Peter Graham

C.A.J. Coady, in his book Testimony: A Philosophical Study (Oxford: Clarendon Press, 1992), offers conditions on an assertion that p to count as testimony. He claims that the assertion that p must be by a competent speaker directed to an audience in need of evidence and it must be evidence that p. I offer examples to show that Coady’s conditions are too strong. Testimony need not be evidence; the speaker need not be competent; and, the statement need not be relevant or directed to someone in need of evidence. I give alternative conditions. Coady was led into the stronger conditions by investigating testimony as it occurs in legal contexts, where special steps are taken to ensure that testimony provides the jury or the judge with evidence by a competent speaker that is relevant to the disputed question of the guilt or innocence of the defendant.

Radical skepticism about the external implies that no belief about the external is even prima facie justified. A theoretical reply to skepticism has four stages. First, show which theories of epistemic justification support skeptical doubts (show which theories, given other reasonable assumptions, entail skepticism). Second, show which theories undermine skeptical doubts (show which theories, given other reasonable assumptions, do not support the skeptic’s conclusion). Third, show which of the latter theories (which non-skeptical theory) is correct, and in so doing show that all of the rival theories of justification, skeptical and non-skeptical alike, are mistaken. Fourth, explain why skeptical doubts are sometimes (or sometimes merely seem) intuitive, and thereby accommodate skeptical doubts without capitulation. Michael Williams has pioneered the very idea of a theoretical reply. A theoretical diagnosis consists in just the first two stages. An adequate reply, which is correct at each stage, would rebut the skeptic entirely. Williams’ own reply, I argue, is inadequate. I offer in its place an exhaustive and accurate diagnosis of skepticism. I distinguish four kinds of skepticism and five theories of justification. I then show which theories do, and which theories do not, support which kinds of skepticism.

Our folk epistemology says that if someone knows that P and tells you that P, then, given the absence of defeaters, if you believe what they tell you, you will come to know that P as well. A speaker's knowledge that P is then, for the most part, enough for a hearer to come to know that P. But there are counterexamples to this principle: testimonial knowledge does not always transfer from the speaker to the hearer. Why should that be so? Because testimonial knowledge arises through the flow of information--from conclusive grounds--from the speaker through the speaker's assertion to the hearer's state of comprehension as of the speaker's assertion. A speaker's belief can be based on conclusive grounds, but the speaker's assertion may fail to provide conclusive grounds to the hearer. Then the hearer won't acquire knowledge, even from a knowledgeable speaker. Knowledge from testimony then works more like knowledge through perception. But wait, you say, that can't be so because testimony involves free will but perception does not. The essay ends with a broadly Hume inspired reply to this objection.

What is the best account of process reliabilism about epistemic justification, especially epistemic entitlement? I argue that entitlement consists in the normal functioning (proper operation) of the belief-forming process when the process has forming true beliefs reliably as an etiological function. Etiological functions involve consequence explanation: a belief-forming process has forming true beliefs reliably as a function just in case forming-true beliefs reliably partly explains the persistence of the process. This account paves the way for avoiding standard objections to process reliabilism and situations epistemic entitlement within a normative framework of functions and functional norms.

This paper argues for the general proper functionalist view that epistemic warrant consists in the normal functioning of the belief-forming process when the process has forming true beliefs reliably as an etiological function. Such a process is reliable in normal conditions when functioning normally. This paper applies this view to so-called testimony-based beliefs. It argues that when a hearer forms a comprehension-based belief that P (a belief based on taking another to have asserted that P) through the exercise of a reliable competence to comprehend and filter assertive speech acts, then the hearer's belief is prima facie warranted. The paper discusses the psychology of comprehension, the function of assertion, and the evolution of filtering mechanisms, especially coherence checking.

Reductionism about testimony holds that testimonial warrant or entitlement is just a species of inductive warrant. Anti-Reductionism holds that it is different from inductive but analogous to perceptual or memorial warrant. Perception receives much of its positive epistemic status from being reliably truthconducive in normal conditions. One reason to reject the epistemic analogy is that testimony involves agency – it goes through the will of the speaker – but perception does not. A speaker might always choose to lie or otherwise deliberately mislead. It is argued that the force of this derives (in part) from Libertarianism about agency, and that Libertarianism, if it undermines the Anti-Reductionist explanation of why we are entitled to rely upon testimony, undermines the Reductionist explanation as..

Are we entitled or justified in taking the word of others at face value? An affirmative answer to this question is associated with the views of Thomas Reid. Recently, C. A. J. Coady has defended a Reidian view in his impressive and influential book. Testimony: A Philosophical Study. His central and most Oliginal argument for his positions involves reflection upon the practice of giving and accepting reports, of making assertions and relying on the word of others. His argument purports to show that testimony is, by its very nature, a “reliable form of evidence about the way the world is.” The argument moves from what we do to why we are justified in doing it. Although I am sympathetic with both the Reidian view and Coady’s attempt to connect why we rely on others with why we are entitled to rely on others, I find Coady’s argument ineffective

Reliabilism about epistemic justification claim that what _makes_ epistemically correct procedures of belief-formation _correct_ is that these procedures produce reliably true outcomes. If correct procedures are necessarily correct, and reliability in the world of use is only contingent, then reliability in the world of use cannot be what makes correct procedures correct. The reliabilist rejoinder shifts from _de facto_ reliability to reliability in a _special set_ of worlds; reliability in special worlds makes correct procedures correct in _all_ worlds. The most popular version identifies special worlds with the actual world; correct procedures _in all worlds_ are reliable procedures in _the actual world_. There are two variants: _absolutist_ actual-world reliabilism and _indexicalist_ actual-world reliabilism. Neither work. Both fail to substantively explain why procedures are made correct by reliability in the actual world. The way forward is to shift from reliability in the actual world to reliability in normal conditions.

Anti-reductionists hold that beliefs based upon comprehension (of both force and content) of tellings are non-inferentially justified. For reductionists, on the other hand, comprehension as such is not in itself a warrant for belief: beliefs based on it are justified only if inferentially supported by other beliefs. I discuss Elizabeth Fricker's argument that even if anti-reductionism is right in principle, its significance is undercut by the presence of background inferential support: for mature knowledgeable adults, justification from comprehension as such plays no active role, and is superseded by inferential warrant. I show that her argument begs important questions. Inferential and non-inferential support combine to over-determine the justification of comprehension-based beliefs

What is the biological function of perception? I hold perception, especially visual perception in humans, has the biological function of accurately representing the environment. Tyler Burge argues this cannot be so in Origins of Objectivity (Oxford, 2010), for accuracy is a semantical relationship and not, as such, a practical matter. Burge also provides a supporting example. I rebut the argument and the example. Accuracy is sometimes also a practical matter if accuracy partly explains how perception contributes to survival and reproduction.