“Your past experiences give you no basis to predict the future.”
– Steven K. Graham
From the archives: 27th December, 1994
Tim wrote:
“The argument against the possibility of induction is Hume’s, based upon a mistaken theory of causality. The causal theory is that events cause other events, so all we can know about what event X will cause is what we observe to be the effects of X under certain conditions.”
Not at all. (Or at least not as Popper presents Hume’s arguments.) Induction is basically the process of generalizing—reasoning from instances to general cases. For example, suppose that everyone you talk to who entered the last lottery, lost. According to induction, you then would (rationally) conclude that no one won the last lottery. But that isn’t a valid conclusion, for it might be the case that the next person you talk to might be the lottery winner. Your past experiences give you no basis to predict the future. Since the supposedly rational induction is invalid, induction is found to be wanting.
For a finite case, you could conceivably check with every participant, but that would no longer be induction, but would instead be conjunction.
“Sarah, what about Robinson Crusoe? Is he incapable of knowing or learning anything alone on his island without anyone to criticize or criticize him? Or is it possible for there to be conflict between his theory and reality as well as conflict between him and another knower? Is it possible that any such conflict between his theory and reality will be settled by reality when it kills him off for being wrong? Is it possible that Mother Nature’s penalty for error is ultimately death?”
(Not Sarah, but responding anyway.) Of course Robinson Crusoe is capable of learning without anyone to criticize him. He is (or should be)1 quite capable of criticizing his own theories. And there is always the possibility of conflict between his theories and reality to help him improve his theories. And the conflict need only be as “severe” as some expectation based on his theories not being met.
Editor’s note
1. Everyone criticises their theories all the time. No one has perfect knowledge, including of how to criticise. So my interpretation of the “or should be” as a categorical imperative is presumably a misinterpretation.
See also:
- Children do not want parental coercion
- Our evolution away from coercive rules and rewards
- Rival theories about early walking and talking
Steven K. Graham, 1994, ‘Induction and criticism’, https://takingchildrenseriously.com/induction-and-criticism