Epistemic Contextualism and Invariantism

Epistemic contextualism is a semantic thesis about the meaning of the word "knows" and its cognates. Invariantism, which is the more traditional view, holds that the truth or falsity of sentences like "Mary knows that the bank is open on Saturday" does not shift from context to context. Contextualists, however, argue that such a sentence can be true in one context but not another. A typical model here is the case of indexical expressions, like "I" or "here." My utterance of "I am President" can be false while Obama's is true. Some contextualists have argued that contextualism solves the problem of skepticism: e.g. "I know I have hands" is true out on the street but false in the philosophy classroom, where the context raises the standards for knowledge. Others have claimed that contextualism yields the best explanation for why our ordinary usage of "knows" varies with what is at stake for us or which error-possibilities we have in mind. Attempting to capture some of the same phenomena as contextualism, various forms of invariantism have been developed. These include subject-sensitive invariantism, moderate and skeptical pragmatic invariantism and psychological error theories. According to subject-sensitive invariantism, “knows” invariably expresses the knowledge relation. The seeming sensitivity of "knows" arises from the supposed fact that this relation is sensitive to non-epistemic features of the subject (such as what is at stake for her or which error-possibilities are salient in her conversational context). According to moderate and skeptical pragmatic invariantism, the variability of “knows” should be attributed not to semantic but to pragmatic factors (such as implicatures). Moderate pragmatic invariantists hold that, semantically, “knows” invariably expresses an epistemic relation we tend to satisfy in many ordinary situations. Skeptical pragmatic invariantists hold that, semantically, “knows” invariably expresses an epistemic relation that is rarely if ever satisfied. According to psychological error theories, the variability of “knows” should be explained neither semantically nor pragmatically but by appeal to psychological biases and similar aspects of the psychology of knowledge ascribing subjects. 

Relevant alternatives theories (e.g. Dretske 1970) may be seen as a forerunner of contextualism. Explicitly contextualist theories, however, begin primarily with Cohen 1986, DeRose 1992, and Lewis 1996. Subject-sensitive invariantism begins with Fantl & McGrath 2002, Hawthorne 2004, and Stanley 2005. Rysiew 2001 is an early defense of moderate pragmatic invariantism. Skeptical pragmatic invariantism is prominently defended in Davis 2007. Psychological error theories begin with Hawthorne 2004, where the mechanism primarily invoked is the so-called "availability heuristic." The view is forcefully criticized in Nagel 2010. More elaborate psychological error theories are defended in Nagel 2008 and Nagel 2010. The data motivating contextualism have also been discussed in experimental philosophy. Schaffer & Knobe 2010 discuss a range of studies.