Evidence One Does Not Possess
Skip other details (including permanent urls, DOI, citation information)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License. Please contact firstname.lastname@example.org to use this work in a way not covered by the license. :
For more information, read Michigan Publishing's access and usage policy.
Unpossessed evidence abounds. There is much to be seen and much to be had. Much of it will never have an impact on our epistemic standings, but some of it does. Some evidence is such that we are blameworthy for not having it, and there is a tricky question about how to delineate this class of evidence. In this paper, I address and propose a solution to the dilemma posed by lazy agents on the one hand and agents facing exceptional evidence on the other. One familiar suggestion is that agents are blameworthy when their conduct results from their vices, and I will make a proposal that further analyses such vices in terms of exceptionality facts.
Keywords: responsibility, evidence, opportunity, exceptionality, vice
Is the quality of your beliefs sensitive to evidence you do not possess, even if you are unaware of that evidence? Some hold that it is, others that it is not. To illustrate, here is a familiar key from Clifford (1877):
- S: Clifford’s shipowner.
- α: The proposition that his ship is seaworthy.
- E: A piece of evidence regarding α.
Setting: S believes α, and did not gather E. Does that matter? Well, it depends. Unpossessed E may have an impact on several of S’s epistemic standings regarding α. It may affect S’s knowing α (as Harman 1974; 1980; and others suggest), or it may affect his being epistemically justified or rational in believing α (as Gibbons and others have it), or his blamelessness in believing α (as it has often been put in the ethics of belief debate). In the following, I will focus on the latter issue: that unpossessed E might render S blameworthy or irresponsible.
How could unpossessed E have such an impact? Famously, Alston (1988) argued that we cannot directly control our beliefs, but we can indirectly influence them via evidence-gathering activities. The shipowner cannot simply decide to believe α, but he can gather or fail to gather evidence regarding his ship, namely support or defeaters for α, and influence his beliefs about α in this indirect manner. On this view, then, if we want to determine whether S is blameworthy for a certain belief, we should consider the evidence he does not possess. To be sure, it is not always the case that S is blameworthy for his belief in α if he failed to gather any arbitrary piece of evidence E regarding α. For perhaps he did not have the duty to gather E in the first place (perhaps he had already seen plenty of evidence regarding α), or perhaps he did have the duty but is excused for violating it (perhaps it was impossible to gather E given his resources). This suggests the following principle:
- If S believes α, S is blameworthy for this belief if there is a piece of evidence E regarding α which is such that
- S had a duty to gather E, and
- S is blameworthy for violating it.
If we are interested in blameworthiness concerning our beliefs, then we should be interested in blameworthiness concerning our evidence-gathering activities. This raises the question, when are we blameworthy for not gathering certain evidence? That will be the leading question in this paper. Given this focus, I will not defend (or criticize) the above principle itself. In this paper, I will argue that we are excused when evidence was ‘exceptional’, in a precise sense to be specified. Specifically, I will defend the following argument:
- (P1)S is blameworthy for his failure X to comply with his duty to gather E if X resulted from S’s vices.
- (P2)X resulted from S’s vices if S was aware he had to take his opportunity to gather E, or if this opportunity was non-exceptional.
- (C) Hence, S is blameworthy for X if S was aware he had to take his opportunity to gather E, or if this opportunity was non-exceptional.
The dialectic is this. Quite a few philosophers have suggested versions of (P1): blameworthiness can be understood in terms of vices (references will be provided in due course). In my view, these philosophers are on the right track, but I also think that virtue and vice talk can be further analysed in other terms. Here comes the role of (P2), my main contribution to the debate. (C) can be derived from the foregoing, and shows why (P2) is relevant: it yields a novel condition of blameworthiness. The plan for the paper is straightforward. In Section 2, I explain why I accept (P1). In Section 3, I will explain and defend (P2). To motivate both premises, I will examine quite a few cases (and close variants) from the epistemological literature.
Before beginning, a note on ‘the duty to gather E’ in this argument. In principle, the argument is compatible with any story on these duties so long as the blameworthiness question is a separate question. But let me say something on what might be understood by them. Duties concerning what evidence one should gather over time are sometimes called ‘diachronic’ duties, as opposed to certain ‘synchronic’ duties, which concern what epistemic standings one should have at a given moment in time (see Chignell 2010; Marušić 2011). Normally we may not say that one has duties to gather specific pieces of evidence (such as E). Rather, we speak of certain duties to keep oneself informed on a given topic, or duties to pay attention to a certain range of things. The shipowner has a duty to keep himself informed about his ship (reviewing certain reports, making certain inspections, and so on), but perhaps he has no duty to look at one particular set of test results. I will assume, though, that such latter duties do make sense. Suppose S did not gather a specific piece E. If we ask whether S is blameworthy for this, this will depend in part on whether S had a duty to gather E in the first place.
Why might S have a duty to gather E? Compare the following two cases. First: the shipowner is about to send his ship to sea full of emigrants (as Clifford’s own case description has it). Second: the shipowner has a ship that is no longer in use. In the first case, the moral stakes are high: α supports a rather risky action (namely, sending out the ship may lead to numerous deaths). Generally, the stakes of α are high iff it matters to the permissibility of one’s actions whether or not α is true. In the second case, there is nothing at stake, because it does not matter whether α is true if S is not going to act on the basis of his belief in α. This difference is clearly relevant: if the shipowner is not going to use his ship, then who cares what he believes and what information he has about α? So it seems that S has the duty to gather E only to the extent that, and indeed because, there is something at stake.
Still, even when there is nothing at stake from a moral perspective (because the shipowner is not going to use his ship and does not act on the basis of α in any other way), there can still be something at stake from some other perspective. Perhaps S’s belief in α does matter, not because it supports certain actions, but because it supports other beliefs. Suppose α is a key proposition in a certain theory: if α is false, then many further propositions are false too. For example, if the Earth is stationary and the center of the universe after all, and if most of present-day cosmology rests on the falsity of that proposition (let us suppose), then most of present-day cosmology is false. In such a case, it matters whether or not the given proposition is true, though this time from an epistemic perspective.
It is a controversial question whether, ultimately, there are such things as purely epistemic stakes (which cannot be reduced to practical concerns). I will not go into this issue, and we may simply assume that S has a duty to gather E when there is something at stake, whether this be morally (which would make the duty moral), epistemically (which would make it epistemic), or something else. To be sure, this is only meant as a placeholder for a fuller story, which is not the topic of this paper. In the following, we will just assume that S did have a duty to gather E (no matter the exact reason), and focus on the question whether he is blameworthy for violating it.
Compare the following two scenarios. First scenario: as the shipowner well knows, he could easily and freely examine the most recent test results on the website of the emigrant organization. They always publish such results on this website. But actually, he does not care much about the seaworthiness of his ship. He is more concerned with making profit, and so avoids carrying out costly repairs, even if this will put numerous lives at risk. Second scenario: this time, the shipowner does care about the safety of his ship, and he has tried to access the website of the emigrant organization. Yet the latter published no test results on its website. In fact, the right tools to discover the flaws in his ship have not yet even been developed.
In the first scenario, the shipowner seems to be blameworthy, while this is implausible in the second. In this section, we will make such verdicts more precise by discussing a number of factors that have a potential impact on S’s blameworthiness (namely, whether S has an opportunity to gather E, or could have created one, whether he is aware that he could and should gather E, and whether he is familiar with E’s source). Meanwhile, we will gather support for (P1) of the argument.
Does it matter whether S had the opportunity to gather E? Compare the following two cases. First: the shipowner could easily and freely have consulted the website of the emigrant organization to view the test results. Second: no test had been carried out, and hence no results were available yet. In the first case, but not the second, S had the opportunity to gather E. This matters: if S has no opportunity to gather E (no resources, or no capacities), then S need not be blameworthy for not doing it.
Generally, S has an opportunity to gather E whenever E is available to S, which means that S can acquire E by undertaking certain steps, such as asking questions, following news reports, making inferences, recalling something, paying attention, or carrying out experiments. Such steps must be open to the agent given his capacities and resources: he should be competent enough to ask the relevant questions, for example, and someone else should be around to answer them. Furthermore, if E is available to S, then this does not imply that S is always aware of E’s existence, but merely that S could become aware of E’s existence and content if he were to take such steps.
Importantly, in some cases S could have created an opportunity to gather E. Compare the following two cases. First: the test had not been carried out (and so the results are not available), but the shipowner could have ordered such a test (so that there would have been results). Second: the test had not been carried out, because the tools to test his ship have not yet been developed. In the first case, but not the second, S had the opportunity to make E available. So even if E is not available to S, S could still be blameworthy for not gathering E if he could have made E available. Strictly speaking, then, S is blameless if he had no opportunity to gather E and could not have created such an opportunity.
Does it matter whether S was aware that E was available and that he should gather it? Compare the following two cases. First: the shipowner knew that the emigrant organization carried out a test of his ship, and that he had a duty to consult it, but he did not ask for the results. Or perhaps the envelope with the test results are lying in front of him on his desk, but he does not open it. Second: unbeknownst to the shipowner, an emigrant organization carried out a test of his ship. Given that he was unaware of this, he did not ask for the results. The difference is that in the first case, but not the second, S is aware of E (i.e., of its existence but not of its content; if S were aware of E’s content, he would have gathered E; by hypothesis, however, S did not gather E).
This factor does seem relevant, but it involves a dilemma. On the one hand, the awareness condition cannot matter because of the Lazy Agent Problem that I will discuss next. On the other hand, the awareness condition cannot simply be dropped because of a problem that I will consider at some length in Section 3. The Lazy Agent Problem concerns agents who have only a very limited body of evidence because of their laziness and inattentiveness. Here are two examples of such lazy agents from the literature. Baehr’s George believes that
exposure to second-hand smoke poses no significant health risks. Given his extremely narrow and practical focus, George is entirely oblivious to all of the well-publicized research indicating the dangers of second-hand smoke. In fact, George’s belief that no such link exists is positively supported by his evidence. (Baehr 2009: 547)
Next, Dougherty’s Craig holds a certain creationist belief according to which the human species shares no common ancestor with other animals but was created separately:
The problem, though, did not seem to be that his beliefs did not fit his evidence⎯they did seem to fit his evidence, for he had read very narrowly on the subject and had been raised and schooled all his life in an apparently reliable community which sustained this belief in the usual social ways, and which had reasonable-sounding stories for why people deny their views. Rather, the problem seemed to be precisely that he only had the very limited evidence he had, since I had often recommended books challenging his views. (Dougherty 2012: 538)
These examples suggest that there are two distinct types of lazy agents. The first type is aware, or has evidence, that there is further evidence he should have (but does not gather). Craig is aware of the existence of further evidence that might challenge his creationist belief, but he dismisses it without considering it. The second type is unaware, or has no evidence, that there is further evidence he should have. George is not aware of the studies on second-hand smoke (or of any further evidence that might undermine his belief).
The behaviour of both types can be explained in terms of certain vices. Generally, vices (as well as virtues) are character traits of an agent. In the present context, vices such as laziness, inattentiveness, indifference, and arrogance explain why agents do not gather further evidence in the face of opportunities to do so (even if they are aware of such opportunities), and also why they lack evidence about the existence of such opportunities in the first place. They fail to have relevant evidence because they are lazy and inattentive.
Furthermore, both types may still be blameworthy for not gathering E. Craig may be blameworthy for not reading the books that challenge his creationist view, and George may be blameworthy for not updating himself about the research indicating the dangers of second-hand smoke. It is an open question which of the two types is more lazy, and more irresponsible. On the one hand, George is so lazy that he is not even aware of the existence of evidence he does not have. On the other, Craig is aware of such evidence, but is so lazy that he does not make any effort to gather it. At any rate, both seem blameworthy for their failures exactly because they result from their vices. Generalizing this thought yields (P1): S is blameworthy for his failure to comply with his duty to gather E if this failure resulted from S’s vices.
Two qualifications. According to many philosophers, such a condition cannot be sufficient for blameworthiness, since certain unfortunate histories may render one blameless (see Wolf 1987). Indeed, if S had no past opportunities to work on his vices and be less lazy and more attentive, then it seems he is blameless for them, and for whatever results from them (including his omissions to gather E). In this paper, I abstract from this debate by assuming that agents lack such unfortunate histories.
Second, (P1) is a sufficient condition. Whether it is also necessary for blameworthiness is a further question, and not relevant for the argument from (P1) and (P2) to (C). Moreover, it is doubtful whether it is necessary in all cases. For that would imply that virtuous agents (agents who are generally attentive and open to evidence they do not possess) can never be blameworthy in particular cases.
Central to my purposes is that George (i.e., the second lazy type, who is unaware of the need and opportunity to gather E) can be blameworthy as well. After all, if this is so, we should accept that awareness cannot be necessary for blameworthiness. Even so, in the next section I will show that dropping the awareness condition creates a serious problem which has received scarce attention.
Before turning to this, I will need one further factor. Does it matter whether S was familiar with E’s source? Compare the following two cases. First: the shipowner knew about the tests conducted by the emigrant organization and knew they were reliable. Second: suppose the shipowner could have informed himself by consulting an oracle (which, unbeknownst to him, is reliable). He is unfamiliar with this oracle (or perhaps he is familiar with it, yet considers it unreliable). In the first case, but not the second, he is familiar with E’s source. That is, he is aware that evidence of that kind (such as test results) is available from sources of that kind (such as online reports by such organizations). This familiarity condition is broader than the awareness condition discussed above given that S could be familiar with E’s source without being aware that a specific new piece of evidence from that source is available. The familiarity condition seems to matter. E has to be available via sources that the agent is familiar with (in this case, familiar testimonial sources such as online reports or letters).
I just argued that if S has the duty to gather E, and S has the opportunity to gather E, then it does not matter whether or not S was aware of that opportunity. This does not matter because of the Lazy Agent Problem. As said, though, dropping the awareness condition creates a problem to which we will turn now. If this problem had no solution, we would be stuck with a dilemma, for neither lazy agents nor the upcoming problem seem acceptable. However, I will argue that the problem does have a solution.
To introduce the problem, consider a variant of The Movie Times case from Feldman (2003: 47–48). A professor wants to go to the cinema. He knows that the movie he wants to see played at 8 p.m. yesterday, and on that basis believes that it will start at 8 p.m. today as well. However, the movie starts today at 7 p.m. Unbeknownst to him, had he looked in today’s paper that he was carrying with him, he would have known this. In this case, E is the paper reporting that the movie starts at 7 p.m., and, unless we have further information, we should not consider the professor blameworthy for not gathering it. After all, in contrast to Craig and George, the professor’s failure does not seem to result from his vices.
It would not be difficult to reach this verdict on the basis of an awareness condition. For the professor is unaware that he has to check the paper, and if awareness is required for blameworthiness, the professor is blameless for not checking it. Yet, if our discussion of lazy George in the previous section is right, there is no awareness condition. Hence, the problem arises whether there is any other condition that can render the professor blameless. That is, is there a suitable condition that is weaker than the awareness condition (and which does not let George off the hook), but still strong enough to excuse agents such as the professor?
Clearly, from what I have said it does not follow that the professor is blameworthy for not gathering E no matter what. Certain conditions have to be in place. For one thing, he should have had a duty to gather E. But let us assume that, for some reason, seeing the movie was sufficiently important for him. In addition, he is familiar with E’s source, and he had the opportunity to gather E. If all these factors are in place, is the professor blameworthy for not gathering E, even if he was unaware it existed?
To see what the problem does and does not involve, it is instructive to consider an example by Moon (2010). In Moon’s scenario, Bertrand believes that God does not exist, but when he arrives in the afterlife, God tells him that he did have the opportunity to know better:
“Well Bertrand, there was relevant, available evidence for my existence at various points in your life. For example, did you know that during the evening of June 16, 1956, when you were alone in your room reading your book, I had an angel silently hold up a sign with the message “God does indeed exist!” right behind your head for five minutes? If you had turned your head and checked, you would have had strong, relevant evidence for my existence; indeed, you would have come to know that I exist!” (Moon 2010: 147)
There is an important difference between this case and The Movie Times above. In The Movie Times, the professor is familiar with E’s source: he is aware that evidence of that kind (concerning movie times) comes from sources of that kind (the paper). In Moon’s case, in contrast, it is unlikely that Bertrand is familiar with E’s source (that is, the angel, with E being the sign the angel holds up). Indeed, it is not usually the case that evidence for someone’s existence can be found on the basis of messages held up by angels behind one’s head. Moreover, if Bertrand is unfamiliar with E’s source, that might explain why he is blameless for missing it. There is no such potential explanation in The Movie Times, which makes it more puzzling.
As I will suggest next, whether agents such as the professor are blameworthy depends on a further factor not yet noted, namely on E’s exceptionality. In The Movie Times, one might ask whether the movies in that cinema have usually started at the same time in the past. If so, the change of time would be exceptional, and hence the professor could not suspect that E would make a difference, and he is blameless. In contrast, if the movies started at different times in the past, then the change of time was non-exceptional, the professor could suspect that E would make a difference, and he can be considered blameworthy.
The problem can also be explained in terms of Clifford’s shipowner. Suppose the emigrant organization carries out a standard, yearly test of all emigrant ships, and the shipowner knows this. All previous tests have been unproblematic: no repairs were required. This year, however, the emigrant organization conducted an extra, unannounced test during the season. They planned this extra test because they wanted to try out some new, more reliable tools. According to the results of this extra test, which they published online, it turned out that the shipowner’s ship is not seaworthy. Given that the shipowner is unaware of this test, he does not examine its results. Or suppose the organization sent the results to his office. Given that he does not suspect any new results, he does not look in the envelope. That is, he neglects to view its content not because he is afraid he has to make repairs to his ship, but just because he has other things on his mind as he prepares for the new journey.
Call this case ‘The Unannounced Test’. Even though one could agree that the shipowner does have a duty to view the test results in this case, and that he does have easy opportunities to inform himself (he could consult the website, or open the envelope, which constitute two familiar sources), not everyone would consider him blameworthy for not taking these opportunities. Again, the problem is that we need a condition that can reach this verdict (i.e., a condition that is different from the awareness condition).
There is a similar problem with exceptional opportunities to make E available. Suppose that every year the shipowner asks the emigrant organization to test his ship. However, some other, familiar organization has been developing different and more reliable tools, and their current tests are not more expensive than the tests conducted by the emigrant organization. Given that the shipowner is unaware of this better option, he does not ask them to test his ship. Intuitively, the shipowner need not be blameworthy, but again it is unclear how to establish that verdict.
There appears to be a simple solution. Namely, one might suggest that agents facing exceptional evidence are blamelessly unaware that they should gather E, while lazy agents are culpably unaware that they should do so (and that S is excused for not gathering E if S was blamelessly unaware that he should do so). But the question is what that means in the present context. The foregoing cases suggest that the agents are blameless because the evidence was exceptional. Hence, we could say that, generally, S is blameless for not gathering E if E was exceptional (and if no further grounds for blameworthiness obtain). Given that E was exceptional in The Movie Times or The Unannounced Test, neither the professor nor the shipowner is blameworthy for not gathering it.
This raises the question what ‘E was exceptional’ amounts to. Does it admit of a further analysis? To explain my answer, let us run through a number of issues and see what makes E special or exceptional in The Unannounced Test. I will use the following key:
- α: The proposition that his ship is seaworthy now.
- E: The results of the extra test published online.
- E*: The results of the extra test sent to his office.
First issue: could S have gathered multiple pieces of evidence that undermined α? In The Unannounced Test, the shipowner could have gathered both E and E*. Thus, this is not what makes the evidence exceptional.
Second issue: did S have repeated opportunities to gather E and E*, or did he have just a single opportunity? In The Unannounced Test, the shipowner has various opportunities to go online and check the site of the emigrant organization, as well as various opportunities to open the envelope. So, this is not what makes E and E* special.
Third issue: did S have past opportunities to gather evidence similar to E and E*? Let us say that pieces of evidence are similar, roughly, iff they support or undermine the same proposition. So the question is whether S had past opportunities to gather further evidence that also undermined α (perhaps the test results from last year already indicated potential problems). He did not. E and E* are two new opportunities, and there was no further evidence available to S in the past that also undermined α (at least, the case description is silent about that). This is the first thing that makes E and E* special.
Fourth issue: did S have opportunities to gather evidence regarding propositions similar to α that undermined those propositions? Let us say that propositions are similar, roughly, iff they support the permissibility of similar actions or similar beliefs of the agent. Consider the following propositions:
- α*: The proposition that his ship was seaworthy last year.
- α**: The proposition that his other ship is seaworthy now.
These propositions are similar to α because they all support the action of sending out a ship (the first or second ship, now or then). Did S have opportunities to gather evidence that undermined these propositions? Again, the answer is negative: there was no evidence available to him that his ship was not seaworthy last year (and the case description is silent about other ships). This is the second thing that makes E and E* special.
Fifth issue: did S have opportunities to gather evidence from a single source, or from distinct sources (though familiar ones)? E and E* derive from the same source: the emigrant organization. It is right that this organization offered the test results in two different ways (via the website and via a postal delivery service), but the evidence would be less exceptional if other familiar organizations had provided similar test results. This is the last thing that makes E and E* special.
The problem of exceptional evidence can be solved by making blameworthiness sensitive to the latter three factors. S is blameless, then, in cases with exceptional evidence (The Movie Times, The Unannounced Test) because the following clause is not satisfied:
Non-exceptionality. S had further past opportunities to gather evidence, distinct from E, that undermines α and propositions similar to α, which derive from distinct familiar sources.
If this clause is satisfied, then both the need and opportunity to gather E were non-exceptional. Non-exceptionality does not imply that S was aware of the need and opportunity to gather E. For example, if the professor had a series of past opportunities to gather evidence that the movie times change everyday, or even that the movie started at 7 p.m., and if he could have figured this out on the basis of today’s paper plus other familiar sources (more on this below), then it need not be that the professor was aware of the need to check today’s paper concerning the movie times. After all, he might not have paid attention to any of those sources. (Still, in many cases it might well be true that S is aware of the need and opportunity to gather E if E is non-exceptional.)
The main idea here is that non-exceptionality can be used to check whether E is an exceptional case that might render S blameless. Let us see whether this condition meets all the desiderata. On the one hand, it should be weak enough to consider lazy agents blameworthy; on the other, it should be strong enough to excuse agents facing exceptional evidence. I just illustrated the latter part. In The Unannounced Test, the shipowner did not have the specified opportunities. Moreover, what goes for the shipowner goes for the professor, and all other agents facing exceptional evidence.
Have lazy agents had such past opportunities? Recall George and Craig (from Section 2). George had plenty of chances to check the well-publicized research on second-hand smoke, that is, a source that was familiar to him (let us suppose that he knew that information about health risks can be found in well-publicized research), and which would have undermined his belief that second-hand smoke poses no health risks. Thus the non-exceptionality clause is satisfied. The fact that he fails to gather this non-exceptional evidence is a clear sign of laziness, inattentiveness, or other vices (and he can be considered blameworthy for his failure).
The same does not carry over to Craig. For Dougherty’s case description suggests that Craig did not have past opportunities to learn that his creationist belief is mistaken. Unlike George, however, Craig is aware of the evidence he does not possess: he is aware of the existence of books that he had to read. Again, the fact that he fails to read these books is a clear sign of laziness, arrogance, or other vices (and he can also be considered blameworthy for his failure).
Thus, the non-exceptionality clause just formulated should be taken in disjunction with the awareness condition, which yields (P2): S’s failure to gather E resulted from S’s vices (and S is blameworthy for this failure, as per (P1)) if either S was aware that he had to take his opportunity to gather E, or if E was non-exceptional in the specified sense. This disjunction, then, seems both strong and weak enough to account for all the cases discussed in the foregoing.
The question arises whether the non-exceptionality clause is really sufficient for vices and blameworthiness. Specifically, what if S is unaware of the whole series of past opportunities to gather E-like evidence (specified in the non-exceptionality clause), and S is not culpable for this ignorance?
In response, I will suggest that there are no cases that fall under this description. First, there are many cases in which agents satisfy the non-exceptionality clause, but are no longer ignorant. Suppose that, due to no fault of his own, the shipowner missed a series of opportunities to find out that his ship had not been seaworthy in past years. In such a case, his ship would never have returned home, and from this he would surely have figured out that it had not been seaworthy (and thus would no longer be ignorant).
In this respect, The Movie Times presents a more challenging case. For suppose that, due to no fault of his own, the professor has been unaware of a whole series of opportunities to gather evidence that the movie actually started at 7 p.m. in the past month. If he never went to the cinema that month, he would never have figured out that it started at 7 p.m. In such a case, the evidence in today’s paper (namely, that his movie starts at 7 p.m.) is non-exceptional, though the professor does not seem vicious or blameworthy.
I agree that the professor might be blameless in such a case. Yet I do not think that the non-exceptionality clause is satisfied. For E to be non-exceptional, it is not enough that S had a series of past opportunities to gather E-like evidence (the paper from day n-1 reporting that the movie starts at 7 p.m. on day n-1; the paper from day n-2 reporting that the movie starts at 7 p.m. on day n-2; etc.). Rather, he should have had a series of past opportunities to gather distinct forms of E-like evidence, such as yesterday’s paper reporting that the movie starts at 7 p.m.; the cinema’s website stating that the movie starts at 7 p.m. today; a friend suggesting that the movies in that cinema always start at different times; or the travel guide reporting that the movies in that country always start at different times. In such a case, E is non-exceptional, and we would not want to say that the professor is blameless. Indeed, that S fails to gather E in the face of such a series of opportunities is a clear sign of inattentiveness and other vices. And if (P1) is correct, S might well be considered blameworthy in such a case.
Even so, the distinctness condition raises tricky issues. Suppose that an art-house cinema publishes their movie times in their trendy magazine. Suppose that the professor has no friends in the scene and he knows he will only get information about movie times from this magazine. Suppose further that every Friday the magazine prints that their one movie of the day is at 8 p.m. In this case, the evidence seems still non-exceptional for the professor, even in the absence of distinct sources.
To account for such cases, one cannot just say that the distinctness condition does not apply in cases where there is only one familiar source. For consider the following variant. The professor is a newbie to the art-house scene, and knows he will only get information about movie times from the magazine. He looks in it on Monday, which says the movie starts at 8 p.m. that day. Yet he does not feel like going. He looks again on Tuesday, and the magazine says the movie starts at the same time that day, 8 p.m. But in the afternoon his friend calls and he decides to go out for drinks. The professor checks again on Wednesday, and Thursday, and on both days the magazine says the movie starts at 8 p.m., though again he is not in the mood. On Friday, he finally decides to go to the movie at 8 p.m., but does not check. Had he checked, he would have found out that the movie started at 7 p.m. In contrast to the previous case, the evidence is exceptional (and the professor blameless for not gathering it), even though the number of available sources is just the same.
In response, it is important to note that evidence can be exceptional (or non-exceptional) in multiple dimensions. In the foregoing, we have considered several such dimensions. For example, it does not only matter if the professor could have consulted various sources on the issue (i.e., whether the movie starts at 8 p.m. on Friday), it also matters if he had opportunities to gather evidence regarding similar propositions (that the movie starts at 8 p.m. on Monday, or on Tuesday, etc.) that undermined those propositions. Clearly, in the latter case, the professor has no such opportunities (given that the movies actually started at 8 p.m. on the other days of the week), and this is a relevant factor that makes the evidence on Friday exceptional in the latter case.
Given that evidence can be exceptional in one dimension, yet less exceptional in another dimension, exceptionality is a matter of degree. Evidence can be exceptional to the extent that there is only one familiar source, and in both cases just described it still holds that the evidence would be less exceptional if there were more familiar sources available to the professor (sources he all fails to consult).
Does evidence you do not possess matter? As we have seen in the foregoing, any answer to this question is vulnerable to two sets of problem cases, namely, cases with lazy agents on the one hand, and cases with agents facing exceptional evidence on the other. Consider a piece of evidence E, any piece, that S does not possess. When is S blameworthy for violating a duty to gather E? S is blameworthy for this, as I have argued, if he was aware that he had to take his opportunity to gather E, or if this opportunity was non-exceptional, in the precise sense discussed. This is a substantial result. For according to it, excuses can be explained in terms of exceptionality facts. If E was exceptional in the relevant sense, one might be excused for not gathering E (as in The Unannounced Test). But if E was non-exceptional in the relevant sense, one might well be blameworthy (as in the case of lazy George).
My proposal adds a new dimension to any case of unpossessed evidence. Think for example of Harman’s (1974) classic cases (the book thief, the unopened letter, and the assassination case). In each of these cases, the question would be whether the unpossessed evidence was exceptional or not. If it was exceptional, the agents in question might well be blameless. As a final illustration of my point, consider Gibbons’s breakfast case:
The other morning, I went downstairs to make a mushroom, jalapeno, and cream cheese omelette. I had checked the night before to make sure we had all of the ingredients. Since Sunny rarely eats breakfast, it was reasonable for me to believe that the ingredients were still there. I went to the refrigerator and pulled out the eggs and mushrooms. While chopping, I firmly believed that I would soon have a mushroom, jalapeno, and cream cheese omelette. Unfortunately, in plain sight on the door of the refrigerator, there was a note [that I didn’t notice]. ‘We’re out of cream cheese.’ (2006: 22)
Gibbons’s claim is that he should have known that there was no cream cheese. Still, at one point Gibbons makes a similar point in terms of responsibility: “When the facts are both relevant and available, we can say that you are epistemically responsible for them. When the note is on the door, I am responsible for seeing it” (2006: 23). The question we have been asking corresponds to this, namely, is John blameworthy for not noticing the note on the door of the refrigerator?
As I have argued, all depends on whether Sunny’s note was exceptional. If she usually leaves her notes on another door, or if she never communicates on the basis of such notes, then her note was exceptional, and I do not think we should consider John blameworthy for not noticing it. That is, in such circumstances we should consider him blameless even if the note is obvious to all others for whom such evidence is not exceptional. Gibbons implicitly assumes that the note is non-exceptional when he adds the following information: “I didn’t notice the note, but I should have. After all, this is where we leave notes of this sort in our house” (2006: 22). In my view, however, this shows precisely that it is not just relevance and availability which determine whether we are responsible for certain unpossessed evidence (as Gibbons proposes), but also exceptionality.
As always, it remains to be seen whether the non-exceptionality clause formulated in Section 3 is precise enough. Yet it provides at least further grip to evaluate agents concerning their evidence-gathering activities. Given any case description, if we have no information about the agent’s vices, how do we determine that he is blameworthy or not? As proposed, we can determine this on the basis of certain exceptionality facts. Moreover, if the principle from Section 1 is correct, then we know not only what blameworthiness for evidence-gathering amounts to, but also what blameworthiness for belief could mean. For one to be a blameless believer, then, is at least for one to be blameless according to the account here defended.
Thanks to the editors and referees of the journal, my colleagues at Vrije Universiteit Amsterdam, and the participants of the Doxastic Agency and Epistemic Responsibility workshop in Bochum. This paper has been written within the context of the research project ‘We should know better. An inquiry at the crossroads of ethics and epistemology’ supported by the Netherlands Organization for Scientific Research.
- Alston, William P. (1988). The Deontological Conception of Epistemic Justification. Philosophical Perspectives, 2, 257–299. https://doi.org/10.2307/2214077.
- Axtell, Guy (2011). From Internalist Evidentialism to Virtue Responsibilism. In Trent Dougherty (Ed.), Evidentialism and Its Discontents (71–87). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199563500.003.0005.
- Baehr, Jason (2009). Evidentialism, Vice, and Virtue. Philosophy and Phenomenological Research, 78(3), 545–567. https://doi.org/10.1111/j.1933-1592.2009.00255.x.
- Baehr, Jason (2011). The Inquiring Mind: On Intellectual Virtues and Virtue Epistemology. Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199604074.001.0001.
- Chignell, Andrew (2010). The Ethics of Belief. In Edward N. Zalta (Ed.), Stanford Encyclopedia of Philosophy.
- Clifford, W. K. 1877. The Ethics of Belief. Contemporary Review, 29(1), 289–309.
- Cloos, Christopher M. (2015). Responsibilist Evidentialism. Philosophical Studies, 172(11), 2999–3016. https://doi.org/10.1007/s11098-015-0454-9.
- DeRose, Keith (2000). Ought We to Follow Our Evidence? Philosophy and Phenomenological Research, 60(3), 697–706.
- Dougherty, Trent (2012). Reducing Responsibility: An Evidentialist Account of Epistemic Blame. European Journal of Philosophy, 20(4), 534–547. https://doi.org/10.1111/j.1468-0378.2010.00422.x.
- Dougherty, Trent (2014). The Ethics of Belief is Ethics (Period). In Jonathan Matheson and Rico Vitz (Eds.), The Ethics of Belief (146–168). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199686520.003.0009.
- Feldman, Richard (2000). The Ethics of Belief. Philosophy and Phenomenological Research, 60(3), 667–695. https://doi.org/10.2307/2653823.
- Feldman, Richard (2003). Epistemology. Prentice Hall.
- FitzPatrick, William J. (2008). Moral Responsibility and Normative Ignorance: Answering a New Skeptical Challenge. Ethics, 118(4), 589–613. https://doi.org/10.1086/589532.
- Gibbons, John (2006). Access Externalism. Mind, 115(457), 19–39. https://doi.org/10.1093/mind/fzl019.
- Gibbons, John (2010). Missing the Obvious. Mind, 119(473), 153–158. https://doi.org/10.1093/mind/fzp150.
- Goldman, Alvin I. (2001). The Unity of the Epistemic Virtues. In Abrol Fairweather and Linda Zagzebski (Eds.), Virtue Epistemology: Essays in Epistemic Virtue and Responsibility (30–48). Oxford University Press.
- Greco, John (2005). Justification Is Not Internal. In Matthias Steup and Ernest Sosa (Eds.), Contemporary Debates in Epistemology (257–270). Blackwell.
- Guerrero, Alexander A. (2007). Don’t Know, Don’t Kill: Moral Ignorance, Culpability, and Caution. Philosophical Studies, 136(1), 59–97. https://doi.org/10.1007/s11098-007-9143-7.
- Haack, Susan (1997). The Ethics of Belief Reconsidered. In Lewis E. Hahn (Ed.), The Philosophy of Roderick M. Chisholm (129–144). Open Court.
- Hall, Richard J. and Charles R. Johnson (1998). The Epistemic Duty to Seek More Evidence. American Philosophical Quarterly, 35(2), 129–139.
- Harman, Gilbert (1974). Thought. Princeton University Press.
- Harman, Gilbert (1980). Reasoning and Evidence One Does Not Possess. Midwest Studies In Philosophy, 5(1), 163–182. https://doi.org/10.1111/j.1475-4975.1980.tb00403.x.
- Kornblith, Hilary (1983). Justified Belief and Epistemically Responsible Action. Philosophical Review, 92(1), 33–48. https://doi.org/10.2307/2184520.
- Levy, Neil (2007). Doxastic Responsibility. Synthese, 155(1), 127–155.
- Lycan, William G. (1977). Evidence One Does Not Possess. Australasian Journal of Philosophy, 55(2), 114–126. https://doi.org/10.1080/00048407712341141.
- Mason, Elinor and Alan Wilson (2017). Vice, Blameworthiness, and Cultural Ignorance. In Philip Robichaud and Jan Willem Wieland (Eds.), Responsibility: The Epistemic Condition (82–100). Oxford University Press. https://doi.org/10.1093/oso/9780198779667.003.0004.
- Marušić, Berislav (2011). The Ethics of Belief. Philosophy Compass, 6(1), 33–43. https://doi.org/10.1111/j.1747-9991.2010.00368.x.
- Montmarquet, James A. (1993). Epistemic Virtue and Doxastic Responsibility. Rowman and Littlefield.
- Moon, Andrew (2010). Gibbons on Epistemic Internalism. Mind, 119(473), 143–151. https://doi.org/10.1093/mind/fzp151.
- Nottelmann, Nikolaj (2007). Blameworthy Belief: A Study in Epistemic Deontologism. Springer.
- Peels, Rik (2013). Does Doxastic Responsibility Entail the Ability to Believe Otherwise? Synthese, 190(17), 3651–3669.
- Rosen, Gideon (2004). Skepticism about Moral Responsibility. Philosophical Perspectives, 18(1), 295–313. https://doi.org/10.1111/j.1520-8583.2004.00030.x.
- Wolf, Susan (1987). Sanity and the Metaphysics of Responsibility. In Ferdinand D. Schoeman (Ed.), Responsibility, Character, and the Emotions (46–62). Cambridge University Press.
- Zagzebski, Linda (1996). Virtues of the Mind. Cambridge University Press. https://doi.org/10.1017/CBO9781139174763
- Zimmerman, Michael J. (1997). Moral Responsibility and Ignorance. Ethics, 107(3), 410–426. https://doi.org/10.1086/233742.
Proponents include Harman (1974; 1980), Kornblith (1983), DeRose (2000), Goldman (2001), Greco (2005), Gibbons (2006; 2010), Baehr (2009; 2011), Axtell (2011), Cloos (2015). Opponents include Lycan (1977), Feldman (2000; 2003), Levy (2007), Moon (2010), Dougherty (2012; 2014). Each of these authors, to be sure, endorses a different version of these positions. This is not the place to discuss the differences, though I will identify some subtleties in due course.
Any proposition that supports or undermines α’s truth will be counted as evidence regarding α (excluding clearly false propositions). We need this non-mental notion of evidence in order to include evidence that S does not yet possess.
If S is blameworthy, he deserves blame, i.e., irrespective of whether this reaction is also useful for improving his behaviour. For the moment, I will not add further qualifiers such as ‘epistemic’ or ‘moral’ to this notion.
There might be further forms of doxastic influence that do not ultimately come down to evidence-gathering, broadly construed, yet still constitute a source of doxastic blame. This means that the analysis is only sufficient, or that the right-hand side should be read as ‘S is blameworthy for his actions or omissions of doxastic influence regarding α’. Alston (1988: 282) proposes a counterfactual variant of this clause. Alston’s proposal has not received much attention, cf. Nottelmann (2007: Chapter 10) and Peels (2013) for exceptions.
Cf. Clifford’s own case description: “[the shipowner] watched [the ship’s] departure with a light heart, and benevolent wishes for the success of the exiles in their strange new home that was to be; and he got his insurance-money when she went down in mid-ocean and told no tales” (1877: 289).
Of course, even if S had no opportunity to make E available, S could still be blameworthy for not gathering E if he could have created an opportunity to make E available (and so on). I will not consider such possibilities in what follows.
For philosophers who endorse (P1) or a close variant, see FitzPatrick (2008), Baehr (2009; 2011), Axtell (2011), Mason and Wilson (2017), among others, as well as the earlier studies by Montmarquet (1993) and Zagzebski (1996).
The different clause ‘S could have been familiar with E’s source’ is too broad. After all, the shipowner could have been familiar with the oracle. Still, there may be cases where S is not actually aware of a certain source, even though she should have been aware of it. In such cases, the clause ‘reasonably informed peers of S are familiar with E’s source’ might be a good approximation.
Feldman uses this case not to make a point about blameworthiness, but to argue that the epistemic justification of the professor’s belief is not undermined by the available evidence of which he is unaware. Indeed, if he is justified in believing that the movie will start again at 8 p.m., he may also be justified in believing that he need not consult any further sources on the issue. I have set the debate in terms of blameworthiness rather than justification, but if these notions are connected, then I would say that we do not know yet whether the professor is justified in believing that the movie will start at 8 p.m. (mainly because we do not know yet whether or not the evidence that it starts at 7 p.m. is exceptional, as I will argue below).
My analysis will be a non-modal one. A modal analogue might be suggested along the following lines. For both lazy agents and agents facing exceptional evidence it holds that it could easily be the case that they miss E, that is, in a wide class of nearby possible worlds, they miss E. The difference is that lazy agents are such that, in a wide class of nearby possible worlds, their non-lazy friends do not miss E, while agents facing exceptional evidence are such that, in a wide class of nearby possible worlds, all their friends (lazy and non-lazy alike) also miss E.
For completeness, this clause reads more precisely: S had, or reasonably could have had, such opportunities. Also, further qualifications might be added regarding degrees of exceptionality and regarding S’s capacities (for S needs to have the capacity to remember that past evidence and to draw conclusions from it).
Hence, the exceptionality clause of (P2) can be read more precisely as follows: X resulted from S’s vices if, and to the extent that, this opportunity was non-exceptional in the various dimensions discussed in this section.
These form two major sufficient conditions that track vices and render agents such as Craig and George blameworthy. Yet this does not exclude that there are further such conditions. Let me mention two candidates. Suppose S was raised in a hostile environment with many killings, so that his opportunity to gather E (here, evidence that killing is wrong) is exceptional for him. Even so, S might be blameworthy for not gathering E, and generally for not gathering evidence about moral truths. Or suppose S never cares about evidence he does not possess. In such a case, he might still be blameworthy for not gathering E, that is, even if E was exceptional.
My response to Gibbons differs from Moon (2010). According to Moon, John should have known that there was no cream cheese only if some awareness condition holds (e.g., he was aware of the note after all, or at least of the fact that Sunny always leaves such notes on the refrigerator and that there might be one about the cream cheese). As discussed, I reject such an awareness condition because of the Lazy Agent Problem. On my analysis, agents may be blameworthy for not gathering non-exceptional evidence (given that the latter is due to certain vices), even if no such awareness condition is fulfilled.