Transcript Introduction to Cognitive Science

Introduction to
Cognitive Science
Lecture #1 :
INTRODUCTION
Joe Lau
Philosophy
HKU
What is Cognitive
Science?

An interdisciplinary science of mind and
behavior.




It is a science
It studies mind and behavior
It is interdisciplinary
Cognitive science theories and explanations
often invoke computations and
representations.
Mental Phenomena

From the mundane ...


To the abnormal and the bizarre


Perception, language, reasoning, action, ...
Cognitive impairments ( e.g. autism,
prosopagnosia, Cotard delusion, … )
and non-human species

Animal cognition
Find your child ...
Cape Cross, Namibia
Why computation?

Assumption #1 :


Assumption #2 :


Mental processes involve complex information
processing.
Complex information processing requires
computations.
Conclusion :

Computations are necessary for explaining
mental processes.
What is computation?

Computation is (roughly) a rulegoverned process which manipulates
representations.

We have strong evidence that the brain is
capable of carrying out massively parallel
computations.
Mental representations


MR = objects or states in the brain
which encode information.
Don’t confuse :
TREE
Mundane reminder


“Tree” is a word, a representation.
It is a representation of a tree, a living
thing, not a representation of “tree”.
TREE
represents
Role of mental representations

Information has to be physically encoded so
that they can be manipulated.



Memory / knowledge : storing representations
Thinking : causal sequence of representations
Two examples
Visual Perception
Topographical
representation of
visual stimulus in
area V1
Syntactic Disambiguation


“We shall discuss violence on TV.”
Two interpretations :
VP
VP
V
NP
PP
discuss
N
P
NP
violence
on
NP
V
TV
discuss
N
violence P
on
PP
NP
TV
Three Levels of Description




Three (kinds of) levels in describing a
computational system :
Task : what the system is capable of
doing (capacities)
Algorithm (software) : which
computation procedures are used
Implementation (hardware) : how the
computations are implemented
Why Cogsci is Interdisciplinary

Horizontal and vertical diversity in
mental capacities :



The mind can carry out lots of different
tasks in different areas.
Each of these capacities can be studied at
different levels.
Cogsci and psychology
Relevance of cogsci




Scientific understanding
Education
Psychiatry
IT, AI



Design of computer interface
Voice recognition, data mining
Cybernetics
Necessity and Sufficiency


Computations might be a necessary part of
most if not all explanations of mental
processes.
But computations themselves might not be
sufficient (enough) to explain all mental
processes.


Maybe some special features of the mind are due
to neuro-physiological properties.
Sleep? Hormonal effects?
Possibility of AI


AI = artificial intelligence
Computations might still be sufficient
for mentality even if some aspects of
the human mind can only be explained
neuro-physiologically.
Three Problems

Philosophers’ three major concerns :



Intentionality
Consciousness
Freewill
Intentionality

Intentionality = aboutness, meaning,
content



Language, knowledge, reasoning, beliefs,
perception
The belief that 2>1 is a belief about
numbers.
To be explained in terms of mental
representations.
Phenomenal Consciousness


Feelings, sensations, experience
Some mental states are both conscious
and intentional

e.g. conscious thinking
Freewill

What is freewill?



(A) Capable of making decisions + (B) ???
Presumably (A) can be explained
computationally.
What more is required?
Can computers do X?


“Computers cannot have emotions /
creativity / understanding / humour …”
To decide whether computers can have
a mental state X, we need to :


Identify the conditions required for having
X.
Decide which of the conditions are easy /
difficult to be satisfied, and how.
Can computers be creative?

What are the preconditions for
creativity?



Generating ideas and hypotheses
Selecting and modifying the useful ones.
Both cognitive / intentional processes

No in-principle obstacles?