Smithian Intrinsic Value

Since G.E. Moore coined the phrase and developed the concept of intrinsic value at the beginning of the 20^th century (Moore 1902 and 1912),[1] the notion of intrinsic value has continued to influence ethical theory, from W. D. Ross’s moral realism (Ross 1930) to more recent revivals of similar theories.[2] Lately, intrinsic value has dominated discussions in applied ethics, especially environmental ethics, where it plays central but contested roles in establishing moral considerability for non-human entities. As one environmental philosopher put it, “How to discover intrinsic value in nature is the defining problem for environmental ethics” (Callicott 1999: 241).[3] And “intrinsic value” play roles in ordinary ethical debates about topics as diverse as abortion, modern art, and whether to go to college.[4] However, the notion of “intrinsic” value is often used with imprecision. Christine Korsgaard (1996: 250) rightly distinguishes between “intrinsic” value, which a thing “has . . . in itself” (vs. extrinsic value), and “final” value, being “valued for its own sake” (vs. instrumental value). These are different concepts, but “intrinsic value” is often used for both.

There are three main senses of “intrinsic” value: non-instrumental, trumping, and non-relational value. Something has non-instrumental value if pursued for its own sake, making it an “an end-in-itself” in Aristotle’s sense, rather than “a means to something else” (Nicomachean Ethics, 1097a). Trumping value is that by virtue of which something overrides other considerations; it is what Kant calls “dignity” that “is raised above all price and therefore admits of no equivalent” (Kant 4:434).[5] Finally, non-relational value is the Moorean value that something has “solely in virtue of its intrinsic properties,”[6] independent of relations to other things.

Cutting across these senses of intrinsic value is the general claim that intrinsic value implies moral considerability. For Aristotle, that something is an end-in-itself gives it a privileged place within ethics, although Aristotle allows that one often ought not promote certain ends-in-themselves when other ends are more pressing. An end-in-itself is worth pursuing for its own sake, all other things being equal. Trumping value is even more closely linked to moral considerability; such values “trump” precisely within moral deliberation. And non-relational value, at least for Moore, is the basic category of moral considerability. Something’s intrinsic value is a simple property of it that ought to be promoted in one’s actions; “all moral laws . . . are merely statements that certain actions will have . . . effects [with intrinsic value]” (Moore 1902: §89).

This paper develops a Smithian account of these types of intrinsic value. I start by showing how Smith’s Theory of Moral Sentiments (TMS) provides a normative framework for a rich “proper-attitude” account of value. I then show that Smith can account for all three sorts of intrinsic value. I end by suggesting an even richer Smithian approach to intrinsic value accounts for the wide variety of values implicit in proper attitudes.

1. Smithian Value

Smith’s approach to value is part of a family of proper-attitude theories of value, according to which things have value just in case they are objects of proper valuing attitudes. Proper-attitude theories are distinct from both subjectivism about value, according to which things have value just in case they are valued, and objectivism about value, where things’ values are objective properties independent of attitudes that anyone might take towards them. Like subjectivism, proper-attitude theories ascribe value to something because it is valued. This helps make transparent the connection between values and human interests and desires, something that can seem obscure on objectivist accounts. But like objectivism, proper-attitude theories are normative rather than merely psychological. Whereas subjectivism rules out error about value (since if one values something, then it has value), proper-attitude theories can accommodate this straightforwardly (one is wrong insofar as one’s valuing attitude is improper).

The general description of proper-attitude theories laid out above can be made specific by identifying which attitudes are normatively evaluable and how one evaluates those attitudes’ propriety.[7] Christine Korsgaard’s neoKantian “rationalism” about value, for example, takes rationality to be the basis of propriety and choice as the only (or primary) value-conferring attitude (Korsgaard 1996). Alternatively, Elizabeth Anderson includes a wide range of favorable attitudes within the range of morally evaluable attitudes and lays out a social approach for determining propriety (Anderson 1993:93f.). One could even transform this proper-attitude account into subjectivism by claiming that all actual attitudes are proper or into objectivism by claiming that a valuing attitude is proper only if responsive to objective value-properties. The rest of this section uses Smith’s TMS to specify a distinctively Smithian theory of value based on Smith’s account of the scope of normatively evaluable valuing attitudes and the nature of propriety.

a. The range of evaluable attitudes

For Smith, any possible human attitude is morally evaluable. Smith includes attitudes such as love, hate, gratitude, anger, hunger, esteem, and even “small vexations” (I.ii.5.3); passions like hunger that “take their origin from the body” (I.ii.1); and those like infatuation or one’s interest in “our own studies” (II.ii.2.6) that “take their origin from a particular turn or habit of the imagination” (II.ii.2). He distinguishes between “unsocial passions” like anger, hatred, or resentment (I.ii.3), “social passions” like love, benevolence, and esteem (I.ii.4), and “selfish passions” like grief, joy, uneasiness, and satisfaction, “when conceived upon account of our own private good or bad fortune” (I.ii.5.1, p. 40).

The broad range of Smith’s account is an important advantage over alternative proper-attitude theories. For instance, rational desire theories of value such as Korsgaard’s limit evaluable attitudes to those oriented towards “bringing about” a “state of affairs” (Korsgaard 1996:226). Many have criticized such views because “we find many things to be valuable besides the possible objects of rational desire” (Anderson 1993:129-30) and, more generally, “There is much that we judge normatively and regulate by norms other than action, for example, reasoning, . . . emotions, responses, feelings, . . . and attitudes . . . Respect . . . cherishing, and sympathy . . . are all ways of valuing something intrinsically which to not reduce to valuing the state of that thing’s existence” (Darwall 2003: 478, 482).[8] Smith’s moral theory rejects this emphasis on intentional action at its core by focusing not on agents but on “person[s] principally concerned” (I.i.3.1, p. 13). Such persons can be agents but are often those to whom something happens; attitudes judged proper or improper are often passive responses. Grief and sorrow (primarily passive) along with animosity and generosity (primarily active) are capable of moral evaluation (see I.i.3.3), and Smith mentions admiration for a poem or amusement at a joke as examples of valuing attitudes that can be proper or improper (I.i.3.3). Of course, attitudes often cause actions, which can then be evaluated, but because attitudes are Smith’s primary locus of evaluation, he includes passive responses within ethics. This extension not only seems ethically appropriate in that we can and do appraise people’s responsive attitudes, it also provides a richer normative basis for thinking about value because something can have value even if no one aims or even ought to aim to “bring it about.”[9]

b. Propriety

Propriety is Smith’s fundamental ethical concept, and an attitude is proper when an impartial spectator can sympathize with it. Smith begins by observing that human beings sympathize by attentively imagining themselves in another’s situation and responding emotionally to that imaginative change of place.[10] Judgments of propriety are based on mutual sympathy between spectator and “person principally concerned”:

When the original passions of the person principally concerned are in perfect concord with the sympathetic emotions of the spectator, they necessarily appear to this last just and proper . . .. To approve of the passions of another . . . is the same thing as to observe that we entirely sympathize with them. (I.i.3.1, p. 16)

One soon learns, however, that people sympathize to different degrees due to inattentiveness, partiality, or prejudice. One thus learns to judge propriety of sentiments – one’s own and others’ – by appealing not to particular spectators’ sympathy but to sympathetic responses that would be present in attentive, well-informed, and impartial spectators (III.2.36, p. 129).[11] Attitudes are judged proper if impartial spectators attentively imagining themselves in the place of the person principally concerned would have those attitudes.

It is important here to clarify two points. First, “impartial” does not mean purely rational or distant from concrete particulars of life.[12] Smith’s impartial spectator is not an “ideal observer” who is “dispassionate” “in the sense that he is incapable of experiencing . . . such emotions as jealousy, self-love, . . . and others which are directed towards particular individuals as such” (Firth 1952: 55). The impartial spectator is sympathetic, imaginatively entering into particular situations and responding emotionally, and thus must be “rich in feeling” (Nussbaum 1990: 338). The spectator’s impartiality requires not apathy but only that the emotional response be entirely sympathetic, not tainted by one’s own purely personal interests. Such impartiality makes possible the “concord” of sentiments that determines propriety.

Second, Smith’s impartial spectator does not have transcendent access to a realm of moral values on the basis of which to make ethical evaluations. The mechanism for evaluating the propriety of attitudes is the impartial spectator, who appeals only to her own sentiments. When one seeks to determine the propriety of an attitude, one imagines oneself as impartially, attentively, and vividly as possible in the place of the person principally concerned, and one reacts to that situation. One must be attentive to relevant features of the situation, lest sympathy or its lack be due to ignorance rather than impropriety. And one must abstract from or eliminate partial considerations that affect one’s emotional response to the imaginative change of place. But features of situations do not carry normativity with them; normativity comes through affective responses of spectators free from personal (partial) interests.[13]

Smith builds several nuances into this account. He points out the important role of moral evaluation of “selfish” sentiments, in contrast to his contemporary Hutcheson. He discusses custom’s role in corrupting sentiments, and he outlines the importance of general rules to alleviate the effects of unseen prejudice in one’s judgments of propriety. He derives notions of “merit and demerit” – and thereby justice – from propriety, and he discusses the origin of “duty.” But the essence of his account is the sympathetic response of impartial and attentive spectators, and this is sufficient to develop a Smithian theory of value.

For Smith, something has value when it is the object of a valuing attitude with which impartial spectators can sympathize. This Smithian account of value avoids the pitfalls of both objectivist and subjectivist accounts of value. Because Smith does not ground the propriety of valuing attitudes on pre-existing “objective” values, his account is free from metaphysical and meta-ethical worries (e.g., about queerness, cf. Mackie 1977) that come with adherence to moral realism. But the impartial spectator also frees this account from subjectivism, since someone valuing something is insufficient for ascribing it value. The impartial spectator incorporates normative constraints that allow Smith to insist that certain valuing attitudes are called for, even if no actual person feels them, and other valuings are wrong, even if widely felt. Smith’s critiques of slavery in his time (see V.2.9, p. 206-7; WN I.viii.41, III.ii.9, IV.ix.47) and his extended discussion of the approval of infanticide by “almost all the states of Greece, even among the polite and civilized Athenians” (V.2.13-6, pp. 209-11; WN intro.4, I.viii.24) provide excellent examples of anti-subjectivism.

The details of Smith’s account of propriety highlight further ways in which Smith’s theory is distinct from alternatives, such as those of Korsgaard and Anderson. On Korsgaard’s account of value, reflection on what can be a rational object of desire leads to the conclusion that (only) the capacity to value is unconditionally valuable and other goods have value by virtue of being objects of a properly constrained valuing will.[14] For this Kantian, universalist approach, one can determine a priori what has intrinsic or unconditional value. By contrast, Smith’s approach is particularist, sentimental, and social. While Smith allows for the development of general rules, all value-ascription is ultimately context-specific. It might turn out that human beings are always ascribed value, but one cannot know anything this through a priori reflection on the mere process of valuing. To discern whether something has value, one must actually imagine oneself in the place of the person principally concerned, responding to all of the concrete particulars of that person’s situation. Moreover, one responds to these particulars not through reason alone, but with one’s whole self. Value is not a matter of thinking through what ought to have value but a matter of feeling value when one is made fully aware of the relevant features and freed from one’s partiality. Finally, because Smith is acutely aware of the tendency to partiality and because there is no a priori, rationalist method for deducing value, Smith’s ethics requires intensive engagement with others in concrete dialogue and interaction in order to discern one’s hidden partiality and better imagine the situation of persons principally concerned.

In these respects, Elizabeth Anderson’s approach is more Smithian. Anderson emphasizes the importance of “non-propositional” attitudes (129), attention to particular details, and the role of social life in determining what has value. But Anderson’s account is importantly different from Smith’s in two respects. First, Anderson builds into her account a substantive moral realism that Smith can forego. She argues against any naturalism that would “try to substitute for the question: do these facts merit this attitude? the question: do these facts cause this attitude? . . . [N]o matter how the facts are presented to a person however naturalistically constituted, she always has room to ask whether her resulting attitudes are rational or merited or endorsable” (139). For Smith, by contrast, what it means to ask whether someone’s attitudes (including one’s own) are endorsable or merited is simply to ask whether impartially imagining oneself in the place of that person actually causes one to feel those attitudes. Contra Anderson, for Smith there is no room for well-informed, impartial spectators to ask whether their sympathetic attitudes are rational. (Of course, one can always ask whether one is well-informed and impartial.) And Smith therefore has no need to appeal to realist “norms for feeling” (Anderson 139) that would be prior to feelings themselves.

Paradoxically, Anderson’s “realism” about value leads her to an account of value that is pluralist and even relativist in ways that Smith’s naturalism lets him avoid. Without a well-defined stance from which to question attitudes’ propriety, Anderson’s approach is social to the point that the “objective valid[ity]” of a value is established by a “process of justification” in which “participants can reach significant agreement or progress on the matters under discussion” (Anderson 1993: 93). Whereas Smith posits the hypothetical stance of impartial and well-informed spectators, Anderson has only concrete discussions in the context of various norms of interaction (93f.). Thus Smith ascribes “objectivity” to values only when all impartial spectators “entirely sympathize with them” (I.i.3.1, p. 16), while Anderson allows that even when conversations about divergent values give rise to “the response [of] sticking to one’s guns,” there can still be an improved understanding of one’s values that gives “a legitimate claim to objectivity” (Anderson 1993: 97), such that opposed sets of values could both emerge out of rational dialogue having an “objective” status. Smith’s impartial spectator thus provides both a naturalist alternative to Anderson’s realism and a universalist alternative to Anderson’s moderate relativism about objectivity.[15]

2. From Value to Intrinsic Value

Even if Smith provides an account of value that is neither crudely subjectivist nor queerly objectivist, this conception might not make sense of intrinsic value. This section shows how Smith can account for the three sorts of intrinsic value of greatest relevance to contemporary value-theory.

a. Non-instrumental Value

Smith’s approach to non-instrumental value is straightforward and provides a general framework for any proper-attitude approach to intrinsic value: things have non-instrumental (intrinsic) value just in case they are objects of proper, non-instrumentally (intrinsically) valuing attitudes. Any proper attitude account can be extended to cover any form of intrinsic value if, on that account, there is a proper attitude that ascribes that sort of intrinsic value to its objects. In this context, non-instrumental value is particularly easy for proper-attitude theories to accommodate. As Aristotle argued with respect to desire in his Nicomachean Ethics, instrumental goods are desired as means to something, and proper instrumental desires must be oriented towards the pursuit of something desired non-instrumentally. Smith adds that even goods that are valuable for further ends are often properly valued for their own sakes as well:

With regard to all those ends [e.g. social-, self-, or species-preservation] which, upon account of their peculiar importance, may be regarded . . . as favorite ends of nature, she has . . . not only endowed mankind with an appetite for the end which she proposes, but likewise with an appetite for the means [e.g. punishment, food, or sex] by which this end can be brought about, for their own sakes, and independent of their tendency to produce it. (II.i.5.9n, p. 77, emphasis added)

The appetite for various goods “for their own sakes” includes goods essential for thriving societies and individuals, but also other goods (Smith is not being exhaustive here).

Non-instrumentally-valuing attitudes are not limited to desire. Martha Nussbaum provides an excellent example of non-instrumental valuing in describing grief over her mother’s death (Nussbaum 2001:19-88).

My mother has died. It strikes me . . . that a person of enormous value . . . is there no longer. It seemed to me as if a nail from the world had entered my insides; it also felt as if life had suddenly a large rip or tear in it, a gaping hole. (Nussbaum 2001:39).

Nussbaum’s mother strikes her as having “intrinsic worth or value” that is inconsistent with seeing her mother “simply as [a] tool… of [her] own satisfaction” (Nussbaum 2001:31).

Of course, other emotional responses – missing a thing, or being upset at its loss – ascribe objects only instrumental value. When my bicycle was stolen, I missed it (and felt some resentment towards its thief), but I did not grieve over it. Were I to have felt anything resembling grief, this would have indicated that the bike had what we generally call “sentimental value,” which is a sort of non-instrumental value.

Grief is not unique in taking its objects to have non-instrumental value. One generally feels grief for what one loves, and love too values non-instrumentally. Likewise awe, delight, and cherishing generally value non-instrumentally. Even when felt towards something that has instrumental value – say, the Hoover Dam – awe transcends and abstracts from instrumental value. Negative attitudes, too, can be non-instrumental. Hatred and resentment do not generally view their objects instrumentally. One who seeks revenge or to harm an object of hatred does not generally do so for the sake of further ends; resentment and hatred precisely consist in seeking another’s harm for its own sake. Of course, one can take instrumentally valuing attitudes towards objects. One might desire wealth for the sake of pleasure or feel sadness over the loss of something because it was particularly useful. One might have hope or fear directed towards objects of merely instrumental value. But many attitudes, from grief, love, and awe to hatred and resentment, generally value objects non-instrumentally.

For a Smithian account of non-instrumental value, showing that attitudes ascribe non-instrumental value is not sufficient. This would be adequate for a subjectivist account, but it is all-too-clear that people sometimes improperly value things non-instrumentally. Adam Smith describes those who “ruin themselves by laying out money on trinkets of frivolous utility . . . . [and] walk about loaded with a multitude of baubles, . . . some of which may sometimes be of some little use, but all of which might at all times be very well spared, and of which the whole utility is certainly not worth the fatigue of bearing the burden” (IV.i.6). One could multiply examples. Children who throw tantrums when denied desired toys, even when given equivalent ones, improperly ascribe intrinsic value. Adults who intrinsically value money are further from propriety, since money’s value is essentially instrumental. Awe towards ordinary pebbles or grief at pruning trees are improper valuing attitudes. Failing to feel awe at the Grand Canyon or grief at a beloved parent’s death reflect improper attitudinal deficiencies. The fact that one has or fails to have particular attitudes towards something is insufficient for determining its intrinsic value. What matters is what attitudes towards an object are proper.

Still, given that some attitudes value non-instrumentally, one need only explain when those attitudes can be proper to complete one’s account of non-instrumental value. Applying Smith’s account of propriety requires looking at the details of particular cases, but even my brief account of Nussbaum’s grief over her mother indicates that one can at least partially sympathize with this grief. And while full assessments would require becoming well-informed about relevant details, there are strong presumptions in favor of some Smithian, non-instrumentally valuing attitudes, including not merely desires for food and sex but also love for caring parents and grief at the death of loved ones.

b. Trumping Value

Things with non-instrumental value need not have trumping value. I may value a painting non-instrumentally, but still be willing to sell the painting if I need money. The fact that my appreciation of the painting is non-instrumental does not imply that it trumps other values (including even instrumental values such as the desire for money).[16] Nonetheless, proper-attitude explanations of trumping value are similar to those of non-instrumental value. Certain attitudes ascribe trumping value to their objects, and when such attitudes are proper, their objects have trumping value. The sort of appreciation one has for paintings does not generally imply that trumping value, but the distinctive sort of appreciation – approaching reverence – that one has for historically, culturally, or artistically unique and important paintings does. One can properly appreciate a fine painting and still sell it, but one’s appreciation of Raphael’s Sistine Madonna would be insufficient, and hence improper, if it did not imply at least some prima facie trumping value of that painting over other concerns. More obviously, respect for others and love for parents or children ought be not only non-instrumental, but trumping. Proper objects of these kinds of reverence, respect, or love have trumping value.

Unfortunately, Smith’s account of propriety poses a challenge for making sense of this sort of intrinsic value. For Smith, propriety depends upon the particular situation that prompts an attitude. Even general rules in ethics “are ultimately founded upon experience of what, in particular instances, our moral faculties . . . approve of” (III.4.8, p. 159).[17] Because situations differ and impartial judgments about particular situations are ultimate arbiters of propriety, Smith seems unable to say that anything has intrinsic value in every situation, which seems to make it impossible for Smith to account for trumping value. This problem is not a general problem for proper attitude accounts, which can be non-particularist (e.g. Korsgaard). But it seems to be a problem for Smith.

Fortunately, Smith can address this problem. First, even if Smith were unable to provide for absolute trumping value, he could explain the sort of prima facie trumping value advocated by philosophers such as J. Baird Callicott, who allows that intrinsic (trumping) value “may be overridden by considerations of public or aggregate utility” (Callicott 2002: 14). Smith’s account of proper attitudes supports this sort of trumping value, since it may generally be proper to take an intrinsically-valuing attitude towards something, though particular cases make this attitude improper.[18]

But Smith also offers an account of general rules that allows for absolutely trumping values that cannot be overridden by any considerations whatsoever. Smith explains the need for general rules using an example:

The man of furious resentment, if he was to listen to the dictates of that passion, would perhaps regard the death of his enemy, as but a small compensation for the wrong, he imagines, he has received; which, however, may be no more than a very slight provocation . . . . [T]he fury of his own temper may be such, that had this been the first time in which he considered such an action, he would undoubtedly have determined it to be quite . . . proper, and what every impartial spectator would approve of. (III.iv.12, p. 160)

Smith mentions this “man of furious resentment” as an example of “self-deceit” (III.4, pp. 156-61): “the violence and injustice of our own selfish passions are sometimes sufficient to induce the man within the breast [our attempted impartial view of ourselves] to make a report very different from what the real circumstances of the case are capable of authorizing” (III.iv.1, p. 157). Even “when the action is over . . . and the passions which prompted it have subsided,” although “we can enter more coolly into the sentiments of the impartial spectator,” we “often purposely turn away our view from those circumstances which might render the judgment [of ourselves] unfavorable” because “it is so disagreeable to think ill of ourselves” (III.iv.4, pp. 157-8). The ideal standpoint from which one ought to evaluate objects’ values is difficult, especially when interests are particularly acute. Because people misjudge values that interfere with their interests, we should develop “general rules” based on truly impartial judgments. With the furious man,

unless his education has been very singular, he has laid it down to himself as an inviolable rule, to abstain from [“sanguinary revenges”] upon all occasions. This rule preserves its authority with him, and renders him incapable of being guilty of such a violence . . . [R]everence for the rule . . . checks the impetuosity of his passion, and helps him to correct the too partial views which self-love might otherwise suggest. (III.iv.12, pp. 160-1)

General rules based upon particular but impartial judgments trump particular judgments in contexts where partiality causes self-deception. Where values are liable to be systematically/universally misjudged through interested self-deception, one should ascribe absolute trumping values.

This Smithian account of trumping values seems, in one sense, weak. General-rule-based trumping values are in principle revisable in the light of further reflection. But this revisability does not imply that value does not trump, only that one’s judgment that it trumps is revisable. Willingness to revise judgments requires acceptance of one’s own limited perspective, not diminution of objects’ values. Moreover, Smith’s account of trumping value helpfully shows that trumping value is so important because it imposes limits on casuistry, preventing one from seeking ways around the requirements to respect something’s value. This need for clear lines is a particularly human need that comes from our tendency to misuse casuistry (a tendency so universal that “sophistry” is given by the OED as a synonym for casuistry). Ascribing trumping value via Smithian general rules not only provides for robust conceptions of trumping, but also shows what trumping value really is and why it is so important.

c. Non-relational value

Even something with non-instrumental (and/or trumping) value might lack non-relational value.[19] Something has objective or non-relational value if its value does not depend upon its relationship to anything else. The precluded dependence could be instrumental, as when the value of gold depends upon its capacity to purchase goods; but value might depend upon relationships non-instrumentally, as with the special value of children to their parents or the value of a benefactor due to past benefits. In these cases, value is non-instrumental but still relational.[20] Nussbaum’s example of grief over her mother exemplifies non-instrumental, relational value.

What inspires grief is the death of someone . . . who has been an important part of one’s own life. This does not mean that the emotions view these objects simply as tools or instruments of the agent’s own satisfaction; they may be invested with intrinsic worth or value, as indeed my mother surely was. . . . But what makes the emotion center around this particular mother . . . is that she is my mother, a part of my life. (Nussbaum 2001: 31)

This example makes clear the distinction between instrumental and relative goodness. Many valuing attitudes – filial love, gratitude, and grief – are appropriate with respect to parents to a degree that would be inappropriate with respect to others. These attitudes do not ascribe instrumental value to one’s parents. To love or feel gratitude towards parents only for what they can do for one would be deeply improper. But these attitudes are proper only because of a relationship between their object and oneself. They are deeply relational valuings, and one’s parents’ value is, in that sense, a relational value.

Gratitude is the paradigm case of non-instrumental, relational valuing. Hobbes famously defended the importance of gratitude for instrumental reasons (cf. Hobbes 1651/1996: 105), but Smith is much closer to the mark, seeing gratitude as a “sentiment, which . . . immediately and directly prompts us to reward . . . another” who has benefitted oneself (II.i.2.1). Gratitude is an immediate response to benevolence, involves immediate concern for another’s well-being, and takes this other’s well-being as an end-in-itself. But gratitude is nonetheless relational. One properly feels gratitude towards only those by whom one is “assisted, protected, relieved” or benefited in some other way (II.i.2.4).

Unfortunately, while Nussbaum rightly distinguishes non-relational from non-instrumental value, she uses the relational nature of her grief to argue that all emotions are “eudaimonistic” (Nussbaum 2001: 31) in the sense that although “they insist on the real importance of their object,” “they also . . . have to do with me and my own, my plans and goals, what is important in my own conception . . . of what it is for me to live well” (Nussbaum 2001: 33):

Let us now return to my central example. My mother has died . . . [A] person of enormous value, who was central in my life, is there no longer. . . . [This emotion] is evaluative and eudaimonistic; it does not just assert “Betty Craven is dead.” Central to the [grief] is my mother’s enormous importance, both in herself and as an element in my life. (Nussbaum 2001: 39).

Nussbaum’s self-analysis here is (I suspect) accurate and insightful. The emotional force of “my mother is dead” is completely different than the emotional force of “Betty Craven is dead.” The grief wrapped up with former ascribes importance to its object both in-itself and in relation to oneself. This grief is, as Nussbaum puts it, eudaimonistic, in that it ties the value of its object to one’s own life in a non-instrumental way.

But Nussbaum goes too far in using this analysis for emotion in general or even grief in general. Images of war or disease, especially when victims are children, inspire grief. News reports of tragedy and even cold statistics can inspire grief in one sufficiently attentive, who has not grown callous, who is willing to deeply reflect on the news. To some extent, these feelings may be due to a sense of distant community, that we are all in this world together. To some extent, they may be due to associations with those closer to one’s own life, stirring up fear or sadness at future dangers or past misfortune. To some extent, one might connect them with oneself through a sort of self-oriented sympathy with others, where one grieves over the pain that one feels in contemplating those situations. But it is proper to feel grief at others’ tragedies even when those others have no special importance in one’s life. Innocent lives lost provoke grief, and in grief one ascribes to those lives “enormous importance,” an importance that is often simply “in-itself” and not at all “as an element in my life.”

Of course, the attitudes that ascribe value non-relationally will be different – in degree and often in kind – from those that ascribe value relationally. Grief at the death of a parent is different, and properly so,[21] from grief at the death of a stranger. Love for one’s children differs from love for parents, lovers, and strangers in need. But love is proper in all of these cases, and love for strangers in need will at most be very weakly relational, and it will not – or at least need not – involve any appeal to the role of those strangers in one’s life. In that sense, attitudes can properly ascribe value non-relationally, and proper-attitude theories of value can explain objects’ non-relational value.

But there are at least two ways in which non-relational value poses special problems. First, Smith’s particularism seems to pose challenges for a Smithian conception of non-relational value. The non-relational character of intrinsic value laid out by Moore required its universality: “it is impossible for what is strictly one and the same thing to possess that kind of value at one time, or in one set of circumstances, and not to possess it in another” (Moore 1922: 260). Again, Smith might appeal to general rules here, although it is not clear that they will be as effective in this case as in the case of trumping values. Second, one might question whether any account of value that essentially ties the value of an object to humans’ attitudes towards that object can be properly non-relational.

Fortunately, both problems can be alleviated by making an important distinction between the situation-dependence of attitudes (Smith’s particularism) and attitudes of situation-dependent value. The propriety of ascribing value may be situation-dependent, but the value thus properly ascribed need not be. Smith even emphasizes that while every attitude’s propriety is situational, values ascribed by those attitudes are typically not based on relations to oneself. “Passion[s] excited by objects peculiarly related to ourselves” (I.ii.intro.1) are merely one sort of passion among others, and Smith specifically highlights the pleasure that impartial spectators take in “affections . . . towards those who are not peculiarly connected with ourselves” (I.ii.4.1, pp. 38-9). The situation-dependence of an attitude is not the same as an attitude of situation-dependent value.

To illustrate: Grief at the suffering/death of young innocents ascribes value to others independent of their relation to oneself, or even to anything else. The grief-value of suffering innocents is non-relational. Nonetheless, the attitude of grief over innocents is deeply situation-dependent. Grief is proper when one becomes aware of suffering, especially when suffering is especially needless and one’s awareness particularly vivid, but such grief is usually not proper at the birthday party of a neighbor’s child. This impropriety is due neither to the fact that no innocents are dying – unfortunately, they are in large numbers – nor even to one’s ignorance of that fact – one may be far from ignorant. But the situation calls for something else, for putting one’s grief to the side, for enjoying the festivities of the moment. This situation-dependence relates to the propriety of feeling the attitude, not a situation-dependence constitutive of the attitude. The fact that feeling grief is inappropriate in certain contexts does not mean that grief, when properly felt, ascribes contextual value to its objects. Grief ought not have the form, “innocents’ suffering and death leaves a vacant hole in our world . . . unless I get to go to a birthday party.”[22]

Like grief, attitudes such as resentment or gratitude are situation-dependent in that one ought not feel intense resentment at global injustice while at the birthday party of a neighbor’s child, nor gratitude for a promotion when comforting a grieving friend. Unlike grief, however, these attitudes also properly ascribe situation-dependent value to their objects. Resentment “prompt[s] us to desire . . . that [the object of our resentment] should be . . . upon account of that particular injury which he had done to us” (II.i.1.6). And gratitude seeks good for another because the other has done good to oneself (II.i.1.5). The value (either positive or negative) of the objects of gratitude and resentment is situation-dependent value.

Like grief, gratitude, and resentment, respect for persons and awe at the Grand Canyon are proper attitudes that are situation-dependent. But like grief and unlike gratitude and resentment, they are not attitudes of situation-dependent value. One wandering alone through a vast and beautiful wilderness need not consciously have respect for persons, and one not in the presence of or contemplating the Grand Canyon need not feel awe. But when in the right context, these attitudes ascribe situation-independent value to their objects. One who truly respects another person ascribes to her a value that does not depend upon situation. Likewise, one in awe sees the Grand Canyon as being awesome simpliciter. The Grand Canyon is not awesome simply because of situational properties (its place in space or time, its designation as a National Park, its being viewed by me, etc), but in itself. Situational prompts[23] may be needed to activate awe, but the awesomeness one ascribes when those prompts are present does not depend upon those prompts. Proper awe at the Grand Canyon ought not have the form, “With me standing here on its edge, this Canyon is pretty awesome.”

The distinction between situation-dependent attitudes and attitudes of situation-dependent value is related to another distinction that is crucial for explaining the sense in which Smithian intrinsic value is non-relational. Moore developed the notion of intrinsic value to defend meta-ethical realism about value. Smith’s proper-attitude theory is precisely offered as an alternative to this meta-ethics, but Smith can still accommodate normatively non-relational value. Normatively non-relational value is value that an object has by virtue of being the proper object of a valuing attitude, where the attitude that one ought to take towards the object ought to be taken without regard to any relations between that object and anything else. A non-relational meta-ethical theory of value, unlike this normative non-relationality, does not pick out a particular sort of value. Rather, it posits that things have value independent of any relations, including normative ones, to human beings. While Smith could support some forms of meta-ethical non-relationalism about value (e.g., the idea that the value of a thing need not depend upon its actual relation to any other actual beings), proper-attitude theories are designed precisely to offer an alternative to meta-ethically non-relational (objectivist or realist) treatments of value. And this section has shown – contra Moore – how non-realist meta-ethics can justify normatively non-relational values.

3. Conclusion

This paper started by noting that “intrinsic value” can refer to at least three distinct sorts of value, and I have shown that Smith’s proper-attitude theory of value can make sense of all three of these core senses of intrinsic value. In this conclusion, I show how the notion of intrinsic value might be enriched even further through Smith’s account. Recently, the notion of “thick” ethical concepts has increasingly gained traction.[24] The general idea is that concepts such as cruelty, disloyalty, humility, and generosity are less abstract and more descriptive than more “reflective” ethical concepts such as “good” or “right.” For Bernard Williams, thick concepts are better candidates for “convergence” (Williams 1985: 140f.) because they have more contentful connection to the world. For example, one can get wide agreement that torturing animals is “cruel” without prior agreement about whether such action is “wrong.” “Cruelty” describes the action and ascribes normative weight to it, but it assigns normative weight in a “thick” way. [25]

Proper-attitude theories can extend this notion of “thick” concepts into discussions of value while avoiding Williams’s relativism about thick conceptual schemes. “Value” and “intrinsic value” are thin concepts, abstract and lacking natural descriptive sense. Proper-attitude theories thicken value concepts because values reflect specific, potentially thick, normative attitudes. Terms like “awesome,” “pitiable,” “loveable,” “fearsome” and “worthy of gratitude” are all thick. Calling the Grand Canyon “awesome” and suffering children “pitiable” (as in “worthy-of-pity”) is more descriptive, less controversial, and better connected to normative claims on attitudes than saying that they have “intrinsic value.”[26]

In other words, Smithian intrinsic value is pluralist.[27] Distinguishing non-instrumental, trumping, and non-relational value begins to highlight that pluralism, but the full range of thick intrinsic values – awesomeness; fearsomeness; worthiness of respect, grief, or gratitude – describe intrinsic values more accurately and foster more productive ethical dialogue than abstract debates about “intrinsic value” itself.

Proper-attitude accounts of thick values also provide for connecting thick and thin value concepts. On proper-attitude theories, value concepts are thick and might even (as for Smith) be irreducibly thick, and these thick concepts are connected to thin concepts. Value concepts are thick in that ascriptions of value are based on particular attitudes. They are irreducibly thick when, as in Smith, there is no fixed rule for deciding which descriptive accounts warrant particular attitudinal responses. At the same time, Smithian thick value judgments connect to thin concepts in three important ways. First, Smith emphasizes “propriety,” a thin ethical concept. Second, one can abstract thin notions of intrinsic value from more specific ones, as I did extensively in the last section; and these abstract notions may play roles in explaining concrete ethical judgments. Finally, Smith’s proper-attitude theory elucidates universal ways of arbitrating disputes about thick values based on thin ones. In the case of Smith, one arbitrates disputes about thick values by means of the “thin” concept of the impartial spectator. Smith thus avoids relativistic implications that pervade Williams’s use of thick values and that affect even Anderson’s more modest value-pluralism.

The final result is a Smithian theory in which proper objects of various valuing attitudes have value. Because of the various ways these attitudes ascribe value to objects, one can abstract different sorts of “intrinsic” value, including but not limited to the non-instrumental, trumping, and non-relational values that play important roles in contemporary ethics (both theoretical and applied).

---

[1] Prior to Moore, the term “intrinsic value” was primarily used – as it was by Smith (cf. WN IV.ii.4) – in economic contexts, to distinguish the nominal value of a thing (especially coins) from its “intrinsic value” (e.g., real weight of coins).

[2] See Lemos 1994, Smith 1998, Zimmerman 2001 and 2007, Audi 2004. For further discussion, cf. D’Arms and Jacobsen 2000 and Rabinowicz and Ronnow-Rasmussen 2004.

[3] Defenders of intrinsic value in environmental ethics include Callicott, Rolston, Naess, and Agar. Critics include Weston, Hargrove, Norton, Regan, Gruen, and Light.

[4] For references to intrinsic value in relation to abortion, see Dworkin. For relation to modern art, see http://query.nytimes.com/gst/fullpage.html?res=9F02E5D7103FF934A25751C1A9669C8B63 (accessed 9-2-2008). Relating to going to college, see http://essay.blogs.nytimes.com/2007/09/24/why-college-matters-7/ (accessed 9-2-2008).

[5] How trumping this value is varies. Contrast Dworkin 1984 and Callicott 1999, 2002.

[6] O’Neill 2001:165. Cf. Korsgaard 1997: 249-74, Moore 1902 and 1912, Ross 1930.

[7] One could even transform this proper-attitude account into subjectivism by claiming that all actual attitudes are proper or into objectivism by claiming that a valuing attitude is proper only if responsive to objective value-properties.

[8] See too Gibbard 1990; Solomon 1976; Nussbaum 1990, 2001.

[9] Like Smith and unlike Korsgaard, Elizabeth Anderson defends a pluralist approach to value according to which a wide variety of different attitudes confer value on their objects. But Anderson’s resistance to rational desire models of value leads her to limit her pluralism in ways that Smith does not. Smith sees all human attitudes as potentially value-conferring, while for Anderson desire itself is not a source of value (Anderson 1993:130). Thus Smith can, with Korsgaard, see mere desire to realize a state of affairs – if proper – as conferring value on that state of affairs, but Smith also recognizes values of objects that cannot be “realized or brought about,” such as the Grand Canyon (through awe) or other people (through respect, gratitude, grief, love, and even hatred and resentment). By being open to value from any quarter, Smith’s ethics provides for a full range of sorts of value.

[10] See Campbell 1971, Fleischacker 2004, Frierson 2006a, Griswold 1999, and Otteson 2002.

[11] Smith assumes that the reactions of truly impartial and attentive spectators will be the same. See Heath 1995: 452 and Frierson 2006a.

[12] Re: impartiality, see Griswold 1999 and Campbell 1971. My contrast of Smith with Firth largely follows Griswold, but I take Smith’s impartial spectator more ideal than Griswold does in that I see the impartial spectator as primarily an imaginative construct.

[13] Some Smithian expressions might seem to imply that impartial spectators respond to objective values. Smith says, for example, that a spectator should “approve of the passions of another . . . as suitable to their objects” (I.i.3.1, p. 16). But even here, he makes clear that this approval is based on the spectator’s sympathetic response, not vice versa: “When the original passions of the person principally concerned are in perfect concord with the sympathetic emotions of the spectator, they necessarily appear to this last just and proper, and suitable to their objects . . . . To approve of the passions of another . . . as suitable to their objects, is the same thing as to observe that we entirely sympathize with them” (ibid.). Passions being suitable to their objects amounts to no more than that impartial spectators fully sympathize with them.

[14] See Korsgaard 1996: 240.

[15] Smith’s overall approach to intrinsic value has another advantage in the context of recent critiques of the notion of intrinsic value by Justin D’Arms and Daniel Jacobsen (2000a, 2000b, cf. Rabinowicz and Ronnow-Rasmussen 2004). By defining something’s value in terms of proper attitudes, Smith simply leaves no room for gaps between emotions’ “fittingness” and its moral “appropriateness.” This gap depends either upon realism about fittingness, moral evaluation, or both, or upon a moral consequentialism that judges the appropriate of emotions in terms of something external to their fittingness. While a full discussion of Smith’s response to D’Arms and Jacobsen is beyond the scope of this paper, Smith’s ethical theory not only avoids these alternatives but also explains the intuitions that generate conflicts in D’Arms and Jacobsen’s examples. In particular, we can have “moral maxims” that conflict with concrete moral judgments, as in the case of judging in abstracto that the same acts should have the same merit regardless of consequence but in concreto judging acts with bad consequences more harshly (Smith’s example) or judging that racist jokes should not be funny in abstracto but finding particular ones funny (D’Arms and Jacobsen’s).

[16] The independence of trumping and non-instrumental value is in principle reciprocal. Someone might ascribe trumping value to her biological life, for example, while still seeing life as merely a means – but an absolutely necessary one – to other goods. In fact, anything perceived as a necessary means to all other goods will be trumping, whether or not it is valuable for its own sake.

[17] Re: particularism, cf. Dancy 2004; Hooker and Little 2001. Re: Smith’s particularism, cf. Griswold 1999; Fleischacker 2004; Frierson 2006a.

[18] Jonathan Dancy defends a similar account in terms of “default value.” Cf. Dancy 2003, especially p. 638.

[19] Here independence seems non-reciprocal. Something that has non-relational value must have non-instrumental value, since value that is independent of any relation cannot be value merely as a means to some further end. However, non-relational value need not be trumping (and vice versa).

[20] A thing’s relationship to something else might give it value by virtue of the value of that to which it is related, as Smith’s example of gratitude towards “the plank upon which [one] had just escaped from a shipwreck” (II.iii.1.2, p. 94). But relational value might also not depend upon the relata’s value, as in parental love. Although parents see their parental relationship as central to their children’s value for them, they should not see their own value as the source of their children’s value. This account of non-relational value differs from Anderson’s, according to which “the judgment that x is intrinsically valuable entails that . . . x is properly intrinsically valued, independent of the propriety of valuing any other particular thing” (Anderson 1993: 3). Insofar as children are properly valued by their parents, independent of the propriety of valuing any other particular thing but only because of their relation to their parents, they lack intrinsic (non-relational) value.

[21] For now I simply stipulate this. If it is not properly different, then one has an even stronger argument for non-relational value on a proper-attitude account. But for Smith, at least, grief over parents will be properly different from grief over strangers. If for no other reason, the gratitude that is appropriate in the case of parents will require a different (more intense) sort of grief, but Smith also takes personal relationships to be directly relevant in assessing proper attitudes.

[22] The grief-value ascribed to dead innocents is situation-dependent in another sense. If the innocents over whom one grieves have not actually suffered or died, then not only is it improper for one to feel an attitude of grief, but the innocents lack the property ascribed to them in grief. They lack the value of being worthy-of-being-grieved-over. One might quite properly say, “the loss of these children leaves a vacant hole in our world, unless, of course, they haven’t really suffered or died.” In this sense, grief is an attitude of situation-dependent value. Of course, there is a derivative sense in which they still have this value. Insofar as they are such that one ought to grieve over them were they to suffer or die, we can say that they are grief-worthy in a secondary sense.

[23] Dancy’s discussion of “enabling conditions” (Dancy 2003: 637) complements the account here. Cf. Dancy 2003, 2004.

[24] See Williams 1985.

[25] The role of thick concepts has been taken by some to suggest a break-down of the fact/value dichotomy (cf. Williams 1985, Putnam 2004) and by others to imply that ordinary ethical concepts involve both factual and ethical components (cf. Blackburn 1998). Some, such as Linda Zagzebski (2003) and Bernard Williams (1985), define thick ethical concepts to be such that the descriptive and evaluative components cannot be pulled apart, but it is not necessary to do this in order to recognize a principled distinction between ethical concepts that are closer to the texture of ordinary life with all its detail, and those that are more abstract. Regardless of one’s judgment about metaethical issues such as cognitivism, “thick” ethical concepts such as cruelty and courage clearly play important roles in human life, and these ethical concepts are more descriptively rich than “thin” ethical concepts such as “good,” “right,” and “ought.”

[26] Although knowing that the Grand Canyon is awesome or a child pitiable does not give rise to immediate action-plans – it does not directly tell us what to do with the Grand Canyon or the child – it provides substantially more guidance than simply saying that it has intrinsic value. In fact, the Grand Canyon’s being awesome in part explains why value concepts, as applied to the Grand Canyon, do not immediately imply action. Awe is precisely a passive attitude, one that limits action, and – at least in its purest form – limits action unconditionally.

[27] Cf. Anderson 1993: 98-103.