An easy way to tell if someone saying \(P\) means \(\phi\) rather than \(\psi\) by some observable concept is to show them \(\phi\)’s (witness them say \(P\)) and \(\psi\)’s (and witness them not say \(P\)). Call this witnessing the distinction via dispositions of the \(P\)-speaker. Disjunctivitis is a thought experiment that says this is not a sufficient tool for us to distinguish \(\phi\)’s and \(\psi\)’s generally. This is a Kripkensteinian problem.
We’re worried about whether the word ‘porcupine’ really means porcupine, or whether it means porcupine or echidna. Because most of us porcupine users can’t tell porcupines from Australian echidnas. And, although everything I’ve ever seen in called a ‘porcupine’ was a porcupine, everything I’ve ever seen called a ‘porcupine’ was also a porcupine or echidna. And the question is, well, what what fact is it in virtue of which ‘porcupine’ means porcupine, and not porcupine or echidna?
You can’t tell it from my dispositions on the output side, and on the input side, though, it might have been equally true that they’ve all been porcupines they’ve also been porcupines or echidnas. (Furthermore, for all I know, they all happened to have been been male porcupines, in which case we have to worry that my word porcupine means male, porcupines and again, I can’t tell them from the female ones.
Solution from Sellars: the people who taught you the word porcupine, what pattern were they trying to inculcate in you? Was it to say ‘porcupine’ when confronted with a porcupine or echidna? This is a crucial piece of how the noise ‘porcupine’ coming out of my mouth means porcupine.