The final frontier: Inference to the Best Explanation (IBE)

1
The final frontier Inference to the Best
Explanation (IBE)

• The primary issue in the defence of scientific
  realism is selective scepticism epistemic
  discrimination against unobservables
  unobservable rights
• Michael Devitt (220)

2
Constructive Empiricism vs. Realism on
explanations

• Constructive Empiricist the explanatory power of
  a theory is pragmatic and not because the answer
  is compelled by the way the world is.

• The realist, of course, claims the opposite.
  Successful theories explain because they are true
  or almost true. Their argument against
  constructive empiricism rests on the use of
  Inference to the best explanation in cases of
  unobservables.

3
An example

• Consider the following example from chemistry.
• Realists would say that given all the data we
  have about chemical reactions and our ability to
  build instruments to manipulate them, the
  molecular account is literally true that there
  are different chemical atoms.
• Not only can the theory of molecules make correct
  predictions, but it does explain the phenomena.
• Hydrochloric acid is HCl and ethanol is C2H5OH,
  and there are hydrogen, oxygen and chlorine atoms.

4
Constructive Empiricism on IBE reasoning

• Consider the following example using observables
• There is a constant running sound behind the
  walls
• Bags containing food have been clawed open
• There are mice behind the wall

• Contrast the previous example with the following
  argument using unobservables
• 1) Tracks in a cloud chamber
• 2) Electrical phenomena (i.e. electricity works!)
• 3) There are electrons.

5
Constructive Empiricism on IBE reasoning

• What would be the realists position?
• The realist would say that it is true that the
  mice and electrons exist.

• What would van Fraassen say about these two
  examples?
• van Fraassen might say that mice (observable)
  exist, but not electrons (unobservable)

6
Scientific realism and inference to the best
explanation

• Realists would say IBE works for both the realm
  of observables and unobservables because the
  positive argument for realism is that it is the
  only philosophy that doesnt make the success of
  science a miracle (Putnam cited on p. 213).
• Why no miracles?

7
No miracles

• Consider biology. Not only do we have a theory
  involving cells, but we have also built
  instrumentsmicroscopes, electron microscopes,
  etcwith which to intervene successfully at the
  cellular level. It would be a case of cosmic
  coincidence for all those theoriescells,
  optics, electromagnetismto work at the same
  time.

8
Constructive Empiricists response to the no
miracles argument

• How might van Frassen respond?
• Coincidences can be explained.
• van Frassen It was by coincidence that I met my
  friend in the marketbut I can explain why I was
  there, and he can explain why he came, so
  together, we can explain how this meeting
  happened. We call it a coincidence, not because
  the occurrence was inexplicable but because we
  did not severally go to the market in order to
  meet (216)
• Is this answer satisfactory?

9
Constructive Empiricists response to the no
miracles argument

• Van Frassens response misses the point of the
  realist argumentthe repeated success of
  scientific theories cannot be explained in terms
  of the kind of coincidence cited in van Frassens
  response.

10
Another Constructive Empiricist response to the
no miracles argument

• Remember the constructive empiricist cite
  pragmatic reasons for solving underdetermination
  Their aim is only for empirical adequacy at the
  observable realm.
• So, constructive empiricists can claim that we
  have from the history of science pragmatic
  grounds for believing that the theoriescells,
  optics, electromagnetism, etc.will continue to
  be empirically adequate

11
The realist counter-argument on no miracles

• The realist could point out that the argument
  from pragmatic grounds is compatible with their
  own.
• Consider the distinction between phenotype and
  genotype.
• The constructive empiricist offers a phenotypic
  argument (surface explanation) for why certain
  theories have survived, but that is compatible
  with a genotypic argument (deep features) of why
  those theories are successful.
• Here think about a phenotypic and genotypic
  explanation for why giraffes have long necks

12
Phenotypic, genotypic,

• How would the constructive empiricist respond to
  the realist attempt to co-opt them?
• They will not be swayed by the realist attempt
  because one of the tenets of their anti-realism
  is to avoid explanations that appeal to deep
  structures (i.e. unobservable).

13
van Frassens criticisms of IBE

• van Frassen has two criticisms against IBE
• The argument from indifference given a set of
  empirically equivalent but ontologically
  incompatible theories to choose from, it is
  highly improbable that the true theory is in that
  set. So it is highly improbable that the best
  explanation is true.
• The argument from the best of a bad lot Given
  the first argument, the best explanatory
  hypothesis may just be the best of a set of false
  theories. How do we know that other hypotheses
  not being considered are not as good or better
  than what we have?

14
The realists response

• First, the set of candidates is not arbitrary.
  Theory choice includes background knowledge which
  narrows the possible candidates of hypotheses. It
  is explanatory considerations of the candidates
  that determine the best hypotheses.
• Furthermore, the same problem applies to the
  constructive empiricist. How do they know that
  the best empirically adequate theory is included
  in the set of candidates?

15
The realist strikes back selective scepticism

• Furthermore, the realist can point to a problem
  for the constructive empiricist What is the
  reason for claiming that realm of observables is
  immune to the problem of underdetermination?
• If the constructive empiricist (i.e. van
  Fraassen) says mice exist but electrons do not,
  then she is inconsistent in rejecting realists
  claims, using IBE, about unobservables.

16
Constructive empiricism and Inference to the best
explanation

• The realist argues that the constructive
  empiricist cannot make the claims they do about
  observables.
• Remember for van Frassen, dinosaurs are
  observables, even though they are not observed.
• But hypothesis about unobserved objects and
  events are just as susceptible to the problem of
  underdetermination as theories about unobservable
  objects and events
• Why?

17
The underdetermination argument

• Consider the hypothesis that a meteor struck Io,
  one of the moons of Jupiter, in 2001.
• The astronomer could point to satellite pictures
  of Io before 2001 which does not show the crater.
• But is that the only explanation?
• How about geothermal activity on Io? What about
  problems with the camera?
• How do we solve the problem of underdetermination
  in this case?

18
The underdetermination argument

• The point of the realists argument is that if
  the underdetermination argument is applied
  consistently, then we can only have knowledge of
  things that have been observed!
• Wouldnt the result be that most of what passes
  for knowledge in the sciences isnt knowledge at
  all, since they concern objects and events that
  havent been observed?
• The constructive empiricist needs to give us a
  principle for allowing her to make the claims she
  does about those observables which are not
  observed which does not apply to unobservables
  (i.e. theoretical entities).

19
The underdetermination argument

• The issue here is that for realists, if a
  particular explanation is agreed to be the best
  explanation of the phenomena in question, it is
  irrational not to adopt it because the best
  explanation is linked to truth.
• van Frassen on the other hand is presenting
  constructive empiricism not as a theory that one
  must adopt if one were rational, but rather as an
  epistemic position that may be adopted which
  accounts for all that we need to know about
  science.

20
The Constructive Empiricists response

• van Frassen likens his position to a particular
  conception of rationality.
• He contrasts the Prussian and the English points
  of view
• Prussian forbids that which is not specifically
  allowed.
• English allows anything that is not specifically
  forbidden.
• van Frassen likens his position (which he calls
  voluntarism) to the English point of view what
  is rational to believe includes anything that one
  is not rationally compelled to disbelieve (223)

21
Constructive empiricism and inference to the best
explanation

• He claims IBE may be indispensable in acquiring
  reasonable expectations and thus may be
  pragmatically indispensable (223).
• For the constructive empiricist, IBE is
  acceptable as a practice of inference for
  empirical adequacy, NOT because it is a rule of
  reasoning which leads to truth.
• Remember what Hume said about our use of
  induction.

22
Constructive Empiricism and inference to the best
explanation

• For van Frassen, IBE does not compel us to be
  scientific realists.
• The realist claims that van Frassens denial of
  IBE for unobservables is arbitrary
• van Frassen argues that to insist, as the
  realists do, to go beyond the empirical adequacy
  of theories to account for the nature of science,
  we are taking unnecessary risks (i.e. making
  claims about unobservables) for no extra
  empirical gain.

23
Constructive Empiricism and inference to the best
explanation

• What is the extra bit that scientific realism
  gives from an empiricist point of view?
• The way the world is but thats just more
  metaphysics. van Frassens rejection of realism
  rests on his empiricism (in Constructive
  empiricism) empiricists should repudiate
  beliefs that go beyond what we can (possibly)
  confront with experience, and this restraint
  allows them to say good bye to metaphysics
  (225).

24
The final word

• The realist counters that van Fraassens position
  commits him to skepticism not only of the
  unobservable realm, but also the realm of
  commonsense realism (the observables) if
  underdetermination were applied consistently.
• van Fraassens response Just as the success of
  science needs no metaphysical explanation, the
  realm of common sense, the everyday world, needs
  no metaphysical explanation.
• The everyday world is a brute fact. We do not
  need to infer or argue for its existence.

25
One final challenge

• The realist on the other hand is susceptible to a
  challenge themselves with regards to IBE. The
  challenge is as follows since it is IBE
  involving unobservables that is in question in
  the realism debate, it is circular to appeal to
  the explanatory power of scientific realism at
  the meta-level i.e. using IBE again to account
  for the overall success of science (218)
• How might the realist respond?

26
One final challenge

• Recall the discussion earlier with the problem of
  induction (section 2.2.8) can there be any
  non-circular justification of inductive and
  deductive reasoning?
• What other position is available to us?