissociation between category & similarity
        Underlying theory not just similarity based
    2. Kiel
      1. Developmental aspects
        Hybrid zebra w/ horse insides
        4yrs = zebra (appearance); 7yrs = horse (lineage)
      2. Characteristic defining shift w/ age
  3. Criticisms
    1. Just because deeper knowledge suggested (not just similarity) doesn't make it more correct
    2. "theory" poorly defined; common sense not compatible w/ science theory
      1. More like 'knowledge' - Murphy
    3. No explanation for complex concepts
    4. May just replace 'similarity' w/ 'theory'
      1. Both woolly terms
4. 4) Psychological Essentialism
  1. Assumes deeper principles
  2. Objects categorised according to unobservable shared 'essential' properties
    1. Superficial properties known, may not be able to define essential properties
    2. Medin & Ortony
      1. 'Placeholder' for essential properties - empty if cant define
  3. Evidence for:
    1. Innate potential
      1. test kids belief re innate properties
      2. Animal switch from its bio parents to new ones
        Pre-school kids: Kangaroo & goat = pouch & hop (ie innate kangaroo property)
      3. Children more nativist
      4. Gellman & Wellman
        Young kids: insides more import re id of category
  4. Criticisms
    1. Context & perspective
      1. Malt
        Categorisation depend on context & goals of perceiver
        Water = puddles or tears
    2. Expert opinion
      1. Malt
        Ps ask expert re borderline natural item than for artefact (eg soap)
        Natural categories may involve psych essentialism as Ps recognise info experts bring
  5. Gelman
    1. Children
      1. Look beyond the obvious in various ways
        Learning words
        Generalising knowledge to new category members
        Reasoning about insides of things
        Beliefs re nature vs nurture
        Constructing causal explanations
    2. Certain categories (eg lion; female) have underlying reality - cannot be directly observed
    3. Psych essentialism = early cognitive bias?
    4. Against view that children are concrete thinkers
Where next?
1. All theories flawed!
  1. Classical
    1. Necessary & sufficient conditions only id for few categories
  2. Prototype
    1. Not explain context sensitivity or complex concepts
  3. Theory theory
    1. Imprecise, no definitions, cant explain complex concepts
  4. Psych essentialism
    1. Mixed evidence, much doesn't map to 'essences'
2. All categorisation the same?
  1. Each approach for different scenarios ie not a single approach
  2. Determinate # of different kinds of category
    1. Classical - for definitions
    2. Prototype - rapid categorisation or fuzzy matching
    3. Theory based - considered judgements, explanation for category membership
    4. Psych Essentialism - integrate scientific knowledge, expert view
  3. Smith & Sloman
    1. Replicate Rips (pizza)
    2. Same dissociation is Ps think out loud
    3. 2 modes
      1. Similarity based
      2. Rule based
  4. Language
    1. Categories have labels
      1. Words & concepts linked
      2. Some words to label items in different categories
        eg stone lion
    2. Malt
      1. "Shampoo bottle"
        Not a bottle!
3. All concepts the same?
  1. Well defined
    1. amenable to definition
    2. Classical view
    3. Not explain typicality!
  2. Fuzzy
    1. Prototype
      1. eg "red"
  3. Common-sense
    1. "sparrow"; "introvert"
      1. Theory or Essentialism
4. All categorisers the same?
  1. Different approaches depend on knowledge/skill?
    1. Medin
      1. Trees
      2. Taxonomists, maintenance workers, landscapers
        Different approach
  2. Lynch et al
    1. Typicality ratings vary between experts & novices

Next up

Description

Resource summary

Similar

	Created by Ken Adams about 10 years ago