Abstract
Process reliabilists hold that in order for a belief to be justified, it must result from a reliable cognitive process. They also hold that a belief can be basically justified: justified in this manner without having any justification to believe that belief is reliably produced. Fumerton (1995), Vogel (2000), and Cohen (2002) have objected that such basic justification leads to implausible easy justification by means of either epistemic closure principles or so-called track record arguments. I argue that once we carefully distinguish closure principles from transmission principles, and epistemic consequences from epistemic preconditions, neither version of this objection succeeds.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
Notes
I shall henceforth talk about propositional justification—justification to believe, and not about doxastic justification—justified belief. The reasons for this are explained in fns. 4 and 6. The connection is this: if you believe a proposition for which you have justification, and your belief is based on that which justifies that proposition, then you have a justified belief. According to process reliabilism, justification is typically only a property of beliefs; in fact, as Kvanvig (2003) has argued, it can seem hard for reliabilists to make much of propositional justification. I shall waive these difficulties here. What is vital is, as Goldman (1976) stressed, that nothing much is read into the locution ’S has justification to believe’. Take this to mean that there is justification for S to believe, and not that S possesses something that she can produce if asked
Although there are dissidents such as Zalabardo (2005), Greco’s agent reliabilism (1999 requires sensitivity to, without knowledge of or belief about, the reliability of my evidence. This is achieved by restricting S’s cognitive processes to those that have a basis in stable and successful dispositions constituting S’s intellectual character. By supplementing objective reliability with this kind of subjective appropriateness, agent reliabilism may be better placed to meet the easy justification objection, but in order to keep things tidy I consider a fairly simple-minded form of reliabilism
Goldman (1986: 62–3, 111–2)
We shall henceforth follow Hawthorne (2005: 27–28) and others in taking the indicative conditional in (Closure) to express strict implication. See also fn. 9. Note also that if the closure principle is formulated in terms of doxastic justification, counterexamples arise when S fails to form the belief that q. One could require merely that S be in a position to form the belief that q, but I suspect this will have to be understood so loosely that one better talk of propositional justification instead. In fact, it matters little whether there’s some notion of doxastic justification that doesn’t entail actual belief such as ‘S is justified in believing p’ as opposed to ‘S justifiably believes p’. For only a fairly weak conception of propositional justification can avoid counterexamples where S knows the entailment from p to q, but fails to infer q from p, thus having justification to believe p, but not that q. That’s another reason for taking ‘S has justification to believe’ to mean that there’s justification for S to believe, and not that S actually possesses some justification
Note that if we substituted ‘having justification to believe’ with ‘having knowledge’ in (4)–(6), as Cohen proposes, it would be open to the process reliabilist to fall back on an epistemic property short of knowledge as not implying the untoward possibility of possessing that property too easily. The point is that by choosing an epistemic property in the premises that is as weak as possible consistent with the target position, we ensure that the conclusion is as strong as possible. In particular, if knowledge entails doxastic justification, and doxastic justification entails propositional justification, we should prefer to put the argument in terms of propositional justification
Wright (2003: 60–63) argues that similar cases exhibit failure of (Transmission), but don’t furnish reason to jettison (Closure)
The claim isn’t that (12) is less intuitive than (11) as a way of understanding the consequent in (7). As an anonymous referee correctly pointed out, (12) may well be the most natural way of understanding that consequent. The claim is rather that if (12) is the correct rendition then the easy justification objection doesn’t get off the ground. Of course given that the red table actually doesn’t have any red lights shining on it, the following material, truth-functional, implication is true:
(13*) [(The table is red) → ((The table isn’t white) & (The table isn’t illuminated by red lights)]
This means that if (Transmission) were to be understood in terms of material implication, then that objection could at least be validly put in terms of (12). Here I am grateful to an anonymous referee.
Cohen (op. cit.)
Klein (2004: 180–181) thinks the only problem about easy knowledge can arise at the first step. He also thinks knowledge is true belief based upon defeasible but undefeated reasoning. He then argues that there can be no easy knowledge, because if the father knows that the table is red, he will not only have good but defeasible reason to believe the table is red provided by the table looking red, it must also be the case that his reason isn’t defeated. And of course the father isn’t in a position to offer any reason why his reason to think the table is red isn’t defeated. But Klein accepts that merely looking at the table is sufficient to provide me with adequate reason to believe the table is red at least “on occasions when the probative value of how things look is not up-for-grabs”. So, forget about the son. Imagine the father running the argument in isolation without any antecedent doubt about any of its steps. In that case, on Klein’s assumptions, it seems the father will acquire easy adequate reason to believe the table is red
Compare with Dretske (1970, 1015–1017) who was clear that if p → ¬q, then a contrast consequence of p has the form ¬ (q & r); where q is the alternative and r the defeater. He said you can know on the basis of visual perception that the animals are zebras, but that you can’t know on that basis that “they are not mules cleverly disguised to look like zebras”, even though “something’s being a zebra implies that it is […] not a mule cleverly disguised by the zoo authorities to look like a zebra.” The present point is that the intuition that you can’t know the proposition expressed by ‘the animals aren’t mules cleverly disguised to look like zebras’ on the basis of visual perception derives partly from failing to note that it is the reliability, not phenomenology, of such perception that confers knowledge but no ability to produce adequate reasons, and partly from the fact that this sentence gives the misleading impression of being in an epistemic position to speak to the defeater, and not just the alternative. See also fn. 13
Here’s why Dretske (1970, 1016) didn’t think you could come to know that the animals aren’t cleverly disguised mules on the basis of inference from perceptually-based knowledge that the animals are zebras: “…think a moment about what reasons you have, what evidence you can adduce in favor of this claim.” Moreover, Dretske thought that knowledge of “general uniformities [or] regularities” would be to no avail. What is needed is rather concrete evidence, e.g. checking with the zoo authorities, close examination of the animals. So, if you share Dretske’s intuition about why you can’t know the animals aren’t cleverly disguised mules on the basis of looking (at some distance), then being told that visual perception is reliable isn’t going to upgrade your belief to knowledge. But then it looks as if the problem about easy justification, if indeed there is one, is orthogonal to the possibility of (Basic Justification)
The distinction between the conditions for having justification and the conditions for warrantedly asserting that I have justification is different from the abovementioned distinctions drawn by Bergmann (2004), Pryor (2004) and Davies (2004). In the case of Bergmann, for instance, the former distinction applies within questioned and unquestioned contexts alike. In particular, even if I have no antecedent doubt as to whether the table isn’t white with red lights shining on it, I may well satisfy the conditions for having justification to believe the table isn’t white with red lights shining on it, but not satisfy the conditions for warrantedly asserting that I have justification to believe just that. The fact that I am not in the dialectical business of overcoming doubt or settling the question doesn’t mean that I cannot satisfy the former conditions without satisfying the latter conditions
For a different bullet-biting response see Van Cleve (2003). Basically, Van Cleve first argues that unless one embraces skepticism, one must endorse reliabilism, and the possibility of easy knowledge via track-record arguments, and secondly he rejects three alternatives: Reidianism, coherentism and two-levels views. Instead I have maintained that it is indeed plausible that we have reliabilist justification in the closure version of the easy justification objection—a version Van Cleve doesn’t address. And I will now go on to argue that the reliabilist can and should reject the transmission principle in the track-record version of that objection—a fourth alternative if you like
Pace Fumerton (1995, 180) who thinks we can never “use a kind of reasoning to justify the legitimacy of using that reasoning”. Alston (1986) argues that epistemic circularity in itself needn’t pose any threat to the justificatory structure of such arguments. For instance, it’s possible to have justification for belief in the reliability of perception as such by relying in part on perception. Bergmann (2004) argues that belief in the existence of non-inferentially justified belief is independently plausible, that non-inferentially justified belief entails the possibility of epistemically circular track record arguments, and hence that belief in the existence of such arguments is independently plausible.
The contentious part is the ‘already’ qualification. Suppose A = justification for the belief that p, B = justification for the belief that r is reliable, C = justification for the belief that r produced the belief that p, and D = justification for the belief that r produced a true belief that p. Then the suggestion is that ¬D ⟹ ((¬B & C)→ ¬A), where ‘¬p ⟹ q’ reads ‘unless antecedent p, q’. But of course ¬D → ((¬B & C) → ¬A) is logically equivalent to (A & ¬B & C) → D, which is just the move to (17) from (14)–(16).
While reliabilist justification typically isn’t individuated in terms of what else I have justification to believe, when faced with putative counterexamples posed by, say, perfectly reliable clairvoyance, reliabilists have readily conceded additional doxastic and evidential constraints on justification—as witnessed by (Non-Undermining). There should thus be no general injunction against imposing some such constraints.
In other words, bearing in mind the notation from fn. 17, maybe (A & C) → D is sanctioned by (Transmission), but (A & ¬B & C) → D isn’t. It’s true that the former entails the latter, but (Closure) suffices to underwrite the cogency of that inference.
As it turns out with so many entrenched examples in philosophy, chicken sexing is actually a skill that can be acquired without extensive practice, and that has nothing to do with smelling. See Bierderman and Shiffrar (1987). But the example serves to highlight a conceptual possibility that could easily have a real—world counterpart
I’m grateful to an anonymous referee for pressing this point
One might object that such properties as epistemic blameworthiness and responsibility are alien to a reliabilist who is concerned only with how justification properly probabilifies beliefs. But the proposal here isn’t that reliabilist justification should be deontological in nature as opposed to truth-conducive. It’s rather that once reliabilist justification, as traditionally understood, is reflective in the abovementioned way, then the question about whether it instantiates these deontological properties becomes pertinent. Epistemological deontology shouldn’t be a prerogative of the epistemological internalist. As Greco (2005) and others have reminded us, in order to determine whether justification is epistemically responsible we need to go beyond evidentialist and other internalist notions. For instance, we need to appeal to the basing relation
The claim isn’t that truth-conditions of justification-ascribing/denying sentences vary with shifts in contextual parameters. The current proposal is agnostic with respect to epistemological contextualism. Nor am I saying that reflection necessarily destroys justification. The idea is merely that when I possess a certain kind of reflective justification, I need to acquire another kind of independent justification
An anonymous referee suggested that the shift from non-reflectively to reflectively relying on r in the case of perception seems too easy. The father says “The table is red”. The son worries “How do you know?” The father replies: “Well, don’t you see it!” The father’s reply may indicate that he reflectively attends to color perception qua belief source, so that justification to believe the table is red is accompanied by justification to believe that color perception produced that belief. But an equally plausible take on that exchange suggests instead that the father appeals to the visual experience he is currently undergoing, and this experience is arguably transparent: introspection of what it’s like reveals only something about the table, and no awareness of either intrinsic properties of the experience itself or indeed of extrinsic properties about its causal history. So, not just any reflective attention will trigger (Basic Constraint)
I would like to thank audiences at the Universities of Edinburgh, St. Andrews and Aberdeen for discussion, and an anonymous referee for helpful comments
References
Alston, W. (1986). Epistemic Circularity. Philosophy and Phenomenological Research, 47(1), 1–30.
Bergmann, M. (2004). Epistemic Circularity, Malignant and Benign. Philosophy and Phenomenological Research, 69, 709–727.
Bierderman, I., & Shiffrar, M. (1987). Sexing Day-Old Chicks: A Case Study and Expert Systems Analysis of a Difficult Perceptual Learning Task. Journal of Experiential Psychology: Learning. Memory and Cognition, 13(4), 640–645.
Cohen, S. (2002). Basic Knowledge and the Problem of Easy Knowledge. Philosophy and Phenomenological Research, Vol. LXV, 2, 309–328.
Cohen, S. (2005). Why Basic Knowledge is Easy Knowledge. Philosophy and Phenomenological Research, 70(2), 417–430.
Fumerton, R. (1995). Metaepistemology and skepticism. Lanham, MD: Rowman and Littlefield.
Davies, M. (2004). Epistemic Entitlement, Warrant Transmission, and Easy Knowledge. Aristotelian Society Supplement, 78, 213–245.
Dretske, F. (1970). Epistemic operators. Journal of Philosophy, 67, 1007–1023.
Goldman, A. (1976). What is justified belief. In G. S. Pappas (Ed.), Justification and knowledge (pp. 1–23). Dordrecht: D. Reidel.
Goldman, A. (1986). Epistemology and cognition. Cambridge, MA: Harvard University Press.
Greco, J. (1999). Agent Reliabilism. Philosophical Perspectives, 13, Epistemology, 273–296.
Greco, J. (2005). Is Justification Internal? In M. Steup & E. Sosa (Eds.), Contemporary debates in epistemology (pp. 257–270). Malden MA: Blackwell Publishing.
Hawthorne, J. (2005). The case for closure. In M. Steup & E. Sosa (Eds.), Contemporary debates in epistemology (pp. 26–43). Malden MA: Blackwell Publishing.
Klein, P. (2004). Closure Matters: Academic Skepticism and Easy Knowledge. Philosophical Issues, 14, Epistemology, 165–184.
Kvanvig, J. (2003). Propositionalism and the Perspectival Character of Justification. American Philosophical Quarterly, 40(1), 3–18.
Lehrer, K. (1990). Theory of knowledge. Boulder, Co: Westview.
Pryor, J. (2004). What’s Wrong with Moore’s Argument?. Philosophical Issues, 14, Epistemology, 349–378.
Sosa, E. (1999). How to defeat Opposition to Moore. Philosophical Perspectives, 1–18.
Sosa, E. (2004). ‘Relevant alternatives (pp. 35–65). Philosophical Studies: Contextualism Included’.
Van Cleve, J. (2003). Is Knowledge Easy—or Impossible? Externalism as the Only Alternative to Skepticism’ in The Sceptics, (ed.) Luper, S., Ashgate, 45–59.
Vogel, J. (2000). Reliabilism Levelled. Journal of Philosophy, 97, 602–623.
Zalabardo, J. (2005). Externalism, Skepticism and the Problem of Easy Knowledge. Philosophical Review, 114, 33–61.
Wright, C. (2002). Anti-Sceptics Simple and Subtle: Moore and McDowell. Philosophy and Phenomenological Research, 65, 330–348.
Wright, C. (2003). Some Reflections on the Acquisition of Warrant by Inference. In S. Nuccetelli (Ed.), New essays on semantic externalism, scepticism and self-knowledge (pp. 57–78). Mass: MIT Press.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Kallestrup, J. Reliabilist Justification: Basic, Easy, and Brute. Acta Anal 24, 155–171 (2009). https://doi.org/10.1007/s12136-009-0053-5
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s12136-009-0053-5