Infallibility

The aim of this paper is to analyse and develop how scepticism becomes an intelligible question starting from requirements that epistemologists themselves aim to endorse. We argue for and defend the idea that the root of scepticism is the underdetermination principle by articulating its specificitya respectable epistemic principle and by defending it against objections in current literature. This engagement offers a novel understanding of underdetermination-based scepticism. While most anti-sceptical approaches challenge scepticism by understanding it as postulating uneliminated scenarios of mass (...) deception, or as endorsing unnatural epistemic requirements, we argue here that both contentions are mistaken. Underdetermination-based scepticism targets our beliefs by issuing a genuine question about the rational support they enjoy. If we cannot establish that the sources of our beliefs provide them the required epistemic merit and authority, they lack non-arbitrary grounds. This has a sizable impact on what constitutes a satisfactory anti-sceptical strategy. Strategies that merely focus on the scenario-based aspect of scepticism, or on the truth-functional evaluation of our beliefs, are shown to miss the mark of the sceptical threat. The proposed analysis ultimately provides a shift in perspective concerning the character and reach of philosophical doubt. (shrink)

I advance a novel argument for an infallibilist theory of knowledge, according to which we know all and only those propositions that are certain for us. I argue that this theory lets us reconcile major extant theories of knowledge, in the following sense: for any of these theories, if we require that its central condition (evidential support, reliability, safety, etc.) obtains to a maximal degree, we get a theory of knowledge extensionally equivalent to infallibilism. As such, the infallibilist can affirm (...) that, when their conditions are suitably interpreted, most post-Gettier theories of knowledge offer necessary and sufficient conditions for knowledge. The infallibilist can thus reconcile major theories of knowledge, and is in a better position to explain the intuitive appeal of these theories than the fallibilist who only accepts one of them, and rejects the rest. (shrink)

The role of certitude in our mental lives is, to put it mildly, controversial. Many current epistemologists (including epistemologists of religion) eschew certitude altogether. Given his emphasis on certitude, some have maintained that John Henry Newman was an infallibilist about knowledge. In this paper, we argue that a careful examination of his thought (especially as seen in the Grammar of Assent) reveals that he was an epistemic fallibilist. We first clarify what we mean by fallibilism and infallibilism. Second, we explain (...) why some have read Newman as an infallibilist. Third, we offer two arguments that Newman is at least a fallibilist in a weak sense. In particular, the paradox he seeks to resolve in the Grammar and his dispute with John Locke both indicate that he is at least a weak fallibilist. We close with a consideration of whether Newman is a fallibilist in a much stronger sense as well. (shrink)

An aspect of Peirce’s thought that may still be underappreciated is his resistance to what Levi calls _pedigree epistemology_, to the idea that a central focus in epistemology should be the justification of current beliefs. Somewhat more widely appreciated is his rejection of the subjective view of probability. We argue that Peirce’s criticisms of subjectivism, to the extent they grant such a conception of probability is viable at all, revert back to pedigree epistemology. A thoroughgoing rejection of pedigree in the (...) context of probabilistic epistemology, however, _does_ challenge prominent subjectivist responses to the problem of the priors. (shrink)

Hope, in its propositional construction "I hope that p," is compatible with a stated chance for the speaker that not-p. On fallibilist construals of knowledge, knowledge is compatible with a chance of being wrong, such that one can know that p even though there is an epistemic chance for one that not-p. But self-ascriptions of propositional hope that p seem to be incompatible, in some sense, with self-ascriptions of knowing whether p. Data from conjoining hope self-ascription with outright assertions, with (...) first- and third-person knowledge ascriptions, and with factive predicates suggest a problem: when combined with a plausible principle on the rationality of hope, they suggest that fallibilism is false. By contrast, the infallibilist about knowledge can straightforwardly explain why knowledge would be incompatible with hope, and can offer a simple and unified explanation of all the linguistic data introduced here. This suggests that fallibilists bear an explanatory burden which has been hitherto overlooked. (shrink)

There is a widespread attitude in epistemology that, if you know on the basis of perception, then you couldn't have been wrong as a matter of chance. Despite the apparent intuitive plausibility of this attitude, which I'll refer to here as “stochastic infallibilism”, it fundamentally misunderstands the way that human perceptual systems actually work. Perhaps the most important lesson of signal detection theory (SDT) is that our percepts are inherently subject to random error, and here I'll highlight some key empirical (...) research that underscores this point. In doing so, it becomes clear that we are in fact quite willing to attribute knowledge to S that p even when S's perceptual belief that p could have been randomly false. In short, perceptual processes can randomly fail, and perceptual knowledge is stochastically fallible. The narrow implication here is that any epistemological account that entails stochastic infallibilism, like safety, is simply untenable. More broadly, this myth of stochastic infallibilism provides a valuable illustration of the importance of integrating empirical findings into epistemological thinking. (shrink)

For Kant, knowledge involves certainty. If “certainty” requires that the grounds for a given propositional attitude guarantee its truth, then this is an infallibilist view of epistemic justification. Such a view says you can’t have epistemic justification for an attitude unless the attitude is also true. Here I want to defend an alternative fallibilist interpretation. Even if a subject has grounds that would be sufficient for knowledge if the proposition were true, the proposition might not be true. And so there (...) is sometimes still rational room for doubt. The goal of this paper is to present four different models of what “certainty” amounts to, for Kant, each of which is compatible with fallibilism. (shrink)

I argue that the Hume of the Treatise maintains an account of knowledge according to which (i) every instance of knowledge must be an immediately present perception (i.e., an impression or an idea); (ii) an object of this perception must be a token of a knowable relation; (iii) this token knowable relation must have parts of the instance of knowledge as relata (i.e., the same perception that has it as an object); and any perception that satisfies (i)-(iii) is an instance (...) of knowledge. I then apply this account to the case of sense perception. I argue that Hume holds that relations of impressions can be intuited, are knowable, and are necessary. For Hume, these relations constitute sensory knowledge. While Hume is rightly labeled an empiricist for many reasons, a close inspection of his account of knowledge reveals yet another way in which he deserves the label. (shrink)

I present a cumulative case for the thesis that we only know propositions that are certain for us. I argue that this thesis can easily explain the truth of eight plausible claims about knowledge: -/- (1) There is a qualitative difference between knowledge and non-knowledge. (2) Knowledge is valuable in a way that non-knowledge is not. (3) Subjects in Gettier cases do not have knowledge. (4) If S knows that P, P is part of S’s evidence. (5) If S knows (...) that P, ~P is epistemically impossible for S. (6) If S knows that P, S can rationally act as if P. (7) If S knows that P, S can rationally stop inquiring whether P. (8) If S knows each of {P1, P2, … Pn}, and competently deduces Q from these propositions, S knows that Q. -/- I then argue that the skeptical costs of this thesis are outweighed by its explanatory power. (shrink)

According to Probability 1 Infallibilism (henceforth, Infallibilism), if one knows that p, then the probability of p given one’s evidence is 1. Jessica Brown (2018, 2013) has recently argued that Infallibilism leads to scepticism unless the infallibilist also endorses the claim that if one knows that p, then p is part of one’s evidence for p. By doing that, however, the infalliblist has to explain why it is infelicitous to cite p as evidence for itself. And yet, the infallibilist doesn’t (...) seem to have a satisfying explanation available. Call this the Infelicity Challenge for Probability 1 Infallibilism. By exploiting the distinction between the justifying and the motivating role of evidence, in this paper, I argue that, contrary to first appearances, the Infelicity Challenge doesn’t arise for Probability 1 Infallibilism. However, after anticipating and resisting two objections to my argument, I show that we can identify a different version of infallibilism which seems to face a problem that is even more serious than the Infelicity Challenge. (shrink)

This paper outlines a new type of skepticism that is both compatible with fallibilism and supported by work in psychology. In particular, I will argue that we often cannot properly trust our ability to rationally evaluate reasons, arguments, and evidence (a fundamental knowledge-seeking faculty). We humans are just too cognitively impaired to achieve even fallible knowledge, at least for many beliefs.

In this paper I consider the prospects for a skeptical version of infallibilism. For the reasons given above, I think skeptical invariantism has a lot going for it. However, a satisfactory theory of knowledge must account for all of our desiderata, including that our ordinary knowledge attributions are appropriate. This last part will not be easy for the infallibilist invariantist. Indeed, I will argue that it is much more difficult than those sympathetic to skepticism have acknowledged, as there are serious (...) problems with regarding paradigmatic, typical knowledge attributions as loose talk, exaggerations, or otherwise practical uses of language. So, I do not think the pragmatic story that skeptical invariantism needs is one that works without a supplemental error theory of the sort left aside by purely pragmatic accounts of knowledge attributions. In its place, I will offer a compromise pragmatic and error view that I think delivers everything that skeptics can reasonably hope to get. (shrink)

According to the doctrine of infallibility, one is permitted to believe p if one knows that necessarily, one would be right if one believed that p. This plausible principle—made famous in Descartes’ cogito—is false. There are some self-fulfilling, higher-order propositions one can’t be wrong about but shouldn’t believe anyway: believing them would immediately make one's overall doxastic state worse.

Knowledge is good, ignorance is bad. So it seems, anyway. But in this dissertation, I argue that some ignorance is epistemically valuable. Sometimes, we should suspend judgment even though by believing we would achieve knowledge. In this apology for ignorance (ignorance, that is, of a certain kind), I defend the following four theses: 1) Sometimes, we should continue inquiry in ignorance, even though we are in a position to know the answer, in order to achieve more than mere knowledge (e.g. (...) understanding) while minimizing the effects of confirmation bias. 2) It’s false that we should believe every proposition such that we are guaranteed to be right about it (and even such that we are guaranteed to know it) if we believe it. 3) Being in a position to know is the norm of assertion: importantly, this does not require belief or (thereby) knowledge, and so proper assertion can survive speaker-ignorance. 4) It can be permissible and conversationally useful to tell audiences things that it is logically impossible for them to come to know: Proper assertion can survive (necessary) audience-side ignorance. Cumulatively, this project suggests that, properly understood, ignorance has an important role to play in the good epistemic life. (shrink)

Jeder Mensch irrt – ausgenommen der Papst, wenn er Glaubenssätze verkündet. So jedenfalls befand einst das erste Vatikanische Konzil. Nun waren die Kardinäle, so bemerkt Keil frech, selbst keineswegs Träger der päpstlichen Unfehlbarkeit. »Woher wussten sie dann, dass der Papst unfehlbar ist?« Niemand weiß vorher, wann und wo er sich irren wird. Viele Philosophen haben daraus geschlossen, dass Menschen nichts wissen, sondern immer nur vermuten. Das ist aber ein Irrtum, den dieser kluge und kurzweilige Essay aufklärt.

This is a reply to Howard Sankey’s comment (“Factivity or Grounds? Comment on Mizrahi”) on my paper, “You Can’t Handle the Truth: Knowledge = Epistemic Certainty,” in which I present an argument from the factivity of knowledge for the conclusion that knowledge is epistemic certainty. While Sankey is right that factivity does not entail epistemic certainty, the factivity of knowledge does entail that knowledge is epistemic certainty.

In this discussion note, I put forth an argument from the factivity of knowledge for the conclusion that knowledge is epistemic certainty. If this argument is sound, then epistemologists who think that knowledge is factive are thereby also committed to the view that knowledge is epistemic certainty.

This reply provides further grounds to doubt Mizrahi’s argument for an infallibilist theory of knowledge. It is pointed out that the fact that knowledge requires both truth and justification does not entail that the level of justification required for knowledge be sufficient to guarantee truth. In addition, an argument presented by Mizrahi appears to equivocate with respect to the interpretation of the phrase “p cannot be false”.

I present an argument for a sophisticated version of sceptical invariantism that has so far gone unnoticed: Bifurcated Sceptical Invariantism (BSI). I argue that it can, on the one hand, (dis)solve the Gettier problem, address the dogmatism paradox and, on the other hand, show some due respect to the Moorean methodological incentive of ‘saving epistemic appearances’. A fortiori, BSI promises to reap some other important explanatory fruit that I go on to adduce (e.g. account for concessive knowledge attributions). BSI can (...) achieve this much because it distinguishes between two distinct but closely interrelated (sub)concepts of (propositional) knowledge, fallible-but-safe knowledge and infallible-and-sensitive knowledge, and explains how the pragmatics and the semantics of knowledge discourse operate at the interface of these two (sub)concepts of knowledge. I conclude that BSI is a novel theory of knowledge discourse that merits serious investigation. (shrink)

When spelled out properly infallibilism is a viable and even attractive view. Because it has long been summary dismissed, however, we need a guide on how to properly spell it out. The guide has to fulfil four tasks. The first two concern the nature of knowledge: to argue that infallible belief is necessary, and that it is sufficient, for knowledge. The other two concern the norm of belief: to argue that knowledge is necessary, and that it is sufficient, for justified (...) certainty. With such a guide in hand infallibilism can be evaluated on its own merits. The most controversial parts are the first and fourth. The idea that knowledge requires infallible belief is thought to be excessively sceptical. The idea that knowledge warrants certainty is thought to be excessively dogmatic. The present paper addresses the first. It argues that knowledge requires infallible belief. (shrink)

ABSTRACTAccording to the Dogmatism Puzzle presented by Gilbert Harman, knowledge induces dogmatism because, if one knows that p, one knows that any evidence against p is misleading and therefore one can ignore it when gaining the evidence in the future. I try to offer a new solution to the puzzle by explaining why the principle is false that evidence known to be misleading can be ignored. I argue that knowing that some evidence is misleading doesn't always damage the credential of (...) the evidence, and therefore it doesn't always entitle one to ignore it. I also explain in what kind of cases and to what degree such knowledge allows one to ignore evidence. Hopefully, through the discussion, we can not only understand better where the dogmatism puzzle goes wrong, but also understand better in what sense rational believers should rely on their evidence and when they can ignore it. (shrink)

Saul Kripke argued that the requirement that knowledge eliminate all possibilities of error leads to dogmatism . According to this view, the dogmatism puzzle arises because of a requirement on knowledge that is too strong. The paper argues that dogmatism can be avoided even if we hold on to the strong requirement on knowledge. I show how the argument for dogmatism can be blocked and I argue that the only other approach to the puzzle in the literature is mistaken.

The sciences occasionally generate discoveries that undermine their own assumptions. Two such discoveries are characterized here: the discovery of apophenia by cognitive psychology and the discovery that physical systems cannot be locally bounded within quantum theory. It is shown that such discoveries have a common structure and that this common structure is an instance of Priest’s well-known Inclosure Schema. This demonstrates that science itself is dialetheic: it generates limit paradoxes. How science proceeds despite this fact is briefly discussed, as is (...) the connection between our results and the realism-antirealism debate. We conclude by suggesting a position of epistemic modesty. (shrink)

A variation of Fitch’s paradox is given, where no special rules of inference are assumed, only axioms. These axioms follow from the familiar assumptions which involve rules of inference. We show (by constructing a model) that by allowing that possibly the knower doesn’t know his own soundness (while still requiring he be sound), Fitch’s paradox is avoided. Provided one is willing to admit that sound knowers may be ignorant of their own soundness, this might offer a way out of the (...) paradox. (shrink)

Among the key factors that play a crucial role in the acquisition of knowledge, Buddhist philosophers list (i) the testimony of sense experience, (ii) introspective awareness (iii) inferences drawn from these directs modes of acquaintance, and (iv) some version of coherentism, so as guarantee that truth claims remains consistent across a diverse philosophical corpus. This paper argues that when Buddhists employ reason, they do so primarily in order to advance a range of empirical and introspective claims. As a result, reasoning, (...) in particular inductive reasoning on the testimony of perception, is based on a theory of causation. (shrink)

Suppose for reductio that I know a proposition of the form <p and I don’t know p>. Then by the factivity of knowledge and the distribution of knowledge over conjunction, I both know and do not know p ; which is impossible. Propositions of the form <p and I don’t know p> are therefore unknowable. Their particular kind of unknowability has been widely discussed and applied to such issues as the realism debate. It hasn’t been much applied to theories of (...) the nature of knowledge. That is what I’m going to do here. (shrink)

Anthony Brueckner argues for a strong connection between the closure and the underdetermination argument for scepticism. Moreover, he claims that both arguments rest on infallibilism: In order to motivate the premises of the arguments, the sceptic has to refer to an infallibility principle. If this were true, fallibilists would be right in not taking the problems posed by these sceptical arguments seriously. As many epistemologists are sympathetic to fallibilism, this would be a very interesting result. However, in this paper I (...) will argue that Brueckner’s claims are wrong: The closure and the underdetermination argument are not as closely related as he assumes and neither rests on infallibilism. Thus even a fallibilist should take these arguments to raise serious problems that must be dealt with somehow. (shrink)

Can we find propositions that cannot rationally be denied in any possible world without assuming the existence of that same proposition, and so involving ourselves in a contradiction? In other words, can we find transworld propositions needing no further foundation or justification? Basically, three differing positions can be imagined: firstly, a relativist position, according to which ultimately founded propositions are impossible; secondly, a meta-relativist position, according to which ultimately founded propositions are possible but unnecessary; and thirdly, an absolute position, according (...) to which such propositions are necessary. In this short essay I show that under the premise of modal logic S5 with constant domain there are ultimately founded propositions and that their existence is even necessary, and I will give some reasons for the superiority of S5 over other logics. (shrink)

This paper is a companion piece to my earlier paper “Fallibilism and Concessive Knowledge Attributions”. There are two intuitive charges against fallibilism. One is that it countenances the truth (and presumably acceptability) of utterances of sentences such as “I know that Bush is a Republican, though it might be that he is not a Republican”. The second is that it countenances the truth (and presumably acceptability) of utterances of sentences such as “I know that Bush is a Republican, even though (...) I’m not certain that he is”, or “I know that Bush it a Republican, even though it isn’t certain that he is.” In “Fallibilism and Concessive Knowledge Attributions”, I argue that fallibilism in epistemology does not countenance the truth of utterances of sentences such as “I know that Bush is a Republican, though it might be that he is not a Republican”. In this paper, I argue that there are independent reasons for thinking that utterances of sentences such as “I know that Bush is a Republican, though I’m not certain that he is” and “I know that Bush is a Republican, though it’s not certain that he is” are unassertible. More specifically, I argue that these are simply instances of Moore’s Paradox, such as “Dogs bark, but I don’t know that they do.” The right account of Moore’s Paradox does not involve the falsehood of the semantic content of the relevant utterances, but rather their pragmatic unacceptability. So the anti-fallibilist intuitions turn out to have pragmatic, rather than semantic import, and therefore do not tell against the truth of fallibilism. Fallibilism in epistemology is often thought to be theoretically desirable, but intuitively problematic. My purpose with these two papers is to show that fallibilism is not intuitively problematic. (shrink)

The Contingency Postulate of Truth. - Is there a statement that cannot be false under any contingent conditions? Two well-known philosophical schools have given contradictory answers to this question about the existence of a necessarily true statement: Fallibilists (Albert, Keuth) have denied its existence, transcendental pragmatists (Apel, Kuhlmann) and objective idealists (Wandschneider, Hösle) have affirmed it. Dieter Wandschneider has (following Vittorio Hösle) translated the principle of fallibilism, according to which every statement is fallible, into a thesis which he calls the (...) contingency postulate of truth (CPT). It says: <For every true statement there are contingent conditions.> If this postulate were true, it would mark an insurmountable boundary of knowledge: a final epistemic justification would then not be possible. Wandschneider has therefore developed a counterargument to show that the contingency postulate of truth cannot be formulated without contradiction and implies the thesis that there is at least one necessarily true statement. This essay deals with the systematic question whether the contingency postulate of truth really cannot be presented without contradiction. To this end I will first present the contingency postulate and the associated problems (I.). Then I will analyze Wandschneider's argument against the consistency of the contingency postulate (II.) and finally reject it with the help of some considerations from the field of epistemic logic (III.). (shrink)

Infallibilism is the view that a belief cannot be at once warranted and false. In this essay we assess three nonpartisan arguments for infallibilism, arguments that do not depend on a prior commitment to some substantive theory of warrant. Three premises, one from each argument, are most significant: (1) if a belief can be at once warranted and false, then the Gettier Problem cannot be solved; (2) if a belief can be at once warranted and false, then its warrant can (...) be transferred to an accidentally true belief; (3) if a belief can be at once warranted and false, then it can be warranted and accidentally true. We argue that each of these is either false or no more plausible than its denial. Along the way, we offer a solution to the Gettier Problem that is compatible with fallibilism. (shrink)

Generally, an epistemic fallibilist considers it reasonable to claim, “I know that P, but I may be wrong.” An epistemic infallibilist, on the other hand, would consider this claim absurd. I argue initially that infallibilism presents more advantages in its assertion of the claim’s absurdity than fallibilism does in making the claim. One, infallibilism is not faulted with the propensity for violations of epistemic closure that beleaguers some fallibilist accounts, due in part to the latter’s problematic shunting of fallible epistemic (...) standards across inferential chains. Two, infallibilism is more easily modelled doxastically than fallibilism, as the former’s understanding of certainty is more tractable than the latter’s idea of what counts as a viable standard of fallibility. A rectification of these fallibilist issues may then be called upon to motivate gradualist variants of fallibilism. For epistemic gradualism, the problematic modellability and shunting of standards is curtailed via an awareness that the standard for knowledge is just one out of many along a gradient that includes, at one extreme, the infallibilist standard of certainty. Specifically, first, gradualism, along with infallibilism, can be modelled doxastically through elucidating upon an obtaining relation between doubt and knowledge; next, the problem with violating closure, although not generally applicable to infallibilism, can be answered by gradualism with a reasonable denial of closure altogether that makes precise which epistemic standard is relevant for which part of an inferential chain; lastly, these modelling resources can be appropriated, by both gradualism and infallibilism, to successfully address a doxastically pertinent form of closure violation called rational self-doubt. (shrink)

Epistemological contrastivism can model how infallible knowledge functions by employing the explanatory resource of structural differences between contrastive propositions, e.g., “P rather than Q”, and orthodox propositions, e.g., “P”. In doing so we notice that how this difference factors into our conception of infallible knowledge depends on two aspects: one, whether belief acts as a necessary condition for knowledge, and two, whether epistemic justification is construed as consciously internalist or non-consciously externalist. We further leverage the notion of phenomenal resolution, conceived (...) as an outcome of one’s discriminative capacities in accessing their evidence, to clarify in what sense it becomes reasonable to say that the truth of P mutually excludes the truth of Q when the latter is contrasted to the former. Importantly then, there is a way of infallibly knowing that P that is indistinguishable from infallibly knowing that P rather than Q, and a way that is not. (shrink)