Cognitive Psychology

1
Semantic Memory Memory for meaning Permanent
memory store of general world knowledge Mental
thesaurus, dictionary, or encyclopedia Language,
concepts, decisions, etc. Whereas episodic
memory differs widely from individual to
individual, semantic memory is similar across
individuals
2
Two Models of Semantic Memory Collins and
Quillian Network Model Smith Feature Comparison
Model Each makes two assumptions of semantic
memory 1) Structure 2) Process of retrieval
3
Collins and Quillian Network Model Two
fundamental assumptions of semantic
memory Structure Nodes in a network Process
of retrieval Spreading activation
4
Collins and Quillian Network Model Structure
Nodes in a network
Each concept in semantic memory is represented by
a node, a point or location in semantic space
5
Collins and Quillian Network Model Structure
Nodes in a network
Each concept in semantic memory is represented by
a node, a point or location in semantic space.
6
Collins and Quillian Network Model Structure
Nodes in a network
Nodes are linked together by pathways,
directional associations between concepts. Every
concept is related to every other concept.
7
Collins and Quillian Network Model Structure
Nodes in a network
Each pathway has a label defining the
relationship between the concepts Isa statements
and property statements form propositions
8
Collins and Quillian Network Model Process of
Retrieval Spreading Activation
An individual concept becomes activated
9
Collins and Quillian Network Model Process of
Retrieval Spreading Activation
This activation spreads to adjacent nodes,
activating them as well
10
Collins and Quillian Network Model Process of
Retrieval Spreading Activation
Activation continues to spread through the
network, but the level of activation decreases
with each step
11
Collins and Quillian Network Model Process of
Retrieval Spreading Activation
Consider the activation caused by Can a robin
can breathe?
12
Collins and Quillian Network Model Process of
Retrieval Spreading Activation
The nodes ROBIN and BREATHE spread activation
through the network.
13
Collins and Quillian Network Model Process of
Retrieval Spreading Activation
The intersection of two spreads of activation is
found indicating critical concepts a decision
stage operates to determine validity of
intersection
14
Smiths Feature Comparison Model Two fundamental
assumptions of semantic memory Structure
Feature lists Process of retrieval Feature
comparison
15
Smiths Feature Comparison Model Structure
Feature Lists Semantic memory is a collection of
Feature Lists Each concept represented as a list
of semantic features simple, one-element
characteristics of the concept Features are
ordered in a list in terms of definingness The
most defining features for a concept are at the
top of the list
16
Smiths Feature Comparison Model Structure
Feature Lists Defining features Features
absolutely essential to the concept (e.g. Birds
are living objects) Characteristic features
Features common to, but not essential to, a
concepts meaning (e.g. Birds fly)
17
Smiths Feature Comparison Model Process of
Retrieval Feature Comparison General example
True or false An A is a B? Stage 1 Global
Feature Comparison Access required concepts and
randomly select features about each
concept Features compared and similarity score
determined High Yes, Low No, or
Intermediate Go to Stage 2
18
Smiths Feature Comparison Model Process of
Retrieval Feature Comparison General example
True or false An A is a B? Stage 2 Comparison
of Defining Features Access defining features of
each concept Determine if defining features
match Features match Yes Features
mismatch No
19
Smiths Feature Comparison Model Process of
Retrieval Feature Comparison
20
Clashing Evidence for the Models General Task
Sentence Verification Key issues Cognitive
Economy Property Statements Typicality Effects
21
Clashing Evidence for the Models Cognitive
Economy (Bad for C Q Good for S)
Wings
Wings
Bird
Bird
Feathers
Feathers
Robin
Robin
Red Breast
Wings
Blue Eggs
Feathers
22
Clashing Evidence for the Models Property
Statements (Bad for S Good for C Q) E.g. A
canary is a small bird with yellow
wings According to Smith Look up feature lists
for five concepts CANARY, BIRD, SMALL, YELLOW,
and WINGS This requires a list of Things that
are small Things that are yellow Things
with wings Collins Quillian incorporate
property statements into their network so it
doesnt face this problem
23
Clashing Evidence for the Models Typicality
Effects (Bad for C Q Good for S) Not all
members of a category are equal Typical members
of a category can be judged faster This is
captured with Smiths similarity score but not
explained by Collins Quillian
Robin
Bird
Chicken
Sparrow
Penguin
24
A Hybrid Model
No strict cognitive economy Property statements
available Typical members of a category stored
more closely Properties more important to
concept stored more closely
25
A Final Wrinkle Recent ERP research is now
suggesting that RT effects (e.g. typicality
effects) in semantic memory may be associated
with decision processes rather than retrieval
processes Whats the problem? Such effects have
lead to model revisions that add semantic
distance between nodes (e.g. the hybrid model
just described.) Such revisions may not be
appropriate. Current models of semantic memory
have yet to adequately address this finding
26
Categorization Concept Formation Traditional
Research Show subjects a series of arbitrary
patterns and have them judge whether each is an
example of the concept being tested. Limitations
are that they are not related to the real world
27
Categorization Natural Categories Concepts and
categories that occur in the real world Members
do not belong to their categories in simple
yes/no fashion Categories have fuzzy boundaries
with ill-defined membership for many category
instances No single feature is absolutely
necessary as a criterion of category
membership Membership in a category is a matter
of degree
28
Back to Spreading Activation Four important
principles associated with this idea 1)
Activation spreads 2) Spreading takes time 3)
Activation becomes diffuse as it spreads 4)
Activation decays over time If semantic
relatedness is the organizing principle of
semantic memory, then relatedness should play a
big role in these principles The test Priming
29
Priming in Semantic Memory In essence How does
the processing of a prime affect the processing
of a target? Does thinking about one concept
bring to mind other concepts? If so, they are
connected in semantic memory
30
Priming in Semantic Memory How can we use this to
test the association between semantic relatedness
and spreading activation? - Distance of
spread - Speed of spread
31
Priming in Semantic Memory Distance of Spread
Vary steps between prime and target
STIMULUS 1 ROBIN
Activation added to Robin
32
Priming in Semantic Memory Distance of Spread
Vary steps between prime and target
STIMULUS 1 ROBIN
Activation spreads through network
33
Priming in Semantic Memory Distance of Spread
Vary steps between prime and target
STIMULUS 2 BIRD
Does activation get this far?
34
Priming in Semantic Memory Distance of Spread
Vary steps between prime and target
STIMULUS 2 FEATHERS
Does activation get this far?
35
Priming in Semantic Memory Distance of Spread
Vary steps between prime and target
STIMULUS 2 BREATHES
Does activation get this far?
36
Priming in Semantic Memory Speed of Spread Vary
time between prime and target
How long does it take activation to go from Robin
to Bird?
37
Priming in Semantic Memory There are two major
ways to set up these experiments
38
Empirical Demonstrations of Priming Freedman and
Loftus (1971) Name a member of a category
defined by a prime and a target Conclusion
Category faster than letter or color
Prime Target Result
P Fruit No Priming Red
Fruit No Priming Fruit
P Priming Fruit
Red Priming
39
Empirical Demonstrations of Priming Loftus and
Loftus (1974) Same methodology as Freedman
Loftus Difference Trials have various SOAs
(within trials) and sometimes repeated the
category of a previous trial (across trials)
TRIAL 1 TRIAL 2 TRIAL 3 TRIAL 4 Lag 0
Fruit-P Fruit-B Animal-D Building-L Lag
2 Fruit-P Animal-D Building-L Fruit-B
40
Empirical Demonstrations of Priming Loftus and
Loftus (1974)
41
Empirical Demonstrations of Priming Loftus and
Loftus (1974) Within trials Category more
facilitation than letter or color Across trials
Facilitation less at longer lags SOAs
Facilitation better at longer SOAs
42
Empirical Demonstrations of Priming Rosch
(1975) Are two things members of the same
category? Prime Category name (related) or
Blank (neutral) Targets Typical or atypical
category members
No priming
Lots of priming
Less priming
43
Empirical Demonstrations of Priming Meyer and
Schvaneveldt (1971) Lexical Decision Task
44
Empirical Demonstrations of Priming Neely (1977)
Insert figure 7-13
45
Empirical Demonstrations of Priming Neely
(1977) At small SOAs, there is facilitation
between related words even tough it is
unexpected. At larger SOAs, this facilitation
disappears.
46
Empirical Demonstrations of Priming Marcel
(1980) Immediately after prime, present a
mask Prevents conscious awareness of seeing
prime Even without conscious awareness, prime
affects target
Child
Infant
47
Summary of Priming Related primes speed
processing Activation spreads from one concept
to another Reduced relatedness or typicality of
concepts decreases priming Activation spreads to
most related concepts Longer SOAs increase
priming Spreading activation takes time
48
Summary of Priming Longer lags decrease priming
Activation decays Priming at very short SOAs and
despite conscious expectations Priming is
automatic Priming occurs without awareness
Priming is implicit
49
Context and Priming Work in context is interested
in large-scale semantic representations that
involve episodic and semantic knowledge -
Comprehension of sentences and
paragraphs - Comprehension of spoken
conversations
50
Contextual Ambiguity and Priming What does the
word bank mean?
1. We had trouble finding the bank.
2. We were swimming at the bank.
3. We were making a deposit at the bank.
Sentence 2 primes RIVER Sentence 3 primes
MONEY Context primes a particular concept of bank