Lynch on Horwich

Michael P. Lynch's response to Paul Horwich is good. He writes:

maybe truth doesn’t have just one nature, or none, but more than one. If so, then the sin of traditional views is precisely taking a good idea and overgeneralizing it. But that doesn’t mean that there is nothing more to say about truth. Far from it — there is lots to say, it just depends on what we are talking about. So no uniform reductive explanation perhaps, but illumination just the same.

And:

Locke’s view that there are human rights, for example, didn’t leave the world as it was, nor was it intended to. Or consider the question of what we ought to believe – the central question of epistemology. [...] In getting more people to adopt new evidence-based standards of rationality — as the great enlightenment philosophers arguably did —philosophers aren’t just leaving the world as they found it. And that is a good thing. 

And then:

[Horwich's Wittgenstein], I suspect, would grant all this,

So Wittgensteinian philosophy, or Lynch's Horwichian Wittgensteinian philosophy, rejects over-generalization, is nevertheless potentially illuminating, and allows for visionary reforms of language use, especially in matters of value. Sounds good to me. Where's the snag?

In order to free us from [certain bad] thoughts, the philosopher must not only show the error in such definitions. She must also take conceptual leaps. She must aim at revision as much as description, and sketch new metaphysical theories, replacing old explanations with new.   

Lynch does not say why. To take his example, if we show the error in defining truth as Authority, why is this not enough to free people from that idea? Especially if we show the error in the kind of illuminating, lots-to-say way that he described earlier?

There are two tasks described here: the illuminating exploration of diverse uses of language that shows the error of over-generalization, and the leaping creation of new uses of language. The former requires mastery of a certain kind of method or methods, but the latter requires creation ex nihilo, or something close to it. There can be no method for that. And there is no obvious reason why people who are good at the one task will also be good at the other. Nor why the creative types (I might call them poets or propagandists) must come from the ranks of the explorer types (call them philosophers). So I don't think that Wittgenstein would include 'poetry' among the tasks of the philosopher, but there is no reason why one could not do both.