Computationalism: The Very Idea

1
Computationalism The Very Idea

• David Davenport
• Bilkent University,
• Ankara 06533 Turkey.
• Email david_at_bilkent.edu.tr

disclaimer (non-standard!)
2
Introduction
C O M P U T A T I O N A L I S M

• Wealth of detail
• lacks solid foundations
• Understand the big-picture
• organising principle -vocabulary

101
3
Outline

• Introduction
• What is Computation?
• What is Cognition?
• Is Computationalism correct?
• Issues
• Meaning
• Symbols Rules
• Representation
• Timing
• Concluding Remarks

4
Models

• Physical Replicas
• Answer questions
• Make predictions
• Implementation independent
• Observer-relative
• Cost, time, difficulty of production

5
Models

• Abstract computational
• Answer questions, make predictions!
• Map questions to answers
• All equivalent

af(q)
gt(q)for 2 times hs(g) gh ar(h)
ar(s(s(t(q))))
au(v(q))
6
Implementation

• Map states seq.
• Find existing system
• Construct one anew
• Use digital computer
• Rely on causation
• Program/algorithm/computation is
• abstract spec. of causal system.

7
What is Computation?

• Models
• Answer questions, make predictions!
• Physical/abstract computational
• Computational Model
• Is a function! questions to answers
• Equivalent views
• Implementation
• Find existing system
• Construct one anew use computer
• Rely on causation
• Program/algorithm/computation
• abstract spec. of causal system.

8
What is Computation?

• Answer questions, make predictions!
• A function! Abstract, but physical
  implementation.
• Program/algorithm/computation
• abstract spec. of causal system.

9
Computation
Implementation
Targetsystem
Abstraction
10
Putnams Argument

• Explain refute

11
What is Computation?

• About Prediction
• Program/algorithm/computation
• abstract spec. of causal system.
• Concept is not vacuous!
• But, it is in the eye of the beholder
• Interpretation
• State system!

12
What is Cognition?

• Agents parts of world
• Cognitive Agents use knowledge
• Prediction is the key!
• Models predict
• Choose most appropriate action based on present
  situation, past experience, knowledge of effects
  of own actions.

13
Computationalism

• Can cognition be described or simulated by
  computations?

YES

• Different levels!
• Assuming physical basis
• occurs elsewhere

14
Computationalism

• Is cognition literally computation?

PRESUMABLY

• Computation is part of cognition (the agents
  world model)
• The rest, I/O, goals, matching decision making
  mechanism?
• Semantics?
• Subjective aspects?

15
Computationalism

• Can the notion of computation help explain
  cognition?

YES

• Computation is clearly
• Of practical historic importance!
• Related to cognition
• Not vacuous!
• Does lack semantics, timing is not embedded,
  but

16
Meaning

• Representations acquire it
• From other symbols?
• Chinese Room Argument
• Foreign language dictionary
• From causal links to environment?
• Symbol Grounding, but
• State has meaning for the agent
• If it has predictive value!
• May be wrong, unreal, undemonstrated, etc.

?
?
?
?
17
Symbols Rules

• Model viewed as formal system
• Symbols rules may be
• Innate or Learnt
• Primitive symbols
• Disjunctions/Conjunctions
• Internal or External
• Rules logical mechanism!

18
Representation

• Symbolic, Connectionist,Dynamicist?
• Functional accounts ofCognition computation
• Organisation/Implementation issue.

19
Timing

• Too quick
• Too slow
• Unstable!

Maybe time is sequence!
20
Concluding Remarks

• Computationalism is right!
• Computation, is about prediction
• Cognition, demands prediction
• Computation - states representational
• only to observer
• Cognition - states representational
• to observer
• for agent itself (i.e. has predictive value.)

21
Classes of Systems
22
(No Transcript)