Cognitive Psychology, 2nd Ed.

1
Cognitive Psychology, 2nd Ed.

• Chapter 8
• Semantic Memory

2
Representing Concepts

• Concepts are general ideas that enable the
  categorization of unique stimuli as related to
  one another.
• Concepts are characterized by dimensions of
  variation among exemplars.

3
Contrasting Types of Concepts

• Rule governed concepts specify the features and
  relations that define category membership on an
  all or none basis.
• Classical view holds assumes defining features
  are related by a conjunctive rule.

• Object concepts refer to natural kinds and
  artifacts that violate the classical view.
• Characteristic features are disjunctively
  related, creating a family resemblance structure
  and a fuzzy boundary.

4
(No Transcript)
5
(No Transcript)
6
Prototype

• The best or most typical example of a category
  that serves in the mental representation of a
  concept.
• The range of feature variation on a stimulus
  dimension and feature frequency of occurrence
  define in part the gradient of category
  membership.
• The gradient creates typicality effects in
  categorization speed, acquisition order, and
  priming.

7
(No Transcript)
8
(No Transcript)
9
Schema

• A schema is a cognitive structure that organizes
  related concepts and integrates past events.
• Frames organize the physical environment (e.g.,
  an office frame).
• Scripts represent routine activities (e.g., a
  restaurant script). Cumulative recall of script
  events is linear whereas object exemplars follow
  a negatively accelerated curve.

10
Meta-Representation

• Defined as a mental representation of another
  mental representation. Thinking about thinking
  requires meta-representation.
• Pretending a banana is a telephone requires a
  meta-representation linking the two object
  concepts. Meta-representation thus affords
  flexible and creative cognition.
• Between ages 2-4 the use of meta-representation
  develops.

11
Theory of Mind

• Theory of mind refers to the human ability to
  infer that others, like ourselves, have mental
  states. It helps account for why we are not all
  adherents of solipsism.
• By age 4 children can not only pretend but can
  predict the consequences of another having false
  beliefs.
• Mindblindness is an inability to understand that
  others possess mental representations and is
  characteristic of autism. An autistic child is
  socially isolated and treats others as robots
  without feelings and thoughts.

12
(No Transcript)
13
Propositions vs. Images

• Abstract means of mental representation.
• Schematic and verbal.
• Each proposition is an assertion that may be true
  or false.
• Coded as a relation and arguments (e.g., Fred is
  tall).

• Perceptual means of mental representation.
• Concrete and nonverbal.
• One image conveys Represent multiple features and
  relations.
• Can images be decomposed into propositions?

14
(No Transcript)
15
Functional Equivalence Hypothesis

• Visual imagery, while not identical to
  perception, is mentally represented and functions
  the same as perception.
• An image is isomorphic to the referent object
  (second-order), meaning spatial relations are
  analogous.
• An image is an analog representation of the
  object, as shown by mental rotation and image
  scanning.

16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
The Nature of Propositions

• Fred is tall is a single proposition coded as a
  relation with two arguments (is, Fred, tall).
• The ants ate the sweet jelly that was on the
  table expresses four propositions.
• Latent Semantic Analysis is a mathematical
  procedure for extracting and representing the
  meanings of propositions expressed by a text. It
  represents the co-occurrence of words and their
  contexts. Using a database of co-occurrence
  relations, it can compute the similarity in
  meaning of two words or texts.

22
(No Transcript)
23
Semantic Network vs. Feature Comparison Models

• Hierarchical network of concepts .
• Cognitive economy stipulates features are
  represented only once in the hierarchy.
• Used in WordNet to represent word meanings.

• Feature vector defines each concept for each
  level (e.g., robin, bird, animal).
• Stages of feature search (characteristic
• vs. defining) explains typicality effects.

24
(No Transcript)
25
(No Transcript)
26
Category Size EffectAll collies are dogs
faster than All collies are animals.

• Network model assumes that feature search must
  proceed from level 0 to level 1to confirm dog.
• Must proceed to level 2 to confirm animal, taking
  more time.

• Feature comparison assumes search of
  characteristic features is sufficient to confirm
  dog.
• Must proceed to Stage 2 search of defining
  features to confirm animal.