Normative reasons and motivational capacities

Razões normativas e capacidades motivacionais

ABSTRACT

A very influential idea on the nature of normative reasons is that the existence of normative reasons for action depends on the motivational capacity of the agents whose reasons they are: there are reasons for an agent to act only if she has the capacity to be moved to act for those reasons. Many theories of reasons developed in recent years imply at least some version of that idea, and many find it attractive since it incorporates some widespread assumptions about the role of reasons. In this paper, I argue that that idea, understood in any of its prominent readings, is ultimately false. First, I endorse the claim that there is at least one case that shows the possibility of the so-called ‘elusive reasons’, that is, roughly, reasons whose existence depends on the agent’s ignorance of the facts that constitute them. Then, I argue that all the crucial objections recently elaborated to the possibility of elusive reasons ultimately fail. The upshot is that we should not aim to explain the nature of normative reasons in terms of motivational capacities of the agents whose reasons they are.

Keywords:
normative reasons; elusive reasons; internalism about reasons; deliberative constraint; ‘ought’

RESUMO

Uma ideia muito influente sobre a natureza das razões normativas é a de que a existência de razões normativas depende da capacidade motivacional dos agentes para os quais elas são razões: existem razões para uma agente agir somente se ela tem capacidade de ser movida a agir por essas razões. Muitas teorias das razões desenvolvidas recentemente implicam pelo menos alguma versão dessa ideia, e muitos a consideram atrativa visto que ela incorpora algumas assunções amplamente difundidas acerca da função das razões. Neste artigo, eu argumento que essa ideia, compreendida em quaisquer das suas leituras proeminentes, é ao fim e ao cabo falsa. Primeiro, eu endosso a tese de que há pelo menos um caso que mostra a possibilidade das denominadas ‘razões elusivas’, isto é, grosso modo, razões cuja existência depende da ignorância do agente acerca dos fatos que as constituem. E então, eu argumento que todas as importantes objeções recentemente elaboradas contra a possibilidade de razões elusivas ao fim e ao cabo fracassam. O resultado é que nós não devemos buscar explicar a natureza das razões normativas em termos de capacidades motivacionais dos agentes para os quais elas são razões.

Palavras-chave:
razões normativas; razões elusivas; internalismo de razões; restrição deliberativa; ‘dever’

1 Introduction

In recent literature on the nature of normative reasons, many philosophers argue or assume that the existence of normative reasons for an agent to act constitutively depends on the agent’s capacity to be moved to act for those reasons. This idea can be called Constraint. For, it can be conceived as a constraint on normative reasons for action: they are facts that count in favor of a particular action, but only those for which the agents whose reasons they are can somehow be moved to act¹. We can understand the Constraint in many ways, depending on how we understand the notion of ‘capacity’ at issue and the notion of ‘acting for a reason’. Still, the Constraint is an important claim in any of its prominent readings. It accommodates some widespread ideas about the role of normative reasons - namely, that normative reasons for action are essentially action-guiding and that normative reasons are potential premises in good reasoning. Moreover, some popular views about reasons entail at least some version of it - such as what has been called ‘Internalism about Reasons’, and the ‘Deliberative Constraint’ on reasons. Since those popular views are the heart of many theories about reasons in recent literature, it is not hard to see how Constraint’s plausibility is crucial to the current debate on the nature of normative reasons.

My main aim in this paper is to argue that Constraint, understood in any of its prominent readings, is ultimately false. I proceed as follows. In the next section, I elucidate the Constraint and delimit its prominent readings. In section 2, I highlight the theoretical importance of the plausibility of the prominent versions of the Constraint. In section 3, I argue against the Constraint by claiming the possibility of so-called ‘elusive reasons’. This particular argument echoes the intuition shared especially by Schroeder (2007), Markovits (2014), and Rossi (2021a). However, I take some important steps forward. In section 4, I argue that all the recent crucial objections raised to the possibility of elusive reasons ultimately fail. The upshot is that we should not aim to explain the nature of normative reasons in terms of motivational capacities of the agents whose reasons they are.

2 The Constraint

What is Constraint supposed to mean?

First, Constraint is a claim just about the nature of reasons for action. So, it is not a claim about reasons for other responses - such as intention, feelings, and beliefs. Second, Constraint is not about every kind of reasons for action. Constraint is only concerned with normative reasons for action and so not merely with motivating or explanatory reasons. Roughly, as many take them generally to be, normative reasons are facts, or truth propositions, that count in favor of or justify someone’s action. In contrast, motivating reasons are taken to be considerations that count in favor of someone’s action, but only subjectively so, and explanatory reasons are taken to be considerations that merely explain someone’s action (Alvarez, 2017). Third, since normative reasons for action are relational properties, constituted by a relation between a fact, an agent, and an action, we can understand Constraint as a claim about the relation that must be obtained between these relata. Finally, Constraint is a claim about the necessary conditions for that relation to be obtained. So, Constraint is not a complete explanation of the nature of normative reasons for action - even if it might constitute part of such explanation. So, we can formulate Constraint as follows (where ‘p’ stands in for a fact, ‘A’ for an agent, ‘ϕ’ for a verb of action, and ‘reason’ means a normative reason for action):

Constraint: p is a reason for A to ϕ only if A has the capacity to be moved to ϕ for the reason that p.

In other words, according to Constraint, the existence of normative reasons for an agent to act depends on her capacity to be moved to act for those reasons².

Beyond those preliminary remarks, to understand more accurately the plausible meanings of that claim, we should delimitate the relevant notion of the ‘capacity’ at issue and how we should understand the notion of ‘acting for a reason’.

The capacity at issue is motivational: a capacity to form a motivation to act in light of some reason. We can understand the modal notion of motivational capacities to act for reasons in two ways. An agent might have an actual or a counterfactual motivational capacity to act for reasons (Paakkunainen, 2018). On the one hand, an actual motivational capacity is, plausibly, a capacity that an agent has that must be capable of being manifested in this actual world. It is not ‘actual’ because her current mental states completely constraint it, but rather because she could manifest it in this actual world - that is, a lot of nearby subjunctive conditionals are true of her (Paakkunainen, 2018, pp. 153-154). So, when an agent has an actual capacity to be moved to act for a reason, there are nearby worlds in which she, being aware of that reason, would be moved to act for that reason.

On the other hand, we can also have a counterfactual motivational capacity. We can conceive it as a capacity to be moved to act for a reason that can be manifested in some possible world, constrained only conceptually and within the boundaries of agency and rationality. This capacity might involve some very different scenarios of the world and the agent’s psychology, which can be drastically different from the present ones. So, when an agent has a counterfactual capacity to be moved to act for a reason, there must be some possible world in which she, being aware of the reason, would be moved to act for that reason.

I suggest, then, that we can understand Constraint in two general ways:

Constraint - Actual Capacity: p is a reason for A to ϕ only if there is a nearby world in which A would be moved to ϕ for the reason that p.

Constraint - Counterfactual Capacity: p is a reason for A to ϕ only if there is a possible world in which A would be moved to ϕ for the reason that p.

Note that Constraint - Actual Capacity implies Constraint - Counterfactual Capacity. If the existence of reasons depends on what agents would be moved to do in this actual world, the existence of reasons depends on what agents would be moved to do in some possible world. However, the inverse surely does not hold.

Now, on the notion of acting for a reason. That notion has two standard readings: the Belief View and the Normative Recognition View³. We can put them as follows:

Belief View: A ϕ’s for the reason that p only if (i) p, (ii) p is a reason for A to ϕ, (iii) A believes that p, and (iv) A believing that p causes in a non-deviant way A to ϕ.

Normative Recognition View: A ϕ’s for the reason that p only if (i) p, (ii) p is a reason for A to ϕ, (iii) A regards p as a reason to F, and (iv) A regarding p as a reason to ϕ causes in a non-deviant way A to ϕ.

Conditions (i) and (ii) in both views highlight that when someone acts for a reason, the thing she acts for must be a normative reason. Condition (iii) states the relevant epistemic relation between the agent and the reason - a belief relation in Belief View and the recognition of the normative relation in Normative Recognition View. Condition (iv) emphasizes that when an agent acts for a reason, the relevant epistemic relation must be causally connected with the action in a non-deviant way in order to rule out actions performed by a deviant process of motivation as actions that are done for a normative reason (Sinclair, 2016, §V).

Belief View is the most basic reading. This is intuitive: when someone acts for a reason, he acts at least being aware of that reason-constituting fact, and it is plausible that such an awareness involves a belief. But Normative Recognition View is also a plausible reading. For, acting for a normative reason can involve the recognition of the normativity of the reason-constituting fact and not only a mere awareness of the reason-constituting fact itself. Nevertheless, there is significant dispute about what it is to regard a fact as a reason, and it is unclear what it should mean. Perhaps for an agent to regard p as a reason to ϕ she must believe that p is a reason for her to ϕ, or use p in good practical reasoning towards ϕ-ing, or he must be in a position to manifest competence to use p as a reason to ϕ, etc.⁴ However, we might set aside that discussion here - we better take the essential reading. After all, Normative Recognition View implies Belief View. For, if you do something regarding a fact as a reason to do that, you must, at least, believe the fact in the first place. After all, we can only recognize the normativity of a particular consideration we take to be true. So, in the following, I shall take as central to our discussion the Belief View on the notion of acting for a reason.

These are the most prominent readings of the Constraint. In particular, though, I will argue against its more fundamental and attenuated reading, which conjugates the Counterfactual Capacity and the Belief View readings. This version of Constraint can be put as follows:

Constraint*: p is a reason for A to ϕ only if there is a possible world in which A would be moved to ϕ by the belief that p.⁵

Since Constraint - Actual Capacity implies Constraint - Counterfactual Capacity, and Constraint read in terms of the Normative Recognition View implies Constraint read in terms of the Belief View, the argument against Constraint* can work against all those readings of Constraint.

3 The importance of Constraint

Constraint is important in each of its prominent readings because it is intimately connected to popular views about the nature of reasons. First, note that the popular view, sometimes called ‘Internalism about Reasons’, is a version of Constraint. Consider Setiya’s (2012, p. 4) formulation of that view:

Internalism about Reasons: the fact that p is a reason for A to ϕ only if A is capable of being moved to ϕ by the belief that p.⁶

Internalism about Reasons just is Constraint formulated in terms of the Actual Motivational Capacity reading and the Belief View (Paakkunainen, 2018, pp. 153-155). So, if Internalism about Reasons is true, Constraint* must also be true.

Moreover, Constraint is also important for a popular view that indexes reasons with sound deliberations of the agents whose reasons they are⁷. That view is now usually called ‘Deliberative Constraint’, which we can put as follows:

Deliberative Constraint: p is a reason for A to ϕ only if there is a sound deliberative route such that, if A underwent it, A would thereby be moved to ϕ because p.⁸

Deliberative Constraint implies Constraint*. For, if the existence of reasons for an agent to act depends on some sound deliberation that she is capable of undergoing with those reasons towards a motivation to act in light of them, the existence of reasons for her to act depends on her capacity to be moved to act for those reasons at least in some possible world. So, the plausibility of Constraint is indeed crucial for some popular views about the nature of reasons.

In addition, Constraint is relevant because it accommodates some alleged general ideas about the role of reasons. First, it incorporates the idea that reasons are action-guiding (Way & Whiting, 2016, p. 214). For instance, the idea is that if there are reasons, say, to quit smoking, there must be facts capable of guiding one to quit smoking. Constraint incorporates and supports that widespread idea. For, if there can only be reasons for an agent to act if she is somehow capable of being moved to act for those reasons, then reasons are essentially action-guiding.

Moreover, Constraint also incorporates the idea that reasons are premises in good reasoning. For instance, if there are reasons to quit smoking, those reasons must be able to constitute a good piece of reasoning towards some motivation to quit smoking - e.g., ‘Since I want to be healthy, and smoking is bad for health, I shall quit smoking’. Constraint provides a general ground for that idea because if the existence of reasons depends on the agent’s capacity to be moved to act for those reasons, and in a non-deviant way, then there must be a piece of good reasoning from the reason that p to a motivation to ϕ (e.g., if p, then I shall ϕ).

Nevertheless, despite its popularity and its alleged attractions, I will, for now on, argue that Constraint is false in any of its prominent readings. For, there might be elusive reasons.

4 Elusive Reasons

The argument from the possibility of elusive reasons against some version of Constraint, put forward especially by Schroeder (2007, p. 33, 165 - 166), Markovits (2014, ch. 2), and Rossi (2021a; 2021b), is that if there can be elusive reasons, at least some version of Constraint must be false. I shall argue in this section that the existence of elusive reasons gives us reasons to abandon Constraint* - and, because of that, any prominent version of it. But what are elusive reasons?

Roughly, elusive reasons are taken to be reasons whose existence depends on the agent’s ignorance of the facts that constitute them. More precisely, elusive reasons are reasons constituted by blindspot propositions, that is, propositions that an agent cannot believe without changing their truth value (Rossi, 2021a, pp. 82-83). So, we can formulate it as follows:

Elusive Reasons: p is an elusive reason for A to ϕ iff there is no possible world in which p is a reason for A to ϕ, and A believes that p at the same time that p is true.

But why should we suppose that if there can be elusive reasons, Constraint* must be false? Well, because Constraint* entails that there must be some possible world in which the agent is moved to act by believing the reason-constituting fact. If there is no possible world in which the agent believes the fact that constitutes her reason, there is no possible world in which she is moved to act by that true belief. So, if there might be elusive reasons, there might be some fact, p, which is a reason for A to ϕ, even though there is no possible world in which that fact is a reason for A to ϕ and A is moved to ϕ by the belief that p.

A case that intuitively shows us the possibility of elusive reasons is Mark Schroeder’s now-famous Surprise Party case, which we can put as follows:

Surprise Party. Nate hates all the parties except for successful surprise parties thrown in his honor, which he loves. Given Nate’s situation, the fact that there is a surprise party waiting for him now at his home is a reason for him to go home. But it is not a reason Nate could know about or act on. (Schroeder, 2007, p. 33).

The powerful intuition that underlies Surprise Party is that the fact that there is a surprise party waiting for Nate at his home is a reason for Nate to go home. After all, given Nate’s subjective scenario, the fact that a surprise party is waiting for him at his home intuitively counts in favor of Nate going home. But - and that is the main point - Nate cannot be moved to go home by believing the reason-constituting fact while that fact constitutes a reason for him to do so. After all, there is no possible world in which Nate is moved to go home by believing the fact that constitutes his reason. In any possible world, as soon as Nate believes in the proposition that constitutes his reason, that proposition will turn out to be false - for, the party will no longer be a surprise for him -, so that fact would cease to be a reason for him to go home.

Therefore, since there is no possible world in which the fact that there is a surprise party waiting for him at his home is a reason for Nate to go home and in which Nate believes in that fact, that fact actually constitutes an elusive reason for Nate to go home. Since that is an elusive reason, we have at least one case that shows that Constraint* is false. Nate has no counterfactual capacity to be moved to go home by believing that there is a surprise party waiting for him at his home.⁹

5 Objections

Surprise Party indeed does not keep the popular views about the nature of reasons as they were. On the one hand, some suggest that Surprise Party shows the implausibility of Constraint read in terms of the Belief View (Schroeder, 2007; Markovits, 2014, ch. 2; Rossi, 2021a). On the other hand, some insist that some attenuated non-doxastic version of it is still plausible (McKeever, Ridge, 2012; Way, Whiting, 2016; Sinclair, 2016). I shall not discuss the latter enterprise here because it reformulates Constraint in other terms that are not at issue here - and which are not prominent readings of the claim¹⁰. Still, some think that Surprise Party does not show the possibility of elusive reasons, so it does not show that any version of the Constraint is actually false. There are four crucial objections to the possibility of elusive reasons in recent literature, and I will, for now on, address them and argue that they all ultimately fail.

5.1 Surprise Party is inconclusive

Some think that it is unclear that we should interpret Surprise Party as a counterexample to claims such as Constraint (Setiya, 2009, p. 538; Kiesewetter, 2016, p. 770; Paakkunainen, 2017, p. 68). For, we can track various normative phenomena in the vicinity of the elusive fact, which are all compatible with some version of Constraint. For instance, we can interpret the case as showing that the elusive fact is a reason for Nate to be glad if he gets surprised (a reason for Nate’s affective response); or that the elusive fact is a reason that explains why it would be good for Nate if he went home (an explanatory reason); or the same fact can be a reason for Nate’s friends to invite him to go home (a reason for someone else’s action); or the fact that it would be good for him could be a reason for Nate to go home (a reason for Nate to act), for which he can be moved to go home (Paakkunainen, 2017, p. 68). So - the objection goes on - Surprise Party is inconclusive. After all, all the other normative phenomena we can intuitively track in the case are either not reasons for Nate to act or non-elusive reasons for him or others to act in that circumstance.

Indeed, we can interpret Surprise Party so that we might track various normative phenomena in it. Still, I do not see why this intuition might rule out the idea that the elusive fact is not a reason for Nate to go home. In virtually any case in which we ascribe a reason to someone, we can track other normative phenomena in the vicinity of the reason in the circumstance, and the fact that we can track them does not rule out the ascription of the specific reason. Consider Bernard Williams’s (1981a) Gin/Petrol famous case. The fact that there is petrol in the glass is intuitively a reason for the agent to refrain from drinking it. In that case, we might track other normative phenomena in the vicinity: e.g., that specific fact is a reason for him to be glad if someone, being aware of the reason-constituting fact, tells him not to drink it (a reason for affective response); that fact also explains why would be good for him if he refrains from drinking it (an explanatory reason); the fact that it would be good for him constitute a non-elusive reason for him not to drink it; and so on. Why should we think this case is inconclusive with regard to the existence of the specific reason that there is petrol in the glass for the agent to refrain from drinking it? The fact that we can track those other normative phenomena in that circumstance does not rule out the intuition that that specific fact constitutes a reason for the agent to refrain from drinking it.

Even if some version of Constraint is compatible with other normative phenomena that we can track in a particular circumstance, it is enough that we identify just one fact that is a reason for A to ϕ in the circumstance, in which the consequent of Constraint* does not hold, to show that Constraint* is false. So, we should not conclude that Surprise Party is inconclusive.

5.2 The Argument from the Guidance Constraint

A similar but principled objection is that the fact that there is a surprise party waiting for Nate at his home is not a reason for action, that is, for Nate to go home. Instead, it is a reason for an affective response of him - again, for Nate to be glad if he went home, or to be dismayed if he did not - or a reason for someone else to act - e.g., for his friends to urge him to go home, etc. For, - the objection goes on - the following principle must be true (Setiya, 2009, p. 538):

Guidance Constraint: p is a reason for A to ϕ only if A is able to ϕ because p.

Guidance Constraint is the principle often used to rule out the possibility of the so-called ‘wrong kind’ of reasons (Setiya, 2009, p. 538; Parfit, 2011, pp. 50-51; Way, Whiting, 2016, p. 215). For example, the fact that I would be happy if I were 6 feet tall cannot be a reason for me to be 6 feet tall, but perhaps merely to want to be 6 feet tall. Guidance Constraint explains this by saying that I cannot be 6 feet tall because of the consideration that I would be happy if I were 6 feet tall. So, the point is that if the Guidance Constraint is true, we have an explanation of why the elusive fact is not - and cannot be - a reason for Nate to go home: Nate cannot go home because of the elusive fact.

However, this strategy does not work. First, it is hard to see how Guidance Constraint does not entail Constraint*. If it does, it simply reaffirms the plausibility of the claim at issue - so it cannot work as a conclusive argument here. Second, we can plausibly rule out the possibility of these wrong kinds of reasons by appealing to another principle that is compatible with the existence of elusive reasons for action, namely¹¹:

Response Constraint: There is a reason for A to ϕ only if A can ϕ for reasons.

If Response Constraint holds, there cannot be reasons, say, for me to be 6 feet tall because I am not able to be 6 feet tall in light of no matter what reasons. So, it does not imply that the specific elusive facts cannot constitute reasons for action. For, while agents cannot perform the action favored by an elusive fact for the elusive fact, agents can do so for other reasons. For instance, Nate can go home because it would be good for him, or because his reliable friend told him to do so, etc.

Now, note that Response Constraint has a crucial advantage over Guidance Constraint. For, Response Constraint gives us a more precise explanation of why there cannot be reasons, for example, for me to be 6 feet tall: the point is not that the consideration that I would be happy if I were 6 feet tall cannot be a reason for me to be 6 feet tall because I am not able to be 6 feet tall for that specific consideration; rather, the point is that the specific consideration cannot be a reason for me to be 6 feet tall because I am not able to be 6 feet tall in light of no matter what the consideration that might be speaking in favor of that. That is, the specific consideration cannot be a reason for me to be 6 feet tall because I am not able to be 6 feet tall for any reason at all - not just that specific one. This seems to be the proper and precise verdict.

So, since Response Constraint has a crucial advantage over Guidance Constraint and is compatible with the possibility of elusive reasons, I do not think this objection can ultimately work.

5.3 Against the pre-theoretical reasons’ ascriptions

Another objection against our intuitions in Surprise Party is that we should not ascribe the elusive reason from a pre-theoretical point of view. For, - the objection goes on - our reason’s ascription should be decided on theoretical grounds: whether reasons depend on the agent’s desires and what kind of dependence is that (Paakkunainen, 2017, p. 68).

From the perspective of a reasons theorist, it might seem implausible to ascribe a reason for someone without knowing at least part of the truth conditions of reasons statements such as ‘the fact that p is a reason for A to ϕ’ or ‘there is a reason for A to ϕ’. However, I think there is something wrong with this assumption. It does not seem to acknowledge that our use of the term ‘reason’ and our use of the reasons sentences should guide our delimitation of the truth conditions of the reasons statements. It seems less contentious to suppose that our pre-theoretical ascription of reasons should be explained by theories of reasons and not rejected by them. As I suggested in the last section, our ascription of the reason in Surprise Party intuitively does not misuse our ordinary use of ‘reason’. So, a plausible view on the nature of reasons should accommodate, not exclude that fact.

Nevertheless, we can bolster our intuitions by considering plausible metatheoretical conditions to reasons’ ascriptions, which the correct theory of reasons should accommodate. I think there are two metatheoretical conditions, and the reason ascribed in Surprise Party satisfies them.

The most trivial condition is that, for a fact to be a reason, an objective favoring relation between the reason-constituting fact, the agent, and the action must be obtained. Now, the fact that there is a surprise party waiting for Nate at his home indeed constitutes such a kind of relation in Surprise Party. For, when someone has a particular desire for something, a fact that explains why a particular action would promote that desire usually counts in favor of that action, at least pro tanto. We would correctly say, without misusing any concept, that if John loves Thai food, the fact that there is a Thai food restaurant in the corner counts in favor, at least pro tanto, John going to that restaurant (Schroeder, 2007, pp. 165-166). The only difference between cases like John’s and cases like Nate’s is that the latter involves elusive facts and the former does not, and it is not clear that in one case the relevant fact could not count in favor simply because it is elusive.

The second metatheoretical condition to reasons ascriptions is related to the deontic status: for a fact to be a reason for an agent to act, it must be able to explain why the agent ought to act in the way favored by it, at least in some circumstances - that is, in circumstances in which it constitutes the set of reasons that is weightier than the set of reasons to perform a conflicting action¹². Now, we can imagine cases in which Nate intuitively ought to go home and in which the elusive fact constitutes the correct explanation of why he ought to do so¹³. We can see this in three complementary ways.

First, note that if the elusive fact cannot correctly explain the deontic status of Nate going home in some circumstances, the elusive fact cannot contribute to making the set of facts it constitutes (which counts in favor of Nate going home) weightier than the set of facts that counts in favor of a conflicting action. However, the elusive fact can play that role. Suppose that today is Nate’s birthday. At his home, a surprise party made by his best friends is waiting for him. At the same time, at his grandmother’s house, a boring familiar reunion is also waiting for him for lunch. But he hates all familiar reunions, and, as a matter of fact, everybody in the familiar reunion has forgotten that today is his birthday, so they have not prepared anything special for him. Intuitively, Nate ought to go home now: the facts that favor going home intuitively outweigh those facts that favor going to his grandmother’s house. In particular, the set that includes the facts that there is a surprise party waiting for Nate at his home, that his best friends made the party, etc. that favor going home now outweighs the set constituted by the fact that there is a familiar reunion waiting for him at his grandmother’s house, etc. that favors going to his grandmother’s house now. So, in that case, intuitively, Nate ought to go home now. Therefore, since the specific elusive fact constitutes the weightier set of facts, the elusive fact surely helps to make the case that Nate ought to go home. But note that the elusive fact has a very strong weight: if there were no surprise party waiting for Nate at his home, it would make no sense to say that Nate ought to go home (other things being equal). The elusive fact can correctly explain why Nate ought to go home in that circumstance. If it does, that fact is likely a reason for Nate to go home.

Moreover, consider a reliable friend of Nate who knows his desires very well and knows the surprise party fact and other facts in the circumstance of Nate’s birthday. Based on those facts, his friend would plausibly conclude that ‘Nate ought to go home’. In light of this conclusion, he could justifiably find ways to make Nate go home without knowing that a surprise party is waiting for him there. It is not clear that his friend would be misusing these concepts: he would be making a justified judgment and sensitive to the normative facts around Nate’s circumstances. If we are comfortable accepting this judgment, we are also comfortable in saying that one thing that explains why Nate ought to go home is the elusive fact.

Still, Nate could surely conclude, by retrospection, that he ought to go home in that circumstance. Suppose Nate did not go to his home - he went to his grandmother’s house. Then, the next day, after being informed of the surprise party thrown in his honor by his friends, Nate could reflect on the circumstances and think, ‘Damn! I should not have gone to my grandmother’s house; I should have gone home!’. Why? Well, it seems that the correct explanation must involve the fact that there was a surprise party waiting for him at his home. Ultimately, it is hard to see why the elusive fact could not plausibly explain why Nate ought to go home in such circumstances.

Thus, the elusive fact satisfies two metatheoretical conditions to reasons’ ascriptions, providing us additional reasons to believe that we are not misusing the concept of a ‘reason’ in our reason’s ascription in Surprise Party.

5.4 The Argument from the Deontic Status

Finally, there is a more abstract and powerful objection to the possibility of elusive reasons, which appeals to the role of reasons in explaining the deontic status. As I already noted, reasons are taken to be the kind of normative property that explains why one ought to perform a particular action: if one ought to ϕ, there must be reasons for him to ϕ, which explain why he ought to ϕ. But - the objection goes on - if one ought to ϕ, he must be capable to conclude that he ought to ϕ in light of the reasons why he ought to ϕ; since elusive facts cannot be part of the agent’s deliberations, elusive facts cannot explain why one ought to do the action favored by them; therefore, elusive facts cannot constitute reasons. So, Surprise Party cannot be a counterexample to claims such as Constraint (Kiesewetter, 2016; Paakkunainen, 2017, 2018).

I have two responses to this objection. The first one appeals to the previous considerations: as I argued, we can construct cases in which Nate ought to go home, and in such cases, the elusive fact must be a fact that explains why he ought to go home; so, if we accept the idea that reasons are the kind of property that explains deontic status, that fact must plausibly be a reason for Nate to go home. Moreover, if those intuitions move us, the main premise of the objection - the idea that if one ought to ϕ, then he must be able to conclude that he ought to ϕ in light of the reasons why he ought to ϕ - also does not hold. After all, we can conceive cases in which Nate ought to go home, even though there is no possible world in which he concludes that he ought to go home because of the elusive fact - he could do so only by retrospection.

Nevertheless, we can set aside our previous intuitions and go further here. Note that if the main premise of the objection holds, all the facts that make the case that someone ought to ϕ must always be accessible before his conclusion to ϕ. For, if the agent must possibly conclude that he ought to ϕ in light of the reasons why he ought to ϕ, those reasons must always obtain before the deliberation start. However, it does not need to be so. We can imagine cases in which at least part of the justification of why someone ought to do something cannot be foreseen. I think Bernard Williams’s (1981b) Gauguin style of cases might illustrate this point. But we can avoid getting into too dramatic and complicated cases. Instead, consider yourself facing the challenge of deciding between two projects to pursue, say, the philosophical or the artistic life. There might be many reasons for you to pursue each one of the projects, and it might be that the reasons for one of them make the case that you ought to pursue one instead of the other. But not all reasons that make the case that you ought to pursue one of the projects you can foresee: the reasons related to your success in one of the projects surely cannot be accessed by you at the time you are deciding which project to pursue. After all, you have not succeeded yet when deliberating about which project you ought to pursue. Still, the reasons you cannot foresee related to the project’s success fundamentally explain why you ought to pursue one of the projects. This is so intuitive because in such circumstances, even if we deliberate correctly with all the relevant facts that we can know about ourselves at the time and about the external conditions for the success of the project itself, there is always a sense in which we do not really know, full-stop, what we ought to pursue. In this sense, the justification related to success cannot be foreseen.

Yet - as Williams suggests in his discussion on Gauguin case - such kind of justification can be accessed by retrospection. For example, suppose you have succeeded in one of the projects, say, the philosophical one. It is plausible to suppose that the facts related to your success at least help to make the case that what you ought to pursue the philosophical life. In other words, those facts help to explain why pursuing the philosophical life was ‘the right thing to do’. In fact, in such cases, only by retrospection would you have a complete sense in your deliberations about why the project you chose was the right one - that is, the one you ought to pursue. Therefore, I do not think that all the facts that make the case that someone ought to ϕ must be accessible before his conclusion to ϕ, and this makes the main premise of the objection ultimately implausible.

Someone might object by saying that those considerations related to success, which - as I argued - constitute the ultimate justification of the project, are not really reasons that explain why you ought to pursue that project: they are not facts when you are deliberating about what to do in the first place, so they cannot be reasons that explain why you ought to pursue one of the projects at all. But why should we suppose that the success in each step of our life projects does not help - conclusively so - to make the case that we are justified in pursuing a particular project after all? From the perspective of the agent pursuing such a project for her life, the success in each step of it is like ‘insurance’ that the project is in fact what she ought to pursue. If success in each step of the project indeed has that role, it must constitute the correct explanation of why she ultimately ought to pursue that project. So, if we agree just on this intuitive thought, we have arrived at the following conclusion: either we should say that the explanation of the deontic status of someone’s action does not depend on his capacity to deliberately conclude that he ought to perform it in light of the justification why he ought, or we should conclude that reasons are not the kind of property that explains the deontic status. Either way, the upshot is that we should reject at least one of the premises of this objection against the possibility of elusive reasons. Thus, I think this objection ultimately fails.

Still, someone might insist and say that we have arrived at a too revisionist conception of the deontic status: the relevant ‘ought’ that we referred to in discussions about the deontic status is the ‘deliberative ought’ - that is, the ‘ought’ related to the question ‘what ought I to do?’ - so the notion of ‘ought’ must be connected to the possible deliberations of the agents to whom the ‘ought’ applies. However, note that our conclusion does not detaches ‘ought’ from the possible deliberations of the agents to whom it applies. Instead, the point is that the agent need not be capable of deliberating before the actions. After all, in the cases we discussed, all agents can deliberate about what they ought to do by retrospection, that is, after some justifications are obtained and possessed. So, we still have such a connection between ‘ought’ and deliberation - though a different one. Whether or not such differences might be accommodated by a complete and correct account of the deontic status is an issue for another time.

6 Conclusion

In this paper, I argued that Constraint is false in any of its prominent readings. I argued that Schroeder’s Surprise Party case shows the possibility of elusive reasons and that all the most crucial objections raised against that possibility ultimately fail. The objection that Surprise Party is inconclusive fails because it is not sufficient to rule out our intuitions about the case. The objection that appeals to the Guidance Constraint fails because we can rule out the possibility of wrong kinds of reasons by appealing to a plausible principle that is compatible with elusive reasons. The objection that we should not ascribe a reason from a pre-theoretical point of view fails because our reasons’ ascription in Surprise Party satisfies metatheoretical conditions on reasons’ ascriptions. Finally, the objection that appeals to reasons’ role in explaining deontic status fails because its assumptions are incompatible with the possibility that one ought to do something, even though he is in no position to conclude that he ought to do that in light of the reasons why he ought.

So, since the general version of Constraint is false, we also have powerful reasons to reject some popular views on normative reasons, such as Internalism about Reasons and Deliberative Constraint. In the end, we better not aim to explain the nature of normative reasons for action in terms of the motivational capacities of the agents whose reasons they are.

References

• ALVAREZ, M. 2017. Reasons for Action: Justification, Motivation, Explanation. In: ZALTA, N. (ed.)The Stanford Encyclopedia of Philosophy Disponível em: Disponível em: https://plato.stanford.edu/archives/win2017/entries/reasons-just-vs-expl/ Acesso em: 10 abril, 2023.
  » https://plato.stanford.edu/archives/win2017/entries/reasons-just-vs-expl/
• BRUNERO, J. 2017. Recent Work on Internal and External Reasons. American Philosophical Quarterly, 51(2): p. 99-118.
• JOHNSON, R. 1999. Internal Reasons and the Conditional Fallacy. The Philosophical Quarterly, 49(194): p. 53-71.
• KIESEWETTER, B. 2016. You ought to ϕonly if you may believe that you ought to ϕ. The Philosophical Quarterly, 66(265): p. 760-782.
• KORSGAARD, C. 1986. Skepticism about Practical Reason. Journal of Philosophy, 83(1): p. 5-25.
• LORD, E. 2018. The Importance of Being Rational 1º ed., Oxford, Oxford University Press, 253 p.
• MARKOVITS, J. 2014. Moral Reason 1º ed., Oxford, Oxford University Press , 208 p.
• MCKEEVER, S.; RIDGE, M. 2012. Elusive Reasons. In: SHAFER-LANDAU, R. (ed.) Oxford Studies in Metaethics , Volume 7. 1º ed., Oxford, Oxford University Press : p. 110-137.
• PAAKUNAINEN, H. 2017. Can there be ‘Government House’ Reasons for action? Journal of Ethics and Social Philosophy, 12(1): p. 56-93.
• PAAKUNAINEN, H. 2018. Internalism and Externalism about Reasons. In: STAR, D. (ed.). The Oxford Handbook of Reasons and Normativity 1º ed., Oxford, Oxford University Press : p. 157-158.
• PARFIT, D. 2011. On What Matters: Volume One 1º ed., Oxford, Oxford University Press , 540p.
• ROSSI, B. 2021a. Elusive Reasons and the Motivational Constraint. Journal of Ethics and Social Philosophy, 20(1): p. 82-110.
• ROSSI, B. 2021b. Introducing a New Elusive Reason. Ratio, 34(3): p. 1-9.
• SCHROEDER, M. 2007. Slaves of the Passions 1º ed., Oxford, Oxford University Press , 224p.
• SETIYA, K. 2009. Reply to Bratman and Smith. Analysis, 69(3): p. 531-540.
• SETIYA, K. 2012. Introduction: Internal Reasons. In: SETYIA, K.; PAAKUNAINEN, H. (ed). Internal Reason: Contemporary Readings Massachusetts, MIT Press: p. 1-34.
• SETIYA, K. 2014. What is a reason to act? Philosophical Studies, 167(2): p. 221-235.
• SINCLAIR, N. 2016. On the Connection Between Normative Reason and the Possibility of Acting for those Reasons. Ethical Theory and Moral Practice, 19(5): p. 1211-1223.
• SOBEL, D. 2001. Explanation, Internalism, and Reasons for Action.Social Philosophy & Policy, 18(2): p. 218-235.
• SMITH, M. 2009. Reasons with Rationalism after all. Analysis, 69(3): p. 531-540.
• WAY, J.; WHITING, D. 2016. Reasons and Guidance (Or, Surprise Parties and Ice-Cream). Analytic Philosophy, 57(3): p. 214-235.
• WILLIAMS, B. 1981a. Internal and External Reasons. In: WILLIAMS, B. Moral Luck 1º ed., Cambridge, Cambridge University Press, 173p.
• WILLIAMS, B. 1981b. Moral luck. In: WILLIAMS, B. Moral Luck 1º ed., Cambridge, Cambridge University Press , 173p.
• WILLIAMS, B. 1995. Internal Reasons and the Obscurity of Blame. In: WILLIAMS, B. Making Sense of Humanity 1º ed., Cambridge, Cambridge University Press , 251 p.

• 1
  For discussions about versions of the Constraint, see Sobel (2001), McKeever & Ridge (2012), Sinclair (2016), Paakkunainen (2018), Way & Whiting (2016), and Rossi (2021a).
• 2
  In what follows, when I use ‘reason’ or ‘reasons’ I mean normative reasons for action unless when emphasized.
• 3
  For discussion about these two possible readings of that notion, see especially Setiya (2012, p. 5), Sinclair (2016), Paakkunainen (2018), and Rossi (2021a).
• 4
  On the discussion about what it is to recognize a fact as a reason, see especially Lord (2018: ch. 4).
• 5
  Note that that is just a simple formulation. For, the notion of ‘ϕ-ing by the belief that p’ is supposed to subsume the conditions stated by Belief View. I opted for the simple formulation to be economical here.
• 6
  Many authors endorse that view. See especially Williams (1981a; 1995) and Korsgaard (1986).
• 7
  Besides the main ‘internalists’ about reasons, Setiya (2014) and Paakkunainen (2017; 2018) seem to endorse some version of that claim.
• 8
  That formulation is very similar to Paakkunainen’s (2018, p. 156) own formulation. However, in my formulation, the Deliberative Constraint is not only about decisive reasons but also about pro tanto reasons - that is why, in my formulation, the conclusion of the deliberative route is a motivation to act, not the action itself.
• 9
  For other cases that are structurally similar, see Johnson (1999), Smith (2009), and Markovits (2014). Unfortunately, I shall not discuss those cases to be economical here. See also Rossi (2021b) for the possibility of another ‘kind’ of elusive reasons.
• 10
  For a discussion on that strategy, see Rossi (2021a).
• 11
  On the distinction between what I have been calling Guidance Constraint and Response Constraint, see McHugh & Way (2022, p. 15, n. 6).
• 12
  That is a widespread idea about the role of reasons. See especially Schroeder (2007, ch. 7), Paakkunainen (2017, §3), and Way & McHugh (2022, ch. 5 and 6).
• 13
  Brunero (2017, §3) argues that Nate ought to go home in Surprise Party case. Instead, I argue that Nate ought to go home in a modification of Surprise Party to make explicit the weighing relation between the considerations for Nate to go home and the considerations for Nate to do a conflicting action, which ultimately must explain why Nate ought to go home.

Publication Dates

• Publication in this collection
  14 June 2024
• Date of issue
  2024

History

• Received
  23 May 2023
• Accepted
  19 Feb 2024