1. Introduction
It seems that when we reason about what to believe, we identify some considerations that are reasons in favor of believing something and then we choose to believe that thing. For example, a detective who is trying to figure out who murdered the mayor thinks that the fact that the butler's fingerprints are on the knife that was sticking out of the mayor's chest is evidence in favor of believing that the butler is the murderer. That is, the detective thinks she has a reason to believe that the butler is the murderer (even if not a decisive reason), which is that his fingerprints were found on the supposed murder weapon. The detective could acknowledge this reason and then, on the basis of that reason, believe that the butler is the murderer.
Suppose that, in addition to the fingerprints on the knife, there is a video recording of the butler stabbing the mayor in the chest. Now, it seems like the detective has decisive reason (assuming there are no defeaters) to believe that the butler is the murderer. But suppose that the detective does not believe that the butler is the murderer, despite all the information pointing to that conclusion. In that case, the detective's assistant could criticize the detective for not believing as she ought. In other words, the detective's evidence obligates her to believe that the butler is the murderer; she is irrational if she does not do so.
The foregoing illustration of what it looks like to be obligated to believe something is representative of a common view. However, this view is probably fundamentally mistaken. However counterintuitive it may seem at first, it is probably a mistake to hold that we can have reasons to believe something, that such reasons obligate us to believe that thing, and that we form beliefs on the basis of such reasons. I say “probably” because I will not argue directly against this view. Rather, I argue for a conditional claim: if doxastic involuntarism is true, then this view is wrong. This is because doxastic involuntarism is inconsistent with there being reasons for belief, and if there cannot be reasons for belief, then there cannot be obligations to believe things. The main burden of this paper will be to establish these latter claims. That is, I focus on arguing that doxastic involuntarism is inconsistent with something that is necessary for there to be obligations to believe – that is, with reasons for belief.
You may think it plausible that when you are obligated to believe that p this is at least partly because you have an undefeated reason to believe that p (cf. Kiesewetter Reference Kiesewetter2021). However, this “plausible” assumption has a lot of contentious terms packed into it. By “obligation,” let's mean a normative obligation, which we can understand as what establishes what you objectively ought to do. That is, when you have a normative obligation to do something, then you ought to do it, full stop. By “reason,” let's mean a consideration that we can reason with and that favors a conclusion or conclusions to varying degrees. Let's understand “consideration” ecumenically. That is, we will not take a stand in this paper on whether considerations are facts, states of affairs, mental states, etc., though the language I use may lend itself better to some understandings than others. One of the basic things we should assume about reasons in this paper is that they favor or support conclusions. Moreover, let's use “reasons” to refer to normative reasons – that is, reasons that, when undefeated and all enabling conditions are present, are necessary conditions for normative obligations.
According to a common view of obligations to believe, when you are obligated to believe that p, this just means that you are obligated to form the belief that p (cf. Feldman Reference Feldman2000, Reference Feldman2008; Littlejohn Reference Littlejohn and Star2018; McHugh Reference McHugh2014; Ryan Reference Ryan2003; Steup Reference Steup2017). I will understand “being obligated to form the belief that p” to mean that you are obligated to perform an act of belief formation. That is, you are obligated to perform an action that just is the formation of a belief (though I write in terms of forming a belief, the same things could be said of getting rid of beliefs).
When you are obligated to believe that p, you have this obligation at least partly in virtue of some reason specifically to do something – to perform an act of belief formation. We can call those reasons that obligate you to believe that p (either on their own or along with other conditions) reasons for belief. Reasons for belief, in this sense, form a subset of the broader category of reasons for action, which we can understand as reasons that obligate us to do something, either on their own or in conjunction with other conditions. For according to this way of understanding doxastic obligations, since you perform an action when you form a belief, reasons that favor you forming a belief are reasons that favor you doing something – they favor you performing an act of belief formation.
The claim for which I argue in this paper is that, if doxastic involuntarism is true – that is, if it is true that the formation of a belief is not an action – then there are no reasons for belief of the sort that is my focus. This is because there is a constraint on what it takes for a consideration to be a reason for action that precludes there being reasons for actions that favor non-actions. If a consideration violates this constraint, then it is not the sort of thing that can be a reason for action; and my claim is that purported reasons for belief violate this constraint (henceforth I will leave the “of the sort that is my focus” qualifier implicit). Furthermore, if there are no reasons for belief, and if my assumption that normative reasons are necessary conditions for normative obligations is correct, then there also cannot be any normative obligations to believe things. This is not to say that we do not have any obligations regarding our doxastic states. In fact, I think we do have obligations regarding our mental lives, such as the obligation to be the sort of person who is sensitive to evidence. Rather, it is just to say that the sorts of obligations we can have regarding our doxastic states do not include obligations to perform acts of belief formation.
I begin the next section with a presentation of my argument and follow that with some clarifications and defenses of how I understand some of the terms. In the final section, I defend my view against some objections.
2. The argument
Here is the argument I explain and defend in the rest of this paper:
(1) If A-ing is not an action, then there are no normative reasons for A-ing.
(2) If doxastic involuntarism is true, then the formation of belief (or believing) is not an action.
(3) Therefore, if doxastic involuntarism is true, then there are no normative reasons favoring forming a belief (or believing).
As a general comment, when I mention normative reasons, I mean normative reasons specifically for rational human agents. We can leave it open that normative reasons, normative obligations, etc., for rational human agents could be different than normative reasons, obligations, etc., for other types of rational beings.
In support of (1), note that if it is not up to anyone whether they A, then there cannot be any normative reasons for them to A. But if A-ing is not an action, then it is not up to anyone whether they A. So, if A-ing is not an action, then there cannot be any normative reasons for A-ing. Now, there are two general ways in which something can be up to someone. First, if someone can A, or if he can refrain from A-ing, then A-ing is up to him (i.e., if A-ing is an action, then if he can perform that action or refrain from performing that action, then A-ing is up to him). For example, it is up to me whether I raise my arm because I can raise my arm (i.e., I can perform the action of raising my arm). Second, if someone can bring about A, then the occurrence of A is up to him (i.e., A is not an action, but a person can make A occur through an action or series of actions). For example, though I cannot fall asleep in the sense of performing an act of falling asleep (since it is not an action), it is up to me in this second sense whether I am asleep because I can bring it about that I fall asleep.
Of course, if A-ing is not up to someone in the first sense, then A-ing is not an action and there cannot be any normative reasons for A-ing. However, you may think that if something is up to you in the second sense, then there can be normative reasons favoring that thing; and you may think that believing is precisely the sort of thing that is up to us, even if it is not an action. However, holding that normative reasons favor the thing that is brought about is to confuse what such reasons favor.
Consider the options for what normative reasons could favor when something is up to us in the second sense: (1) they could favor the actions that bring about A, but not A itself; (2) they could favor A, but not the actions that bring about A; (3) or they could favor both the actions that bring about A and A itself. (2) makes no sense. For example, it makes no sense to affirm that I have a normative reason to sleep, but that reason does not favor me performing the actions that are necessary for me to bring it about that I am asleep.
Further, (3) is less plausible than (1). One reason for this is that the actions that bring about A are sufficient for A to occur. It is overkill to affirm that a normative reason favors you performing the actions that bring it about that you are asleep and that it favors you being asleep, since the former is both logically and temporally prior to the latter, and the former is sufficient for the latter. More precisely, it is simpler to affirm that a normative reason only favors the action (or actions) that brings about A, since a favoring relation between the normative reason and A is unnecessary and explanatorily useless.
You may think that having a normative reason that favors you bringing it about that A depends upon you having a normative reason that favors A itself. If this is right, then a normative reason that favors A is neither unnecessary nor explanatorily useless. For example, you may think that you have a reason that favors you performing actions that bring it about that you are asleep only because you have a reason that favors you being asleep. Similarly, you may think that you have a reason that favors you performing actions that bring it about that you believe that p only because you have a reason that favors you believing that p.
However, it is not necessary to understand a normative reason that favors bringing it about that A to depend specifically upon a reason that favors A itself. Instead, we could understand a normative reason that favors bringing it about that A to depend upon some value (i.e., some value that is distinct from whatever normative value we can say normative reasons have). For instance, you have a reason that favors you performing actions that bring it about that you are asleep because it is good for you to sleep at this time, or because it is good that you be asleep at this time. Further, I contend that it is just as plausible or even preferable to understand the dependence relation to be between a reason that favors bringing it about that A and some value rather than to be between a reason that favors bringing it about that A and a reason that favors A itself. For, compared to what it means for A to be good for you, it is less clear what it means for a reason to favor A, where A is a non-action that is up to you in the second sense, and we are not understanding this in terms of it being a reason that favors bringing it about that A.
So also, even if it is true that believing is up to us in the sense that, though forming a belief is not an action, we can bring it about that we have a belief through actions, we should not take this to mean that there could be normative reasons that favor believing. The reason for this is that, whatever normative reasons we can have that favor us bringing it about that we believe something, it is more plausible to affirm that such reasons only favor us performing those actions that bring it about that we believe something. Yet, what a normative reason favors is what that reason is for: a normative reason that favors performing actions that bring about a belief is a normative reason for those actions, rather than a normative reason for forming that belief. So, what matters for my argument is whether forming a belief is up to us in the first sense, namely, whether forming a belief is an action. For, if forming a belief is up to us only in the second sense, then there are no normative reasons for forming a belief. Let's turn to a defense of the second premise.
(2) could be controversial because you might disagree with how I characterize doxastic involuntarism (but cf. Audi Reference Audi, McCain and Stapleford2020). There are several ways to understand doxastic involuntarism, and I acknowledge that my characterization will not satisfy everyone. The main divide is between understanding the formation of belief to be an action or something other than an action. Despite the possibility of several different views, let me give a few reasons for why I think my characterization of doxastic involuntarism is preferable. First, if we understand doxastic involuntarism as the view that, though the formation of belief is an action, we do not in fact have the ability to perform such an action, then our debate will become mired in competing intuitions over cases. For, there are cases in which it seems that an agent chooses to believe something. If the doxastic involuntarist grants that the agent could, in principle, perform an act of belief formation, but insists that the agent does not in fact have the ability to do so, then it seems that there is nothing more that the doxastic involuntarist could do than to voice a differing intuition about the case. In contrast, understanding doxastic involuntarism as the view that the formation of belief is not an action provides an alternative interpretation of such cases.
A second reason for why I think that my characterization of doxastic involuntarism is preferable is that it requires minimal extra commitments. That is, if we hold that the formation of belief is an action, but one that we cannot perform (either because we lack some sort of control over our doxastic states, because we lack some sort of requisite ability, etc.), then, in addition to a view about the nature of belief, we must also include a view about the nature of control over actions, or a view about the nature of some special doxastic ability, etc. In contrast, the version of doxastic involuntarism according to which the formation of belief is not an action could only require a view about the nature of belief – the formation of belief is not an action because beliefs are not the sorts of things we acquire through actions. Of course, if beliefs do not result from acts of belief formation, then we would still need an explanation of some mechanism that produces beliefs. However, the explanation of such a mechanism could be part of a larger explanation of the nature of belief.
There is another thing that might make you hesitant to move from (1)–(2) to (3), namely, that you might have a qualm with how I am using the term “reason.” You might think that evidence favors beliefs. For example, that blotch in your x-ray is evidence that you have a fractured thumb. This evidence favors the belief that you have a fractured thumb. In this way, there can be reasons for belief, since evidence that favors beliefs can function as reasons for beliefs.
However, there is an ambiguity in saying that evidence favors beliefs. First, this could mean that evidence favors – in the sense of “supports” – the content of beliefs. For instance, the blotch in your x-ray supports the truth that you have a fractured thumb. I have no problem with understanding the function of evidence in this way. However, a second way to understand what it means for evidence to favor beliefs is that it favors agents performing acts of belief formation. For instance, the blotch in your x-ray favors you forming the belief that you have a fractured thumb. It is this second understanding of the function of evidence that (1)–(3) is meant to discredit. If (1)–(3) is a good argument, then we should not understand evidence as functioning as reasons that favor agents forming beliefs.
However, you may think that having such narrow conceptions of reasons for belief and doxastic obligations makes my argument trivial. For, you may think that a better way of understanding these terms is in a non-actional way. That is, rather than holding that reasons for belief are a subset of reasons for action, and that doxastic obligations are obligations specifically to do something (i.e., to perform an act of belief formation), an objector might simply say that reasons for beliefs favor beliefs and that doxastic obligations are non-actional obligations to believe things. That is, doxastic obligations are obligations to which an agent is beholden, but which do not require the agent to perform actions for their satisfaction/fulfillment. This objection centers on a dispute over whether doxastic obligations are actional or non-actional, rather than over whether some normative reasons favor beliefs. I agree that the contents of reasons support the contents of beliefs. For example, if reasons are understood as evidence, I agree that the content of evidence supports the content of some belief in the sense that it supports the truth of the proposition that is believed. My claim is that this does not establish a doxastic obligation. So, the focus should be on the nature of doxastic obligations – whether they are actional or non-actional. In the next section, I defend my argument against this sort of objection.
3. Defense
To recap, here is the rationale for why the success of (1)–(3) supports the claim that there are no doxastic obligations: normative reasons are necessary conditions for normative obligations; so, reasons for belief are necessary conditions for doxastic obligations. If there are no reasons for belief, then there cannot be any doxastic obligations. To insist that there could be doxastic obligations even in the absence of reasons for belief would result in something similar to a Moorean paradox. That is, you would be landed with sentences such as “I am obligated to believe that p, but I have no reason to believe that p.” Such a statement is hard to make sense of. This is perhaps clearer if we consider similar statements about other domains, such as “I am obligated to tell the truth, but I have no reason to tell the truth,” or “I am obligated to give back what I stole, but I have no reason to give back what I stole.”
Here is where the objection from the end of the previous section rears its ugly head: a doxastic obligation is something other than an obligation either to form, alter, or relinquish a belief directly and intentionally. Rather, a doxastic obligation is a non-actional obligation.
Feldman (Reference Feldman2000, Reference Feldman and Steup2001) provides one way of understanding doxastic obligations non-actionally. According to Feldman, we can have obligations to believe things even though we are unable to form such beliefs directly and intentionally. For, we can have obligations that apply to us in virtue of us occupying certain roles, and these obligations apply even when we are unable to fulfill the functions of such roles. Feldman calls such obligations “role-oughts.”
An example of a role-ought is the obligation teachers have to instruct their students accurately. Even if Maria is not capable of instructing her students accurately on some subject for which she is assigned to provide instruction, she still ought to instruct her students accurately, for that is what teachers ought to do, and Maria is a teacher. Her incompetence or inability does not excuse her of her responsibility to instruct accurately, insofar as she occupies the role of a teacher. Another example of a role-ought is the obligation of a renter to pay rent to the property owner. If Arthur decides to move into the ritzy Oasis Luxury Town Homes, then he is obligated to pay the monthly rent, even if he does not have the money to do so. The property owner holds Arthur to this standard when she evicts him for not paying rent.
Others, such as Oliveira (Reference Oliveira, Stapleford and McCain2020), Sullivan-Bissett (Reference Sullivan-Bissett, McCain and Stapleford2020), and Altschul (Reference Altschul2014) accept doxastic involuntarism but maintain that there are things along the lines of doxastic role-oughts that apply to us. Oliveira distinguishes between prescriptive and evaluative doxastic obligations and argues that we can have evaluative doxastic obligations even if doxastic involuntarism is true. Here is how Oliveira puts it (2020: 50):
When we say truly that S ought to have doxastic attitude D toward p, that is, we are always expressing something of a kind with what we express by saying that the clock ought to strike on the quarter hour; we are always merely claiming that S's having doxastic attitude D toward p at t is well-ranked by a certain epistemic standard that we accept or take for granted, and we are never claiming that there is a relation of normative requirement holding between S and the having of D at t.
In this passage, Oliveira states that doxastic obligations are akin to expectations. That is, just as we say that a clock ought to strike at a certain time, meaning that, if the clock is working properly, then it will strike at that time, so also we say that a person ought to have a certain belief, but here we mean that, if the person is functioning properly, then they will have the certain belief. Oliveira calls this sort of “ought” an evaluative doxastic obligation. What we should not mean, according to Oliveira, is that the person is normatively bound to acquire the belief. Oliveira calls these sorts of “oughts” prescriptive doxastic obligations.
Oliveira argues that taking his “evasive” strategy comes at a cost, which is that the sorts of doxastic obligations we are left with are revisionary. For, according to Oliveira, evaluative doxastic obligations lack binding authority – the normative “umph” that is characteristic of practical obligations such as the obligation to keep your promise to your friend (ibid.: 53–54). Nevertheless, he maintains that evaluative doxastic obligations are uniquely doxastic in that they have to do with whether you believe that p, believe that not-p, withhold belief about whether p, and so on (cf. Chrisman Reference Chrisman2008). Oliveira also maintains that evaluative doxastic obligations are genuinely normative even though they are not authoritatively binding, since their function still is to hold agents – as doxastic agents – to doxastic standards.
Sullivan-Bissett is not explicit about the distinction between prescriptive and evaluative doxastic obligations, but she nevertheless clearly has something like doxastic role-oughts in mind with her arguments. According to Sullivan-Bissett, doxastic obligations are similar to the types of norms that are commonly applied to non-human animals (2020: 99–100). For example, we might say that an apple tree ought to grow apples. If it does not, then something is wrong with it – it does not meet the standard to which we hold apple trees.
When it comes to beliefs, we can look to the evolutionary functions of our doxastic mechanisms to ascertain what they ought to do. One such function could be to obtain true beliefs about the world. Given this function, it follows that humans ought to obtain true beliefs about the world. Another evolutionary function of our doxastic mechanisms could be to obtain useful beliefs (which needn't always be true). As with the first function, given that we evolved to obtain useful beliefs, it follows that humans ought to obtain useful beliefs (ibid.: 100–101).
Sullivan-Bissett's idea is similar to recent Neo-Aristotelians, such as Foot (Reference Foot2001), who argue that normativity is “baked into” the natural world. Put differently, the idea is that the natural world (humans included) contains species that have proper functions that serve as standards according to which individuals within each species are evaluated. When a member of a species does not live up to how a member of its species ought to function, then that member ought to function differently. The normativity involved in a bear hibernating during the winter is the same (as far as normative authority is concerned) as the normativity involved in humans believing in accordance with their evidence.
However, Sullivan-Bissett does not go as far as the Neo-Aristotelians in her claims about the normativity of natural functions. For, although the Neo-Aristotelians typically hold that natural normativity has all the authority of what Oliveira terms prescriptive obligations, Sullivan-Bissett (similar to Oliveira) holds that her biological account of belief yields a revisionary understanding of doxastic obligations. Her conception of doxastic obligations is revisionary in that she does not think that doxastic obligations are binding on us. In other words, even though it is true that we ought to obtain true and useful beliefs, when we do not do so, we cannot appeal to the proper function of our doxastic mechanisms to explain any blameworthiness. For, there may not in fact be any blameworthiness, just failure to function properly.
Altschul takes a slightly different tactic than either Oliveira or Sullivan-Bissett. According to Altschul, and similar to Oliveira and Sullivan-Bissett, the truth of doxastic involuntarism rules out prescriptive doxastic obligations. However, Altschul also argues that the truth of doxastic involuntarism rules out evaluative doxastic obligations, at least according to how Feldman, Oliveira, and Sullivan-Bissett understand them.
Consider Feldman's example of involving parents. Feldman argues that doxastic obligations are like parental obligations in that parents are obligated to care for their children even when they cannot do so, and we, as believers, are obligated to form beliefs in accordance with our evidence even though we do not have direct voluntary control over our beliefs. Altschul argues that parenting and believing are too disanalogous for Feldman's theory to work. For, people often voluntarily become parents, and once they are parents they have direct voluntary control over each action they perform in relation to their children. In contrast, we do not voluntarily become believers, and (assuming the truth of doxastic involuntarism) the formation of a belief is never an action. So, even if it is true that parents can be obligated (in virtue of filling the role of a parent) to do things they cannot do, this does not warrant us to think that we have doxastic obligations (in virtue of us filling the role of a believer) despite the truth of doxastic involuntarism.
Altschul argues instead that believing is much more like sleeping and perceiving than it is like either parenting, teaching, or eating (2014: 255–56). For, we do not have direct voluntary control over falling asleep, and once we are asleep we do not have direct voluntary control over how we sleep (i.e., falling asleep and sleeping are not actions). Nevertheless, there is still an intelligible sense in which there are standards to which we can be held as sleepers. For instance, when my friend sees me in the morning with blood-shot eyes, disheveled hair, and a drooping posture, and learns that I did not sleep last night, she might say to me, “What were you thinking? You ought to have gotten some sleep last night!” In this case, my friend holds me to a standard according to which I ought to have slept last night, even though, alas, falling asleep is not an action and so not an action I can perform.
Just as we can be beholden to obligations to sleep, despite our lack of dormic agency, so also, according to Altschul, we can be beholden to doxastic obligations, despite our lack of doxastic agency. For, both our obligations pertaining to sleep and our obligations pertaining to our doxastic states obligate us to perform actions that facilitate us conforming to the standard to which we are held: our obligations to sleep require us to facilitate falling asleep, and our obligations to believe require us to facilitate forming beliefs. Here is how Altschul puts it (ibid.: 255):
Usually, when we hear someone say to us, – You ought to sleep (or – You ought to get some sleep), we know exactly what instructions are being handed down. The instructions are not to be asleep, even if, as matter of social convention, this is something we are often told to do (said to Timmy: – You had better be sleeping by the time I come back to check on you!). Rather the instructions are only to get into bed, turn out the lights, close our eyes and then wait for sleep to wash over us.
And a bit further along he writes (ibid.: 260):
Just as there are bodily actions Timmy can do so as to facilitate the state of sleep (e.g. lie down in bed), there are mental actions epistemic agents like Sam can do so as to facilitate properly responding to one's evidence. In particular, while in deliberations one can willfully choose to pause and ask (or even re-ask) oneself the question: – What are my reasons here?
Altschul's position is different from Oliveira's and Sullivan-Bissett's insofar as the doxastic obligations he posits are authoritatively binding on those to whom they apply. That is, when I fail to facilitate forming beliefs in accordance with my evidence, that is a distinctively doxastic failure on my part, and I am blameworthy specifically for failing to hold certain beliefs, and these things can be true even given the truth of doxastic involuntarism.
The three positions just presented are representative of some of the ways in which people argue for a non-actional understanding of doxastic obligations. Though my list of representatives is not exhaustive, my arguments are intended to cover not just the theories outlined here but also similar theories. I contend that Oliveira's, Sullivan-Bissett's, and Altschul's theories fail to establish uniquely doxastic obligations.
Let's start with Oliveira's and Sullivan-Bissett's theories. Both argue that there can be evaluative doxastic obligations even if doxastic involuntarism is true, but that these obligations do not carry normative authority. For both theories, the idea is that an individual can be evaluated in terms of a standard, where the standard can be understood as identical to some type or category of which the individual is a token. Such an evaluation is essentially normative, even though it does not, on its own, bind the agent to some duty to do something. Oliveira and Sullivan-Bissett both acknowledge that their views of doxastic obligations are revisionary, but they nevertheless hold that the standards are genuinely doxastic obligations – that is, obligations that govern doxastic agency.
The main problem with Oliveira's and Sullivan-Bissett's theories is that they fail to note that most (if not all) evaluative obligations are obligations specifically to be a certain way. When it comes to evaluative doxastic obligations, these are best understood as obligations to be the sort of person who has certain beliefs. Put differently, I suggest that the sorts of obligations we have that concern our doxastic states are best understood either as evaluative obligations that apply to our characters, or as practical obligations to perform actions that put ourselves in positions that are conducive to acquiring true beliefs, not as obligations that apply directly to our doxastic states. If the formation of belief is not an action, then any sort of obligation to believe something (or to have certain beliefs) should be understood either as an obligation to be the sort of person who has some belief or who would gain some belief when in a specified situation, or, if you are already the sort of person who would gain some belief when in a specified situation, as an obligation to perform some action that puts yourself in that situation. But such obligations are not doxastic obligations at all, but rather obligations pertaining either to character – what we can call aretaic obligations – or to practical actions.
In general, obligations are the sorts of things that require something of an agent. The requirement of an obligation to believe could be either a requirement to believe that p or a requirement to be a certain way, which could involve a requirement to bring about the belief that p in some way. If the requirement to believe that p does not consist in the requirement to perform an act of belief formation, then this requirement just amounts to the requirement to be a certain way. Otherwise, it is difficult to make sense of what the requirement of an obligation to believe that p is supposed to be. Yet, if the requirement is to be a certain way, then it is preferable to understand the obligations that concern our doxastic states as either aretaic obligations or practical obligations rather than specifically doxastic obligations.
The aretaic obligations are obligations to be a certain way – to be sensitive to evidence, to be the sort of person who has certain beliefs, etc. The prescriptive practical obligations are obligations to place yourself in a situation that is conducive to acquiring true beliefs. The satisfaction/fulfillment of aretaic obligations would usually not be achieved through performing some isolated action but rather could be through performing actions that change the agent's sensitivities, dispositions, etc., over time (though it is unclear whether Oliveira and Sullivan-Bissett would agree that evaluative doxastic obligations could indirectly be fulfilled through non-belief-forming actions).
Of course, I am not denying that we believe things. I am not denying that there is a sense in which we can make evaluative judgments about human agents specifically concerning our doxastic states. For example, if we were to encounter someone who somehow had no beliefs, of someone whose beliefs were all false, we would rightly think that something was wrong with that person. Similarly, if we were to encounter someone who believed everything without reflection or scrutiny, we would rightly think that something was wrong with that person. So, there is a sense in which I agree with, say, the claim that humans ought to have beliefs, just as I agree, in a sense, with the claim that we ought to have fingers. However, the standard according to which humans ought to have doxastic states, despite how it may seem, does not really have to do with our doxastic states directly.
The claim that humans ought to have doxastic states could be ambiguous between the claim that we ought to be the sorts of people who have certain doxastic states and the claim that we ought to have cognitive mechanisms that form and retain beliefs (this is another way of understanding obligations to “be a certain way”). However, on neither construal of the claim that we ought to have doxastic states is it true that the obligation is specifically doxastic – that is, that the obligation governs our doxastic states. For the ways in which we can indirectly influence our doxastic states, such as performing actions that put us in a better position to recognize certain truths, are not uniquely doxastic but rather practical or aretaic.
On one reading of the claim that humans ought to have true doxastic states, this means that our characters ought to be such that we are properly sensitive to our evidence. However, this is an aretaic obligation, much like the evaluative obligation for us to be generous. Put differently, the obligation in this case is not to exercise some control over our doxastic states, but rather to exercise some control over our characters via actions we can take to make ourselves more sensitive to evidence. The negative evaluation of someone who does not have the sorts of beliefs he should is an evaluation of the sort of person he is, namely, the sort of person who is not sensitive to evidence, who is incoherent, etc. Alternatively, assuming that you already are the sort of person who is properly sensitive to your evidence, the “obligation to have true doxastic states” would just mean that you should perform some action that puts yourself in a situation that is conducive to acquiring true beliefs.
However, another construal of the claim that humans ought to have true doxastic states, according to which we ought to have cognitive mechanisms that form and retain true beliefs, is also not an evaluative doxastic obligation. For, the standard applies to the psychological/biological faculties that underlie the doxastic states rather than to the doxastic states themselves. This is similar to an evaluative obligation according to which we ought to have lungs that enable us to breathe. This standard applies to the structures that enable the breathing (i.e., the lungs) rather than to the breathing itself.
The person who believes everything without reflection or scrutiny does not possess the type of character that is sensitive to situations in which reflection and scrutiny are appropriate. This lack of character does not establish an evaluative doxastic obligation. Rather, if the standard is a character that is sensitive to situations in which reflection and scrutiny are appropriate, then the obligations that result would have to do with the agent performing actions that over time would likely produce that sort of character trait.
My argument so far may seem quite similar to Altschul's position. In fact, Altschul's theory is close to my own. However, he maintains that the only sense in which we have evaluative obligations related to our doxastic states is still genuinely doxastic. That is, Altschul argues that our obligations to facilitate appropriate responses to the evidence we have gathered (Altschul Reference Altschul2014: 260) are specifically doxastic obligations (cf. Chrisman Reference Chrisman2008). Just as we come to fulfill our role as sleepers by putting ourselves in a position to fall asleep, so also we come to fulfill our role as believers by putting ourselves in a position to form appropriate beliefs.
However, Altschul is mistaken to identify the obligations to be the sort of person who holds certain beliefs, or to bring about certain beliefs, as doxastic obligations. For the things we are obligated to do in bringing about certain beliefs, or in being the sorts of people who hold certain beliefs, are not acts of belief formation but rather actions that influence our characters or that enable us to recognize evidence. These are what I have termed aretaic obligations (though some of these might more properly be labeled practical obligations). Further, the aims of the actions we perform in fulfilling the obligations to bring about certain beliefs, or to be the sorts of people who hold certain beliefs, are not directed toward forming certain beliefs but rather at, say, looking up new information, or asking questions of other people, etc. Again, even though these actions have downstream doxastic effects, their most immediate concern is practical or aretaic. Further, the thing that such evaluative obligations evaluate is our characters rather than the beliefs we have or do not have.
In short, the things we are obligated to do (in virtue of the evaluative obligation to be the sort of person who has certain beliefs) and the things we aim at doing in fulfilling such obligations are aretaic obligations and practical actions no different than other commonplace practical obligations and practical actions, such as the obligation to keep a promise, or to tell the truth, which fulfills the obligation to keep the promise.
Altschul is right insofar as he holds that, given the truth of doxastic involuntarism, the only way in which we influence our doxastic states is putting ourselves in better positions eventually to hold appropriate doxastic states. However, Altschul's argument does not go far enough. To summarize: setting prescriptive doxastic obligations aside, the only way in which we have obligations related to our doxastic states is to have obligations to be the sorts of people who have correct beliefs, or to perform actions that place ourselves in situations where we would acquire true beliefs. But such obligations are not best understood as doxastic obligations; rather, they are aretaic obligations or practical obligations.Footnote 1
Open access
