KNOWLEDGE AND ASSERTION: A CRITIQUE OF LACKEY Download PDF

Joshua Anderson (Virginia State University)

© 2020 Joshua Anderson

Correspondence: joshuawanderson81@gmail.com

Orcid: https://orcid.org/0000-0002-3621-9771

Original scientific article – Received: 05/01/2020 Accepted: 7/5/2020

ABSTRACT

In the literature on assertion, there is a common assumption that having the knowledge that p is a sufficient condition for having the epistemic right to assert that p—call this the Knowledge is Sufficient for Assertion Principle, or KSA. Jennifer Lackey has challenged KSA based on several counterexamples that all, roughly, involve isolated secondhand knowledge. In this article, I argue that Lackey’s counterexamples fail to be convincing because her intuition that the agent in her counterexamples both has knowledge and do not have the epistemic right to assert is wrong. The article will progress as follows: In section 2, I present Lackey’s argument. In section 3, I suggest some more general reasons for doubting that the agent in her counterexamples actually has knowledge. I then show that from a virtue theoretic and Edward Craig’s practical explication of knowledge perspectives the agent in Lackey’s counterexamples does not know. Since the agent in Lackey’s counterexamples does not have knowledge, she has failed to convincingly prove that KSA is false. In section 4, I conclude by suggesting that, at most, what Lackey’s counterexamples demonstrate is a problem with a simplistic evidentialist and/or process reliabilist epistemology.

Keywords: Assertion; Jennifer Lackey; secondhand knowledge; virtue epistemology

1. Introduction

In the literature on assertion, there is a common assumption that having the knowledge that p is a sufficient condition for having the epistemic right to assert that p—call this the Knowledge is Sufficient for Assertion Principle, or KSA.^[1] Recently, Jennifer Lackey has challenged KSA based on several counterexamples that all, roughly, involve isolated secondhand knowledge. In this article, I argue that Lackey’s counterexamples fail to be convincing because her intuition that the agent in her counterexamples both has knowledge and does not have the epistemic right to assert is wrong.

Lackey is correct that the agent in her counterexamples does not have the epistemic right to assert, but part of the reason that they do not have that right is because the agent does not, in fact, have knowledge. In other words, the reason that Lackey’s agent does not have the epistemic right to assert is in part because KSA, or something like it, holds. However, it is not just that I do not share Lackey’s intuitions—which I do not. It will be shown that for a variety of reasons, and according to different epistemological theories, the agent does not have knowledge.

This article will progress in the following way. In section 2, I present Lackey’s argument. In section 3, I suggest some more general reasons for doubting that the agent in her counterexamples actually has knowledge. I then show that from a virtue theoretic and Edward Craig’s practical explication of knowledge perspectives the agent in Lackey’s counterexamples does not know. Since the agent in Lackey’s counterexamples does not have knowledge, she has failed to convincingly prove that KSA is false. In section 4, I conclude by suggesting that, at most, what Lackey’s counterexamples demonstrate is a problem with a simplistic evidentialist and/or process reliabilist epistemology.

2. Lackey’s Argument

According to what has been identified in this article as KSA: If a subject S knows that p, then S has the epistemic right to assert that p (Lackey 2011, 252).^[2] To be clear, KSA in its most general form simply maintains that knowledge is sufficient for the epistemic right to assert. KSA does not require one to assert, and there may be additional norms governing assertion. For example, a psychiatrist, over the course of many therapeutic sessions, comes to know that her client, Tom, is an alcoholic; the norms governing doctor-patient privilege prevent the psychiatrist from asserting what the psychiatrist knows to certain people. However, the psychiatrist would not be epistemically blameworthy for asserting that Tom is an alcoholic.

Even though Lackey admits that there is some intuitive plausibility and, potentially, some theoretical power behind KSA, she argues that KSA is actually false. She suggests “that there are various kinds of cases in which a speaker asserts that p, clearly knows that p, and yet does not have the proper epistemic authority or credentials [i.e. the epistemic right] to make such an assertion, thereby showing that knowledge is not always sufficient for epistemically proper assertion” (Lackey 2011, 253). Lackey makes her case through a series of counterexamples that all involve isolated secondhand knowledge.^[3] For simplicity, I will focus, predominately, on one of Lackey’s counterexamples that she calls DOCTOR.

DOCTOR: Matilda is an oncologist at a teaching hospital who has been diagnosing and treating various kinds of cancers for the past fifteen years. One of her patients, Derek, was recently referred to her office because he has been experiencing intense abdominal pain for a couple of weeks. After requesting an ultrasound and MRI, the results of the tests arrived on Matilda’s day off, consequently, all of the relevant data were reviewed by Nancy, a competent medical student in oncology training at her hospital. Being able to confer for only a very brief period of time prior to Derek’s appointment today, Nancy communicated to Matilda simply that her diagnosis is pancreatic cancer, without offering any of the details of the test results or the reasons underlying her conclusion. Shortly thereafter, Matilda had her appointment with Derek, where she truly asserts to him purely on the basis of Nancy’s reliable testimony, “I am very sorry to tell you this, but you have pancreatic cancer”. (Lackey 2011, 253)

The first thing Lackey draws one’s attention to is that in DOCTOR Matilda’s knowledge is isolated and secondhand. It is isolated because the only thing that Matilda “knows” is simply the fact that Derek has cancer—Matilda’s general background knowledge regarding cancer, generally, and the little information she has from her previous meeting with Derek is not significant enough to un-isolate her knowledge. Matilda’s knowledge is secondhand because the only reason she “knows” is based exclusively on another person’s—viz. Nancy’s—testimony. There is nothing inherently wrong with isolated secondhand knowledge;^[4] it is just that in DOCTOR it leads to problems.

Lackey maintains “that Matilda clearly knows that Derek has pancreatic cancer—it is true, she believes it, she has good reason to trust the testimony of her medical student, and Nancy is in fact a reliable source” (Lackey 2011, 254). Further, according to Lackey, it is clear that Matilda does not have the epistemic right to simply assert to Derek that he has pancreatic cancer. There are several reasons that Lackey gives to justify her claim that Matilda lacks the epistemic right to assert. Matilda is an expert oncologist and this expertise carries with it certain epistemic duties and responsibilities to fulfill before asserting a diagnosis. These duties and “responsibilities may include having reviewed the test results firsthand, possessing reasons for choosing one condition over another, knowing details about the size and nature of the cancer, and so on” (Lackey 2011, 254). Further, as an expert it is expected that Matilda should be able to justify and explain her diagnoses, in general, and Derek’s cancer diagnosis, in particular. Such justification and explanation is impossible for Matilda given the isolated secondhand knowledge involved.

Based on the considerations regarding DOCTOR, Lackey concludes DOCTOR is “a case where a speaker knows that p without thereby being epistemically positioned to assert that p, thereby falsifying [what Lackey refers to as] KNA-S*” (Lackey 2011, 258). In the terminology being used in this article: DOCTOR demonstrates that one can know that p, but lack the epistemic right to assert that p, thereby falsifying KSA. More explicitly, Lackey’s argument amounts to this:

If KSA is true, then if one knows that p, then one has the epistemic right to assert that p.
In DOCTOR: One knows that p, but does not have the epistemic right to assert that p.
Therefore, KSA is not true. (by 1, 2 and Modus Tollens)

3. Problems with Lackey’s Argument

Lackey’s argument is simple and straightforward. If she is right, then she has identified a huge problem with much of the literature on assertion.^[5] In this section of the article I argue that premise 2 of Lackey’s argument does not hold, and thus she fails to demonstrate that KSA is false. I will show that in DOCTOR^[6] it is not the case that an agent both knows and lacks the epistemic right to assert.

Although Lackey claims that it is “clear” that Matilda knows Derek has cancer, it certainly does not seem clear that Matilda knows. Further, although Lackey gives some reasons for thinking that Matilda does know Derek has cancer—viz. “it is true, she believes it, she has good reason to trust the testimony of her medical student, and Nancy is in fact a reliable source”—she is really just trading in intuitions (Lackey 2011, 254). It is obvious that she is merely appealing to intuition when she considers some modified versions of her counterexample and states that one can “compar[e] the intuitions elicited from such modified cases with those from the original” (Lackey 2011, 256).

Considering DOCTOR from the perspective of a variety of epistemological theories—e.g., Craig’s practical explication of knowledge, virtue epistemology, inferentialism—it will be shown that Matilda does not actually know Derek has cancer. If Matilda does not know that Derek has cancer, then Lackey’s counterexample fails, and thereby her argument that KSA is false does not go through. Moreover, though it was not gone into above, Lackey gives reasons to justify here claim that Matilda lacks the epistemic right to assert. What is interesting is her reasons for claiming that Matilda lacks the epistemic right to assert—such as the increased epistemic duties tied to Matilda’s status as an expert oncologist and her inability to provide a justification for the diagnosis—are actually reasons to think Matilda does not know. Thus, Lackey is implicitly appealing, in part, to KSA, or something like it, in here counterexample.

The advantage of demonstrating that Matilda fails to have knowledge for a variety of reasons is that it will block a couple of responses that Lackey could make. Such responses would amount to something like this: On the one hand, “I disagree that your general reasons hold”, and on the other hand “Well sure according to theory X Matilda does not have knowledge, but I have independent reasons for thinking theory X is false, therefore it has not been shown that Matilda does not have knowledge, really”. I will be considering DOCTOR in a variety of ways. First, I will give some more general reasons for doubting that Matilda has knowledge. Then I will show that from the perspectives of virtue theoretic account of knowledge and Craig’s practical explication of knowledge Matilda does not know that Derek has cancer. Even if Lackey has independent reasons for rejecting all the considerations put forward, it seems hard to believe that if all of these reasons and theories point to the fact that Matilda lacks knowledge in DOCTOR that Matilda would have knowledge. At the very least the onus would be on Lackey to do more work to prove that Matilda does have knowledge—i.e. more than the few off hand reasons she does give, and her intuitions.

3.1 Why Lackey’s Agent Fails to Have Knowledge

Before turning to Craig and virtue epistemology, in this sub-section I will suggest some reasons for thinking that the agent in Lackey’s counterexample does not actually have the purported knowledge that Lackey thinks the agent does. Importantly, part of my reasoning for claiming that the agent does not know is similar to Lackey’s justification for claiming that the agent does not have the epistemic right to assert.

So first, in DOCTOR, consider the content of what is to be known—viz. that Derek has pancreatic cancer—especially since Matilda is an expert oncologist it is not a straightforward proposition like “grass is green”, or “a square has four sides”. Like cancer itself, the content of the belief is complex and multifaceted,^[7] in fact that is why multiple tests are run before a diagnosis can be made. Saying that an oncologist knows someone has cancer is not to say that the oncologist knows the truth value of a simple proposition. Thus, at least prima facie, Matilda does not actually know that Derek has cancer. The reason she does not know is because all that the reliable testimony of Nancy conveys to Matilda is, at best, the vague understanding that there is something like pancreatic cancer present in the patient Derek, and, at worst, simply the truth-value of a statement.

Cancers are unique. They come in stages, involve different types of tumors, can be more or less deadly, sometimes they run the danger of spreading, and sometimes not. In order to really know, Matilda needs to know what Derek’s cancer is. Nancy’s testimony does not speak to that. Therefore, based simply on Nancy’s testimony Matilda simply cannot know Derek has cancer because she does not know what Derek’s cancer is. Lackey, inadvertently, draws attention to the issue under consideration here when, in justifying why Matilda lacks the epistemic right to assert, she states that “Matilda should be able to (at least partially) explain or justify the diagnosis of pancreatic cancer that she is offering to her patient” (Lackey 2011, 254). One way to understand the explication or justification of the diagnosis is to see it as explaining what the diagnosis is. In other words, to explain a diagnosis is to fill out the content of the proposition “Derek has pancreatic cancer”.

Another way to think about what Lackey is saying when she is denying that Matilda lacks the epistemic right to assert is that she is, implicitly, appealing to a principle that is even stronger than KSA. The reason that Matilda lacks the epistemic right to assert is because she does not know what she is asserting. Matilda cannot explain or justify the assertion that Derek has pancreatic cancer because she does not know that he has cancer. That is to say, one has the epistemic right to assert only if one knows—i.e. knowing that p is necessary in order to have the epistemic right to assert that p. The fact that Matilda cannot explain or justify the diagnosis underscores the fact that she does not know what she is purportedly diagnosing, which, in turn, explains the impropriety of asserting the diagnosis. If that is right, then KSA has not been refuted, because Matilda does not know.^[8]

A further issue is that it seems that the evidence needed to ground knowing that Derek has cancer needs to be fairly significant, especially considering Matilda is an expert oncologist. Again, Lackey draws attention to this increased demand for evidence. Being an oncologist carries with it “certain epistemic duties. […] [T]hese responsibilities may include having reviewed the test results firsthand, possessing reasons for choosing one condition over another, knowing details about the size and nature of the cancer, and so on” (Lackey 2011, 254). Lackey believes that the aforementioned epistemic duties relate only to assertion. However, they seem much more like evidential requirements on knowledge, on the one hand—reviewing test results, possessing reasons. On the other hand, “knowing details about the size and nature of the cancer” speaks to not only the evidential requirements but also the previous issue about the complexity of what is involved in a doctor knowing that someone has cancer (Lackey 2011, 254).

While the above considerations are not definitive, they are at least some prima facie reasons for doubting that the agent in Lackey’s counterexamples, in general, and Matilda in DOCTOR, in particular, does actually have the knowledge required for her argument to work. I now turn to another consideration that suggests that it is far from clear that Matilda knows that Derek has cancer. What was argued above basically amounts to the fact that doctors need stronger justification than the word of a student—no matter how reliable it might be—to make a diagnosis. Thus, isolated secondhand testimony is insufficient for Matilda to know that Derek has cancer. More obviously, however, is the fact that Matilda is unable to make the same appropriate inferences from the reliably produced testimony of Nancy, which she would be if she actually knew that Derek had cancer—from looking at test results and so forth.

Imagine that when Matilda tells Derek he has cancer and he starts asking her questions: “Is his case terminal, is he going to need surgery or chemotherapy or both, how large is the tumor, is there a danger of the cancer spreading?” Based on Nancy’s testimony, Matilda cannot answer any of these questions. The reason she is unable to answer the questions is because she really does not know that Derek has cancer. In fact, the most natural response to Derek’s questions would be for Matilda to say “I don’t know”, and the reason she does not know the answers to his questions is because she does not really know he has cancer. Matilda may have a true belief based on Nancy’s testimony, but one of the things that makes knowledge more valuable than mere true belief is the ability to make appropriate inferences.

Lackey implicitly relies on these inferential considerations in a similar example in order to argue that someone similarly situated to Matilda lacks the epistemic right to assert. Lackey states:

“Suppose, for instance, that he asks [Matilda] what exactly the ultrasound and MRI revealed, or how large his tumor is, or why she thinks it is pancreatic cancer. All she can say […] is that she had been told that he has pancreatic cancer […] that she hadn’t actually seen any of the test results herself, and that she has no additional information to offer about his particular diagnosis. Wouldn’t [Derek] be entitled to resent [Matilda] under such circumstances, to feel that he has been epistemically cheated by his doctor who owes him more than a diagnosis grounded purely in isolated secondhand knowledge?” (Lackey 2013, 38)

What the above quote suggests is a few things; first it underscores what was said above regarding the justification, and evidence, needed for a doctor to know someone has cancer. Second, it suggests that part of Derek’s resentment involves Matilda’s inability to make appropriate inferences based on Nancy’s testimony. Finally, the reason that Derek is entitled to feel “epistemically cheated” is not merely because Matilda lacks the epistemic right to assert, but, more importantly, the information that Matilda is conveying to Derek does not meet the standards for knowledge. Derek is being cheated because Matilda is trying to pass off true belief as knowledge—i.e. Matilda is giving Derek something less epistemically valuable (a true belief) when what he deserves is something more epistemically valuable (knowledge).

Now one could possibly object by claiming something along the lines that DOCTOR is a peculiar case because of the context or stakes involved raises the demands for knowledge. However, even in low-stakes scenarios Matilda still does not know that Derek has cancer, based solely on Nancy’s testimony. Later in her paper, in responding to possible objections to her thesis Lackey gives the following modified DOCTOR case. “Suppose, for instance, that instead of flat out asserting to Derek that he has pancreatic cancer in DOCTOR, she [Matilda] casually asserts this fact to her husband over dinner” (Lackey 2011, 272).^{^[9]}

Lackey thinks that in the modified DOCTOR case that Matilda actually has knowledge and the epistemic right to assert—at least intuitively. If what I have been arguing thus far is right, though, even when talking to her husband, Matilda still does not know. Imagine that when Matilda mentions Derek’s cancer to her husband, he says to her “Is there anything you’ll be able to do for him?” Matilda will be forced to answer “I don’t know”, because even in that context Matilda does not really know that Derek has cancer. Even if the low-stakes lowers the evidential requirements, the fact that she is unable to draw appropriate inferences is enough to justify the claim that Matilda does not know even when talking to her husband.

3.2 Edward Craig and Virtue Epistemology

Having given some more general reasons for doubting that the agent in Lackey’s counterexamples does, in fact, have knowledge, I now turn to some more theoretical considerations. It seems that a virtue theoretic account of knowledge and Edward Craig’s practical explication of knowledge give one reason to think that the agent in Lackey’s counterexamples does not know what is being asserted.^{^[10]} Again, for simplicity, I will focus on DOCTOR. However, the same reasoning can be applied to her other cases.

3.2.1 Virtue Epistemology

According to John Greco on a virtue theoretic account of knowledge

S knows that p if and only if

p is true;
S believes that p; and
S believes the truth because S’s belief is produced by intellectual ability. (Greco 2010, 12)

It seems that Matilda, in DOCTOR, does not satisfy all three of these conditions. Even if 1 and 2 are granted, it does not seem that 3 is satisfied.^{^[11]} In a sense, Matilda’s intellectual abilities are involved—e.g. her ability to hear. However, on a virtue theoretic account it is not simply that any intellectual ability needs to be involved in order to attain knowledge, but the relevant intellectual ability needs to be involved in the right way.

In order to clarify 3, and why Matilda has not satisfied it, consider that 3 really amounts to something more like this: “when we attribute knowledge to someone we mean to give the person credit for getting things right. Put another way, we imply that the person is responsible for getting things right” (Greco 2003, 111; emphasis added). In other words, in order to be attributed knowledge one has to have done something creditworthy, or one has to have done something that makes one responsible for having the true belief. What creditworthiness, and responsibility, amount to can of course vary. However, it does not seem that Matilda has done anything creditworthy, nor is she really responsible for having a true belief regarding Derek’s cancer.^{^[12]}

If anyone deserves credit, it would be Nancy. Nancy used her excellent visual, reasoning and other abilities in deciding that Derek has pancreatic cancer. Those are the type of abilities relevant for a cancer diagnosis. Matilda used no abilities like that; she simply believed what Nancy told her. Thus, Matilda is neither creditworthy nor responsible for the true belief, and therefore does not have knowledge.

To underscore the point that Matilda is neither creditworthy nor responsible reduce Nancy’s reliability for cancer diagnosis. Even if it is intuitively plausible that Matilda could have knowledge based on Nancy’s highly reliable testimony, as one reduces Nancy’s reliability there will come a point where Matilda does not have knowledge. Further, other than a naïve evidentialist, that point will probably far exceed a .5 probability.^{^[13]} Here is the idea. Nancy’s testimony is too unstable to transfer credit or responsibility to Matilda. If Nancy’s testimony is too reliable, the credit that accrues to Nancy will swamp any credit that could be assigned to Matilda. If Nancy’s testimony is too unreliable, then a true belief is not transferred at all.

More importantly, what the variability of Nancy’s reliability illustrates, is precisely what the relevant abilities are, and how a doctor can be responsible for knowing someone has cancer. When Nancy is highly reliable it is because of her education, specialized visual abilities needed for reviewing test results and excellent reasoning abilities, for example. When Nancy is less reliable, it is because those same abilities are not as good. Further, when Nancy is less reliable one does not give credit to Matilda because Matilda did not exercise the abilities—education, visual and reasoning abilities—and in fact would be blamed for not using them. By the same reasoning then, even when Nancy is highly reliable, Matilda does not deserve credit nor is she responsible precisely because Matilda did not exercise the relevant abilities.

Another way to think about the issue under consideration here—that Matilda is not creditable—is to note that Matilda’s abilities are not a salient part of the causal story leading to the true belief. Greco discusses two ways that something can be a salient part of the causal story. Only one of which is relevant here. According to Greco, a “major factor governing salience is our interests and purposes” (Greco 2003, 118). The interests and purposes involved in a cancer diagnosis, predominately, all relate to knowing what to do with the diagnosis. Based on what was said above regarding Matilda’s inability to make appropriate inferences, and Derek feeling epistemically cheated, it seems clear that Matilda’s abilities—e.g. hearing—do not serve the relevant interests and purposes.

Alternatively, and relatedly, one can think of the creditability along more Sosaian lines and see that Matilda is not creditable because she has not manifested her abilities qua oncologist (Sosa 2007). Remember that Matilda is an expert oncologist giving a cancer diagnosis. Therefore, even if Matilda has a true belief her coming to hold that true belief does not manifest her abilities. It does not manifest her abilities because, on the one hand, the nature of what is to be known, demands certain abilities—like the ones mentioned above—and, on the other hand, it does not manifest her abilities as an expert oncologist. Thus, again, Matilda is not creditable for the true belief, and therefore does not have knowledge.

Finally, consider Linda Zagzebski’s virtue epistemology. Unlike Greco and Sosa who understand epistemic virtues more along the lines of a type of reliabilism—i.e. cognitive abilities like reasoning, and perception—Zagzebski understands virtues more along the lines of traditional virtue ethics, and she refers to them as “intellectual virtues”. Intellectual virtues are motives or dispositions like intellectual courage—standing by one’s beliefs—intellectual humility—being open to contrary evidence—and so forth. For Zagzebski, “[k]nowledge is belief arising out of acts of intellectual virtue” (Zagzebski 1999, 109). In DOCTOR, Matilda, by basing her diagnosis on Nancy’s testimony, was exercising a kind of intellectual vice—something like intellectual laziness. Thus, Matilda fails to know that Derek has pancreatic cancer if knowledge is understood along Zagzebski’s line because her belief did not arise out of an act of intellectual virtue. In fact, Matilda’s belief arose out of an act of intellectual vice, which explains, in part, why Derek would feel epistemically cheated.

3.2.2 Edward Craig and the Practical Explication of Knowledge

Having discussed why Matilda fails to have knowledge according to a virtue theoretic account of knowledge, I now turn to Edward Craig’s practical explication of knowledge. According to Craig, “the concept of knowledge is used to flag approved informants” (Craig 1986-7, 215). While that is Craig’s, so to say, definition of knowledge, what is helpful for present purposes is “that the principal candidates [for] the analysis of the everyday concept of knowledge all lie very close to the concept constructed by adopting the point of view of the inquirer” (Craig 1986-7, 225).

By adopting the point of view of the inquirer, if one wants to know whether Matilda knows Derek has pancreatic cancer, then one needs to adopt Derek’s perspective. Fortunately, Lackey has explained what Derek’s perspective is, or would be.

“Derek reasonably has the right to expect his doctor to fulfill such a duty [viz. reviewing test results firsthand, possessing reasons for choosing one condition over another, knowing details about the size and nature of the cancer]. […] Wouldn’t Derek be entitled to resent Matilda under such circumstances [i.e. not fulfilling here epistemic duties], to feel that he has been epistemically cheated by his doctor who owes him more than a diagnosis grounded purely in isolated secondhand knowledge.” (Lackey 2011, 6-7)

The issues in the above quote have been discussed already, but Craig gives them some more theoretical substance.

If one, again, imagines that Derek begins asking Matilda questions about the diagnosis, and Matilda is unable to respond, at first Derek would probably be a bit confused. Matilda then “reveals to Derek that she had been told that he has pancreatic cancer from her student Nancy, that she hadn’t actually seen any of the test results herself, and that she has no additional information to offer about his particular diagnosis” (Lackey 2011, 6). It seems the most natural response that Derek might give—beyond resenting Matilda and feeling epistemically cheated—would be something like: “So, you really don’t know I have cancer. Your student believes I may have cancer. That’s what you’re telling me”.

What Craig’s practical explication of knowledge gives is yet another reason to doubt that Matilda both has knowledge and lacks the epistemic right to assert. By adopting the perspective of the inquirer, namely Derek, it becomes abundantly clear that what is epistemically problematic in DOCTOR is the fact that Matilda is asserting something she does not know, as if she does know it. It is not, as Lackey maintains, merely that Matilda lacks the epistemic right to assert—which she does—but the reason why she lacks the epistemic right to assert is, in part, that KSA, or something like it, holds and Matilda lacks knowledge. At a minimum, the way Lackey explains the fact Matilda lacks the epistemic right to assert seems much more like she is explaining why Matilda does not know, but if that is right then she clearly has not shown that KSA is false.

A further implication of Craig’s view is that knowledge flags good informants for actionable information.^{^[14]} Thus, part of the problem in DOCTOR is that merely asserting to Derek that he has pancreatic cancer is not an actionable piece of information. He does not know if he needs more tests, or needs to start taking medicine, or change his diet, or prepare for surgery. What would make the information actionable is answers to the types of questions he would ask that Matilda cannot answer. Therefore, yet again, in order to know Matilda would have to have looked at the test results and so forth, in order to make the information she was asserting actionable for Derek, and thereby knowledge on her part.

All of the considerations put forth in section 3 may not, individually, tell against Lackey’s counterexamples.^{^[15]} However, by taking everything together it seems clear that the assertor in Lackey’s counterexamples does not actually have knowledge, and therefore does not disprove KSA.^{^[16]} Lackey is correct that the assertor in her counterexamples lacks the epistemic right to assert, but she is wrong about why they lack that right. It is unclear why Lackey would think they do not have the epistemic right to assert, but the real reason is, in part, because of KSA, or something like it, and the fact that the assertor does not know what they are asserting. Still, even if Lackey is not relying implicitly on KSA it does not seem that she has shown that KSA is false.

3.3 A Fall Back Position

All the reasons discussed above seem to point to the fact that the agent in Lackey’s counterexamples does not have knowledge, demonstrating that Lackey has failed to disprove KSA. However, some still might not think that anything that has been said actually proves that the agent lacks knowledge. In this sub-section of the article, I present a, kind of, fallback position. The idea is that even if I have not established that the agent does not know, the reasoning above makes it plausible that the agent might not know. If the agent might not know, that is sufficient to undermine the force of Lackey’s counterexamples, which, in turn, gives one good reason to think that she has not really established that KSA is false.

Lackey has made a very bold claim. The problem is that bold claims need to be well established. Lackey is not unaware of the fact that she is endorsing a rather strong position. In order to falsify KSA Lackey needs to, in her words, establish “that there are various kinds of cases in which a speaker asserts that p, clearly knows that p, and yet does not have the proper epistemic authority or credentials to make such an assertion” (Lackey 2011, 253). The crux is that the speaker, or agent, asserting “clearly knows that p”. It could be argued that the above considerations do not establish that the speaker, or agent, in her counterexample clearly does not know that p. However, they certainly show that the speaker, or agent, does not clearly know that p. Yet, clearly knowing that p is exactly what Lackey needs to draw the conclusion that she does. Therefore, the onus is on Lackey to do much more work to establish that it is clear that the agent has knowledge in the cases she describes. Until she has done that, there is little motivation to accept her conclusion.

Another fallback position is to point out that Lackey’s counterexamples, especially DOCTOR, are too unrealistic to really have enough force to undermine a general norm of assertion, like KSA. At least intuitively, it does not seem reasonable that an expert oncologist, like Matilda, would base a cancer diagnosis solely on the testimony of a student, no matter how reliable the student might be. At least intuitively, it does not seem that Matilda would take herself to know that Derek has cancer, or at least know in the right way necessary for assertion, based simply on the word of her student.

If Nancy were not a student, but rather another competent oncologist, perhaps Matilda would accept that as sufficient for knowing that Derek has cancer. However, then the intuition that Matilda lacks the epistemic right to assert is not so clear. On the other hand, if Nancy were another expert oncologist it does not seem that the scenario of DOCTOR, mutatis mutandis, would play out exactly the way Lackey describes. It seems reasonable to think that instead of flat out asserting to Derek that he has cancer, Matilda might add the preamble that her colleague had looked at Derek’s test results and concluded that Derek had cancer, or she might bring Nancy in to meet with Derek as well. But by adding the preamble or bringing Nancy with her to the meeting, which seems like at least something that Matilda might do, points to the fact that Matilda might not take herself to actually know that Derek has cancer, and hence might just be more evidence that it is not clear that Matilda knows.^{^[17]}

Many of Lackey’s other counterexamples seem a bit unrealistic for the same type of reasons. Consider two others, PROFESSOR and FOOD.

“PROFESSOR: Judith is a professor at one of the best law schools in the country, and today’s lecture is on U.S. copyright law. While she is generally quite knowledgeable of this topic, she has failed to keep up with some recent developments in this area. Over lunch yesterday, one of her colleagues briefly expressed his belief that it is extremely improbable that the Supreme Court will consider a case challenging the addition of 20 years to the original copyright protection of 50 years after the death of authors. Though Judith does not know any of the reasons or considerations underlying this claim, she asserts to her students in class, “The Supreme Court is unlikely to hear the upcoming challenge to the recent extension of U.S. copyright protections to 70 years after the author’s death”. While this assertion is in fact true, it is based purely on the basis of the reliable testimony of Judith’s colleague.” (Lackey 2011, 254)

“FOOD: My neighbor Ken is a connoisseur of fine dining. As we were leaving Starbucks this afternoon, he told me that the food at a new local restaurant about which I was previously quite unfamiliar, Quince, is exquisite, though being in a hurry prevented him from offering any details or evidence on behalf of this claim. While talking to my friend Vivienne later in the day, she was fretting over where to take her boyfriend to dinner for Valentine’s Day. I promptly relieved her stress by truly asserting, “The food at Quince is exquisite.”” (Lackey 2011, 257)

It seems that it is more realistic for the speakers in each of the cases to preface their assertion with something like “I have been told that […]” or “According to […]”, than for the speakers to just flat out assert what Lackey has them assert. If it is likely that the speakers would preface their assertion, then it seems reasonable to think that the speaker does not take themselves as clearly knowing the propositions under consideration.

Here, I wanted to just draw attention to the fact that even if the arguments put forward in 3.1-2, are not as convincing as I take them to be, they are still sufficient to undermine Lackey’s argument. Lackey needs it to be the case that the agent in her counterexamples clearly knows, but she has not demonstrated that the agent clearly knows. Since the agent does not clearly know her conclusion does not follow; at least it does not follow in the right way to truly undermine KSA. I also drew attention to the fact that Lackey’s counterexamples, as presented, seem highly unrealistic. Making them more realistic seems to make the intuition that the agent in her counterexamples both has knowledge and the epistemic right to assert less clear, which is sufficient to undermine the force of her argument.

4. What Lackey’s Argument Might Show

Despite the fact that Lackey has failed to prove that KSA is false, her paper is instructive. What Lackey’s paper points to is the inadequacies of various epistemological theories. Before concluding, then, I will briefly discuss the positive upshot of Lackey’s paper. It will be suggested that what Lackey’s counterexamples demonstrate are problems with simplistic evidentialist and process reliabilist theories of knowledge.

For evidentialism, knowledge is, roughly, simply justified true belief. Further, “S is justified in believing p at [time] t if and only if S’s evidence for p at t supports believing p” (Mittag 2020).^{^[18]} What counts as “support” varies according to different theories of evidentialism, and can be context sensitive. Further, support basically amounts to something like the evidence makes it highly likely that p is true. Imagine, as Lackey does in responding to some objections, that Nancy and her testimony might be quite truth conducive. Here is the problem then for the evidentialist if Nancy’s testimony is extremely truth conducive, then Matilda is going to get knowledge. Yet, if what has been said above is right, then Matilda does not have knowledge. Thus, Lackey has not shown that KSA is false, but that a straightforward evidentialism does not do justice to how one understands whether or not someone can know—at least not in all cases.

The same type of argument can be made against a simplistic form of process reliabilism. According to process reliabilism, “S knows that p if and only if S believes that p, p is true, and S’s belief that p is formed by a reliable process” (Becker 2020).^{^[19]} What counts as a reliable process can vary, but is, roughly, a process that is highly truth conducive. Therefore, if Nancy and her testimony are highly reliable, then Matilda will get knowledge based on Nancy’s testimony, but that is the wrong result. The problem is thus with process reliabilism, not KSA.

What both evidentialism and process reliabilism have in common is their focus on truth-conduciveness, and an implicit acceptance of knowledge being about simple propositions. Because of this—i.e. the focus on truth conduciveness and acceptance of knowledge being about simple propositions—evidentialism and reliabilism, at least simplistic forms of them, are not fine grained enough to notice that things like cancer diagnoses are more complex than a simple true or false proposition. The upshot is that something like a virtue theoretic account, whether a virtue reliabilism like Greco and Sosa or a virtue responsibilism like Zagzebski, has more theoretical power, and thus, is plausibly a superior epistemological theory. That, then, is what the real lesson of Lackey’s paper is.

Now, if one did not find the arguments put forth in 3.1.-2. to convincingly show that the agent in Lackey’s counterexamples lacked knowledge, then what has been said in this section might not be convincing either. On the other hand, what has been said in this section can still be instructive. At most what Lackey has shown with her counterexamples is that KSA and simplistic evidentialist, or process reliabilist, epistemology, are at least prima facie incompatible, or more strongly KSA is perhaps false only if one holds a simplistic evidentialist or process reliabilist epistemology. That in itself—KSA being perhaps false, conditionally—is telling against Lackey. Remember, Lackey needs it to be the case that the agent in her counterexample clearly does not know, but if her argument only works on certain assumptions about what counts as knowledge, then she has failed to make the case that KSA is actually false.

5. Conclusion

In this article, I have argued that Jennifer Lackey has failed to prove that the KSA is false. According to the KSA, if a subject S knows that p, then S has the epistemic right to assert that p. Lackey’s argument proceeded by putting forth various counterexamples where it seemed, at least intuitively, that a subject both knew that p but lacked the epistemic right to assert that p, thereby falsifying KSA. It was shown that, in fact, the subject under consideration actually did not know. Since the subject lacked knowledge, the subject’s lack of an epistemic right to assert was not proof that KSA is false, but rather, perhaps, that KSA is true. The idea is that if the agent lacks the epistemic right to assert, then the agent fails to know, by Modus Tollens and KSA. However, at a minimum, it seems that Lackey is implicitly working with something that corresponds to KSA, where one can assert that p only if one knows that p. Since the subject in Lackey’s counterexample lacks the epistemic right to assert then the subject does not know; this seems to be the case because the reasons that Lackey gives for claiming the subject in her counterexamples does not have the epistemic right to assert are better understood as reasons for thinking the agent does not know. Notice that, even if Lackey is not relying on KSA, if the agent in the counterexample does not know then she has not falsified KSA.

To avoid the quagmire of intuition bumping, it was demonstrated that from a variety of perspectives, and for a variety of reasons the agent in Lackey’s counterexamples does not know. First, it was shown that, for example, cancer diagnoses are complex and require special evidence. Second, it was shown that part of what distinguishes knowledge from mere true belief is the ability to make appropriate inferences. Then, it was shown that from the perspective of a virtue theoretic account of knowledge and Edward Craig’s practical explication of knowledge the agent in the counterexamples lacks knowledge.

Finally, it was concluded that although Lackey failed to successfully prove that KSA is false, there were some valuable lessons. It was suggested that a simplistic evidentialism and process reliabilism, are theoretically less powerful than a virtue epistemology. Thus, what Lackey truly gives in her paper is a prima facie argument in favor of, perhaps, something like a virtue epistemology.

REFERENCES

Becker, K. 2020. Reliabilism. Internet Encyclopedia of Philosophy. https://www.iep.utm.edu/reliabil/. Accessed January 4, 2020.

Becker, K. 2006. Reliabilism and safety. Metaphilosophy 37: 691-704.

Conee, E., and R. Feldman eds. 2004. Evidentialism: Essays in Epistemology. Oxford: Clarendon Press.

Craig, E. 1986-7. The practical explication of knowledge. Proceedings of the Aristotelian Society 87: 212-226.

DeRose, K. 2002. Assertion, knowledge, and context. The Philosophical Review 111: 167-203.

Goldman, A. 1979. What is justified belief? In Justification and Knowledge, ed. G. Pappas, 1-24. Boston: D. Reidel.

Greco, J. 2003. Knowledge as credit for true belief. In Intellectual Virtue: Perspectives from Ethics and Epistemology, eds. M. DePaul and L. Zagzebski, 111-134. Oxford: Oxford University Press.

Greco, J. 2010. Achieving Knowledge: A Virtue-Theoretic Account of Epistemic Normativity. Cambridge: Cambridge University Press.

Hawthorne, J. 2004. Knowledge and Lotteries. Oxford: Oxford University Press.

Hawthorne, J., and J. Stanley. 2008. Knowledge and action. The Journal of Philosophy 105: 571-590.

Heller, M. 1995. The simple solution to the generality problem. Nous 29: 501-515.

Lackey, J. 2007. Why we don’t deserve credit for everything we know. Synthese 158: 345-361.

Lackey, J. 2011. Assertion and isolated secondhand knowledge. In Assertion: New Philosophical Essays, eds. J. Brown and H. Cappelen, 251-276. Oxford: Oxford University Press.

Lackey, J. 2013. Deficient testimonial knowledge. In Knowledge, Virtue, and Action: Putting Epistemic Virtues to Work, eds. T. Henning and D. P. Schweikard, 30-52. New York: Routledge.

Mittag, D. M. 2020. Evidentialism. Internet Encyclopedia of Philosophy. https://www.iep.utm.edu/reliabil/. Accessed January 4 2020.

Sosa, E. 2007. A Virtue Epistemology: Apt Belief and Reflective Knowledge, Volume I. Oxford: Oxford University Press.

Simion, M. 2016. Assertion: Knowledge is enough. Synthese 193: 3041-3056.

Zagzebski, L. 1999. What is knowledge? In The Blackwell Guide to Epistemology, eds. J. Greco and E. Sosa, 92-116. Malden, MA: Blackwell Publishers.

Endnotes

For example, DeRose (2002), Hawthorne (2004), Simion (2016), amongst others, can all be seen as either advocating or sympathetic to something like KSA. ↑
What I have identified as KSA, is Lackey’s second formulation of the sufficient condition of the Knowledge Norm of Assertion (KNA-S*). “KNA-S*: One is properly epistemically positioned to assert that p if one knows that p. According to the KNA-S*, then, knowledge is sufficient for possessing the epistemic authority for assertion” (Lackey 2011, 252). KSA simply cashes out “possessing the epistemic authority for assertion” as the epistemic right to assert. Similarly, Simion (2016, 3043) refers to it as the sufficiency claim for a Knowledge Norm of Assertion (KNA-Suff) and words it as follows: “One’s assertion that p is epistemically proper if one knows that p”. KSA is just clarifying “epistemically proper”. ↑
What Lackey means by “isolated secondhand knowledge” will become clear as I explicate her view, below. ↑
An example of unproblematic isolated secondhand knowledge is Lackey’s own Chicago visitor case. See Lackey (2007). ↑
See for example, DeRose (2002), and if one takes assertion to be a kind of action, see Hawthorne and Stanley (2008). ↑
Again, I will only be primarily arguing against Lackey’s counterexample DOCTOR. Although she gives several counterexamples that she thinks refute KSA, they are all roughly the same. They all involve what she calls isolated secondhand knowledge and are cases where, purportedly, the agent involved has knowledge but lacks the epistemic right to assert. The reasoning I will use to refute DOCTOR, I think can be extended to all her cases. Thus, for brevity I have only focused on the one counterexample. ↑
This same type of issue holds for Lackey’s other counterexamples as well. For example, when one is speaking to the quality of a restaurant, or a student’s writing ability, these are things that are multifaceted and complex, not simple propositions Lackey (2011). ↑
My thanks to an anonymous reviewer for drawing attention to the fact that I might not have been clear on this point. What I am suggesting is that because KNA-N fails in the case means that Matilda does not have knowledge, which means that DOCTOR is not a case where the agent has knowledge but lacks the right to assert and eo ipso means KSA is not refuted. ↑
Certainly, it would be inappropriate for Matilda to tell her husband that Derek had cancer based on the norms governing doctor-patient privilege. However, as I mentioned above those norms are not of concern for this article—only the epistemic norms governing assertion are of concern here. ↑
This line of argument is not definitive, but it does the work that it is supposed to do. For Lackey’s counterexample to work it needs to be the case that it is obvious that Matilda has knowledge but lacks the right to assert. Presenting multiple theories that would indicate that Matilda does not have knowledge switches the burden to explain why, beyond the intuition. Of course, one might disagree with Greco or Craig, but that is actually fine. Simply the existence of such theoretical accounts undermine Lackey’s claim. My thanks to an anonymous reviewer for pointing out that my argumentative strategy might not have been clear on this point. The same reviewer also pointed out that from the perspective of “speech-act theory” there is something odd about Lackey’s counterexample. As I am unfamiliar with the particulars of the theory, I have decided to leave considerations of that perspective to someone better suited to the task. ↑
I am not even sure that 1 and 2 are satisfied because of what was said above about the complexity of knowing something like someone has cancer. However, even granting 1 and 2 still does not get Matilda knowledge. Therefore, I will focus on 3. ↑
Again, it is not even clear to me that Matilda has a true belief, but, for the sake of argument, it will be granted. ↑
Actually, to me even 100% reliability would not meet the standards necessary. For precisely the reasons which will be explained. ↑
John Greco has brought this to my attention by. ↑
I actually think many of them do, but the strength of the argument against Lackey is the force of everything taken together. ↑
Recall, I have focused on DOCTOR for simplicity. The same reasoning can be equally applied to all of her counterexamples. ↑
Actually, it is not unreasonable that Matilda might not do something like this even when Nancy is just a student. Instead of flat out asserting to Derek that he has pancreatic cancer, she would instead reschedule his appointment until she has looked at the test results, or preface the assertion with something like “It appears that […]” and then add “[…] but I will need to review the results more carefully”. ↑
Mittag’s Encyclopedia entry is used here for efficiency. For more complete accounts of evidentialism, see, for example, Conee and Feldman (2004). ↑
As with Mittag (2020), Becker’s Encyclopedia entry is used here for the sake of efficiency. The literature on Reliabilism is vast, its foremost proponent being Goldman (1979), but see also Heller (1995), and Becker (2006). ↑