The Heart’s Reasons: Intuition as an Authority in Practical Reasoning

 

By Ronni Sadovsky, Swarthmore College

 

 

Since the seventeenth century, Western philosophy has tended to polarize, not only mind and matter but also reason and feeling — to treat these as separate aspects of life, not intelligibly related.        

—Mary Midgley1

 

The heart has reasons that reason cannot know.                                       

—Blaise Pascal2

 

 

The dichotomization of mental life into opposed dominions of reason and feeling is a staple of our cultural inheritance. But although we speak of feelings as if they were irrational, we recognize that they often serve as the reasons for our actions. Whenever someone investigates a hunch or goes with her gut, she is letting her intuitive feelings guide her decision. Unlike desires or tastes, intuitions are not treated merely as idiosyncratic preferences for one action or another; rather, we treat the heart (and the gut) as a compass, responsive to facts about the world in a way that the mind often cannot be. Intuitive feelings are not merely related to practical reasoning; in fact, they are intrinsic to it. My goal is to make the relation between intuitive feeling and practical reasoning intelligible.

I will argue that intuition plays an indispensable role in proper reasoning — the role of an authority. Like an expert’s testimony, intuitive feelings serve as authoritative reasons to judge and act in a certain way. And like every authority, intuition has its limits. We advise our friends to listen to their hearts just as often as we instruct them to use their heads. Through my account of intuition’s authoritative role, I hope to explain why we are justified in heeding our feelings, and also to mark out the limits of their authoritative legitimacy.

 

The Structure, Justification and Scope of Authority

 

In giving this account, I will rely on the theory of authority given by Joseph Raz in his book, The Morality of Freedom. Raz carves legitimate authority into two constitutive properties: preemption and dependence. The directive of an authority is preemptive when “the fact that an authority requires performance of an action is a reason for its performance that… should exclude and take the place of some of the [other reasons relevant to the decision]”.3 For instance, if my mother asks me to be home by 7:00, she surely will have done so for some reason. But if I come home on time because I find that reason compelling, I will merely be complying with her request, and not obeying it. My mother’s directive is only authoritative if I come home on time because she told me to. When a directive motivates our action authoritatively, it preempts the reasons that motivated the authority to issue the directive in the first place.

How could the bare fact of someone’s being an authority justify our obedience to him? One central part of that justification is the second constitutive property of an authoritative directive: dependence. A directive is dependent when it is “based on reasons which already (1) independently apply to the subjects of the [directive] and (2) are relevant to their action in the circumstances covered by the directive” (Raz, 47; numerals added). My mother’s request that I be home before seven is dependent if, say, my parents plan to go out at seven and I am responsible for taking care of my little sister while they’re gone. My responsibility as a member of the household is a reason that applies to me independently of my mother’s request, so the curfew satisfies the first component of Raz’s criterion of dependence. The responsibility is relevant to the action that she asked me to perform because I cannot meet my responsibility if I don’t come home on time. Thus, the curfew satisfies the second component of the dependence criterion. Because it meets both the criterion of preemption and that of dependence, this curfew constitutes a legitimate exercise of my mother’s authority.

All authorities are fallible, and even the best of them will occasionally make mistakes in evaluating the reasons that apply to their subjects. Raz permits that an authority can fail to meet the criterion of dependence some of the time and still be legitimate, as long as it acts on dependent reasons often enough (47). But this caveat highlights an apparent problem arising from the conjunction of authority’s preemptive and dependent properties: if an authoritative decree depends on reasons that already apply to its subjects, why should the decree preempt those reasons? Why shouldn’t the subjects simply act according to those reasons that apply to them and ignore the authority altogether? In other words, what justifies the authority’s authoritativeness?

Raz’s answer to this question forms the crux of his explication of authority’s legitimacy: the Normal Justification Thesis. A would-be authority (whose decrees are preemptive and dependent) legitimately commands its would-be subject when

 

the alleged subject is likely better to comply with reasons which apply to him (other than the alleged authoritative directives) if he accepts the directives of the alleged authority as authoritatively binding and tries to follow them, rather than by trying to follow the reasons which apply to him directly.(Raz, 53)

 

If my mother’s authoritative decrees are dependent on reasons having to do with, say, my household responsibilities, then they are authoritative as long as obeying them improves my ability to respond to those reasons. In other words, her authority is legitimate if and only if I meet more of my household responsibilities when I let her rules guide my actions than I would just by trying to identify and meet my responsibilities by myself.

In one breath, the Normal Justification Thesis legitimizes authority and restricts its scope. Where Normal Justification obtains, the authority serves the interest we have in responding properly to the reasons that apply to us, and (barring extenuating circumstances) we are clearly right to obey its commands. Where Normal Justification does not obtain, however, the authority does nothing for us and cannot legitimately command us.

 

Challenging Authority

 

Here, Raz’s explanation of authority raises another apparent paradox. How can a subject determine whether the authority that commands her meets the criterion of Normal Justification? Her only option, it seems, is to use her own best reasoning to evaluate the authority’s decree. But if she obeys the authority only when it coheres with her own independent reasoning, then the Normal Justification Thesis is empty, and the authority is doing no work at all. The problem grows stickier still when we notice that even an authority that once met the criterion of Normal Justification may suddenly cease to do so. How can a subject ever justify her belief in an authority’s legitimacy?

Though it seems intractable, this problem actually poses no major difficulty for Raz. Consider the authority of a pocket calculator. When I rely on my calculator as an authority, I am much more likely to do arithmetic correctly than I would be if I were to work math problems out by hand. Hence, my calculator meets the criterion of Normal Justification. I do not need to check each operation by hand to be justified in this belief, and in fact, if my calculator were to suddenly start malfunctioning, I might accidentally continue to accept its authority after Normal Justification has ceased to obtain. On the other hand, if my calculator told me that the sum of two even numbers is an odd number, or if I suddenly found myself inexplicably getting questions wrong when I used that calculator to do math homework, I would have reason to believe that it no longer meets the criterion of Normal Justification. Such an event would drive me to check its calculations by hand and reevaluate the legitimacy of the calculator’s authority.

For Raz, an obvious error on the part of an authority gives its subject a good reason to investigate the authority’s performance with respect to the Normal Justification Thesis. There are also a number of other grounds on which an agent might challenge the legitimacy of an authoritative decree. For example, an authority may fail to respond to all the reasons it purports to preempt. If a judge systematically fails to consider divorcees’ assets in awarding alimony, the legitimacy of his authority can and should be called into question. A would-be authority can also be challenged for issuing a directive on the basis of reasons that are not appropriately related to the actions it requires of its subjects. Catch-22’s General Peckem and Colonel Scheisskopf are guilty of this abuse of authority when they order their fighter pilots to bomb towns in order to create bomb patterns that will look pretty in aerial photographs. An authority could also overextend its command by issuing orders in domains where it doesn’t belong. My mother’s authority would be subject to this objection if she tried to give my friends curfews. Even if an authoritative decree is legitimate, the reason it gives to act in a certain way may be defeated by a sufficiently strong reason to do otherwise. A professor may require that I write a paper by a certain date, but if some emergency befalls my roommate and she urgently needs my help, her needs may override the professor’s (albeit legitimate) demands.

My analysis of intuition’s authoritative role in reasoning will justify its legitimacy using the same criteria that Raz uses to justify the legitimacy of authority in general. With these justifications come the same limitations that apply to the authorities Raz describes. Like any authoritative decree, an intuition can be challenged for any of the reasons enumerated above, and should not be taken to guide action outside of the strictly limited scope of its legitimacy.

 

Practical Reasoning and Intuition

 

Before I can apply Raz’s theory of authority to intuition, I need to say a few words about the structure of practical reasoning. In Raz’s account of authority’s legitimacy, he makes reference to reasons that apply to subjects. I will use the same language — there exist reasons (in the world, as it were) which apply to us and which we ought to acknowledge and obey by performing certain actions. These reasons are different from the reasons that motivate our actions; whereas the reasons that apply to us exist independently of our acknowledging them, the reasons for which we act are psychological entities, which attain their status as reasons by contributing to our deliberation. These, in turn, are different from the deliberative process itself, which combines and balances all of our psychological reasons to act and produces an overall judgment about what to do. For clarity’s sake, I will use the term “f-reasons” to refer to the reasons that apply to us (factive reasons), “m-reasons” to refer to the reasons for which we act (motivating reasons), and “judgment” to refer to the output of the deliberative process. Note that, by the account given earlier, authorities depend on our f-reasons and provide us with m-reasons to act.

Where do feelings fit into this picture? Although our inherited wisdom tells us that feelings are opposed to reasons, I will treat feelings instead as a type of reason. Specifically, feelings can be either m-reasons or judgments. For instance, when my brother visited CalTech and Stanford and found that he felt more at home among Stanford students, his feeling was an m-reason to prefer Stanford. This reason responded to certain f-reasons having to do with his personality and various details of Stanford’s social culture, and it ultimately became a component of his judgment that Stanford was his best choice. That judgment also came to him as an intuitive feeling: my brother could not have given numerical weights to his various m-reasons to prefer each institution, nor could he have given a conclusive argument that explained why the balance of his m-reasons ultimately sided with Stanford. Nevertheless, when he considered all of the m-reasons to prefer Stanford alongside the m-reasons to prefer CalTech, Stanford felt right. His judgment was a matter of intuition.

Although all intuitive feelings are reasons, it is clear that not all reasons are intuitive feelings. Robert Solomon distinguishes two types of reasons: prereflective and reflective.4 Reflective reasons are the kinds of reasons philosophers usually deal with. When Socrates interrogates Euthyphro about his decision to prosecute his father, for example, he is asking for an account of Euthyphro’s reflective m-reasons: that his father killed a man unjustly, that it is pious to prosecute murderers, and so on. Like m-reasons, judgments can also be reflective. Utilitarianism gives its adherents a prescriptive methodology for reflective judgment: consider all the reasons that have to do with the action’s consequences, and weigh those reasons based on the principle of utility maximization. Whatever we may say about utilitarianism, it surely is the very model of reflective judgment!

Prereflective reasons, on the other hand, are intuitive — they are felt, rather than thought through. Just like reflective reasons, prereflective reasons may be m-reasons or judgments. When Raskolnikov confesses to murder in Dostoevsky’s Crime and Punishment, he is acting on a prereflective judgment, informed by prereflective m-reasons. He cannot put his m-reasons into words or articulate his judgment as a logical argument, but his judgment and his m-reasons are nevertheless responsive to the facts of Raskolnikov’s situation — his f-reasons.

The difference between reflective and prereflective m-reasons seems to come down to their referential content: whereas reflective m-reasons refer to the f-reasons they depend on, prereflective m-reasons do not. Contrast a reflective m-reason to avoid a used-car dealer (“he lied about the car’s gas mileage”) with a prereflective m-reason to do the same (“there’s something fishy about him”). Both depend on f-reasons to distrust the man, but the reflective m-reason names the f-reason it depends on, while the prereflective m-reason does not.

The distinction between reflective and prereflective judgment is similar: a reflective judgment is like an argument that explains why the m-reasons that the agent is considering support one action more strongly than any other. This is not to say that the agent who makes a reflective judgment needs to make such an argument in words, or even use words to think it through. If her judgment is to be reflective, the agent must reflect on the justification for deriving her conclusion from the m-reasons at hand, but she need not use any particular mental device to do it. Like a reflective judgment, a prereflective judgment follows from some chain of justification. But that justification is not available to the agent who makes a prereflective judgment.

The fact that prereflective reasons do not have referential content might make us wonder why we obey them. A reflective m-reason tells us why it should motivate our action — it says that such-and-such an f-reason applies to us. A reflective judgment also wears an explanation on its sleeve. But a prereflective m-reason or judgment can only prod the agent in one direction or another; the only justification it gives is “because I say so.” Perhaps it is this mysterious quality of intuition that has made some philosophers so suspicious of its legitimacy. At any rate, intuition’s prereflective nature invites us to treat it as an authoritative decree. Having distinguished reflective reasons from prereflective reasons, I am finally equipped to examine intuition using the framework of authority developed earlier.

 

Intuition’s Authority

 

In Raz’s theory, the decree of a legitimate authority is an m-reason that depends upon (and preempts) some set of f-reasons that apply to its subject. Its legitimacy depends primarily on the Normal Justification Thesis, which we can now rephrase using the terminology of reasons established above: an authority is legitimate if and only if its subject is likely better to comply with the f-reasons that apply to him if he takes its directives as m-reasons that motivate his action, instead of trying to discover those f-reasons for himself and thereby formulate m-reasons independently of the authority.

Phrasing the Normal Justification Thesis in terms of this ontology of practical reason should help to clarify the link between authority and intuition. This link is clearest with respect to prereflective m-reasons (though, as we shall see, it also has interesting implications for prereflective judgments). Just as I obey my mother’s authority because I will respond better to my preexisting f-reasons by heeding her than I would on my own, I listen to my heart in those situations where I have reason to believe that I will respond to my f-reasons better by doing so than I would have if I had analyzed my circumstances to pieces. The observation that intuition may be legitimated by the Normal Justification Thesis lends some promise to the enterprise of cloaking intuition in authority’s robes, but by itself it is insufficient to defend my thesis. To count as an authority in Raz’s sense, intuition must be dependent and preemptive. In what follows, I will show that dependence and preemption are constitutive properties of intuitions.

The dependence of intuition is what differentiates it from a preference or an impulse. A preference for year-round sunny weather over cooler climates is just an f-reason (which, when recognized by the agent, gives rise to an m-reason) to live in a warm climate. One need not like sunny weather for any particular reason; it can just be a taste one has. Intuitions are unlike mere preferences in that they respond to f-reasons that apply to the subject independently of the intuition’s existence. If my brother feels more at home at Stanford than he does at Cornell, his feeling is more than just a preference for Stanford, which he would weigh alongside all of his other m-reasons to go there or to Cornell; rather, it depends on a number of f-reasons independent of the feeling itself — reasons having to do with the sunny California weather, the southwestern architecture of the campus buildings, or the students’ relaxed social culture. Intuitive feelings are dependent reasons in that they respond to facts.

Intuitions also preempt those facts that they respond to. Suppose my brother feels apprehensive about applying to Dartmouth. He may realize that his apprehension arises from his belief that Dartmouth’s fraternities run the college’s social scene. But my brother cannot now consider himself to have two separate reasons to not to apply to Dartmouth: first, the extent of its Greek life, and second, his feeling of apprehension. My brother’s opinion of Dartmouth’s fraternities can explain his feeling, but it does not motivate his decision alongside his apprehension. If he takes his apprehension as a reason against applying to Dartmouth, it must be a preemptive reason. This is not to say that he must take his apprehension as a preemptive, prereflective m-reason. He may ignore it instead, and take Dartmouth’s Greek life as a reflective m-reason not to apply there. The point is that he cannot do both. An intuition cannot motivate action alongside the reasons it depends on. It can serve as a preemptive reason, or no reason at all.

Given that choice, when should an agent go with his gut, and when should he think things through? If intuitions really are authoritative, then Raz’s Normal Justification Thesis provides the answer to this dilemma: an agent should treat his intuitions as preemptive m-reasons whenever doing so improves his ability to respond to the f-reasons that apply to him. Hence, we should let our intuitions preempt the f-reasons they depend on when they are the best way to respond to those reasons, but we should ignore them if we have good grounds to suspect that reflective reasoning would serve us better. The kinds of grounds one can use to challenge any traditional authority (obvious failure to meet the requirements of Normal Justification, failure to respond to all the reasons it would preempt, and so on) can also undermine the legitimacy of someone’s intuitions. And just like any other authority, a prereflective m-reason can be defeated by any sufficiently strong opposing m-reason. Even if my brother feels certain that Stanford is the place for him, a meager financial aid package might force him to look elsewhere.

 

The Next Step

 

I have shown that the relation between a prereflective m-reason and the f-reasons behind it is just like the relation between any authoritative decree and the reasons that it depends on. But not all intuitions are prereflective m-reasons. Some intuitions are prereflective judgments. Can intuitive judgments also be authoritative?

Because Raz formulated his theory with external authorities in mind, his account deals only with authoritative m-reasons that preempt f-reasons. But my analysis deals with authorities internal to a single agent’s reasoning process, and therefore makes it possible for us to extend the domain of authority from m-reasons to judgments. Prereflective judgments are just as dependent and preemptive as prereflective m-reasons are: they depend upon and preempt the m-reasons that relate to the performance of some action. While judgments cannot, strictly speaking, meet the criterion of Normal Justification (because Raz’s Normal Justification Thesis is formulated in terms of f-reasons and m-reasons), they can meet an analogous criterion of justification: a prereflective judgment is legitimately authoritative if and only if the agent will be better able to assess the m-reasons relevant to the action he is deliberating about if he takes his intuition as an authoritative, instead of trying to weigh his m-reasons reflectively. Since this criterion is analogous to the criterion of Normal Justification, it seems to me that, together with the dependent and preemptive properties of intuitive judgment, it gives us sufficient grounds to treat intuitive judgments as authoritative.

This analysis of intuitive judgment shows how Raz’s theory of authority might be extended into new domains. My argument that prereflective m-reasons are authoritative showed that one part of a single agent’s reasoning can play an authoritative role with respect to another part of his reasoning. Extending this idea to judgment requires us to revise Raz’s criterion of Normal Justification, but also enables us to generalize his theory in an interesting way, making it more broadly applicable to the problems of practical reasoning. If my argument about the authoritative quality of intuitive m-reasons and judgments has been at all compelling, its success suggests that a broader treatment of authority’s role in practical deliberation may be a productive philosophical project.□

 

©Ronni Sadovsky, 2007