Introduction to
Cognitive Science
Sept 2005 :: Lecture #1 :: Joe Lau :: Philosophy HKU
Topics




About this course
What is cognitive science?
Methodology
The computer model of the mind
About this course
About this course



Course coordinator
 Joe Lau
Other teachers
Course tutors


Course web site


Antonio Cheung and Elaine Lau
http://philosophy.hku.hk/courses/cogn1001
Assessment




60% exam -- 2 hours.
25% problem sets -- 5 problem sets; one for each topic.
10% tutorial participation.
5% tutorial attendance.
What is cognitive
science?
What is cognitive science?
 Answer
: Cognitive science is the
science of mind and behavior.
Feature #1


Cognitive science studies MIND AND BEHAVIOR
Mental states and processes inside the brain


The behavior caused by these processes



Emotions, knowledge of language, reasoning …
Facial expressions, speech …
The normal mind
The abnormal mind

Autism, Cotard delusion …
Feature #2

Cognitive science is a SCIENCE.


Theories and hypotheses have to be tested.
How? Check whether they can explain the
data from experiments and observations.
Why study cognitive science?
“ Man is the Measure of all Things. ”
The Mind


Intellectual value
Practical value





Education
AI and technology
Medical application
Educational value
Entertainment value!
Protagoras of Abdera
( c. 480-410 B.C.)
Methodology of
cognitive science
Methodology

Some distinctive features about research
methods and explanations in cogsci.




Brain-based explanations
Functional explanations
Interdisciplinary approach
The computational model of the mind
400 BC - Hippocrates


Founder of Western medicine
“Men ought to know that from the
brain, and from the brain only,
arise our pleasures, joys, laughter
and jests, as well as our sorrows,
pains, grievances, and tears.
Through it...we...think, see, hear,
and distinguish the ugly from the
beautiful, the bad from the good,
the pleasant from the unpleasant.”
Brain-based explanations


The mind is explained in terms of physical
processes in the brain. (pace Aristotle)
What about Dualism?



Dualism - The mind is a soul.
Not to be decided a priori. Need experiments.
BBE is the default hypothesis because it has
been more successful and has better predictive
power.
However, the brain is a very, very,
very complex system. (
100 billion neurons vs. 6 billion people
)
Some of the
connections between
the different visual
systems.
How to deal with this complex system?

1. Functional approach

Understand the functions of different systems of
the brain and see how they interact.



Visual areas, language, emotions …
Cognitive science is like reverse engineering.
2. Inter-disciplinary approach
INTERDISCIPLINARY approach

Division of labour





Psychology – cognitive psychology,
developmental psychology …
Linguistics – syntax, semantics, phonology …
Neuroscience – brain structures, localization …
Computer science – AI, computer models …
Philosophy – theoretical foundations …
The computer model of the mind


The mind is like a computer.
A distinctive feature of cognitive science.
What is the computer model?


The mind is an information processing
system.
Information processing is best explained by
computations and symbols.


Information processing in the computer =
programs operating on symbols.
Information processing in the brain = neural
computations involving mental representations.
A typical computer

Inside a computer, we have :

Symbols
e.g. HTML color codes
symbols are objects to which meaning can be assigned.

Programs
e.g. Str_replace( “I have a cat” , ”c” , ”h” )
programs are procedures for manipulating symbols.
Mental representations


Mental representations
are symbols in the
brain that have
meaning or encode
information.
Thinking P ~ Activating
a mental
representation that
means P.
Boring!
Thinking as neural computation
Is-in-love
Peter
Is-happy
Why should we accept
the computational model
of the mind?
Some reasons



Information processing does seem to be a
distinctive feature of the mind.
Mental representations are useful in
explaining lots of mental phenomena.
We can observe mental representations.
Information processing in the mind

Perception


Language use


combining different sources of information, deriving new
information, testing consistency of information, etc.
Action


making use of information about syntax, semantics and
phonology.
Reasoning


acquiring real-time information about the surrounding
environment.
making use of information in action planning and guidance.
Memory

storing and retrieving information
They help us explain lots of things

Example: Syntactic Disambiguation


“We shall discuss violence on TV.”
Two interpretations :
VP
VP
V
NP
PP
discuss
N
P
NP
V
NP
discuss
N
PP
violence P
violence
on
TV
on
NP
TV
A real representation
Topographical
representation of
visual stimulus in
area V1
Two methodological consequences of
the computer model


Computer models can be built to test
theories of mental processes.
There are different levels of analysis for
a complex information processing system.
Three Levels of Description
(David Marr)




A complete understanding of a computational
system has to involve three (kinds of) levels :
Task : what the system is capable of doing
(capacities)
Algorithm (software) : which computational
procedures are used
Implementation (hardware) : how the
computations are implemented
Example



Task: Multiplication
Algorithm: input
numbers x and y. Output
number in row x and
column y.
Implementation: human
being and paper.
1
2
3
4
5
2
4
6
8
10
3
6
9
12
15
4
8
12
16
20
5
10
15
20
25
Different algorithms
Same task
: calculate X2-1
2
 Algorithm #1 : X -1 = (X+1)•(X-1)
 52-1 = (5+1)(5-1) = 6·4 = 24
2
 Algorithm #2 : X -1 = X•X-1
 52-1 = (5·5)-1=25-1=24
 Task
Same task
Different hardware implementations
1
2
3
4
5
2
4
6
8
10
3
6
9
12
15
4
8
12
16
20
5
10
15
20
25
Exercise: Can you give an example where the same algorithm
is implemented in different hardware?
Application: linguistic understanding

Task


Algorithm


Identify grammatical structure and meaning
of speech sounds.
What kind of computation and
mental representations?
Implementation

Which part of the brain?
VP
V
discuss
NP
PP
N
P
NP
on
TV
violence
How to think about cognitive science
Language
Task
Algorithm
Hardware
Vision
Reasoning
Other
areas
If you get lost
 Which

Language, reasoning, emotions, …
 Which

mental process?
level?
Task, algorithm, neural implementation
Summary



Cognitive science as an inter-disciplinary
science of mind and behavior.
The computational approach : using
computations and representations to
explain mental processes.
Three levels of descriptions.
Finally, to end …
Scope of the computer model?


Computations are NECESSARY for
understanding the mind because the mind
processes information.
Are computations SUFFICIENT for
understanding everything about the mind?
What about
consciousness?