Knowledge Flashcards
Conceptual Knowledge
Allows us to recognize objects and categorize them
Allows you to know how to interact with things
Make inferences about them
What is a Concept?
The mental representation of
a class or individual and the categories of objects, events, and abstract ideas
Category
Includes all possible items that fall under a certain concept
“pointers to knowledge”
Name the functions of concepts
Classify into categories
Allows us to communicate with each other – we all have similar concepts and make similar inferences
To learn concepts by association (generalization)
Create new knowledge by putting together existing concepts
Family resemblence
Members of a category resemble each other in a variety of ways
more resemblance to your category = closer association (Typicality)
Typicality
An effect
Sentence verification task
most common members of a category should have faster RT
Sentence Verification Test
People identify whether sentence is accurate (Yes or no)
Naming
People tend to name common members of a category first
then name less common
Priming
If a cue for a stimulus increases one’s ability to recognize it increases
Definitional Approach Theory
States that we have definitions of what belongs in a category and not in others
*NOT most supported, lacks specificity, works best for geometric objjects
Feature Set Theory
Suggests we compare the features of an item with the features of a category
Defining features and characteristic features
Defining Features
Member must have these features to belong to a category
e.g. Cat MUST have whiskers
Characteristic features
Typical features of a category, not necessary to have
e.g. fur and a cat (hairless)
Prototype Theory
Suggests that we don’t focus on features but focus on the ideal prototype for a category
Explains typicality (still lacks some clarity)
Prototype and schematic average
Deals with fuzzy boundaries
Prototype
Idealized form of the whole category, not present in real life (idealized)
Schematic Average
Average of every feature, characteristic, and dimension of all members of the category – If an item is sufficiently similar to this, we can categorize it as the same
Exemplar Theory
Suggests that we compare instances to actual exemplar of the category
Exemplar
An actual instance of a member in that category– compare until a match is found or list is exhausted
Is often an average member of the group
ALL categorization is done by…
Episodic memory
Which theories of categorization have most support?
Prototype and exemplar
Both account for all important findings in different ways
Which is the categorization theory is most likely to be used when learning a new category?
Prototype
Which categorization theory is used when you know the category is known well?
Exemplar
Levels of categorization
Superordinate
Basic
Subordinate
Superordinate
Broadest (furniture or animal)
Basic
Instances
Chair or dog – typically used in teaching
Subordinate
Subtypes
e.g. beanbag chair or doberman
Two main principles of categorization
The larger the category, the more searching involved and longer the response time (subordinate takes longer)
More levels, more searching involved
Explain: More levels to go through, the more searching involved
RT is longer if you go up two levels rather than one
Concepts: Experience and knowledge
The more experience with a concept, more specific categories used
Networks
Things that are linked and connected (usually in a meaningful and effective way)
Stronger connection = faster RT
Semantic Networks
States that information is organized semantically (by meaning)
Information at different levels is responded to differently
“Is a” levels - longer RT due to processing
Criticisms of Semantic Networks
Does not explain typicality
Some hierarchal orders are not quicker, and doesn’t account for everything
Connectionist Networks
AKA parallel processing
Views all information as connected and accessed by a pattern of activation
Similar to computers
Parallel, distributed, nodes/units
Connections are weak or strong
Parallel connection
Not serial
Happens all at once
Distributed Connection
Not localized
Node/unit connection
Bits of info
excitatory or inhibitory
Graceful Degredation
Occurs when the input signal is incomplete (degraded)
Or damage within the system (brain damage)
Can still activate if not entirely there, but there is a point
Generalization
Knowledge about one thing can generalize to another – e.g. fruit, a lot of berries are sweet and yummy, you haven’t had a blackberry, assume they are the same.
Back propagation
When an error is made, the signal gets transmitted back through the network
effects weight of connections
efficient way of learning by unlearning incorrect info
Pros of Parallel processing/connectionist networks
Pros:
Mimics actual brain activity
Explains how fast process occurs
Explains how incomplete information can be recognized
Cons of parallel processing/connectionist networks
Some processes occur serially
Relationships among nodes are unclear
To understand brain damage due to encephalitis:
Loss of ability to name objects in categories
Ability to identify objects but not living things
Sensory-Functional Hypothesis
Some people with damage can have category-specific memory impairment
Sensory - Animal
Functional - Artifacts/tools
Have a problem with sensory or functional
Semantic category Approach
Some brain areas are responsible for different categories (identified based on meaning)
Networks may have evolved to promote survival
circuit for living things could be damaged
Multiple-Factor Approach
Conceptual organization
How similar the members of category are
Telling from leopard to a cheetah
E.g. animals are identified by movement and colour
Objects associated with the actions performed with them
Mechanical objects = both
Crowding
How similar properties are between the members of the categories
makes differences more difficult
Embodied Approach
Suggests that our knowledge of the concepts is based on the reactivation of the sensory and motor processes that occur when we interact with the object
Bring up experience
more distinct with artifacts than living things
