History of Philosophy Quarterly, 1993, Vol. 10, pp. 271-277.
James Fieser

The intuition behind evolutionary ethical theories is that human social behavior is an extended development of biological evolution. Ethical behavior, then, parallels the complexity and development of the society in which it occurs. Such theories are bold attempts to give moral theory an empirical referent and to include ethics among the social sciences. In spite of these high hopes, many evolutionary ethical theories are visibly flawed, such as the following by E.D. Cope:

In reaction to such evolutionary attempts at naturalizing ethics, G.E. Moore criticizes evolutionary ethics in the second chapter of Principia Ethica.(2) Moore's target is Herbert Spencer's theory as it appears in Spencer's popular work, The Data of Ethics.

Moore offers two lines of attack. The first is that Spencer wrongly uses the terms "more evolved" (a natural term), and "higher" and "better" (ethical terms) as though they were equivalent, and thereby commits the naturalistic fallacy. The second is that since Spencer believes that pleasure is the same as goodness, then he is once again committing the naturalistic fallacy. It will be argued here that both of Moore's attacks fail since he misunderstood Spencer's meanings of the terms "higher" and "good." In light of Moore's failure, it is suggested that the naturalistic fallacy must be qualified so it does not reduce simply to an allegation that an ethical theory denies moral realism. The present analysis begins with a sketch Spencer's ethical theory, and Moore's account of the naturalistic fallacy.

Spencer and Moore

Much of Moore's argument against Spencer is in reference to the following summary passage from Spencer's The Data of Ethics:

Spencer's point is that as the complexity of the organism increases from insect to humans so, too, increases the complexity of conduct. Humans, the most complex organisms, have the most complex conduct, and ethics is part of that conduct. Thus, ethical conduct (1) has emerged only in the most developed life form, that is, human life, (2) is the most developed form of human social conduct, and (3) has emerged only in the most advanced human societies.

Spencer argues further that all conduct, whether of humans or animals and whether moral or nonmoral, is the adjustment of acts to ends. Since deeming actions as good or bad is part of ethical conduct, "always, then acts are called good or bad according as they are well or ill adjusted to ends." An action is good if it helps bring about an end, while it is bad if it does not help bring about an end. But what is the "end" to which Spencer refers? Spencer writes that, "the end of good acts is their being conducive to life in oneself and in others. Life, itself, is good or bad according as it does or does not bring a surplus of agreeable feeling."(4) He argues further that "it becomes undeniable that, taking into account immediate and remote effects on all persons, the good is universally the pleasurable."(5) Good actions, then, are those which have universal pleasure as their end.

In short, Spencer's ethics is evolutionary since ethical behavior is found only in the most highly evolved form of behavior. Further, by linking good actions with universal pleasure Spencer endorses hedonism. Against both the evolutionary and hedonistic aspects of Spencer's ethic, Moore accuses Spencer of committing the naturalistic fallacy.

Moore's classic explanation of the naturalistic fallacy appears in Chapter one, Section ten of Principia. Briefly his point is that the term "good", like the term "yellow", is a simple, and thus irreducible to any constituent parts. Therefore, "good" cannot be defined in terms of other properties, such as pleasure or highly evolved conduct. The name "naturalistic fallacy" implies that it is a fallacy to define good (which is non-natural) in terms of natural properties. For as Moore writes, if someone "confused 'good'... with any natural object whatever, then there is reason for calling that the naturalistic fallacy."(6) However, there is more to the naturalistic fallacy than simply identifying "good" with a natural object. In Chapter four of Principia, Moore argues further that it is improper to define "good" in terms of any non-natural quality as well as natural quality. For instance, "good" may not be defined as doing the will of God, which is a nonnatural quality. Accordingly, William Frankena argues that Moore is actually describing a definist fallacy which has as subsets the naturalistic fallacy and the metaphysical fallacy.(7)

Assume, for example, that all morally commendable actions (such as charity) have the following attributes: (1) they promote universal pleasure, (2) they constitute the most highly evolved conduct, and (3) they are good actions. The definist fallacy would occur if we identified the third attribute with either of the first two attributes. This would happen, for example, if we claimed that goodness and the promotion of universal pleasure were the same attributes. However, it would be no fallacy to claim that universal pleasure and highly evolved conduct are qualities which always accompany goodness. Indeed, "it is a fact, that Ethics aims at discovering what are those other properties belonging to all things which are good."(8)

Identifying "Good" with "More Evolved"

Moore first attacks Spencer for identifying good behavior as highly evolved behavior: "...what a very different thing is being 'more evolved' from being 'higher' or 'better'.... But Mr. Spencer does not seem aware that to assent the one is not to assent the other."(9) Moore's reasoning is that in the natural evolutionary process, some organisms and behavioral traits are less evolved, while others are more evolved. By contrast, some human conduct is ethically lower or worse while other conduct is ethically higher or better. Moore argues, though, that noting the degree to which some conduct is ethical is an entirely different process that noting the degree to which some conduct is evolved.

Spencer indeed does identify higher conduct with more evolved conduct as Moore charges. However, Spencer uses the term "higher" in a restricted sense and only in the context of one of three areas of evolution:

Each of these three areas have their own evolution. Thus, when Spencer writes of the higher complexity of an organism, he means only that an organism is more evolved. Similarly, to say that some conduct is higher he means only that it is more evolved conduct. For Spencer, conduct is higher if and only if that conduct is ethically significant (i.e. that conduct may be deemed either good or bad). By contrast, ethically superior conduct for him would be conduct which produces universal pleasure. Since the term "higher" for Spencer means only more evolved, and does not denote ethical superiority, Spencer does not commit the definist fallacy.

In another passage, Moore offers a second argument attacking Spencer's identification of "good" with "more evolved":

For Moore, the above phrase, "that conduct gains ethical sanction" means that some conduct is recognized as good, or is ethically commendable. Moore's complaint in this passage is that Spencer identifies more evolved with ethically commendable which commits the definist fallacy. That is, Spencer is confusing goodness (ethical commendability) which is undefinable, with a natural property (more evolved).

But, Moore is attacking a straw man since Spencer has in mind a weaker claim than the identification of ethically commendable with more evolved. Spencer's point is that the highest form of conduct is mutual cooperation in a society. The result of this conduct is that more pleasure is produced than would be the case if mutual cooperation did not take place. Thus, ethical commendability is an attribute which always accompanies the most evolved conduct since such conduct produces the most pleasure. Spencer, then, is not claiming that goodness (ethical commendability) is identical to more evolved conduct, but that being good is an attribute of more evolved conduct. As noted above, this would not commit the definist fallacy. If anything, Spencer would seem to be linking ethical commendability with the production of pleasure, which will be addressed subsequently.

Spencer's Hedonism

With regard to the evolutionary character of Spencer's ethics, then, Spencer cannot be accused of committing the definist fallacy. Moore's second attack on Spencer derives from the fact that Spencer "imagines [that] 'pleasant' or 'productive of pleasure' is the very meaning of the word good...."(12) A similar charge was launched against Spencer by Henry Sidgwick. In Principia, Moore rightfully acknowledges his debt to Sidgwick by noting that Sidgwick was the only other ethical writer who recognized the unanalyzable nature of goodness.

But Sidgwick's attack differs slightly from Moore's. Sidgwick agrees "with Mr. Spencer in holding that 'pleasure is the ultimate good,'" which is to say that pleasure is a necessary accompanying feature of goodness. However, Sidgwick disagrees with Spencer who defines goodness tautologically as pleasure, thus identifying the two. The foundational problem, for Sidgwick, is that "a tautology cannot be an ethical principle."(13) Moore objects to Sidgwick's endorsement of Spencer's hedonism and argues instead that good things do not necessarily have the attribute of being pleasurable. Nevertheless, Moore agrees with Sidgwick that Spencer errs by defining goodness tautologically since "propositions about the good are all of them synthetic and never analytic."(14) Both Moore and Sidgwick, then, would agree that Spencer commits the definist fallacy in this regard.

However, even when granting that Spencer identifies productive of pleasure and goodness, Spencer's specialized use of the term "good" raises questions about whether he commits the definist fallacy as Moore and Sidgwick charge. Spencer's notion of goodness is made clear in the following:

For Spencer, the meaning of "good" is not found in an intrinsic property, but is instead indicated by human desires and conventions. Moore, on the other hand, argues that he is not concerned with how people generally use the word "good" and how it is established by custom. Instead, he is concerned with discovering the nature of the idea, or, its intrinsic properties.(16) John Hill notes the fundamental difference between Moore's notion of goodness and the notion suggested by evolutionary ethics:

This difference can be seen throughout both Spencer's and Moore's writings where they describe things which are good. Spencer uses phrases such as "acts we call good", "we ascribe goodness," and "we apply them [i.e. 'good' and 'bad']."(18) Moore, by contrast, speaks in terms of something being good.

Given their differing notions of the term "good", a case may be made showing that Spencer does not commit the definist fallacy. It may be argued that "good" for Moore is undefinable only because goodness is seen as a property some things have intrinsically (wisdom, for instance, would be good intrinsically). Since Spencer rejects the idea that things are intrinsically good, and the term "good" is by itself an empty placeholder, then it would be no fallacy for him to define "good" as pleasure. Clearly, for Moore, this would not resolve the issue. Spencer would be criticized for misusing the term "good" when giving it a non-intrinsic, naturalistic definition.

At this point, the issue is no longer one of identifying two properties (i.e. pleasure and goodness), but one of denying the notion of intrinsic goodness. It makes little sense to say that Spencer commits the definist fallacy when, by denying intrinsic goodness, there is only one property under consideration (i.e. pleasure), and not two. The dispute is not definitional, but ontological and concerns the issue of moral realism. For, the issue of moral realism involves the assertion that goodness is an independent, objective quality of the world which right actions possess intrinsically, and wrong actions lack. To deny moral realism, then, as Spencer does, entails that there is no objective quality of goodness which is intrinsic to right actions.

However, it was neither Moore's intention, nor is it in the spirit of Moore to reduce the issue of the definist fallacy to the denial of moral realism. Accordingly, the following parameters should be set on the definist fallacy such that it is committed only when someone (1) asserts intrinsic goodness, and (2) identifies goodness with a second property (e.g. pleasure) which always accompanies a morally commendable action. Since Spencer's theory falls outside the first parameter above, then he cannot be accused of committing the definist fallacy.


It has been argued that the fault of evolutionary ethical theories, such as Spencer's, is not one of identifying highly evolved conduct with intrinsic goodness as Moore charges. Nor is the fault one of denying intrinsic goodness, for in this regard evolutionary ethics merely anticipates many contemporary accounts of value theory which also deny intrinsic goodness. Instead, the fault of evolutionary ethics involves the lack of objective criteria for distinguishing less evolved conduct from more evolved conduct. Little suggests that behavioral interaction in primal cultures is socially less evolved than behavioral interaction in technological societies. Without a clearly defined evolutionary spectrum of social behavior, no correlation can be drawn between such behavior and the degree to which it is morally commendable.


1. E.D. Cope, "On the Material Relations of Sex," Monist, 1890, Volume 1, pp. 39-40.

2. G.E. Moore, Principia Ethica, (London: Cambridge University Press, 1903), p. 58.

3. Herbert Spencer, The Data of Ethics, (New York: D. Appleton and Company, 1895), p. 20.

4. Ibid, p. 27.

5. Ibid, p. 30.

6. Principia Ethica, p. 13.

7. William Frankena, "The Naturalistic Fallacy," Mind, 1939, Vol. 48, pp. 464-477.

8. Principia Ethica, p. 10.

9. Ibid, p. 49.

10. The Date of Ethics, p. 8.

11. Principia Ethica, pp. 48-4.

12. Ibid, p. 53.

13. Henry Sidgwick, Lectures on the Ethics of T. H. Green, Mr. Herbert Spencer and J. Martinear (London; MacMillan and Co., 1901), p. 145.

14. Principia Ethica, pp. 7, 58.

15. The Data of Ethics, p. 21

16. Principia Ethica, p. 6.

17. John Hill, The Ethics of G. E. Moore: A New Interpretation, (Amsterdam: Van Gorcum, 1976), pp. 92-93.

18. The Data of Ethics, pp. 44, 22.