Abstract
Quine is usually read as arguing either for a non-factualism about analyticity (1) ... Or, at the very least, for an error thesis about it: (2) ... These attributions — including the stronger non-factualist thesis — seem licensed by many passages, including the famous one which concludes Quine's discussion in "Two Dogmas" ... Nevertheless, Paul Horwich does not wish to read Quine as endorsing either (1) or (2). He certainly does not wish to attribute (1) to him. And he wishes to attribute only a restricted version of (2). What this restriction is and why Paul wants to make it are matters I'll come to in just a moment.