The deliberative and epistemic dimension of legitimate authoritative directives *

What reasons do legitimate authoritative directives generate in their subjects? Raz identifies two relevant reasons: (a) a first-order reason to do as directed; and (b) a second-order, exclusionary, reason not to be motivated by at least some of the first-order reasons which conflict with the corresponding legitimate directive. In this chapter, I put forward two claims. The first claim, which I develop in the first section of this chapter, introduces a revision to the Razian framework: I argue that legitimate authoritative directives also exclude a particular type of reasoning about the balance of first-order reasons which is similar to the type of reasoning involved in the reconsideration of decisions. This captures the deliberative dimension of legitimate authoritative directives. The second claim, which I develop in the second section, is that the reasoning-excluding reasons legitimate authoritative directives generate have important epistemic value. This captures their epistemic dimension. My revisionism does not affect Raz’s account of legitimate authority and more specifically his normal justification thesis. Rather I presuppose it in developing my arguments.


1
The deliberative and epistemic dimension of legitimate authoritative directives *

Antony Hatzistavrou
What reasons do legitimate authoritative directives generate in their subjects? Raz identifies two relevant reasons: (a) a first-order reason to do as directed; and (b) a secondorder, exclusionary, reason not to be motivated by at least some of the first-order reasons which conflict with the corresponding legitimate directive. In this chapter, I put forward two claims. The first claim, which I develop in the first section of this chapter, introduces a revision to the Razian framework: I argue that legitimate authoritative directives also exclude a particular type of reasoning about the balance of first-order reasons which is similar to the type of reasoning involved in the reconsideration of decisions. This captures the deliberative dimension of legitimate authoritative directives. The second claim, which I develop in the second section, is that the reasoning-excluding reasons legitimate authoritative directives generate have important epistemic value. This captures their epistemic dimension.
My revisionism does not affect Raz's account of legitimate authority and more specifically his normal justification thesis. Rather I presuppose it in developing my arguments.
1. Reconsideration and reasoning-excluding reasons * Earlier drafts were presented at the IVR Workshop on Legal Normativity and the Philosophy of Practical Reason and a seminar of the Philosophy Department of the University of Stirling. I am grateful for comments received on both occasions. What exclusionary reasons exclude has been a matter of debate. 1 We can take them to exclude motivation by, consideration of, or reasoning about a certain range of first-order reasons.
(1) Motivation-excluding reasons. Raz takes exclusionary reasons to be normative reasons which favour 'not being motivated in one's actions by certain (valid) considerations' and 'exclude reasons from being one's motivation for action ' (1999, 185). For example, Leonidas' command to the 300 Spartans that they hold Thermopylae is a reason for the latter not to be motivated by a desire to save their lives. 'Motivation' is ambiguous in this context. On the one hand, we may take Leonidas' command to put forward a rather strong demand on the souls of his soldiers: they should not even be tempted by considerations of the sweetness of life and experience a pull or an inclination towards throwing down their shields and running away. On the other hand, Leonidas' command may be less demanding: we may understand Leonidas' command to exclude only that considerations about survival will not be the motive which gets his soldiers to act. I have argued elsewhere that the second, less stringent reading of motivation-excluding reasons is more plausible. 2 Given that motivation-excluding reasons are not the focus of this chapter I will not dwell further on this distinction.
(2) Consideration-excluding reasons. Alternatively, we may understand exclusionary reasons to rule out that the agent consider certain reasons. Consideration of reasons may take two forms. First, it may mean entertaining certain thoughts in one's mind or preoccupying oneself with elements of one's mental imagery. For example, one may think of or mentally visualise a romantic dinner with one's sweetheart. In some cases there may be good reasons that one does not entertain such thoughts or mental visualisations. For example, if one has decided to become a monk (and let us assume for the sake of the argument that this is a good decision) one may have good reasons not to think of romantic dinners. Similarly, if one has dedicated oneself to the pursuit of certain ends which require one's single-minded devotion (climbing Mt Everest, or trying to prove a mathematical conjecture), the achievement of these ends may be threatened by thoughts or mental representations of various forms of joie de vie.
Second, we may understand consideration as equivalent to taking certain factors into account when one is deliberating. In this case, consideration-excluding reasons exclude taking into account certain factors. For example, once Leonidas has issued his command to hold Thermopylae in deciding how to position themselves in the battlefield the Spartans should not take into account the possibility of leaving open a route for escape. The point is not that the thought of an escape route should not cross their mind or that they should not mentally visualise their running away from the field or mentally replicate the feeling of relief from the stress of the battle correlated with their imaginary escape. The second understanding of considerationexcluding reasons does not require guarding against thoughts that crop up in one's mind or mental quasi-perceptions. It only requires that these thoughts are treated as irrelevant in the context of one's deliberations.
Consideration-excluding reasons are not equivalent to motivation-excluding reasons. First, thoughts which crop up in one's mind or mental quasi-perceptions need not have motivational clout. The monks or the dedicated mathematicians who think of romantic dinners need not be tempted by them to change course in life (one's sweetheart may be far away, one may be too old or too shy to seriously consider amorous relationships a real option). One's thoughts or mental images may be nothing This material has been published in Reasons and Intentions in Law and Practical Agency by/edited by G. Pavlakos & V. Rodriguez-Blanco http://doi.org/10.1017/CBO9781107707573.010. This version is free to view and download for private research and study only. Not for redistribution or re-use. © Cambridge University Press. 4 more than a distraction from the normal activities of the life one has dedicated oneself to (fingering the prayer beads while one counts the repetitions of prayers, or going through the steps of an elusive proof).
Second, the mental action of taking something into account is conceptually distinct from the condition of being motivated by something. More importantly, factors one may register in one's deliberations may turn out to be motivationally inert.
We often take into account certain factors in our deliberations while being sceptical about whether they are actually going to play a role in our decisions or even while being convinced that they will not. For example, I may initially take into account the torque power of a certain car model in the process of deciding which car to buy even though I may not have made up my mind yet about whether it is an important consideration for me (i.e. a consideration which may incline me towards buying the car) or even about whether it would be the decisive reason for selecting a car (i.e. a consideration which may get me to buy the car). Or I may take it into account even though I know it is not important for me simply because I want to have a clear picture of the engine capacities of the car. So, in one sense consideration-excluding reasons exclude more than motivation-excluding reasons: they exclude taking into account certain considerations even though the latter might turn out to lack motivational clout.
Third, in another sense motivation-excluding reasons exclude more than taking into account certain factors. Psychological elements other than factors we register in our deliberations may motivate us. For example, we may be moved by sudden urges or unreflective desires.
(3) Reasoning-excluding reasons. Finally, we may understand exclusionary reasons to exclude not simply taking into account certain factors while deliberating but engaging in certain types of reasoning. We may distinguish types of reasoning by reference to Some valid exclusionary reasons exclude a particular type of reasoning, namely, reasoning about a particular end and of a particular modality. This type of reasoning is paradigmatically involved in the reconsideration of decisions or personal rules. It is also similar in kind to the reasoning involved when one questions the legitimacy of authoritative directives. For convenience I will call the exclusionary reasons valid decisions and legitimate authoritative directives generate 'reconsideration-excluding reasons' and 'reconsideration like reasoning-excluding reasons' respectively. Though Raz insists that exclusionary reasons exclude only being motivated to act by certain considerations, he provides an analysis of the exclusionary function of decisions which implies that they provide (also) reconsideration-excluding reasons. Raz claims that 'decisions are exclusionary reasons in the sense that it is logically true that if x has decided to do A then x believes that his decision is a reason for him to disregard further reasons for or against doing A.' (1999, 68; Raz's italics) 3 Disregarding further reasons for or against doing A is treated as equivalent to a refusal to 'go on looking for more information and arguments and to decline to listen to them when they crop up in one's mind or are suggested by other people. '(1999, 67) Raz suggests that only decisions the agent is justified in making are valid exclusionary reasons (1999,68). This account of the exclusionary function of decisions captures a defining feature of valid decisions. Valid decisions exclude reconsideration, i.e. they exclude deliberation about the first-order reasons which the decision is meant to reflect with a view to reassessing the decision.
This normative relation between valid decisions and non-reconsideration is explanatorily more basic than the logical relation between decision and non-reconsideration. I mean the following: it is necessary that the person who has made a decision believe (at least implicitly) that she has normative reasons not to reconsider her decision -in the absence of this belief the person cannot be considered to have made a decision.
The emphasis on the normative relation between decisions and reconsideration is probably motivated by a desire to avoid bootstrapping. My belief that I have exclusionary reasons not to reconsider a decision does not create a reason not to reconsider it. The decision must be valid. The content of the decision may be a specific action (do A at t), or a conditional action (do A if p occurs) or a general policy or personal rule (do A whenever circumstances C obtain). These criteria may yield different results when applied to the same decision. Though a decision may satisfy the first criterion it may fail to satisfy the other two. For example, though it may have been rational for me to decide to buy bonds of a specific bank in August given the information I possessed and the information I could have been reasonably expected to collect at the time of making my decision, it may still both be a disastrous financial investment and fail to improve my decision-making ability. Similarly a decision may satisfy the second criterion but fail to satisfy the first. For example, if the suggestions of the university's Quality Office may turn out to have adverse effects on the educational experience of the students then it was not rational for me to decide to accept them (even though at the time of deciding I did not have any evidence that there were problems with them). Finally, a decision may satisfy the first two criteria but not the third. The reasoning guidance that the decisions provide relates to another intrinsic feature of theirs. Decisions are not simply time and labour saving mental devices but also assurance conditions. They furnish the agent with a significant degree of epistemic assurance that she will not deviate from her plans because of a change of mind and that she will continue building on and expanding them. For it is not only external conditions or bad luck which may frustrate one's plans but also tricks one's mind plays. Doubts may lurk in one's mind and may lead to continuous retraction or hesitation. In this case the point of the normal justification thesis is that the deferring party is more likely to conform to the balance of reasons which is evidentially valid by following the judgement of the ruling party rather than relying on her own collection and assessment of the relevant evidence. Alternatively the balance of reasons may be understood to be assessed by reference to whether it corresponds to the relevant practical reality. In this case the normal justification thesis claims that the deferring party has normative reasons to conform to the judgement of the ruling party which is more likely than her own judgement to possess practical truth-tracking validity. Finally, the judgement of the ruling party may be more likely to be systemically justified since it may be more likely than the judgement of the deferring party to enhance the overall decision-making ability of the latter. In this case the deferring party has normative reasons to conform to a judgement reliance on which in future deliberations is more likely to improve her overall decision-making ability than reliance on her own judgement. There are two arguments in favour of the thesis that the deferring party has normative reasons not to engage in this type of reasoning. The first is that the link between legitimate authoritative directives and exclusion of this type of reasoning is, like the link between valid decisions and non-reconsideration, normative cum logical. Unless the deferring party believes Similarly, the reconsideration like reasoning-excluding reasons which legitimate authoritative directives generate are not equivalent to consideration-excluding reasons. On the one hand, like reconsideration-excluding reasons they exclude only a type of reasoning and not mental visualisations or thoughts. On the other hand, the reconsideration-like reasoning excluded by legitimate authoritative directives has a specific end, that is, it is about the balance of reasons which support the authoritative directives, and modality, that is, it is accompanied by willingness to rely on one's own judgement in one's deliberations. There are types of reasoning about legitimate authoritative directives which involve taking into account reasons that conflict with legitimate authoritative directives but are not excluded by reconsideration like reasoning-excluding reasons: (1) Reasoning about the defeasibility conditions of general authoritative directives.
(2) Reasoning about the applicability conditions of general authoritative directives.
(3) Reasoning about the execution of occasion-flexible authoritative directives. 9 (4) Reasoning about the relevance of conditional authoritative directives. To sum up, in this section I have argued that valid decisions generate reconsiderationexcluding reasons and legitimate authoritative directives (at least on a particular account of legitimate authority) generate reconsideration like reasoning-excluding reasons. 10 In the next section I turn to the issue of the epistemic value of legitimate authoritative directives.
2. Legitimate authoritative directives and epistemic agency I argued that the deferring party has conclusive reasons to treat refraining from challenging the legitimate authoritative directives as her default position in her deliberations.
I explained the value of this refraining attitude by reference to the importance it has for practical agency. 11 The agent is better off not challenging the legitimate authoritative directives since she is more likely to act in accordance with the balance of reasons by following the authoritative directives rather by following her own reasoning and is thus more likely to promote her practical goals. The contribution of this refraining attitude to the attainment of one's practical goals is indirect. By this I mean that it has negative causal influence, that is, it contributes to the attainment of one's practical goals in the way in which the absence of an impediment may causally influence an outcome. Other things being equal, refraining from challenging a legitimate authoritative directive significantly raises the probability that the agent will act on the directive and thus attain her practical goals; while, other things being equal, challenging a legitimate authoritative directive raises the probability that the agent may not act on the directive, since she might follow her own judgement after examining the balance of reasons, and thus fail to attain her practical goals.
This suggests that the value of reconsideration like reasoning-excluding reasons depends in an important sense on the value of motivation-excluding reasons. Refraining from challenging legitimate authoritative directives is a means to preventing acting on certain conflicting considerations. One has reason to refrain to the extent that one has reason not to be motivated to act by certain considerations. This does not entail that reconsideration like reasoning-excluding reasons are equivalent to motivation-excluding reasons for reasons I explained in the previous section. But it makes motivation-excluding reasons the primary reasons legitimate authoritative directives generate.
However, the perspective of practical agency does not fully capture the value of refraining from challenging legitimate authoritative directives. There is another perspective from which to assess its value: the perspective of epistemic agency. We have not only practical but also epistemic goals. We aim not only at fulfilling our practical plans but also at attaining knowledge or at least a comprehensive body of true beliefs. This refraining attitude contributes to the attainment of our epistemic goals. On the one hand, it has instrumental epistemic value. When one follows an authoritative directive, then, absent cases of akrasia, one forms a corresponding judgment about what is best for one to do. For example, the colonel who follows the general's order to capture the hill forms the belief that he should act so as to lead the regiment's attack on the hill. If the directive has legitimate authority, then one has epistemic reasons to be unwilling to challenge it. For by challenging it, one enhances the probability that one may lose an epistemically good state. If the legitimacy of the relevant authoritative directive depends on how the corresponding balance of reasons which it reflects is assessed ex ante, this state consists in the preservation of a belief which is justified by the available evidence at the time the directive is issued. In this case the agent has epistemic reasons to treat the relevant justified practical belief as the default position of her deliberations. For one should give precedence to beliefs justified by the available evidence. If the legitimacy of the relevant authoritative directive depends on whether the corresponding balance of reasons which it reflects correctly represents practical reality, this state consists in a belief which tracks a practical truth. In this case again the agent has normative reasons to be unwilling to challenge the authoritative directive. For one should retain beliefs which track truths. Finally, if the authoritative directive is systemically justified then the relevant refraining attitude should remain the default position of the agent. One's decision-making capacity has clearly an epistemic dimension, since it enables one to reach practical truths in the future. So, it is rational for one to retain a relevant practical belief which corresponds to the authoritative directive and has systemic value for one's epistemic agency because it raises the probability that one will track a large number of important practical truths in the future. In all these cases refraining from challenging legitimate authoritative directives is valuable because it is an instrumental means to the preservation of a state which is epistemically good.
On the other hand, this refraining attitude has value in itself as a constituent of good epistemic agency. Unwillingness to challenge legitimate authoritative directives is constitutive of intellectual trust in others since it exhibits respect for the epistemic authority of others and in particular those who in an important sense deserve to be treated as epistemic authorities. Trusting others in general is a prerequisite of my epistemic agency and my ability to reach accurate and comprehensive beliefs. Epistemic agency is impossible without epistemic trust in my current self, my current opinions and decisions. 12 But epistemic trust in my current self is inextricably linked with intellectual trust in others. Given the social construction of our system of beliefs there is an intellectual pressure to trust the opinions and testimony of others. In order for us to be able to form our own judgements we need to rely on 12 For the importance of self-trust see Foley 2001 andLehrer 1997. This material has been published in Reasons and Intentions in Law and Practical Agency by/edited by G. Pavlakos & V. Rodriguez-Blanco http://doi.org/10.1017/CBO9781107707573.010. This version is free to view and download for private research and study only. Not for redistribution or re-use. © Cambridge University Press. 20 information provided by others which we either have no means or no time to check and assess. We have to take them on trust. More importantly we cannot plausibly trust our own reasoning capacities unless we at least implicitly have trust in the beliefs of others. Foley (2001,102) puts this point succinctly: '...our most fundamental concepts and assumptions, the material out of which our opinions are built, are not self-generated but rather are passed down to us from previous generations as part of our intellectual inheritance. We are not intellectual atoms, unaffected by one another. Our views are continuously and thoroughly shaped by others. But, then, if we have intellectual trust in ourselves, we are pressured also to have prima facie intellectual trust in others. For, insofar as the opinions of others have shaped our opinions, we would not be reliable unless they were.' Foley speaks of prima facie intellectual trust. I think that the prima facie jargon fails to capture correctly the intrinsic link between intellectual trust in others and intellectual selftrust. First, the link is best understood as being normative cum logical in a way familiar from my discussion in the previous section. That is, unless I at least implicitly believe that I have normative reasons to intellectually trust others I cannot be thought to intellectually trust myself. Intellectual self-trust involves trust in my beliefs, that is, beliefs which are born out of If I have normative reasons to trust by default the valid beliefs of others, then I have a fortiori normative reasons to trust by default the judgement of those who are more likely than I am to get it right about the balance of reasons. These people may possess a higher level of intellectual expertise or competence than I do. But they need not to. Their judgement about the balance of reasons may be more trustworthy than mine because it is less likely to be blurred by contravening factors (stress, fear, emotional attachment) than my judgement even though we share the same level of intellectual expertise or competence.
The mental attitude of refraining from challenging legitimate authoritative directives (understood along the lines of the normal justification thesis) may be reasonably considered an aspect of or a way of expressing default intellectual trust in the valid beliefs of others.
Furthermore, since self-trust shapes my epistemic agency and is in it itself shaped by default intellectual trust in others, default intellectual trust in the valid beliefs of others may be considered a constituent of good epistemic agency. It is thus reasonable to regard the mental attitude of refraining from challenging legitimate authoritative directives as having value as a constituent of good epistemic agency. 13 This view, which is in essence a modified version of Foley's modest epistemic universalism, is contrasted to what Foley labels as 'epistemic egotism' and 'epistemic egoism' (Foley 2001, 83-92). The first claims that I have no normative reasons to trust the views of others and thus my default position should be to always determine for myself the validity of their views. The second claims that I have normative reasons to trust the beliefs of others only if I believe in their reliability, that is, the fact that their views are valid is not a reason for me to trust them unless it is backed by the fact that I believe that they are reliable. I do not have the space here to argue against these two theories. I will only note that they both appear to rest on the view that only beliefs the agent has can be reasons for her to adopt the epistemic attitude of trusting the beliefs of others. By contrast I hold that only facts (like the fact that the beliefs of others are valid or that they are more likely than the agent's beliefs to be valid) are reasons for adopting the aforementioned epistemic attitude. For a defence of the view that reasons are facts and not beliefs see Gardner and Macklem 2002. I have argued so far that this refraining mental attitude has both instrumental epistemic value and value as a constituent of good epistemic agency. I believe that these considerations provide us with adequate reason to identify default intellectually trust of legitimate authorities as an intellectual virtue. I do not mean that whenever one intellectually trusts legitimate authorities one exhibits an intellectual virtue. For one could do this unreflectively or by accident. Rather I mean that one can through habituation and training develop an ability to identify legitimate authorities (which includes an ability to identify their jurisdiction) and treat their directives as providing deliberative guidance. I will not provide a detailed account of the form and function of this intellectual virtue. I will simply offer a sketch of it which I hope will make its existence at least plausible.
A contrast between epistemic and practical agency is a good starting point. A good practical agent is not simply someone who does the right thing consistently. It is also someone who has a specific character, that is, certain mental attitudes towards her actions.
For example, a courageous person is not simply someone who holds her position in the battlefield and does not run away to save his life. It is someone who performs a certain range of actions or refrains from others out of a certain motive, say, out of a desire to defend her country as opposed to a desire to accumulate wealth, or overcome a childhood trauma; who engages in certain mental actions, for example, resisting fears or the temptation of daring actions; and who appreciates certain pleasures, for example, she enjoys the confidence generated by her not succumbing to fear or the temptation of daring actions. In a similar manner, a good epistemic agent is not someone who accumulates true beliefs and acquires pieces of knowledge in a consistent manner. It is someone who has in addition the right mental attitude towards her belief-formation processes: she approaches a cognitive field out of a certain cognitive motive, for example, a desire for knowledge, engages in particular types of mental actions, such as resisting quick conclusions or the temptation of elaborating on trivial issues and appreciates certain cognitive pleasures, like the pleasure of doubting an unclear argument and the confidence associated with a firm grasp of a truth or the development of her own cognitive abilities.
These mental attitudes may be considered intellectual virtues. And the attitude of showing default intellectual trust of legitimate authorities may be among them. By refraining from challenging legitimate authorities the agent treats her cognitive resources in a prudent manner (since she avoids engaging in unnecessary and potentially faulty reasoning), exhibits relevant epistemic humility, and cultivates an attitude of epistemic respect for the beliefs of others which is essential for the advancement of knowledge. These intellectual qualities may become after a process of habituation and training more or less stable features of her epistemic agency. The initial focus of habituation and training may be certain relevant intellectual tendencies human beings have which one may regard as intellectual proto-virtues.
As one may regard certain tendencies human beings have towards their emotions, for example, the tendency to disregard fear when angry, as practical proto-virtues which can be properly calibrated, one may equally regard a tendency to rely on other people's views as a relevant intellectual proto-virtue. Once this proto-virtue is properly calibrated and related to other intellectual proto-virtues, for example, a tendency for creativity and independence of thought, it may become a proper intellectual virtue.
To sum up, in this chapter I have explored two dimensions of legitimate authoritative directives. On the one hand, legitimate authoritative directives have a deliberative dimension since they exclude a particular type of reasoning which is similar to the type of reasoning involved in the reconsideration of valid decisions. On the other hand, they have an epistemic