Knowledge and credit

Abstract

A widely accepted view in recent work in epistemology is that knowledge is a cognitive achievement that is properly creditable to those subjects who possess it. More precisely, according to the Credit View of Knowledge, if S knows that p, then S deserves credit for truly believing that p. In spite of its intuitive appeal and explanatory power, I have elsewhere argued that the Credit View is false. Various responses have been offered to my argument and I here consider each of these objections in turn. I show that none succeeds in undermining my argument and, thus, my original conclusion stands—the Credit View of Knowledge is false.

A widely accepted view in recent work in epistemology is that knowledge is a cognitive achievement that is properly creditable to those subjects who possess it. More precisely, according to the Credit View of Knowledge, if S knows that p, then S deserves credit for truly believing that p. So, for instance, Ernest Sosa claims that “[b]elief amounts to knowledge when apt: that is to say, when its correctness is attributable to a competence exercised in appropriate conditions” (Sosa 2007, p. 92). Similarly, John Greco argues that “…knowledge attributions can be understood as credit attributions: when we say that someone knows something, we credit them for getting it right” (Greco 2007, p. 57). And Wayne Riggs holds that “S knows that p only if being right about p in this instance is attributable to S as a cognitive agent” (Riggs forthcoming, p. 1 of ms.).¹

But what exactly does it mean for a subject to be deserving of credit for a true belief? Different answers to this question are provided by the various proponents of the Credit View. According to Sosa, “…we might understand success due to an agent’s competence as success that manifests that competence, a special case of the manifestation of a disposition” (Sosa 2007, p. 80, original emphasis). In contrast, Greco understands credit in causal/explanatory terms, such that a subject deserves the relevant kind of credit when she believes truly because of her reliable cognitive faculties. More precisely, S deserves credit for her true belief that p when S’s reliable cognitive faculties are the most salient part of the cause explaining why S holds the true belief in question. And Wayne Riggs claims that “Knowing that p…requires truly believing p in a way that is neither veritically lucky nor accidental. Under those conditions, one is creditworthy for having come to a true belief” (Riggs forthcoming, p. 3 of ms.). Veritic luck, according to Riggs, is what is present in a lucky guess, and “one avoids accidental true belief when something like a ‘desire for one’s belief to be true rather than false’ is sufficiently operative in regulating one’s doxastic performance in the production of the belief” (Riggs forthcoming, pp. 2–3 of ms.).

Despite these different conceptions of credit, however, all proponents of this view of knowledge agree that credit is deserved for those true beliefs that qualify as knowledge, but not for those acquired in Gettier or Gettier-type situations. Such a claim has a great deal of intuitive appeal, which is often emphasized by considering other kinds of achievements: when Allen Iverson, for instance, makes a much needed 3-point shot in a basketball game, this is a success that is due to his impressive athletic ability. Now compare this to my making a 3-point shot, which would undoubtedly be purely the result of good luck. In both cases, the ball ends up in the basket after being shot from behind the 3-point line, but only in the former is the player deserving of credit for this success. Similarly, according to proponents of the Credit View, true belief acquired in a case of knowledge is a success that is due to the subject’s cognitive ability, but true belief in Gettier or Gettier-type situations is a success that is due to good luck. Thus, while true belief is acquired in both cases, it is only in the former that the subject is deserving of credit for this cognitive achievement.

It should be clear that the Credit View, if correct, has enormous explanatory power. First, it provides an explanation of the widely accepted thesis that knowledge is incompatible with luck. If a subject’s true belief is the result of good luck, then this success is not properly attributable to her cognitive faculties and is, thus, not an instance of knowledge. Second, and related, such a view sheds light on what is absent in Gettier and Gettier-type cases: when a success, cognitive or otherwise, is disconnected from a subject’s ability, then it is not an achievement creditable to the subject herself. And, finally, it explains the additional value that knowledge has over justified belief that is merely accidentally true: just as a basket made through athletic ability is more valuable than one made via good luck, so, too, hitting upon the truth through cognitive ability is more valuable than doing so via good luck.

In spite of its intuitive appeal and explanatory power, I have elsewhere argued that the Credit View is false.² One of my central arguments against this view relies on the following type of case:

CHICAGO VISITOR: Having just arrived at the train station in Chicago, Morris wishes to obtain directions to the Sears Tower. He looks around, approaches the first adult passerby that he sees, and asks how to get to his desired destination. The passerby, who happens to be a lifelong resident of Chicago and knows the city extraordinarily well, provides Morris with impeccable directions to the Sears Tower by telling him that it is located two blocks east of the train station. Morris unhesitatingly forms the corresponding true belief.

I have argued that while Morris clearly knows on the basis of testimony that the Sears Tower is two blocks east of the train station, he does not deserve the requisite kind of credit for truly believing this proposition, thereby showing that the Credit View of Knowledge is fundamentally incorrect. In a nutshell, I argued that whatever notion of credit the proponent of the Credit View invokes, it has to be robust enough to rule out subjects in Gettier and Gettier-type situations from deserving credit for their true beliefs, yet weak enough to allow subjects in testimonial cases, such as Morris in CHICAGO VISITOR, to be deserving of credit for their true beliefs. And this, I argue, is a task that is doomed to failure.³

Various responses have been offered to my argument and I shall here consider each of these objections in turn. I shall show that none succeeds in undermining my argument and, thus, my original conclusion stands—the Credit View of Knowledge is false.

1 Credit and knowledge go hand in hand

One of the more common responses to CHICAGO VISITOR can be put in the form of a dilemma: either Morris doesn’t deserve credit for the truth of the belief that he acquires on the basis of the passerby’s testimony, but then neither does he acquire the relevant testimonial knowledge; or Morris does acquire the testimonial knowledge in question, but then so, too, does he deserve credit for the truth of the belief about the whereabouts of the Sears Tower. Either way, credit and knowledge go hand in hand.

The first horn of this dilemma is defended by Riggs in the following passage:

The first task…is to call into question the attribution of knowledge to Morris in Lackey’s example. I am surprised that it is offered as an uncontroversial example of testimonial knowledge. Why on earth would we say that Morris knows where the tower is when he has picked a stranger at random, and unhesitatingly (and, one assumes, unreflectively) accepted what that person said? On the face of it, this is terrible epistemic practice. Intuitively, more is required on the part of the hearer than simply opening his brain and putting into it whatever some random stranger has to say. (Riggs forthcoming, p. 10 of ms.)

According to Riggs, then, since Morris does not acquire the testimonial knowledge in question in CHICAGO VISITOR, it clearly fails to provide a counterexample to the Credit View of Knowledge. The second horn of the dilemma is advanced by Greco. He begins by arguing that his preferred virtue-theoretic view provides a unique approach to analyzing testimonial knowledge. He writes:

Often theories of testimonial knowledge are divided into two camps. On the first kind of theory, what is important for testimonial knowledge is that the source of testimony is in fact reliable. On the second kind of theory, it is also important that the believer knows, or at least justifiably believes, that the source is reliable. From a virtue-theoretic perspective, however, a third kind of theory becomes plausible. Namely, testimonial knowledge requires that the believer is a reliable receiver of testimony. (Greco 2007, p. 63)

Given this conception of testimonial knowledge, Greco argues that if Morris indeed possesses the relevant testimonial knowledge in CHICAGO VISITOR, then he must be a reliable receiver of testimony, which in turn requires “…reliable capacities for discriminating reliable sources of testimony from unreliable sources” (Greco 2007, p. 63). According to a virtue-theoretic account of testimonial knowledge, then, if Morris knows the whereabouts of the Sears Tower, he does deserve credit for this true belief. In particular, “his success is grounded in his ability to discriminate good from bad testimony and is therefore attributable to him” (Greco 2007, p. 63).

By way of response, there are several points I shall make which, together, address both horns of this dilemma. Let me begin with Greco’s purportedly alternative, virtue-theoretic conception of testimonial knowledge. Now, the first point to notice is that his claim that the acquisition of testimonial knowledge requires a reliable receiver of testimony is neither novel, nor unique to virtue-theoretic accounts. Indeed, in my own work on testimony, I have argued that a “properly functioning or reliable recipient of testimony” condition needs to be added to any plausible account of testimonial knowledge, regardless of whether it is non-reductionist, reductionist, virtue-theoretic, and so on.⁴ While I have argued that such a requirement involves a hearer’s exercising the capacity for being appropriately sensitive to defeaters and possessing minimal positive reasons,⁵ Greco argues that it involves reliable capacities for discriminating reliable sources of testimony from unreliable sources. The point I wish to emphasize here, however, is that if a virtue-theoretic approach opens new doors in the epistemology of testimony, requiring that testimony be reliably received is not one of them.

This leads to my second point: given my own requirements on testimonial knowledge, I certainly never presented CHICAGO VISITOR as involving a hearer who is “simply opening his brain and putting into it whatever some random stranger has to say,” much like a robot would. On my view, testimonial knowledge requires that Morris exercise the capacity to be appropriately sensitive to defeaters—if, for instance, he would accept the passerby’s testimony even if he appeared highly intoxicated or told him that the Sears Tower was in France, then he would clearly not acquire the knowledge in question. In addition, I require the presence of minimal positive reasons for rational acceptance of testimony—if Morris had no relevant beliefs about humans’ general testifying habits, or about the reliability of humans when offering directions, or about Chicago, and so on, then, once again, I would deny that testimonial knowledge has been acquired. If these rudimentary abilities are what Greco means by “…reliable capacities for discriminating reliable sources of testimony from unreliable sources,” then we agree on whether Morris acquires knowledge in CHICAGO VISITOR. However, this is not sufficient for creditworthiness. For surely Morris’s being reliable at discriminating between the intoxicated and the sober, and between those who believe that the Sears Tower is in France and those who believe that it is in Chicago, does not explain in any substantive sense why he ends up with a true belief about the precise whereabouts of the tower. Instead, it is the passerby’s familiarity with the city of Chicago and her experience with the Sears Tower that explains his true belief. In other words, Morris’s knowledge is not success because of his cognitive ability but success because of the passerby’s cognitive ability.

But what if Greco means something more robust by “…reliable capacities for discriminating reliable sources of testimony from unreliable sources” than the rudimentary abilities I require for testimonial knowledge? What if he instead means that, relative to a given domain, I am able to reliably discriminate between the competent and the incompetent, the sincere and the insincere? For instance, I have enough experience with stories in The New York Times being confirmed by multiple other independent sources, and with the ludicrous cover stories of The National Enquirer, to deliberately choose the former and avoid the latter in my consumption of news. In this way, I am playing a significant role in the acquisition of these testimonial beliefs. What if something similar were required of Morris in CHICAGO VISITOR? If Greco generally requires this sort of reliable discrimination for testimonial knowledge, he will be forced to embrace a limited form of skepticism regarding our testimonial beliefs. To see this, let us ask: what does an honest, competent passerby look like that would enable Morris to reliably distinguish her from a dishonest or directionally challenged one? Sure, when asking for directions, most of us would reliably choose an adult passerby rather than a toddler, a sober person rather than one who appears intoxicated, one with a native accent rather than one with a camera and a guidebook, and so on. But among adult, sober, humans not carrying guidebooks, there can be substantial variation in the competence and sincerity of their testimony. Thus, on any ordinary reading of CHICAGO VISITOR, Morris could have plausibly approached a competent-looking liar or a directionally challenged speaker in much the same way that he did an honest, knowledgeable, Chicago resident when asking for assistance in finding the Sears Tower. But if Morris’s behavior is equally compatible with all of these outcomes, then it is clear that he is not reliably discriminating reliable sources of testimony from unreliable ones in the robust sense under consideration. Hence, Greco will be forced to deny that knowledge can be acquired in the sort of scenario envisaged in CHICAGO VISITOR.

Some may think that it is not necessarily a problem if we deny that knowledge is acquired when we find ourselves in new cities asking for directions. But situations similar to that found in CHICAGO VISITOR arise with respect to countless other areas in which testimony is not only frequently accepted, but is also intuitively acceptable. Most of us, for instance, are not very reliable at discriminating reliable testimony from unreliable testimony when people whom we first meet report their names, occupations, family histories, and so on. For in such circumstances, liars and incompetents typically fail to have identifiable marks announcing their deception and incompetence, and those who are honest and competent in such matters rarely can be picked out as such. To put this more concretely, how on earth would you be able to tell that the woman next to you on the airplane is lying when she tells you that her name is Amanda, or that she is a nurse, or that she has 3 children, or that she lives in Albuquerque? Similar considerations apply when we find ourselves confronting entirely new areas of inquiry for the first time. How, for instance, would a high school student learning U.S. history for the first time be able to tell that her teacher is slightly off with respect to most of his dates about wars? And would the student who is taking physics for the first time be in any position whatsoever to assess the reliability of his teacher? Moreover, infants and young children often engage in even less discriminatory behavior on a wider range of topics than many adults do. For instance, many 3-year-olds would not be able to reliably discriminate reliable preschool teachers from unreliable ones, or reliable televisions programs from unreliable ones, or reliable books from the public library from unreliable ones. Thus, the Credit View of Knowledge may be saved by making the requirement about what the reliable reception of testimony amounts to extremely demanding, but it does so at the cost of embracing a limited version of skepticism about testimonial knowledge. I, for one, think this price is far too high to pay.

Thus, I have argued for three central conclusions: (1) on any plausible conception of testimonial knowledge, Morris can be said to know the whereabouts of the Sears Tower in CHICAGO VISITOR; (2) Morris does not deserve the requisite credit for the truth of this belief; and (3) denying Morris knowledge in CHICAGO VISITOR leads to a limited form of skepticism about testimonial knowledge.

2 Credit can be shared

A second general response offered to CHICAGO VISITOR is that my argument implicitly relies on assuming that credit for true belief cannot be shared. For instance, Greco argues that:

…credit for success, gained in cooperation with others, is not swamped by the able performance of others. It’s not even swamped by the outstanding performance of others. So long as one’s own efforts and abilities are appropriately involved, one deserves credit for the success in question. (Greco 2007, p. 65)

In a similar spirit, Riggs argues:

Why do we suppose that someone has to get all the credit? Why not just say that both the parties involved get some credit for the recipient’s true belief? It is vanishingly rare for any human being to accomplish anything completely on the basis of his own powers and abilities alone. And yet, even in many of those cases, we unhesitatingly attribute such accomplishments to people. (Riggs forthcoming, p. 17 of ms.)

Thus, once it is acknowledged that credit can be shared by both speaker and hearer, it is thought that there is no longer a problem posed by CHICAGO VISITOR for the Credit View of Knowledge.

By way of response, let us begin by examining the second part of Greco’s claim above; namely, that so long as one’s own efforts and abilities are appropriately involved, one deserves credit for the success in question. Now, the crucial task for Greco, and indeed for any proponent of the Credit View, is to flesh out the precise sense in which a subject’s own efforts and abilities must be “appropriately” involved for creditworthiness. But there is a serious problem lurking: a subject’s own efforts and abilities are often importantly involved in the acquisition of true beliefs in Gettier-type cases, which stand as a paradigm of the sorts of true belief that are supposed to lack creditworthiness. The sense in which a subject’s own efforts and abilities must be “appropriately involved,” then, must be strong enough to rule out deserving credit in Gettier-type cases, yet weak enough to render subjects deserving of credit in ordinary cases of mundane knowledge, such as that found in CHICAGO VISITOR. The prospects for successfully striking this delicate balance are, I suggest, bleak.

To see this, consider the following familiar Gettier-type case: while driving from Iowa to Illinois, Fiona looks out her window, sees a barn in the distance, and forms the corresponding true belief. As it happens, the barn she sees is the only real one, completely surrounded by barn façades that the local farmers have erected to make their community appear prosperous. Now, in this Gettier-type case, Fiona’s own efforts and abilities are surely importantly involved in her truly believing that there is a barn in the field: she trusts her reliable faculty of vision and she forms her belief on the basis of her veridical perceptual experience of a barn. Yet, according to the proponent of the Credit View—who maintains that deserving credit for holding a true belief is what renders knowledge different from and more valuable than those beliefs that are true merely by luck—Fiona does not deserve credit for truly believing that there is a barn in the field. In contrast, Morris, whose own efforts and abilities have very little to do with why he has a true belief about the whereabouts of the Sears Tower, does deserve credit for his true belief. Here is the problem: Fiona’s own efforts and abilities are far more intimately and importantly involved in the true belief that she acquires than Morris’s are in the true belief that he acquires. For Fiona’s reliable vision and veridical perceptual experience shoulder much of the explanatory burden for why she formed the true barn belief, while Morris’s ability to choose a conscious, adult human to ask for directions shoulders very little of the explanatory burden for why he formed a true belief about the whereabouts of the Sears Tower. Sure, Fiona is lucky to have looked at the one real barn, but Morris is also lucky to have chosen a Chicago resident who knows the city extraordinarily well when asking for directions. The central difference in luck in these two cases is that Morris’s environment is far better suited than Fiona’s is to the formation of true beliefs. But this difference has nothing to do with the epistemic effort, virtues, or faculties of the respective subjects, and hence it has nothing to do with whether they deserve credit for their true beliefs.

To my mind, this is the central problem afflicting the Credit View of Knowledge. Let us formulate this objection even more precisely as follows:

Creditworthiness Dilemma: Either the notion of creditworthiness operative in the Credit View of Knowledge is robust enough to rule out subjects from deserving credit for the truth of their beliefs in Gettier-type cases, but then neither is credit deserved in CHICAGO VISITOR-type cases; or the relevant notion of creditworthiness is weak enough to render subjects deserving of credit for the truth of their beliefs in CHICAGO VISITOR-type cases, but then so, too, is credit deserved in Gettier-type cases.

As should be clear, either horn of this dilemma undermines the Credit View of Knowledge at its core. For, on the first horn, credit may be adequately blocked in Gettier-type cases, but only at the expense of also blocking credit in countless cases where testimonial knowledge is intuitively present despite minimal work being done on the part of the hearer. And, on the second horn, credit is secured in cases of testimonial knowledge where such minimal work is done by the recipient, but only at the expense of also securing credit in Gettier-type cases. Either way, the Credit View not only fails to shed light on what is absent in Gettier-type cases, but it also fails to explain the additional value that knowledge has over merely accidentally true belief.

It should also be clear that my argument against the Credit View does not at all depend on assuming that credit for true beliefs cannot be shared. Surely, there are all sorts of ways in which this can happen. You and I may jointly work on a scientific experiment and thus both be equally responsible for the truths that we uncover. Or you and I may collaborate while bird watching and, through our combined ornithological knowledge, together correctly identify the bird in the distance as a golden eagle. There are also cases of testimonial knowledge where shared credit between speaker and hearer does not seem entirely misplaced. If I do extensive research on your background in medicine and choose to consult you because of your outstanding credentials, it may be appropriate to say that I deserve some credit for the truth of the belief I form on the basis of your medical testimony because my cognitive faculties were so intimately involved in my choice of source. None of this, however, addresses in any way the fundamental problem that the Creditworthiness Dilemma poses to the Credit View of Knowledge.

3 Sosa’s response

A third response to CHICAGO VISITOR can be found in recent work by Sosa, where he takes up the problem that testimony generally poses for the Credit View of Knowledge. He writes:

Any belief that is knowledge must be correct, but must it be correct due to an epistemic competence? That seems strained at best for knowledge derived from testimony…. Others no doubt made the relevant discovery—perhaps a historian, or a detective, or a scientist, or a physician—and the information was then passed on, resulting in some later recipient’s belief, whose correctness then owes little to his own individual accomplishment, if all he does is to receive the information. (Sosa 2007, p. 93)

By way of answering this problem posed by testimonial knowledge, Sosa argues that despite the minimal role played by the epistemic competence of the hearer in the acquisition of a true testimonial belief, such a subject still deserves partial credit for the correctness of her belief, and this suffices for the relevant notion of creditworthiness operative in the Credit View of Knowledge. For instance, a quarterback may exercise a competence by throwing a touchdown pass, but this individual accomplishment is part of a broader competence in the whole offensive team. Thus, while the individual player certainly deserves credit for this great pass, it is only partial given the crucial role of his other teammates. Similarly, Sosa argues that in the case of testimonial belief, the individual hearer exercises a competence by receiving the testimony in question, but this individual performance is part of a broader competence in a collective social group. Hence, “[t]he correctness of one’s belief is still attributable in part to a competence seated in oneself individually, but the credit that one earns will then be partial at best” (Sosa 2007, p. 95).

But if partial credit is sufficient for the Credit View of Knowledge, then doesn’t Sosa face the second horn of the Creditworthiness Dilemma? In particular, while partial credit may be weak enough to render subjects deserving of credit for the truth of their beliefs in cases of testimonial knowledge, won’t credit likewise be deserved in Gettier-type cases? Sosa provides a threefold response to this question. First, he distinguishes among at least two different kinds of Gettier cases. On the one hand, there are what we may call traditional Gettier cases, a paradigm of which is the following:

NOGOT/HAVIT: Anya has ample evidence supporting her belief that Nogot, who works in her office building, owns a Ford. She has, for instance, repeatedly seen Nogot driving a Ford to and from work, Nogot frequently wears a T-shirt that boasts, “Proud owner of a Ford,” Nogot showed Anya sale papers that indicated that he had purchased a Ford, and so on. From her justified belief that Nogot owns a Ford, Anya draws the existential conclusion, “Someone in my office building owns a Ford.” It turns out that Nogot does not in fact own a Ford—he has been driving his sister’s car and forged the sale papers he showed to Anya. But the existential conclusion that Anya drew from her false belief is nonetheless true because Havit, who also works in her office building, does indeed own a Ford.

On the other hand, there are what we may call extended Gettier cases, a paradigm of which is the barn façade case discussed in the previous section. What is the precise difference between these two kinds of Gettier cases? Various proposals have been offered in the literature. Duncan Pritchard, for instance, argues that “intervening luck” is found in traditional Gettier cases, where this is understood as “luck that intervenes between ability and success, albeit in such a way that the success is preserved” (Pritchard forthcoming, p. 3 of ms.). In contrast, Pritchard claims that extended Gettier cases involve “environmental luck,” where this is understood as luck that “…concerns the environment in which ability generates that success” (Pritchard forthcoming, p. 4 of ms.). For our purposes, we can simply grant at this point that there is a difference between traditional and extended Gettier cases without settling what exactly it amounts to. Now, given this distinction, Sosa moves to his second point where he argues that not even partial credit is deserved for the truth of the beliefs acquired in traditional Gettier cases. To this end, he distinguishes between a competence explaining the existence of a belief and a competence explaining the correctness of a belief. For instance, consider again the NOGOT/HAVIT case. According to Sosa, “[t]he reasoning by way of Nogot does of course help explain why the believer has that belief, but it does not in the slightest help explain its correctness” (Sosa 2007, p. 96, emphasis in the original). Finally, Sosa grants that, because extended Gettier cases involve apt belief that is correct in a way that is sufficiently attributable to the exercise of the subject’s competence in its proper conditions, knowledge is indeed acquired in such cases.⁶ Thus, with respect to extended Gettier cases, he essentially accepts the second horn of the CREDITWORTHINESS DILEMMA.

In response to Sosa’s view, I shall raise three concerns and offer one main argument against this sort of defense of the Credit View of Knowledge. Let us begin with the concerns. First, Sosa introduces the problem as one involving testimonial knowledge in general. But to my mind, focusing on the issue of testimony at this level of generality misses the force of CHICAGO VISITOR-type cases in particular. There are countless ways in which we go about choosing our testimonial sources and forming beliefs on the basis of what they say. For instance, in seeking an answer to one’s question about World War II, one may thoroughly investigate all of the expert historians on this topic, evaluate their reputations, credentials, and areas of specialization, and then deliberately choose the best to consult on this matter. Or when choosing one’s source of news, one may do extensive research on the reliability, backgrounds, and experience of the writers for all of the major newspapers and then specifically choose The New York Times as one’s regular newspaper. In such cases, it does not seem strained to say that such recipients of testimony deserve some credit for the truth of the relevant beliefs that they acquire, and thus it is not clear that they even pose a problem for the Credit View of Knowledge. At the other end of the spectrum, however, is a case such as CHICAGO VISITOR, where the hearer in question chooses the first conscious adult passerby that he sees in a new city to ask for directions. Here, the absolutely minimal work being done by the recipient of testimony casts serious doubt on the plausibility of him deserving credit for the truth of his belief. And, of course, there are all sorts of cases in between, where more than the minimal work in CHICAGO VISITOR, but less than that involved in the WWII and New York Times beliefs, is done by the hearer in a testimonial exchange. Thus, I think that treating the problem posed to the Credit View of Knowledge as one involving testimonial beliefs in general groups together under a single category importantly different epistemic phenomena.

This leads to my second concern: it is not clear that even partial credit is deserved by Morris of the true belief that he acquires in CHICAGO VISITOR, nor is it obvious that most of the credit in such a case is due to a “complex social competence” (Sosa 2007, p. 97). To see this, recall that the analogy that Sosa relies on in elucidating the notion of partial credit at issue is that of the quarterback who throws a touchdown pass but shares credit of this accomplishment with the other members of his team. It is quite intuitive that credit is shared in such a case: the quarterback is exercising his competence as a football player in successfully throwing the ball to one of his teammates, and yet the touchdown pass would not happen without the other players, one of whom must actually catch the ball in order for it to even be a touchdown pass. But what competence does Morris exercise in asking the first adult, conscious passerby that he sees for directions that justifies granting him partial credit for the truth of the belief in question? Of course, as already noted, he knows not to ask a toddler, an obviously intoxicated adult, and so on. This minimal cognitive work, however, seems more analogous to a quarterback who knows not to throw the football to a member of the opposing team or to the fans watching the game, but is quite unaware that the player who in fact catches his pass is well-situated to do so. In other words, just as Morris is not at all responsible for choosing a lifelong resident of Chicago who knows the city extraordinarily well, the parallel situation in football would be that of a quarterback who is not at all responsible for choosing a player who is in an excellent position to catch his pass. But then attributing partial credit in either case for the respective successes seems misplaced. Moreover, why is the correctness of Morris’s belief primarily creditable to a complex social competence rather than simply to the individual passerby’s competence? After all, it is the passerby’s extensive experience with the city of Chicago, and with the whereabouts of the Sears Tower in particular, that shoulders most of the explanatory burden for Morris’s true belief. Indeed, everything could be exactly as it is in CHICAGO VISITOR, except Morris asks a passerby who always confuses the Sears Tower with the John Hancock building and thus ends up conveying incorrect directions to him. The broader complex social competence is the same in both cases, yet Morris acquires a true belief in only one, depending on whether the passerby in question knows the layout of the city of Chicago. To my mind, this casts serious doubt on attributing creditworthiness to a broader social competence in CHICAGO VISITOR rather than simply to the individual testifier.

Third, granting knowledge in extended Gettier cases is arguably an unwelcome concession. While defending this claim at length lies beyond the scope of this paper, it is worth pointing out that such a conclusion flies in the face of some deeply held intuitions in epistemology. For instance, the barn beliefs of a subject who is surrounded by barn façades will fail to be both sensitive and safe; that is, such a subject would still believe that there is a barn even if she were in fact seeing a barn façade, and such a subject would believe that there is a barn without it being so that there is one. Thus, granting knowledge in such cases is incompatible with any epistemological view that includes as a necessary condition for knowing either sensitivity or safety.⁷ Moreover, it is a widely accepted thesis that subjects who are unable to discriminate among relevant alternatives in a given domain do not possess the knowledge at issue.⁸ For instance, if I am completely unable to distinguish between a Siberian Husky and an Alaskan Malamute, then it is quite plausible to conclude that I do not know that the dog next door is a Siberian Husky even if in fact it is one. Once again, then, granting knowledge in extended Gettier cases is at odds with a deep commitment in epistemology since subjects in barn façade cases are unable to discriminate between the one real barn and the many surrounding fakes.

At this point, I have focused on some broad concerns with appealing to partial credit to vindicate the Credit View of Knowledge. I shall now turn to my direct argument against Sosa’s defense of this view. To begin, consider the following:

TWO SHEEP: While taking a walk in the country in late October, Nolan stops in front of a farm to admire the animals. After seeing what appear to be two sheep grazing in the field next to a large rock, he forms the belief, “There are two sheep in this field.” It turns out that while one of the sheep Nolan sees is real, the other is a goat that the farmer has dressed up as a sheep for Halloween. However, behind the rock and out of Nolan’s sight, there is a second real sheep, thereby rendering his belief that there are two sheep in the field true.

Notice, first, that, by all accounts, TWO SHEEP seems to be a traditional Gettier case. If one wishes to adopt Pritchard’s distinction, for instance, the luck involved is of the intervening rather than the environmental sort; that is to say, the luck at issue intervenes between Nolan’s perceptual ability and his success in acquiring a true belief, albeit in such a way that the success is preserved. This stands in contrast to the barn façade case, where the luck is in the environment in which the subject’s cognitive ability successfully acquires a true belief. But, however, one wishes to cash out the difference between traditional and extended Gettier cases, TWO SHEEP seems to clearly fall in the former camp.

Moreover, if one wishes to quibble about this conclusion, it is certainly not difficult to construct a modified version of one of Gettier’s own counterexamples along lines similar to those found in TWO SHEEP. Consider, for instance, the following:

TWO HIRES: Bennett’s office is currently in the process of hiring two new employees, and he has excellent evidence for believing that the jobs will be offered to two of his friends, Smith and Jones. For instance, Bennett overheard his boss saying on the phone that Smith and Jones will get the positions, he knows that they are both exceptionally well-qualified for the jobs, and he saw employee papers on the secretary’s desk with the names of Smith and Jones written on them. Bennett also has excellent reason for thinking that Smith and Jones each have ten coins in their respective pockets, having seen both of then empty the contents of their pockets, count the coins, and then place them back in their pockets. From this, Bennett concludes that the two men who will get the jobs have ten coins in their pockets. It turns out that while Smith will get the job and does have ten coins in his pocket, the boss misread the name of the other applicant when he spoke on the phone and had the secretary write up the employee papers—in fact, the job is going to James not Jones, who also just so happens to have ten coins in his pocket.

As was the case in TWO SHEEP, Bennett’s belief about the new hires is true, justified, and yet not an instance of knowledge. The similarity of TWO HIRES to Gettier’s own cases, however, completely precludes trying to deny that it is a traditional Gettier case.

The second point to notice about TWO SHEEP is that Nolan clearly deserves partial credit for the truth of his belief that there are two sheep in the field in question. For at least part of the justification for his belief—namely, that involving the first real sheep—is grounded in a veridical visual experience, which in turn results from the exercise of a competence in proper conditions. Otherwise put, part of Nolan’s true belief is successful precisely because of his cognitive ability. Similar considerations apply in the case of TWO HIRES: Bennett clearly deserves partial credit for the truth of his belief that the two men who will get the jobs have ten coins in their pockets since part of the justification for this belief is based on excellent evidence that appropriately grounds a justified inference. Of course, in both cases, part of the belief in question is not creditable; namely, that part that is riddled with accidentality. But the thesis under consideration is whether “[p]artial credit might hence suffice for aptness, and so for animal knowledge, without risk of Gettier refutation” (Sosa 2007, p. 97). And this conclusion is precisely what is called into question by TWO SHEEP and TWO HIRES.

It is worth pointing out that the conclusion here defended can be applied to a response offered by Pritchard on behalf of the Credit View of Knowledge. According to Pritchard, while proponents of this conception of knowledge are right about intervening luck—which he claims is incompatible with achievement—they are wrong about environmental luck—which is compatible with achievement.⁹ According to Pritchard, then, if one is in a traditional Gettier case, one is denied credit for successfully acquiring a true belief and this, it is urged, goes some way toward explaining the value that knowledge has over accidentally true belief. But, as we have seen above, there can be traditional Gettier cases, such as TWO SHEEP and TWO HIRES, that involve true beliefs properly regarded as achievements that are nonetheless riddled with intervening luck. Pritchard’s strategy, then, fails to vindicate even the spirit of the Credit View of Knowledge.

But, one might argue, there are two different senses in which one might be said to deserve partial credit for the truth of a given belief. On the one hand, this may be understood in terms of part of the correctness of a given belief being fully attributable to a competence seated in the subject. On the other hand, this may be understood in terms of the full correctness of a given belief being partially attributable to a competence seated in the subject. For instance, it may be argued that in TWO SHEEP, part of the correctness of Nolan’s true belief is fully attributable to a competence seated in him—namely, the part grounded in the veridical experience of a real sheep. In contrast, the model of the football analogy employed by Sosa to explain testimonial knowledge may be understood as suggesting the latter sense, in which the full correctness of a testifiee’s belief is partially attributable to a competence seated in her and also partially attributable to the individual testifier or to a broader social competence. Given this, it may be argued that TWO SHEEP and TWO HIRES fail to provide relevant counterexamples to Sosa’s view since the notion of partial credit at work in these cases is not the same notion at work in his view.

By way of response to this objection, there are two central points I should like to make. First, it is questionable whether the distinction between these two notions of partial credit can be rendered clear enough to underwrite this defense of the Credit View. For instance, let us take an ordinary case of testimonial knowledge: I come to believe that the bird on the tree is a Harris’s hawk, in part because I am able to identify it as a hawk—rather than, say, an eagle or a falcon—and in part because your expert ornithological testimony enables me to specifically classify it as a Harris’s hawk. There is a perfectly reasonable sense in which the truth of my Harris’s hawk belief here seems to involve partial correctness that is fully attributable to me. The generic hawk part of my belief is fully attributable to me, and the Harris’s part of my belief is fully attributable to your expert testimony. Yet, according to the objection under consideration, this notion of partial credit is not relevant to the notion at work in the Credit View of Knowledge, and thus the proponent of this view will be forced to maintain that all cases of this sort exemplify full correctness that is partially attributable. This, I take it, will strike many as ad hoc. Moreover, it is questionable whether sense can be made of full correctness being partially attributable to a subject. Doesn’t the truth of our beliefs nearly always rely in part on non-attributable features of our situation, such as the epistemic suitability of our environments? Given this, is the full correctness of our beliefs ever partially attributable to us?

Second, and more importantly, even if a tenable distinction could be made between partial correctness being fully attributable and full correctness being partially attributable, there are still problems for the Credit View of Knowledge involving instances of the latter kind of partial credit. To see this, consider the following:

TWO JOKES: Isabella has been working on a complicated and lengthy mathematical proof in her college dorm room for several days. After she stopped working on the first day, her roommate, Catherine—who happens to be a bit of a jokester—tiptoed over to Isabella’s desk and removed a negation sign from one of the steps of the proof. After Isabella finished work on the second day, Catherine again crept over to her desk, but this time she added a negation sign to one of the steps of the proof. It just so happens that Catherine’s two jokes—the removal of a negation sign and the addition of a negation sign—offset one another, and thus enabled Isabella, who was completely unaware that her work had been tampered with, to correctly proceed with her proof and ultimately end up with a true conclusion at the end of her third day of work.

There are several features of this case that are worth noting. First, TWO JOKES is clearly a traditional, rather than an extended, Gettier case. The good luck involved in Catherine’s two jokes offsetting one another intervenes between Isabella’s ability and her success, but in a way that preserves the truth of her belief. This stands in contrast to the luck involved in extended Gettier cases, where the environment is ill-suited for success, but the subject in question just so happens to arrive at a true belief. Second, Isabella clearly deserves partial credit for the true mathematical belief that she comes to accept as a result of completing her proof. For arriving at the true belief in question involved successfully working through numerous steps of a complicated and lengthy proof over the course of 3 days. Indeed, without all of Isabella’s meticulous and competent work, she never would have arrived at the true mathematical conclusion that she ends up holding. Given this, the correctness of her belief is surely at least partially attributable to an individual competence seated in her. Third, the sense in which Isabella deserves partial credit for the true belief in question does not involve partial correctness being fully attributable, at least not in the sense found in TWO SHEEP and TWO HIRES. In these latter two cases, the subject at issue holds a complex belief whose justification is in part grounded in a veridical experience that results from the exercise of a cognitive excellence and in part the result of good luck. The composite nature of this belief may be what underwrites the intuition that such a case involves partial correctness being fully attributable to the subject. In TWO JOKES, however, Isabella’s mathematical belief does not in any sense have such a nature. She holds a simple belief that owes it correctness to various sources, including her own competence, Catherine’s two offsetting jokes, and good luck. Thus, to the extent that a distinction can be clearly made between partial correctness being fully attributable and full correctness being partially attributable, TWO JOKES is an instance of the latter, thereby ruling out a defense of the Credit View that relies on rejecting the former notion of partial credit.

Of course, the proponent of the Credit View of Knowledge may argue that partial achievement is not good enough for creditworthiness; full or complete achievement is necessary for properly deserving credit for success. But then the question arises: how does a testifiee, whose belief is true almost entirely because of the competence of the testifier, deserve credit for the truth of the belief that she acquires via testimony? And this brings us right back to where we started.