Peter Lipton Induction

1
Peter Lipton - Induction

• Diederick Spee

2
Problems concerning induction

• Justification
• Description
• Underdetermination

3
Underdetermination

• Initial conditions and rules are not enough for a
  unique solution
• There must be hidden rules

4
Justification

• Methods of inference must be taking us to the
  truth
• Not 100 per cent reliable
• Show that they usually take us to true conclusions

5
Justification

• Underdetermination shows that there are other
  worlds where our principles would fail us
• But we have good reason to believe the actual
  world is one where they are reliable.
• For a skeptical argument we also need circularity!

6
Justification

• We can not rule out the possibility of
  unreliability by using the same principles that
  are under investigation
• Descartes first meditation
• Hume

7
Justification

• Governing principle of induction is that the
  world is uniform
• Conservative induction has an impressive
  track-record
• The question is will it be succesfull, not if it
  has been!
• Conservative and revolutionary induction are both
  arguing circulair!

8
Hume

• If IP can be shown to be justified, there is an
  argument that shows it.
• Arguments are either deductively valid or
  inductive.
• No deductively valid argument can justify IP
  (underdetermination)
• No inductive argument can justify IP (because of
  circularity)
• IP cannot be shown to be justified!

9
Justification

• It looks bad!
• We can ask for justification of a particular
  principle of induction

10
Description

• One might still be interested in how our
  inferences work, even if they are unjustifiable
• There are no problems with circularity
• Description is hard! There is a general gap
  between what we can do and what we can describe
• Principles of induction are not observable, they
  must be shown indirectly

11
Description

• Many different mechanisms produce the same
  pattern, but it is often hard to come up with
  even one
• Try to describe one part of our inductive
  practice
• It is tricky! Maybe not describable in isolation
• Habit formation maybe assumed to be!

12
Description

• Attempts to tackle the descriptive problem
• More of the same
• Instantial model
• Hypothetico-deductive model
• Causal inference
• Method of agreement
• Method of difference