|
 Current day Popperians, the “critical rationalists”, espouse the following epistemic principle CR:[i] |
(CR) it is reasonable to adopt or believe a claim or theory P which best survives serious criticism.
|
| What justifies CR? To merely declare it is a reasonable epistemic principle without giving evidence that following it advances any epistemic goals is entirely unsatisfactory, and decidedly un-Popperian in spirit. |
Alan Musgrave (1999), leading critical rationalist, mounts a defence of CR that he openly concedes is circular, admitting, as he does, that such circular defences could likewise be used to argue for principles he himself regards as ‘crazy’.
However, he also gives a subtle and clever argument that it’s impossible to do better, that such a circular defense is the only kind possible. So since we’re reading Popper this week (some of us), and since an analogous argument arises in defending principles of statistical inference, try your hand at this conundrum.
“Even if it is accepted that CR withstands criticism better than rival epistemic principles (a big ‘if’), another objection immediately presents itself. All this is circular! The critical rationalist is saying that it is reasonable to adopt CR by CR’s own standard of when it is reasonable to adopt something!” (Musgrave, p. 330)
|
|
| Nevertheless, Musgrave declares that since all epistemic principles can only be defended circularly, it is no special reason to find fault with critical rationalism! According to this, we know in advance that no epistemic principle could be faulted, thanks to the availability of its surviving a circular defence. It would follow that claims about methods are non-testable! Could this really be the long-sought for defence of Popper? |
|
| Musgrave’s Remarkable Argument |
|
| It is easier to express horror at the final destination of Musgrave’s reasoning than it is to show just where one is warranted in getting off his train of argument. His argument, while unsound, or so I argue (in Mayo 2006), is subtle and interesting. His argument is this: |
“Any general epistemic principle is either acceptable by its own lights (circularity), acceptable by other lights (hence irrational by its own lights and inviting an infinite regress), or not rationally acceptable at all (irrational again). So even though the rational adoption of CR involves circularity, this cannot be used to discriminate against it and in favour of some rival theory of rationality”. (p. 331)
|
| Although our interest is in CR, let us analyse this striking general argument. Any general epistemic principle is either |
|
| (A) acceptable by its own lights (circularity), or |
| (B) acceptable by other lights (hence irrational by its own lights and inviting an infinite regress), or |
| (C) not rationally acceptable at all. |
| So if a general epistemological principle is rationally acceptable (i.e., (C) is false), he concludes, either (A) or (B) is the case (i.e., its acceptability will be circular, or irrational and inviting a regress). |
So is there a way out? If not, what are the implications for criticisms of general statistical principles and methods? Share your thoughts.
(For my own solution, see p. 15 starting with section 4.2 of “Critical Rationalism and its Failure to withstand Critical Scrutiny” Mayo 2006.)
[i] While Popper also settled for the comparative principle CR, the principle he really wants is stronger:
“Observations or experiments can be accepted as supporting a theory (or a hypothesis, or a scientific assertion) only if these observations or experiments are severe tests of the theory—or in other words, only if they result from serious attempts to refute the theory, and especially from trying to find faults where these might be expected in the light of all our knowledge”. (Popper, 1994, p. 89)
My work attempts to provide a non-comparative account of severity, because being “best tested so far” may still be horribly tested.
- Mayo, D. (2006). “Critical Rationalism and Its Failure to Withstand Critical Scrutiny,” in C. Cheyne and J. Worrall (eds.) Rationality and Reality: Conversations with Alan Musgrave, Kluwer Series Studies in the History and Philosophy of Science, Springer: The Netherlands: 63-99.
- Musgrave, A. (1999), “Essays in Realism and Rationalism” (chapter 16), Rodopi.
- Popper (1994), The Myth of the Framework.
Error Statistics Philosophy 
I feel like if we just inserted the word “tentatively” into CR before “adopt” all would be well. But I’m a statistician and not a philosopher.
Pingback: Verbatim | Ross Neir