Popper and Hypothetico-deductivism

Popper famously declared that induction is a myth. This thesis, if true, makes nonsense of the current volume. But is the thesis true? And, before we get to that, what precisely does it mean? Popper is a deductivist. He thinks that whenever we reason, we reason deductively or are best reconstructed as reasoning deductively. Most philosophers disagree. Most philosophers think that most reasoning is non-deductive. To understand why most philosophers think this, we have to look at the functions of reason or argument, and see that deduction seems quite unsuited to serve some of those functions. What are the functions of argument? Why do people reason or argue? One function of reason or argument is to form new beliefs or come up with new hypotheses. Another is to prove or justify or give reasons for the beliefs or hypotheses that we have formed. A third is to explore the consequences of our beliefs or hypotheses in order to try to criticise them. We need a logic of discovery, a logic of justification, and a logic of criticism. It is usually accepted that deductive logic is fine so far as the logic of criticism goes. “Exploring the consequences of our hypotheses” means exploring the deductive consequences of our hypotheses. Criticism proceeds by deducing some conclusion, showing that it is not true (because it does not square with experience, experiment, or something else that we believe), and arguing that some premise must therefore be false as well. Criticism only works if the reasoning is deductively valid, if the conclusion is ‘contained in’ the premises, if the reasoning is not ‘ampliative’. That is what entitles us to say that if the conclusion is false some premise must be false as well. If our argument were ampliative, criticism would not work. Showing that the conclusion is false would not entitle us to say that some premise must be false as well. But deduction’s strength so far as criticism is concerned seems to be a weakness as far as discovery and justification are concerned. In a valid deduction the conclusion is contained in the premises, does not ‘amplify’ them, says nothing new. If we want to come up with new beliefs or hypotheses, deduction obviously cannot help us. And if we want to justify a belief, deduction cannot help us once more. Deducing the belief we want to justify from another stronger belief is bound to be question-begging. The logics of discovery and justification must be non-deductive or ampliative. The conclusions of the arguments involved cannot be contained in the premises, but must ‘amplify’ them and say something new. Or so said the critics of deductive logic, down the ages.