Cognition,
Language, and
Intelligence
Cognition
What is Cognition?
• Literally “thinking”
• The brain’s representations of information
in the world around us
• The way the mind processes this
information
• Sensation → Perception → Cognition
Categories
Categories: mental groupings of similar objects, events, and people.
“Chair”
Categories
 Organize lots of information
with a minimum amount of effort
 Allow us to solve problems
quickly
What determines category membership?
Aristotle: Defining features
 look for features that every member of the that
category possesses
 Either is or is not a member of a category
Examples
 Triangle
 Bachelor
But do categories really fit this definition?
“Chair”
What is common to all of them?
Wittgenstein: “Family resemblance”
Eleanor Rosch studied how people actually use categories:
Prototype model
Category membership determined in terms of similarity to
“prototypical” exemplars
Examples
• Chair
• Red
Prototype model
Categories are not “all or none”
Instead, they have graded structure
Good example
Poor example
Prototype model
Categories have “fuzzy” boundaries
Chair?
Chair?
Red?
Red?
Chair?
Red?
Chair?
Chair?
Red?
Red?
Prototype properties apply to
Natural categories:
e.g., Birds, trees, red
Artificial categories:
e.g., Chairs, Aircraft, Weapon
What’s the category?
•
•
•
•
•
•
•
Human infant
Photo albums
Jewelry
Dog
Insurance documents
Data discs
Tapestries
What’s the category?
Human infant
Photo albums
Jewelry
Dog
Insurance documents
Data discs
Tapestries
“Things to take from a burning house” (Barsalou, Ad Hoc Categories)
“things to take on a camping trip”
“ways to make friends”
“possible costumes to wear to a Halloween party”
“emergency firewood”
 Categorizations made with respect to a goal
 Things can be cross-categorized
Levels of Categorization
Superordinate
Animal
Furniture
Plant
Vehicle
Basic
Fish
Chair
Tree
Train
Subordinate
Trout
Rocking Chair
Oak Tree
Freight train
Superordinate: Broadest and least specific category. (e.g., animal, furniture)
Basic: More specific, the level of prototypes. A common level of
categorization. (e.g., fish, chair)
Subordinate: Most specific type of category, it has the most distinguishing
about of information. (e.g., trout, rocking chair)
Bandwidth-Fidelity Tradeoff
Bandwidth
• Number of members included
• Superordinate categories have the highest bandwidth
Fidelity
• Number of properties
• Subordinate categories have the highest fidelity
Basic level categories
• Represent best solution to Bandwidth-fidelity tradeoff
• Generated first in naming tasks
• Learned first by children
Language
• Language
– our spoken, written, or gestured words
and the way we combine them to
communicate meaning
Animal language
Why study language in animals?
• Communication with animals allows us
to learn more about other species
• By examining differences across
species we can learn about language
Language vs. Communication
Many animals (especially social animals) communicate
Examples
• Bees—“dances” specify distance, direction, & quality of food
• Vervet monkeys—Different signals for different predators
• Parrots
=> animals can communicate. BUT can they use language?
What is language?
Two conditions must be met:
SEMANTICS - ARBITRARY UNITS (WORDS)
WHICH HAVE MEANINGS.
 Hissing cat?
 Vervet monkey?
SYNTAX - WORDS ARE ORGANIZED
TOGETHER ACCORDING TO RULES
Chomsky: it is syntax that is innately human. Animals
can learn words (perhaps) but cannot have a grammar.
Experimental Evidence: can animals be
taught to use language?
William Furness (1916) –
taught orangutan to speak
English
3 words: "Cup" "papa" and
"th"
Hayes's (1951): raised a chimp-Vicki-as a human child
• tried to teach her English
4 words: "mama" "papa" "cup" & "up"
• perhaps chimps do not have coordination and control
required to produce spoken language.
Gardners (1969) raised chimp—Washoe—as a human child
• Tried to teach A.S.L. (has discrete units and a grammar)
• Washoe lived in a trailer behind house for 5 years
• Had human companions during all her waking hours
Washoe
• Not specifically rewarded for performing a sign
• After 4 years had developed 132 A.S.L. signs
• apes can meet 1st criterion for a language - use arbitrary
symbols as referents for objects (semantics)
BUT could apes do syntax? (organize words together according
to grammatical rules)
• Not clear whether Washoe could do this
• Washoe prompted investigators to look at syntactic competence
• Premack trained female chimp, Sarah, to use plastic chips to
communicate.
• E.g., Blue triangle = apple
• Was required to give signs in correct order
• Rumbaughs trained several chimps to use “lexigrams” to
communicate with each humans & with each other
=> But evidence controversial
Used caged chimps
Rewarded animals
Nim Chimpsky
• 125 ASL words
• Made many 2-word combinations
• But did not move on to more complex sentences
• 3-word sentences had no more info than 2-word sentences
• Most utterances prompted by experimenter
• Unlike human infants
• Chimps merely repeating what E signed
• Chimps prodded by experimenters
Rumbaughs
• Examined Bonobos (Kanzi)
• Learned early in life
• Comprehension of 2.5 year old
• Well controlled studies
The experimenter prepares to test Kanzi by placing
headphones on his head.
Kanzi listens for spoken English, which comes from another room.
Kanzi generally displays a very intent expression at the moment
he recognizes the word.
The experimenter prepares to open a "test booklet" which hold 3
items, one of which corresponds to the word that only Kanzi can
hear.
Kanzi gestures with his left hand in a upward motion to signal that
he is ready for the test booklet to be shown to him.
Kanzi looks over the photos in the test booklet
Kanzi selects a photograph to indicate which English word he heard
through the headphones.
The experimenter congratulates Kanzi on his correct choice.
Human Language
• Language related to our thought processes
• Behaviorists: thought = language
But
• People with vocal cords paralyzed can think
• So can children and nonhuman adults
• So can brain-damaged adults
• Today, relation between thought & language more complex
=>Language facilitates thought
• Some people try to get us to think about differently by using
different words.
E.g., politicians: “revenue enhancement”= tax increase
Human Language
Other Euphemisms
• Oral hygiene appliance
• Media courier
• Negative patient care outcome
• Vertical transportation corps
• Social expression products
• Preemptive counter attack
• Given an alternative career opportunity
• Being economical with the truth
• Unscheduled terrain encounter
Human Language
But do different languages facilitate different thoughts?
E.g., Dani people of New Guinea
• 2 basic words for color: mili & mola
• So do they have difficult discriminating colors?
• “Whorf” hypothesis: Language determines the way we think
=> people who speak different languages think differently.
=> Dani should not be able to think about different types of
colors.
• BUT research (Rosch) shows Dani make as many
discriminations between colors as people who have many more
basic color terms
Language Development
• Genes design the mechanisms for a language, and
experience activates them as it modifies the brain
Environment:
spoken language
heard
provides
input to
Genes
Brain:
design
mechanisms for
understanding and
producing language
Behavior:
mastery of
native
language
Learning Foreign Languages
Percentage
correct on
grammar
test
100
New language
learning gets
harder with age
90
80
70
60
50
Native 3-7
8-10 11-15 17-39
Age at school
Intelligence
What is intelligence?
• Cognitive ability → the ability to perform
well in cognitive tasks
• The ability to use knowledge, solve
problems, understand complex ideas, learn
quickly, and adapt to environmental
challenges
Issues in the study of intelligence
• Can intelligence be quantified?
• Is it fair to use measures of intelligence in
situations that can affect people’s life
outcomes?
• Is intelligence more determined by genetics
or environmental factors?
• Are their racial differences in intelligence?
Francis Galton
British scientist who wrote
“Hereditary Genius” in 1869,
suggesting that genius is heritable
because it runs in families
1. Intelligence is quantifiable (i.e.,
numerical values can be assigned
to distinguish between people’s
level of intelligence)
Frequency
2. Differences among people
form a Bell shaped curve
Low
Intelligence
High
Francis Galton
3. Intelligence can be measured by
objective tests
4. People’s scores on different tests are
correlated
Founded “Eugenics”: advocates improving human species by
applying evolutionary theory to encourage “biologically
superior” people to interbreed
Alfred Binet
France, 1904:
Commissioned to develop
techniques to identify
children in need of special
education
Designed first intelligence test (series of short tasks not
relying on reading skills)
Did not believe intelligence was fixed or inherited
Lewis Terman
Stanford University, 1916:
Brought Binet’s
intelligence test to US →
Stanford-Binet test
Designed to be administered in a standardized fashion to
large numbers of people
E.g., Immigrants at Ellis Island; 1.75 million soldiers in
WW2
Immigrants measured according
to country of origin
Nordics > Southern Europeans & Slavs
> Blacks
=> stricter immigration policy in 1924
But this was overlooked:
Blacks from Northern states > Blacks from Southern states
Europeans in US for longer > those in US for shorter period
Examples of items on Stanford-Binet Test
Age
Sample test item
2
Here are some pegs of different sizes and shapes.
See whether you can put each one into the correct hole
4
Why do people live in houses?
6
Here are some candies. Can you count how many there are?
10
Why should people be quiet in the library?
12
What does “regret” mean?
14
What is the similarity between high and low?
What does an IQ test reflect?
Intelligence Quotient (IQ) = Mental Age
X 100
Chronological Age
E.g., If you had a mental age of 10 when you were 8 years old, you
IQ would be 125
IQ = M.A. X 100 = 10 X 100 = 125
C. A.
8
IQ range
100
90-110
above 120
below 70
average
normal range
superior
developmentally disabled
What does an IQ test reflect?
An example of an IQ test (though not a good one) is the SAT
2 sets of multiple choice items (verbal & quantitative)
Can we improve our SAT scores with coaching?
Yes, but not by much: 10-20 points on the normal 200-800 range
What do we mean by “intelligence?”
• What does the Stanford-Binet test actually
measure?
• What do we want to measure with IQ tests?
• What do modern IQ tests measure?
• Are IQ tests valid? (Do they measure what
they are intended to measure?)
The “g” factor
Spearman
Logical
s
Mechanical
s
g
Arithmetical
s
Spatial
s
What does “g” determine?
•
•
•
•
•
•
Educational achievement
Occupation
Income level
Probability of divorce
Health
Longevity
What determines “g”?
•
Which of the following factors do you think are associated with schoolaged children’s performance on standardized test scores?
– The child has highly educated parents
– The child’s family is intact
– The child’s parents have high socioeconomic status
– The child’s parents recently moved to a better neighborhood
– The child’s mother was thirty or older at the time of her first child’s
birth
– The child’s mother didn’t work between birth and kindergarten
– The child had low birthrate
– The child attended Head Start
– The child’s parents speak English in the home
– The child’s parents regularly take him to museums
– The child is adopted
– The child is regularly spanked
– The child’s parents are involved in the PTA
– The child frequently watches television
– The child has many books in his home
– The child’s parents read to him nearly every day
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
The child has highly educated parents
The child’s family is intact
The child’s parents have high socioeconomic status
The child’s parents recently moved to a better
neighborhood
The child’s mother was thirty or older at the time of
her first child’s birth
The child had low birthweight
The child attended Head Start
The child’s parents speak English in the home
The child’s parents regularly take him to museums
The child is adopted
The child is regularly spanked
The child’s parents are involved in the PTA
The child frequently watches television
The child has many books in his home
The child’s parents read to him nearly every day
What determines “g”?
• Genetics?
• Environment?
• Genetics + environment
(education + nutrition +
culture + family + brain
exercise…)
• But how great is the role
of genetics?
Genetic influences on intelligence
Monozygotic
(identical) twins
Dizygotic
(franternal) twins
Are there racial differences in
intelligence?
• Reasons to consider the question:
–
–
–
–
–
Genetics do affect intelligence
Races differ genetically, though not by much
Reliable group differences have been found
What causes group differences?
It is a testable question
Are there racial differences in
intelligence?
Reliable differences have been found on IQ scores between racial
groups. In America today:
Average IQ score
Blacks
85
Hispanics
94
Whites
100
Asians
106
Ashkenazi Jews
109
[keep in mind that these scores are averages for entire population -some Blacks score higher than some Asians on IQ tests]
Herrnstein & Murray (1994): “The Bell Curve”
Evidence from large group of white subjects, followed for 10 years
high IQ scores predicted getting into a good college, success
in college, getting a good job, and success at the job
low IQ scores predicted having illegitimate children,
going on welfare, and committing a crime
Pointed out 2 things we have already noted:
Intelligence is heritable
There are differences in IQ between racial groups
Conclusion: racial differences in IQ are due genetic differences and
that these genetic differences explain behavioral differences between
races.
Possible explanations for observed racial
differences in IQ scores
1. Blacks are genetically predisposed to have lower IQ
BUT within-group differences do not explain between group genetic
effects
Variation within group
Variation within group
Poor soil
Fertile soil
Differences between groups
Possible explanations for observed race differences
in IQ scores
1. Blacks are genetically predisposed to have lower IQ
2. Blacks & whites are raised in different environments
(physical & psychological)
When blacks & whites matched for socioeconomic factors,
IQ differences diminish
But could matching ever be complete?
Claude Steele
Stanford University
“Stereotype threat”: fear of doing something that
would inadvertently confirm a negative stereotype
Experiment:
IQ-matched Black & white subjects take difficult test
Condition A
Told nothing
Condition B
Told no race differences
Blacks do relatively
poorly
No differences between
Blacks & Whites
Possible explanations for observed race differences
in IQ scores
1. Blacks are genetically predisposed to have lower IQ
2. Blacks & whites are raised in different environments
(physical & psychological)
3. Tests are racially biased
Gosling’s Intelligence Test (GIT)
1. In a game of cricket, David scored all of his 22 runs from sixes and
boundaries. He got one six. How many boundaries did he score?
2. Who designed St. Paul’s cathedral in London?
3. What is the official address of the British Prime Minister?
4. Red is to the Central Line as
is to the Northern Line
5. Neil Kinnock was to Margaret Thatcher as
is to Tony Blair.
6. In David’s second innings he had to run all 5 of his runs. How
many yards did he run?
7. What kind of programs would you expect to hear on BBC radio 3?
8. David Gower was to cricket as Will Carling was to
?
9. Andrew is to Scotland as
is to Wales
10. If Ophelia traveled at the speed limit on a motorway for two
hours, how many miles will she have traveled?
Possible explanations for observed race differences
in IQ scores
1. Blacks are genetically predisposed to have lower IQ
2. Blacks & whites are raised in different environments
(physical & psychological)
3. Tests are racially biased
In terms of common IQ tests, evidence for bias is
mixed – most modern tests are not biased, though
historically, many tests have been
Conclusions: Genetic factors clearly play a role in people’s
intelligence, but there is no clear evidence that genetic factors account
for observed IQ differences in Black & White Americans
Genetic Relatedness and IQ Correlation
Text
Text
Text
Descargar

Cognition, Language, and Intelligence