Presentation to the Ayn Rand Society
APA Eastern Division
New York City, December 2000
Tara Smith's Viable Values is a valuable addition (so to speak) to a newly-emerging literature on the philosophical thought of Ayn Rand. Its contribution, as I see it, is two-fold. On the one hand, it lays out the essential features of Ayn Rand's Objectivist Ethics in a clear and persuasive way; on the other hand, it presents a series of bold challenges, both direct and indirect, to virtually every type of mainstream ethical theorizing in the `analytic' tradition. The book as a whole adds up nicely to the sum of these parts.
Though there is much to say about virtually every section of the book, my aim in this paper is confined to outlining and assessing two of the book's most important positive claims: the conditionality of life as the basis of moral value, and the content of moral prescriptions as egoistic rather than impersonal. These two claims are central both to Smith's project and to Ayn Rand's distinctive ethical vision; they also mark a faultline that differentiates the Objectivist conception of ethics from that of the mainstream. Focusing on them, I think, should bring into sharp relief the issues that divide Objectivism from mainstream ethical thought.
Viable Values begins with the age-old question, "Why should I be moral?," and with a provocative discussion of the failure of mainstream approaches to the question (VV, ch. 2). (For ease of reference, let's call this the Question.) One of the unique features of her account, Smith says, is success at answering the Question where other theories have failed (VV, pp. 83, 187). Given this, it would help to know exactly what the Question is asking.
As many philosophers have noted, the Question "Why should I be moral?" is highly ambiguous; in fact, a discussion of its ambiguities could take a book in itself. Given the ambiguities of the Question, and given the fact that Viable Values describes itself as an "elaboration and defense" of Ayn Rand's ethical views (VV, p. 83), it may help to turn to the opening of Ayn Rand's "The Objectivist Ethics" to pin down the Question's specifically Objectivist interpretation. In the first few paragraphs of the essay, Rand poses a similar but not identical question, as follows:
The first question that has to be answered, as a precondition of any attempt to define, to judge or to accept any specific system of ethics, is: Why does man need a code of values?
Let me stress this. The first question is not: What particular code of values should man accept? The first question is: Does man need values at all—and why?
The first thing to note here is that Rand asks four different Questions, each significantly different from those typically discussed in the mainstream literature. They are:
1. Do rational beings need values in the first place?
2. Assuming that rational beings need values, why do they need them?
3. Assuming we know that they need them, why do they need a code of values?
4. Assuming they need a code of values, which code do they need?
I will in due course clarify the terminology here. But stated in this form, our list draws attention to the methodological self-consciousness of Ayn Rand's approach to ethics, which I think subtly informs Smith's approach as well (VV, pp. 13-19). Rand insists that ethical theorizing has a specific order, involving questions posed and answered in a specific sequence. In particular, note Rand's special insistence, echoed by Smith (VV, p. 15), that we not begin by assuming that our "considered beliefs" embody moral knowledge, and that ethical theorizing merely consists in putting those beliefs into reflective equilibrium. That procedure, as Smith aptly puts it, simply allows one to seek "a rationalization for one's existing beliefs rather than a justification of true beliefs" (VV, p. 15). And truth, of course, is what we're after.
From Needs to a Code of Values
Let's revisit our Questions and consider the Objectivist answers to them in turn. Question (1) was: Do rational beings need to value? Is there anything that entities like us have to do?
At first glance, the answer might seem to be "no." On the Objectivist view, a rational being is a self-determining entity, with the power of free choice. "A course of thought or action is `free'," writes Leonard Peikoff, "if it is selected from two or more courses of action possible under the circumstances. In such a case, the difference is made by the individual's decision, which did not have to be what it is, i.e., which could have been otherwise." Given this conception of free action, how can an agent be necessitated to do anything? Any act of valuation could after all have been otherwise.
On second thought, however, it's fairly obvious that our actions are necessitated in a different sense of "necessitation." Though we're metaphysically free, we do have needs. In common parlance, a "need" is a condition that "has to be" satisfied by action; satisfaction leads to reward, non-satisfaction leads to the agent's suffering adverse consequences. I need to eat to meet my metabolic requirements; I need to acquire knowledge to deal competently with the world around me. And so on. A need so construed is a conditional necessity. If I need something, X, the necessary and available means to X all become practical requirements for me. If I need to eat, I must obtain food somehow; if I need to obtain food, I have to acquire knowledge; if I'm to acquire knowledge, I must cultivate rationality; and so on. So the concept of a need, or a conditional necessity, gives us a kind of prima facie answer to Question (1). A rational being needs to value because it has needs; if it is to satisfy those needs, it has to act accordingly by free choice.
Despite its initial plausibility, however, this explanation can only go so far. Question (1) asked whether we have to value at all, and the answer was that we do because we have needs. But why do we have needs at all? Why are there conditional necessities? And that brings us to Question (2). Supposing, provisionally, that we need values, why is this so?
Smith's answer, which comes from Rand, turns on the conditional character of life as the underlying generator of practical requirements. Following the Aristotelian tradition, Rand characterizes life as a conditional process of self-generated and self-sustaining action. In short, what this means is that a living being's very existence as the entity it is depends on its engaging in a complex repertoire of continuous action across its lifespan, generated and directed by an energy source internal and integral to itself. The nature or identity of an organism determines the kind of action appropriate to it. Precisely because each organism has a determinate nature, each has a determinate mode of life and valuation, appropriate to its needs, capacities, and environment. Given the determinacy of modes of life so conceived, there is such a thing as acting in accordance with, and acting against, the requirements of an organism's life—the most obvious requirements being those of physical health and functionality. An organism's default on its mode of life undermines or impairs its functionality by degrees; severe default undermines its existence and identity altogether, and leads ultimately to death. Let's call this complex fact the conditionality of life, and the thesis expressing it, the conditionality thesis.
Contrary to a common misinterpretation, the central idea behind the conditionality thesis is not that life is a necessary condition for valuation, given that one can only value while alive (or no longer can if dead). That may well be true, but it's not the point of the thesis. The point of the conditionality thesis is that it provides the ultima facie explanation for which the notion of "needs" constituted at best a prima facie one. A need demands action on its behalf. An entity with needs faces an alternative: either meet the need (and be rewarded), or don't meet the need (and face adverse consequences). Obviously, needs so construed are enormously varied and complex: the need to eat is vastly different from the need for knowledge, for instance. What the conditionality thesis asserts is that the alternative of life and death is the most fundamental alternative underlying the entire structure of needs. Life, in other words, is a kind of second-order need—not merely a need, but the basic need that explains why all other needs exist and exert practical pressure on us. Likewise, the conditionality of life is not merely a conditional necessity, but the ultimate conditional necessity in virtue of which the phenomenon of conditional necessitation exists at all. The reason why the proposition "If I'm to obtain X, then I must enact Y" is true is that at the deepest level, it codifies the structure of life itself. At every level, whether rational or non-rational, conscious or non-conscious, it's true that: if an organism is to preserve its existence and identity, then it must act appropriately to its nature and environment.
That, in short, is the answer to Question (2). Ultimately, I must engage in valuation because as a living being, I'm a conditional being, and as a conditional being, my life depends on self-generated and self-sustaining action. Nor does it ever stop depending on it. If life requires continuous action, then every need I satisfy will bring another one in its wake, at a higher level of complexity. I can ignore or evade that fact about myself, but I can't escape it—at least so long as I aim to exist. If so, I have to identify what to pursue, and how to pursue it; I have to find what's valuable, and resolve to track it.
What, then, about Question (3)? So far, we've learned that the conditionality of life requires constant valuation of us. But why should it require a code of values?
First, a terminological detour: what is a "code"? The Oxford English Dictionary defines a `code' as a systematic collection or listing of statutes, rules, regulations, or signals. A code of values, we might infer, is a set of interconnected principles by means of which values are sought, in contexts that require the pursuit of those values. Judges by definition value justice; they need a legal code to secure it, and the judicial temperament to follow the code. Fire inspectors by definition value fire-safe buildings; they have to insist that every building be "up to code," and cultivate the Javert-like habits appropriate to the task. The idea of a "code" in this context captures two things: (a) the systematicity of a collection of complex principles geared to realizing specific purposes in the world, and (b) the traits required to follow those principles in the concrete circumstances of the world.
Now let's return to our question. Human beings must value because they have needs, and they have needs because they are living, conditional entities. Conditional entities must value to remain in existence and to maintain their identities. Is having a code itself a value for beings in this predicament?
The answer, it turns out, depends on the kind of being we're talking about. As Ayn Rand first suggested, conditionality is an attribute of all and only organisms—from the lowliest prokaryote all the way across the biological taxa. Since a person is an organism, a person is therefore a conditional being like all the others. Given the distinctiveness of human nature, however, the conditionality thesis applies in a special way to human beings—a way of particular importance to ethics. The human being's distinctive nature, for Rand as for Smith, arises from the special features of human consciousness, which differentiate the organism that possesses it from all of the other organisms.
In the other organisms, valuation is a thoroughly deterministic phenomenon, genetically coded by natural selection, and triggered by features of the environment. Whatever their simplicity or complexity, the lives of non-human organisms proceed more or less automatically—from metabolism, homeostasis and growth, through locomotion, learning, and cooperation. Non-human organisms are, in short, automatic, deterministic value-trackers that automatically take life as their ultimate value, and an automatic awareness of, and propensity to act on, their needs. The problem of the conditionality of life is for them solved by genetic coding.
By contrast, human beings are volitional, rational agents born tabula rasa. The fact that we're tabula rasae entails that we have no automatic means of discovering what our life-requirements are, or of maintaining a consistent commitment to them. Lacking innate knowledge, we have to discover our means of survival for ourselves; lacking innate behaviors or traits of character, we have to cultivate a commitment to tracking the good. Our distinctive moral task, then, is to employ volition and reason in the service of both ends. The problem posed by the conditionality of life is for us a matter of explicit, volitional and rational codification.
The problem in question is complicated by two interconnected facts, which Smith discusses in her illuminating account of flourishing in chapter 5. The first is the sheer temporal extension of a human life. A full human life consists of the total expected length of an entire lifespan—say, eighty or ninety years. So we need a code that will last the distance, and traits that will do so as well. The second is the unity of a human life. A human lifespan is not just a series of disaggregated and unrelated sequences, but an integrated whole, each of whose parts contributes to the sum. So we need a code that will serve to integrate the parts into a whole, and traits that will do the same.
A code of values, then, is a system of principles, adopted by choice and internalized in character, which fully satisfies the uniquely human version of the problem of discovering and tracking values. We need a code of values so construed for reasons analogous to those that explain why sunflowers are genetically coded to seek the sun, why amoeba have to engage in perpetual phagocytosis, why squirrels have to be compulsive hoarders of nuts, why lions roar, and why dolphins must echolocate. We need to preserve ourselves across a lifespan, as they do, but we lack the automatic means for doing so that they have. We need morality to do for us what genetics does for them. That answers Question (3).
I can only touch here on Question (4): what code of values does man need? A code of values is a matter of discovering the principles and inculcating the traits that will promote the survival of a rational being so conceived. The principles take the form of what Smith calls "fundamental prescriptive generalizations" (VV, p. 165), identifying the values that preserve and unify a full human life and enjoining action on their behalf. The traits take the form of self-beneficial virtues, geared to achieving the values in question.
The most fundamental principle on which the entire code is based is its ultimate standard of value—survival qua man. In Ayn Rand's words, "Man's survival qua man means the terms, methods, conditions, and goals required for the survival of a rational being through the whole of his lifespan—in all those aspects of existence which are open to his choice." This standard, on Rand's view, gives rise to the need for three cardinal values: Reason, Purpose, and Self-Esteem, along with their corresponding virtues, Rationality, Productiveness, and Pride. The cardinal virtues and values as Rand puts it are jointly "the means to and realization of" the ultimate value, life. They promote life by constituting the core of its structure.
The standard of survival qua man also gives rise to other virtues, whose function in each case is to give us a constant, automatized (though not automatic) awareness of the facts required to pursue our values and maintain a commitment to them. Among the major virtues Rand mentions are independence, integrity, honesty, justice, and benevolence; some of the particular values to be sought are dramatized in her novels.
This short outline of the Smith-Rand argument should alert us to the sheer distance between the Objectivist view of what ethics is about and the view we find in mainstream philosophy. In fact, a mainstream ethicist might well wonder whether I've been talking about ethics at all for the past few pages. The reason for the puzzlement, I suspect, derives from the distance between the claims of the conditionality thesis and the standard operating concept of mainstream ethics, "the moral point of view." I turn to that issue next.
The standard approach to morality in the mainstream literature is not to begin with the Question "Why be moral?" or anything like it, but to define morality circularly in terms of what is usually called "the moral point of view." The moral point of view is a distinctive outlook on the world, which literally transforms the way it looks to the agent who adopts it. In doing so, one brings out the world's morally salient features at the expense of its non-moral ones, filters out the latter, and adopts principles to govern one's actions—usually in a social context.
One of the differentiating features of specifically moral principles on this view is their impersonality. In taking the moral point of view, one explicitly excludes considerations of self-benefit in assessing the moral worth of an action; one sees oneself as merely one member of the moral community, each one of whom is equally and agent-neutrally significant in one's deliberations. Impersonality so conceived is typically thought to constitute morality as such, and is therefore neutral between conceptions of morality—neo-Kantian, Utilitarian, Contractarian, etc. The differences between these mainstream theories are differences within the moral point of view, not alternatives to it. On the mainstream conception, then, though self-interest is at least occasionally a permissible motive of action, it remains conceptually at odds with morality. Though we may make concessions to self-interest, it's not itself a moral motive; to the extent that we insist on acting from a specifically moral motive, we must put self-interest aside.
While philosophers in the past few decades have begun to take issue with the stringency of the impersonality constraint, and with some of its particular demands, none goes nearly as far as Smith in rejecting the constraint itself. Against the impersonality thesis, Smith argues that the conditionality of life entails a commitment to ethical egoism: every moral obligation serves our objective self-interest; none fails to do so. More precisely, ethical egoism may be defined as the view that a moral agent should act with the intention of being the ultimate beneficiary of his or her own act. As Smith interprets it, ethical egoism entails that morality just is self-interest, and vice versa: "principled egoism is the only way to live" (in her words) because it's the only way to be moral. This claim is perhaps the most controversial one in Viable Values, and is also the substantive ethical thesis for which Ayn Rand is most notorious.
How do we get from the conditionality thesis to ethical egoism? Smith puts the point succinctly. If life is the standard of value, self-benefit is the only rational and permissible motive: "Human beings survive by acting for their own benefit." (VV, p. 155). Rational beings deliberate, then, from a conception of their own good, formulated by the standard of life described previously. To depart from that conception, whether from the motive of impersonal duty or altruism, is to sacrifice self-benefit to some other concern. But no such action could be sanctioned by the standard of survival qua rational being, so every such action is to be ruled out. Following Smith, let's call a life lived by this standard, a life of flourishing.
Flourishing so conceived involves a plan for an entire life, which itself involves a commitment to principled action internalized by self-beneficial virtues. I mentioned some of the specific virtues earlier, but a general point about the role of virtue is worth stressing. The virtues, recall, give us an automatized grasp of the requirements of survival qua rational being. In doing so, they call for a principled commitment to virtue across the whole of one's lifespan, a commitment that requires that one make principled action an ineliminable part of one's moral identity. As Smith argues, this commitment to virtue is strong enough to rule out the (apparent) value of "ill-gotten gains" derived from violations of moral principle: "Something is in a person's interest only if it offers a net benefit to the person's life. Since a person's life is not reducible to any isolated element of his condition, we cannot fasten on such elements to draw conclusions" about the person's interests (VV, p. 168). Note, however, that virtue is not "an isolated element" of one's condition, but rather something one needs across a whole lifespan. The requirements of virtue, then, are always part of the calculation of "net benefit." So this momentary departure from moral principle—this isolated, secret, juicy act of vice--no matter how tempting or apparently beneficial, is never in my interest.
Egoism so conceived is a "positive-sum game" (Smith's term) that holds out the possibility of a harmony of interests in which the virtue of each agent becomes an asset for that of every other. The harmony is underwritten by the fact that each virtuous agent benefits from the virtues rather than the vices (or involuntary weaknesses) of others. One benefits neither by sacrificing one's interests to others, nor by preying on their vices or vulnerabilities, but by interacting with others in order to gain from their virtues and strengths, and by implication, from their self-interest. The rational pursuit of self-interest, then, is never the cause of conflicts of interest, because virtue constitutes our self-interest, and virtue can never be the cause of conflict. Conflicts arise not through the pursuit of objective interests, but through violations of the egoistic principles that secure our interests.
This picture of egoism is very distant from the one we find in the mainstream literature. Kurt Baier's argument excluding egoistic considerations from the moral point of view is typical of the mainstream conception of egoism:
Morality is designed to apply in just such cases, namely, those where interests conflict. But if the point of view of morality were that of self-interest, then there could never be moral solutions of conflicts of interest. However, when there are conflicts of interest, we always look for a `higher' point of view, one from which such conflicts can be settled. Consistent egoism makes everyone's private interest the `highest court of appeal'. But by the `moral point of view' we mean a point of view which is a court of appeal for conflicts of interest. Hence it cannot (logically) be identical with the point of view of self-interest.
In a similar vein, James Rachels's polemic against egoism, which we can take to be representative of a type, is notorious. An egoist, Rachels argues, would have no reason not to burn down a building on whim, just for the spectacle of the fire and thrill of burning everyone inside to death. After all, if a person believed the action self-beneficial, what principle of egoism could one invoke to condemn it?
These arguments are the standard ones in the literature, repeated in virtually identical form wherever they are found. Despite their ubiquity, however, they are, I think, remarkably weak arguments, proceeding as they do from a common premise that none demonstrates—viz., that consistent egoism does in fact lead to conflicts of interest. The assumption that underwrites this claim is the belief that egoism is a form of unconstrained preference-satisfaction. Since it is, "egoism" becomes a kind of black box, in both metaphorical senses of "black": inscrutable and nefarious. One's "self-interest" consists in doing whatever one wants, however one wants.
The problem with such accounts of egoism is not merely that the claim about conflicts of interests goes undefended; it's that we've been given no way of giving any content to the concept of "self-interest" at all. Nothing in these discussions tells us what egoism prescribes (or why), or why anyone would think it rational, much less why it must lead to conflicts of interest. The ultimate court of appeal about the content of egoism reduces instead to linguistic consensus: "we" all speak as though consistent egoism led to conflicts of interest; since "we" generally think that "our" speech mirrors the way the world is, "we" should assume that ordinary language proves that egoism leads to conflicts of interest.
In the space I have available, let me simply offer a laundry list of reasons why I think these arguments fail, taking Smith's account and discussion of egoism as my point of departure.
For one thing, it's unclear why the claims of ordinary (English) language should assume such extraordinary epistemic powers in defining the nature of human self-interest. If the conditionality thesis is true, then our lives are structured by a determinate set of interests, whose nature is as discoverable as the requirements of health. We would surely be alarmed to discover that medical schools were teaching their students that the principles of clinical diagnosis and practice were based on nothing but the "linguistic intuitions of the medical community in reflective equilibrium." Why should it be different in ethics?
Second, and relatedly, virtually nothing in the literature on egoism focuses on what actual egoists—e.g., Aristotle, Spinoza, Ayn Rand—have actually said about what our interests are. The preference is to concoct notional confrontations with notional egoists, bearing no relationship to actual defenders of the view. When this literature occasionally does make reference to actual figures, like Ayn Rand, it typically distorts their claims in fantastic ways, or ignores important aspects of what they say. (Perhaps Smith's book will remedy this.)
Third, there is little appreciation in this literature of the need or nature of planning an egoistic life-plan, taking one's life as a single unit of long-term deliberation, as opposed to thinking of one's life as a series of disaggregated and discrete moments. Consequently, the focus in the literature is always on problems of choice in the context of range-of-the-moment situations, as in Rachels's arson example. But if Smith's argument is right, an egoist is someone who adopts value-tracking principles and virtues so as not to live by the range-of-the-moment. Confronted by a supposed egoist with a yen for arson, as in Rachels's example, the obvious question is: Why would arson promote anyone's life as described in Smith's or Rand's theory? This question is never posed, even by implication.
Fourth, there is little or no discussion of the place of virtue in such a life or life-plan; when the egoist's values are not straightforwardly depicted as psychopathological, discussions of egoism almost always concern the unjust seizure of external goods (or exploitation of persons for this purpose). Such claims miss the point: on an egoistic view, the external goods are only valuable when they're gotten by virtuous means (VV, pp. 164-187). So such examples are irrelevant unless proven otherwise.
Finally, there is little appreciation for how large a generalization the conflicts of interest claim is. Egoism, Baier tells us, is "among the main causes of our social problems." But where does one get evidence of such a thing? None is ever provided, apart from adducing discrete examples of conflict-situations. But isolated examples don't prove that something is "among the main causes" of something else. If virtue is in our self-interest, how could it be among any of the causes of our social problems? The conceptual incoherence of the suggestion itself requires explanation.
One of the great frustrations of anyone exposed to the literature on the Question "Why be moral?" is the fact that there are so few genuine answers to it. What one so often gets in the literature is the admonition that the Question ought not to be asked because it's immoral to ask it—or the claim that what the Question ultimately requires is "the secular equivalent of faith in God" or recourse to "the sanctions of law" to persuade the recalcitrant.
One of the merits of Viable Values is that, in offering a clear and accessible exposition of Rand's theory, it offers us neither faith nor force as an answer to the question it poses, but precisely what I take to be a satisfying and responsive one. Our own lives, as Smith puts it, are "the sum of morality's claims on us." "Life sets the standard of value, life is the goal of morality, life is the reward of morality. What stronger answer can one imagine to the question of why we should be moral?" (VV, p. 187). For my part, I can't think of one.Irfan Khawaja Dept. of Philosophy
University of Notre Dame
Notre Dame, IN 46556
As I suggest in the body of the paper, mainstream ethics is based on "the moral point of view." One of its differentiating features, as remarked there, is a commitment to impersonality. Another, which I lacked the space to discuss in the body of the paper, is the categoricity of moral obligation. In this Appendix, I address that issue vis-à-vis what Smith and Rand call "the choice to live" (VV, ch. 4). As Rand puts it, "To live is [a rational being's] basic act of choice. If one chooses to live, a rational ethics will tell one what principles of action are required to implement that choice. If one does not choose to live, nature will take its course." Smith writes, "the choice to live is prerational. It is a presupposition of the standards of rationality" (VV, 107). Moral obligation, then, is contingent on a choice.
On the standard conception of morality, by contrast, whatever their superficial grammatical form, moral principles are categorically binding, as opposed to conditionally so: morality consists of an important class of duties that one owes without having been incurred by any voluntary act. One just has them, period. As Rawls puts it,
[I]t is characteristic of natural duties that they apply to us without regard to our voluntary acts…Thus if the basic structure of society is just…everyone has a natural duty to do his part in the existing scheme. Each is bound to these institutions independent of his voluntary acts, performative or otherwise.
Likewise Stephen Darwall:
While other elements of Kant's theory of morals…perhaps reflect rather poorly any very widely shared view about morality, his insistence that moral requirements are categorical imperatives expresses our common sense about an important part of ethics.
And James Rachels:
Commonsense morality would say, then, that you should give money for famine relief rather than spending it on the movies.
This way of thinking involves a general assumption about our moral duties: it is assumed that we have moral duties to other people—and not merely duties we create, such as by making a promise or incurring a debt. We have natural duties to others simply because they are people who could be helped or harmed by our actions.
Such claims are widespread in mainstream ethics. The language of "binding" involves the metaphor of a sort of cord that "ties" the agent inescapably to various duties. But the very description of a natural duty suggests that there is no way to explain how the cord came to be placed around us.
This categorical conception of obligation poses a challenge to the Smith-Rand view of moral obligation, which makes moral obligation ultimately conditional on a "choice to live." If the Objectivist view is really "objective," how can morality's binding force rest on a choice? Doesn't it then collapse into subjectivity? We can make this objection more specific by subdividing it into two sub-objections. The first objection concerns the escapability of conditional obligations; the second concerns their apparent arbitrariness.
The escapability objection goes as follows. Either moral obligation is categorical or it's merely conditional on a choice. If it's categorical, the agent has unchosen obligations. If it's conditional, the agent's obligations are chosen. But if obligation depends on choice, what if the agent doesn't make the relevant choice? Then there is no obligation. But surely morality, and especially justice, must be "inescapable." Hence moral obligation must be categorical, not conditional.
The arbitrariness objection goes as follows. Either moral obligation is conditional or categorical. If it's conditional, it depends on a choice. But someone who decides not to make that choice is beyond the reach of morality. Imagine now that we try to persuade such a person to be moral. Clearly, we must appeal to reasons to make the relevant choice. But in the nature of the case, there are no reasons to give; the choice is "prerational." Hence the choice must be arbitrary, and must make morality arbitrary. Since morality is not merely arbitrary, its binding force must not rest on a choice.
It's clear that for both Smith and Rand, though the conditionality of life explains the overridingness of morality, its binding force—indeed, its overridingness for any given individual—is conditional, not categorical. So if the preceding objections were sound, they would apply to, and refute, the Objectivist view. Though Ayn Rand discusses the issue in her essay "Causality Versus Duty," (where she articulated the view), she never addressed the preceding objections as such. In Viable Values, Smith couches her own account of the choice to live as a series of discrete responses to objections (VV, pp. 103-117) without, in my view quite offering a positive interpretation of the idea. In what follows, I offer my own tentative thoughts on the issue, which I believe are an extension of Rand's view, and are compatible with Smith's.
The key to understanding the "choice to live," as I see it, is to think of the binding force of an ultimate value by analogy with the binding force of a logical axiom, on the Aristotelian conception of an axiom (cf. VV, p. 107). So conceived, an axiom can be thoroughly conditional in its binding force without being either escapable or arbitrary in the senses implied by the objections just described. Though ideally, I would like to have compared Smith's account of the choice to live with Ayn Rand's conception of "axiomatic concepts," in the interests of accessibility to a general audience, I'll use the more familiar example of Aristotle's defense of the Principle of Non-Contradiction (PNC).
At Metaphysics IV.3, Aristotle enunciates the PNC: "A thing cannot be and not-be at the same, in the same respect." This is an undeniable and foundational truth; its truth is merely re-affirmed in the attempt to doubt or deny it. Note, however, that the Principle is not a categorical injunction to engage in thought. In fact, it says nothing at all about thought, nor is it a prescription of any kind. It merely states a fact about the world—one that becomes a guide for thought when and only when one chooses to think. In choosing to engage in thought, one sees in one's own case that if one is to do so successfully (i.e., at all), one must obey the Principle without exception. Any isolated attempt to evade the Principle would be self-subverting. A wholesale attempt to evade it would render one entirely unable to think. In choosing to think, then, one is therefore bound by the Principle without ever having been forced or enjoined to do so.
Is the PNC "escapable"? In one sense, yes; in another sense, not really. One can escape the PNC-- if one is willing to pay the price. The PNC binds all thought; one way to evade it, then, is simply to stop thinking. And refusing to think is a way of escaping the PNC. The PNC applies if (and only if) one aims at thought. It doesn't apply to a non-thinker. On the other hand, its non-application to the non-thinker is hardly a threat to its logical or epistemic authority. A non-thinker can't raise an objection (or even have one), and thus cannot constitute a problem for the PNC. So the "problem" of escapability is no problem at all: the price of escape is such that it poses no threat to the Principle.
Is the PNC "arbitrary"? Not at all. To be sure, there's no argument for the PNC that proceeds from principles that are epistemically prior to and independent of it. But that's because there are no such principles. The PNC is an epistemically basic principle; all principles presuppose its truth. The Oxford English Dictionary provides six relevant entries for the term "arbitrary": "to be decided by one's liking," "dependent on one's will or pleasure," "at the discretion or option of anyone," "derived from mere opinion or preference," "capricious, uncertain, varying," ""unrestrained in the exercise of will." I think it's clear from these entries that the concept of "arbitrariness" only applies in contexts where we face two or more alternatives and choose one while ignoring the established principle that adjudicates between them. In the nature of the case, this stricture couldn't apply to the PNC: putting aside the Law of Identity (which is equally basic), the choice between accepting or rejecting the PNC couldn't be governed by our choosing some other principle which told us to accept or reject it. There is no such principle, and couldn't be one. So the PNC isn't arbitrary, because it applies prior to the proper application of the very concept of "the arbitrary." The concept "arbitrary" only has meaning subsequent to accepting the PNC's axiomatic status.
The analogy to moral obligation applies as follows. As a matter of non-prescriptive fact, life can only be kept in existence by a constant process of self-sustaining action. Moreover, life is unique in this respect: it's the underlying generator of practical requirements that explains why there are practical requirements at all, themselves requiring self-sustaining action. So life is the ultimate value. Life's conditional character, however, is not by itself a prescription. It's simply a fact about the world. The fact by itself generates no categorical duty to keep one's own life (or anyone else's life) in existence, or indeed, to value or do anything at all. Life's conditional character becomes a guide for action when and only when one chooses to value and live, i.e., chooses to engage in goal-directed action, and chooses to promote one's life by goal-directed action.
In choosing to live, one is conditionally bound by the requirements of life. The momentary attempt to evade that fact would lead one to practical inconsistency: if I will life as an end, I must will the means to it; if I refuse to will the means, I must give up the end. So a momentary act of evasion would stand condemned as a single violation of the principles of rational choice. The wholesale attempt to evade the conditionality of life, of course, leads the evader to non-existence. In choosing to live, then, one is bound to realize one's life without ever having been required to do so by anything besides the choice to live.
Is the choice to live escapable? Again, yes and no. It's escapable in the sense that one can, in principle, fully opt out of the task of aiming at one's self-preservation. But such a person is no more a threat to the authority of moral norms than the non-thinker is to the PNC. The person who denies the PNC is rendered incapable of thought; the person who denies the conditionality of life is rendered incapable (in the same sense of "incapacity") of goal-directed action.
Is the choice "arbitrary"? No. "Arbitrariness" is a matter of flouting some established principle. But once we take a regress of justifications back to an ultimate principle, there are no principles left in terms of which to adjudicate the choice to adopt the ultimate principle. Here, the ultimate principle is survival qua rational being, so the ultimate choice is the choice to live. The choice, however, is hardly arbitary. If life really is ultimate, then only it justifies normative standards and rational action. If one accepts it, acceptance makes rational action possible. If one rejects it, one has no legitimate claim to valuation based on standards at all. There's nothing arbitrary about the former course of action, in any legitimate sense of "the arbitrary."
 Thanks to both Allan Gotthelf and Gregory Salmieri for extensive written comments on earlier drafts of this paper.
 I use the term "mainstream" throughout to denote "the currently dominant mode of philosophical research and inquiry in the United States according to its avowed practitioners." As John Searle puts it, "The dominant mode of philosophizing in the United States is called analytic philosophy. Without exception, the best philosophy departments in the United States are dominated by analytic philosophers, and among the leading philosophers in the United States, all but a tiny handful would be classified as analytic philosophers….Analytic philosophy has never been fixed or stable, because it is intrinsically self-critical and its practitioners are always challenging their presuppositions and conclusions." John Searle, "Contemporary Philosophy in the United States," in The Blackwell Companion to Philosophy, (Cambridge, MA: Blackwell Publishers, 1996), edited by Nicholas Bunnin and E.P. Tsui-James, pp. 1-2. See also Bernard Williams, "Contemporary Philosophy: A Second Look," in the same volume, p. 26.
 "Objectivism" is the name Ayn Rand gave to her philosophy as a whole. Given Smith's claim to offer an elaboration and defense of Rand's account (VV, p. 83), it follows trivially (I think) that her book is a defense of the Objectivist Ethics. In saying this, I don't mean to imply that Smith is an official spokesperson for Objectivism, or that her book (or this paper) represent "official doctrine."
In the space at my disposal, I've had to give a very compressed account of Smith's argument; I've also felt free to move back and forth between Smith's book and Rand's original claims, without (I hope) distorting the content of either.
 Rand's use of the term "man" to stand for "male and female human being"—which I more or less follow here—needs some comment. Ideally, one wouldn't use the term "man" to stand for both sexes or genders; in my view (though probably not Rand's), the usage can be both confusing and sexist. In the present context, however, I'm not convinced that there is a usable alternative. Rand uses the term "man" to refer to what characterizes Homo sapiens sapiens as such, presupposing her theory of concepts and essences in Introduction to Objectivist Epistemology. The linguistically natural alternatives to "man" don't quite capture the exact shade of meaning she had in mind, and my aim in the text is to preserve that meaning in explicating her theory.
What are the alternatives? (1) One might use the term "person," but the notion of a "person" has Cartesian connotations that don't quite work: personhood, as it's typically conceived, is not the kind of thing that could conceptually be connected with the conditionality of life as Rand conceives of it. The concept of a "person," I think, abstracts away from an agent's biological properties; it's a purely mentalistic concept. (2) Another possibility might be "human," but it doesn't work, on grammatical grounds; it makes no grammatical sense to ask, "Why does human need values?" (3) A third possibility might have been "humans." However, "humans need values" sounds like a summary description or rule of thumb observation about many human beings; it doesn't assert the same proposition as "Man needs virtue," which says that virtue characterizes the very structure of the distinctively human way of life. (4) A fourth possibility, suggested to me by Hilary Persky, is "The human individual" or "The human being," but though it gets the meaning right, it sounds even more awkward than "man." So I reluctantly stick with "man."
 Ayn Rand, "The Objectivist Ethics," in The Virtue of Selfishness: A New Concept of Egoism, (New York: Signet Books, 1964), pp. 13-14, italics in original.
 Leonard Peikoff, Objectivism: The Philosophy of Ayn Rand, (New York: Dutton, 1991), p. 55.
 Cf. Aristotle, De Anima II.2-4, and Generation of Animals II.1.
 Harry Binswanger, The Biological Basis of Teleological Concepts, (Los Angeles: ARI Press, 1990), ch. 4.
 Peikoff makes this point as well, Objectivism, pp. 212-213.
 For a more sustained argument, see Binswanger, Biological Basis, especially chs. 8-9.
 (note removed by author)
 Ayn Rand, "The Objectivist Ethics," p. 26.
 Ayn Rand, "The Objectivist Ethics," p. 27.
 The locus classicus is Kurt Baier's The Moral Point of View, (Ithaca, NY: Cornell University Press, 1958); see also Baier's "Moral Reasons and Reasons to be Moral," in Values and Morals, edited by Alvin I. Goldman and Jaegwon Kim, (Dordrecht, Holland: Reidel, 1978).
Other notable accounts include R.M. Hare, Freedom and Reason, (Oxford: Oxford University Press, 1963); Thomas Nagel, The Possibility of Altruism, (Princeton, NJ: Princeton University Press, 1970), and The View from Nowhere, (Oxford: Oxford University Press, 1986), chs. 8-10; John Rawls, A Theory of Justice, (Cambridge, MA: Harvard University Press, 1971); William Frankena, Ethics, (Englewood Cliffs, NJ: Prentice Hall Press, 1973); J.L. Mackie, Ethics: Inventing Right and Wrong, (Harmondsworth, UK: Penguin Books, 1977), chs. 4-5; Alan Gewirth, Reason and Morality, (Chicago: University of Chicago Press, 1978); Peter Singer, Practical Ethics, (Cambridge, UK: Cambridge University Press, 1979); David Gauthier, Morals by Agreement, (Oxford: Oxford University Press, 1986): Kai Nielsen, Why be Moral? (Buffalo, NY: Prometheus Books, 1989); Christine Korsgaard, The Sources of Normativity, (Cambridge, UK: Cambridge University Press, 1996).
For a useful overview of the debate, see The Definition of Morality, (London: Methuen & Co., Ltd, 1970), edited by G. Wallace and A.D.M. Walker.
 I think this remains largely true of so-called Virtue Theory as well, e.g., as described by Greg Pence in his entry "Virtue Theory" in A Companion to Ethics; (Malden, MA: Blackwell Publishers, 1993), pp. 249-258; see also the references he cites therein. At best, Virtue Theorists strike me as ambivalent about the impersonality constraint, not opposed to it.
I suspect a commitment to impersonality at least partially explains why Pence finds it natural to open his essay by describing the dilemma faced by Dorothea Brooke in George Eliot's Middlemarch, rather than, say, the dilemma faced by Dagny Taggart at the beginning of Part III of Atlas Shrugged. (Readers of both novels will note that at least superficially, the dilemmas are similar!) On the impersonalist conception, Dorothea's dilemma looks like a real moral dilemma faced by a real moral agent; Dagny's doesn't.
 Classic expressions of this view include Bernard Williams, "Persons, Character, and Morality," in Moral Luck: Philosophical Papers 1973-1980, (Cambridge, UK: Cambridge University Press, 1981), pp. 1-19; Susan Wolf, "Moral Saints," Journal of Philosophy 79 (August 1982), pp. 419-439; John Cottingham, "The Ethics of Self-Concern," Ethics 101 (1991), pp. 798-817; and Jean Hampton, "Selflessness and the Loss of Self," in Self-Interest, edited by J. Paul, F. Miller, and E. F. Paul, (Cambridge, UK: Cambridge University Press, 1993), pp. 135-165.
 Kurt Baier, The Moral Point of View, p. 190.
 I am unfortunately missing this reference as I write, but I believe it is James Rachels, "Two Arguments Against Ethical Egoism," Philosophia, vol. 4 (1974), p. 297.
 Except in emergency situations, as discussed in Rand's "The Ethics of Emergencies," in The Virtue of Selfishness, essay 3.
 Baier, "Egoism," in A Companion to Ethics, (Malden, MA: Blackwell Publishers, 1993), edited by Peter Singer, p. 197.
 The first quotation comes from Annette Baier, "Secular Faith," pp. 204-5 in Revisions: Changing Perspectives in Moral Philosophy, (Notre Dame: University of Notre Dame Press, 1983), edited by Stanley Hauerwas and Alasdair MacIntyre. The second comes from Peter Singer, Practical Ethics, p. 220.
 Ayn Rand, "Causality Versus Duty," in Philosophy: Who Needs It, (New York: Signet Books, 1982), p. 99.
 Rawls, A Theory of Justice, pp, 114, 115.
 Stephen Darwall, Impartial Reason, (Ithaca, NY: Cornell University Press, 1983), p. 173.
 James Rachels, The Elements of Moral Philosophy, (New York: McGraw Hill, 1993), p. 76.
 I owe the inspiration for my view to Allan Gotthelf's 1990 Ayn Rand Society paper, "The Choice to Value," but he is not responsible for my way of formulating the issues.
 On axioms, see Aristotle, Posterior Analytics, I.2., and Nicomachean Ethics VI.2.