I’m in the middle of editing, and writing a contribution for, Reason Papers’s forthcoming symposium on the epistemology, ethics, and politics of emergencies. (Though I’ll be contributing an essay to it, credit for the idea of the symposium goes to Carrie-Ann.) The topic turns out to be an amazingly interesting and fertile one, with ramifications in a number of directions, including the epistemology and semantics of definition; the application, scope, and stringency of moral principles; the scope and content of rights; the proper role of government; the rule of law; and so on.

The topic has been made particularly timely by a number of recent events–the recent Ebola outbreak; the state of emergency declared this past summer in Ferguson, Missouri after the riots; the states of emergency declared during Hurricane Sandy in 2012; the states of emergency that have been thought to obtain after the terrorist attacks of 9/11, etc. But those phenomena hardly exhaust the applications of ‘emergency’: once you start looking for emergencies, you find them everywhere, often in the most unlikely places. And then there are emergency-cognates, like disasters, catastrophes, crises, epidemics, and pandemics, as well as pseudo-emergencies of every kind.

At the most basic level, emergencies are events that are ontologically and normatively discontinuous from “normal” events. It’s part of the very concept of an emergency that it’s an exceptional sort of event, that it involves danger, and that it demands urgent attention and action. Given those facts, emergencies drastically affect the content and application of normative principles–epistemic, ethical, and political: principles that apply to ‘ordinary’ life don’t apply, or at least don’t apply in the same way, to emergencies. Ordinarily, we expect people to form their beliefs on the basis of the evidence for them–but in emergencies, guesses and hunches become acceptable. Ordinarily, we expect people not to kill, torture,  kidnap, steal from, trespass on, cheat, deceive, or manipulate others–but in the right kind of emergency (some would say) just about anything goes. Ordinarily, we’re governed by the rule of law, including constitutional law, but in emergencies, we either bend the rules, or are governed by open-ended edict or fiat. And so on.

Given this, emergencies give rise to a difficult dilemma in need of solution. On the one hand—call it the rigorist hand–there’s the danger that the invocation of “emergency exceptions” to moral principles serves as a rationalization or excuse for ad hoc exception-making. The concept of an “emergency” seems ill-defined and elastic, and “emergency exception” seems like a magic wand by which moral principles are undercut and effectively ignored. If we define “emergency” too broadly, every form of duress becomes an “emergency,” so that the concept loses its meaning, and emergency exceptions serve to subvert the stringency of the moral principles that are supposed to govern normal life. At a certain point, a moralist has to put his foot down and insist that duress requires the exertion of moral strength and endurance in the name of normality rather than exception-making in the name of abnormality. But the appeal to “emergencies” is one of the most seductive ways of evading this responsibility. From the rigorist perspective, then, emergencies are philosophically important primarily because we have to put them in their place. Emergencies, the rigorist insists, are marginal exceptions in life. We should cordon them off as essentially irrelevant to normal life, and focus our attention on the non-exceptional rules that govern normal life.

On the other hand—call it the pragmatist hand—there’s the danger that if we ignore emergencies, or minimize their significance or frequency, we ignore a real phenomenon that affects the proper application of moral principles. All principles, even the most stringent, apply within a specific and in-principle specifiable context. If we ignore the difference between emergencies and non-emergencies, we apply principles designed for one context to a context where they lack application. In doing so, we run the risk of sacrificing things of greater value to things of lesser value on the basis of a robotic commitment to empty verbal formulations masquerading as moral principles. In doing that, we risk the dangers of imposing pointless burdens on those suffering great duress, invoking “morality” as a pseudo-justification for our dogmatism. From this pragmatic perspective, we ought to take emergencies seriously because emergencies are an extreme instance of duress, and duress is a ubiquitous but easily-ignored part of our moral life. Goodness is fragile, and emergencies underscore that fact in an acute way.

The rigorist and pragmatic views are, of course, one-sided caricatures of possible positions. Each view gets something right, but each view ignores the merits of the other view, and thereby gets something wrong. A good dialectician would have to work through the two views, integrating their best insights, discarding what each view gets wrong, and fashioning an alternative to them that avoids the pitfalls of either view. The task would be to strike the mean between rigorist and pragmatist extremes.

I’ve argued elsewhere—very briefly and tentatively, in a footnote (p. 219 n.29)—that Ayn Rand appears to have been the first writer to have ‘thematized’ the topic of emergencies in twentieth-century English-speaking philosophy. (If I’m wrong about that, as I very well could be, I’d be interested to hear about it in the comments.) Her discussion appears in “The Ethics of Emergencies,” a 1963 essay published in her 1968 essay collection, The Virtue of Selfishness, and anthologized every now and then in philosophy textbooks. I have mixed feelings about the essay, as I do about much that Rand wrote.

On the “pro” side, I think she gets four or five things basically right.

(1) Most fundamentally, she gets the basic framing issue right:

It is important to differentiate between the rules of conduct in an emergency situation and the rules of conduct in normal conditions of human existence. This does not mean a double standard of morality: the standard and the basic principles remain the same, but their application to either case requires precise definitions. (Ayn Rand, “The Ethics of Emergencies,” The Virtue of Selfishness, p. 54).

In other words, some principles straddle emergencies and non-emergencies, retaining their identity across both contexts, but taking different forms in the one context as opposed to the other. Good theorizing keeps both facts in mind. It’s not enough for a theorist to insist on the sameness of the principles that apply to emergencies and non-emergencies; she has to explain how a principle can prescribe opposite courses of action in emergencies as opposed to emergencies and yet be the same principle. Nor is it sufficient for a theorist to insist on the need for exception-clauses in emergencies, based on the dissimilarity of emergencies to non-emergencies. She has to explain why the exception clauses exemplify the general and universally applicable principles that apply to both contexts. In other words, if it’s wrong to lie in ordinary life, but right to lie in an emergency, what needs explanation is what single principle is exemplified in both contexts, and how that single principle demands truth-telling in the one case and lying in the other. The constraint seems to me a plausible one, even if Rand herself doesn’t explain exactly how it works, and even though most theorists (in my experience) seem to violate or ignore it.

Having said that, it’s worth adding that some philosophers have emphasized the constraint. For better or worse, Mill does so in the last few paragraphs of Utilitarianism. A more recent example is Alasdair MacIntyre’s trio of papers on moral dilemmas and truth-telling in Part II of Ethics and Politics: Selected Essays, Volume 2 (2006). Another interesting discussion is Thomas Hill’s “Making Exceptions Without Abandoning the Principle: or How a Kantian Might Think about Terrorism,” in Dignity and Practical Reason in Kant’s Moral Theory (1992). There are, of course, many others.

I think Rand gets some other things right.

(2) She’s right, I think, that the philosophical literature misuses emergencies and moral dilemmas to the point of systematically indulging in fallacious appeals to the emotions in order to induce readers to accept otherwise under-argued moral claims. (Peter Singer’s drowning child example is a paradigm of this approach–an approach my graduate school roommate Patrick Kain once aptly dubbed “The Argument by Surely Operator,” as in, “Surely, we would all save the child…”  Interestingly and counter-intuitively, both Singer and Rand regard it as utterly uncontroversial that one has an obligation to save the drowning child, but neither of them has an argument for it. I have never, in twenty years of reading on the subject, seen or heard a bona fide argument for saving the child beyond table-pounding appeals to intuition, consensus, or plain old emotion.)

(3) Unlike much of the literature, Rand insists on defining “emergency.” It’s amazing how much of the literature discusses emergencies without ever defining the term.

(4) Again, unlike most of the contemporary analytic literature, Rand’s definition of “emergency” takes the form of a definition by genus and specific difference (“An emergency is an unchosen, unexpected event that…”) rather than biconditional equivalence (“An emergency takes place if and only if…”). This is a large topic, but I tend to think that the traditional Aristotelian format for definitions has advantages over the newer analytic one. I’m also inclined to think that Rand’s definition of “emergencies” is basically right, while admitting that she says nothing in defense of it, and admitting that any any adequate defense of it would probably require revisions to it.

Unfortunately, I also find Rand’s essay seriously defective. I see at least seven or eight basic problems.

(1) For one thing, though the essay is titled “The Ethics of Emergencies,” it’s not primarily about the ethics of emergencies at all. In my edition, the essay is about eight pages long, and is about a variety of topics, none of them particularly well developed. Rand opens the essay with a page-and-a-half-long well-poisoning polemic against the misuse of emergency scenarios in ethics. Another four pages go to a paradoxically illuminating but ill-argued explication of the egoistic basis of human relationships. A mere page and a half discusses emergencies. Another half page concludes the polemic against the misuse of emergency scenarios in ethics, focusing on the role that such scenarios play in defending altruism. Despite the title of the essay, then, the discussion of emergencies ends up being a mere afterthought to, and application of, the claims about the egoistic basis of human relationships. The cumulative result is that the essay is torn between at least two different topics–relationships and emergencies–and fails to do justice to either topic.  For that reason, though I think the essay has some worthwhile things to say, it is, on the whole, a failure.

There are other serious failures of argumentation in it, failures that have essentially gone undiscussed in the “literature” on the subject, almost all of it written by writers sympathetic to Rand. (For a notable exception, read pp. 100-104 of Carrie-Ann’s 2008 Reason Papers review of Tara Smith’s Ayn Rand’s Normative Ethics.)

(2) In discussing egoistic relationships, Rand repeatedly stresses that a rational egoist forms a hierarchy of values, incorporates other people’s well-being into that hierarchy, and then treats others well-being as part of his own. What she doesn’t explain is how or why anyone would do that. Yes, if you incorporate others’ well-being into your own, you come to treat their well-being as part of your own, and thereby promote their good while promoting your own. But why would an egoist do that? What benefit does an egoist get by making the initial choice? She doesn’t explain. Her failure to explain would, by itself, merely be an omission rather than a defect in her view (or in the essay), but having omitted a discussion of that crucial topic, she then goes on to make very strong claims about the structure that a rational hierarchy of values would have to take, as well as the actions that would have to follow from its adoption. But unless she specifies the benefit that an egoist gets from doing as she prescribes, she’s not entitled to such claims. What we need but don’t get in the essay is a derivation of other-regard from egoistic self-interest, not a rhetorical deflection of the accusation that egoists are mean people.

(3) Rand divides her discussion of other-regard into two parts: how an egoist deals with his intimates, and how he deals with strangers. As for intimates, she says, we incorporate their welfare into ours based on their past track record of virtue. Roughly speaking, the more virtuous they are, the more deeply integrated into our own welfare; the less virtuous, the less so. Though there’s deep truth in this claim, it’s also a gigantic oversimplification. But never mind that for now; let’s accept it ex hypothesi. One problem is that it explains how egoistic relationships continue, not how they come to be. I can continue to incorporate your welfare in my hierarchy of values based on your past track record insofar as you have a past track record (with me). But there has to be some initial point at which you lacked a track record (with me). How then do relationships begin?

The same problem applies even more problematically to strangers. A stranger by definition has no track record with another stranger. So we can’t incorporate the welfare of strangers into our own on the basis of their past track record of virtue. What then do we do? Rand acknowledges the problem here, and claims to deal with it, but what she says is very compressed and obscure (p. 53). We grant a stranger a “generalized respect and good will,” she says, on the basis of “the potential value he represents” (my italics). Later she describes the rationale for this respect and good will as “a consequence, an extension, a secondary projection” of one’s own self-esteem (p. 53, my italics).

There’s an intriguing idea lurking here, but taken at face value Rand’s claims are pretty puzzling. Why does a stranger “represent” any value to me at all, especially if I’ll never see him again? And why does he “represent” value rather than straightforwardly being valuable to me (or not being so)? Elsewhere, in the context of the debate about abortion, Rand derides the idea that potentialities are morally significant (“Of Living Death,” reprinted in The Voice of Reason). There, she says, the embryo and fetus are merely potential human beings, hence not rights-holders. Here, however, the argument asserts that a stranger is potentially valuable, hence valuable. The latter inference seems ad hoc, and the two claims together seem inconsistent. Rand doesn’t seem to have had a consistent position on the normative significance of potentialities, a problem, I suspect, that lies at the heart of the common accusation that she has no adequate ethical account of children, the family, and moral patients.

Finally, the language of “projection” is equivocal and potentially problematic. What does it mean? Here are some possibilities:

• We project the value of the stranger in the Goodman-Quine sense of treating his being-valuable-to-us as a “projectible” predicate.
• We project the value of the stranger in the Humean or Freudian sense of ascribing to him traits that belong to us.
• We project the value of the stranger simply in the sense of expressing our self-value in relation to him, treating him as its proper object and beneficiary.
• Some combination of the preceding.
• None of the preceding.

Unfortunately, she doesn’t say a word to decide between the options.

(4) Rand rather arbitrarily asserts that we ought to give assistance to strangers only in emergencies, but she never explains either the restriction or the ought:

• Why help only in emergencies (as opposed to elsewhere)?
• Why help in emergencies at all (as opposed to not helping even there)?

Again, she says little or nothing to clarify.

(5) Rand equivocates as to whether helping strangers in an emergency is a moral obligation (an “ought”) or merely a permission (a “may”). Sometimes she suggests that we ought to help strangers in an emergency (as long as the risks to us are “minimal”), implying that anyone who doesn’t help is morally defective, or lacking in virtue (specifically, lacking in integrity). Sometimes she suggests that we may help others in an emergency if we wish, but only if we do so from a sense of good will, and only if we want to, implying that we are not defective if we don’t help. Both claims can’t be right, and the latter claim contradicts Rand’s insistence, elsewhere, that her Objectivist Ethics is an ethic of conditional “necessities,” according to which every morally right act is a conditional imperative necessitated by “the conditionality of life” (“Causality Versus Duty,” in Philosophy: Who Needs It). Prima facie, a necessitated permission is a contradiction in terms.

(6) Having told us that we ought only to assist strangers in an emergency, Rand abruptly contradicts this claim, and tells us that it’s permissible to assist strangers in cases of illness or poverty that aren’t emergencies (p. 55). What she’s saying here may be a concession to commonsense, but in the context of the essay, it really seems like nonsense. Either we help others only in emergencies, or we help others in emergencies and elsewhere. We can’t have our “only” and eat it. An additional problem here is that her definition of “emergency” leaves no clear way of conceptualizing medical emergencies. But medical emergencies are a paradigm case of emergencies.

(7) Though I’m inclined to agree with her definition of “emergency,” she shows no awareness of how controversial it is, or of the most obvious objections that a reader might have to it (namely, that it seems too narrow). Even if I ended up agreeing with her definition, however, I’d have to disagree with her applications of it to cases.

(8) Rand conspicuously fails to meet her own adequacy-condition for a discussion of the subject, as described above.

For present purposes, as well as for purposes of my RP symposium contribution, I want to continue the task that Rand started–defining “emergency” by genus and specific difference. As I see it, the two basic definitional questions about emergencies are these:

(a) What are emergencies?

(b) Of what kinds of thing are emergencies properly predicated?

The two questions are verbally distinct but conceptually interconnected. Any answer to (a) probably constitutes an answer to (b), though I don’t think that the reverse is bound to be the case.

Rand defines an emergency as “an unchosen, unexpected event, limited in time, that creates conditions under which human survival is impossible” (“Ethics of Emergencies,” p. 54). She adds, in elaboration of the definition, that “[i]n an emergency situation, men’s [sic] primary goal is to combat the disaster, escape the danger, and restore normal conditions” (p. 54). Presumably by “men” she means “rational agents acting rationally”: men aren’t the only people who respond to emergencies, and some men fail to respond to them at all.

In any case, Rand takes “emergency” to contrast with “events that take place in normal conditions,” so that either (i) “emergency” and “normal-condition event” are correlatives, or (ii) “emergency” is defined in terms of (and as a basic deviation from) some conceptually prior notion of a “normal-condition event.” I’m inclined to think that (ii) is the case. If so, a great deal of the definition of “emergency” depends on spelling out that prior conception of metaphysical “normality.”

“Normality” is a notoriously difficult concept to define in a non-circular, non-statistical way.  The concept is one to which Rand often helps herself, and which bears some unclarified relation to what she elsewhere calls “the benevolent universe premise.” But it also finds its way fairly often into the non-Randian literature, whether in discussion of emergencies or other topics. Here’s one example from the scholarly literature on Aristotle’s ethics: The virtuous agent, Terence Irwin writes, “will correctly regard as dominant those rational and rigid states of character that secure complete happiness in moderately favourable external circumstances” (“Permanent Happiness: Aristotle and Solon,” in Aristotle’s Ethics: Critical Essays, p. 15, my italics). I take Irwin’s reference to “moderately favorable external circumstances” to pick out a concept similar to Rand’s notion of “normal-condition events.” “Favorable circumstances” seems to do analogous work for Aristotle as for Rand.

There’s a lot going on behind Rand’s definition–a lot of questions to be asked of it, and a lot of refinements to be made to it even on the most charitable reading. Here are a few questions that seem to me worth asking, and which I hope to address in my RP essay.

1. Suppose that “emergency” is parasitic on some prior notion of metaphysical normality. What is metaphysical “normality,” and how is it to be conceptualized prior to and independently of “emergency”? On the other hand, if “metaphysical normality” and “emergency” are correlatives, is there some other concept that allows us to break free of the conceptual circle they form?
2. Rand says that emergencies are unchosen, but why, or in what sense, can’t you bring an emergency on yourself? Does Rand mean that you can’t self-consciously bring about an emergency under that description (“I’m going to bring about an emergency right now”), or does she mean that you can’t chose to bring about an emergency, full stop? If I plan to commit suicide, and then start to enact the plan, is the enactment not an emergency even if it’s genuinely life-threatening? There’s something right about her claim, but it has to be made more precise and explicit than she makes it.
3. Rand says that emergencies are “unexpected.” Does that mean that they’re unpredictable? If so, why, or in what sense, can’t they be predicted? Mutatis mutandis, the same follow-up questions apply here as in #3. If, for instance, I correctly predict a hurricane, and correctly predict that it will lead to fires in a certain city, is Rand’s point that the subsequent fires aren’t emergencies? Or is her point that while “fires” are predictable of the hurricane as event types, the corresponding event-tokens are unpredictable emergencies? In other words, what is expected (but not an emergency) is “fires resulting from the hurricane”; what is unexpected and is an emergency is, say, the particular fire that breaks out on Broad Street at 10:43 pm during the hurricane. If so, Rand’s view seems to imply that while the hurricane itself is not an emergency, unexpected micro-events caused by it could be.
4. That, however, raises another set of questions. How predictable does an event have to be to qualify as “predictable”? I may not be able to predict that I’ll have a serious traffic accident today, but if I’m on the road and see a bus hurtling toward me–and have nowhere to go as it does–I may be able to predict, a few seconds before the event, that a serious traffic accident (construed as an emergency) is about to happen. Does that sort of predictability count or not? Here again, Rand’s definition needs some “chisholming.”
5. What justifies the stricture that emergencies are “limited in time”? The claim seems plausible, but needs an argument. Among the considerations that make it plausible: human pre-history lasted a long time but involved an enormous degree of pain, suffering, and duress; we might be inclined to regard the latter phenomena as emergencies if we suffered them, but there is something problematic about describing hundreds of thousands of years of human history as one long “emergency.” Supposing that emergencies are limited in time, how are the limits set? After all, some wars (or totalitarian dictatorships) last decades or centuries, and create life-threatening conditions. Does Rand want to deny that wars or dictatorship are emergencies? (It’s worth remembering that her novel We the Living features a protagonist who lies and cheats her way through life in order to survive as best as she can under Soviet socialism.) Perhaps, as in the hurricane case, only unexpected micro-events within wars or dictatorships are emergencies, but the macro-events themselves are not. If so, recurring causal chains of emergencies under long-lasting emergency-prone conditions are not (qua chains) emergencies, even if they require massive adjustments to the adoption and practice of “normal” moral principles (as they obviously do). Similar issues arise about life in, say, prisons, concentration camps, totalitarian states, and epidemics–generally, of phenomena unfavorable to life but durably long. (For an interesting novelistic portrayal of life in an epidemic, I’d recommend reading or watching Somerset Maugham’s The Painted Veil. [Afterthought, October 23: Obviously, another one is Camus’ The Plague!]).
6. Rand says that emergencies “create conditions that make human survival impossible.” Prima facie, this seems too strong. Most dictionaries tell us that emergencies threaten life, health, and/or property, or raise the probability of threats to survival.  Why insist on the strong modal claim–that emergencies create survival-impossible conditions? Further, does survival have to be threatened at all in an emergency? Imagine a case where someone gets her hand terribly stuck in a glass pickle jar, and calls 911. Is that an emergency? (That’s a real example described to me by a paramedic.) What about bathroom emergencies, e.g., desperately needing to use the bathroom but not being able to find one?
7. Suppose that we accept Rand’s definition as stated. Does the definition entail that “emergency” and “non-emergency” is a proper distinction, i.e., mutually exclusive and jointly exhaustive? Or is it compatible with the existence of borderline events between emergencies and non-emergencies, or events that fall into neither category?
8. Suppose that emergency/non-emergency is a proper distinction. Is it a sufficiently  fine-grained distinction, or are there non-emergencies that are sufficiently like emergencies to justify our thinking of them as quasi-emergencies, and to justify our creating a precise, fine-grained vocabulary to describe them (e.g., ‘crises’)?

I’m curious what readers think about any or all of this. In later posts, I’ll offer some answers of my own.

P.S.: Here’s a directly relevant article from out of today’s New York Times science section: “Ethicist Calls CPR Too Risky in Ebola.” I have no specialized knowledge of the issue, but Fins’s advice strikes me as nearly self-evident. I also agree with Fins that the issue needs to be discussed more widely than it has.

I was, incidentally, somewhat surprised to discover that medical ethicists don’t just regard “self-sacrifice” as virtuous, but as a special virtue of its own. Does anyone know of a good argument for that?

I’m curious to know whether Arthur Caplan has an argument for the claim that it would be morally wrong for health care workers to invoke Fins’s advice in cases of patients who are asymptomatic for Ebola but have (or even may have) come into contact with an Ebola patient. It simply isn’t obvious to me that health care workers have a duty to resuscitate in cases where there is–symptoms or not–a reasonable basis for believing that someone might have Ebola. Medical ethicists seem to me in the bad habit of offering prescriptions about the risks that health care providers are obliged to bear without really being able to explain why they should or must bear them. Relatedly, it’s worth pointing out that you can’t prove that p by asserting that you “can’t help thinking that p must be true.”