Concept and Knowledge Structures
Concepts
categories by which we organize the world. These include natural (animals) and man-made (furniture) . It’s a fundamental aspect of cognition. High similarity across cultures and languages. It affects memory, thought, and language
Classical View:
Concepts are all well-defined
There are necessary and sufficient criteria to determine category membership
Example: bachelor ——— “unmarried” and “male” ——-both are necessary and jointly sufficient (but what about your infant nephew?)
Modern View:
Many concepts are fuzzy. They lack clearly defined boundaries.
Proposed by Lugwig Wittgenstein - Austrian philosopher who argued that many concepts share no one criterion. Famous analysis of the concept “game”
What is shared?
Wigenstein: think of members of a category possess a family resemblance . They resemble each other in the way that members of a family resemble each other. Membership can be thought of as being probabilistic (as opposed to absolute).
Psychological evidence
McCloskey and Glucksberg (1978) - Asked subjects questions like: Is an avocado a vegetable or a fruit? How about a tomato?
Is a lampshade an example of furniture? Are books furniture?
Results:
Subjects didn’t always agree with each other. Inter subject disagreement and was consistent with modern view.
Then they brought the subjects back again and tested them again with same questions. For many of the questions, the subjects had shifted their opinions! (intra-subject disagreement) Suggests these concepts are not very stable
Structure of Categories:
Rosch proposed that categories have a hierarchical organization
Many categories possess three levels
(1) Highest level – superordinate level
Example: musical instrument
At this level few attributes are actually shared.
(2) Middle level: basic level
Examples: drum, flute, table, lamp
The level where members are most differentiated from each other
Basic level terms tend to be short
Concepts that children learn first
Level used most frequently in conversation
(3) Lowest level – subordinate level
Examples: bass drum, snare drum; dining room table and coffee table
At this level, many attributes are shared
Typicality (again)
Rosch: some members of a category are “better” representatives than others. (robin for bird as opposed to turkey or ostrich)
These best examples are called exemplars, and we organize concepts around them. When we think about a concept in the abstract, we tend to think of its exemplars.
Example: Fruits
Knowledge Structures
Another way of organizing semantic memory
Pioneered by Frederick Bartlett (1920’s)
Reaction against Ebbinghaus’ passive approach to remembering
Bartlett: When we remember something, we reconstruct it. (actively build it from its parts)
He gave subjects stories to read. Chose stories from different cultures. Most famous example: “the War of the Ghosts” – a north american indian story
Then subjects would reproduce the story in their own words.
Changes to the story:
(1) Distortions:
Unfamiliar → Familiar:
Canoe → boat
Hunt Seal → fishing
(2) Omissions
Strange Content left out:
something black came out of his mouth = just don’t mention
(3) Additions
Reasons for why things happen:
writhing and shrieking to explain death
Results inconsistent with Ebbinghaus’ approach. If memory worked the way he suggested, we should simply remember less over time. We shouldn’t add or change or add things.
Bartletts explanation
Proposed the existence of schemata in memory (singular: schema). They are mental frameworks based on cultural experiences.
New events are fitted into these frameworks
Those that do not fit a schema are distorted or forgotten
Event schemata are referred to as scripts. Example: restaurant scripts
Strong scripts are those where order is important. Weak scripts are those where order is unimportant.
Implications of Scripts
Eyewitness Testimony (Loftus and Palmer 1974) -
Subjects saw a movie of a car accident.
Subjects asked questions afterwards including “How fast were the two cars going when they _______ each other?” (“contacted” [neutral], “collided”, “smashed”)
subjects saw the same video. They just had a different word in that blank.
may be the cause of some “recovered” memories (childhood sexual abuse)
Results
Subjects in the contacted condition estimated the speed as 32 mph.
Subjects in the smashed condition estimated the speed as 41 mph.
Subjects condition subjects were more likely to believe that there was broken glass at the scene of the accident (there was none)
Implications:
Memory is malleable or changeable and can be affected by information long after the fact.
False memories can be “implanted” in subjects (Loftus) – being lost in mall when very young.
Related articles
- Mental Imagery (andrewhoff.com)
- Models of Memory (andrewhoff.com)
- Process Model of Memory (andrewhoff.com)
- Semantic Memory (andrewhoff.com)
- Concept and Knowledge Structures (andrewhoff.com)