Dissertation, The University of Queensland (1989)
Popper's account of refutation is the linchpin of his famous view that the method of science
is the method of conjecture and refutation. This thesis critically examines his account of
refutation, and in particular the practice he deprecates as avoiding a refutation. I try to
explain how he comes to hold the views that he does about these matters; how he seeks to
make them plausible; how he has influenced others to accept his mistakes, and how some
of the ideas or responses to Popper of such people are thus similarly mistaken. I draw
some distinctions necessary to the provision of an adequate account of the so-called
practice of avoiding a refutation, and try to rid the debate about this practice of at least one
red herring. I analyse one case of 'avoiding' a refutation in detail to show how the
rationality of scientific practice eludes both Popper and many of his commentators.
Popper's skepticism about contingent knowledge prevents him from providing an
acceptable account of contingent refutation, and so his method is really the method of
conjecture and conjecture. He cannot do without the concepts of knowledge and refutation,
however, if his account of science is to be plausible or persuasive, and so he equivocates
between, amongst other things, refutation as disproof and refutation as the weaker notion of
discorroboration. I criticise David Stove's account of this matter, in particular to show how
he misses this point. An additional advantage Popper would secure from this equivocation
is that if refutations were mere discorroborations they would be easier to achieve, and
hence more common in science, than is the case. On Popper's weak notion of refutation, it
would be possible to refute true theories since corroboration does not entail truth.
There are two other related doctrines Popper holds about refutation which, if accepted,
make some refutations seem easier to obtain than is the case. I call these doctrines 'Strong
Popperian Falsificationism' (SPF) and 'Weak Popperian Falsificationism' (WPF). SPF is
the false doctrine that if a prediction from some theory is refuted then that theory is refuted.
Popper does not always endorse SPF. In particular, when confronted with a counterexample
to it, he retreats to WPF, which is the false doctrine that if a prediction from some
theory is refuted then that theory is prima facie refuted. WPF , or even SPF, can seem
plausible if one has in mind predictions derived from theories in strong or conclusive tests
of those theories, which I suggest Popper characteristically does.
Popper is disposed to describe any such case of predictive failure which does not lead to
the refutation of the theory concerned as one in which that refutation has been avoided. To
reinforce his portrayal of the refutation, or the attempted refutation, of major scientific
theories as the rational core of scientific practice, Popper treats the so-called practice of
avoiding a refutation as untypical of science, and much so-called avoidance he dismisses as
unscientific or pseudo-scientific. I argue that his notion of avoiding a refutation is
incoherent. Popper is further driven to believe that such avoidance is possible, however,
because he conflates sentences with propositions and propositions with propositional
beliefs. Also, he wishes to avoid being saddled with the relativisim that is a consequence
of his weak account of refutation as discorroboration.
Popper believes that ad hoc hypotheses are the most important of the unscientific means of
avoiding a refutation. I argue that his account of such hypotheses is also incoherent, and
that several hypotheses thought to be ad hoc in his sense are not. Such hypotheses appear
to be so largely because of Popper's use of rhetoric and partly because these hypotheses are
unacceptable for other reasons. I conclude that to know that a hypothesis is ad hoc in
Popper's sense does not illuminate scientific practice. Popper has also attempted to
explicate ad hocness in terms of some undesirable, or allegedly undesirable, properties of
hypotheses or the explanations they would provide. The first such property is circularity,
which is undesirable; the second such property is reduction in empirical content, which is
not. In the former case I argue that non-circularity is clearly preferable to non-ad hocness
as a criterion for a satisfactory explanation or explanans, as the case may be, and in the
latter case that Popper is barking up the wrong tree.
Some cases of so-called avoidance are obviously not unscientific. The discovery of
Neptune from a prediction based on the reasonable belief that there were residual
perturbations in the motion of Uranus is an important case in point, and one that is much
discussed in the literature. The manifest failure of astronomers to account for Uranus's
motion did not lead to the refutation of Newton's law of gravitiation, yet significant
scientific progress obviously did result. Retreating to WPF, Popper claims that Newton's
law was prima facie refuted. In general, astronomers have never shared this view, and they
are correct in not doing so. I argue that the law of gravitation would have been prima facie
refuted only if there had been good reason at the time to believe as false what is true,
namely, that an unknown trans-Uranian planet was the cause of those Uranian residuals.
Knowledge of the trans-Uranian region was then so slight that it was merely a convenient
assumption, one which there was little reason to believe was false, that the known
influences on Uranus's motion were the only such influences. I conclude that in believing
or supposing that it was this assumption that was false, rather than the law of gravitation,
Leverrier and Adams, the co-predictors of Neptune, were acting rationally and intelligently.
Popper's commentators offer a variety of accounts of the alleged practice of avoiding a
refutation, and of this case in particular. I analyse a sample of their accounts to show how
common is the acceptance of some of Popper's basic mistakes, even amongst those who
claim to reject his falsificationism, and to display the effects on their accounts of this
acceptance of his mistakes. Many commentators recognize that anomalies are typically
dealt with by changes in the boundary conditions or in other of the auxiliary propositions
employed. Where many still go wrong, however, is in retaining the presupposition of WPF
which encouraged Popper to hold the contradictory view about anomalies in the first place.
Thus Imre Lakatos and others, for example, have developed a 'siege mentality' about major
scientific theories; they see them as under continual threat of refutation from anomalies,
and so come to believe that dogmatism is essential in science if such theories are to survive
as they do. I examine various such doomed attempts to reconcile Popper with the history
of science. It is a common failure in this literature to conflate or to fail to see the need to
distinguish a belief from a supposition, and an epistemic reason from a pragmatic reason. I
argue that only if one does draw these distinctions can one give an adequate account of
how anomalies are rationally dealt with in science.
The other important strand in Popper's thinking about 'avoidance' of refutation which has
seriously misled some of his commentators is his unfounded belief in the dangers of ad hoc
hypotheses. I examine the accounts that a sample of such commentators provide of the
trans-Uranian planet hypotheses of Leverrier and Adams. These commentators imply or
assert what Popper only hints at, namely, that there is something fishy about this
hypothesis. I provide a further defence of the rationality of entertaining this hypothesis at
I conclude with a few remarks about Popper's dilemma in respect of scientific practice and
his long standing emphasis on refutations.