A Noncontextualist Account of Contextualist Linguistic Data

The paper takes as its starting point the observation that people can be led to retract knowledge claims when presented with previously ignored error possibilities, but offers a noncontextualist explanation of the data. Fallibilist epistemologies are committed to the existence of two kinds of Kp-falsifying contingencies: (i) Non-Ignorable contingencies [M-contingencies] and (ii) Properly-Ignorable contingencies [PIcontingencies]. For S to know that p, S must be in an epistemic position to rule out all NI-contingencies, but she need not be able to rule out the PI-contingencies. What is required vis-a-vis PI-contingencies is that they all be false. In mentioning PI-contingencies, an interlocutor can lead S mistakenly to think that these contingencies are NI-contingencies, when in fact they are not. Since S cannot rule out these newly mentioned contingencies and since she mistakenly takes them to be NI-contingencies, it is quite natural that she retract her earlier knowledge claim. In short, mentioning NI-contingencies creates a distortion effect. It makes S think that the standards for knowledge are higher than they actually are, which in turn explains why she mistakenly thinks she lacks knowledge. Conclusion: The primary linguistic data offered in support of contextualism can be explained without resorting to contextualism.