SciELO - Scientific Electronic Library Online

vol.2 special edition author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Kriterion: Revista de Filosofia

Print version ISSN 0100-512X

Kriterion vol.2 Belo Horizonte  2006


The problem of the naturalist fallacy for evolutionary ethics


O problema da falácia naturalista para o projeto de uma ética evolucionista



Karla Chediak*

Translated by Thomas Hasek
Translation from Kriterion, Belo Horizonte, v.47, n.113, p.147-157, June 2006.




One of the most difficult problems for those who defend evolutionary ethics is the naturalist fallacy. In this article, we examine the solutions proposed by W. Rottschaefer and R. Richards. We believe that these solutions are not good enough to completely eliminate the problem of the naturalist fallacy without compromising the specificity of morality.

KEY WORDS:  Naturalist fallacy, evolutionary ethics, ought, moral


Um dos problemas mais difíceis enfrentados pelos defensores de uma ética evolucionista é a falácia naturalista. Neste artigo, analisamos as soluções propostas por W. Rottschaefer e por R. Richards e verificamos não serem elas suficientes para eliminar o problema da falácia naturalista sem comprometer a especificidade da moral.

Palavras-Chave: Falácia Naturalista, Ética Evolucionista, Dever, Moral



One of the theses most discussed among biological philosophers is the one which affirms the existence of a close relationship between morality and evolution. It is based fundamentally on the understanding that man is a natural species like any other and, therefore, in order to explain man's origin, it is only necessary to use his natural history, that is, the process of human evolution. The evolutionary process alone would explain the origin and the formation of competencies in the human species.  

In general, the evolutionary thesis holds that the last 5 million years were fundamental for the development of human skills, be they cognitive or practical, and that biological history would have continued to exercise influence despite the fact that the cultural history of the species had begun. 

With this, there would be many links between the human species and other animal species. Although for other domains such as social life, intelligence and language, the influence of evolutionary biology is accepted, in relation to morality, the problem is quite complicated. In spite of this, some biological philosophers who affirm that there is a very close relationship between morality and evolution, conceive of a kind of evolutionary ethics.  They hold that an empirical investigation based on the theory of evolution can clarify the nature of morality, to the point of being capable of supplying a justification for some of our moral norms. This conception of morality has been subject to severe criticism, partly because morality is frequently understood as the maximum expression of the indetermination and the independence of man in relation to the rest of nature, something which expresses the capacity of the human species for self-determination and which has never been achieved by any other animal species.

Among the numerous criticisms of the link between morality and evolution, one stands out, not only because it is the most frequently cited, but because it is, possibly, the most difficult to refute—the naturalist fallacy. This term—the naturalist fallacy—was coined by Moore to combat the naturalist and metaphysical conception of morality.  

However, before Moore dealt with the issue, David Hume touched on the problem, for the first time, in a passage from A Treatise of Human Nature. He says:

In every system of morality, which I have hitherto met with, I have always remarked, that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a God, or makes observations concerning human affairs, when of a sudden I am surprised to find, that instead of the usual copulations of propositions, is and is not, I meet with no proposition that is not connected with an ought, or an ought not. This change is imperceptible; but is, however, of the last consequence.

 (Hume, 1978, p. 469).

Hume's passage is usually understood to mean that it is not legitimate to draw a moral conclusion, of a prescriptive nature, from factual premises insofar as these statements are of a different nature.  Hume would be denouncing the inconsistency of passing from "is" to "ought." One of the fundamental meanings of the naturalist fallacy is attributed to this passage of Hume.

The other meaning of the naturalist fallacy was presented by Moore and becomes clear in the following passage from Principia Ethica:

Far too many philosophers have thought that when they named those other properties [belonging to all things which are good] they were actually defining good; that these properties, in fact, were simply not "other," but absolutely and entirely the same with goodness. This view I propose to call the "naturalistic fallacy." ( Moore, 1978, p. 10).

According to Moore, it is fallacious to define goodness in terms of natural properties since goodness is not a natural property. It is also fallacious to define it using something else, as for example, happiness. In short, it is fallacious to define goodness which is, by its own nature, indefinable. For Moore, goodness is an unnatural, unique, and indefinable property.  

When analyzing the problem of the naturalist fallacy, Frankena, in the article "The Naturalistic Fallacy," makes a distinction between and correlates the two fundamental meanings of fallacy. The first one presents the naturalist fallacy as a logical fallacy, the derivation of an ethical conclusion from non-ethical premises, and the second as a fallacy of definition, in which one would define "goodness" using another property, for example, a natural quality.

According to Frankena, the naturalist fallacy in the sense of a logical fallacy, that is, the derivation of an ethical conclusion from non-ethical premises, can be easily solved with the introduction of an ethical premise. For example, the enunciation "pleasure is good since all men seek it" is fallacious because it stems from a supposed fact—"all men seek pleasure"— and concludes that "because of this, pleasure is good." In reality, this would be an entimema, an argument where a premise is suppressed, which could be made valid by making the premise clear. In this example, the ethical premise would be: "Whatever all men desire is good." So there would be nothing logically wrong with the following argument: "All men seek pleasure. Whatever all men seek is good. Therefore, pleasure is good." However, according to Frankena, the naturalist fallacy doesn't disappear from the argument, and this is because the naturalist fallacy is not, strictly speaking, a logical fallacy since this would only be the case if it were not possible to make the occult premise clear. The problem, then, is transferred to the way by which the ethical premise that was made clear is obtained, that is, to the way the premise was obtained: "Whatever all men seek is good." If it is obtained by another deduction, the problem is merely postponed. If it is by definition, in Moore's point of view, there is a fallacy since one would be identifying, that is, defining, a property through another. The naturalist fallacy would be, in reality, a type of fallacy of definition.

Frankena observes that for enunciations like "what is pleasurable is good" to be consider fallacious, it is necessary that one accepts the fact that good cannot be defined in terms of pleasure; therefore, this is the starting point of the argumentation and not exactly its conclusion. Otherwise, one runs the risk of committing petition principii since "one must know that the characteristic is non-natural and indefinable in natural terms before one can say that the definists are making a mistake" (Frankena, 1977, p. 60). The problem is that the opponent may not allow himself to be convinced that goodness is a unique and indefinable property and, therefore, he would not have committed a fallacy.

In fact, ever since it was presented, the problem of the naturalist fallacy has been the Achilles heel of all the theses that attempt to defend an evolutionary origin for morality, principally because, whether one accepts the thesis of Moore or not, the distinction between "is" and "ought" is almost universally accepted. This acceptance is due in part to the obvious distinction between enunciations which state how things are―descriptive enunciations, and those that state how they ought to be―prescriptive enunciations. The specificity of moral enunciations would be in the fact that they are normative and non-descriptive; the problem, mentioned by many philosophers, is that, when the two are mixed together, there is the risk of annulling the normative character of moral enunciations. This would occur, as Luc Ferry says, when "ought" is reduced to "is" in the following passage:

Or we reduce the normative to the descriptive, the right to the fact, the moral to the history and to the nature that determines it; however, in this case, it is necessary to renounce the idea of normative ethics and limit ourselves to describe moral behavior in a neutral way, as we do with animal behavior. Therefore, ethics no longer exists; there is only ethology which, with no judgment of value, is limited to showing why and how animals, human or not, conduct their lives. (Ferry and Vicent, 2000, p. 86)  

The risk of reductionism and, therefore, of the decharacterization of the normativity of moral enunciations, without a doubt, casts a shadow on the doctrine of moral evolution, but it is necessary to inquire whether all the conceptions of evolutionary ethics are necessarily reductionist, as Luc Ferry believes.

In relation to the problem of the naturalist fallacy by definition, it is reasonable to suppose that the evolutionist who defends the existence of a correlation between morality and evolution would hardly embrace Moore's conception that goodness is a simple property which is unnatural and indefinable in terms of another property. It doesn't make sense—for one who considers that human skills, whether cognitive or practical, had an evolutionary origin—to understand the concept of "goodness" as something so autonomous. Besides, as Frankena declares, if one does not initially accept such an understanding, there is no fallacy, since the very notion of unnatural properties would not be accepted.  After all, the affirmation of its existence is not self-evident and is precisely what is in question and what the evolutionist would probably deny.

In this way, we are left with the problem related to the derivational meaning of fallacy, namely, the passage from factual premises to normative conclusion. As Frankena said, this question is resolved by inserting a normative premise. However, the evolutionist must show how this premise is obtained without committing the naturalist fallacy. It would do no good to create a new deductive argument because the problem would be the same. The only option remaining is for the evolutionist to explain how he arrived at this normative premise to make the argument that associates facts and values valid. As Barrett observes in his article "Really Taking Darwin and the Naturalist Fallacy Seriously," there is a dilemma here for, if this premise is not explainable in factual terms, it is not relating the domain of ethics with that of facts, and the question nullifies itself because the theory of evolution would not have any relevance to morality. And if it were explainable in factual terms, one would be committing a derivational fallacy.   

An attempt to resolve the problem of the naturalist fallacy so as to defend evolutionary ethics with consistent arguments is carried out by William Rottschaefer in his article "Evolutionary Ethics: An Irresistible Temptation: Some Reflections on Paul Faber's The Temptation of Evolutionary Ethics." When considering the following fallacious deduction: "Evolution has shaped humans to pursue the community good. Therefore, the community good is morally valuable," he looks for a way to correct it by inserting another premise.  The argument goes like this: What is morally valuable is what evolution has shaped humans to pursue. Evolution has shaped humans to pursue the community good. Therefore, the community good is morally valuable (Rottschaefer, 1997, p. 372). The question is, then, one of explaining how we arrive at this inserted premise: "What is morally valuable is what evolution has shaped humans to pursue." For Rottschaefer, one of the ways to avoid the naturalist fallacy in its derived form is to debate the relationship between explanation and justification.  It is generally accepted that factual enunciations are subject to explanation but normative enunciations are not. The explanations given to a normative enunciation are not relevant to the acceptance of this enunciation since normative enunciations must be justified in such a way as to produce its acceptance. For this reason, by furnishing an explanation for the premise, one would be committing the genetic fallacy, for it is one thing to furnish the causes that explain the origin of a phenomenon, but it is another to furnish the reasons that justify it. Moral enunciations require justifications and not explanations.

However, for Rottschaefer, the theory of evolution would have a role not only in the explanation but also in the justification of morality without confusing them or considering the explanation and the justification distinct. The author accepts this distinction in certain cases, such as the well-known example of Kekule that the idea of the structure of the benzene ring would have been generated through a dream of a snake biting its tail. This could have been the origin of the idea but could not justify the structural formula of benzene. However, in other cases, the way that an idea is acquired would be sufficient to justify the belief. For example, in the case of perceptive beliefs, the perception is the origin of the belief and is sufficient to justify the belief. This case would be analogous to that of the relationship between evolution and morality. If it is really possible to maintain that some motivational, behavioral and cognitive capacities related to morality were selected during the evolution of the human species, then one can affirm that they ascertain at least in part, human values and one can also have recourse to them to justify actions. (Rottschaefer and Martinsen, 1991, p. 376).  

One example presented by the author is that helping someone who is  suffering, in pain, or sad is considered to be a morally good act. In addition, preserving his or her life helps the individual to have children and to be able to care for his or her children, should that be the case. Evolution would have selected types of behavior that favor survival and caring for children. So, survival and caring for children are generally considered good things and would justify  helping someone in difficult situations.

Rottschaefer states: "If we can find the mechanisms that reliably generate morally good stances, then we can appeal to them in our justifications as well as our causal explanations." (Rottschaefer, 1997, p. 376). Identifying the mechanisms that generate certain beliefs of moral order is, for the author, an empirical question and, therefore, refutable and of a probable nature, based on a posteriori not a priori mechanisms. Besides, Rottschaefer says that he does not defend the view that the theory of evolution is sufficient to justify moral practice, but he insists that it is one of its elements. Strictly speaking, no theory by itself would be sufficient to justify moral beliefs (Rottschaefer and Martinsen, 1991, p. 374).  

The heart of Rottschaefer's argument is to imagine that the evolutionary thesis can be both a source of explanation and a source of justification; in this way, that which was favored in adaptive, survival, and reproductive terms can justify its transformation into a moral value. The problem, in our opinion, is that this cannot be generalized without producing some absurdities. It is not only possible but very plausible that motivations and types of behavior have evolved and, therefore, favored the survival and reproduction of the individuals in a group; but, even so, they would not be considered moral. And if this is true, we cannot distinguish what is part of a moral system and what is not by using evolution as the basis. Evolution becomes insufficient to furnish a justification for morality, and that which evolved and which is also considered morally good has to be subject to another type of criteria, no longer of an evolutionary nature.

Another biological philosopher who tries to refute the argument of the naturalist fallacy is R. Richards in A Defense of Evolutionary Ethics. For him, the evolutionary theory should not repeat the errors of the past by affirming that if something evolved, then it is good, or by supposing that hierarchies can be established in terms of what is more or less evolved. Every specific culture would answer for its own social arrangement. However, Richards defends that there is a direct relationship between evolution and moral norms. The naturalistic fallacy would not, in reality, describe a fallacy, it being possible to derive norms from facts without committing a fallacy. Thus, it would be possible to justify the ethical premise of a mixed argument appealing to evolution (Richards, 1995, p. 269).

 Richards argues that any ethical system deals with at least three considerations that can and should be treated in an empirical way. First, an understanding of human nature—of what man is, of his capabilities and of the relationships between individuals. Secondly, an understanding of how the considerations about what is morally good are elaborated in human society.  And, thirdly, the way in which the justifications of moral systems and principles should be constructed (Ibidem, p. 271).  It is in relation to the third aspect that the important question arises because Richards recognizes that it is necessary to find a way to justify a norm that has an evolutionary origin, for example, a norm that says that the well-being of the community is the highest moral good. This would be possible if one understands that to justify is "to demonstrate that a proposition or system of propositions conforms to a set of acceptable rules, a set of acceptable factual propositions, or a set of acceptable practices" (Richards, 1995, p. 276). The argument is based on the fact that a system cannot justify its own principles. In this way, even if the propositions can conform to more general principles, the chain has to stop at some point. If it is not accepted that they are postulated, evident by themselves, or based on authority, then there is no way to avoid the appeal to experience. And this is what he intends to do in relation to the ethical principle cited above that says: "The well being of the community is the highest moral good." Evolution would have shaped the human species so that individuals protect their life, that of their children and that of the community, and this appears in the moral norms and values.  

The heart of Richards' argument is in the analogy established between reasoning processes like "if there is lightening, then there must be thunder" or  "if I am a member of the university, I must adequately prepare my classes" or even "if the human species evolved in such a way as to promote the well being of the community, then each one must act for the well being of the community." The argument that would justify the thesis would be:

The evidence shows that evolution has, as a matter of fact, constructed human beings to act for the community good; but to act for the community good is what we mean by being moral. Since, therefore, human beings are moral beings – an unavoidable condition produced by evolution – each ought act for the community good (ibidem, p. 281).

Richards relates the use that one makes of the term "ought" in the natural sciences—where it indicates what is probable, what is expected—with the moral sense of duty.  In a certain way, he believes that neither meaning is as distant from the other as is usually thought. Richards states that he took his model from Alan Gewirth, whose argument, according to Richards, is that the concept of "ought" means fundamentally what is required for reasons that originate from a structured context — what is expected to happen under normal conditions if there is no external interference. Therefore, there would be no fallaciousness in justifying the supposed ethical premise "Man should act keeping in mind the well-being of the community" by means of the following affirmation: "Evolution prepared the human species in such a way that it acts for the well-being of the community."   

 Justification by means of using facts as a basis is supported by this assimilation of the moral sense of "ought" and the sense of "ought" used by the sciences. However, it is not evident that this identification can be done without severe harm to the sense of moral duty. We agree with K. Ferguson's criticisms in Semantic and Structural Problems in Evolutionary Ethics where he says that, in making this assimilation, what is lost is the prescriptive character associated with moral enunciations. For him, there would be two distinct meanings of "ought," one indicating what is probable and which would be the scientific use and the other indicating what is required and which would account for the moral  sense (ibidem, p. 281).  In this way, to accept that the sense of the term "ought" meaning what is expected can be equivalent to the moral sense of the term, meaning what is demanded, even if they have similarities insofar as each one supposes a structured context in which it is formed, doesn't appear very convincing.  

It is notable that the two analyzed proposals that attempt to answer the problem of the naturalist fallacy are subject to consistent criticism. The establishment of a relationship between evolution and morality cannot be done at the cost of compromising moral specificity, that is, by means of indistinctness between justifying and explaining or by means of indistinctness between moral and non-moral duty. In this sense, it seems to us that it is difficult for the project of evolutionary ethics to rid itself of the problem of the naturalist fallacy in its form of derivational fallacy or even genetic or definition fallacy. However, this does not mean denying the relationship between morality and evolution; it means recognizing its limits.

In fact, defending evolutionary ethics of an objective character has been very problematic since it does not avoid the naturalistic fallacy. Because of this, as Michael Ruse states, the subjective conception of morality seems to be more compatible with the evolutionary thesis of morality (Ruse, 1995, p. 321). For this author, conceiving ethics as subjective would avoid the naturalist fallacy as definition fallacy because, by not being objective, moral values could not be assimilated by natural facts. It would also avoid the naturalist fallacy in its derived form because, if evolution alone furnishes an explanation, it could not justify moral values and norms.

In this way, a subjective perspective of morality, because it does not recognize the existence of a foundation or a justification for moral values, could only be based on the emotional inclinations of the species. However, subjective evolutionary ethics is not sufficient to account for human experience because the evolutionist cannot eliminate this dimension of ethics. It is not possible to negate that morality requires justification that can furnish reasons in favor of certain norms.

It is true that, for evolutionists who defend a relationship between morality and evolution, there would not be an a priori basis for morality. This does not mean that morality must be conceived as irrational in the sense of being only a direct manifestation of emotional dispositions of the species, like feelings of obligation, guilt, and compassion. These would be important insofar as they generate behavior of approval and disapproval of certain behaviors but would be insufficient to account for human moral behavior that requires justification.

If the proposal of understanding morality by means of evolution cannot find a basis that furnishes necessary reasons for our norms being as they are, it can, based on the contingent history of the species, furnish information and explanations for why they are the way they are. So, reflections with a basis in evolution have important contributions to make since they can explain the motivations that were and still are parts of the present norms in human societies, even if these norms, as improbable as it may be, may cease to exist. The fact of having been generated by evolution explains their probable permanence, but this fact alone does not seem sufficient to justify their permanence.



Barrett, J. "Really Taking Darwin and Naturalistic Fallacy Seriously," Biology and Philosophy, 6, p. 433-437, 1991.

Cosmides, L.; Tooby, J. & Lieberman, D. "Does Morality Have a Biological Basis? An Empirical Test of the Factors Governing Moral Sentiments Relating to Incest," Proc. Royal Society Lond., 2003.

Ferguson, K. G. "Semantic and Structural Problems in Evolutionary Ethics," Biology and Philosophy 16, p.69-84, 2001.

Ferry, L.; Vicent, J-D. Qu'est-ce que l'homme? Sur les fondamentaux de la biologie et de la philosophie. Paris: Odile Jacob, 2000.

Frankena, W. K. "The Naturalistic Fallacy," in Foot, P. Theories of Ethics. Oxford: Oxford University Press, 1977.

HUME, D. A treatise of human nature. Oxford: Clarendon Press, 1978.

McIntyre, A. After Virtue. Indiana: University Notre Dame Press, 1984.

MOORE, G. E. Principia ethica. Cambridge: Cambridge University Press, 1978.

Richards, R. J. "A Defense of Evolutionary Ethics," in Thompson, P. (ed) Issues in Evolutionary Ethics. New York: State University New York, 1995.

Rottschaefer, W. "Evolutionary Ethics: An Irresistible Temptation: Some Reflections on Paul Farber's The Temptation of Evolutionary Ethics," Biology and Philosophy 12, p. 369-384, 1997.

Rottschaefer, W.; Martinsen, D. "The Insufficience of Supervenient Explanations of Moral Actions: Really Taking Darwin and the Naturalistic Fallacy Seriously," Biology and Philosophy 6, p. 439-445, 1991.  

Ruse, M. "Evolutionary Ethics: a Phoenix Arisen," in Thompson, P. (ed.) Issues in Evolutionary Ethics. New York: State University New York, 1995.    



* Professor of  Philosophy at Rio de Janeiro State University. E-mail: