Marla asked

Which philosopher believed we find new laws by guessing?

Answer by Craig Skinner

The name’s Popper. Karl Popper (1902-1994). Born in Vienna, one of the most influential and controversial thinkers of the 20th Century.

He believed that truth is elusive rather than manifest, and that we increase our knowledge not by reaching truth but by avoiding error. We proceed by trial and error. Not random trial and error – we learn from our mistakes.

So we advance by making a guess (conjecture) as to how the world works, and then check it against the world to see if the guess is supported or refuted. If supported we run with it for now. If refuted (falsified) we try again with a different or amended guess, check it, and so on.

In science, repeated application of this ‘conjecture and refutation’ procedure yields scientific laws – conjectures currently accepted because all observations are in line with them and no observation has refuted them. A law is held tentatively because a future observation or experiment may refute it. For example Newton’s Law of Gravitation held for nearly 300 years till Einstein’s theory of gravitation predicted different results, and subsequent observations favoured Einstein. However, Einstein’s theory is incompatible with quantum mechanics, so that one or both of these theories will have to give way to a better theory which unifies gravity and quantum mechanics – current guesses as to the unifying laws include string theory and loop quantum gravity.

Popper felt that refutability of a guess (falsifiability of a hypothesis) was the key distinction between science and pseudoscience. This overstates the case – there is no absolutely clear-cut demarcation – but falsifiability is certainly a key feature of a scientific hypothesis or conjecture.

Appropriately, one of Popper’s famous books on philosophy of science and the growth of scientific knowledge is titled ‘Conjectures and Refutations’, first published in 1963.

His notion of learning by trial and error or conjecture and refutation is a fair description of how organisms with brains live. Thus we constantly interpret our sensory input as a guess as to what’s going on, act accordingly, get feedback, amend the guess if necessary, act again and so on. We proceed by iterated conjecture and testing. We are hypothesis-testing machines. Sometimes called ‘Popperian machines’, as opposed to ‘Darwinian’ ones which act by random trial and error (flexibility but no learning) or ‘Pavlovian’ ones which produce automatic responses to stimuli (no flexibility, no learning).

Finally, Popper thought that induction was a myth. Rather it’s all a matter of conjecture and testing. It’s not that repetition (constant conjunction) induces expectation of more of the same (as Hume suggested was the habit of mind producing inductive inference). We don’t wait passively to let the world impose regularities on us. Rather expectation comes first, we have a propensity to see regularities, a conjecture if you like, which repetition will corroborate (or not). I think Popper is right here. In effect, past regularities have driven the evolution of adaptive hard-wired cognitive strategies, our ancestors have done the induction for us, as it were, we don’t each have to start at square one. Of course the traditional philosophical problem of induction is not about what cognitive mechanisms are behind it, but rather what justifies it, and I won’t go into that. Ultimately Popper accepted a ‘whiff of induction’ otherwise he couldn’t explain why we should prefer a theory with a good past performance (many corroborating observations, none refuting) to one with no corroborating observations.