Reasons for Belief in Context

Abstract

There is currently a lively debate about whether there are practical reasons for belief, epistemic reasons for belief, or both. I will argue that the intuitions on all sides can be fully accounted for by applying an independently motivated contextualist semantics for normative terms. Specifically, normative terms must be relativized to a goal. One possible goal is epistemic, such as believing truly and not believing falsely, while another possible goal is practical, such as satisfying desires, or maximizing value. I will argue that we have practical reasons given the practical goal and epistemic reasons given the epistemic goal. Disagreement disappears when we make the context explicit. The result is an independently motivated version of pluralism.

1. Introduction

Consider three views about the nature of reasons for belief:

Anti-Pragmatism

All reasons for belief are epistemic reasons.

Austere Pragmatism

All reasons for belief are practical reasons.

Pluralism

There are practical reasons for belief and epistemic reasons for belief.¹

This paper defends pluralism. Pluralism faces all the challenges to epistemic reasons for belief, plus all the challenges to practical reasons for belief, plus the further challenges of combining epistemic and practical reasons. Nevertheless, I think pluralism is true.

My argument is based on normative contextualism – the thesis that the meaning of normative terms like ‘ought’ and ‘reason’ vary with context. Specifically the thesis, independently supported by work in linguistics, is that context supplies a parameter for a goal and a modal base (or set of propositions).

Section 2 explains the three views about the nature of reasons to believe, two arguments for anti-pragmatism, two arguments for austere pragmatism, and one argument against pluralism. Section 3 introduces the contextualist semantics and section 4 explains how it allows us to reject all the arguments of section 2. Section 5 concludes.

I aim to show that a wide range of problems can be solved by contextualism. I will not offer direct arguments that contextualism is the correct way to deal with the problems. Arguments that contextualism is the correct linguistic theory are to be found in the linguistics literature and I have nothing to add to them. I will argue that contextualism can explain our competing intuitions, and I take this to be a powerful (indirect) argument in favour of contextualism. Compare: contextualism about knowledge is supported by arguments that it can explain our intuitions (that we do not know we have hands in some siutations and that we do know we have hands in other situations).²

Have similar ideas have been suggested before? One important precedent is Richard Feldman's (2000) introduction of the concept of an epistemic ought. This can be understood as ‘ought’ relativized to the epistemic goal. But although I've found Feldman's work very helpful, I find myself disagreeing on important points. First, Feldman holds that the epistemic ought only applies to beliefs and, though he isn't explicit, it would appear Feldman denies that the practical ought applies to beliefs. Neither of these claims follow from contextualism. I allow that the epistemic ought applies to actions (section 4) and that the practical ought also applies to beliefs. Second, Feldman defends evidentialism (roughly, S epistemically ought to have the attitude that is supported by S's evidence), while I remain neutral. Furthermore, Feldman does not connect his view with contextualism.

A quick clarification: I take it ‘what one ought to do’ is synonymous with ‘what one has most reason to do’ and ‘what one should do’ (Berker 2018; Broome 2013; Finlay 2014) so will move freely between them.

2. Background

We can distinguish three mutually exclusive positions. The first is that reasons for belief must be closely connected to truth, justification, or knowledge. Call these epistemic reasons. Defenders of this view often say that inducements to belief such as bribes are the wrong kind of reason to believe.³ The thought that the wrong kind of reasons are not genuine reasons leads to anti-pragmatism:

Anti-Pragmatism

All reasons for belief are epistemic reasons.

Perhaps the paradigm anti-pragmatist view is evidentialism,⁴ which says that only evidence provides reasons to believe.

The second view is the mirror image, which says that all reasons for belief are practical:

Austere Pragmatism

All reasons for belief are practical reasons.

Previously out of fashion, there have been several defences of austere pragmatism in the last few years.⁵ The core idea is that whether you should believe P depends only on whether believing P will contribute to achieving some practical goal. In most cases, having true beliefs will help achieve practical goals, and so you should have true beliefs. But you have reasons to have true beliefs only to the extent that doing so will further your practical goals. If believing falsely is the best way to further your practical goals, you should believe falsely. I'll say more about epistemic and practical reasons and goals in section 3.

The third view, my view, allows both types of reasons⁶:

Pluralism

There are practical reasons for belief and epistemic reasons for belief.

Pluralism has the benefits of both theories, and also the problems of both theories. Furthermore, pluralism faces the charges of being metaphysically profligate and non-unified, and also has problems caused by the interaction of the two types of reason.⁷

To motivate pluralism we can start with a brief rehearsal of arguments for the existence of epistemic and practical reasons for belief. Starting with epistemic reasons for belief, suppose I overhear someone revealing how a film ends before watching it. Plausibly, I now have good reason to form a belief about how the film ends, despite the fact that I do not want to know how it ends. My practical goals will be frustrated by forming a belief about the ending, but it seems I have good reason to do so nonetheless.⁸ This is an argument for the existence of epistemic reasons, and therefore against austere pragmatism.

Arguments for the existence of practical reasons for belief date back at least to Pascal's argument that we should believe in God because, if God exists, disbelievers will be sent to hell.⁹ More recent arguments concern reasons not to believe that your friend behaved unethically (even if there is evidence that they did),¹⁰ reasons not to believe that a member of an ethnic minority is more likely to have committed a crime (even if there is statistical evidence that they are)¹¹ and reasons not to believe something when the stakes for being wrong are high.¹² Arguments for the existence of such reasons are arguments against anti-pragmatism. I can endorse all these arguments for the existence of practical reasons and epistemic reasons to believe.

Let's get on the table five arguments that I need to reject. Start with two arguments for anti-pragmatism (and so against practical reasons for belief).

2.1. Truth-based argument for anti-pragmatism

The anti-pragmatist might take the connection between belief and truth to make their case. Specifically, consider:

(i) Truth is the sole norm of belief i.e. a belief is correct iff it is true.

(ii) A reason to believe that P explains why the belief that P is likely to be correct.

(C1) Therefore: a reason to believe that P explains why the belief that P is likely to be true.

(iii) Only evidence explains why the belief that P is likely to be true.

(C2) Therefore: only evidence can be a reason to believe P.¹³

The most controversial of the three premises (and the one I will deny in section 4.1) is (i). Nevertheless, the thesis that a belief is correct iff it is true is highly plausible and widely held. Jim Joyce (1998) writes:

The relevant success criterion for full beliefs is well-known and uncontroversial.

The Norm of Truth (NT):

An epistemically rational agent must strive to hold a system of full beliefs that strikes the best attainable overall balance between the epistemic good of fully believing truths and the epistemic evil of fully believing falsehoods (577).

Alvin Goldman (1999, 2001, 2015) has defended something similar for many years, calling it ‘veritism’¹⁴:

Veritism

Our only ultimate epistemic goals are

(i) the accumulation of true beliefs and

(ii) the avoidance of false beliefs.

2.2. Deliberation-based argument for anti-pragmatism

Shah (2006) argues against the existence of any practical reasons for belief based on the claim that when we deliberate about whether to believe p, we only consider whether p is true. Any thought that it is beneficial to believe p does not seem to be capable of guiding our deliberation. Add a principle such as ‘if a consideration cannot guide an agent's deliberation about whether to believe p then it cannot be a reason to believe p’, and anti-pragmatism follows. The argument can be reconstructed as (with the names Shah uses for closely related theses in parentheses)¹⁵:

(i) If a consideration is a reason to believe then it is capable of guiding deliberation (deliberative constraint).

(ii) Only evidence is capable of guiding deliberation (transparency).

(C) Therefore only evidence provides a reason to believe.

The conclusion entails anti-pragmatism (i.e. all reasons for belief are epistemic reasons). Specifically, the conclusion entails the version of anti-pragmatism that says that all reasons for belief are constituted by evidence (other versions of anti-pragmatism might connect reasons for belief with epistemic justification, truth, or knowledge).

2.3. Simplicity argument for austere pragmatism

Let's move on to two arguments for austere pragmatism. Rinard (2017) suggests that positing epistemic reasons for belief would be metaphysically profligate¹⁶:

Most philosophers agree that we should aim for simplicity and elegance in our theorizing. These considerations tell in favor of [Austere Pragmatism]. After all, [Austere Pragmatism] demands a theory of rationality on which all states are treated alike. Any version of Exceptionalism [e.g. Pluralism], on the other hand, presents us with a fractured and thereby more complicated theory of rationality (123–24).

The idea is that it is widely agreed that there are practical reasons e.g. for action. Given that one is already committed to practical reasons, it is simpler to use practical reasons to explain reasons to believe as opposed to positing some new type of reason. The result is austere pragmatism.

2.4. Univocality argument for austere pragmatism

A related, but separable argument is that rationality should speak with one voice. That is, there is always a single answer to the questions: what should I do? Or, what should I believe? And only austere pragmatism delivers this result.¹⁷ Rinard writes:

According to [Austere Pragmatism] rationality speaks in a single voice. Insofar as we look to rationality for guidance, and conceive of it as an ideal to which one might coherently aspire, this tells in favor of [Austere Pragmatism]… [Austere Pragmatism] is a more unified approach to rationality: it treats all states, including belief states, alike, whereas [alternative theories] postulate special complexities in the case of belief (2017: 128).

Thus the argument seems to be that rationality provides guidance, providing guidance requires univocality, and univocality requires austere pragmatism.

2.5. Combinatorial argument against pluralism

Finally, there is an argument against pluralism that arises from the interaction of epistemic and practical reasons. Berker (2018) argues that purported practical reasons for belief exhibit permissive balancing, whereas epistemic reasons for belief exhibit prohibitive balancing.

Suppose I have a strong practical reason to perform action A, an equally strong practical reason to perform alternative action B, and no other practical reasons that bear on the matter. Then I have sufficient practical reason to perform A, have sufficient practical reason to perform B, and lack sufficient practical reason to do anything else. In other words, A and B are both permitted by the overall balance of my practical reasons…

Epistemic reasons for belief, on the other hand, exhibit a phenomenon I call prohibitive balancing. Suppose I have a strong epistemic reason to believe P, an equally strong epistemic reason to disbelieve P, and no other epistemic reasons that bear on the matter. Then I lack sufficient epistemic reason to believe P, lack sufficient epistemic reason to disbelieve P, and have decisive epistemic reason to suspend judgment on P. In other words, rather than being permitted by the overall balance of my epistemic reasons to believe P or to disbelieve P, neither option is epistemically permissible (430–1).

Berker considers and rejects various ways of allowing practical and epistemic reasons to interact, and argues that none succeed. Berker accepts that there are epistemic reasons for belief, and so concludes that there are no practical reasons for belief (anti-pragmatism).

Obviously this does not exhaust the arguments against pluralism. But I take them to be a representative sample of the most influential arguments that pluralists need to engage with. Next I will introduce the semantic machinery that I will use to defend pluralism.

3. Contextualism

It is a familiar thought that whether someone is correctly described as tall depends on the details of the conversation. For example, Michael Jordan, at 1.98 m, is tall for an ordinary person, but not tall for a basketball player. So the truth of ‘Michael Jordan is tall’ depends on the conversational context. It is true given a context in which ordinary people are being discussed, but false given a context in which basketball players are being discussed.

A popular theory in linguistics is that normative terms like ‘ought’ are context-sensitive in a similar way.¹⁸ They have hidden parameters that are fixed by the conversational context, resulting in different semantic values in different contexts. In fact it is plausible that there are at least two parameters needed to fix the content of ‘ought’ – a standard and a set of live possibilities. In this section I will explain the view, and also separate the core commitments from stronger positions we need not be committed to.

3.1. Propositions/possible worlds

The first parameter is a modal base that determines a proposition or set of live possible worlds. The live worlds are those compatible with the modal base. If the modal base is empty then all worlds are live. As the modal base grows, the set of live worlds is restricted. On the standard theory of modals (Kratzer 1981), ‘it must be that p’ means, roughly, that in all the live worlds, p.

The modal base need not correspond to the beliefs of the subject of the sentence. We can allow that the modal base consists of the beliefs of the subject of the sentence, or the speaker, or some third party or the collective beliefs of a group, or the propositions known by any of the former, or any of these plus a number of fixed propositions, and endless further options. This parameter is often called the ‘information set’, but using information here is too restrictive as information is naturally taken to imply truth. We can allow that the modal base includes falsehoods.

I mention the modal base because it is part of the standard semantics, but it will play no role in my argument. We can remain neutral on whether the modal base varies with context.

3.2. Standard/goal

The second parameter is a standard or goal that determines an ordering of the live possible worlds. The standard proposal is that ‘S ought to A’ is true iff S A's in every live world are at the top of the ranking. Although the semantics literature focuses on ‘ought’ I take contextualism about all normative terms to follow naturally. For example, if what you ought to do is relative to a goal, it is natural to say that what reasons you have are relative to a goal.¹⁹

The standard need not be one that the subject cares about. If I say ‘you ought to start with the cutlery on the outer edge’, the standard might be the rules of etiquette. The more explicit sentence is ‘by standards of etiquette, you ought to start with the cutlery on the outer edge’. This sentence can remain true even if you don't care about etiquette. This allows us to say to the psychopath ‘you shouldn't kill people’; the full sentence is ‘by standards of morality, you shouldn't kill people’, which is true even if the psychopath doesn't care about morality.

For our purposes we only need to distinguish two standards: those corresponding to the epistemic ought and the practical ought. The epistemic ought determines a ranking of worlds based on which is epistemically best. We can get a grip on the epistemic ought by thinking about contexts where the conversation concerns some epistemic standard such as having true beliefs. Typical sentences might be ‘you ought to be uncertain’, or ‘we ought to expect defeat’. Again, the standard need not be one that the subject cares about, so we need not assume that agents care about any epistemic goals.

There is disagreement about what the epistemic standard is. Leading contenders for epistemic standards include having beliefs that are (a) true, (b) justified, and (c) knowledge. The differences between these positions won't matter here, so we remain neutral. And we can remain neutral on whether the standard (e.g. truth) is constitutive of belief or whether something can be a belief without having such a standard.

This brings us to the practical ought. The practical ought determines a ranking of worlds based on which is practically best. We can get a grip on the practical ought by thinking about contexts where the conversation concerns what is best to do. Typical sentences might be ‘you ought to stay in school’ or ‘should I boil or steam the vegetables?’.

There is also disagreement about what the practical standard is. Humeans might hold that the practical standard is grounded in one's desires. Non-Humeans might hold that the practical standard is grounded in objective value. There are further debates about whether the practical standard is to maximize actual value or expected value, and whether expected value is determined by beliefs or evidence. We can remain neutral on these issues.

We can leave open whether there are other standards. For example, one might argue that there is a moral standard, according to which holding beliefs can be morally good or bad. Positing a moral standard for belief seems to me at least as plausible as positing a practical standard for belief, so I will use moral standards occasionally in examples below. (Notice that we could incorporate a moral standard into a practical standard e.g. if agents care about morality then a Humean practical standard could include a moral standard.)

Someone might object that standards could be multiplied arbitrarily. For example, perhaps there is an aesthetic standard giving aesthetic reasons for belief. And perhaps there are similar standards for the coolest belief, or the zaniest belief. The objection would be that the contextualist has no way to block this implausible explosion of standards.²⁰

In response, I agree that the semantics allows arbitrary standards, but it doesn't follow that the English word ‘ought’ can be relativized to arbitrary standards. Of course, one could easily introduce such a standard, but the resulting ‘ought’ would be a new word, not a word of our language.

We can also remain neutral on whether there are further parameters that determine the content beyond standards and propositions. For example, Carr (2015) argues that ‘ought’ must be relativized to a decision rule. This may be so, but it will not play a role below.

We need not be committed to any further details of Kratzer's theory. Indeed, contextualism is not the only way to generate the semantic variation I will use. One alternative is a polysemic theory according to which ‘ought’ has multiple meanings.²¹ I think contextualism is more plausible given the wide range of meanings that would need to be posited, but my arguments do not depend on using contextualism. I use contextualism for concreteness, but if you have a preferred alternative semantic theory you should be able to plug it into my arguments.

4. A contextualist defence of pluralism

Once we posit different goals we are naturally led to:

Pluralism

There are practical reasons for belief and epistemic reasons for belief.

Practical reasons are associated with practical goals and epistemic reasons are associated with epistemic goals. There are practical reasons to do anything that promotes a practical goal and there are epistemic reasons to do anything that promotes an epistemic goal.

This leads to the rejection of the view that there are only epistemic reasons to have beliefs and only practical reasons for actions. Instead, there are epistemic reasons to have true beliefs and also epistemic reasons to take actions that make it more likely that you'll have true beliefs e.g. performing experiments.²² Likewise, there are practical reasons to have beliefs if having those beliefs will promote your goals. For example, if having positive beliefs will improve your mood and motivate you to achieve your goals, you have practical reason to have positive beliefs.

Let's return to the arguments of section 2.

4.1 Truth-based argument for anti-pragmatism

Recall the argument for anti-pragmatism based on the connection between belief and truth:

(i) Truth is the sole norm of belief i.e. a belief is correct iff it is true.

(ii) A reason to believe that P explains why the belief that P is likely to be correct.

(iii) Therefore: a reason to believe that P explains why the belief that P is likely to be true.

(iv) Only evidence explains why the belief that P is likely to be true.

Therefore: only evidence can be a reason to believe P.

But contextualists will happily reject that (i) truth is the sole norm of belief. What is correct/best depends on the goals in the context. In a context where some practical goal is at stake, we can assess how well believing that P promotes the practical goal. Such practical goals explain the arguments for the existence of practical reasons to believe that we saw in section 2. Some argue that we have practical reasons to believe in God.²³ Some argue that we have moral reasons not to have racist beliefs.²⁴ Some argue that that we have moral reasons to have positive beliefs about our friends.²⁵ Some argue that high stakes provide reasons not to believe.²⁶

Contextualism predicts and explains the conflict between the truth norm and these arguments. It is plausible that truth is the sole norm of belief given the epistemic goal.²⁷ But if we open things up and allow other goals then we can reject the claim that truth is the sole norm of belief. (An objector might bang the table and ask ‘Is the belief correct?’. But the contextualist will reject the question, holding that beliefs are correct or incorrect only relative to goals.)

4.2. Deliberation-based argument for anti-pragmatism

Recall Shah's (2006) argument against practical reasons:

(1) If a consideration is a reason to believe then it can guide deliberation (deliberative constraint).

(2) Only evidence can guide deliberation (transparency).

(3) Therefore only evidence provides a reason to believe.

But the deliberative constraint is not compelling. Shah gives the following argument for it:

(4) If a consideration is a reason to believe then it is capable of being a reason for which one believes.

(5) If a consideration is a reason for which one believes then it can guide deliberation

Therefore

(1) If a consideration is a reason to believe then it can guide deliberation (deliberative constraint).

Shah's defence of (4) is based on Williams' (1979) influential argument that ‘the function of reasons is to guide agents’ (Shah: 486). Very roughly, Williams argues that if an agent has no motivation to join the army and is not capable of being guided by purported reasons in favour joining the army, then the agent has no reasons to join the army.

But Williams' claim is highly controversial, and there are numerous purported counter-examples involving reasons that cannot guide agents. One counter-example is that of psychopaths who are not moved by moral reasons, but of whom we are inclined to say that they have moral reasons not to kill people.

Another counter-example in which there seem to be reasons that cannot guide is the problem of surprise parties²⁸:

Nate hates most parties, but not all of them – he thoroughly enjoys successful surprise parties held in his honour. So let us imagine that a surprise party is waiting for Nate at his home, and suppose also that he is unaware of it. (Risberg 2016: 104)

The fact that there is a surprise party waiting for him at home is a reason for Nate to go home, but this reason cannot guide him. If the fact that there were a party at home were guiding Nate's deliberation then the surprise would be ruined and Nate would not have this reason to go home.

Another type of counter-example comes from the truth norm e.g. a belief is correct iff it is true.²⁹ Assuming a plausible connection between what is correct and what we have reason to believe, it follows that the fact that some proposition p is true is a reason to believe it.³⁰ And, arguably, such reasons cannot guide, as in order to be guided I must first form a belief as to whether p; but that was the very question I wanted guidance on! Once I have formed a belief as to whether p, I simply do not need such guidance anymore. Furthermore, the only guidance I would receive would be to continue believing, as my belief that p implies that I think that p is true.

What does all this have to do with contextualism? Contextualism can explain the intuitive appeal of Williams' view that the function of reasons is to guide agents, as well as the counter-examples. Specifically, reasons guide agents in some contexts. In fact, we can specify what the modal base and goal parameters must be in order for reasons to be able to guide agents.

First, the modal base must contain no more than the beliefs of the agent. Thus, if an agent has not determined whether it is true that p, the modal base cannot determine whether it is true that p. So if the agent has not formed a belief about whether p (is true) then the modal base cannot include the claim that p (is true). Similarly, if Nate does not believe that there is a surprise party waiting for him then the modal base cannot include the proposition that there is a surprise party waiting for him. Second, the goal must be a goal of the agent. If a psychopath does not have the goal of behaving morally then they cannot be guided by moral reasons.

The counter-examples above are all cases where the parameters being invoked do not match the goals and beliefs of the agents. However, in common uses of reasons talk, the parameters invoked do match the goals and beliefs of the agent, and so the reasons can guide the agent. We rarely engage in moral discussions with psychopaths. And if we were to realize that this is what we were doing we would stop, as we would recognize that our moral goals are not shared by the psychopath.

The upshot is that contextualism explains why the deliberation-based argument for anti-pragmatism is tempting, but why it ultimately fails. It fails because it is too strong to say that the function of reasons is to guide agents. Where speakers invoke parameters of modal base or goal that do not match the agent's beliefs and goals, there are reasons that cannot guide the agent. But in typical conversations where reasons talk is used – about whether we have good reasons to be vegetarian, or study for exams, or get some exercise – the parameters invoked in the conversation do match the beliefs and goals of the agent, and the reasons will be able to guide.

4.3 Simplicity argument for austere pragmatism

Recall Rinard (2017) argues that epistemic reasons for belief would be metaphysically profligate. I think the best response is to agree with the letter of Rinard's objection, but downplay its significance. That is, the pluralist does end up with a more complicated normative theory than the Austere Pragmatist, but the complication is not troubling.³¹

The first thing to say is that parsimony is never a fatal objection by itself. Everything should be made as simple as possible, but not simpler. If a more complex theory is needed to explain the phenomena, so be it. My position is that the more complicated pluralist theory is needed, and the extra complications are anyway not significant.

Let's first get clear on the way in which pluralism is complicated. Distinguish ontological and ideological parsimony (Quine 1948; Sider 2013). Ontological parsimony concerns parsimony with respect to things posited by a theory, while ideological parsimony concerns parsimony with respect to concepts, or primitive concepts, employed by a theory.

Start with ontological parsimony. Is pluralism ontologically complex? I suggest not. Pluralism posits various standards or goals, but surely everyone posits such standards or goals. For example, even the austere pragmatist will posit a standard of etiquette. The difference is that the austere pragmatist will not use this standard in their normative theory. So perhaps the austere pragmatist will claim the virtue of having an ontologically simpler normative theory. But this does not seem to be a significant virtue. The overall worldview of the austere pragmatist and the pluralist is the same – both are committed to various standards. The pluralist is using tools that are already available, and this does not seem to be metaphysically costly.

Is pluralism ideologically complex? Pluralists certainly have more complex normative concepts i.e. concepts with a hidden parameter. But this extra complexity is independently supported by work in semantics. It would not be very convincing for the philosopher to say to the linguist ‘no thank you, your concepts are too complicated, I prefer my own simpler concepts’.

It is worth comparing contextualism about knowledge. The contextualist about knowledge says that ‘S knows P’ has a hidden parameter, such as a set of relevant alternatives. So the sentence is more fully expressed as ‘S knows P given alternatives Q’. This is of course a more complex theory, but it is noteworthy that few philosophers have criticized contextualism for being either ontologically or ideologically profligate. I suspect that this is for the reasons just discussed. The contextualist is using an ontology that the non-contextualist should admit exists e.g. alternatives; and the more complex ideology is independently motivated by semantics.

The extra complexity would be troubling if these plural normative goals came with some heavy-duty metaphysics attached. One form of heavy-duty metaphysics says that normative goals necessarily motivate us. Another, related, form of heavy-duty metaphysics is non-naturalism, which might leave us committed to one form of non-natural facts corresponding to practical reasons, and another form of non-natural facts corresponding to epistemic reasons. But neither type of metaphysics is needed here.

In fact, the pluralist could well make a case that they have a pleasingly simple theory, as they offer a reductive naturalistic normative theory in terms of goals and modal bases (Finlay 2014), promising to solve what Frank Jackson (1998) called the location problem – the problem of finding a place to put troublesome ideas such as ‘ought’ and ‘reason’ in a physical world. If we cannot solve this problem we might have to posit normativity as a fundamental part of the world; in offering a reductive theory in terms of goals, we retain a simpler naturalistic theory of the world. But I should stress that contextualism does not entail naturalism. As a semantic theory, contextualism is compatible with any metaphysics.

This is a good place to point out that pluralism is neutral on whether there is a metaphysically privileged sense of ‘ought’ that has authority. Alex Worsnip writes:

we should be careful to separate the question of whether (e.g.) the law …has genuine normative authority from whether there is a robustly normative usage of the legal ‘ought’. The former requires the law to actually possess normative authority, whereas the latter only requires there to be speakers who take the law to possess normative authority. So even if only a handful of the above ‘oughts’ reflect a genuine source of normativity, many more of them might nevertheless be robustly normative usages of ‘ought’. (Worsnip 2019: 3103–4; see also Worsnip 2020)

If there is a metaphysically privileged sense of ‘ought’ that has authority then the anti-pragmatism vs pragmatism debate can return as a debate about which ‘ought’ is authoritative. I offer no arguments against this here. But positing a metaphysically privileged sense of ‘ought’ brings with it familiar metaphysical (what makes it privileged?) and epistemic problems (how do we know it is privileged?).³² The contextualist need have no such problems, and can explain that it is natural to think that there is a metaphysically privileged sense of ‘ought’ because the parameters are usually implicit.³³

4.4. Univocality argument for austere pragmatism

The univocality argument is that rationality should speak with one voice. Rinard motivated univocality by emphasizing that we look to rationality for guidance, which suggests that there is always a single answer to the questions: what should I do? Or, what should I believe?

Here the pluralist can simply deny that rationality always provides guidance. Consider cases where there is a conflict between moral and prudential reasons.

A trolley is on its way to kill 5 innocent people. You can save their lives by jumping in front of the trolley and sacrificing yourself.

Should you sacrifice yourself?

Is there a fact of the matter about whether you should sacrifice yourself? I have no inclination to say yes. I can just about see that ethics might provide an answer in terms of most-ethical-thing-to-do, but not that rationality will provide an answer in terms of overall-most-rational-thing-to-do. So it seems that there are cases where rationality does not provide guidance.

Furthermore, pluralism can explain the sense of irresolvability we sometimes feel when faced with such choices. There might be a fact about what we ought to do given some particular practical goal, and there might be a fact about what we ought to do given the ethical goal, but there is no fact about what we ought to do simpliciter.

4.5. Combinatorial argument against pluralism

Berker (2018) argues that purported practical reasons for belief exhibit permissive balancing, whereas epistemic reasons for belief exhibit prohibitive balancing, and that there is no acceptable way to combine practical and epistemic reasons. Berker's response is to reject practical reasons for belief.

But there is only a problem if we expect there to be a way to combine practical and epistemic reasons. Pluralism says that there are practical reasons and epistemic reasons, but does not say that there is a way to combine them. In fact, the contextualist theory suggests that there is no way to combine them, as the normative words have different meanings (content). There is the question of what one ought-given-the-practical-goal to believe, and the question of what one ought-given-the-epistemic-goal to believe, and there is no presumption that there will be a question of what one ought-overall to believe. As noted in section 4.4, the contextualist can posit such an ‘ought’, but has little motivation to do so.

Again, compare contextualism about knowledge. The contextualist says that we know-in-a-normal-context that we have hands, and that we don't know-in-a-sceptical-context that we have hands, but there is no presumption that there will be a question of what one knows-simpliciter. The contextualist can reject that question as ill-formed. Similarly, the normative contextualist can explain why it is so difficult to combine epistemic and practical reasons – it is because the question of what one ought to believe simpliciter is ill-formed.

5. Conclusion

I have argued that there are practical reasons for belief and epistemic reasons for belief. Pluralism is sometimes seen as an easy option for the soft-headed who want to agree with everyone, and who overlook the inelegance of the resulting theory. I have tried to avoid this accusation by arguing from an independently motivated semantics. Furthermore, many of the arguments in the literature are against the existence of either practical or epistemic reasons, so pluralists have lots of work to do. I have tried to show that once we introduce contextualism, the arguments against practical reasons and epistemic reasons are not convincing, and that the criticisms specific to pluralism have little bite given the contextualist semantics.³⁴