Semantic Memory Flashcards Preview

PY2CN > Semantic Memory > Flashcards

Flashcards in Semantic Memory Deck (36):
1

How is conceptual knowledge organised and represented in semantic memory

Classical defining feature (collins and quillian 1967)

Feature comparison model (smith, shoben and rips 1974)

Prototype model (rosch 1975)

2

Concept as defined in classical defining feature theory

Represented in memory and defined by its set of features
mental representation of category/class of objects wjhich stored vast amounts of representational information

3

Describe classical defining feature theory (collins and quillian)

Concept defined by features - features that are integral to making it what it is (can also have additional features that vary)
Features are 'primitive' to the concept, of equal importance and all examples are equally representative of that concept
organised into a hierachy of super and subordinate groups

4

Hierarchical organisation of concepts in classical defining feature theory

Hierarchy of super ordinate and subordinate categories -
Super ordinate - all primitively defining features
Subordinate - the individual, independent examples/cases that can vary

5

Collins and quillan defining feature theory hierachial structure

Knowledge organised in hierarchy in semantic network

Network has nodes (of concepts) and connections between the nodes which connects concepts and concept properties together in terms of their relation to each other

Make semantic judgments by spreading activation across connections - when reach intersection between concepts must verify the relation- eval path for relevance and decide if satisfies constraints
- links can have different critereonalities dependent on how essential each link is to the meaning of a concept

6

Ways nodes are connected in network (Collins and quillan defining feature)

'Is a' - is a member of... - super and subordinate links

'F' - is a feature of...

7

Define cognitive economy Collins and quillan defining feature

Some general features which may not fully apply to ALL examples of a concept is attached to general categories instead of every individual node - those which deviate have own later nose
Ie birds - 'can fly'
weak theory of cog economy - everytime learn that x is a bird - do not link to all the properties of a bird - instead, make inference of the links with other features unless directly confronted with the info

8

What does Collin and quillians hierarchical organisation of defining features predict about retrieval of info

Time it takes To acces the info depends on the distance that is travelled - the length of the connection between the two concepts
if reach an intersection between two notes then triggers the decision to verify the relation

9

How can you test Collins and quillans prediction

Sentence verification task - give and instance/example and ask if it fits into a category

Ie
Canary is a canary (one level)
Canary is a bird (two level)
Canary is an animal (three level)

10

collins and quillians 1969/1970)

performance on sentence verfication task predicted by how many ‘isa’ and ‘property’ pathways one needed to traverse to verify a sentence
activation retrieval mechanism that spread across
links within the network, and the more links traversed
the slower the retrieval time
clear typicality effect - faster for spiders are bias (no) than something which has slight overlap ie butterfly bird?

11

Problems with 'features' defining feature theory

Not all features are equal
Some more closely associated to a concept than others

12

Problem with 'instances' in defining feature theory

Some instances/examples are more typical of a concept than others

Smith et al 'typicality theory' - quicker to identify robin than ostrich as bird in sentence verification task

13

Problems with 'defining features' in defining feature theory

Hard to identify set of features that defines all examples and excludes non members

14

Problem with categorical boundaries in defining feature theory

Differences between categories are not always clear cut - some atypical instances mean people can't agree with others or themselves or tomato as a fruit?

15

Problem with prediction of defining feature theory - rate of response

Sometimes fail - faster to verify ostrich animal than ostrich bird

16

Describe feature comparison model (smith et al 1974)

Two types of feature:
Defining features - essential to a concept
Characteristic features -
Non essential features of some instances but not all
Judge similarity between categories by 1 or 2 stage processing mechanism
emophasis on the decision process - if high degree of overlap between concepts then faster response than if less so - faster also when no overlap

17

Describe the 1 or 2 stage processing mechanism of feature comparison theory

In judging similarity between an instance/example and a category

Stage 1
Rapid and general comparison of defining features of both concepts

~ if intermediate ie not completely obvious ~

Stage 2
Slow and careful comparison of inly the defining features

18

Feature comparison theory and the typicality effect

Typicality effect - faster for instances typical of an overall category

Feature comparison accounts for this - take longer to process (2nd stage) if instance is atypical

19

Feature comparison in explaining problems with defining feature hierarchy (faster to verify ostrich animal than ostrich bird)

Reason faster is the that ostrich animal requires only stage 1 processing so FASTER while ostrich bird atypical of bird category and therefore stage 2 processing

20

Problem with feature comparison theory and noun order

Why robin bird faster than bird robin? - NOUN ORDER

Should change performance of judgement on comparing features

21

Feature isolation problem in feature comparison

Features treated as single instances and doesn't expand on how relate to other features and how this may have an influence

22

Defining feature problem feature comparison

Same as defining feature theory - what constitutes as a defining feature?

23

Describe the prototype model (rosch)

Category judgements not based on similarity between instance and a categories features

Form a prototype example of a category that is averaged from prior experience of category members
- compare with new instances to judge membership of fitting the prototype

Accounts for typicality effect - prototype most typical of category BUT allows for fuzzy boundaries where distinction not completely clear as can vary across categories and people have different prototypes for categories

24

Prototype view and family resemblance

Category membership by similarity not definition - not all features are shared by all instances but all have things in common that group together (the smith brothers)

25

Define a self schema

General info of traits attitudes abilities as goals

Way we think about ourselves influence the way we process into about self nd others - ie remember more self relevant info and attend to info that fits with self schema

more likely to recall self relevant info and attend to self-congruent>incongruent info

26

Define a script


schema: Describe knowledge of frequently occurring events - generic knowledge about wheat usually happens when you go there based on overall instances

27

Need for schemas and scripts

Understand events and form expectations

Make sense of events and draw inferences

Short hand communication (don't need to detail to others EXACTLY what did as all know basics)

facilitate retrieval

28

Brewer and treyens 1981 script evidence

Pps in office of 30 secs

Take to another room and write down everything you can remember

Recalled more scenarios consistent than inconsistent

Falsely recalled more schema consistent

29

Failures in semantic memory - tip of the tongue phenomenon

Momentarily unable to reveal info that know have stored - strong feeling of knowing but not recall completely

Semantic info activate but not spread to phonological info to produce the word

30

Define a concept

Mental representation of a category or class of objects

Represented in semantic memory

31

Collins and Loftus (1975) development of defining feature theory

developed a network that was not forced into a hierarchical framework
with pathways between concepts that are related - strength of the relationship reflected by the length of the pathways
links between nodes dependent on semantic
similarity (e.g., items from the same category, such as
DOG and CAT, would be linked), or from lexical level factors, such as cooccurrence in lang/context

32

problem with prototype theory
ad hoc categories
barsalou 1983

people construct ad hoc categories to achieve goals - differ from common categories in that violate correlational nature with environment and not well established in memory
unlikely to have pre existing protoypes for ad hoc categories- how is catergory membership determined?

33

example barsalou 1983

eem to be easily
generated from traces that do not inherently have
natural category structure; for example, what do
photographs, money, children, and pets have in common?
On the surface, these items do not appear to be
similar – they do not belong to the same taxonomic
category, nor do they share many features. However,
when given the category label ‘‘things to take out of
the house in the case of a fire,’’ these items seem to fit
quite naturally together because our knowledge base
can be easily searched for items that are in the house
and are important to us

34

problems with prototype theory

know the relation between attributes even when not captured by the representation of a single prototype

category membership not just based on similarity

dissociation between similarity and catergory membership not readily explained by prototype or feature theory

35

similarity and category membership limitation of prototype
rips 1989

pps given two everyday categories
1- 25 cent coins (have a fixed diameter)
2- pizza (have a variable diameter)
imagine an object with a diameter somewhere between that of a coin and a pizza
group 1 - which category is the object most similar to? - most likely say coin
group 2 - which category does the object most likely belond to? - most likely say pizza
- dissociation between similarity and category membership of objects

36

bower et al 1979

high agreement between people when asked to list what do in different circumstances ie going to a restaurant