Abstract
In his book Normative Reasons (Logins A in Normative reasons: between reasoning and explanation. Cambridge University Press, 2022), Artürs Logins accepts that a normative reason to do A is always an answer to a ‘Why A?’ question, but rejects the unifying explanationist theory which identifies reasons always as explanations. On his Erotetic Theory, ‘Why A?’ questions sometimes seek an explanation (in No-Challenge contexts) but sometimes seek rather an argument (in Challenge contexts). This article defends a unifying, end-relational explanationist theory by interpreting ‘Why A?’ as being elliptical for different questions, with different explananda. I also respond inter alia to Logins’ claim that end-relational explanationism is extensionally inadequate because it fails to account for normative reasons for attitudes. Finally, I consider the objection that explanationism fails to account for normative reasons’ characteristic functional role of settling deliberation, introducing a “chicken or egg” dilemma over the order of discovery of reasons and options, resolved by suggesting that in open-ended deliberation reasons guide us to options without being represented under the guise of “reasons”.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
Artürs Logins’ book Normative Reasons (2022) makes a state-of-the-art contribution to metaethics, championing a unifying and reductive theory of normative reasons as answers to normative Why questions. Superficially this would seem to indicate substantial agreement, as I and others have said the same.Footnote 1 But Logins describes his view as a “radically different proposal” (p. 159), arguing against extant treatments, including a generous section of his third chapter critically engaging with my account (as laid out in Finlay 2006, 2010, 2014, 2019). Here I explore the basis for the claimed disagreement, and examine the case: are Logins’ objections to my account of normative reasons cogent? How does his own proposal stand up to scrutiny?
The disagreement arises because Logins splits a hair that I and most others do not.Footnote 2 I identify answers to normative ‘Why?’ questions universally with explanations of normative facts, yielding an Explanationist analysis of normative reasons. While Logins agrees that answers to ‘Why?’ questions are sometimes explanations, he argues that at other times, answers to ‘Why?’ questions are not explanations but rather (premises of) arguments providing the basis of good reasoning to a normative conclusion. This yields a disjunctive view of normative reasons as either explanations or arguments, which Logins embraces as a satisfying reconciliation of the two main reductivist camps in metaethical theorizing about normative reasons: Explanationism and the Reasoning View.Footnote 3 This result is to be welcomed, Logins holds, because both Explanationist and Reasoning views face fatal shortcomings as complete theories of normative reasons.Footnote 4 In this article I offer an Explanationist rejoinder.
1 Are ‘Why?’ Questions Ambiguous between Explanations and Arguments?
Logins suggests that ‘Why?’ questions have two different kinds of answers because there are relevantly two different kinds of ‘Why?’ question, and subsequently two different kinds of normative ‘Why?’ question. One kind of ‘Why Φ?’ question is asked when the interrogandum Φ itself is not saliently in doubt (a No-Challenge context), and is appropriately answered (familiarly and uncontroversially) by providing an explanation of Φ. A distinct and overlooked kind of ‘Why Φ?’ question, Logins claims, is asked rather when Φ is in doubt (a Challenge context), and is appropriately answered by providing a premise, or part of an argument, that constitutes good reasoning to Φ. Consider the question ‘Why are dolphins not fish?’ (p. 166f). If asked in a No-Challenge context where it is accepted that dolphins are not fish, Logins suggests, then an appropriate answer may be an evolutionary explanation. But if asked in a Challenge context where we don’t already accept that dolphins are not fish, Logins suggests, then an appropriate answer may rather be ‘Because that’s what biologists have shown to be the case’, which he suggests is a premise of good reasoning to the conclusion that dolphins are not fish, but not an explanation of this fact.
I have concerns about this general account of ‘why’. One concern is whether the observation that ‘Why?’ questions are asked in different kinds of context for different purposes supports the claim that these are different kinds of question, or that ‘why’ is ambiguous between them. To infer directly from the use observation to the meaning claim would be to commit the Pragmatic (or Speech Act) Fallacy: the mistake of conflating the meaning of a word or sentence with some uses to which it is sometimes put.Footnote 5 We should consider whether Logins’ two kinds of ‘Why?’ question might really be (more parsimoniously) a single kind of question put to different uses in different contexts. In particular, whether providing an explanation why Φ might not only be of relevance for deepening our understanding of Φ, but also for settling a question of whether Φ.Footnote 6
Fueling this worry is the observation that we can and do ask for arguments for Φ when Φ isn’t in question (e.g. Descartes looking for a proof for God’s existence while (albeit disingenuously) denying it needs any), and we can and do ask for explanations why Φ when Φ is in question. Consider Logins’ examples of “Moore-paradoxical” reasons (e.g. the fact that the building is on fire but John doesn’t believe it is, is a reason why John should check on the building (p. 180-1)) and surprise party reasons (the fact that there is a surprise party planned for John is a reason why he should be home on time), which Logins identifies as explanatory reasons only (since John cannot appropriately reason from them to the normative conclusion), but which he allows can appropriately be cited to address someone else’s uncertainty about what John should do.Footnote 7
I am unconvinced that we need a separate Reasoning reading of ‘Why Φ?’ Every example Logins offers of a premise-reason seems to me (if plausibly a reason at all) also an excellent candidate, given an appropriate informational background, for being an explanation (e.g. dolphins are not fish because they are mammals (p. 167)). The undergeneration objections against Explanationism seem to assume a narrow (causal or scientific) understanding of what qualifies as an explanation. What would be needed to establish the inadequacy of Explanationism are persuasive examples of appropriate answers to ‘Why?’ questions that are premises of good arguments but not plausibly also relevantly explanations. Notice that while ‘Biologists have shown that dolphins are not fish’ may indeed be a premise of a sound argument to the conclusion that dolphins are not fish, it is not really a premise one could use to settle a genuine doubt about whether dolphins are not fish, since ‘shown’ is factive and therefore the premise covertly assumes the conclusion, i.e. the argument would be question-begging, and so by Logins’ own lights any context in which it was granted would be a No-Challenge context.
Perhaps the clearest candidates for premises which are not also explanations seem to involve testimonial evidence (Kearns & Star, 2009). For example, the fact that marine biologists all say that dolphins are not fish, which seems a good premise of a non-question-begging argument to the conclusion that dolphins are not fish while seemingly not an explanation of it. But it isn’t clear that we can say felicitously that this is a reason why dolphins are not fish, and indeed Logins agrees (p. 145f) that testimonial evidence that p (along with other forms of nonexplanatory evidence such as enabling conditions of p and derivative normative facts) are not reasons why p. Simple Reasoning (or Evidence) theories, he agrees, overgenerate reasons in these cases, but the Erotetic theory is claimed to avoid this problem on the grounds that ‘Why?’ questions put in place additional appropriateness constraints on premises (p. 180f). We are not told much about where these appropriateness constraints come from, but my suspicion is that they invariably arise precisely because ‘Why?’ questions demand explanations.Footnote 8 If so, Explanationism is being covertly smuggled in the back door to bail the Erotetic theory out of the trouble it gets into every time it ostensibly departs from Explanationism.Footnote 9
2 Interpreting Normative ‘Why?’ Questions
Considering specifically normative reasons and ‘Why?’ questions raises further issues. The canonical form of these questions in Logins’ treatment is ‘Why F?’ (e.g. ‘Why jump in the pond?’), where F is a verb phrase/ names an action or behavior. This prompts two immediate questions. First, how are we to make sense of ‘Jump in the pond’ being either something in need of explanation, or the conclusion of an argument? Second, what makes this a normative question, and its answers normative reasons? The answer to both questions is that Logins takes ‘Why F?’ to be universally equivalent to an explicitly normative question of the form ‘Why should/ought I jump in the pond?’ The account is ostensibly unifying because both kinds of normative reasons are answers to questions of this form. This would seem to require that the word ‘why’ itself is ambiguous, composing with its interrogandum Φ in two different ways. One might worry, then, that the unity Logins claims for his account is no deeper than the unity of nephrite and jadeite in virtue of both being called ‘jade’, a mere contingency of language.
Observe however that ‘Why F?’ has many other possible readings. For example,
Why should she/they/we jump in the pond?
Why should I have jumped in the pond?
including some readings that are not normative at all, like
Why did you/he jump in the pond?
(Imagine an exasperated parent berating a muddy toddler). Rather than postulate massive ambiguity and a different sense of ‘why’ for each question, it is natural to see such ‘Why F?’ questions as elliptical, leaving significant parts of the question implicit. (Compare also ‘Why him?’) But then we should be careful about leaping to the conclusion that a normative or practical ‘Why F?’ question is necessarily equivalent to ‘Why should I F?’ Recognizing the possibility of ambiguity in the interrogandum Φ brings to light the possibility that different readings of ‘Why F?’ questions don’t involve any ambiguity in ‘why’. We’ll return to this observation numerous times.
I suggest that a better, more parsimonious (and Explanationist-friendly) diagnosis of Logins’ observation about the ambiguity of ‘Why F?’ questions is that they allow interpretation both as
Why should I F?
(requesting an explanation for why one should F), and also as, e.g.
Why do you say/think I should F?
(or ‘Why should I think I should F?’)Footnote 10 An appropriate answer to this second question need not be an explanation why one should F, simply because it needs to be an explanation rather of the speaker’s saying or thinking it. Similarly with non-normative ‘Why?’ questions; I suggest it is only on an interpretation of ‘Why are dolphins not fish?’ such as ‘Why do you claim dolphins are not fish?’Footnote 11 (or ‘Why are dolphins not fish in your opinion?’) that an answer like ‘Because that’s what biologists have shown’ is appropriate, precisely because in that context it is indeed an explanation.
This diagnosis has the advantage of avoiding positing any ambiguity in ‘why’.Footnote 12 Does it also have the philosophically more substantial advantage of enabling us to vindicate a more robustly unifying, Explanationist theory of normative reasons? I believe so, but to show this we must consider Logins’ case against the adequacy of Explanationist theories. Fortunately, I think that Logins’ objections to at least one existing Explanationist theory do not succeed.
3 Defending End-Relational Explanationism
According to my favored analysis of (pro tanto) normative reasons, they are all explanations why the action or attitude (etc.) in question would be good to some degree, in relation to some salient end. On this view, normative reasons statements are contextually relative to both an end parameter (fixing the kind of value at issue) and a background information parameter (fixing what can count as an explanation of that value). Logins therefore labels it Axiologic Contextualist Explanationism, but a less distracting label would be End-Relational Explanationism (ERE). Logins’ presentation of my view is meticulous, and he generously observes numerous virtues before presenting his grounds for dismissing it as unsatisfactory. I don’t think these objections succeed, and therefore believe that ERE emerges unscathed from the carnage Logins skillfully metes out on the rest of the field.
I discriminate three broad areas of complaint directed at ERE.Footnote 13 The first is an objection about undergenerating reasons: that ERE (like all Explanationist theories) has a “hard time explaining how reasons for attitudes are possible” (p. 181). Logins acknowledges that unlike most other Explanationists, I do offer a vindicating analysis of (“right-kind”) reasons for attitudes: that they are explanations why having the attitude is good for (i.e. raises the probability of) thereby achieving the constitutive ends of that attitude. For example, supposing we say that the constitutive end of the attitude of belief (believing that p) is truth (believing that p iff p is true); then on this account a proposition q is a “reason to believe p” just in case q is an explanation why believing p would be good for thereby believing p iff p is true. Logins judges this account unsatisfactory: “it is not clear in what sense, if at all, constitutive aims can count as one’s salient ends and be of genuine use for a value-based account of reasons” (p. 119), and presents a cluster of objections.
These objections start from an assertion that the sense in which (e.g.) truth is the “constitutive end” of belief is that “truth sets the standard of correctness or fittingness for belief” and therefore that belief is defined as “an attitude that is correct just in case it is true” (p. 119f). This is much more controversial than Logins’ presentation suggests. Indeed as I am much more confident in the existence of beliefs than I am in their sometimes having a property of objective correctness, distinct from their merely being true, I find it quite implausible. But supposing we grant it, what is the problem?
Logins writes first that “if we understand the aim of belief in this sense, it is not clear how it could be one of the salient ends that can be plugged into [ERE’s] definition of reasons” (p. 120). The worry (if I understand it) is that a “constitutive aim” is not an end, in the sense of an actual goal being pursued.Footnote 14 I agree, but this is not the intended sense of “end”. Rather, an “end” is just a potential state of affairs, conceived as a possible outcome. So defined, ends can be salient in numerous ways. One source of salience for ends is being the object of a salient person’s goal-directed activity (e.g. of the speaker’s desires). But there are other sources of salience. My account of (“right-kind”) reasons for attitudes appeals to the salience due to the attitude in question being associated with a constitutive end. Hence I claim that it really doesn’t matter for this account what being a constitutive end involves (2019: 69). All this account needs is that talking about attitudes of various kinds makes them salient, which doesn’t seem controversial.
A second objection against ERE is that “allowing for standards of fittingness…to play an indispensable and irreducible role in an analysis of reasons is giving up on the purely ‘axiological’ aspect…[making] it unclear in what sense the view would still count as a value-based view of reasons” (p. 120). There are several points to be made in reply,Footnote 15 but the most pertinent is that I (and ERE) am not committed to a narrowly axiological Explanationism. My appeal to explanations of value is in the service of accounting specifically for pro tanto reasons.Footnote 16 I also admit pro toto reasons analyzed as explanations of ‘ought’s, as well as merely permissive reasons as explanations why one may. It is therefore fully compatible with this broader Explanationist picture to accommodate reasons for attitudes also as explanations of fit.
Logins acknowledges that appeal to fit is not “a problem per se” (p. 120), but observes that such a “hybrid view” might be inferior to non-hybrid views “on account of simplicity and parsimony”. I disagree. Clearly requirements, permissions, values, and fit all (at least sometimes) have explanations (and arguments!), so if the general notion of a normative reason is just that of an explanation of a normative fact, then it would be quite ad hoc to admit value-based normative reasons but not ought-based or fit-based normative reasons (etc.), or vice versa. Given that there are normative ‘Why F?’ questions involving normative concepts other than ‘ought’, this should be no less expected on the Erotetic approach.
Logins presents one further complaint about ERE’s ability to accommodate reasons for attitudes: that it cannot accommodate what he calls (following Robert Gordon) epistemic reasons for emotions. He illustrates this with the case of worry. One kind of reason for worry (“connected to fit-making”) is a proposition about something bad, e.g.
Tom’s reason for being worried that his wife was on the two o’clock flight is that it was hijacked.
Another kind of reason for worry is “connected to the purely epistemic status of an emotion”. For example,
Tom’s reason for being worried that his wife was on the two o’clock flight is that she said she’d be arriving early in the evening.
The example is a bit opaque, but I take the idea to be that because this fact warrants some intermediate credence that Tom’s wife is on the two o’clock flight, it makes it appropriate for him to worry in light of some unspecified bad thing (her being stranded at the airport, perhaps). Logins classifies this as a “wrong-kind” reason in virtue of not being connected to fit-making, but argues that because it involves epistemic warrant rather than value it poses a difficulty for Explanationism.
I agree that there is an epistemic dimension to the appropriateness of worry, but disagree with Logins’ analysis. Certain emotional attitudes—including worry and hope—include epistemic factors in their constitutive ends. Approximately, hope aims at the good and uncertain, while worry aims at the bad and uncertain. It is this sensitivity to uncertainty that differentiates hope from joy, and worry from sorrow. Accordingly, when you are certain that something is or isn’t the case, it then isn’t fitting (or as I suspect, possible) to hope or worry that it is the case. By contrast, other emotional attitudes such as pride and amusement do not include epistemic factors in their constitutive ends, and so there are no corresponding “epistemic reasons” for these attitudes. The fittingness of pride or amusement towards p is not similarly contingent on being uncertain whether p.Footnote 17
If this analysis is correct, then Logins’ “epistemic reason” is in fact of the fit-making “right kind”. It just targets worry’s uncertainty condition rather than its badness condition. The ERE template for an Explanationist accommodation of right-kind reasons for attitudes then applies straightforwardly: the constitutive end of worrying that p is: worrying that p iff p is bad and uncertain. A (right-kind) reason for worrying that p will then be an explanation why worrying that p is good for (raises the probability of) thereby worrying that p iff p is bad and uncertain.
I think we have now addressed all Logins’ extensional objections applying to ERE—which all target reasons for attitudes, not reasons for action. If these responses are successful then he has not identified any genuine reasons that ERE omits, or any false predictions ERE makes that something is a reason. However, Logins also raises two more general philosophical objections, one against Contextualism more generally, and one against Explanationism more generally.
4 Is Contextualism Grounds for Disqualification?
The very flexibility that enables a contextualist view like ERE to deal with the problems for Explanationism, Logins claims, is itself “somewhat problematic” (p. 117). Why? He writes,
One needs to explain how we are supposed to make sense of a view on which normative concepts are always relativized to some further aspect… The worry here is the apparent lack of independent theoretical motivation for such an extreme contextualism given that our pre-theoretical judgments about reasons…don’t seem on the face of it to be so radically context-dependent.
There are a lot of responses I’m tempted to offer here to this rather sweeping complaint. I disagree emphatically about the “lack of independent theoretical motivation”, but can only refer readers to other work.Footnote 18 One might also question the alleged extremity of ERE’s contextualism, given the extensive appeals to the context-sensitivity of reasons-talk throughout Logins’ book. Logins suggests (before the Erotetic theory rides to the rescue) that given the challenges facing the views he surveys, we might reasonably be tempted to the eliminativist view that there are no normative reasons (p. 164). By contrast, swallowing a bit more context-sensitivity than one might pre-theoretically have expected seems a small price to pay if a theory makes all the intuitively correct predictions about what is and is not a normative reason.Footnote 19
The most pertinent response here however is that the ERE account of normative reasons does not need the allegedly extreme claim that normative concepts are always relativized to some further aspect. All that is needed is to make sense of normative concepts sometimes being relativized in these ways. While it is true I personally favor a radical contextualism, those who believe in properties of value (or ought)-simpliciter can add them to this Explanationist framework without rejecting anything else in the picture. The real question is whether normative concepts do allow the kind of flexibility needed for ERE’s accommodation of the full extent of our normative reasons claims. I agree that if the Erotetic theory—or some other rival—could match or beat ERE’s extensional adequacy without such a high level of context-relativity, then it would be ipso facto superior. But I don’t see that challenge as met, and theoretical simplicity plays second fiddle to theoretical adequacy.
5 Reasoning with Explanations
We come finally to Logins’ most general objection to Explanationism: that it cannot do justice to the role of normative reasons in practical reasoning. He writes,
The exclusive focus on the role of reasons in the explanation of [normative facts] leads to overlooking the importance of the role of reasons in good/fitting reasoning. (p. 5)
[Explanation views] have a hard time explaining…that there has to be some connection between reasons and reasoning broadly understood. In a sense, reasons are understood to be right foundations on which our attitudes can be based; that is, reasons are often perceived as considerations that can lead us to hold fitting responses… Views that limit their accounts of normative reasons only to the explanatory function fail to appreciate this important aspect… (p. 135).
Explanation views…are not really apt to account for the reasons-reasoning connection. (p. 164)
But why not?
Notice that in asking this ‘Why?’ question, I am at once challenging Logins’ claim, and also demanding an explanation why it would be true, i.e. what would make it true if it were—an explanatory demand in a Challenge context. If a satisfactory explanation is given, then the claim will be vindicated and we should accept it as true. But if no satisfactory explanation is offered, then we are warranted in rejecting the claim as groundless or presumptively false. As hereby performatively demonstrated, if there is a problem for Explanationism in this vicinity it is not that explanations cannot function as premises of good reasoning. As noted, at least many of Logins’ examples of premise-reasons are also eminently plausible candidates for explanations. (E.g. dolphins are not fish because they are mammals, the army should go North because that will put it in a better position to defeat the enemy). This should be no surprise: we should normally be able to reason to a proposition’s truth from the facts that make it true.
Perhaps the problem is supposed to be that Explanationism is unable to fully capture the extension of normative reasons, both overgenerating and undergenerating reasons (p. 164) because it doesn’t make the “connection between reasons and reasoning” tight enough? Extensional adequacy is a hard thing to prove, but if there is any problem for End-Relational Explanationism in this vicinity I don’t believe it has been located. I’ve argued above that Logins’ particular extensional objections to ERE (concerning undergenerating reasons for attitudes) don’t succeed, and I am not persuaded that any examples have been provided of (i) plausibly “appropriate” premise-reasons which are not also in relevant contexts plausibly explanations of the normative status of the action or attitude, or (ii) normative explanations which are not also in relevant contexts appropriate premises for good reasoning to the normative conclusion. By contrast, we have seen that the Erotetic theory—like pure Reasoning theories—faces trouble wherever it encounters non-explanatory premises (e.g. testimony, enabling conditions), needing to be bailed out by an underexplained “appropriateness” constraint that may just prove to be explanationist. I conclude that no extensional problem has been identified for ERE here.
There is however an intensional version of the complaint which deserves our attention. This stems from the observation that in practical reasoning our goal is to figure out what the normative facts are, which is just a different activity than trying to explain what makes them true.Footnote 20 The central functional role of normative reasons is to provide considerations we can use to figure out what the normative facts are (or perhaps better, to decide what to do, think, feel), which is fundamentally distinct from an explanatory functional role. So one might think, no matter how much Explanationism might get extensionally right, it is still fundamentally wrong about the concept of a normative reason, which is rather the concept of a premise of good reasoning. Supporting this line of thought is the obvious lexical connection between the terms ‘reasons’ and ‘reasoning’, which Reasoning theorists not uncommonly treat as decisive evidence for their view.
This is an appealing line of thought, but I think it is ultimately mistaken. In addition to its extensional problems (that the clearest cases of good premises that are not also saliently explanations, like normative testimony, are not intuitively identified as reasons to F), the overall linguistic evidence is also overwhelmingly against it and in favor instead of a conceptual link between reasons and explanations. We use the same word ‘reason’ for explanations in other contexts (see Finlay, 2014: ch.4), a pattern that is strongly cross-linguistic, in contrast to the connection of ‘reasons’ and ‘reasoning’ which is largely idiosyncratic to English. In any case this link could equally be explained the other way around: that reasoning owes its name to being a process characteristically involving thinking about reasons (as grounds or explanations).
Yet the basic observation about different functions seems correct. Why would it be that practical reasoning or deliberation looks to normative explanations, given that its function is to settle what to do rather than to explain it? I suggest a promising answer can be found in the supervenient or “consequential” character of (particular) normative facts. Whenever an option is good or ought to be chosen, this is always in virtue of empirical, non-normative facts which make it a good or choiceworthy option. The epistemic upshot is that we do not perceive normative facts directly, but rather discern them on the basis of their subvening truth-makers. Generally, therefore, our access to normative truths is via their grounds (testimony being a rare exception), and we have no means of settling what the normative facts are except by looking to those grounds. This may suffice to explain why practical reasoning looks to reasons as explanations despite explanation not being its aim.
But if the very concept of a normative reason is that of an explanation, then why do we think in terms of “reasons” when engaged in deliberation, when our goal is to find out what we ought to do (or better, decide on what to do).Footnote 21 Reflection on this challenge prompts an intriguing thought: perhaps we don’t. Or not as much as is assumed. The concept of a “normative reason” figures centrally in philosophical discussions of practical reasoning and deliberation, which are reflective contexts focused on explanation rather than on decision-making, so perhaps its centrality in our thinking about deliberation is owed to this.Footnote 22 It is questionable whether it is similarly central in practical reasoning itself for these considerations to be represented in their guise as “reasons”.
Open-ended deliberation (at least) begins with a recognition of some threats or opportunities, without any actions or options yet identified. But to be a “reason” is always to be a reason for some option or other. So deliberation cannot begin with identifying one’s “reasons” as such (reasons to do what?). Indeed, despite the characterization of arguments as “forward-looking” versus explanations as “backward-looking” this is no less an issue for premises than for explanations. Something is only a premise or argument in relation to a particular conclusion. It is therefore implausible to conceive of open-ended deliberation as a search for “premises” as such. Accordingly, in discussing deliberation, Logins switches from ‘Why?’ questions to ‘What?’ questions: ‘What am I to do/believe?’ (p. 220f). Indeed there is no ‘Why?’ question corresponding to open-ended deliberation.
We face a chicken-and-egg problem: which comes first, identifying the option (or conclusion), or identifying the reason (or premise)? On one hand, it surely can’t be that we identify what to do prior to any consideration of our reasons. But on the other hand, it is only in relation to particular options or conclusions that a consideration can even be conceived of as a reason (or explanation, or premise). I tentatively conclude that our thoughts must initially be guided to options by (some of) the considerations that are reasons by recognizing what follows from them (what they ground), without us already representing them in their guise as “reasons” (for those options).
This is not to deny the centrality of normative reasons to deliberation, merely to challenge the centrality of conceptualizing them as “reasons”. But neither is it to deny that the concept of a normative reason has any role to play in deliberation. Once we have identified some candidate options A, B… we can then profitably ask, ‘Why choose A?’, ‘Why choose B?’ These are calls for reasons as explanations—of value.Footnote 23 What makes A a good option? (and how good does it thereby make it?) What makes B a good option? Those grounds can then be weighed for how much value they contribute, to reach a should, or a decision (an answer to ‘What shall I do?’). When Logins dismisses the relevance of explanatory reasons to practical reasoning, as backwards-looking and therefore requiring a conclusion already reached (p. 221), he overlooks explanations of value rather than of ‘shoulds’. But the question ‘Why should I do A?’, I suggest, needn’t and perhaps doesn’t normally figure significantly in deliberation at all.
I therefore remain confident that normative reasons are always and only normative explanations.
Notes
In Finlay, 2001 (p. 108f) I also identified normative reasons both as facts that explain, and as facts that reveal, normative facts, but this was due to an unintentional failure to differentiate explanations and evidence.
Adherents of Reasoning theories are more likely to find this reconciliation amenable, since Logins holds that the only kind of normative reasons to F relevant for settling the question of whether one ought to F are premise reasons. Explanatory reasons are deemed relevant only to the issue of what to think about what we ought to do (p. 224f), and so would seem not to be normative for action at all.
Logins’ cited sources for the distinction predate the Gricean distinction between semantics and pragmatics, with the Erotetic Logic of Prior and Prior (from 1955) just predating Geach’s (1960) seminal charge of pragmatic fallacy against noncognitivism.
Logins seems (p. 169) to reject this on the basis of the factivity of explanation: ‘What explains p?’ presupposes that p, so (allegedly) it cannot be coherently asked except in No-Challenge contexts. But we can plausibly interpret ‘Why p?’ as ‘Why would it be true that p (if it were)?’ (‘What would explain p’) without presupposing p. (See Nebel, 2019 for examples of seemingly nonfactive explanations). It is also unclear that Reasoning views are any better off in this respect: does ‘Why p?’ as a request for an argument presuppose p? Arguments can be good in various ways (e.g. valid, persuasive) without being sound, but presumably ‘Why p?’ wouldn’t be satisfied by providing an unsound argument.
Advocates of Reasoning views lean hard on an intuition that we rightly deny these can be reasons for John to act. Contextualist Explanationism can accommodate this intuition: in the relevant contexts these facts cannot provide explanations for John, of why he should F. In this sense they therefore cannot be reasons for John, although they can still be reasons for others, for why John should F (Finlay, 2014: 104f).
Logins says just that appropriateness “depends on the further aims and presuppositions in a conversation”. Explanatory considerations seem to be the only source of these constraints presented in the critique of Evidentialist theories of reasons on p. 144f.
I agree with Kearns and Star that (given the right context), merely testimonial evidence that “S should F” can be correctly identified as a reason why “S should F”—due to a context shift. Evidence that S should F relative to a more complete information-base (e.g. objectively should) can be an explanation why S should F relative to a less complete information-base (e.g. subjectively should), and therefore appropriately cited as a reason to F precisely when it is such an explanation (see see Finlay, 2019).
Note ‘Why should I F?’ can itself be elliptical for this.
The different sentence position of ‘are’ is merely a grammatical transformation required for questions in English (satisfied by ‘do’ in the longer sentence): “Are dolphins not fish?” “No, dolphins are not fish”.
This issue has ramifications for Logins’ application of the Erotetic Theory to the debate over whether there are both epistemic and pragmatic reasons for belief. Epistemic reasons, he suggests, are premise-reasons why we should believe, while pragmatic reasons are explanations why we should believe. Unfortunately this solution runs into problems. Reasons for believing p should on Logins’ theory always be answers to questions of the form ‘Why believe p?’ But here we get a switch: epistemic reasons to believe p are analyzed instead as premise-answers to the question, ‘Why p?’ (i.e. ‘why is it true that p?’) (p. 216). This is not a normative question, and so its answers won’t (qua answers to this question, at least) be normative reasons to believe. Rather, they are (explanatory!) reasons why p. A premise-reason for believing p, according to Logins’ framework, should be a premise of an argument to the conclusion I should believe p. Evidence that p may well qualify as such a premise (given enthymemes such as I should believe what is true.) But the problem here is that on Logins’ account, pragmatic considerations should also be viable premises for such arguments. If considerations about the benefits of believing p can be legitimate explanations of why I should believe p (as Logins argues to secure the existence of pragmatic reasons), then it is hard to see how they could not also be premises of good arguments to the conclusion that I should believe p. Two moves one might be tempted to make: (i) one could claim that pragmatic premises are “inappropriate” in Challenge contexts. This seems implausibly ad hoc: people can and do make such appeals (e.g. Pascal: should I believe in God?). (ii) Logins emphasizes the rational character of epistemic normativity, so a distinction could be drawn between epistemic reasons as answers to ‘Why is it rational to believe p’, and pragmatic reasons as answers to ‘Why should I believe p’ (for a nonrational ‘should’). But this would abandon the Erotetic solution.
Although ultimately setting it aside, Logins also voices concerns about my “commitment to a version of” the Deductive-Nomological (DN) model of explanation (p. 113-5). However, ERE has no such commitment. In Finlay, 2019 I “assume” a version of DN as a “model”, acknowledged to be a rough approximation (see n20, n22), for the purpose of demonstrating that on a fairly orthodox, off-the-shelf theory of explanation, ERE straightforwardly accommodates the existence of various kinds of reasons that Explanationism’s opponents claim it cannot. Logins’ objections to the DN model turn on the problem of irrelevancies (explanatorily otiose propositions that DN predicts to be explanations), and my claim that this problem “doesn’t arise” for my purposes. What I meant there was that irrelevancies wouldn’t be a problem for my demonstrations that ERE accommodates the normative reasons it allegedly cannot, because none of those proofs intuitively involve irrelevancies. It is neither here nor there that DN faces problems with irrelevancies in other cases: the issue for ERE is whether all normative reasons are explanations in fact, not whether they are explanations according to DN.
Logins expresses puzzlement about how the constitutive ends of attitudes could “compete” for salience with the ends that we desire or prefer (pp. 121-2). My proposal here is simply that our association of attitudes with their constitutive aims can cause those “ends” (in my non-psychological sense) to be more conversationally salient for interpreting “reasons” claims than anybody’s desires or intentions. One possible way of fleshing out this story is to construe “constitutive ends” as constitutive norms such that it is impossible to (directly) form the attitudes except in response to “object-given” considerations. In this case it will normally be unhelpful to cite reasons of the “wrong kind”, making them conversationally infelicitous. (Sure, given the demon’s threats it would be good for me to admire the demon, but how is that of any use seeing as I can’t admire at will?) Logins here also presses an “incoherence” (category mistake?) objection: constitutive aims are definitions of attitudes, but definitions as abstracta can’t be ends (pp. 119–120). This seems to confuse a de re identification of the constitutive ends of particular attitudes, with a functional description of what it is to be a constitutive end.
(i) As observed, Logins’ reduction of constitutive ends to fittingness is contestable, and my account supports a reduction in the opposite direction, of fittingness to value relative to constitutive ends. This does require an independent analysis of constitutive ends, but there are promising alternatives. E.g. dispositional (and see n. 14 above). (ii) In the ERE analysis, the good for relation is intrinsic to the relation of being-a-reason-for, whereas the fit relation figures at most extrinsically, in making the relevant end salient, and each individual right-kind reason can be fully articulated without any reference to fit. This seems sufficient to qualify the analysis as value-based rather than fit-based. So I don’t accept the claim my explanatory analysis of normative reasons “appeals crucially to irreducible standards of fittingness” (p. 121).
Surprisingly, Logins on p. 181 identifies as a virtue of his Erotetic theory that it denies the existence of pro tanto reasons. Since he gives an account of pro tanto premise reasons on p. 184, I assume he means only to deny pro tanto explanatory reasons. This is still puzzling, given that earlier my theory’s ability to generate pro tanto explanatory reasons was acknowledged as a virtue, and we can give weighing explanations of normative facts.
Superficially, Logins’ examples conflict with this diagnosis. In the case he identifies as involving a right-kind reason for Tom’s worrying that his wife is on the two o’clock flight (because it is bad), the appropriateness of the worry is not plausibly contingent on its being uncertain whether she is on the flight. (To the contrary!) However, this just shows we need to take more care with such reasons-reports. While Tom may have a reason to worry because his wife is on the flight, strictly speaking this isn’t a reason to worry that his wife is on the flight. The correct (bad and uncertain) target of his worry is that his wife will be harmed. But if Tom knows that the hijackers have killed his wife, then worry would not be the fitting response!
Especially my 2014. In my (biased) opinion, no rival view has as much independent motivation, particularly from considerations of linguistics (“extreme” contextualism being almost universally taken for granted by contemporary linguists) and psychology.
I have also met the “challenge” (p. 118) of providing an “error theory” for pre-theoretic rejections of contextualism; see e.g. Finlay, 2024. Pre-theoretical obliviousness to context-sensitivity seems the rule rather than the exception.
Logins p.167f; see also Kearns & Star, 2009.
See Björnsson & Finlay, 2010: 15f.
Two supporting observations: (i) Anecdotally, I have found it harder to get students to recognize the idea of a normative reason than concepts like should and good. (ii) A clearly discriminated concept of a normative reason for action (fully disentangled from “motivating reasons”) was a relatively late development in philosophy, emerging around the early 1970s (e.g. Nagel, 1970) after years of being overshadowed by the concept of explanatory/motivating reasons for action.
Or (what comes to the same thing): of pro tanto/ceteris paribus should.
References
Broome, J. (2004). Reasons. In J. Wallace, M. Smith, P. Scheffler, & P. Pettit (Eds.), Reason and Value: Themes from the Moral Philosophy of Joseph Raz (pp. 28–55). Oxford University Press.
Bjornsson, G. & Finlay, S. (2010). Metaethical Contextualism Defended. Ethics 121(1), 7–36.
Finlay, S. (2001). What Does Value Matter? The Interest-Relational Theory of Value. University of Illinois PhD Dissertation.
Finlay, S. (2006). The Reasons that Matter. Australasian Journal of Philosophy 84(1), 1–20.
Finlay, S. (2014). Confusion of Tongues: A Theory of Normative Language. Oxford University Press.
Finlay, S. (2019). A “Good” Explanation of Five Puzzles about Reasons. Philosophical Perspectives 33: 62–104.
Finlay, S. (2024). Beyond Normativity: Can Metaethics Escape Samsara’s Wheel? In S. Kirchin (ed.), The Future of Normativity. Oxford University Press.
Geach, P. T. (1960). Ascriptivism. Philosophical Review 69(2): 221–225.
Hieronymi, P. (2005). The Wrong Kind of Reason. Journal of Philosophy 102(9), 437–457.
Kearns, S., & Star, D. (2009). Reasons as Evidence. Oxford Studies in Metaethics 4, 215–242.
Logins, A. (2022). Normative reasons: between reasoning and explanation. Cambridge University Press.
McHugh, C., & Way, J. (2016). Fittingness First. Ethics 126(3): 575–606.
Nagel, T. (1970). The Possibility of Altruism. Princeton University Press.
Nebel, J. (2019). Normative Reasons as Reasons Why We Ought. Mind 128(510), 459–484.
Schroeder, M. (2007). Slaves of the Passions. Oxford University Press.
Setiya, K. (2007). Reasons Without Rationalism. Princeton University Press.
Thomson, J. J. (2008). Normativity. Open Court.
Toulmin, S. (1950). An Examination of the Place of Reason in Ethics. Cambridge University Press.
Wedgwood, R. (2009). Intrinsic Value and Reasons for Action. Philosophical Issues 19(1), 321–342.
Acknowledgements
Thanks to Artürs Logins, Mitch Green, and Matt Lutz.
Funding
Open access funding provided by SCELC, Statewide California Electronic Library Consortium
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Finlay, S. Why Reasons Are Explanations. Philosophia (2024). https://doi.org/10.1007/s11406-024-00767-9
Received:
Accepted:
Published:
DOI: https://doi.org/10.1007/s11406-024-00767-9