Semantic Long Term Memory NEED TO FINISH Flashcards Preview

Language and Memory > Semantic Long Term Memory NEED TO FINISH > Flashcards

Flashcards in Semantic Long Term Memory NEED TO FINISH Deck (57)
Loading flashcards...
1
Q

What is Semantic long-term memory comprised of?

1.

2.

A
  1. Declarative long-term memory.

2. Procedural long-term memory.

2
Q

What is Declarative long-term memory?

A

Memory for things that can be talked about and consciously re-experienced.

3
Q

What are the 3 types of Declarative long-term memory?

A
  1. Episodic
  2. Semantic
  3. Lexical
4
Q

What is Episodic LTM?

A

Autobiographical memory, rich in contextual detail.

5
Q

What is Semantic LTM?

A

Memory of facts, concepts, language. Often devoid of contextual detail.

6
Q

What is Lexical LTM?

A

Memory for words, including their meanings, spellings, pronunciations, syntactic categories. A mental ‘dictionary’.

7
Q

What is Procedural long-term memory?

A

Memory for how to do something.

8
Q

What are the 3 types of Procedural long-term memory?

A
  1. Priming
  2. Conditioning
  3. Misc.
9
Q

What is priming?

A

Processing a stimulus facilitates the process of another, similar stimulus.

10
Q

What is conditioning?

A

Operant or classical (Skinner and Pavlov).

11
Q

What is misc. types of learning?

A

Taste-aversion learning.

12
Q

Semantic concepts: a classical view.

What did Aristotle argue about concepts?

A

Aristotle argued that concepts are defined by necessary and sufficient features.

13
Q

What are the 2 major flaws in Aristotle’s argument about concepts?

A
  1. This view of concepts is very limited in scope: most concepts don’t have the necessary features and conditions.
  2. Cannot explain many aspects of conceptual knowledge, such as how it interacts with Episodic LTM.
14
Q

In 1969, who developed a semantic network model (computer program) of conceptual knowledge?

A

Collins and Quillian

15
Q

Using a program that was used to answer questions for practical purposes (artificial intelligence), Collins and Quillian used it to develop what instead?

A

A semantic network model of conceptual knowledge.

16
Q

What are Semantic Network Models?

A

A hierarchy or organised networks of propositions (elementary units of meaning) that represent your knowledge.

17
Q

Semantic Network Models:

What are propositions?

A

Semantic primers that hold little bits of meaning.

18
Q

Semantic Network Models:

What are propositions composed of?

A

1 predicate + multiple arguments

19
Q

Semantic Network Models:

What is an argument/node?

A

Concepts or features.

20
Q

Semantic Network Models:

What is a predicate?

A

Specifies what kind of relationship propositions (semantic primers) have to arguments/nodes (concepts or features).

21
Q

What are the 2 ways propositions can be represented?

A
  1. Predicate calculus.

2. Linked graphs.

22
Q

What is predicate calculus?

A

A formal way of representing logical relationships.

E.g.
has (robin, red breast)
predicate (node, node)

23
Q

What are linked graphs?

A

A formal way of visually depicting relationships.

E.g.
robin - has - red breast
node - predicate - node

24
Q

What are the 3 properties of Semantic Networks?

A
  1. Cognitive economy.
  2. Inheritance.
  3. Access to info.
25
Q

Properties of Semantic Networks:

What is a cognitive economy?

A

Only useful info is represented in the network.

26
Q

Properties of Semantic Networks:

What is inheritance?

A

Concepts inherit or automatically share features from concepts higher up in the network.

E.g. the concept robin inherits the features feathers, from he concept bird, and breathes from the concept animal.

27
Q

Properties of Semantic Networks:

What is access to info?

A

If one concept is activated, its activity automatically spreads to nearby concepts via Automatic Spreading Activation (ASA).

28
Q

Properties of Semantic Networks:
Access to info:

What are the 3 properties of Automatic Spreading Activation? (ASA)

A
  1. Is loosely based on neural activation.
  2. Takes time to propagate between nodes.
  3. Dissipates as it propagates across the network (activation decreases with time and distance).
29
Q

What is the method used in Sentence-Verification tasks?

A

Subjects view individual statements (sentences) displayed one at a time. Instructed to indicate whether each statement is true or false via button presses.

E.g. a robin lays eggs = true.

30
Q

What is the logic in Sentence-Verification tasks?

A

Reaction times are used to map out the organisation of semantic networks.

31
Q

What conclusions can we draw from Sentence-Verification tasks?

A

Semantic networks exist, using economy, inheritance and access to info.

32
Q

What are the 4 key problems with Semantic Networks theory?

A
  1. Explains how people makes ‘yes’ responses, but not how they say ‘no’.
  2. Doesn’t explain how concepts are learned.
  3. Doesn’t explain ad hoc strategies (things you haven’t thought about before).
  4. Doesn’t explain typicality effects.
33
Q

Give an example of how Semantics Networks theory explains how people makes ‘yes’ responses, but not how they say ‘no’.

A

Can’t explain why responses to statements like ‘a robin is a fish’ take longer than statements ‘a robin is a tool’.

34
Q

Give an example of how Semantics Networks theory doesn’t explain how concepts are learned.

A

How does a child who initially sees a robin with wings then later learn the fact that all birds have wings?

35
Q

Give an example of how Semantics Networks theory doesn’t explain ad hoc strategies (things you haven’t thought about before).

A

Things that are good to remove from a burning home.

36
Q

Give an example of how Semantics Networks theory doesn’t explain typicality effects.

A

Subjects are faster at responding to ‘a robing is a bird’ than to ‘a penguin is a bird’ because robins are more typical birds than penguins.

37
Q

Who proposed an alternative, feature-list model of semantic memory in 1974?

A

Smith

38
Q

What did Smith propose about semantic memory?

A

An alternative, feature-list model of semantic memory. According to this model, concepts are lists comprised of 2 features.

39
Q

What are the 2 features in Smiths alternative feature-list model to semantic memory?

A
  1. Defining features.

2. Characteristic features.

40
Q

What are the Defining Features in Smiths alternative feature-list model to semantic memory?

A

Features that are necessary to be a member of a category.

E.g. birds, feathers, beaks, warm blooded.

41
Q

What are the Characteristic Features in Smiths alternative feature-list model to semantic memory?

A

Features that are common of a category, but not absolutely necessary.

E.g. birds, small, fly, eat bugs, seeds, sing.

42
Q

How do we explain Sentence-Verification Task results?

A

Via a 2-stage feature comparison process.

43
Q

What is Stage 1 of the 2-stage feature comparison process?

A

Stage 1: Rapid sampling of random features from both defining and characteristic, used to identify overlap.

44
Q

What occurs is there is large overlap during Stage 1 of the 2-stage feature comparison process?

A

Large overlap = yes response.

45
Q

What occurs is there is no overlap during Stage 1 of the 2-stage feature comparison process?

A

No overlap = no response.

46
Q

What occurs is there is moderate overlap during Stage 1 of the 2-stage feature comparison process?

A

Moderate overlap = go to Stage 2!

47
Q

What is Stage 2 of the 2-stage feature comparison process?

A

Careful checking of defining features for overlap. Reaction times are necessarily slower to account for this extra level of processing.

48
Q

What occurs is there is perfect overlap during Stage 2 of the 2-stage feature comparison process?

A

Perfect overlap = yes response.

49
Q

What occurs is there is mismatched overlap during Stage 2 of the 2-stage feature comparison process?

A

Mismatched overlap = no response.

50
Q

What are the pro’s of the Feature-List Model?

A

Can address many of the logical problems Semantic Networks Model cannot.

E.g. typicality effects.

51
Q

What are the cons of the Feature-List Model?

A

Most people can’t tell you what the defining features of concepts really are.

52
Q

What is Prototype Theory?

A

The theory that natural concepts can be understood via 3 basic principles.

53
Q

What are the 3 basic principles used in Prototype Theory?

A
  1. Fuzzy definitions.
  2. Family resemblance.
  3. Prototypes.
54
Q

What are ‘fuzzy definitions’ in Prototype Theory?

A

No necessary or sufficient features. All features can instead be used to describe somethings characteristic features.

Depending on the number a characteristics a given thing has, it can belong to a category to a varying degree.

55
Q

What is ‘family resemblance’ in Prototype Theory?

A

Category members vary in how similar they are to the best example of a concept.

56
Q

What are ‘prototypes’ in Prototype Theory?

A

The best example, or central tendency.

E.g. the ‘mean’ example.

57
Q

What is the Wharf-Sapir hypothesis?

A

The use of language shapes conceptual knowledge.