Karl Popper’s deductivism

Popper thought science aimed at truth. He was anti-conventionalist in that regard. He held that we can never know if a theory is true. Furthermore, we can’t know if a basic observation report is true. That’s because he rejected the distinction between observational and theoretical statements held by the logical positivists.  Popper explains:

“… in a statement like ‘here is a glass of water’, necessarily transcends experience. It is due to the fact that words like ‘glass’ or ‘water’ are used to characterize the law-like behavior of certain things; which may be expressed by calling them ‘dispositional words’. Now since every law transcends experience—which is merely another way of saying that it is not verifiable … It is for this reason impossible … to define it in purely experimental or observational terms”

For Popper, a basic statement is a singular existential statement (e.g., There is a raven in spatiotemporal region k). What made a theory scientific was that it sorted basic statements into two classes: (1) basic statements that are inconsistent with the theory (potential falsifiers), and (2) basic statements that are consistent with the theory (statements that corroborate). Popper rejects confirmation, so how are we to choose basic statements?  Popper thought that the choice of the basic statement was not arbitrary: the scientific community would need to make a decision. But under Popper’s deductivism, induction couldn’t be used to justify the choice of basic statements: so the basic statements are also conjectures without confirming evidence. But if all this is right, and it comes down to a unconfirmed decisions, then Popper’s system is not much different than the “mob psychology” of Kuhn. This is an unwelcome outcome, since Popper held to the correspondence theory of truth and was anti-conventionalist. (On a conventionalist picture the “truth” would be grounded in society, rather than an external reality.)  Popper seems to realize this:

“The empirical basis of objective science has thus nothing ‘absolute’ about it. Science does not rest upon solid bedrock.  The bold structure of its theories rises, as it were, above a swamp. It is like a building erected on piles. The piles are driven down from above into the swamp, but not down to any natural or ‘given’ base; and if we stop driving the piles deeper, it is not because we have reached firm ground. We simply stop when we are satisfied that the piles are firm enough to carry the structure, at least for the time being.”

Popper rejected inductive reasoning as a part of science. Science would rest on deductive reasoning in the form of conjecture and refutation. If you’re curious as to what science might look like without induction, look no further than Popper. Popper held falsifiability to demarcate science from pseudoscience. The key to this deductivism is the falsifiability, for falsification is deductively valid:

  • If Hypothesis, then Observation.
  • Not Observation (Falsification)
  • Therefore, Not Hypothesis.

While verification is inductive and not deductively valid:

  • If Hypothesis, then Observation.
  • Observation (Verification)
  • Therefore, Hypothesis???

Post Duhem-Quine, it’s well known that theories aren’t tested in isolation. Instead, there are auxiliary hypotheses. So reconstructing the inferences above, we have

  • If (Hypothesis & Auxiliary Hypotheses), then Observation.
  • Not Observation (Falsification)
  • Therefore, Not Hypothesis.

But in order for this to be deductively valid, the Auxiliary Hypotheses need to be true. If we can’t verify the auxiliary hypotheses, then we can’t falsify the hypothesis. This is a serious problem for Popper’s deductivism.

Popper seems to think that well corroborated theories should be favored over less corroborated theories.  A corroborated theory is a theory that has passed some test. A corroborated theory is like a report card in that it only shows the past success, without recommending the theory. A confirmed theory is like a letter of recommendation.  It’s hard to see how, without induction and confirmation, Popper could justify choosing a corroborated theory over an untested theory.  After all, all we have are conjectures and refutations.   But it seems obvious that we should build a bridge with a well corroborated theory over one with an untested theory.   Popper explains:

“But I want to make quite clear that the degree of corroboration of a theory (which is something like a measure of severity of the tests it has passed) cannot be interpreted simply as a measure of its verisimilitude. At best, it is only an indicator”

I wonder what the difference is between a measure of verisimilitude and an indicator of verisimilitude. Either way, it seems that there is a dreaded whiff of induction here if past success of theories is to serve as an indicator to verisimilitude.

Godfrey-Smith, Peter. Theory and Reality.
Newton-Smith, W.H.. The Rationality of Science.
Popper, Karl. Logic of Scientific Discovery.
Sober, Elliott. Philosophy of Biology.

This entry was posted in Philosophy of Science. Bookmark the permalink.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s