Tom Swann Brandom Articulating Reasons Ch 3

Brandom - Articulating Reasons Chapter 3: Insights and Blindspots of Reliablism

I’m pretty sure that I was talking about pretty much the same stuff as this chapter in the bit of my Joyce essay where I interpreted his argument as an attribution of unreliability.<<<

##I

Platonic JTB recently turned into TB produced reliably

Founding Insight of Reliablism: true beliefs can be knowledge even if justification condition is not met P97 “(In the sense that the candidate knower is unable to produce suitable justifications)” provided arrived at by a reliable process.

ever? in all possible worlds? tension introduced by the modalised qualification.<<<

_J_TB was to prevent accidents from being K. enough to show that the belief “could…have been expected or predicted to be true”. fn2 ~(Kp->KKp)

The C American archaeologist. Reliable detector, but doesn’t affirm her disposition to judge. but others do. reasonable to say she already knows, even before she checks her evidence.

Against justificatory internalism, that is against the restriction of knowledge to cases where the candidate knower can offer reasons. fn3 the j.i. claim was never that K must be arrived at inferentially. but that reasons could be cited. e.g. one’s reliability as a noninferential reporter, esp perception. Reliabilism a kind of externalism.

but does not simply contradict JTB epistemology. a generalisation. reasoning is one potentially reliable process among others. reliability ought to trump justification. Where justification offered but process is unreliable, ~K

prefers to talk about reliability of perception, memory and testimony rather than as these offering reasons. because they give us knowledge at most in the cases where they are reliable.

temptation is to say that reliability can just supersede justification in epistemology. recentering. traditionally, beliefs noninferentially arrived-at would be knowledge iff the judger could cite her reliability as a reason. now reliabilism takes the citing of reasons as one process among other candidates for reliability.

why care about the truth of our beliefs? success of actions. why also care about knowledge? cause we want to rely on them. reliability legit object of epistemolo even if it doesn’t cohere with natural usage of ’k”

Three Questions here - p & Bp & R(Bp) & ~B(R(Bp))* -> Kp ? >>>*tricky. is it meant to be can’t be B(R(Bp))?<<< - warrant replacing J with R in JTB? - warrant replacing J concepts with R concepts, treating J only as mark of R?

Bradom: 1st but ~2nd and so ~3rd

##II Chicken Sexing and Super Blindsight

If S is Rp and S B(Rp), then if S Bp, then S has reasons to Bp.

Acquired noninferentially, but reasons could be given. So the examples have to get a subject who ~B(Rp) but Bp and R(Bp)

Chicken sexers good example. they get the mechanism they use wrong when they try to explain. quite unable to offer (good?) reasons for why they believe the chicks are the sexes they believe them to be. Ditto blindsighters. not believers, only ‘guessing’. need super blindsighters who actually believe. >>>or blindsighters who have learned they are reliable at the given task? … No he needs ~B(R(Bp))<<<

but both examples can be folded into J internalism. any * -> Kp ? >>>*tricky. is it meant to be can’t be B(R(Bp))?<<< - warrant replacing J with R in JTB? - warrant replacing J concepts with R concepts, treating J only as mark of R?

Bradom: 1st but ~2nd and so ~3rd

##II Chicken Sexing and Super Blindsight

If S is Rp and S B(Rp), then if S Bp, then S has reasons to Bp.

Acquired noninferentially, but reasons could be given. So the examples have to get a subject who ~B(Rp) but Bp and R(Bp)

Chicken sexers good example. they get the mechanism they use wrong when they try to explain. quite unable to offer (good?) reasons for why they believe the chicks are the sexes they believe them to be. Ditto blindsighters. not believers, only ‘guessing’. need super blindsighters who actually believe. >>>or blindsighters who have learned they are reliable at the given task? … No he needs ~B(R(Bp))<<<

but both examples can be folded into J internalism. any ?<<< - warrant replacing J with R in JTB? - warrant replacing J concepts with R concepts, treating J only as mark of R?

Bradom: 1st but ~2nd and so ~3rd

##II Chicken Sexing and Super Blindsight

If S is Rp and S B(Rp), then if S Bp, then S has reasons to Bp.

Acquired noninferentially, but reasons could be given. So the examples have to get a subject who ~B(Rp) but Bp and R(Bp)

Chicken sexers good example. they get the mechanism they use wrong when they try to explain. quite unable to offer (good?) reasons for why they believe the chicks are the sexes they believe them to be. Ditto blindsighters. not believers, only ‘guessing’. need super blindsighters who actually believe. >>>or blindsighters who have learned they are reliable at the given task? … No he needs ~B(R(Bp))<<<

but both examples can be folded into J internalism. any <<<

but both examples can be folded into J internalism. any J I Ism that didn’t allow noninferentially acquired B to be K would be nonstarter. >>>Davidson? or is it enough that he allows such to be knowledge iff it coheres.<<<

cognitively irresponsible to Bp if ~B(R(Bp)). still intelligible cases? Br thinks so, thinks they can be K, but still wants to call them cognitively irresponsible. >>>!!!<<< cognitively responsibility here just = don’t Bp if can’t offer any reasons. to treat these examples as incoherent is to say that Bp involves being able to offer reasons for Bp. >>>this is . still intelligible cases? Br thinks so, thinks they can be K, but still wants to call them cognitively irresponsible. >>>!!!<<< cognitively responsibility here just = don’t Bp if can’t offer any reasons. to treat these examples as incoherent is to say that Bp involves being able to offer reasons for Bp. >>>this is Mc Dowells view, isn’t it?<<< to impose this condition is to beg the question against the reliabilist. >>>unless we have some reason for it? or we could just disambiguate senses of B…<<<

Faith is surely not incoherent concept. reliable true faithfuls should be knowers.

not incoherent, but in principle exceptional. local phenomenon, not global.

##III Epistemology and Semantics

What would it be like for all knowledge to be like this? a possible autonomous set - this without any other? the J I Ist’s knowledge is autonomous in this sense. can imagine believers who know only in this way. condition for K is only that they form Bp only when B(R(Bp)) >>>is this really so concievable? how to individuate the processes candidate for reliability? any actual act of knowing will fall in indefinitely many reference classes.<<<

to have no reasons would be to make no judgments about reliability P107 “any attribution of reliability [with respect to some range of belief and some processes for getting to them] … inferentially underwrites a conclusion.

imagine a community precluded from giving reasons. cannot have the concept of reliability or knowledge. of course, its members can serve as measuring instruments. “But they cannot treat themselves and others as doing that. For they do not discriminate between reliable indication and unreliable indication. Absent such discrmination, they cannot be taken to understand themselves or one another as indicators at all. For the very notion of a correlation between the states of an instrument and the states that it is a candidate for measuring is unintelligible apart from the assessments of reliability.”

a lot being packed into the preclusion of reason-giving. Argument is: If no giving of reasons, no assessment of epistemic reliability, for that would give reasons. But what about assessments of reliability that are not offered as reasons?

what concept of unreliable indication?

either the relevant notion of correlation is already precisely that of reliability? or it isn’t, in which case there is room for a notion of correlation that is purely neutral on what is correlated with what - frequencies. the non-reason giving community could make judgments of this sort? <<<

thinks this gives reasons to deny that they have knowledge. but only ‘probative, not dispositive’.

but P108 “states that do not serve as reasons one for another, are not recognizable as beliefs at all.” what makes the difference between the parrot and the believer? application of concepts minimally: the inferential articulation of the response. beliefs are inference-apt. they are propositional and so conceptual. >>>but open question whether application of concepts without as part of propositions<<< believing is adopting a stance with extra commitments. P109 “No response that is not a node in a network of such broadly inferential involvements, I claim, is recognizable as the application of concepts.”

Not open to the reliabilist to refuse to draw a line between belief and all other products of processes. If so, no longer talking about knowledge.

Conceptual Blindspot of reliablism: failure to realise that “what distinguishes propositionally contentful and therefore conceptually articulated beliefs, including those that qualify as knowledge, from those the merely reliable responses or representations of noncognitive creatures… is (at least) that they can both serve as and stand in need of reasons.” P109

P110 “The examples of knowledge based on reliability without the possibility of offering reasons, which motivate the Founding Insight, are essentially fringe phenomena. Their intelligibility is parasitic on that of the reason-giving practices that underwrite ordinary ascriptions of knowledge—and indeed of belief tout court

I think I want to deny the original case. No reliability without being able to assess as reliable and so have this reliability provide one with reasons. In the cases where this possibility is not actual, the knower is cognitively irresponsible, and so does not know. The <<<

##IV Reliabilism and Naturalism

The semantic replacement version of reliabilism is a Strawman?

but there is another temptation: to naturalise epistemology via reliabilism. P111 “if and insofar as the concept of (propositionally contentful) belief can be naturalized, so can the concept of knowledge.” a conditional claim, but not a trivial one. “if the notions of belief and truth have been explained physicalistically or naturalistically—a substantive task, to be sure, but perhaps not a distinctively epistemological one—then one of the promises of reliabilism in epistemology is that all one needs to extend those account so encompass also knowledge is a naturalistic story about objective likelihood.”

Br thinks that many, even those who deny the antecedent, think the conditional is true. even obviously so. Br denies the conditional.

##V Barn Facades and Goldman’s Insight

Objective probabilities are relative to reference classes. P113 “The proper choice of reference class is not itself objectively determined by facts specifiable in a naturalistic vocabulary. So there is something left over.” >>>wtf?<<<

Goldman’s barn facade (1976). ironic as it was introduced in order to undermine causal in favour of reliability theories of knowledge.

internal and external knowledge ascriptions? why think there is one concept?<<<

barn facade county is enough to show that classical justificatory epistemological internalism is inadequate. causal connection also inadequate - here the causal connection is as it would be in a knowledge case. what is needed is the condition of reliability.

what is special about Barn Facade cases is that what undermines the knowledge is external to the subject’s beliefs and causal connections.

Goldman’s Insight is an epoch making move >>>really??<<< but does it support naturalism in epistemology?

Br thinks not. Goldman’s example ‘literalizes’ the boundary of a reference class. what if Barn Facade County were one of many counties, the rest of which had real barns? or a country with mostly facades. or a world with mostly real barns. or the reference class of one…

maybe one can get around this by thinking about relationships between different possibilities? it was very likely that the belief would not be true. is this the ‘safety’ condition? but the very likely safe faces the same kind of problems?

goal-relative?<<<

P116 “nothing in the way the world is privileges one of those reference classes” reliability varies as we describe the mechanism and as we describe the believer. P117 “no naturalistically statable facts pick out one or another of these descriptions as uniquely privileged or correct.” Naturalistic Blindspot of reliabilism.

##VI Inference and the Implicit Insight of Reliabilism

Implicit Insight: “concern with reliability should itself be understood as concern with the goodness of a distinctive kind of inference.”

epistemology typically concerns attribution of knowledge. formal mode rather than material. not just asking “what is knowledge” but “when is ‘knowledge’ properly applied?”.

what do we ascribe in ascribing knowledge? inferential commitments. entitlement. truth. to ascribe truth is to endorse the claim oneself. if we think of truth ascriptions as ascribing properties, describing a statement, we get spurious metaphysics. truth ascriptions are expressions of one’s stance.

expressivist. but surely there is room for understanding some sort of property ascription? that the sentence says how things are? <<<

knowledge as a normative social status in Making it Explicit.

what are we doing when we take someone to be a reliable noninferential believer of some p.  so P120 “endorsing a distinctive kind of inference: an inference, namely, from the attribution to another of a propositionally contentful commitment acquired under certain circumstances to the endorsement or undertaking oneself od a commitment with the same content… Endorsing such an inference is just what being prepared to rely on someone else as an informant consists in”.

relibaility inferences play central role in the game of giving and asking for reasons. as much as the related but distinguishable role of truth ascriptions. Implicit Insight: that “concern with reliability is not opposed to concern with what is a reason for what, but actually a crucially important species of it”. is externalism, but only in the sense that we can ascribe knowledge even where the subject assessed does not possess reasons. one must thereby have one’s own reasons for the ascription. reliability assessments not “external to the game of giving and asking for reasons”. reliabilism points to the social game.

tenuous link, surely? the point could be made with any theory of knowledge? any condition that we ascribe such as to ascribe knowledge will thereby give us reason to believe it ourselves. and what link especially with the social game?<<<

Inferentialist expressivist account of knowledge helps clarify Goldman’s insight P121 The “relativity to reference class of assessments of reliability (and hence of knowledge) that seemed so puzzling when viewed in a context that excluded concern with what is a reason for what falls naturally into place once we understand assessments of reliability as issues of what inferences to endorse. The different reference classes just correspond to different (true) collateral premises or auxillary hypotheses that can be conjoined with the attibution of noninferentially acquired perceptual belief in order to extract inferential consequences the assessor of reliability (and knowledge) can use as premises in her own inferences.”

given the premise of where the barn perciever is, she can infer whether she should take her perception to be reliable or not. so knowing about the state will license an inference to her reliability, knowing about the country…

must assume something about the distribution of probability for the barns / facades in the different ref classes that one is looking at them.<<<

all the premises are true, each gives a different conclusion. but this is fine because they give different inferences.

“The relativity to description that is threatening to an understanding of reliability and knowledge that ignores reason giving, justification and inference can be taken in stride once we see concern with reliability as arising in just such contexts. For we expect the goodness of inferences to be sensitive to differences in how the items we are reasoning about are described. The intensionality of assessments of reliabilty is just a mark of their membership in the inferential order rather than the causal order.”

should expect them to be nonmonotonic. see Chapter 2.