What’s so Hard about Hard Choices?

: What, exactly, is so hard about hard choices? I suggest that what is distinctively hard about hard choices is that they present us with the volitional difficulty of putting ourselves behind an alternative and thereby making it true of ourselves that we have most reason to do one thing rather than another. Making it true through your commitments that, for instance, you have most reason to be a philosopher rather than a lawyer makes the choice between the careers hard. This answer is in contrast to that of Sergio Tenenbaum, who understands the hardness of a hard choice as a deliberative difficulty in specifying our alternatives and ends in ways that conform with certain proposed constraints of rationality. For Tenen-baum, the hardness of hard choices is not distinctive to such choices but is a general difficulty rational agents face when they need to further specify their alternatives and ends, even if the choice is easy.


I. INTRODUCTION
Sometimes we find ourselves with a choice that seems hard.We do not have most or sufficient reason to choose either of the options: our reasons 'run out'.Or, with respect to what matters in the choice, neither option is better than the other and nor are they equally good: the options fail to be 'trichotomously related'. 1When reasons run out or the options 1 A note on terminology since it is important for regimenting the general area of research.Many philosophers have used the term 'incommensurable' to mean 'incomparable', and some of the more recent instances of this practice explicitly follow Joseph Raz (1986), who once told me that, as a non-native English speaker, he looked up 'incomparable' in the English dictionary, and since one of the meanings was 'much better' he decided to use 'incommensurable' to mean 'incomparable'.I suggest that we depart from this contingently sourced practice and instead use 'incomparable' to mean incomparable and 'incommensurable' to mean what it has meant in line with its etymology from the Greek, namely, 'lacking (precise) cardinal comparability'.In the past few years, philosophers have used 'incommensurable' more cautiously to mean 'neither better nor worse and not equally good', that is, the failure of the standard trichotomy of relations.However, this fail to be trichotomously related, our ordinary way of making a rational choice-namely, by choosing what we have most or sufficient reason to choose or choosing what is evaluatively at least as good as the other options-seems to break down.Now suppose, further, that in such situations, despite our knowing that our reasons have run out or that our options fail to be trichotomously related, we also know-or are practically certain-that we are rationally permitted to choose either alternative.Three things are true in such cases.
1.The choice seems hard, in some intuitive sense of 'hard'.2. The reasons for choosing have run out, or the options fail to be trichotomously related with respect to what matters in the choice between them.3. We know that it is rationally permissible to choose either alternative.
But this seems puzzling.If we know that it is perfectly rational to choose either option, how can the choice nevertheless be intuitively hard?
In "The Hardness of the Practical Might: Incommensurability and Deliberatively Hard Choices", Sergio Tenenbaum confronts such cases.His aim is to answer the question: Why does a choice between two options that fail to be trichotomously related seem intuitively hard, even if one knows that it is rationally permissible to choose either option?Since he frames the question evaluatively, we follow suit.(On certain plausible views about the relation between reasons and values, the failure to be trichotomously related is equivalent to reasons running out, and so one expression can be substituted for the other).
What is important to answering this question is the idea of intuitive hardness.Curiously Tenenbaum never tells us what it is for a choice to be 'hard' in the intuitive sense, but we need to have a grip on this idea before we are in a position to assess the plausibility of his-and, I will argue, a preferable alternative-answer to the question.So the first order use, which Tenenbaum adopts, is also problematic.Since there is much of philosophical interest in the idea of lacking precise cardinal comparability, and we should avoid creating two distinct spheres of research concerned with two distinct ideas that travel under the same label, I suggest we reserve 'incommensurable' for the idea that items cannot be measured by the same (precise) cardinal scale of value and use the admittedly awkward 'failing to be trichotomously related' or 'trichotomously unrelated' to denote the idea that neither item is better, worse, or equal to the other.Moreover, as I argue in the text, being trichotomously unrelated does not 'carve at the joints', and so it makes sense to assign to the idea the least felicitous label.
of business is to fill in the intuitive sense of 'hard' that is at work in tackling the question.
I propose that we understand intuitively hard choices-just plain 'hard' going forward-as follows.In a hard choice: (i) neither option is better than the other and nor are they equally good with respect to the 'covering consideration', the consideration that gives what matters in the choice between them; and (ii) there is nevertheless a rational choice to be had in the situation; but (iii) getting to a rational choice is, in a way yet to be determined, 'difficult'; where (iv) the difficulty in arriving at a rational choice is distinctive to cases that meet conditions (i)-(iii).The point of condition (iv) is that the explanation of the hardness of a hard choice should involve appeal to a difficulty that is not also present in easy choices.Tenenbaum should, I believe, accept this relatively neutral framework for hard choices in the intuitive sense.
With the idea of hard choices in place, we need to ask whether the failure of trichotomous relatedness, which is what Tenenbaum says makes a choice seem hard, makes a choice in fact hard.The answer is 'no', and thus the 'seeming hardness' of a choice when the options are not trichotomously related is not the right place to rest when trying to investigate what is so hard about hard choices.Instead, we must go deeper and ask, what phenomenon holds when the options are not trichotomously related?As it turns out there are three rather different phenomena that are compatible with trichotomous unrelatedness, and in only one can the choice be hard in the intuitive sense outlined above.Marking and distinguishing the three phenomena is important because understanding what does hold of the options-as opposed to what does not hold-puts constraints on how we can go about explaining what is hard about hard choices.
When options fail to be trichotomously related, there are three logical possibilities.
First, the options might be incomparable, that is, no positive relation holds between them with respect to the covering consideration.If options cannot be compared, then it is widely thought, rational choice between the options is off the table. 2Thus in a choice between incomparables, there is no point bringing to bear one's rational faculties in trying to determine what to do since rational choice is precluded.Instead, the appropriate response in the face of incomparables is to 'plump', that is, nonrationally select-as opposed to rationally choose-an option.Acting on a whim or impulse-or, as Rosalind Hursthouse famously argued, throwing your malfunctioning computer against a wall in frustration-are exercises of nonrational agency.Plumping is another.It is the nonrational selection of-a nonrational 'going for'-an option outside of the scope of rational agency.
Since rational choice is precluded when options are incomparable, such cases are by that very fact not hard.There is, of course, a degenerate sense in which a rational choice is possible between incomparable options.We could say that nonrational selection between them is 'permitted' by rationality in the same sense we could say that brushing your teeth three times a day is 'permitted' by the law: everything outside of the jurisdiction of a norm is permitted by it.But even if we allow that in cases of incomparability it is rationally permissible to choose either option in this degenerate sense, getting to such a rational choice is easy: just plump.Thus, incomparability between options does not make a choice hard.
Second, it might be indeterminate, or vague, how the options relate with respect to the covering consideration; it could be indeterminate, for example, whether one is better than the other.If it is indeterminate whether one is better-in much the same way it might be indeterminate whether Jason Statham is balder than Will Smith-then rational resolution of the indeterminacy is a matter of 'picking' among equally good or eligible sharpenings of the relevant concept or world (Ullman- Margalit and  Morgenbesser, 1977).Such sharpenings are equally good 'intrinsically', that is, on the basis how the options relate without regard to 'extrinsic' factors that may favor some sharpenings over others.
Picking should be distinguished from plumping.The former is an exercise of rational agency while the latter is not; picking is the arbitrary choice of one item over another that is appropriate when options are equally good.If options are equally good with respect to the covering consideration, they are in that way interchangeable, and so it is rationally permissible to pick between them.When options are indeterminately related, there are multiple eligible sharpenings of their relation, each the formal conditions for possible comparison but fail to be comparable on substantive grounds (Chang 1997, 2017).
intrinsically as good as the other and each of which provides a resolution of the indeterminacy.By picking a sharpening, one arbitrarily stipulates a resolution and can, on that basis, make a rational choice.If it is indeterminate whether A is better than B, a rational agent can simply stipulate that A is better, and on that basis rationally choose it. 3ut indeterminacy does not make the choice is hard in our intuitive sense.Although options fail to be trichotomously related, and there is a rational choice to be had, getting to a rational choice is easy.Just stipulate a resolution and, on that basis, make a rational choice.
There is a final possibility.Two items that are not trichotomously related could be on a par.Parity is a fourth sui generis basic relation that can hold between options with respect to the covering consideration.In contrast to nonnormative comparisons of items that can be adequately represented as quantities and therefore proceed in terms of 'more', 'less', and 'equal', normative comparisons of items typically involve both quantities and qualities that require a fourth way in which they can be normatively related: they can be on a par.Roughly, items are on a par when they are qualitatively different but nevertheless in the same 'neighborhood' of the covering consideration.
Crucially, a 'neighborhood' of value or other normative element is not a trichotomously-related, hierarchical category such as 'excellent', 'very good', 'good' and so on, each of which represent an increasingly better or worse category of the covering consideration.Neighborhoods are essentially different from trichotomously-related, hierarchical categories; one neighborhood can be on a par with another.In short, neighborhoods afford nonhierarchical evaluation of packages of qualitative and quantitative features that determine the merits of something with respect to the covering consideration, without forcing them into artificially constructed, trichotomously-related, hierarchical categories that mimic the quantitative categories of 'more', 'less', and 'equal'.
So far we have seen that incomparability and indeterminacy fail to makes choices hard in our intuitive sense.What about parity?If options for choice are on a par, is the choice hard?As it turns out, being on par is the only phenomenon underlying the failure of trichotomous relatedness that could make a choice hard.But before turning to what I think is an explanation of what is so hard about hard choices when options are on a par, we should look at what Tenenbaum has to say about what is hard about hard choices.

II. HARDNESS AS A MATTER OF THE NEED FOR FURTHER SPECIFICATION
Although Tenenbaum says he is focused on hardness that derives from how options for choice relate, in fact he seems mostly interested in hardness due not to what relation holds between two options but hardness due to some feature of the options, namely their underdetermination, or what he calls their 'indeterminacy'.(Since underdetermination of the covering consideration, ends, circumstances, and so on can always be expressed in terms of the underdetermination of options, I will stick with talk in terms of options).The 'indeterminacy' he has in mind does not always seem to be vagueness, and if he means by 'indeterminacy' vagueness, then, as we have seen, vagueness makes hard choices easy, and so the fact "that our ends are (nearly always) indeterminate" (Tenenbaum,  2024, 186) in the sense of vague would not help explain what is hard about hard choices.
I believe by 'indeterminacy' Tenenbaum has in mind something more akin to 'underdetermination', that is, the lack of full specification of an option for choice.And when options are not sufficiently specified, Tenenbaum appears to hold, neither is better than the other and nor are they equally good (since a small improvement in one does not settle the matter as to which is better). 4What counts as living a low-risk life?Does watching a 4-hour black-and-white indy film make for a better evening than watching a short Marvel flick followed by dinner with friends?What counts as having a good evening, anyway?Should I retire now or retire later?'Retiring now' and 'retiring later' are too underspecified for us to determine which we are justified in choosing.We need to engage in the kind of specification that Aristotle (and contemporary specificationists like Henry Richardson (1997) and Elijah Millgram (1999)) thought was central to practical deliberation.Does retiring now give me time to work on my novel?Is there a way that I could retire later that would allow me to both work on my novel and play with the grandchildren?What impact do these options have on other things I care about?Is there a way to implement retiring now that is Pareto superior to all my ends, and if so, what is it?
These are difficult deliberative questions, questions whose answers require expenditure of both time and cognitive resources and make the choice "deliberatively hard", which Tenenbaum glosses by its contrary, "deliberatively easy", cases that "do not require much reflection and deliberation to arrive at the right choice" (Tenenbaum, 1-192).
More specifically, the further deliberation that Tenenbaum thinks is required must satisfy two desiderata, which makes the choice what he calls 'deliberatively hard'.It must, first, conform to The Tenenbaumian Principle of Instrumental Rationality: "Take some sufficient to all the ends [you] pursue" (186, emphasis added), which is the "only fundamental principle" of his proposed Extended Theory of Rationality ( 186), to which we will return later.This means that spending your evening in way A rather than way B is not justified if there is a way C that is Pareto optimal with respect to all your ends.So not only must you specify A and B sufficiently to ensure that each is a sufficient means to your end, but also you must scan the horizon for any possible option C that does just as well as A and B in achieving your end but better with respect to achieving something else you care about.Difficult deliberation, indeed.Moreover, Tenenbaum urges, our rational choices must conform to the Kantian Condition, the condition that we make "the right choice for the right reasons" (200), and it is hard deliberative work ascertaining which reasons are the right ones.
In sum, according to Tenenbaum, when choices seem hard because they fail to be trichotomously related, even though we know that it is rationally permissible to choose either, the choice is nevertheless hard because there is difficult deliberation ahead, deliberation that we are required to undertake to ensure that our choice is in conformity with i) the Tenenbaumian Principle of Instrumental Rationality and ii) the Kantian Condition.
Here are three remarks about Tenenbaum's proposal.First, the idea of a 'well-formed choice situation' might be helpful.A well-formed choice situation contains some limited set of circumstances, a finite, smallish set of options for choice, and a covering consideration that gives what matters in the choice, where, for the purposes of making a rational choice, each of these elements is 'fully determined', that is, not in need of further specification. 5In a well-formed choice situation, a rational choice is relativized to the situation 'as is', even though there may be uncertainty and a whole host of unanswered questions about the elements of the choice.
Choice situations that are well-formed provide a stopping point for what could otherwise be a never-ending quest for further specification of the elements of choice.If you are wondering whether to take a pleasurable break from work by taking a walk in the park or reading a short novel, there are in principle an infinite number of ways you could further specify those options that would be relevant to what matters in the choice.A wellformed choice situation might leave open questions about the direction of the wind, the pollen count, the probability of meeting your arch enemy, and so on, all of which might affect how well the options fare with respect to the covering consideration.
Exactly how to demarcate well-formed choice situations from still-inchoate-ones-in-need-of-further-specification-before-counting-aswell-formed is a substantive question about which philosophers have interesting things to say, such as when it is rational to stop gathering more information about options and how they normatively relate (see Andrew Sepielli (2009, 2018); Jake Ross (2006).Tenenbaum himself allows that "deliberation must come to an end" (2024, 204) and that "no matter how much effort we put in specifying our ends, a great deal of indeterminacy will remain" (2024, 204).But once we allow that deliberation about the elements of choice comes to an end in well-formed choice situations, we see that in such situations, the choice can be hard.In short, if there are well-formed choice situations, discovering that some are hard forecloses the kind of further specification that Tenenbaum thinks is required in the face of hard choices and explains their hardness.
Second, Tenenbaum claims not only that we are required to deliberate further in hard choices but also that this deliberation is of the ordinary kind that led to our recognizing that the items were not trichotomously related in the first place (194).It seems that, on Tenenbaum's view, it is specification all the way down.I doubt, however, that the rational response in intuitively hard choices is to carry on with business as usual; indeed thinking so might be thought to misunderstand the distinctiveness of hard choices or, indeed, to deny their existence altogether.At the very least, an account of the hardness of hard choices should arguably allow that something different is permitted of a rational agent who faces a hard choice.We will examine a view that makes just such allowances in due course.
This leads us to a third worry.Tenenbaum's description of the hardness of hard choices does not make the hardness of such choices distinctive.Indeed, as Tenenbaum says, as far as deliberation is concerned, there is no difference between how a rational agent should respond in the face of options (that fail to be trichotomously related) and how she should respond in the face of uncertainty about whether it is permissible to choose either: "she deliberates exactly in the same way she would deliberate had she been uncertain about whether both options are permissible ones".(Tenenbaum, 2024, 205)  If an agent is uncertain whether both options are permissible, that could take many different forms: she could be uncertain about whether the options are equally good, incomparable, vague, on a par or even whether one is better.So the deliberative hardness of hard choices, on Tenenbaum's account, also seems to be present in cases of uncertainty.
Indeed, Tenenbaum's deliberative hardness might also be present in easy cases.Suppose it is clearly and definitively better that you retire now than later, where that will pretty much guarantee you a good pension and time to write your novel and play with your grandchildren-the choice is easy.But how are you going to 'implement' the chosen option?Here is a question: what counts as 'now'?If you retire today, you might miss out on some further option floated by your employer tomorrow, a possible specification of 'now', that would be Pareto superior with respect to all your ends.It is deliberatively hard work hunting for Pareto superior sufficient means to all your ends at every turn in life, whether or not your options fail to be trichotomously related.The Tenenbaumian Principle of Instrumental Rationality and the Kantian Condition presumably operate as constraints on all rational choices, easy or hard, and it is unclear why conformity with those constraints is especially difficult when the options fail to be trichotomously related.Tenenbaum does not seem to provide an explanation of the hardness of hard choices that makes them distinctively hard.6

III. HARDNESS AS A MATTER OF MAKING SOMETHING NORMATIVELY TRUE
We argued earlier that when options for choice fail to be trichotomously related, there is only one underlying possibility, parity, that allows for the choice between them to be hard.Is there an explanation of the hardness of hard choices when options are on a par?
In other work, I have argued that if options are on a par, choosing either is rationally permissible.I have also argued that there is distinctive response available to a rational agent in such cases.I now want to suggest that this distinctive response is what explains what is so hard about hard choices.
Rational agents have the normative power to commit to some feature of an option that, as a logical matter, counts in favor of choosing the option with respect to the covering consideration.Commitment is not a mental state but a volitional activity of putting your very self behind something. 7This activity is that in virtue of which some feature can be a 'will-based' reason for choosing an option.In other words, by putting your very self behind some feature, like the fact that retiring now will allow you to avoid endless meetings, you can make the fact that you would avoid endless meetings a will-based reason to retire now.In this way each of us, as rational agents, can quite literally create new 'will-based' reasons to choose one option over another.We 'create' will-based reasons in the sense that we have direct volitional control over their grounds.
It is important to underscore that commitment is not itself a choice or decision guided by reasons.In particular, it is not guided by the existentialist question, 'What kind of agent do I want to be?' 8 After all, reasons having to do with being one kind of agent or another can be part of what matters in at least some choices, and it would be odd if in those cases any hardness is to be resolved by such reasons counting twice.Commitment has as an upshot the constitution of one's rational identity, roughly, all the reasons one has in one's life, including most crucially, what one has most reason to do, but it should not be confused with aiming at, deciding 7 It might be wondered whether commitment and the creation of reasons is the exercise of a skill, like playing tennis, or of judgement, like aesthetic judgment.Walden (forthcoming) fruitfully mines Kant's views on aesthetic judgment to moot the interesting idea that we can 'shape' our reasons in hard choices in the way we 'shape' the aesthetic features of a work of art through a faculty of judgment.Commitment, however, is different; there is no skill or sensibility of which commitment is an expression.Commitment is the exercise of a rational capacity, but the capacity is not borne of a sensibility that can be vulgar or refined and whose value constrains its proper expression.Commitments are, we might say, the deepest expression of rational freedom not governed by reasons or values.In any case, as an explanation of the hardness of hard choices, Walden's view, like Tenenbaum's, insofar as it is understood as an account of the hardness of hard choices, would not make the hardness distinctive. 8Contrast Edna Ullman-Margalit's (2006) and L. A. Paul's (2014) account of transformative experience and its effects on standard decision-theoretic accounts of rational choice.In the face of a transformative experience, Paul suggests that one contemplate what sort of person one wants to be-someone who knows what it is like to, for example, be a vampire.Margalit suggests one just takes a 'leap of faith'.to be, or deliberating about what rational identity to have or what kind of person to be.That question, the question of who to be, is one of the questions of human life in which human commitment can figure.
When options are on a par, standing behind something through making a commitment is not the only appropriate rational response.It is also rational for the agent to drift into one of the options, that is choose it for a reason, but without standing behind it or any of its features.Because drifting is choosing an option for a reason, viz., for one or more of the reasons that support it, it is not plumping.And because it is choosing an option that is qualitatively different from the other options and not interchangeable with it, it is not picking.When a rational agent exercises her rational agency by drifting in a hard choice, she does not create any reasons or values for herself.She simply chooses one of the options for any of the reasons that make it meritorious with respect to the covering consideration.But like committing, drifting has upshots for the agent's rational identity.
By drifting in the face of choice between options that are on a par, an agent shapes the reasons she has in her life; she leaves the options on a par.She could have committed to some feature of one of them and created new reasons for herself.But she did not.So she constitutes her rational identity by letting her reasons be.
The normative power we have to commit to something and endow it with normativity is a rational freedom we have in both hard and easy choices.But the exercise of this power is distinctively hard in hard choices because only by committing to an option in a hard choice can we make it true that we have most reason to choose one option over another.We cannot do this in any other kind of choice situation. 9Putting your rational agency behind something so that it might now be true that you have most reason to do one thing rather than another is an exercise of rational agency that is volitionally hard. 10This is not the hardness of gearing yourself up to do what you know you have most reason to do as in what would otherwise be akratic action-the hardness of getting your will to match your judgement that you really ought not to have another slice of chocolate cake.The volitional hardness involved in commitment in a hard choice is deeper: it is the difficulty of getting your very self to take one normative direction rather than another when the usual resources of rational agency-reasons or values-do not instruct you in the matter. 11It is the difficulty of being the author of your own life, of forging one path through life rather than another, of taking charge in creating values for yourself in the face of an open sea of reasons and values.What is distinctively hard about hard choices, then, is the volitional difficulty of making it true that you have most reason to live your life like this rather than like that.This is a difficulty we face in hard choices and only in hard choices. 12e can now see how the 'Parity View' ticks all the boxes of our intuitive gloss of hard choices.According to this view, in hard choices the options (i) fail to be trichotomously related-they are on a par, (ii) there is a rational choice to be had, and yet (iii) getting to a rational choice is difficult: the agent faces the volitional difficulty of committing to something and thereby making it true that she has most reason to do one thing rather than another.Finally, this volitional difficulty is something she faces only in hard choices; (iv) it is a distinctive difficulty.
Note that the Parity View also satisfies the Kantian Condition-you choose A over B for the right reason, namely either the will-based reason you created through your commitment or, if you drift, one of the reasons that any account would recognize as a right reason, namely a reason that supports choosing A when A and B are on a par.But the Parity View does not conform to the Tenenbaumian Principle of Instrumental Rationality.
11 It might be thought that there is another kind of hardness when options are on a par, namely the difficulty of determining whether to commit or drift.But there is no hardness here independent of the hardness of committing.After all, what one does when one commits or drifts is either exercise one's rational capacity to commit or exercise one's rational capacity to drift.There is no rational choice involved in determining which to do; one simply exercises one rational capacity or the other.If you were the kind of rational creature who was volitionally incapacitated so that you were unable to commit, then hard choices would be easy-you would always drift.I think this is roughly the state of many nonhuman animals.We might say they do not face any genuinely hard choices, choices in the intuitive sense of 'hard' outlined above. 12Of course, we might always drift in hard choices.Drifting is easy.When we face something hard-be it a move to a new job, an impending divorce, the death of a loved one, or the volitional difficulty of being the author of our own lives-we can always do things to avoid or mitigate the hardness.But this is not to say that it was never hard to begin with.
It can conform to a close cousin, what we might call the Ordinary Principle of Instrumental Rationality.According to this more humdrum principle, you have a pro tanto reason to take a sufficient means to your ends in a well-formed choice situation.You do not have to keep track of all your ends, but just what matters in each well-formed choice situation as you come to be in it.
But how do we come to be in one well-formed choice situation as opposed to another, and can we be justified in being in the one that we are in at any point in time? 13 We now arrive at what I believe is the core matter raised by Tenenbaum's article: how should we conceptualize the domain of practical reason?Underlying Tenenbaum's approach to explaining the hardness of hard choices is a conception of practical rationality in which the basic elements of rationality are not the usual mental states like preferences, desires, or intentions often thought to be the result of practical reasoning, but temporally extended plans, aims, options, and so on (Tenenbaum, 2024, 185, 186, and 206; Tenenbaum 2020).Such a view essentially conceives practical reasoning as a constrained but perpetual specification of "pursuing the end of φ-ing" (Tenenbaum, 2024, 186).We have suggested that one worry about such a view is that it cannot explain what is distinctively hard about hard choices.But Tenenbaum's view offers us an important insight: there is more to practical reason than making rational choices in well-formed choice situations.We might think of well-formed choice situations as islands dotting what would otherwise be an incessantly churning sea of values, reasons, capacities, and circumstances-elements of practical reason that call out for structuring and organization by rational agents.Well-formed choice situations can be the results of such structuring work; they are sufficiently specified 'safe harbors' in which rational agents can make choices that shape their lives.
Much of rational life is getting from one island to the next, and it is an open question in the philosophy of practical reason how we do and should do so.Tenenbaum's account provides a helpful way forward. 14By specifying our temporally extended ends through deliberation constrained by certain principles, we move ourselves towards certain series of well-formed choice situations over others.In some of these well-formed choice situations, the choice will be hard.But we should not mistake the specification that occurs in the interstices of well-formed choice situations as what is hard about hard choices.What is hard about hard choices is that they present us with the opportunity to make it true that we have most reason to do one thing rather than another.And making ourselves into creatures who have most reason to live one way rather than another-without the guidance of reasons-is distinctively hard.