Taking Heed: Between Consideration and Deference

I. BACKGROUND: RESPECT, AUTHORITY, AND DEMOCRACY

A few clarifications to begin, first regarding what “authority” in this context means to capture, second about how I understand respect to connect to authority and authority to democracy, and third about the scope of my discussion.

Regarding “authority,” my central concern here is with conceptions of democracy that treat authority as an input into the democratic process—as a kind of influence owed to citizens in deciding matters of common concern, sufficient for them to count as co-authors of those decisions, throughout the political process. This kind of authority may well be connected to that of the outputs of democratic procedures, i.e. legitimate law, and in turn to the special authority of the democratic state.⁷ But it is not my aim to account for the latter here.⁸

Regarding how respect connects to authority and in turn democracy, some maintain that respecting others (as autonomous and/or equals) involves⁹ granting them certain kinds of authority in one’s deliberation, in particular by treating their preferences, practical judgments, or expressions thereof¹⁰ about some matters as reasons for action.¹¹ This kind of authority can ground norms against paternalism¹² and/or interference more generally¹³—in other words, liberal claims to decide for ourselves how to live—but it is also deployed to comprehend respect’s claim to a voice with others in deciding matters of joint or common concern.¹⁴ Herein lies one potential way that respect connects to democracy, particularly if the latter is understood as a condition in which citizens have the relevant kind of authority with each other and the state in deciding matters of common concern.

Why, then, go for such an account of democracy? As I read the literature, one significant appeal of such accounts is their ability to compellingly describe democratic influence in a variety of settings, formal and informal, including those in which voting seems inappropriate. In Democratic Autonomy, for instance, Henry Richardson develops such an account in order to theorize the realization of democratic autonomy in the administrative state.¹⁵ Respecting citizens’ autonomy, he argues, requires equal consideration of their views, which involves not just soliciting their input in deliberation but also treating them as—he quotes Rawls—“self-originating source[s] of claims” which “carry weight of their own,”¹⁶ such that “what individual citizens think should be done intrinsically matters as a consideration bearing on what ought to be done.”^17,18 In Democratic Equality, James Lindley Wilson argues that political equality in general is best understood as a condition in which each citizen is “equally entitled to render authoritative judgments as to how to organize and regulate all citizens’ common life,”¹⁹ not least because such an account captures the realization of equal respect throughout the extended political process—i.e. not just in voting but in agenda setting, deliberation, representation, etc.²⁰ Wilson terms the kind of authority to which citizens are entitled “appropriate consideration,” and he defines it in Razian terms: to give others appropriate consideration is to treat their political expressions as preemptive reasons to rationally attend and reply thereto and, sometimes, unprotected reasons to do as they say (more on these terms below).^21,22

Finally, regarding scope, Richardson and Wilson both maintain that citizens have claims to consideration from each other and from officials, which, accordingly, manifest in different democratic practices and institutions in different ways.²³ I will for my part mostly set aside the latter issue and, more generally, the complexities of democratic institutional design, focusing instead on what I take to be the basic normative claim that Richardson and Wilson share: that citizens are owed authority over matters of common concern and that, should they receive it robustly and with assurance throughout the political process, this will suffice to realize democracy per se.

That said, I follow Richardson and Wilson in thinking that both citizen-citizen and citizen-official relations may be governed, at least partly and at some high level of abstraction, by the same basic normative claim to authority. This follows from the idea of respect sketched above, which makes no distinction between citizens and officials but rather is owed to and from persons in general, as makers and receivers of claims. It also seems to follow from the general idea of democracy as equal self-rule, for if citizens are understood as equal co-authors of decisions that meaningfully affect each other’s lives, then it seems plausible that they could have the same basic claim on each other in this capacity as they have on officials. We can, of course, imagine different democratic systems fulfilling this claim at different points and in different ways. But we need the right account of authority to begin.²⁴

II. THE DEMANDINGNESS DILEMMA

Consider, then, Raz’s account, according to which A has (practical) authority with B in some context in case A’s directives (“φ!”) to B in that context constitute preemptive reasons for B to do as directed (φ).²⁵ A preemptive reason is a combination of two reasons: (1) a first-order content-independent reason for B to φ²⁶ and (2) a second-order exclusionary reason which excludes from B’s deliberation his other reasons for and against φ-ing.²⁷ Raz later calls this a “protected” reason.²⁸ I’ll help myself to both terms, understanding preemptive reasons as one kind of protected reason, but identifying a novel kind I call “resistant.”

For now, Razian authority obviously cannot be the kind that constitutes democratic citizenship. If each citizen has Razian authority, Citizen A’s expressions of his preferences preempt B’s, but B’s also preempt A’s. This, obviously, is not what democracy involves. It would seem to make social choice through voting impossible, since it’s not remotely clear that preemptive reasons can aggregate. It would also undermine deliberation; A’s expression of his preferences would often preempt other considerations that might arise in deliberation about, e.g., which policy to choose. So citizens’ authority understood in preemptive terms is conceptually unworkable; we cannot each have preemptive authority over matters of concern to all; claims to this authority are not compossible. Unsurprisingly, democratic life is poorly understood as a series of commands.²⁹

The most obvious way to weaken Razian authority is to abandon exclusion—to see citizens’ “authority” as the normative power to give mere pro tanto (i.e., weighty-but-unprotected³⁰) content-independent reasons. This, though, will not do. Raz rightly maintains that requesting creates weighty-but-unprotected content-independent reasons for others to do things: I can request a glass of water or a glass of cola, thereby giving you a reason to get me one or the other.³¹ But, of course, the normative power to request is insufficient for democratic citizenship. That I have a claim to make requests of my king does not require that he be anything but. Through requesting φ of my king, I could at most change what is right for him to do by changing the balance of his reasons for action (presumably in concert with many other citizens). This might in one, though I think fairly minimal, way displace his judgment, since he should now treat my judgment in favor of φ-ing as a reason to φ even if he would have otherwise judged he had no reason to φ.³² But it does not shape his deliberation any further, since it remains up to him to determine the weight of this reason vis-à-vis the others he judges himself to have. Thus it does not, in any deep way, change who decides what is to be done. It does not even require any kind of positive responsiveness to citizens’ political expressions from the state.³³ These are significant problems, for democracy is not essentially about what is right substantively, but procedurally, in who decides, and it characteristically involves positive responsiveness to citizens at some point in the political process.

The strategy of downgrading the authority of citizens to the power to give unprotected content-independent reasons does, though, exemplify a more general strategy for weakening authority which is worth exploring. It fiddles with the kind of reason which authorities can give, where their being able to give it constitutes authority. Call it, then, a “kind-of-reason” strategy—it’s the one I’ll ultimately pursue. But first note that other authority-weakening strategies are available: restricting the scope of the relevant authority, and what we can call a “reason-for-what?” strategy.

Scope restrictions are not especially relevant for my purposes; I’m interested in the kind of authority citizens have, not its limits.³⁴ But for an example of the reason-for-what? strategy, suppose that we take citizens’ expressions of their preferences not as coming with preemptive reasons to act as they say, but rather for something else, like giving them rational attention—actively listening to and considering their perspective—or an objectively acceptable justification for decisions ultimately made.

This also seems too weak for democratic citizenship. Obligations to respond appropriately to citizens’ authority thusly understood could in principle be discharged in a monarchy that institutionalized consultative or justificatory practices. While such a government may be “decent”³⁵ or even liberal,³⁶ it is not democratic; the people do not decide for themselves how to live. A hybrid view like Wilson’s, on which citizens can issue content-independent reasons for action and/or preemptive reasons for attention and justification, is also too weak.³⁷ This kind of authority might be satisfied in a consultative/justificatory monarchy in which the king and his officials take their subjects’ expressions of will as issuing requests.^38,39 So, insofar as democracy involves a claim to take part in deciding political matters, and, relatedly, to positive responsiveness from the state, this kind of authority still seems too weak.

In a different way, this reason-for-what? strategy is plausibly overdemanding, too, in a way that undermines its plausibility as an account of what citizens are owed in deliberation specifically. For instance, it may well be that citizens and/or state officials owe certain others greater attention on the basis of, say, their special stakes in a particular issue.⁴⁰ Conversely, it is conceivable that some views are so profoundly unreasonable⁴¹ or even hateful that others have claims against giving them attention.⁴² But it is hard to see how these claims can be satisfied if all citizens’ expressions of their perspectives constitute preemptive reasons to attend. This is not because preemptive reasons to attend necessarily constitute preemptive reasons to attend equally. Rather, it’s because preemptive reasons to attend exclude those considerations that could rationally guide determinations of which perspectives are owed more or less attention. If, say, A clearly has high stakes in some policy, but both A and B’s expressions of their preferences constitute preemptive reasons to attend, B’s expression preempts considerations like A’s high stakes.

Put more generally: preemptive reasons to attend and reply/justify have the troubling effect of severing what I’ll call the “motivation-worthiness” of another’s claim or its content—i.e., the degree to which one should be moved to act as claimed—from its “attention-” and “reply-worthiness.” Obviously, attention-/reply- and motivation-worthiness are not 1:1. Respect can require that we rationally attend to mistaken perspectives, for instance. But attention should not be insensitive to motivation-worthiness, either. Attending is the first moment in valuing,⁴³ and attention is a precious and increasingly scarce resource.⁴⁴

So, the Razian conceptual tools get stuck between the preemptive authority of a command (either to act or to attend) and the merely content-independent normative power of a request. Thus they mishandle both democratic responsiveness and democratic deliberation. At the level of natural language, deference and consideration (and obedience, requesting, listening, etc.) are all flawed or incomplete concepts for capturing democratic citizenship. This is not, of course, to say that we don’t ever owe others deference—we do, for instance, about matters that are solely “theirs to decide”—or that mere listening and consideration have no place in a healthy democracy. But, clearly, none of these concepts alone are sufficient for democratic citizenship, and it seems doubtful that they are jointly sufficient.

At very least, I hope I’ve said enough to show that democratic theorists would benefit from a richer conceptual vocabulary. At the level of natural language, I favor the concept Anderson deploys but does not elaborate: heed. At the level of the theory of authority, capturing the meaning of a claim to heed will require identifying novel tools: the related ideas of a resistant and an insulating reason.

III. “HEED” BETWEEN “CONSIDER” AND “DEFER” (AND “LISTEN” AND “OBEY”)

As with any concept, heed’s natural meaning varies. But I think there is a distinct and self-contained concept, here, which is not reducible to either consideration or deference.⁴⁵ To be clear, I am not endeavoring for an analysis so much as trying to impress upon the reader the concept’s natural distinctiveness. To that end, I’ll focus on what I take to be the paradigmatic usages of considering, heeding, and deferring and simply flag felicitous uses that do not seem to perfectly cohere with my account. I am also not yet discussing claims to heed, or, for that matter, any kind of authority. Just as A can consider what B says or defer to B without B having a claim to this kind of treatment, so too can A heed B absent such a claim. The nature of claims to consideration, deference, and heed—thus of the kind of normative power or authority each claim encodes—will turn on the nature of the underlying dispositions.⁴⁶ We need to get clear on these first. In Section IV, then, I’ll discuss cases in which Pete does have practical authority with Sam, in the form of claims to deference vs. heed.

To begin it will help to fix terminology already introduced. Let practical deference, then, paradigmatically involve doing as another says (claims to which involve taking what another says as a preemptive reason to do it), and let consideration paradigmatically involve giving careful rational attention to what another says (claims to which involve taking what another says as a preemptive reason to rationally attend/reply).

To see that heeding differs from both consideration and deference, consider now two hypothetical cases of warning—i.e., practical speech that regards one’s own, rather than the other’s, interests.⁴⁷ First,

Deep Wood: Suppose Sam is heading down a mountain path towards the Deep Wood, listening to music in his headphones. Up the path from the Wood, he passes Pete. “Turn back!” Pete calls. Sam listens to and carefully considers Pete’s warning, but it leaves him cold. There’s a highly reviewed bakeshop just through the Wood, and he really wants to try it. He puts back on his headphones and saunters into the Wood.⁴⁸

Though Sam considered Pete’s warning, I think it’s inapt to say that he “heeded” it. For one thing, his deliberation seems heedless: Pete’s warning does not move him beyond calling his rational attention; he rationally attends to it but sets it aside. Relatedly, Sam’s behavior seems heedless, too: he does not only act contrarily to Pete’s advice but does this in a way that expresses that it did not move him.

What’s missing, it seems to me, is a motivational ingredient. Had Pete’s warning moved Sam and had Sam’s behavior expressed this, I think it would be much more natural to say that he “heeded” Pete. If so, then perhaps to heed another is to defer to them practically or in a practical context. But this is not right either. To see heed’s motivational aspect and also that heeding is not quite deferring, consider

Dark Wood: Sam is heading towards the Dark Wood; “Turn back!” calls Pete. Sam considers Pete’s warning and is moved by it. He takes off his headphones, out his flashlight and gun, and proceeds cautiously down the path.

Obviously, Sam has not deferred to Pete. But it is natural enough to say that, in this case, he has heeded Pete’s warning. Granted, supposing that Sam got in trouble somewhere in the Wood, Pete might complain, “he should have heeded my warning!”, suggesting that he did not in fact do so. But supposing Sam was left unscathed thanks to proceeding carefully, Pete might naturally take credit: “it was good that he heeded my warning!” And notice that, although the natural language seems to pull in different directions here with “heed,” it seems much less controversial to say that Sam “took heed” of Pete’s warning.⁴⁹ So I think there is a sense of “heed” according to which Sam could proceed cautiously and still be said to have “heeded” Pete’s warning, and that’s the one that interests me here.

Now, obviously, Sam’s heeding Pete can involve doing what Pete says—he certainly could have heeded Pete’s warning in turning back—but it seems this is not necessary for heed. It’s not sufficient, either. Obeying and deferring both involve turning back but needn’t involve heed. For “heed” typically suggests rational activity that is characteristically absent from obedience and typically absent from deference.

Of course, deference, more directly than obedience, can be performed willingly and from reason. But when I decide rationally to defer to expertise, say, I am deciding to go by the expert’s judgment, not my own. Deference may also involve the kind of rational activity required to interpret a directive or determine how to best adhere to it in a particular case. But unlike “heed,” it strains the natural concept of deference to say that I defer to a directive when I behave in a way that directly contradicts it.⁵⁰ Moreover, though deference can involve these limited kinds of rational activity, it needn’t involve them; one doubtlessly can defer in a way that’s passive, and, to my ear, “defer” typically has a passive sense.⁵¹ By contrast, heeding seems to constitutively involve some kind of rational activity of one’s own. It seems to involve feeling independent pressure from others’ judgment or speech, but making something of that pressure oneself, rather than acceding—or, to again use the word that is quite natural here—deferring to it.

To lend this point more concrete support, consider “work-to-rule” strategies for industrial action.⁵² Instead of striking outright, workers follow their employer’s rules and managers’ directives to the letter, usually causing significant disruption. In a famous action in the 1930s, French rail workers, legally prohibited from striking, worked-to-rule by adhering rigorously to rules requiring them to inspect each bridge before passing over it. The trains ran catastrophically late. In the UK more recently, postal workers—who would typically arrive to work early and leave late, use their own cars to deliver mail, and take out mailbags heavier than the safety guidelines allowed—forsook these heroics for working-to-rule to protest layoffs. Now in practice, work-to-rule strategies often involve the refusal to work more than is required to accomplish the goal at hand and sometimes veer into sabotage. But as the historical examples make clear, the core strategy does not involve disobedience but rather unthinking obedience to rules taken to the letter, as if without any understanding of their practical purpose. Thus, while it seems natural to say that workers working-to-rule “obey” the rules and “defer to” the directives in question, it is much less natural to say they “heed” or “are heedful of” them.

This difference between heed, on the one hand, and obedience or deference, on the other, is also borne out in their typical objects in natural language. As noted, we usually say that we “heed” or “take heed” of things like warnings, advice, and instructions, and also things like calls (as in, “nobody heeded his call!”), claims, and complaints. By contrast, we don’t as often say we “heeded” a command, much less an order, and we certainly don’t say we “took heed” of either.⁵³ We could say this, but it seems to me somewhat beside the point. Unlike your advisor, who might be glad you took heed of her advice, you probably do badly to try to placate your commanding officer by telling her that, no, you really did take heed of her order. That you did does not matter, because taking heed of it is not the essential thing to do with the order. What matters is that you obeyed or disobeyed it. By the same token, we don’t usually say that we “obeyed,” and it rings a bit oddly to say that we “deferred to,” a warning, piece of advice, set of instructions, or a call, claim, or complaint.

It is highly telling, here, that warnings, advice, etc., all engage the practical capacities of those who receive them in a way that commands and orders do not. The idea that respect for persons in receiving their claims—recall Anderson—would be similar is intuitive and instructive. Respect is an attitude towards other persons as persons, as rational natures like us. It would, I think, be odd indeed for that attitude to paradigmatically require our going rationally offline by simply deferring to those we’re supposed to respect. I am not saying respect never demands this of us, but rather that it is much more natural to think it paradigmatically involves engaging another as a rational nature via that capacity in oneself. Herein, I think, is one draw of “heed.”

But I am getting ahead of myself. Heed, I hope I’ve shown, differs not only from deference but also from mere consideration. Unlike with deference, Sam could still continue into the Wood while heeding Pete’s warning against doing so. But unlike with consideration, he could not continue into the Wood as before—headphoned, sauntering. His heeding Pete’s warning turned not only on his paying attention to it but also it moving him to do as Pete says, albeit not necessarily decisively. The challenge is to capture this.

Before trying to do so, a few notes. First, recall that heeding another can involve doing as they say—just not passively, letting their practical judgment preempt one’s own. To keep this terminologically distinct from deference, then, let me distinguish decisive from non-decisive heed. We heed another decisively when we heedfully do as they say, but we can heed another non-decisively without doing just as they say.

Second and relatedly, there is a general question about whether heeding necessarily involves behaving in a way that positively reflects the content of the speech one heeds, or, further, doing so legibly. For suppose Sam received Pete’s warning heedfully—that is, in a heedful attitude—without his behavior expressing this at all. Did he heed Pete? For the inverse, what if he behaved in a way that seemed to express heed but without experiencing the real attitude—if, say, he merely humored Pete?⁵⁴ This is a common set of questions about practical attitudes, which are typically understood to include dispositions to act in ways that express the relevant attitude. We can, here, simply distinguish the attitude of heed from heedful action, the latter of which I’ll take to normally express and to be from the attitude of heed.⁵⁵ Accordingly, merely humoring another—while perhaps manifesting a worthy concern to express heed one does not actually experience—is insufficient for heeding them.

When I say the state owes citizens heed, I’ll be referring not only to the heedful attitude (which is in any event motivational) but also heedful action. This is typically what we mean when we say that persons are owed, e.g., respect: not just a free-floating attitude but (one which includes) a disposition to treat others respectfully.⁵⁶ Heedful deliberation among citizens, by contrast, need only internally manifest heed’s deliberative dimension, since deliberation aims to settle the question about how to act (collectively). Citizens may then express their mutual heed in action down the line, in, e.g., the way they respond to public deliberation.⁵⁷

IV. HEEDING WARNINGS VS. HEEDING CLAIMS

The preceding section yields a few aims for a conception of heed. First, heeding is modally open relative to deferring in that Sam can heed Pete’s “φ!” without φ-ing, but by acting in ways that positively reflect Pete’s “φ!”. Second, relatedly, heeding must be rationally active whereas deferring is paradigmatically passive.⁵⁸ “Heed” implies no such passivity at any point. Third, heeding involves more than just giving rational attention—it also involves being moved and thus disposed to act. So it differs from “consider” on its natural meaning.

So, then, how should we understand Deep Wood and Dark Wood? As a first pass, consider an epistemic reading of these cases: perhaps Sam is treating Pete’s warning as information that is practically relevant, but only in light of his prudential reasons to avoid harm. This information isn’t only evidence in favor of Sam abandoning his plans but also proceeding cautiously if he does not.

This may be alright for understanding heeding warnings, advice, and so forth. When the reasons at hand are B’s (rather than when A is expressing his own claims), B is permitted to treat what A says just as information. This is what is prudentially rational for him, and it’s only his reasons that are relevant, so prudential rationality seems to almost exhaust what matters normatively.⁵⁹ But the same cannot be said of cases in which another’s reasons also bear on one’s choice. In those cases, treating what another says merely as practically relevant information often strikes us as disrespectful.⁶⁰

Suppose Sam offers to get Pete something from the bakeshop and Pete says he wants the croissant, yet Sam somehow knows that Pete prefers the Cardamom Bun. It seems to me that Sam should (1) suggest the cardamom bun but (2) get Pete the croissant if he insists. But epistemic normativity—thus a view on which what Pete says is just information about what independently matters—strains to explain why. For one thing, it’s not clear that Sam should treat what Pete says as evidence at all, since he knows Pete’s true preference contradicts it. And even if he should, say, downgrade his credence for the cardamom bun, it’s unlikely he should downgrade enough for this to tip the practical scales in favor of the croissant in light of all of his other evidence. In any case, respect doesn’t obviously require any of this. What it does obviously require is that Sam get Pete what he says he wants. This would suggest Pete’s ask is poorly read as just practically relevant information. It also, more importantly, has irreducibly practical authority.⁶¹

In Cardamom Bun, it is plausible that Sam should simply defer (maybe decisively heed) Pete—if what Pete will eat is all that’s the matter, the matter is Pete’s to decide.⁶² But we can also disrespect people by treating their claims or other self-regarding practical testimony merely as information even when it bears on matters of joint or common concern, in which the practical matter at hand is not solely theirs.⁶³ Suppose, for another variation, that Pete is growing delicate orchids along the path through the Forest of Flowers and is reasonably worried Sam will damage them. In this case respect plausibly demands that Pete have practical authority of some kind either directly with Sam or over the norms governing cases like these. But it is implausible to understand this authority in preemptive terms. Pete is owed a voice, not a veto, not least because the same thought applies equally to Sam, who has reasons of his own to proceed into the Forest. In this case the notion that Pete is owed heed is attractive. Pete’s claim should move Sam to deliberate with Pete to find the best solution, but also to modify his intentions to better accord with it—at least proceeding down the path with due care. If he does not, he has failed to heed Pete.

So instead of trying to capture the wide range of cases of heed all in terms of a response to practically relevant information given by another, it makes better sense to distinguish cases of epistemic from practical heed, and, correspondingly, heed in forming beliefs vs. intentions. I confess this distinction, though ubiquitous in the literature on authority, is murky. But addressing the challenges in drawing it cleanly is beyond the scope of this paper. For now, then, I’ll restrict my focus to cases like Forest and Bun—cases, that is, of heeding practical authority.

V. HEEDING PRACTICAL AUTHORITY (OR, NOVEL VARIETIES OF PROTECTION)

Recall, now, the strategies for weakening Razian authority discussed in Section II. The best strategy for capturing “heed” is the kind-of-reason strategy. But simply downgrading the preemptive authority of a command to the content-independent power to request will not do. Heeding another’s claim involves treating it as more than a weighty-but-unprotected content-independent reason, but less than a preemptive one. What we need is a reason that is protected, but less aggressively than the preemptive one is—via exclusion of other reasons.

It helps here to take a step back. With the idea of exclusion Raz identified a novel kind of defeat among reasons—R can exclude S rather than simply outweigh or undercut it.⁶⁴ They say the best defense is a good offense, but even so, this seems like a fairly aggressive way to protect a reason given by another’s utterance in one’s deliberation. And this aggression has a number of troubling effects when trying to capture the kind of authority that constitutes relations of mutual respect and, in turn, democratic citizenship. But I think a more defensive kind of protection is available.

Heeding another’s practical authority, I’ll contend, is best understood as involving treating another’s practical speech as constituting a kind of protected reason I call “resistant,” which, unlike merely content-independent but unprotected reasons resists defeat by shaping the deliberation of the receiver, but, unlike a preemptive reason, does not do so by excluding other reasons from their deliberation. This resistance is achieved in the following way. Instead of traveling with an exclusionary reason, the first-order content-independent reason given by the authoritative utterance travels with a positive second-order reason—an insulating reason—against (typically rebutting) defeat by the other reasons at hand in the receiver’s deliberation.⁶⁵ Thus, the reason to act given by the authoritative utterance is protected via insulation rather than exclusion. This insulation is rarely complete—it is not as if the protected reason cannot be defeated, only that it is harder to defeat it or defeat it fully.

If such insulation is really on offer, it would achieve the modal openness and rational activity characteristic of “heed” while demanding more than rational attention. So, is it? How, in other words, might a first-order content-independent reason resist defeat? What kind of reasons might “insulate” it other than an exclusionary one?

There are a number of subtly different ways of reading the cases above, which we may take to correspond to different accounts of insulation or ways that a first-order content-independent reason R to φ might be insulated. Consider the following effects and insulating reasons that would induce them.

• (1) Sticky salience: R travels with a second-order insulating reason requiring that it remain practically salient even after defeat in some instance of deliberation about whether to ϕ, thereby achieving a kind of temporal protection for the utterance past defeat.

This differs from a preemptive reason’s effects, which are in the first instance to settle the matter about whether to φ.⁶⁶ Two closely related kinds of insulation may offer different ways of understanding what’s going on in Forest of Flowers and other cases of claims to heed:

• (2) Instrumental transmutation: R to φ travels with an insulating reason to treat R as an instrumental reason to ψ in a way maximally compatible with R after defeat about whether to φ. Put differently: R to φ becomes a reason to ψ φ-wise. Thus R gets a second life instrumentally, bearing on our means rather than ends.⁶⁷

• (3) Reverberation: R travels with an insulating reason to find and act on ends nearby our current ones that are better compatible with the reason given by the other’s utterance. Thus R to φ can bear on our choice of ends, not just means, without preempting the other considerations bearing on our choice.^68,69

Both instrumental transmutation and reverberation lean on an idea of compatibility that needs more detail. What guides judgments of compatibility? In Forest (as in Deep and Dark Wood), it seems that Sam recognizes that Pete has a reason for saying what he says and acts in a way that is positively responsive to that reason rather than a reason to act just according to the direct content of his utterance.⁷⁰ Compatibility judgments thusly guided seem respectful qua rooted in recognition of and a positive disposition towards the rational nature of the other person. That said, insofar as Pete is available to speak for himself, it may also be disrespectful to avoid consulting him further. In any event, instrumental transmutation and reverberation thusly guided seem to constitute deliberation-shaping effects less demanding and constraining than the preemptive one.⁷¹ Importantly, although these different effects could all explain how Sam can proceed cautiously and count as having heeded Pete, this will not always be so. (1)–(3) pick out genuinely different deliberative processes which may express in different courses of action.⁷²

In the background of this exercise is the hunch that defeat (of one reason by another) is complex: not a singular or on/off condition, but rather a cluster of effects which may come in degrees. A reason can be partially defeated if one or more, but not all of, the normal effects of defeat are present. (1)–(3) in particular traded on the idea that a reason can be defeated in deliberation about whether to φ without losing broader practical salience or effects, e.g. without being defeated in deliberation about how to φ. Classical Razian authority tends to occlude this possibility, for the kind of authority it most vividly captures and is designed to capture is that of a valid command, requiring one to do what another says. This kind of authority is end-oriented and characteristically settles in the affirmative the question of whether to perform the commanded action. Once we scrapped that strong condition in trying to capture the intuition that Sam could in principle heed Pete’s claim while still continuing down the path, we were left wondering how to capture the requirement that Sam be positively moved and responsive to Pete’s practical speech other than via obedience or deference. The idea of a resistant reason helps do so.

VI. CONCLUSION

In general, the idea of a resistant reason helps to capture claims to heed better than the received Razian resources. It does so in a way that illuminates the concept’s value to democratic theory. Since resistant reasons do not preempt other reasons but instead leave them practically intact, they do not strain to capture heed’s openness and rational activity. This explains how each citizen can enjoy a claim to heed over matters of joint and common concern, avoiding the incoherent condition in which their political speech preempts others’.

Yet unlike claims to (robust, assured) consideration, claims to (robust, assured) heed are demanding enough to require democratic responsiveness—yet not so demanding as to undermine democratic deliberation. Unlike consideration, heed requires positive responsiveness to citizens’ expressed judgments and will. This is enough to proscribe forms of government like considerate monarchies and oligarchies, which only really guarantee rational attention from, and the power to make requests of, the state. We can, of course, imagine monarchies and oligarchies that incidentally heed their citizens, but assurance of this treatment seems to preclude these political forms. It requires that political institutions be designed to give citizens rational attention but also to assure positive responsiveness to their claims.

As in the cases above, heed, unlike practical deference, does not require 1:1 responsiveness to the expressed preferences of any individual. This allows for various ways of harmonizing and executing citizens’ (often conflicting) political judgments and wills. If so, democratic states can at least possibly generate compromises without obviously contravening the authority of citizens who disagree deeply about matters of common concern.

Specifying more specifically how (non-normative) power must be distributed in order to realize the heedful state is a further question, and beyond the scope of the discussion here.⁷³ Answering it will require specifying more precisely than I do above how claims to heed are themselves distributed. While much of this will have to wait for future work, let me say something briefly about who is, and what kind of claims are, owed heed in concluding, to sharpen the picture of a democracy of heed and to assuage potential worries that claims to heed are overdemanding on individual citizens, particularly in deliberation.⁷⁴

To begin, it bears note that, unlike preemptive claims to consideration, claims to heed do not obligate citizens to consider all other citizens’ views and claims equally regardless of their assessment of merit or strength. Thus they are on their face naturally less demanding on individuals than consideration. Since heeding involves giving rational attention, claims to heed will still require citizens to deliberate with others, but not in a self-effacing way.

In thinking through just who is owed heed and when, it may help to recall the discussion of claims to deference vs. heed in Section IV above. There, I suggested that respect requires us to treat A’s practical expressions about matters that meaningfully affect him as authoritative, but authoritative in different ways. To put it no more strongly than is needed: that some practical matter meaningfully affects only A seems to be a good reason that A decide it—though there may in principle be other good reasons as well. By contrast, in cases in which some matter meaningfully affects A but also B and C, this is often taken to ground a voice for each. This, I think, suggests a degree of proportional sensitivity baked into respect: that A has much higher stakes than any others can ground a respect-based claim to deference (or decisive heed), whereas cases in which stakes are more widely distributed call for (sometimes mutual, never decisive) heed. Such proportionality would, at least, powerfully explain how respect—a singular moral attitude—can ground both liberal claims to author one’s own life and democratic claims to co-author decisions about common concerns.

But if so, it’s not obvious why proportionality would not obtain within democracy as well, at least where the prudential dictates of institutional design best permit. For while some matters of common concern are of equal concern, this is not always so. In some matters, some citizens have obviously higher stakes.⁷⁵ Without saying anything too controversial here, I think this can help us decide whose perspectives demand heed and whose may be set aside in deliberation, when time is limited and attention scarce.^76,77 Of course, democratic deliberation should also help reveal who has high stakes, and what counts as “high stakes” is of course open to reasonable contestation.⁷⁸ But, in an iterated deliberative process, such a proportionality principle could help us (and the institutions and practices which scaffold deliberation⁷⁹) home in on the perspectives that most demand attention and heed in a way that helps ease deliberative burdens on individual citizens and also supports heedful deliberation “within.”⁸⁰

I hope to say more about this in future work. For now, I think I have said enough to show that heed avoids the demandingness dilemma better than consideration or deference and that it is a distinct and attractive concept for democratic theory.

ACKNOWLEDGEMENTS

This article is from a larger project and was long in the works. Particular thanks goes to Annie Stilz, who read and gave invaluable feedback on many drafts. Thanks also to Chuck Beitz, Will Combs, Dan Friedman, Steve Macedo, Jan-Werner Müller, Temi Ogunye, Aidan Penn, Philip Pettit, and Matthieu Queloz for written feedback. For helpful conversations and correspondences about “heed” over the years, thanks to Elizabeth Anderson, Elvira Basevich, Colin Bradley, Daniel Browning, Mala Chatterjee, Andrew Chignell, Eli Frankel, Eleanor Gordon-Smith, Andrew Hahm, Mie Inouye, Sonny Kim, Maya Krishnan, Dan Layman, Damian Maher, Nikhil Menezes, Gaby Nair, Max Ridge, Ella Street, Ophelia Vedder, Darius Weil, Leif Wenar, Jim Wilson, Jo Wilson, and Daniel Wodak. Thanks finally to audiences at APSA, Bard, and Princeton, and to Bob Goodin and three referees for Political Philosophy for invaluable comments that improved the paper significantly.

COMPETING INTERESTS

The author declares that he has no competing interests.