Book Discussion of Christopher Norris’ “Epistemology: Key Concepts in Philosophy”

Chapter 5: Making for Truth: Some Problems with Virtue-Based Epistemology –

Section V.

Ok, so, apparently “foundationalist” means coherentist, and “anti-foundationalist” means pragmatist. Sosa claims they are a dead-end dispute to how to solve the problems that came up after Quine beat down logical empiricism—but virtue theory clearly inherits the same old dilemmas.

The most crucial dilemma is “the problem as to just what an epistemic virtue should be taken to involve if not—at its most basic—the realist acceptance that truth might always elude our best, most reliable or virtuous means of ascertainment. It is here (as I argued in Chapter 4) that the advocates of response-dependence lean strongly in an anti-realist direction despite their overt professions of even-handedness, and here also that McDowell draws the line against any alethic (objectivist) theory of truth that would place it outside and beyond the realm of jointly operative human ‘receptivity’ and ‘spontaneity’.” Virtue theorists claim to shift the emphasis from “justificatory warrant” to “disposition”—but this puts them in the RD, or anti-realist, camp, since “it equates truth with the best opinion or optimized judgement among those presumptively qualified to know on account of just these imputed qualities.” Not only is virtue theory vulnerable to the counter-argument of the realist about epistemology—it is vulnerable to the counter-argument of the realist about ethics, who is a consequentialist.

The virtue theorist will reply that “caution, self criticism, acceptance of the non-finality of knowledge as we have it—…between them ensure an adequate respect for the standing possibility that truth might elude our present best truth-seeking endeavors. Yet these character traits, no matter how admirable, are still specified in such a way as to reverse the realist order of priority between truth…and truthfulness.”

Yes, virtue theory is better than outright communalist theories and Kripkensteineism, but it must resort to them in order to explain what counts “as an adequate definition of this or that epistemic virtue”—since they refuse to take one side of the dilemma between “a causal-reliabilist approach devoid of normative values, and a deontological approach devoid of credible epistemic warrant.” To avoid communalism and Kripkensteinism, the virtue theorist can:
1. Strengthen the internalist requirement (to responsibly apply standards like “rationality, consistency, intellectual caution and care to avoid erroneous results”).
2. Fall back on reliabilist position which makes room for virtue-talk to deal with lack of normativity.

Those two options (deontological or internalist, versus reliabilist externalist) cannot be reconciled and tend to pull in opposite directions. (However…my paper does reconcile doing and being…hmm…deontology says DO the standards and you will become them…virtue says BE the standards and you will do them…both are right. How are either of them any more/less general than the other? Norris seems to be saying they both suck compared to consequentialism…hmm…interesting.)

Anyway, the irreconcilability between externalist and internalist arguments is apparent “in Sosa’s early suggestion that the intellectual virtues could be held analogous to reliably knowledge-conducive perceptual faculties such as good eyesight, acute hearing or tactile sensitivity” because those things are mostly modular or encapsulated—they mostly run on automatic—and higher-level reasoning does not (Fodor and other modularists have scaled down their claims that higher-level reasoning is also modular or purely computational, being more global). Sosa’s claim does not account for the interpretive processing that is part of perception, making it different than passive sensory uptake…though Quine (and his successors) would naturalize the interpretive processing (the ‘torrential output’ of scientific theories and hypotheses). Wittgenstein and Kuhn also put in that we see through the lens of our background information—to the point that is impossible to distinguish between veridical and delusory appearances. All the motorists and passerby’s in Barn-Façade County have it right, by this account (though, here, Norris refers to an example by Norwood Hanson). So this screws up Sosa’s analogy. The alternative to the Quinean or Wittgensteinian path is the Fodorian path of the modularity of higher thinking, so that it is conducive to knowledge not subject to interpretation. So…what to make of interpretation? Hence, Fodor’s stepping away from massive modularity. Modularity does not account for abduction—or inference to the best explanation.

So—since the virtues of higher thinking cannot be held analogues to automatic processes, that “threatens to deprive those virtues of the normative (ethically oriented) character which they must be taken to possess if they are to count as such in any but a loose or misapplied sense of the term,” – not exactly sure why. So, Sosa has developed two requirements: reliable beliefs, and evaluating those beliefs and their acquirement for their virtue. This pushes him into internalist territory, when he is supposed to “cut out that false disjunction between causally reliable sources of knowledge and ethically responsible ways of construing the best evidence to hand.”

John Greco says he does this when he “argues that the normative deficit of Goldman-type reliabilist theories can best be got over by introducing the further requirement that, in order to possess ‘positive epistemic status’, a subject’s virtues must include their willingness to take responsibility for whatever they advance in the way of purportedly veridical statements.” This harkens back to McDowell’s “responsible freedom” and doesn’t resolve what should count as an epistemic virtue and why it should count. It is in a tug of war between “truth” and “truthfulness” (in short).

Truthfulness matters, but sometimes, despite being truthful, we can fail to get at the truth. Virtue theorists can make the virtues infallibly truth-tracking (realist), but therefore “trivially circular and devoid of substantive (perceptual-cognitive or epistemic) content.” Or—they can define what traits count as virtuous—restricting truth to human response.