I.1 An Examined Life and Reasons

According to an old and venerable philosophical dictum that Plato attributes to Socrates, ‘The unexamined life is not worth living’. Stated as such, this dictum may sound a bit too harsh, especially for those who are incapable of undertaking the method of examination as understood by Socrates. Yet something close to its converse looks like a platitude. Figuring out and understanding how one should act; what to think or believe; what to regret; when, if ever, to get angry or feel guilty; when to be afraid; when to indulge in sadness and melancholy; or when to be grateful and happy are some of the fundamental questions that matter for everyone who aims to lead a meaningful life, a life worth living. In other terms, part of a meaningful life is to aim to figure it out and to arrive at a better understanding; and, in particular, to figure out what to do, or which attitudes to hold when, and to understand better the facts about oughts and shoulds that apply to us.

A related observation is that reasons are central to our lives. That is, it matters to us what reasons there are for us and others to act in certain ways or to believe certain things and to hold other attitudes. Reasons here are to be understood roughly as considerations that count in favour of some act or some attitude (in the contemporary philosophical jargon, these are normative reasons). When examining whether I should take my work with me on vacation, it matters for me what considerations count in favour of this option and what considerations count against it. That I will be able to make progress with my manuscript certainly counts in favour of taking the work with me. However, that I will miss out on spending fun time with my family counts clearly against taking the work with me. That your friend hates pistachio ice cream counts in favour of not buying one for your friend. That you see your partner’s car in the driveway counts in favour of taking it for granted – that is, believing that your partner is at home. Arguably, reasons thus understood matter for us precisely because they help us to figure out what to do and what attitudes to have, and because they help us to understand better why we should do certain things and why having certain attitudes is fitting in a situation. The consideration that taking my work with me on vacation will make me miss out on fun time with my family can help me to figure out whether to take the work with me or not. That your friend hates pistachio ice cream explains why you should not buy one for her.

The talk about reasons to act and to have attitudes is also popular in contemporary philosophy. Indeed, reasons seem to be the ‘new black’ in the so-called normative fields of philosophy – that is, in fields that are concerned with exploring aspects of obligations, values, and virtues, be they moral, political, aesthetic, or epistemic. Reasons prove themselves to be particularly useful for discussing meta-normative questions – that is, questions about the very foundations and principles governing oughts, values, and virtues. According to one prominent approach in recent meta-normative debates, the so-called reasons-first approach, reasons are indeed essential to understanding all other normative statuses and properties (see Scanlon Reference Scanlon1998; Schroeder Reference Schroeder2007; Skorupski Reference Skorupski2010; Parfit Reference Parfit2011). On this view, what one ought to do is, roughly, what one has most reason to do, what is good is what one has sufficient reason to value, what is admirable is what one has sufficient reason to admire, what is justified or rational is what one possesses reasons to do and so on. And crucially, reasons cannot be reduced, on this approach, to any other normative properties (some reasons-first proponents think that this doesn’t mean that they cannot be reduced to some natural properties – for example, one’s desires, though cf. Schroeder Reference Schroeder2007).

One problem with the reasons-first approach thus understood is that it lacks informativeness in characterizing reasons. It doesn’t say much about what reasons are. On the standard reasons-first view, reasons just are considerations that count in favour. But considerations that count in favour of an act or attitude just are reasons. No substantive, explanatory definition of reasons is possible, according to the reasons-first approach. However, such a lack of informativeness about reasons is problematic, since the view seems to end up in taking up arbitrary commitments when it has to distinguish among considerations that count in favour of some act or attitude in different ways. An already classical illustration of this is the problem of the ‘wrong kind’ of reasons (cf. Rabinowicz and Rønnow-Rasmussen Reference Rabinowicz and Rønnow-Rasmussen2004; see also Section 1.4 for more and for further references). A threat can certainly count in favour of admiring a despicable person. But is it a reason to admire the threatener? If the reasons-first proposal is taken literally and considerations that count in favour are reasons, then the threat has to be a reason to admire. And yet, it is not properly connected to admirability; the threatener is not admirable. Thus, if one takes this line of thought at face value, it seems one has to conclude that admirability cannot, after all, be reduced to reasons to admire. But then reasons are not fundamental in the normative realm – admirability is not explained in terms of reasons. On the other hand, one might try to avoid this conclusion by introducing different senses in which something might ‘count in favour’. On this view, there are genuine normative reasons to admire that are connected to admirability (whatever it amounts to exactly), and then there are the ‘wrong kind’ of reasons to admire. However, for such a move to be theoretically acceptable, one needs to provide independent grounds for such a distinction. But the reasons-first view’s lack of substantial definition of reasons prevents its proponents from providing such an independent motivation. Thus, the reasons-first view seems to face a dilemma. And at the heart of that dilemma is the inability of the view to provide a more substantive, informative account of reasons. Reasons are important, but we should be able to say something more about them than just that they are things that count in favour of acts and attitudes.

In the light of the worries affecting the reasons-first programme, reductive accounts of reasons are proliferating within contemporary literature. Indeed, I think it is not an exaggeration to say that nowadays an (academic) article per week is published on reasons. And some of the existing accounts are illuminating. Indeed, a starting point of the proposal to be developed in what follows is to observe some of the significant insights about reasons that we have received from the most promising existing reductive theories of reasons.

I.2 Reasons in Reasoning or Reasons in Explanation?

Once we agree that reasons are important and that the reasons-first approach should be our last resort in theorizing about reasons, the question that naturally arises is: but how do we go about building a reductive account of normative reasons? Where do we start? A reasonable place to start is to consider the role of reasons. Why do we need reasons? What roles do they play? Investigating central functions of our ordinary concept of reason to do something or to believe or to fear and so on might help us advance on this issue. So, what are the central functions of our ordinary, common-sense concept of reasons?

One central function of reasons seems to be to pick out elements that help us to figure out what we should do, believe, fear or what other attitude to have. In other terms, reasons seem to pick out premises in good reasoning/deliberation. As Paul Grice has put it, ‘Reasons […] are the stuff of which reasoning is made’ (Grice Reference Grice2001: 67).

Another central function of reasons appears to be to pick out elements that help us to better understand what we should or ought to do, believe, fear and so on. In other terms, reasons pick out considerations that contribute to explaining why we should or ought to do certain things or to have certain attitudes.

Now, as I see it, most, if not all, existing reductive theories of reasons can be classified as belonging to one of the two following general frameworks. On the one hand, roughly, there are the views that attempt to explain reasons by appeal to the role of reasons in good or fitting reasoning. Views belonging to this approach combine two elements in explaining reasons: reasoning and a normative property (e.g. goodness, fittingness). There is much to be said about this approach, most notably that it does seem to capture the figuring-it-out element that we commonly associate with reasons. Reasons are important to us, since in a sense they help us to figure out what we should do, what to believe, and what other attitudes to have. Reasoning-centred views bring to light this important aspect that we standardly associate with reasons.

On the other hand, roughly, there are views that propose to define reasons by an appeal to the role of reasons in explanations of why one ought to do certain things or to have certain attitudes or, alternatively, why it would be good for one to do certain things or have certain attitudes. Views belonging to this approach also combine two elements in explaining reasons: explanation and a normative property – for example, [facts about] oughts, goodness. Again, there is much to be said in favour of this sort of explanation-centred approach; not least that it does seem to bring to light the other fundamental aspect that we typically associate with reasons, an aspect that makes reasons important for us: that reasons help us to understand better what to do/which attitude to have or what would be good to do/which attitude to have by providing a [partial] explanation of why we ought to do certain things or have certain attitudes. We value reasons since they help us to understand better normative/evaluative facts that concern others and us.

Unfortunately, however, despite their promising and insightful aspects, both approaches also have serious pitfalls. The exclusive focus on the role of reasons in reasoning leads inevitably to overlooking the explanatory role that we commonly associate with reasons. And conversely, the exclusive focus on the role of reasons in the explanation of normative or evaluative facts (or considerations) leads to overlooking the importance of the role of reasons in good/fitting reasoning (towards appropriate actions and attitudes, or conclusions about what one ought to do/which attitude to have). In short, the main insights from both approaches are also their main weaknesses.

I.3 Our Positive Proposal: The Erotetic View of Reasons

In light of the problems with the two most promising reductive views, we might be tempted to draw a pessimistic conclusion that our concept of reasons is incoherent and that it is naïve to expect to find one single, overarching theory of normative reasons (compare to Wedgwood Reference Wedgwood2015). Such a temptation should be resisted, though. A key objective of the present work is to explain why. In short, according to the positive thesis developed in Chapters 5 and 6, there is an overlooked view of reasons that can integrate the lessons from reasoning- and explanation-centred views and can also explain what is the most fundamental common element that both of these views capture only partially. Thus, contrary to what a pessimist about reasons might think, there seems to be a unificatory and well-motivated account of normative reasons at a more fundamental level, such that the apparent failures of reasoning-only- and explanation-only-centred views of reasons are accounted for while their respective insights are well respected. The unificatory idea, simply put, is that most fundamentally normative reasons are appropriate answers to normative ‘Why F?’ questions. Normative ‘Why F?’ questions are of the form ‘Why should/ought one do this or that or have this or that attitude?’. Crucially, answers to normative questions, exactly like answers to any ‘Why?’ questions, come either as premises in arguments/patterns of reasoning or as elements of explanation. This is the essence of our positive view, the question-centred view of reasons, or, as we will call it, the Erotetic view of reasons.

If we need a slogan for the main thesis of the present book, it could be ‘no questions, no reasons’. In other words, we suggest that the point of normative reasons is to answer normative questions. That’s what reasons do; that’s what reasons are for. We need reasons insofar as we deem it important to reply to normative questions, questions like ‘Why do this? Why believe that? Why be angry?’ and so on. The view builds on insights from Pamela Hieronymi’s (Reference Hieronymi2005) view on which reasons bear on questions, as well as on insights from argumentation theory, informal logic, and linguistic observations about questions. Combining these two lines of insights together and reconsidering the role of reasons in good reasoning and explanation, we arrive at the following conclusion. Central functions of our ordinary concept of reasons to F, namely, the function of playing a role in good patterns of reasoning towards F-ing and the function of playing a role in a normative explanation (e.g. explanation of why one ought to F) are subsumed under an even more fundamental function, the general function of playing a role in answering the normative questions.

How is the function of playing a role in answering normative questions a more general function? The insight from informal logic, argumentation theory, and language use has it that it is a general feature of ‘Why?’ questions that they come in two varieties. Or rather, when we ask why such and such is the case, depending on the context of the conversation we may be asking one or the other of the following two things. We might be asking for an explanation of why such and such is the case. Or we might be asking for an argument for the claim that such and such is the case. And pace Hempel and the deductive-nomological model of explanation, we know that arguments and explanations are distinct. Why are dolphins mammals? This question may be understood as a request for an argument for the claim that dolphins are mammals, typically when we don’t yet know or believe the conclusion (appeal to the fact that they are warm-blooded would reply to the question on this reading). But it can also be understood as a request for an explanation, typically when we know or accept the conclusion but want to understand it better (an appeal to the evolutionary history of dolphins would constitute a reply to that reading of the question). Our main contention is that the same holds with respect to ‘Why F?’ questions. When we ask why should I do this or that or why should I believe, fear, hope that such and such and so on, we may be asking either for an argument to the conclusion that I should indeed act in these ways or have these attitudes or, alternatively, we may be asking for an explanation of why I should act/have the relevant attitude. Thus, the fundamental normative question may have either a premise in a reasoning reading or an element of an explanation reading. Sometimes our possible answers to these two readings of ‘Why F?’ questions will coincide, but not always. Insofar as reasons are properly understood as appropriate answers to normative questions, both reasoning and explanation functions of our notion of reasons are understood as two facets of the same, more fundamental phenomenon. Note also that the difference between our view and Hieronymi’s is that, at the end of the day, Hieronymi’s proposal looks very much like a variant of the Reasoning approach, since bearing on questions for her is roughly the same as figuring in a reasoning. The Erotetic view also does justice to the explanation-requesting reading of normative questions.

Reasons matter to us since they enable us to answer normative questions. But they always enable us to answer the normative questions in one or the other reading of ‘Why F?’ questions, either by providing a premise in a good argument/pattern of reasoning or by providing elements of an explanation of the relevant ought. The former helps us figure out what we should/ought to do, believe, and so on. The latter helps us better understand the shoulds and oughts that we may already suspect to hold. We cannot do without reasons insofar as we cannot do without arguments for and explanations of the relevant oughts or shoulds as possible appropriate answers to normative questions. We cannot do without reasons insofar as we cannot stop trying to figure things out and understand the normative facts that apply to us. Asking normative questions just is a part of who we are as agents aiming to live meaningful lives.

The dual life of the normative ‘Why F?’ question explains the duality of normative reasons as the possible appropriate answers to normative questions. Thus, the view to be developed here can both vindicate the insights of the Reasoning approach and the Explanation approach to reasons, and also explain in a theoretically motivated way why neither of these can be accepted as such.

I.4 What’s in the Book?

Here is a brief summary of the content of the chapters to come. Chapter 1 consists in some ground clearing. Here we consider some of the most prominent distinctions and clarifications about reasons – for example, the difference between motivating reasons and normative reasons. We also look (in a historically informed way) at some much-debated issues within the contemporary reasonology – for example, are reasons causes? are all reasons subjective? what is needed to possess reasons? – only to set these venerable debates aside in what follows. We also present tenets of the reasons-first approach and review the much-debated ‘wrong kind’ of reasons problem for the reasons-first approach. Chapter 2 begins a proper investigation into reductionist theories of normative reasons. We begin in Chapter 2 by considering the advantages and problems of the reasoning-centred approaches to normative reasons. Chapter 3 then focuses on the explanation-centred approaches. Chapter 4 then examines the so-called Evidence view of reasons, according to which reasons are evidence that one ought to do something/have an attitude. One might think that the Evidence view is a third possible reductionist account of reasons and doesn’t fit into our overall classification. The main suggestion in exploring that view is that in the most plausible form of the Evidence view, it reduces to a version of the Reasoning view and as such inherits some of its most problematic aspects. Chapter 5 begins developing our positive proposal, the Erotetic view of reasons. Chapter 6 then develops the view further by showing how it can be applied fruitfully to make progress in one notorious debate in epistemology, the debate concerning the possibility of pragmatic reasons for belief. The Erotetic view can be applied to show that both pragmatists and evidentialists can be right within this debate since there is a clear sense in which there can be pragmatic reasons to believe, and a clear sense in which there cannot. If the proposal is on the right track, then this provides an additional consideration in favour of our new proposal.

Reasons matter for us, in ordinary as well as theoretical contexts. Building a viable theory of reasons helps us to better understand some, and perhaps the most fundamental, of our normative concerns. It may even help in complying with Socrates’s dictum about leading an examined life. Advancing this task is what I hope to do in the chapters to come.