|
|
Normative Ethics From pp. 179-221 of Ethical and Religious Thought in Analytic Philosophy of Language, by Quentin Smith, Yale University Press, 1997. 24.
Hurka’s Perfectionism and Linguistic Essentialism In addition to arguing for ethical naturalism, I shall address issues about the ethical meaning of human life that so far have gone unaddressed in this book. Up to this point, the discussion has been confined to meta specifically, to whether moral realism or antirealism is true. But what about normative ethics? If we know that moral realism is true, we know that human life has an objective ethical meaning. But this will not tell us what this ethical meaning is. How should we live? What are the norms that govern the behavior of moral agents? A discussion of the meaning of human life is insubstantial without some arguments in normative ethics, so I shall argue for some theses in normative ethics, specifically, for a version of perfectionism that is partly influenced by Hurka’s book Perfectionism (1993) This argument is a part of my project of discussing and evaluating essentialist ethics, for Hurka’s ethical theory is based in large part on the essentialists’ method of linguistic analysis, and my development and modification of his perfectionism will also be based on the essentialists’ method. The essentialist movement made possible a type of moral theory that had gone virtually undeveloped in the analytic tradition up until Hurka’s Perfectionism. Perfectionist ethics (or at least what Hurka calls “narrow perfectionism,” which is his topic) is based on the idea that the good is the development of human nature. Prior to essentialism, the idea of a thing’s nature was seen as incoherent, obscure, or at least merely relative to a way of describing something. A main pre-essentialist position (whose main champion was Quine) was that necessity does not belong to objects (there is no necessity de re) but is relative to certain ways of describing objects. For example, relative to the description of humans as social beings, it is essential to humans to communicate. And relative to the description of them as animate beings, their essential nature is life. So the perfectionist idea of “developing human nature” has no univocal sense; it gives no one moral ideal, but an indefinite number of different and often conflicting moral ideals, depending on the description we use to pick Out humans. Thus, perfectionism is not a viable moral theory if essentialism is false. But the arguments for linguistic essentialism developed in the 1960s and 1970s provided the needed groundwork for a perfectionist ethics. Hurka writes, “Philosophers used to dismiss perfectionism by saying the concept of human nature is incoherent or obscure. This attitude is less common since Kripke.”[1] Hurka does not elaborate, but one may ask exactly how essentialism or the essentialist’s linguistic method enables us to make sense out of the concept of human nature. Consider Hurka’s interpretation of how the essential properties of humans and other things are known. He says there are two methods, the intuitive, associated with Kripke, and the scientific, associated with Putnam. Hurka describes the intuitive method as involving thought experiments involving candidate members of a kind: “To learn whether its atomic structure is essential to gold, for example, we imagine a series of possible substances with gold’s atomic structure but with a different outward appearance. If we judge all these substances to be gold, our judgement shows that its inner constitution is essential to gold and its phenomenal properties contingent” (34). Hurka describes the scientific method of essentialism as determining what is postulated in explanations given by good scientific theories. Science tells us that the atomic structure of gold explains the phenomenal properties of gold, its yellowness, solidity, weight, and so forth. Because gold’s atomic structure is explanatorily prior to its phenomenal properties, this structure is a nontrivial essence of gold. It seems to me that Hurka is not entirely accurate in saying that Kripke adopted the intuitive method and Putnam the scientific method for knowing the essences of natural kinds. Rather, they both used a certain hybrid of the two methods Hurka distinguishes. For example, they both believed that we know by scientific investigation that a certain chemical structure explains the phenomenal appearance of gold, and that we know by modal intuition that this structure is essential to gold. Hurka uses both his intuitive and scientific methods of knowing essences to determine the nontrivial essential properties of humans. Hurka uses these methods to arrive at the thesis that human nature consists of practical and theoretical rationality, a cardiovascular and digestive system, and so on. Because perfectionism is the thesis that goodness consists in developing human nature, Hurka will draw the conclusion that goodness consists in developing our practical and theoretical rationality and in developing our organic properties (becoming healthier). How do Hurka’s methods of knowing essences show that humans are essentially rational and have a certain organic structure? Regarding the intuitive method of knowing these nontrivial essences, Hurka briefly notes that we would not intuitively consider any beings to be human unless they were rational, that is, could form the relevant sort of justified beliefs and intentions to act. He writes, “If we imagine a species with no capacity for a mental life, or with none more sophisticated than other animals, we do not take ourselves to be imagining humans” (39). This seems to me a plausible reliance on our modal intuitions, given the defense of modal intuitions I gave at the end of chapter 4. But Hurka’s more substantial account appears in his description of the scientific method of knowing human nature, which in the present case involves the science of psychology. The essential property of humans that distinguishes humans from other animals is rationality. (Hurka defines “rationality” in terms of types of highly sophisticated beliefs and intentions that are not possessed by whales, dolphins, and so forth, which may be considered to have rationality in some different definition of the word.) Hurka plausibly takes as the relevant phenomenal properties of humans our observable actions or behaviors. He argues that psychological explanations posit rationality as the explanation of human action, much as chemical explanations posit the atomic structure of element 79 as the explanation of the phenomenal features of gold. The premises of psychological explanations of human actions refer to human beliefs about scientific laws and sophisticated intentions for the distant future. These premises refer to practical rationality, a nontrivial essence of humans. Theoretical rationality is also postulated in psychological explanations, for psychologists explain the beliefs that are relevant to actions in terms of humans having evidence that a certain activity is the most effective means to an end; forming sophisticated beliefs, grounded in scientific evidence, about the most effective means to an end is an instance of theoretical rationality. This account seems plausible as far as it goes, but it seems to me that Hurka should say in addition that theoretical rationality is postulated to explain the observable human behavior evinced in writing books, teaching and learning in educational institutions, and talking about science and philosophy. In short, the most important phenomenal or behavioral evidence that humans have theoretical reason is not their practical behavior, but their theoretical behavior. I want to reformulate Hurka’s theory in terms of our understanding of linguistic essentialism, which Hurka intends to be applying to his ethical theory. We may say that the sentence, “Gold has the Aristotelian essence of having 79 protons” is analogous to the sentence, “Humankind has the Aristotelian essence of rationality and certain digestive, nervous, etc. systems.” The reference—fixing description for the natural kind word “gold” is that this word refers directly and rigidly to what ever is causally responsible for the observable properties of being metallic and yellow and so on. The reference-fixing description for the natural kind locution “human nature” is that this term refers directly and rigidly to whatever nontrivial essences are causally responsible for the observable behaviors of humans. Are the sentences in which these rigid designators appear cases of necessary a posteriori identifications or cases of necessary a posteriori nontrivially essential ascriptions? Although Hurka does not raise these issues, it seems the answer is that both sorts of sentences are involved: (a) A Posteriori Identifications. As I have indicated, “human nature” is a directly referential rigid designator of a certain complex property, and the reference- fixing description is that “human nature” directly refers to whatever essential properties explain the behavior and physical appearance of humans. Let us use “organic structure F” to designate the digestive tract, sentient structure, and so forth that are Aristotelian essences of humans. “Human nature is rationality and F-organicity” then expresses the identity proposition rationality and F-organicity is (identical with) rationality and F-organicity. This proposition is known a priori. But the metalevel proposition that “human nature” directly refers to rationality and F-organiticity is known a posteriori. Empirical knowledge is needed to learn that rationality and F-organicity rather than, say, a robotic structure explain the phenomenal properties of humans. (b) A Posteriori, Nontrivial Essential Attributions. If we consider any particular that exemplifies the property designated by “human nature,” then this will be a nontrivial and Aristotelian essence of that particular. “Jane is rational and has F-organicity” expresses a singular proposition that is synthetic, metaphysically necessary, and a posteriori. Further, “All humans are rational and have F-organicity” is a nontrivial essential ascription, but it expresses a general rather than singular proposition. How does the essentialist method of knowing the Aristotelian essences of humans enable a normative ethics to be developed? In Hurka’s theory, this method enables him to reach such normative conclusions as that right actions are those that maximize the development of human rationality and F-organicity, and that good states are those in which humans’ rationality and organicity are developed. At this juncture, it seems we have a plausible beginning of a normative ethics that is known via the linguistic method of contemporary essentialism. It will appear, however, that there are some foundational problems in Hurka’s theory of human nature that call into question whether his theory is really an instance of perfectionism. I shall argue that a solution to these problems is to adopt a naturalistic version of perfectionism. 24.1
Hurka’s Theory of Human Nature The basic problem with Hurka’s narrow perfectionism concerns his procedure of formulating a definition of human nature. Hurka characterizes narrow perfectionism in terms of “the foundational idea that what is good, ultimately, is the development of human nature” (3). Broad perfectionism is a theory that is based not on the specific idea that the development of human nature is good, but on the more general idea that the good is the development of capacities or the achievement of excellence in art, science, and culture. Hurka proposes to construct a narrow perfectionism, and the key task is to formulate a definition of human nature: “To develop the best or most defensible perfectionism, we need, most fundamentally, the best concept of human nature” (9). Hurka proposes two tests that a definition of human nature must pass to be “the best concept of human nature”: (i) human nature must come out as morally significant and (ii) there must be intuitively plausible moral consequences of the definition. The definition must assign intrinsic value to certain properties; that is, the properties included in the definition of human nature must have intrinsic value, and they must seem to be morally worth developing. This implies that the proper ties are degreed properties, properties capable of admitting more or less, since a nondegreed property cannot be developed. A definition of human nature will fail to pass the moral test if it succumbs to what Hurka calls the wrong-properties objection. This is the objection that a suit able definition of human nature cannot include properties that intuitively seem to be morally trivial (such as developing humans’ essential property of occupying space) or morally bad (for example, being capable of killing members of one’s own species). Hurka examines several definitions to see if they pass the moral tests. Once he selects the best definition, he will proceed to construct the details of a normative ethics that is based on the idea that the moral goal is the development of human nature as specified in that definition. The definitions Hurka considers, and his reasons for rejecting most of them, are presented in what follows: (Dl)
Properties Distinctive of Humans The first definition Hurka discusses is that human nature is the set of properties possessed only by humans. He says this definition succumbs to the wrong-property objection. Humans uniquely make fires and uniquely kill things for sport. Hurka says these are not properties that are morally worth developing, and thus this definition of human nature is not suitable for a perfectionist ethics. Hurka notes a second problem. The best definition of human nature must depend only on facts about humans. But the definition in terms of distinctive properties depends on facts about other species, for example, whether they possess rationality. This definition would imply the implausible consequence that we should conclude that human nature does not include rationality if we happen to determine that dolphins, for example, are rational (in the sophisticated sense of “rational” Hurka has in mind). (D2)
Properties Essential to Humans A property is essential to the kind human if and only if every member of this kind possesses this property in each possible world in which it exists. A property essential to humans depends only on facts about humans; we need consider only what properties humans possess in the possible worlds in which they exist. Hurka says this definition also succumbs to the wrong-properties objection, for it includes the wrong properties in human nature. It includes morally trivial properties, such as being self—identical, being red if red, and being occupiers of space. These properties are not morally worth developing. (D3) Properties Essential and Distinctive
to Humans A property F is essential and unique to humans if and only if (a) each human x possesses F in each possible world in which x exists and (b) only humans possess F in the actual world. Hurka finds this definition unsatisfactory because it too succumbs to the wrong-property objection; it includes the properties of being able to light fires and killing for sport. A second problem is that this definition depends on facts about species other than humans; for example) if it turns out that some other species is rational, rationality is not a part of human nature. (D4)
Properties Essential and Necessarily Distinctive of Humans A property F is essential to and necessarily distinctive of humans if and only if each human possesses F in each possible world in which she exists and there is no possible world in which anything other than humans possess F. This sort of property is analogous to Plantinga’s “individual essences,” except that they are natural kind properties. Hurka notes that this definition excludes from human nature many morally valuable properties that are worth developing, for example, rationality. It is at least possible that there are other beings with rationality, and so rationality is not necessarily unique to humans. Hurka concludes from this that the notion of distinctiveness must be rejected in attempts to come up with an adequate definition. He suggests we instead examine more closely the notion of essential properties and construct a satisfactory definition in terms of several nontrivial essences of humans. (D5)
The Essence-and-Life Definition The essence-and-life definition is the one Hurka eventually concludes is the best definition. Hurka notes that each human has six classes of essential properties, distinguished by the range of objects that possess the property. These essences include trivial essences, Aristotelian essences, and individual essences. If ordered in terms of their extensiveness, they would appear as follows. The properties are 1. Essential to humans qua objects: Essential properties shared by all and only objects, for example, self-identity. 2. Essential to humans qua physical objects: Essential properties shared by all and only physical objects, for example, occupying space. 3. Essential to humans qua living things: Essential properties shared by all and only animate objects (for example, matter organized for functions of nutrition). 4. Essential to humans qua animals: Essential properties shared by all and only animals (Hurka offers no example of an essential property that is essential to animals but not to nonanimal living things). 5. Essential to humans but not to any other animals (rationality is such an essence, given that current empirical theory is correct in supposing that no other animals possess rationality in Hurka’s sophisticated sense). 6. Essential properties that distinguish one human from another (for example, a person’s genetic profile or a person’s origination from these particular sperm and egg cells). The definition that Hurka selects as the best definition of human nature includes the third, fourth, and fifth essential properties of humans; the essential properties we share with living things and animals and the essential properties that distinguish humans from other animals. By “the best definition” Hurka means the best definition for the purposes of constructing a theory of ethical perfectionism. Hurka rules out numbers i and 2 because he holds that our essential properties qua objects and qua physical objects are not morally significant; for example, self-identity and being spatially extended. But he holds that becoming healthy and developing our nature as animals (which Hurka never specifies), which are degreed properties corresponding to 3 and 4, and developing our practical and theoretical rationality (corresponding to 5) are morally significant. Hurka rules out 6, individual essences, because a person’s genetic profile may include abilities for fire-lighting, stealing, or being a serial killer, and thus category 6 succumbs to the wrong-properties objection. 24.2
An Evaluation of Hurka’s Theory of Human Nature I shall argue in this section that Hurka’s perfectionist ethics is implicitly self- contradictory, although many of the plausible theses he develops can be retained if his theory is reformulated to avoid the contradiction. We may approach this con tradiction by first considering some remarks that merely seem to embody a contra diction (the real contradiction will lie in a related aspect of his theory). As I indicated, Hurka mentions two tests of a definition of human nature; these tests involve at least the following two considerations: 1. “Human nature must seem in itself morally significant.” 2. “A perfectionist concept of nature assigns intrinsic value to certain properties” (9). He also makes the following claim: 3. To include a property in human nature (in a definition) is not to make an evaluative claim; it is not to evaluate the property as “important or desirable,” and so to use “prior evaluative standards.” Rather, it is to make a descriptive or factual claim (18). Are (2) and (3) mutually inconsistent? Does their conjunction imply that the perfectionist concept of human nature both assigns and does not assign value to a property F that it includes in its definition of human nature? I think a charitable reading might resolve the problem in this way. Perfectionism (or, more precisely, narrow perfectionism, which is about human nature) is the thesis that developing human nature is good. Accordingly, if we make the descriptive or factual claim that human nature includes the property F, we can deduce from the perfectionist thesis that F (or developing F) is good. That is, 4. Developing human nature is good (the perfectionist assumption). 5. F belongs to human nature (taken as a descriptive or nonnormative claim). Therefore, 6. Developing F is good. Here we are interpreting (3) as (5). We are interpreting (1) as equivalent to (4), and we are interpreting (2) as the claim that the argument (4)-(6) is sound. This seems to avoid the contradiction. Thus, it appears so far that there is no implicit contradiction in Hurka’s theory. But we have approached the contradiction. The contradiction lies in part in Hurka’s procedure for formulating a definition of human nature, specifically, in his reliance on the two “tests”: that the definition must make human nature seem morally significant and must entail attractive moral judgments. These tests are most clearly used in Hurka’s employment of the “wrong-properties objection” as a criterion for rejecting various candidate definitions of human nature. Note Hurka claims that perfectionism is a freestanding morality; he says that if the perfectionist selection of a definition of human nature were not based on purely descriptive criteria, then perfectionism would not be a “free-standing morality” (18) but would depend on the “prior evaluative standards” (18) used to select the definition. Hurka’s perfectionism, however, does depend on prior evaluative standards and is not a freestanding morality: the contradiction is that Hurka’s theses jointly imply that perfectionism both is and is not a freestanding morality. For example, Hurka writes that it counts as an objection to a certain concept of human nature that the “concept of nature includes properties that on their own seem morally trivial—if it gives value to what, intuitively, lacks it. This is a telling objection to the concept. A morality based on the concept will be hard to accept because it flouts our particular judgements of value” (9, emphasis added). The real foundations of Hurka’s ethics are “our particular judgements of value”; his perfectionism stands on these prior judgments of value. This fact is quite explicit in Hurka’s statement: “We also want a concept [ human nature] that avoids the wrong-properties objection, by having fall under it only properties that seem in their own right worth developing” (10). If, however, we allowed prior moral criteria (embodied in what seems in its “own right worth developing”) to determine our definition of human nature, then these moral criteria are the true basis of morality (not the ideal of developing human nature), and perfectionism then becomes equivalent to (a notational variant on) a moral system based on these nonperfectionist moral criteria. Suppose we had the prior moral judgment that it is good to acquire philosophical knowledge. On this basis, we could define human nature (at least in part) in terms of humans’ ability to acquire philosophical knowledge. “It is good to acquire philosophical knowledge” is then redundantly expanded to become, “It is good to develop the natural ability to acquire philosophical knowledge.” We do not need any reference to human nature or developing natural abilities to state the moral truths, and so narrow perfectionism becomes an empty theory. Hurka writes that one of the claims narrow perfectionism has on present moral thought is that “perfectionism, when combined with a well-grounded theory of human nature, entails attractive particular judgements. Many of us believe that states such as knowledge, friendship, and the completion of challenging tasks are good intrinsically, that is, apart from any satisfaction they bring. The best perfectionism entails judgements about these states that either match those we already make or take us beyond them in a way we can recognize as progress” (4-5). Suppose we believe that knowledge, friendship, and the completion of challenging tasks are intrinsically good. We then select, among the several possible factual definitions of human nature, the one that entails, in conjunction with the assumption that developing human nature is good, that knowledge, friendship, and so on are good. What, then, is the justification for believing that knowledge and the others are good? The justification is not that it is entailed by the selected definition of human nature, in conjunction with the premise that developing human nature is good. For the beliefs about the goodness of knowledge themselves constitute the justification for selecting a certain definition of human nature. Because we in dependently believe that knowledge is good, we do not need the perfectionist assumption, 1. Developing human nature is good, to justify this moral belief. This argument can be presented more formally. Suppose we conjoined (i) with a definition of human nature that entails knowledge is good. Thus, 1. Developing human nature is good and 2. Human nature includes rationality jointly entail 3. Knowledge is good. Now despite the deduction of (3) from (1) and (2), (3) is not justified by this argument because it is an antecedently plausible moral thesis and serves to justify the selection of (2) as our definition of human nature. Maybe there is a way to make Hurka’s procedure seem plausible. Could not (3) confirm the conjunction of (1) and (2); perhaps the perfectionist theses are con firmed by our independent moral judgments. This possibility is ruled out by the fact that (2) is a stipulative definition of human nature, not a real definition. We are stipulating that “human nature” means (2) because by virtue of this stipulation (3) can be derived. Because a stipulation cannot be true (stipulations have no truth-value), the stipulation cannot be confirmed. Hurka’s theory is in effect a version of intuitionist ethics, not perfectionist ethics. “Intuitionism” is here understood as the theory that normative ethics is based on articulating the more or less independent particular or general moral judgments that we make in our everyday lives and find intuitively or intrinsically plausible. Moral truths, according to intuitionism, are not justified by a systematic criterion, such as the principle of utility, a Kantian deontological principle, or the criterion of developing human nature. Rather, they need no justification other than themselves; our particular moral judgments (and our general moral judgments) are self-standing. The underlying problem is that while claiming to offer a perfectionist ethics, Hurka in fact offers an intuitionist ethics. He correctly writes that human nature perfectionism has for its “foundational idea that what is good, ultimately, is the development of human nature” (3). But he abandons this foundational idea by basing his normative ethics on the foundational idea of intuitionism, that we determine what is good by relying on the various intuitive moral judgments we make in our daily lives. Is this assessment unfair to Hurka’s theory? Consider his three claims on behalf of perfectionism: (i) He says, “That the human good rests somehow in human nature is, although elusive, also deeply attractive” (4). In order to make this thesis attractive, however, Hurka has to select from among a variety of definitions of human nature one that is designed to appeal to our antecedent moral judgments. What is attractive ultimately is not the perfectionist thesis about human nature, but these antecedent moral judgments. Indeed, because he himself argues that most of the definitions of human nature are not morally attractive, his theory suggests that it is largely false that the idea that developing human nature is attractive. (ii) Hurka’s second claim on behalf of perfectionism is that it entails attractive judgments. But this suffers from the same problem, for we begin with the theses we want to be entailed and then search for some definition of human nature that will entail these judgments. If our moral knowledge is already embodied in the entailed theses, the construction of some thesis about human nature that entails them is superfluous. (iii) The third claim is that perfectionism systematizes our particular moral judgments. If we choose among the variety of definitions of human nature a certain definition that entails our particular moral judgments, then it might seem that we do have a systematization of our particular moral judgments. But the problem is that we are here stipulating that “human nature” is to refer to the human properties F, G, and H, rather than the human properties I, J, and K, because this stipulation enables us to deduce our antecedent particular moral judgments. But because stipulations lack truth-value, there is no systematization. An ethical systematization requires that (a) the theses being systematized be theorems of a theory of which the systematizing principles are axioms and (b) the axioms give us reason to think that the theorems are true. But in Hurka’s case, the axioms lack truth-value and thereby provide no reason for us to think that the theorems are true. Condition (b) is not met by Hurka’s perfectionism and so his third claim, about systematization, is false. Thus, we are forced to the conclusion that Hurka has not developed a perfec tionist theory. He has developed an intuitionist theory. This problem with Hurka’s theory, however, is not as major as it might seem to some philosophers interested in developing a perfectionist ethics, because many of the other ideas developed in Hurka’s book are particularly relevant to a perfectionist ethics and may be profit ably used by a perfectionist theorist. The antecedent moral intuitions upon which Hurka relies accord in many cases with the theses that the perfectionist thinks are evidentially supported by genuine perfectionist premises. Apart from this kinship of Hurka’s intuitionism with genuine perfectionism we should note that Hurka himself does not seem averse to intuitionist or pluralist ethics, as he himself remarks in his book, so the conclusion that his ethics is an intuitionism, with some ideas that a perfectionist could use, would not seem to be a major criticism of his theory. In spite of my criticisms of his foundations as being in fact intuitionist, Hurka’s book seems to me to be the most important contribution to perfectionist ways of thinking since Aristotle’s Nicomachean Ethics, and the perfectionism that I develop in what follows would not have been possible without it. In the next section I shall show how a perfectionist ethics can be developed, using the tools of linguistic essentialism. I argue that there are reasons to believe that this perfectionism is a satisfactory normative ethics and provides us with an objective ethical meaning of human life. 25.
Global, Naturalistic Perfectionism and the Method of Linguistic Essentialism A perfectionist ethics can be
developed if we use the method of linguistic essential ism to a greater extent
than it is used by Hurka. Hurka uses the essentialist method only for the
purpose of showing how we can know essential properties of humans. He does not
apply this method to the moral terms “good” and “right.” Hurka adopts a
nonnaturalist position and takes “is good” to express the sense of “ought
to be desired by agents” and “is right” to express the sense of
“maximizes the good.” Thus, in Hurka’s theory, “x is good because it
ought to be desired by agents” is an analytic, a priori, and necessary
sentence. This nonnaturalist definition of “good” is partly similar to If we apply the essentialists’ method of linguistic analysis to moral terms, we will be able to develop a plausible naturalistic perfectionism. Because the conjunction of ethical naturalism and moral realism is intrinsically more plausible than the conjunction of ethical nonnaturalism and moral realism, as I argued above, this naturalistic perfectionism will give us an even stronger reason to believe that human life has an objective ethical meaning. Hurka believes that any naturalistic perfectionism is necessarily false because it fails to account for the intimate connection between evaluation and action. Hurka believes that to assent to a moral judgment “one must act as the judgement directs, or at least form the intention so to act when circumstances are appropriate” (28). Hurka believes that ethical naturalism entails that there is no such connection between evaluation and action. I believe he is correct, but I believe that no plausible ethical theory should entail there is such a connection. There is no such connection. Hurka’s claim that assenting to a moral judgment requires forming an intention to act on it, when circumstances are appropriate, seems false. I assent to “I ought to get out of bed and start working on the next chapter of my book” and yet, feeling lazy and slightly guilty, lie in bed for another half an hour. In this case, I am not forming an intention to act when the circumstances are appropriate because the circumstances are then appropriate for me to so act—but I do not because I am too lazy. And I am sincerely assenting to the moral judgment, because I feel guilty for not forming an intention to act on the judgment. Hurka is assuming internalism about motives, but as I have earlier argued, and as Brink has argued in chapter 3 of his Moral Realism and the Foundation of Ethics, externalism about motives is the correct position. I conclude that Hurka has given no convincing reason to think that a naturalistic version of perfectionism is untenable. In what follows, I argue for a version of naturalistic perfectionism that relies entirely on the essentialist method of linguistic analysis. “Human,” like “gold,” “water,” “cat,” and “tiger,” is a natural kind word. According to linguistic essentialism, it rigidly and directly designates a natural kind. A natural kind is a nontrivial essence or conjunction of such essences. In many cases, the reference—fixing description for a natural kind word specifies that this word refers rigidly and directly to the nontrivial essences that explain the phenomenal or observable properties that are normal to members of that kind. These nontrivial essences constitute the “kind of nature” of the things or, more briefly, the nature of the things. The reference-fixing description for “human” is that this natural kind word refers directly and rigidly to whatever nontrivially essential properties explain the phenomenal properties of humans (their physical appearance and their behaviors). These properties belong to some of the classes of essential properties Hurka mentioned, that is, what is essential to humans as physical objects, as living objects, as animals, and as distinctive from other animals. I do not include what is essential to humans as mere objects, for these essential properties are either trivially essential properties (for example, being identical with Socrates) or logically necessary properties (being self-identical). Trivially essential and logically necessary properties do not explain the phenomenal properties of humans (their behaviors and physical appearance) and thus are not among the designata of the natural kind word “human.” According to the naturalist perfectionism I shall be defending, “goodness” refers to the natural property being a development of things’ natures. What is said by the statement, “Goodness is a development of things’ natures” is an identity proposition, but we know that the statement expresses this identity proposition only a posteriori (as I argue below). The essences constituting the natural kind to which an
object belongs do not include all nontrivial essences of the object, for
example, its individual essences. Marcel Proust has the nontrivial essence, writing
in a cork-lined room in the actual world a, but this is not a part of Proust’s
“nature.” (Of course, it is a part of Proust’s “nature” if we mean by
this his individual essence, but I am using “Proust’s nature” to mean the
nontrivial essences that make him a member of the natural kind human.) My argument
will proceed by specifying various reference—fixing descriptions of
“good,” “right,” and so forth and showing that a posteriori
considerations indicate that such properties as being
a development of things’ natures satisfy these reference-fixing
descriptions. This method of ethical knowledge was first used (and discovered)
by My argument that the properties countenanced by global, naturalist perfection ism satisfy the reference-fixing descriptions for “good,” “wrong,” and so on cannot be clearly presented unless we first have a reasonably precise idea of the sort of property that I use “being a development of things’ natures” to designate. I shall first explain some of the main theses of global, naturalist perfectionism, and in the next section begin presenting some arguments for this normative ethics. According to naturalistic perfectionism, the reference-fixing descriptions for the relational predicate “is more valuable than” imply there are two uses of this predicate. In one use it designates the relation ( ) is a greater degree of development of the natural property F than ( ). This relation obtains between different degrees of development of a given essential property. An example is that Einstein’s epistemic states exemplify a greater degree of development of the essential property of theoretical rationality than do the epistemic states of an average person. The relational predicate “is more valuable than” also has a different use, in which it designates the relation ( ) has a more developed kind of nature than ( ). An example is that a whale has a more developed kind of nature than a pebble on a beach. The second relation allows global, naturalistic perfectionism to provide a hierarchy of kinds of things based on the degree of value of different kinds of natures. In this sense of “more valuable than,” a thing of kind K1 is more valuable than a thing of kind K2 if and only if K1 is a more developed kind of nature than K2 An explanation of this hierarchy will show what “more developed kind of nature” means. The hierarchy is based on the fact that higher kinds of things include the natures of lower kinds of things as parts of their own nature. At the bottom of the hierarchy are mere physical objects: electrons, stones, mountains, stars, and so forth. The nature of a mere physical object consists of the nontrivial essential properties of being spatiotemporally extended and having mass. Because a mere physical object has a developed kind of nature (even though it is the least developed kind of nature of the various kinds of concrete objects), by virtue of having these two nontrivial essences, a mere physical object has positive value just by virtue of being a physical object. And because its nature consists of these two degreed properties, which vary in degree among different physical objects, a physical object can have more or less value than some other physical object. A physical object x that is more massive and spatiotemporally extended than another physical object y is (all else being equal) more valuable than y. The moon has more intrinsic value than a stone, and the Andromeda Galaxy (even abstracting from whatever life-forms it may include) has more intrinsic value than the moon. A physical object develops its nature by either enduring in time (and thereby becoming more temporally extended), expanding in space (and thereby becoming more spatially extended), or becoming more dense (and thereby becoming more massive). Humans rarely have occasion to make intrinsic value judgments about mere physical objects, in that they are habituated to judging them only in terms of their extrinsic value as means to or conditions of the development of human beings. It would not be an exaggeration to say that theoretical discussion of the intrinsic values of mere physical objects is the least developed part (indeed, a virtually nonexistent part) of normative ethics. This is one respect in which global, naturalistic perfectionism departs from familiar types of normative ethics. It is “global” in that it is about the intrinsic value of all objects, including mere physical objects. (Hurka called his theory “narrow perfectionism” because it was only about the value of human beings.) Even most environmental ethics are not global in that they usually discuss the intrinsic values only of all animals or only of all living things or ecosystems. I am not appealing to antecedent and independent “moral intuitions” to deter mine what is more valuable than what. If I followed Hurka’s method, I would not be developing an ethical perfectionism but an ethical intuitionism. Rather, I am determining the values of things and their states by deducing their values from premises about the nontrivial essences that are the rigid designata of natural kind words. In global, naturalistic perfectionism, material or substantive moral theses about what is good or bad and what is better than what are deduced from premises about things’ natures, and the evidence that the perfectionist theory (axioms and theorems) is a true normative ethics is that this theory satisfies the formal conditions laid down in the reference-fixing rules for moral words. (This aspect of the essentialist method of ethical knowledge will become clearer in the next several sections.) To return to the subject of hierarchy, mere animate things (for example, plants, bacteria) have the next most valuable kind of nature. The nature of an animate thing includes the nature of a physical thing and in addition the nontrivial essences of reproducing and obtaining nourishment. (Here I side with the biologists who take bacteria, rather than viruses [which do not reproduce], as the most primitive known life-form.) This is the sense in which a mere animate thing has a more developed kind of nature than a mere physical thing. The development of the nature of an animate thing includes realizing its reproductive capacities (producing offspring), realizing its nutritive capacities (for example, performing photosynthesis), and in general maintaining increasing these organic functions to the degree that is optimal for the health and reproduction of the organism. Unless there are mitigating circumstances, the development of a mere physical thing’s nature (for example, expanding in space) is less valuable than that of a mere animate thing’s nature (for example, reproducing). It is bad that an ocean (considered only as a mere physical thing) increases its size and in so doing prevents the nutritive and reproductive functions of numerous plants that populate the shore line. But the ocean contains plants and fish, and if its increase results in a significant development of the plant and fish life in the ocean, then the ocean’s increase in size will be good overall, in spite of the harm done to shoreline plants. Moral considerations such as these are neglected in standard normative ethics, which tend to have humans as their sole subject matter, but are integral to global perfectionism. Animals have a more valuable kind of nature than merely animate things. In addition to being essentially massive and spatiotemporally extended, nutrition- obtaining and reproductive, animals have a further nontrivial essence. But here numerous complications arise owing to the various species of lower and higher animals, ranging from the most primitive known kind of animal (some members of the parazoa family, which have only a thousand cells) to mammals. To avoid the complications of a theory that embraces the natures of all known species of animals, I will talk only about animals that are significantly different from mere organisms, which are at least the vertebrates (for example, fish, birds, amphibians, reptiles, and mammals). Vertebrates (“the higher animals”) are distinguished from mere organisms by virtue of having the nontrivial essence of the ability to move in order to satisfy their desires regarding what they can sense or imagine. (Lower animals are self-realized in much the way mere organisms are, for lower animals have nutritive and reproductive abilities and differ from mere organisms in that they develop from egg and sperm cells and have double chromosomes. To simplify matters, I will use “animal” to refer only to the nature of the higher animals.) Some recent work by Arthur Falk has shown that the relevant properties—desire, sentience, and the ability to move so as to satisfy one’s desires—present in humans are present in more primitive forms in much less developed animals, and we may with justification use “desire,” “sentience,” and so on to refer to the properties of a wide range of animals.[2] The distinctive essence of animals is a degreed property, and its development consists in movements that successfully fulfill the animal’s desires regarding what it can sense or imagine. Because successfully fulfilling a desire is a pleasurable feeling of satisfaction, our global perfectionism is able to explain why pleasure is intrinsically good, a normative fact that Hurka’s “perfectionism” is unable to explain and is in fact inconsistent with (171). Because animals have more valuable natures than mere animate things, the preservation of an animal is morally preferable to the preservation of a mere animate thing (all else being equal). It is good that a cow eats the grass it desires to eats, and thereby kills the grass, because a cow is a more valuable kind of thing than the blades of grass it eats. The comparative principle at work here is based on a certain formal ethical principle, the finite difference principle, which differs from the infinite difference principle. According to the (false) infinite difference principle, any one thing or good state of a thing on a higher level of the hierarchy of kinds of things is more valuable than all the things or all the good states on a lower level of the hierarchy of kinds of things. One implication of an infinite difference principle is that all the plants on earth are less valuable than a single one of the billions of insects. Another implication is that the moral goal is always to increase the development of a thing of a higher kind, at the expense of anything lower, so that even the smallest increase in development of a single thing of a higher kind is more valuable than enormous increases in development of an infinite number of things of a lower kind. It is not this principle, however, but the finite difference principle that is the formal moral principal embodied in the reference-fixing descriptions of our comparative moral words. According to the finite difference principle, it can be true that the aggregate of all things, or all good states of things, on a lower level of the hierarchy can be higher in total aggregate value than some things or good states on a higher level. All the merely physical things in the universe are more intrinsically valuable than just one of the billions of bacteria that exist. But this comparison is dependent on the abundance of bacteria and other life-forms; if the only animate thing in the universe was one bacterium, its value would not be outweighed by the aggregate value of all merely physical things. This is an instance of another formal principal, the uniqueness condition, which comes into play in comparing the value of things and is at the basis of the judgment that it is much worse to let the last members of a certain species die out than to let a few members of a highly populated and unendangered species die out (assuming the species are at approximately the same level of nature development.) The finite difference principle also implies that in some cases, great increases in development of a large number of things of a lower kind add up to a greater amount of value than increases in development of some higher kinds of things. The flourishing of all plants on earth is of greater intrinsic value than the flourishing of just one of the billions of insects on earth. But the existence and flourishing of the billions of plants on earth is not intrinsically more valuable than the existence and flourishing of thousands of the scientists and philosophers that live in the twentieth century. (The existence of all plants is extrinsically more valuable than the existence of twentieth-century theorists because an earth without plants would not allow any higher kinds of things to exist.) One species of physical things may have more value than another species if it is more developed in the relevant sense, and the same holds for different species of animate things. Regarding the species of animals, the most developed species is the human species. The nontrivial essence of rationality (as I shall later define it) is possessed only by humans, and this essential property is the most developed kind of essential property that belongs to anything’s nature. At this juncture, we have enough of an initial idea of global, naturalistic perfectionism to present the arguments that this normative ethics satisfies the reference- fixing descriptions for the words “good,” “bad,” “right,” and “wrong.” 25.1 Perfectionism and the
Reference-Fixing Descriptions of Moral Words The reference-fixing descriptions for moral words determine the rules by which competent language—users use these moral words, and these descriptions in addition constitute the cognitive significance of these words. Regarding cognitive significance, we tacitly or explicitly have in mind one or more of these several descriptive senses (the cognitive significance) when we use these moral words. These moral words refer to certain properties, but an acquaintance with these properties is no more a part of the cognitive significance of these words than an acquaintance with being H2O molecules is a part of the cognitive significance of “being water.” Competent language-users do not have in mind as the cognitive significance of “good” that “good” directly and rigidly designates the property a development of things’ natures. Ordinary language analysts may correctly point out that this is not what we ordinarily would say “good” designates; but this only goes to show that ordinary language analysts failed to distinguish the cognitive significance from the semantic content of locutions. In the following, I explain six reference-fixing descriptions relevant to the moral uses of “good,” “bad,” “right,” and “wrong.” I call them conditions because they are formal descriptive conditions that something must satisfy in order to be the designatum of “good” (or some other moral word). After I explain each reference-fixing description, I shall show how the properties mentioned in global, naturalist perfectionism satisfy these descriptions. (R1) The Reaism Condition This reference-fixing rule of use of “good” states that “good” is used to refer to a property that is exemplified by items independently of whether or not we believe it is exemplified by these items (the same for “bad,” “right,” and “wrong”). If something is good, it is objectively good. I have already argued in several places in this book that first—level ordinary moral beliefs imply moral realism and thus I think we are justified in including the realism condition in the reference—fixing descriptions that govern the use of moral words. Global and naturalist perfectionism conforms to the realism condition. If “good” in one of its uses refers to the property a development of things’ natures, and, in another use, refers to a developed kind of nature, then our perfectionism meets this reference-fixing condition. For example, if on a certain occasion eighty million years ago a dinosaur developed its nature qua living thing by becoming nourished by the plants it was eating, the dinosaur exemplified this property independently of whether or not we believe it exemplified this property. According to naturalist perfectionism, development-properties involve a species / genus distinction. The dinosaur being nourished by plants is a species of the genus developing its animate nature, and this genus in turn is a species of the higher genus developing its nature. (R2) The Attitude Condition A rule of use of “good” is that we predicate this word of things or states of things toward which we typically have emotionally positive attitudes. When we predicate “good” of a possible and actualizable state of things, we typically desire to actualize this state or wish or hope that it is actualized or at least find its actualization acceptable or unobjectionable. When we predicate “good” of existing states of things or of existent things, we are typically pleased, satisfied, joyous, or at least accepting of these states or things. Predications of “bad” are typically accompanied by the corresponding negative atiitudes. Global, naturalist perfectionism satisfies the attitude condition; our positive attitudes associated with predications of “good” are attitudes toward developments of things’ natures or toward things that have highly developed kinds of natures. There are many examples of this, but we may choose a seemingly controversial one. When I praise a teacher for being a good teacher, I am praising her for teaching effectively and promoting significant learning in students. Teaching effectively is a way of developing a part of her nature (her theoretical and practical reason), and promoting learning is promoting the development of the students’ natures (their theoretical reason). This example, which is typical of perfectionist ethics, may face this objection: Is “good” in the phrase “a good teacher” a moral use of “good” or a nonmoral, task-related sense of “good”? I think it is a moral use of “good,” for there are the following connections: Suppose we believe that Jane lives a morally good life. If asked to give evidence for our belief, we would include in our evidence that she has chosen teaching as a career and that she is an effective teacher. If she were a criminal or a poor teacher, these facts would not count as evidence that she lives a morally good life. Task-related senses of “good” (“a task well done”) can be types or species of the moral sense of “good.” It is morally good to do well a task (profession, job, social function) whose defined goal is to develop people’s or other things’ natures. It is easy to think of examples of nature developments that meet the attitude condition, but the real test comes from putative counterexamples, such as the following. We admire or respect Jane for learning calculus (which is a development of her theoretical reason) and believe that her learning calculus is a good state of affairs. But suppose Jane is responsible for the welfare of children in a day care center and they all died because she spent her time studying calculus rather than caring for them. In this case, we would be outraged at Jane’s developing her theoretical reason. Does not this show that developments of natures do not satisfy the attitude condition? The answer is negative because a state of a thing that is a development of some thing’s nature can be part of a more complex state of things that consists of a greater amount of prevention of nature development than it does of nature development. The predicate “badness” in one of its moral uses refers to the property being a prevention of the development of things’ nature or, more exactly, being more of a prevention of the development of things’ nature than a development of things’ natures. The complex state of things Jane-learning-calculus-and-thereby-allowing the-children-to-die is bad (prevents more than it develops) even though one part of this complex state, Jane’s learning calculus, is good (is a development). Just as bad states can contribute to the overall goodness of a whole, so good states can contribute to the overall badness of a whole. Joseph Ellin makes a good point that the cases of animals killing and eating each other seem to many to be counterexamples to a naturalistic perfectionism.[3] A wolf realizes its essential nature by killing and eating a caribou; the wolf realizes good-ness by developing its nature, but what about the caribou? Clearly, the caribou’s premature and painful death is bad. So how can the wolf’s development of its nature be considered good? Certainly, the caribou would not agree to that proposition, so to speak. These sorts of questions cannot be adequately answered unless we consider the overall state of affairs, the wolf-eating—the-caribou. Is this complex state of things more bad than it is good or vice versa? There is no clear answer here without considering extrinsic matters. Is the wolf population too great for its territory? Are the caribou dying out, and will they soon be extinct? Is there no other way for the wolf to be nourished? Or is the caribou population becoming so great that unless a few were killed by wolves, the caribou would eat up their own food sources and all die from starvation? Whether this state of affairs is overall good or bad depends on whether answers to these other questions are affirmative or negative. There is also a possibility that there is a moral tie here: the good belonging to this state of affairs may equal the bad. The ambiguity in our emotional attitudes to this state of affairs reflects the fact that it is not clear (without further information) whether it is overall good or overall bad or is a moral tie. Further, if it is overall good, this maybe by only a slight degree; it is just barely more good than bad. This would also reflect our emotionally ambiguous attitude to a state of affairs of this sort. But note that it would have been better if a different sort of universe U1 had existed instead of the universe U that in fact exists, a universe U1 in which all animals were vegetarians, in the precise sense I have argued elsewhere.[4] This counterfactual moral truth is a part of “the probabilistic argument from natural evil” against the existence of God, for no all-good and all-powerful being would have created animals that had to cruelly murder and devour one another just in order to survive.[5] But given that (a) our universe U with its many kinds of carnivores does exist, and that (b) it is better that these carnivores exist than all die Out, and that (c) many animals will die out from scarcity of food if some of their kind are not killed, it is (just barely) overall good in some cases that some animals be eaten by others. Another
possible objection to the claim that a naturalist and global perfection ism
satisfies the attitude condition concerns our attitudes to death. We typically
have negative emotional attitudes to cases of premature death. We regard
premature deaths as intrinsically bad. It may be objected to global, naturalist
perfection ism that it is theoretically ad hoc to characterize a thing’s
premature death as bad in virtue of the fact that it is a prevention of the development of the thing’s nature. This is not ad hoc because it follows logically from a naturalist global perfectionism and characterizes exactly why premature death is intrinsically bad. Pre maturely dying or ceasing to exist is the prevention of a thing developing its nature as a physical thing and higher sort of thing. Being a physical thing entails being a spatiotemporally extended thing. Qua spatiotemporally extended thing, I develop my nature by becoming more spatiotemporally extended, which entails becoming more temporally extended (lasting longer). Premature death prevents this temporal development, and thereby prevents any further development of all the higher parts of my nature. This explains why premature death is bad. But this may suggest a further counterexample to the claim that naturalistic perfectionism conforms to the positive and negative attitudes relevant to our predications of “good” and “bad.” We have a negative attitude to fatness that is justifiable insofar as being fat is unhealthy. Naturalist and global perfectionism, however, implies that developing my spatiotemporal extension is good. Because I can in crease my spatial extension by becoming fat, should I become fat? Global, naturalist perfectionism implies that increase in spatial extension is intrinsically good; but for humans, this increase—if it involves becoming fat—is typically accompanied by a decrease in the degree of development of a person’s nature as a living thing. Becoming unhealthy is a decrease in the effectiveness of my organic functioning. Because my nature as a living thing has higher value than my nature as a physical thing, increases in spatial extension that involve becoming fat are typically part of a state of affairs that contains more bad than good and thereby is a bad state of affairs. Thus, it is false that global perfectionism implies that humans should become as fat as possible. This point, along with the point about increasing my temporal extension, is evidence against Hurka’s claim that a plausible perfectionist ethics cannot include humans’ physical nature in the definition of human nature. One moral use of “bad” is to refer to the property being more of a prevention of a development of natures than a development. A related use of “bad” is to refer to being a decrease in degree of a degreed essential property that belongs to a thing‘s nature. Becoming unhealthy is bad not merely in that it prevents an increase in the effective functioning of an organic degreed property, but is a decrease in this effective functioning. “Bad” and “good” also have additional uses in which they are predicated of things to which we have negative or positive attitudes (as distinct from states of things). How can a naturalist global perfectionism explain our positive attitudes to Jane when we assent to the sentence “Jane is good” or “Jane is a good person”? Our perfectionism implies that the first sentence expresses the singular proposition Jane highly develops human nature (her own and others) (or Jane has a highly developed nature and highly develops the nature of other things). Inductive evi dence suggests that, as a matter of fact, it is to such “developed people” as Jane that we have attitudes of admiration or respect or approval. But what of the increases in development of the nature of insects, weeds, and the like, to which we typically do not have positive attitudes? Does this show that perfectionism fails to satisfy one of the reference-fixing descriptions for “good”? The reason we do not have positive attitudes toward developing weeds and insects is that we normally regard these developments only as parts of more cornplex states of affairs to which we have a negative attitude, such as states of affairs that frustrate our desires (our nature qua animals) to inhabit a suitably clean, healthy house, walk through a forest without being stung or poisoned, or cultivate an aesthetically satisfying garden. (It may be the case, however, that humans’ disenchantment with weeds is not based on any good attitude to the plant kingdom, in which case moral progress would involve changing our value judgments about weeds and manicured gardens and other unnecessary attempts of humans to dominate or control the plant world.) But what about love and friendship, to which we have positive attitudes? How are these developments of our nature? They are developments of our animal natures; humans qua animals have the degreed essential property of being able to move so as to satisfy their desires regarding sensed or imagined things. One of our desires is to relate to other people in a friendly or loving way and to have them relate to us in a friendly or loving way. But are not love relationships of intrinsic positive value apart from the fact that they satisfy our desires to love and be loved? It seems tome this is not the case, for if a species of animals had an essential aversion to love relations and lived fully satisfied lives without them, then love relations among members of this species would have negative value. The primary reason love is good is that it satisfies our desires to love and be loved. I said that “humans qua animals have the degreed essential property of being able to move so as to satisfy their desires.” Is increasing the degree of this essence increasing the ability to move in a satisfying way or is its increasing the quantity or quality of the satisfying movements? In general, are developments increases in the capacity to do something or are they increases in the occurrent realizations of a fixed capacity? Essential developments are both, but most cases of development are increases in occurrent realizations of a fixed capacity. A serial killer has the desire to murder people; this desire (as is any desire) is part of the killer’s essence qua animal. Does it follow from perfectionism that it is good that he moves so as to satisfy this desire? By no we should know how such objections to perfectionism are answered: We consider the overall state of affairs, the killer murdering Jane, Bob, Beth, and Richard, and recognize that this state of affairs is bad. One of its parts is good, the killer moving so as to satisfy his desires, but we are unable to imagine this part without imagining the horribly bad other parts of this state of affairs, the murders of the victims, so this entire state of affairs appears to us to be repugnant. (R3) The Justification Condition Robert Adams suggested that understanding the nature of wrongness would give us one more reason, rather than one less reason, to not want to perform the wrong action. This condition is clearly satisfied by perfectionism, for if we understood that wrongness was decreasing or preventing the development of things’ natures, that would give us one more reason, not one less reason, to not want to engage in actions we regard as wrong. (The same holds, conversely, for right or good states.) This is not, however, what I have in mind as reference—fixing description (R3) the justification condition. This condition is that “good” is used to refer to a property F such that a possible state of affairs’ exemplification of F is typically regarded as a sufficient justification for some agent to actualize that state of affairs. Note that the justification condition does not imply internalism about reasons (in Brink’s sense). Description R3 does not say that “F is good” analytically entails “There is sufficient justification for an agent to actualize F” The reference—fixing concept of “good” instead implies that expressions of the form “F is good,” if believed to be true by humans, are typically regarded by humans as stating sufficient justification for an agent to actualize the property F This is a form of externalism about reasons. Global perfectionism satisfies this condition because if a possible state of affairs is believed de re or de dicto to have the property being a development of things’ natures (or, more fully, being more of a development than a hindrance of things’ natures), then actualizing that state of affairs is typically regarded as a sufficiently justified action. The belief de re that something has this property is crucial to this case, given the distinction we are using between cognitive significance and semantic content. The cognitive significance of “The F-ness of x is good” may be a belief de dicto involving the reference-fixing descriptions governing “good,” and the belief that the F-ness of x is a development of x’s nature may be a belief de re. This may be explained as follows. Belief de dicto does not allow intersubstitutivity in all epistemic contexts; the sentence, “David believes it is sufficiently justified to become a chemist because the advancement of chemistry is good” does not in general allow a substitution of “a development of things’ natures” for “good,” given that David may not know that the semantic content of “good” is a development of things’ natures. If” x believes de dicto that S” is true, x must recognize that the proposition he believes is expressible by the sentence S. But belief de re allows intersubstitutivity in all epistemic contexts because “goodness” and “developing things’ natures” have the same semantic content, even though David does not know this. David is epistemically related (de re) to the proposition, It is sufficiently justified to become a chemist because the advancement of chemistry is a development of things’ natures. If “x believes de re that S” is true, it is sufficient that x believe the proposition expressed by the sentence S, even though x may not realize that the sentence S expresses this proposition. (Analogously, David believes de re that H2O is H2O when he believes that “water is water,” even though he may not recognize that “water is water” expresses this proposition.) The justification condition is met if the possible states of affairs typically believed to be sufficient justifications for actions are believed de re to be developments of things’ natures, even though the associated de dicto beliefs are merely that these actions are generally approved of, are desirable, are what ought to be realized, and so forth. In many cases, however, the justification condition is met by a de dicto belief that a possible state of affairs is a development of things’ natures. We may have in mind as sufficient justifications for acting that some state of affairs consists in a forest or ecosystem developing its nature by becoming more healthy, a horse developing its animal nature by having room to run and exercise, a human having enough material goods so he can devote his time to developing his nature as a theoretically rational person, and so on. (It is worth emphasizing that “typically” does not imply “always,” for it is built into the rules of use of moral words that there can be some moral failures, for example, due to sociopathology or other defects of human moral nature. Thus, my frequent use of “typically” in discussing these rules of use.) Are there counterexamples to the claim that global perfectionism satisfies the justification condition? It is the nature of the AIDS virus to grow and multiply. Is enabling this virus to develop its nature a justified action? We already know the global perfectionist answer; a virus has less intrinsic value than humans, and thus a greater amount of goodness is realized by preventing the virus from realizing its nature and thereby allowing the relevant humans to realize their nature. The virus growing, considered by itself, is good, but the more complex state of affairs of which it is a part is bad. Another possible counterexample is this: There is an air mattress lying on the floor, and it is not blown up. Becoming more spatially extended is a development of its nature as a physical thing, so extending the mattress by blowing it up is actualizing a good state of affairs. It may be alleged that global perfectionism implies the absurd conclusion that increasing the mattress’s spatial extension is a morally sufficient justification for my engaging in the action of blowing it up, and therefore that I ought to go around blowing up any deflated mattresses I see. Global perfectionism accounts for this putative counterexample by noting that the more complex state of affairs, my blowing up the mattress in order that the mattress may increase its spatial extension, is bad. It is bad because I should be spending my time bringing about nature developments that are significantly more valuable than developing the mattress’s physical nature. (R4) The Consensus Condition Part of the reference-fixing descriptions of our moral words is that there be large- scale agreement about which states and things have the properties designated by “good,” “bad,” and so forth. This does not mean there is a certain body of items all agree upon, but that there are significant overlapping agreements among most people. The primary evidence about the moral beliefs people hold is their behavior, not what they report verbally when they are asked about their moral beliefs. Frequently, the behavior and verbal reports of people differ. For example, biographical materials about Nietzsche show that he actually lived by and intuitively believed the moral views he condemned in his books; that is, he was a compassionate, sensitive, and gentle man who often took pity on people. The appeal to behavior has added importance when we recognize the distinction between the cognitive significance and semantic content of people’s utterances, and the fact that belief in global, naturalistic perfectionism is manifested more often in belief de re (exhibited in behavior) than in belief de dicto (exhibited in verbal reports “about what I believe”). Our naturalist perfectionism satisfies the reference-fixing description (R4) for there is widespread agreement (both de re and de dicto) that developments of things’ natures are good and preventions of this development bad. Such nature developments as the acquisition of knowledge, the satisfactions of humans’ and other animals’ desires, the various practical accomplishments that are developments of practical reason, becoming healthy, the flourishing of wildlife, the endurance of the mountains, and so on are regarded as good and their opposites as bad. Notice that I
am not including within the reference—fixing description for “good” a list
of kinds of goods—knowledge, friendship, health—for this would make our
normative ethics an intuitionist ethics rather than a perfectionism. If I did
include this list, I would begin with an antecedent intuitive knowledge of what
is good, a knowledge embodied in the cognitive significance of our moral words,
and our ethical theory would simply be an articulation of these moral
intuitions. Our normative ethical conclusions, in this case, would not
be derived from premises about the development of things’ natures, and
statements about “developing something’s nature” would be as redundant in
our normative ethics as they are in Hurka’s. For this reason, I must reject
Robert Adams’s inclusions of examples of kinds of wrong actions in the
reference-fixing concept associated with “wrong” (see section 22); Adams’s
normative ethics presupposes an antecedent knowledge of what is wrong and is not
derived from premises only about contrariety to divine commands. In this respect, To clarify further this crucial aspect of the linguistic essentialist method, I should emphasize that the examples of good and bad items, or kinds of goods and evils, that I just mentioned and that I mentioned in discussing reference-fixing conditions (R1) through (R4) are not analytic or a priori parts of the reference-fixing conditions. Rather, they are offered as evidence that global, naturalist perfectionism satisfies these conditions. Consider the proposition, P. The state of affairs Dinosaurs developed their natures regardless of whether humans believed they did satisfies the moral realist condition (R1). The mentioned state of affairs is known a posteriori and the concept of this state of affairs includes a concept of a kind of good (developing an animal nature), but neither of these concepts is contained analytically or a priori in any of the reference-fixing descriptions (R1)-(R6) for our moral words. This kind of good is deduced from the axioms of global, naturalist perfectionism (some of these axioms being that all things [concrete objects] have natures in the sense of nontrivial essences that constitute their natural kind, that goodness is identical with developing these nontrivial essences, and so forth). An example of this kind of good, dinosaurs developing their animal nature, is mentioned in proposition (P) for the purpose of providing evidence that global perfectionism meets the reference-fixing conditions (R1) for the word “good. The next two reference-fixing descriptions pertain to uses of “right” and “wrong.” The rules I have already listed, (R1) through (R4) are not very controversial (except for the realism condition, and I have presented several arguments on its behalf in chapters 1-6), but rules (R5) and (R6) are more controversial in that they imply that both consequentialist and deontological moral properties are instantiated. Consequentialists might want to deny that any deontological properties are instantiated, and some deontologists might want to deny that any consequentialist properties are instantiated. But it seems to me that the inductive evidence about our moral talk and beliefs evidence suggests that both sorts of properties are instantiated, and this is the major source of moral dilemmas. (R5) The Consequentialist Condition The rule of use of “right” (in its consequentialist sense) is that this word is predicated of voluntary actions that maximize the good. The predicate “right” expresses a relational property that is formally characterized as maximizing the good or, more exactly, as having the highest expected value, the expected value is a function of the act’s possible outcomes and of their probabilities of being realized. (As I understand the consequentialist condition, the value of the act itself is one of the components—and sometimes the sole component—included in the aggregate amount of value realized by the act.) Note that the rule of use of “right” includes the concept of the good, so that this rule of use is conceptually dependent on the concepts in rules (R1)—(R4) that pertain to the word “good.” The expression “is wrong” (in its consequentialist use) is predicated of a voluntary action that is not maximizing, and this predication is normally associated with attitudes of condemnations or other negative attitudes. There is evidence that global, naturalistic perfectionism satisfies condition (R For example, people admire or approve of Einstein’s theoretical efforts, Beethoven’s musical efforts, Sylvia Plath’s literary efforts, and these efforts maximize (or at least are believed to maximize) the development of human nature, which includes the creator’s nature as well as the audiences’ natures. Many other examples could be adduced: doctors are respected for maximizing health, some psychologists are respected for maximizing happiness (the satisfaction of desires), and firefighters and rescue workers are admired for maximizing the temporal extension of people qua physical beings (that is, “saving their lives,” as these right actions are normally called). (R6) The Deontological Condition The rule of use of the noun phrase “a right” (for example, “a right to life”) is different from the rule of use of the verb phrase “is right.” “A right” has a deontological sense and, unlike “is right,” does not mean maximizes the good. The noun phrase “a right” refers to a state of a thing, such that voluntary destructions or impairments of this state are typically objects of disapproval, outrage, condemnation, and other negative attitudes. (“State of a thing” is meant in the broadest possible sense to include voluntary or involuntary actions, emotions, dispositions, static physical conditions, movements, and so on; that is, it refers to any property or relation of a thing.) According to global, naturalistic perfectionism, the phrase “a right” directly and rigidly designates the property being a necessary condition of developing something’s nature. For example, life, liberty, free speech, freedom of inquiry, freedom from physical harm by others, control over one’s body, and so forth, are all conditions of developing human nature. All rights are defeasible; your right to life is defeated if the only way to stop your random spree of murdering humans is to kill you. Global perfectionism implies that nonhuman animals as well as humans have rights. But they do not have equal rights; for example, “Nonhuman animals have rights” does not imply “It is always wrong to use nonhuman animals in medical experiments, such that their health or life ought never be sacrificed for the sake of advancing the health of humans.” Humans’ rights to life and health defeat the rights of animals with less developed kinds of natures. In general, the rights of higher kinds of things defeat the rights of lower kinds of things. It would be wrong to use animals in medical experiments if such use were not necessary to bring about vital advances in medical knowledge relevant to humans’ health. The right of an individual x (of a certain natural kind K) to do F defeats the right of another individual y (of the same natural kind K) to do G if (all else being equal) x’s doing F is a realization of a higher kind of natural essence than is y’s doing G. Your right to develop your theoretical reason defeats my right to satisfy my desire to amuse myself with material luxuries. But in cases in which y’s being prevented from doing the action 0 of the lower kind would deprive y of abilities to do actions of a higher kind as well, y’s right to do G is not defeated. Your right to pursue knowledge does not defeat my right to life because my being alive, although a lower kind of natural essence than theoretical reason, is a necessary condition of me developing all my higher kinds of natural essences. The finite difference principle, however, shows there are many exceptions to these general hierarchy-based principles. For example, foxes’ rights to life are not defeated by humans’ desires to wear furs as luxury items. The right to life of a less developed kind of thing (the fox) morally outweighs a certain trivial and superfluous desire of a more developed kind of thing (the human’s desire to wear furs). If it were necessary for humans to wear furs as clothing to survive, the foxes’ right to life would be defeated, but this is not necessary, and so it is wrong to kill foxes for their furs. In some cases, rights and consequentiahst considerations about maximizing the good clash and produce moral dilemmas. One of the most important moral dilemmas concerns vegetarianism. Because humans can survive without eating animals and eating meat is not necessary to human health or happiness, it might be brought that global perfectionism implies that it is morally obligatory to be vegetarian. But in fact, global perfectionism implies a moral dilemma about this issue. If all humans became vegetarians, millions of animals that would otherwise be born and live to maturity would be deprived of the chance for life, health, and happiness. These animals’ living a healthy and happy life until their maturity, when they were killed painlessly, would be a more valuable state of affairs than their not living at all. Thus, reasoning on consequentialist grounds, the raising of happy animals to be eaten by humans counts as a right action—it maximizes the good. These animals, once alive, however, have a right to life that is not defeated by a human’s morally insignificant desire to enjoy the taste of meat. This deontological consideration clashes with the consequentialist consideration and produces a genuine moral dilemma (that is, there is no moral fact of the matter as to whether we ought to raise and eat happy animals or not bring them into existence and eat only plants). Global perfectionism does imply that the unhealthy and painful conditions in which cows, pigs, chickens, and so on are typically raised are bad and that farmers are doing something wrong in raising the animals in these conditions. Global, naturalist perfectionism also implies that plants and physical things have natural rights. A tree has a right not to be cut down, even though this right can be defeated if this is necessary as a consequence of the rights of a more developed kind of thing. Humans have a right to pursue the development of their theoretical reason, and if this requires that trees be used for books, then the tree’s right to life is defeated. A mountain has a right to its natural integrity; it is wrong to dig a tunnel through a mountain unless the tunnel is required by more significant rights of a more developed kind of thing. It may be objected to naturalist, global perfectionism that plants and physical objects are not typically regarded as having rights, and therefore that this normative ethics does not satisfy the reference—fixing description of “a right.” I would respond that the reference-fixing description of the word “a right” is merely formal and does not include a list of the kind of things that have rights, so global perfectionism is not inconsistent with a rule of use that limits rights to humans or to humans and animals. It may nonetheless be objected by some ethicists that the reference—fixing description of “a right” does include limitations on the application of this word, and that its application is limited to humans or to animals or at least to living things. I believe the inductive evidence about the use of “right” suggests that there is no such limitation built into its reference-fixing description. First, note that the rights of plants and physical objects are habitually defeated by humans’ rights to life, health, happiness, knowledge, and so forth, and for this reason the rights of plants and physical objects are typically not a topic of moral deliberations. There is no cause to stop and deliberate about a piece of fruit’s right to life because this right is obviously defeated by humans’ right to life. Second, on the relevant occasions, the rights of physical things and plants are recognized. Humans’ predominate way of relating to plants and physical things is to use them. This prevents humans from contemplating plants and physical things as something other than mere materials for human usage, and this contemplation is required to recognize they have rights. On such occasions of contemplation, however, their rights typically are recognized. For example, when humans are sitting on top of a mountain contemplating the beauty of nearby mountains and forests, they are in a position to recognize the intrinsic value of these things, and on these occasions they do tend to recognize the rights of the mountains and forests to their natural integrity or life. For example, the thought of cutting down a forest or strip mining one of the mountains to produce unnecessary luxury items would typically be found to be morally repugnant on these occasions. The fact that humans are typically disturbed by such thoughts on these occasions is best explained by the fact that they recognize that the mountains and forests have rights to be left alone and not to be destroyed needlessly by humans. Some consequentialists and deontologists may object to the claim that the Reference-fixing descriptions of our moral words specify both consequentialist and deontological uses of these words. But this objection seems inconsistent with the inductive evidence. Some evidence is that the predicates of voluntary actions “is wrong” and “is right” have deontological uses as well as the consequentialist use mentioned in (R4) In the deontological use of ”is wrong,” an action is wrong if it violates something’s rights and is right if it respects things’ rights. Consequentialists would object that human moral life does not involve a deontological use of “wrong action” in addition to the consequentialist uses of “wrong action.” But it seems the consequentialists are wrong; consider, for example, the statement, “It is wrong to steal.” What is the cognitive significance and rule of use of “wrong” in this sentence? “Wrong” does not mean here that stealing fails to maximize the good in the relevant situation. Rather, it means that stealing is an action that violates a person’s right, namely, their right to private property. The same holds for “It is wrong to rape,” which does not mean that raping somebody does not bring about the best consequence in the relevant situation. Evidence for the claim that it is wrong to rape Jane is not that this brought about the second-best consequence in the situation, rather than the first-best. Neither does it mean that rape is a kind of action that tends not to bring about the best consequences. Rather, “It is wrong to rape” means that the act of rape has the property of violating the person’s right to control her body. But it also seems evident that this is not the only use of “is wrong.” Deontologists who claim that our moral life does not involve a consequentialist sense of “is wrong” seem to be mistaken. Consider the statement, “Rimbaud was wrong to stop writing poetry and devote the remainder of his life to making as much money as possible.” Here “wrong” has the consequentialist sense of an action that fails to maximize the good (to maximize the development of Rimbaud’s and other humans’ natures). Here Rimbaud is not violating anybody’s rights. Further, “It is wrong to lie around lazily all day” does not mean that some right is being violated; it means that this activity manifestly fails to bring about the best consequence open to the agent. I believe the explanation of most moral dilemmas is that one and the same action simultaneously instantiates “is right” in a consequentialist sense and “is wrong” in the deontological sense (or vice versa). This fact explains most of the data dis cussed in the art in Moral Dilemmas, edited by Christopher Gowans. For instance, Marcus mentions in “Moral Dilemmas and Consistency” the example, originally given by Plato, of a person who has promised to give a cache of arms back to a person who is intent on mayhem. The decision to not give the arms back “is right” in the consequentialist sense (it maximizes the good or minimizes the bad) but “is wrong” in the deontological sense (because it violates the relevant person’s right to have promises made to him abided by). This is a dilemma in that there is no common criterion to resolve the issue—consequentialist and deontological criteria of rightness and wrongness are incommensurable. (This does not mean that we cannot know the true answer to the dilemma, but that there is no true answer, and knowing this is knowing all there is to know about it.) Concluding Remarks about Rules R1-R6 Part A of this chapter is about metaethics, part B about normative ethics. The division corresponds roughly to the distinction between the issue of whether or not human life has an ethical meaning (part A) and the issue of what this meaning is (part B). It is pertinent at this juncture to make some concluding remarks about the role the reference-fixing rules (R1)-(R6) play in the arguments constitutive of normative ethics. The argument I presented has the following structure. We know by empirical induction that (R1)-(R6) are the reference-fixing descriptions for “good” and other moral words; the descriptions (R1)-(R6) require that it be an empirical discovery as to what properties satisfy these descriptions and are the referents of moral words. I further argued that we know by empirical induction that properties involving developments of things’ natures satisfy the reference-fixing descriptions. It might be objected that this “method” of doing normative ethics is misguided from the very beginning. How can mere facts about the senses of words determine what is good and bad? How can the sense of a word make it right or wrong for me to do something? If it turned out that, upon investigation of natural language, that “good” meant random killing, would it follow that it is good to engage in random killing? Consider this argument: 1. The referent of the moral use
of “good” in natural languages is a
development of natures. Therefore, 2. Moral goodness is a development of natures. What shall we make of this inference? On the one hand, it seems valid if “good” is used in the conclusion in the same way it is used in natural languages. But it seems invalid if we allow that the conclusion is a sentence in a metalanguage and that all competent speakers of natural languages may be systematically mistaken in their (de re and de dicto) beliefs about goodness. I think we should allow these latter conditions and thus that the inference of (2) from (1) is invalid. If we add to (1), however, the premise consisting of a relevant version of “the argument from veridical seeming” presented in section 23.3, then we may validly infer the conclusion 3. We are justified in believing that moral goodness is a development of natures. It might be objected that even if a pertinent version of “the argument from veridical seeming” is sound, and even if I have presented some evidence that (a) is true, I have not shown that other types of normative ethics, for example, a Kantian type ethics, egoism, or utilitarianism, or a social contract ethics, or a pluralism of Pritchard’s sort are false. Specifically, the objector may continue, I have not shown that other normative ethics do not satisfy the reference-fixing descriptions (R1)-(R6) or that there is a better fit between perfectionism and these descriptions than other types of normative ethics. I agree that I have not demonstrated this, but I believe I have put the ball in the court of the defenders of these other sorts of normative ethics. I have argued that global, naturalist perfectionism satisfies these rules, and the burden of proof is on the defenders of other sorts of normative ethics to show that their theory equally or better satisfies these rules than perfectionism. But I would briefly note that many of these normative ethics seem ruled out on the face of it. For example, social contractarianism appears to fail to satisfy the realist condition (R1) And utilitarianism (at least in its traditional forms, in which happiness is the good) seems to fail to satisfy the attitude condition because its range of goods seems much too limited to accommodate all that we approve as morally good; egoism fails for the same reason. Further, utilitarianism appears inconsistent with the deontological condition. Kantian ethics fails to satisfy the consequentialist condition. If some nonperfectionist ethics is to satisfy all these rules, it would at least have to be in a form modified from its standard way of being formulated. It seems to me doubtful, however, that any nonperfectionist ethics satisfies these rules as well as global, naturalist perfection ism, and that there is justification to believe that this perfectionism is the most adequate type of normative ethics. 26. The Supreme Ethical Meaning of Human
Life According to global, naturalist perfectionism, the objective ethical meaning of human life is to develop things’ natures, the nature of humans as well as that of other objects. (Here “human life” is used in its intuitive sense in the phrase “the meaning of human life”; I am not referring merely to human organic processes.) For any x, if x is a human, then the ethical meaning of x’s life is to develop the natures of things. The ethical meaning of human life cannot just be to develop the nature of humans, for humans can realize more goodness by striving to develop the natures of humans, nonhuman animals, plants, and mere physical things, than just humans alone. Furthermore, the ethical meaning of my life cannot just be to develop my nature, for in some cases I can bring about more goodness by striving to develop the nature of other things at the expense of my personal development. The extreme case would be to sacrifice my life for the sake of other people, animals, and so forth if this would bring about significantly greater goodness than the further development of my own nature. The questions, “What should I do with my life?” and “How should I live?” are answered in a general way by saying, “You should choose the way of life that would most contribute to developing the nature of all things.” A person who
lives an exceptionally meaningful life is a person who contributes more to the
development of things’ natures than is normal, whether this is done by the
person developing her own nature to an especially high degree or by contributing
an exceptional amount to the development of other things’ nature (typically,
one will involve the other, for example, as in the case of Albert Einstein or
Charles Darwin). This fits in with the empirical data about what people
typically believe to be meaningful lives. We typically regard such people as
Plato, Galileo, Susan B. Anthony, Martin Luther King, Of special interest is the supreme ethical meaning of human life. By this, I mean the supreme distinctively human good. The nontrivial essence that is distinctively human, that is, that distinguishes humans from other animals, is reason. “Reason” is an ambiguous word and only in a certain sense of “reason” does it designate the distinctively human essence. “Theoretical reason” (as I use this phrase) refers to the ability to know the nomological reasons why something exists (or why things of a certain kind exist) or possesses certain properties; that is, it is the ability to construct deductive or inductive arguments whose premises include at least one law of nature and which explain the proposition that serves as the conclusion. The conclusion is a proposition asserting that something exists (or things of a certain kind exist) or possesses certain properties. In short, theoretical reason is the capacity to explain why things are and why things are the way they are. The maximal development of theoretical reason is to know the reason why everything exists and why everything possesses the properties that are in fact possessed. The maximal development is (at least) to know the reason for every fundamental state of affairs that has a reason; it may be that some states of affairs have no reason, such as the fact that the universe exists, and being able to pick out these states of affairs would be a part of the maximal development of theoretical reason. Although the standard naturalist view is that the supreme explanatory laws explain why there are some fundamental kinds of things (for example, atoms), rather than explain why the universe exists at all (rather than nothing), it is at least epistemically possible that humans can attain the latter sort of explanatory knowledge, which would be the greatest explanatory knowledge that is conceivably possible. For example, it is at least epistemically possible we can know a certain law of nature that explains why this universe exists. Perhaps such a law would be a complete “wave function of the universe,” which is a supreme law of nature that involves the unification of Einstein’s general theory of relativity with quantum mechanics. Stephen Hawking mentions such a complete law. He says that the “sum over all the possible histories of a universe” will give us a law that describes everything in the universe: “Each history in the sum over histories will describe not only the space-time but everything in it as well, including any complicated organ isms like human beings who can observe the history of the universe.” |