«Assertion, Practical Reasoning, and Epistemic Separabilism Forthcoming in Philosophical Studies Draft – Please Cite Final Version Abstract I argue ...»
Assertion, Practical Reasoning, and Epistemic Separabilism
Forthcoming in Philosophical Studies
Draft – Please Cite Final Version
I argue here for a view I call epistemic separabilism (ES), which states that there are two
different ways we can be evaluated epistemically when we assert a proposition or treat a
proposition as a reason for acting: one in terms of whether we have adhered to or violated the
relevant epistemic norm, and another in terms of how epistemically well-positioned we are towards the fact that we have either adhered to or violated said norm. ES has been appealed to most prominently in order to explain why epistemic evaluations that conflict with the knowledge norm of assertion and practical reasoning nevertheless seem correct. Opponents of such a view are committed to what I call epistemic monism (EM), which states that there is only one way we can be properly evaluated as epistemically appropriate asserters and practical reasoners, namely in terms of whether we have adhered to or violated the relevant norm. Accepting ES over EM has two significant consequences: first, a “metaepistemologial” consequence that the structure of normative epistemic evaluations parallels that found in other normative areas (namely, moral evaluations), and second, that the knowledge norms of assertion and practical reasoning are no worse off than any alternatives in terms of either explanatory power or simplicity.
1. Introduction Proponents of a knowledge norm of assertion (like Peter Unger (1975), Timothy Williamson (2000), Keith DeRose (2002), and others) argue that it is appropriate to assert proposition p only if one knows that p. Assertions of propositions that are not known are thus in some ways improper, and, in turn, are subject to criticism. Specifically, epistemic criticism: we are deemed insufficiently well-positioned epistemically towards p to assert that p. Almost immediately, however, one can come up with plenty of cases in which the knowledge norm of
assertion seems too strict. The most common cases typically take one of the following forms:
Justified False Belief (JFB): S asserts that p. S is justified in believing that p, but p happens to be false (perhaps for reasons that S would not typically be expected to take into account).
Gettiered: S asserts that p. S is justified in believing that p, and p is true, but S does not know that p (perhaps for reasons that S would not typically be expected to take into account)1.
When S has a justified but false belief that p or has a justified true belief that p that S does not know, we are often not inclined to criticize S for asserting that p, despite S’s lack of knowledge in both cases. Now, one test for the plausibility of a norm is that it generally coheres with intuitive judgments: we expect that if one asserts appropriately, it is because one has adhered to the norm of assertion, and, similarly, if one asserts inappropriately it is because one has violated the norm of assertion. Thus if our judgements about the JFB and Gettiered cases are that one is not epistemically criticisable in these cases, and if intuitive criticisability is taken to be a mark of norm violation, then these cases seem to be counterexamples to the knowledge norm of assertion.
Other norms face the same problem. Proponents of a knowledge norm of practical reasoning (like Stanley and Hawthorne (2008), Fantl and McGrath (2002, 2009), and others) argue that it is appropriate to treat p as a reason for acting only if one knows that p. And again, I take Gettiered to encompass traditional Gettier cases as well as other kinds of justified true beliefs that fall short of knowledge, such as barn facade cases and their kin (see Goldman (1976)). Gettiered is named after Edmund Gettier’s famous (1963).
the knowledge norm of practical reasoning seems too strict: it often seems that someone who treats a proposition as a reason for acting when they justifiably believe that p when p is false, or when they have a justified true belief that p that they do not know, are not liable to be criticized because of how epistemically well-positioned they are towards the relevant proposition. Once again, if a norm of practical reasoning needs to be able to accommodate a wide range of intuitive judgments of propriety in order to be plausible, the above cases seem to be counterexamples to the knowledge norm of practical reasoning.
We might, then, choose to reject the knowledge norms in favor of some other norms that are better able to accommodate a wider range of our intuitive judgments; indeed, much of the debate concerning the norms of assertion and practical reasoning are guided by finding a replacement for the knowledge norm that does not admit of what seem to be such ready counterexamples. Defenders of the knowledge norms, however, deny that there is such a strict connection between criticisability and norm violation. Instead, they have argued that one can be epistemically criticisable for one’s assertions and instances of practical reasoning in two different ways: one, in terms of whether one has adhered to the relevant norm, and another, in terms of whether one is epistemically criticisable for one’s primary level improprieties. The idea, then, is that although the JFB and Gettiered cases are instances in which one’s assertion or treatment of a proposition as a reason for acting are inappropriate in terms of being in violation of the relevant norm, we can still accommodate the intuition that there is something epistemically appropriate about them by noting that by virtue of one’s relationship to the fact that one has violated the norm, one might not be criticisable for it.
Call evaluative separabilism the view that intuitive propriety judgements can be indicative of either one’s adherence to or violation of a norm, or whether one should be, in some way, criticisable for said adherences or violations. Evaluative separabilism is common in other normative areas. For example, it is widely accepted that there are two different kinds of ways that we can evaluate someone as a moral subject, one that pertains to whether our actions are morally permissible (i.e. whether our actions adhere to the relevant moral norms), and another that pertains to whether we are blameworthy for them. For example, my action might be morally impermissible because, say, it causes a significant amount of avoidable pain, but I might not be morally blameworthy for it, perhaps because I had good reason to think that my action was going to have different results. We thus accept a kind of separabilism when it comes to certain kinds of
moral judgments. Call the epistemic version of evaluative separabilism epistemic separabilism:
Epistemic Separabilism (ES): A view of the norms of assertion or practical reasoning admits of epistemic separabilism iff the view accepts that there are two different ways that we can be evaluated as epistemic agents in relation to the norms of assertion and practical reasoning, namely by either adhering to or violating those norms, or in terms of whether we are criticisable in terms of the way in which we have adhered to or violated
Accepting ES is what allows the proponent of the knowledge norm to accept what appear to be intuitive judgements of the propriety of our assertions and actions in the JFB and Gettiered cases, namely, not as judgments not about the permissibility of the relevant act, but rather whether we are criticisable for violating the relevant norm.
The strategy of defending the knowledge norms by appeal to epistemic separabilism is employed by several authors, but in different ways. For example, Keith DeRose (2002) defends the knowledge norm of assertion against judgments about cases like JFB and Gettiered in the
As happens with other rules, a kind of secondary propriety/impropriety will arise with respect to [the knowledge norm of assertion]. While those who assert appropriately (with respect to this rule) in a primary sense will be those who actually obey it, a speaker who broke this rule in a blameless fashion (one who asserted something she didn’t know, but reasonably thought she did know) would in some secondary sense be asserting properly.
(180)2 Stanley and Hawthorne (2008) take a similar tack, although they refer to a notion of one being “excused” for violating the norm of practical reasoning:
If someone asserts that p without knowing it and knowing that they don’t know that p, they will have no excuse for their failure to adhere to the norm that one should assert only if one knows. If on the other hand, they assert that p, do not know that p, but cannot be expected to know that they don’t know that p, we may be willing to deem their failure to comply with the norm excusable. The conceptual structure, one familiar from the normative realm, explains suitable appraisal in terms of a combination of norms and excuses for failure to comply with them. (573) Timothy Williamson (2000) employs the same strategy in defense of the knowledge norm of
assertion, but appeals instead to a notion of “reasonableness”:
As opposed to the other authors considered here, DeRose is a contextualist, i.e. a proponent of the view that the semantic value of a knowledge ascription depends on the context of the ascriber. Contextualism, I think, does not have any better way of handling the JFB and Gettiered cases when it comes to a contextualist-knowledge norm of assertion or practical reasoning: one can still have a justified but false belief in a context, and assert the content of that belief in that context, thus violating the knowledge norm but still seemingly doing something epistemically appropriate. Similarly, I mentioned Fantl and McGrath (2002;2009) as proponents of a knowledge norm of practical reasoning. Fantl and McGrath are proponents of interest-relative invariantism (IRI), i.e. the view that whether one has knowledge that p depends at least in part on practical factors surrounding the truth or falsity of p. Again, the proponent of IRI is going to face problems like JFB and Gettiered, since one might be justified in believing a proposition which happens to be false in a situation in which there are no relevant stakes standing in the way of one knowing that p.
The [knowledge norm of assertion] makes knowledge the condition for permissible assertion, not reasonable assertion. One may reasonably do something impermissible because one reasonably but falsely believes it to be permissible. In particular, one may reasonably assert p, even though one does not know that p, because it is very probable on one’s evidence that one knows that p. (256) All of the proposals from DeRose, Stanley and Hawthorne, and Williamson attempt to account for the judgements that we are liable to have in response to the JFB and Gettiered cases in terms of some other epistemic evaluative notion, one that pertains to the way in which one is situated epistemically towards whether one has adhered to or violated the relevant norm. All of these proposals, then, reflect a commitment to ES.
What kind of evaluative notion, exactly, are we positing when we appeal to epistemic separabilism? The three proposals above conceive of the relation between the conditions for the adherence to a norm and our being criticisable for them in three different ways: in terms of being “excused” from the norm, in terms of being “blameless” for violating it, and in terms of being “reasonable” while violating the relevant norm. So which one is the right one? I will not decide amongst them here. Rather, my interest in epistemic separabilism is a structural one, namely whether there is any additional sense that we can be evaluated as an epistemic agent in terms of our adherences to or violations of norms of assertion and practical reasoning. It is this claim, that there are two different ways we can be evaluated as epistemic subjects, that the defender of the knowledge norm in particular needs to appeal to in order to accommodate those judgments that by asserting or treating a proposition as a reason for acting in a JFB or Gettier case one is doing something that is epistemically appropriate. We can, of course, be “blameless” for our epistemically inappropriate assertions and instances of practical reasoning in ways that have nothing to do with our epistemic relationships or the ways in which we are positioned epistemically towards whether we have adhered to or violated the relevant norm. For instance, my assertion of a known falsehood might violate a norm of assertion, but I might not deserve blame for it, in some sense, because it was the best thing to do overall (perhaps I told a lie to spare someone’s feelings). But this is not the kind of case I am interested in looking at here.
Rather, the driving notion behind epistemic separabilism is to capture a way in which we can still find something epistemically appropriate about an assertion or instance of practical reasoning that violates the relevant knowledge norms.3 With that being said, epistemic separabilism has received considerable backlash. We can divide its main criticisms into three general themes: first, that ES is implausible because there is no “conceptual room” for a properly epistemic evaluation of assertion or practical reasoning that does not consist in the adherence to or violation of the relevant norm; second, that by accepting ES we are unable to give straightforward advice about what to do; and third, that norms that appeal to ES account for our judgments of the epistemic propriety of assertions and actions in an unparsimonious way. Those who deny the plausibility of epistemic separabilism are committed to what I will call epistemic monism (EM) for assertion and practical reasoning: intuitive judgments of epistemically appropriate or inappropriate assertions and actions reflect our adherence to or violation of the relevant norm. We cannot consistently adhere to the knowledge Here I am discussing norms of assertion and practical reasoning, but one might wonder whether we should also add norms of belief to the discussion, since debates concerning the correct norm of belief has faced similar issues in determining the epistemic conditions for its proper formation. I leave out beliefs here for two reasons: first, the literature on the norms of belief is too extensive to do proper justice to in this space. Second, it is not clear to me whether the discussion of the norms of belief are really that similar to those of assertion and practical reasoning.