Theme 4: Semantic memory Flashcards
findings supporting that semantic and episodic memory are separate:
- hippocampus damage usually only impacts episodic memory, just a bit semantic
- retrograde amnesia: semantic usually intact, episodic damaged
-semantic amnesia = problems with semantic amnesia, not episodic
findings for semantic and episodic being interdependent systems
- overlapping neural networks
- tasks designed for episodic memory also involves semantic memory
collins & quillian: hierarchical network model
semantic memory = hierarchical: animal –> bird –> wings, property information is stored as high up as possible
– support: categorization task: RT should be faster for properties on the same hierarchical level, confirmed!
- limitations:
– familiarity was a confounding variable
–typicality effect
–concepts can belong to multiple categories
hierarchical model: individual differences in categorisation
determined by: ambiguity (using different criteria) and vagueness (different cutoffs)
Collins & Loftus: spreading activation model
-organization based on semantic relatedness
-in test: the more often produced is most related to category
-in the network: the closer together, the more closely related
-the spreading activation as an explanation for typicality effect = stronger spreading to closer items than further ones
-limitations: flexible model –> lack of precise predictions & information is not represented as just one node in the brain & implies that each concept has a single fixed representation
findings supporting spreading activation model:
-semantic priming/facilitation effect, faster responses for items that were related more closely
three levels within word hierarchies (roach et al)
-superordinate (item of furniture)
-basic level categories (chair)
-subordinate (easy chair)
–> preference for basic level categories
why preference for basic level categories?
provides balance between informativeness and distinctiveness
Barsalou: situated simulation theory
-processing is influenced by the context/setting
-representations depend on current goals & features of the situation
-limitations:
–exaggerates the extent to which concept processing varies across situations, however, concepts have stable core & structures are situation dependent
–several possible interpretations for the finding that concept processing typically involves perceptual and or motor features
finding for situated simulation theory by barsalou
-objects qualities not visual in real life are harder to imagine
-more accurate object categorisation when full situational information was provided
hub&spoke model
-spoke = modality specific brain areas
-hub = modality independent concept representation in the anterior temporal lobes
-limitations:
–more research needed in the info in hubs
–how is information integrated between hubs and spokes
– non consensus on number and natures of spokes
findings for hub and spoke model
brain damaged patients show category specific deficits & brain areas consistently activated were the anterior lobes
schemas, scripts and frames, stereotypes
schemas = chunk of knowledge about the world
script = knowledge about event
frame = knowledge about structure with fixed & variable information slots
stereotypes = schemas involving simplified generalisations about various groups
rationalizatoin
tendency in story recall to produce errors conform to the cultural expectation
limitations to schema theory
- schema theories are vague
-de-emphasize importance of individual differences - exaggerate the schema driven memory errors in everyday life
