In a not-so-famous passage in his famous paper “The New Riddle of Induction” (which gave us the words Grue and Bleen) Nelson Goodman claimed that:
That a given piece of copper conducts electricity increases the credibility of statements asserting that other pieces of coper conduct electricity, and thus confirms the hypothesis that all copper conducts electricity. But the fact that a given man now in this room is a third son does not increase the credibility of statements asserting that other men now in this room are third sons, and so does not confirm the hypothesis that all men in this room are third sons.
But suppose you got lost in a hotel room and stumbled into a room filled with one hundred men, and you begin mingling, pretending you belong. If the first four men you talked to all told you they were third sons, this would be plenty of reason to suspect that perhaps you hand stumbled into the yearly convention of the Third Sons Society.
Strictly speaking, it’s not clear that Goodman is committed to denying this. Maybe he only wants to deny that finding out about the first thrid son would confirm the convention hypothesis, while he would admit that meeting the second, thid, and fourth third sons would confirm the hypothesis. But if that’s right, it’s easy to wonder if he might be wrong about the very first piece of conductive copper–maybe that also fails to to confirm that copper conducts electricity.
The idea that the first instance of a hypothesis is no confirmation of it is awfully tempting. On the other hand, the idea that there can’t be a difference in significance between the first and second confirmation is also tempting. Yet together those ideas would imply we can never confirm any hypothesis at all.
In his article, Goodman suggests one reason why we think finding out that one piece of copper is conductive is confirmation: our knowledge that samples of the same substance tend to have similar conductivities. This strikes me as basically correct, though Goodman says it fails to address his worries for reasons I don’t quite fathom. I can think of one definite worry about this approach, though: while it does help with cases where we have some background information about what we’re dealing with, it might seem to be unhelpful when we’re tying to figure out the most basic features of the universe, for which no background info is really relevant (i.e. what background information could be relevant to knowing how electrons might behave?)
That’s an interesting worry at the end.
How do you think we should determine whether or not background knowledge is relevant? I’m not sure I see any obvious reason why studying the most basic features of the Universe precludes the use of background knowledge. Why couldn’t it be, for example, that basic features behave rather like more complicated features?
I should say I agree with you that our background knowledge is not relevant in the case of basic features of the Universe (or, say, large scale items like the Universe as a whole). But I wonder what could underpin this intuition.
Doesn’t the possibly random nature of whether or not a man is a third son place it in a different category from whether or not a metal has the physical property of conducting electricity? Or am I missing the point?
Paul: Ordinarily, whether or not a man is a third son *is* random. But, there are circumstances where it could be non-random, as in the case of a Third Sons’ Convention. On the other hand, while the conductivity is not in fact random, this doesn’t seem to be something we can know without going and measuring a bunch of conductivities. Its seems possible that conductivities could very at random, it’s just that science tells us they don’t.