Types of knowledge
- Procedural
- Declarative
Procedural knowledge
- How
- Practice
- Hard to extinguish
Declarative knowledge
- Facts, events
- Acquire with organization
- Easily forgotten
Types of declarative knowledge
- Episodic
- Semantic
Episodic knowledge
- Autobiographical
- Temporal tages
Semantic knowledge
- Factual, words, concepts
- Not referenced to own experiences
Explicit episodic learning vs implicit procedural learning
- Compared normals vs amnesics
- Move disks from one pillar to another, one disk at time, disks can’t go on top of smaller disk
- Normals learn procedure, transfer to other tasks, have episodic memory
- Amnesics learn task, show transfer, no episodic memory
- Shows difference between explicit (episodic) memory and implicit (procedural) knowledge
Henry Molaison
- Retreival of information from LTM intact
- STM intact
- Could not transfer new declarative info to LTM
Mirror tracing task
- HM no recollection of doing task one day to the next
- HM improved in this motor learning task
- Evidence for implicit learning and memory
Semantic memory
- Permanent memory of general world knowledge
- Uses different brain processes than episodic memory
- Semantic memory highly organized
Models of semantic memory
- Semantic feature-comparison model
- Hierarchical network model
- Spreading activation model
Semantic feature-comparison model
Structure
- Knowledge consists of sets of features
- Features weighted on dimensional space
Decision process
- Stage 1: global feature comparison to give high overlap, low overlap, or inbetween
- Stage 2: Only if inbetween, spends more time deciding if enough features overlap
Defining features
Critical, core featuers defining a category
Characteristic features
Less important features
Problem for semantic feature-comparison model
- Disconfirming sentences
- “Collies are birds” stage 1
- “Collies are poodles” stage 1 + 2
Hierarchical network model
Structure - Semantic memory is a vast collection of associated nodes (concepts) - Hierarchy - Cognitive economy Processing - Info retrieval - self-terminating search
Hierarchy
Higher up is more inclusive
Cognitive economy
Common information is stored at only one level
Problem with hiearchical network model
Typicality effects - Why faster to respond to typical vs atypical things (dove vs chicken)
Spreading activation model
Structure
- Not hierarchical
- Related nodes are connected
- Closer things are, more they are related
Process
- Activating one node causes spread of activation to related nodes
- Closer things are semantically, quicker response time is
Lexical decision task
- Given 2 words, are they both words
- Response time for words low to high: related, dinstantly related, unrelated
