Tag Archives: ethics

Neither Naturalistic Nor Fallacious

What follows is an adaptation of an essay I wrote a few years ago during my studies at the University of Edinburgh. The course itself concerned “metaethics”, which is the study of ethical properties, as opposed to ethical conclusions. Essentially metaethics asks not what is good, but what is ‘good’. In order to build a more solid system of morality it is valuable to discuss exactly what we mean when we use ethical terms, and how exactly such terms should be understood. It’s a little heavy on philosophy and light on entertainment, but at very least I can hope that this is enlightening and perhaps of some use to those who do battle in the ethical colosseum.

‘Good’ is a rather mysterious entity. Many of us would easily admit that we know what ‘good’ is, but probe a little further and definitions will often erode to the point that we are reduced to saying that we know what ‘good’ looks like, but we don’t know what it is. To some ‘good’ is a natural property of objects or actions. If we observe the properties of such we will be able to identify instances. To others ‘good’ is something non-natural, but rather a more ethereal quality that permeates good things but cannot be pinned down and studied.

The naturalistic fallacy is a concept in metaethics first elucidated by G. E. Moore in his book Principia Ethica. For Moore ‘good’ cannot be identified with any natural property. ‘Good’ “has no definition because it is simple and has no parts.” Ordinarily we define objects or concepts with reference to their parts. For instance, a knife is a utensil comprised of a handle and a blade. But ‘good’ cannot be simplified in this way; it is irreducible. Moore is writing in opposition to naturalistic methaethicists who attempted to equate good with a property like “fairness” or “justice”, instead claiming that such theorists committed a naturalistic fallacy. However, it is my contention that Moore fails in his criticism precisely because his attack is ill-named and ineffective. It does away with more than his intended target, does not meet the definition of a logical fallacy, and may not even constitute an error in reasoning. These are large and damaging claims to Moore’s thesis, but as I will argue they are entirely valid. To begin with I will show how this naturalistic fallacy is as applicable to non-natural or supernatural properties. Then I shall show that there are definitely valid logical arguments that still fall foul of this supposed fallacy. Finally I shall argue that the naturalistic fallacy, even redefined as a sometimes-naturalistic-sometimes-non-naturalistic non-fallacy, may not even be a simple mistake, and that identification of ‘good’ with a natural property is defensible and reasonable.

Ethical investigation may be defined as the identification of that shared quality held by all things that we consider good, and so a useful starting point is to understand what ‘good’ is and then go searching for examples of those objects that have this property. This property may be something natural that is a feature of the object or it may be something more metaphysical that bears some connection. It may be something simple or it may be something more complicated. For Moore ‘good’ is both non-natural and simple, and is thus indefinable. The mistake that naturalist ethicists make is equating ‘good’ with whatever natural shared property they identify in good things. ‘Good’ is seen as analogous to colour because, though we can state that a certain wavelength of light is responsible for our perception of, say, yellow, the light itself is not yellow. The physical properties associated with ‘good’ are incapable of fully defining it. There is something else going on here. A useful, if lamentable, modern example of the naturalistic fallacy may be found in the current debate over homosexuality as a socially acceptable lifestyle – those opposed to homosexuality will state that since it is not ‘normal’ it is not ‘good’. This holds ‘normal’ and ‘good’ to be synonymous, and so commits the naturalistic fallacy.

Moore supports his contention in a number of ways, but in the interests of fairness I shall attempt to present his strongest arguments, beginning with the “Open Question argument”. This states that any natural property offered as a definition of ‘good’, for instance, “”Good’ is what is pleasurable,” begs a pertinent question: “Is it good?” A definition that leaves unanswered questions is by that very fact a bad definition. This is equivalent to saying that, “‘Good’ is what is pleasurable because pleasure is good.” A circular definition is of use to no-one because it contains only the same information restated. One might as well explain that, “I drank some coffee because coffee is what I drank.” According to Moore there is a further problem with this identification with ‘good’ with a natural property, namely that it takes discussions about ‘good’ outside the realm of philosophy. There is nothing inherently wrong with this, as indeed science and other fields are beginning to make inroads into ethical discourse, but for a philosopher it is slightly embarrassing. If one philosopher states, “‘Good’ is what is pleasurable,” and another states that “‘Good’ is the object of desire,” then the disagreement between the two ceases to be philosophical. It becomes psychological as soon as the first philosopher attempts to prove that the object of desire is not pleasurable. It is rendered semantic when the two start to quibble over whether ‘object of desire’ and ‘pleasure’ are equivalent terms. Both of these may be fascinating debates to have, but in neither can philosophy be of essential use.

Moore is attempting to to argue by the elimination of the competition. “When you have eliminated the impossible, whatever remains, however improbable, must be the truth.” Thus if it is a fallacy to identify ‘good’ with any natural property, it must necessarily be a non-natural property. Because of this Moore’s fallacy becomes useless if we are able to show that it is equally applicable to non-natural properties. Non-natural properties are of various kinds. They may be supernatural, concerning the operations and attributes of things beyond the physical world, or they may be conceptual, and exist as ideas without any material form. Given these options it becomes very simple to render an instance of the naturalistic fallacy occurring in a non-natural definition of ‘good’. We might say that “‘Good’ is consistent with the will of God.” The will of God is emphatically not a natural property and yet it is perfectly coherent to ask, “Is what is consistent with the will of God good?” W. K. Frankena argues that the underlying fallacy is not ‘naturalistic’, but rather ‘definist’. This simply means that it is fallacious to substitute one property for another and treat the two as being the same when they are not. Unlike in Moore’s claimed fallacy, a definist fallacy may be committed without even leaving the sphere of ethical properties. To identify the ethical property ‘rightness’ with the ethical property of ‘goodness’ still commits this logical misstep. Given that those guilty of this fallacious reasoning are not exclusively naturalists, it undermines the ability of non-naturalists to use this for specifically opposing ethical naturalism.

To state that an error is a fallacy is not just to say it is wrong; it is to say that it is wrong in a very particular way. Constructing a logical argument using one of these faulty components renders such an argument invalid. A brief aside on the terminology of logical syllogisms. A logical argument is valid if its conclusion follows necessarily from its premises. It need not be true or even particularly convincing, it need only be consistent. For instance:

Premise 1: All mammals are cats.

Premise 2: I am a mammal.

Conclusion: I am a cat.

This is clearly absurd; despite the ubiquity of cats on the internet, very few of them are writing blog posts or communicating in English. However, if we provisionally allow that the premises are true, the conclusion is inescapable. Thus the argument is valid. The second classification that may be applied to a logical argument is that it is sound. In this case a syllogism is not only valid, but the premises are known to be true and therefore so is the conclusion. For instance:

P1: All humans are mammals.

P2: I am a human.

C: I am a mammal.

With these clarifications we can move on. In order for the naturalistic fallacy to be a true fallacy, its use must render a logical argument invalid. This is certainly true in some cases, as can be seen in the following:

P1: Pleasure is sought by all people.

C: Therefore what is pleasurable is good.

The conclusion does not follow necessarily from the premises and so the argument is invalid. But instead of declaring the victory of the naturalistic fallacy it is possible to alter the argument in order to make it valid. Again I cite Frankena, who notes that the above argument (incidentally the Epicurean argument for hedonism, for those who were curious) contains a suppressed middle premise that can be brought to light in order to treat the argument properly. Fully stated the argument becomes:

P1: Pleasure is sought by all people.

P2: What is sought by all people is good.

C: Therefore pleasure is good.

Clearly this form of the argument is valid, if not conclusively sound, and yet it still contains the fallacy as stated by Moore; identifying the natural property ‘being sought by all people’ with the ethical property ‘good’. Validity assured, the only remaining debate surrounding this argument involves the truth or falsehood of the premises themselves, particularly the second premise. This relegates the ‘naturalistic fallacy’ to the status of ‘possible mistake’, dependent upon whether a particular theorist regards ‘good’ as natural or definable.

In this sorry state it seems cruel to subject the naturalistic fallacy to further assaults, but it is possible to argue that it does not even constitute a mistake. Before we can accuse any move in reasoning to be fallacious, we must show that it is wrong at all. Certainly it is not self-evidently absurd to identify ‘good’ with any particular kind of property, and so claiming that someone has committed the naturalistic fallacy seems to veer into a fallacy of its own, begging the question. Contrary to popular use, this is a more recognisable logical fallacy wherein your interlocutor assumes the truth of their conclusion, and then uses this in order to defend their conclusion. “I know that the pixies exist, because the King of the Pixies told me so.” So rather than stating that the naturalistic fallacy is real and then weaponising it against ethical naturalists, first one must defend it individually before it can be utilised in this way. The claim at the heart of the fallacy is actually an example of a well-known problem in ethics, brought to prominence by David Hume; the identification of ‘good’, which is a value, with a natural property, which is a fact. This is known as the ‘is-ought problem’, which simply means that stating the way things are cannot lead us to a conclusion on how things ought to be. However if a naturalist can formulate a proposition about ‘good’ without falling foul of the is-ought problem it is reasonable to assume that the naturalistic argument is a failure at the outset. The is-ought problem is highly daunting but not insurmountable. In order to produce such a formulation a naturalist may appeal to goal-oriented behaviour as a factual approach to moral action. An agent A has a particular goal B. This is a natural fact. Action C is the most reasonable way to achieve BThis too is a natural fact. Thus it is also a natural fact that in order for A to achieve Bought to do C. This is because a part of the definition of ‘goal’ is that it is something which the agent strives to make a reality. Applying this example to ‘good’, we can show that the is-ought problem is overcome and the naturalistic fallacy loses further ground.

As previously stated, Moore feared that a naturalistic definition of ‘good’ would carry ethics beyond the reach of philosophy. This only remains a problem if we conceive of ‘good’ as something that exists independently, without any reference to the affairs or thoughts of conscious creatures. I have previously stated that my conception of both good and evil is inextricably linked with the flourishing and suffering of conscious creatures, and so I see nothing wrong with moving the discussion of ‘good’ into other arenas, in this case fields like neuroscience, physiology and psychology. Or if ‘good’ is simply  a label which we use to designate certain agreeable things, it is hardly a mistake to approach it from a semantic point of view. To regard ‘good’ as the sole concern of philosophers, to be jealously guarded like Gollum’s ‘Precious’, is to show a regressive and ugly side to philosophy. Just as questions about the natural world were once the province of natural philosophers, but are now considered by scientists, so ‘good’ may become the concern of non-philosophers. However distasteful, this is not apparently a mistake.

At very least the naturalistic fallacy seems to be poorly named. To claim it as ‘naturalistic’ seems calculated to shape it as a sword against Moore’s naturalistic opponents, when in reality the underlying mistake is equally applicable to all of cases of mistakenly identifying two propositions as one, regardless of the sphere in which that property belongs. The accusation of fallaciousness is also unsustainable insofar as it is clearly possible for this apparent mistake to be used in perfectly valid logical syllogisms. It is not even clear that what is referred to as the ‘naturalistic fallacy’ is even mistaken at all, as there remains an argument to be had regarding the charge of begging the question. These weaknesses of the naturalistic fallacy, both in name and intent, would seem to consign the term to uselessness as there are better ways to define the possible mistake which do not suffer from the objections raised by Moore’s framing and phrasing of the problem.


Moore, G. E., 1903. Principia Ethica. New York: Prometheus Books.

Tanner, J., 2006. The Naturalistic Fallacy. Richmond Journal of Philosophy, 13.

Prior, A., N., 1949. The Naturalistic Fallacy: The Logic of its Refutation. Logic and the Basis of Ethics. Oxford: Oxford University Press.

Frankena, W. K., 1939. The Naturalistic Fallacy. Mind, New Series, Vol. 48, No. 192 (October, 1939).


Leave a comment

Filed under Ethics, Philosophy

Objective: Morality

When considering morality, questions about the subject are as important as the ethical conclusions reached within moral investigation. This subject, known as metaethics, asks not what is good but what is good. This question seems basal to the whole process; first we figure out what the hell an ethical conclusion is, then we can attempt to identify instances thereof and derive more practical outcomes. Is the good subject to the whims and feelings of the individual or are there murkier, more holistic and impersonal forces at work? Are morals subjective or objective, and what form does such objectivity take? There are some who deem objective moral values both desirable and unthinkable without a god or gods to send down such edicts and give us lowly primates a foundation upon which to stand. As with so many religious arguments this seems to succeed only in pushing back the problem one more step, and then placing great faith in the power of special pleading.

Objective morality, by its very nature, is something which can be assented to by any rational, dispassionate observer. To be subjective is to be open to debate precisely because there is no fact of the matter. I might say that Star Wars is a better franchise than Star Trek – and defend my conclusion to the bloody end – but since there is no demonstrable correct answer this is purely subjective. But were I to say that the first Star Wars film was released in 1977, there can be no disagreement of this kind. All of the available evidence leads to the inescapable conclusion that this is a fact; it is objectively true. Objective morality is of this kind. Objective moral propositions are verifiably accurate therefore are true regardless of your protestations or support. Based on this definition, objective morality must be based upon objective reality, not upon the conviction of a single being. A statement of an objective moral value will end in, “… because facts xy and z are factually accurate to any rational observer.” A statement of a subjective moral value will ultimately end with, “… because I/he/they say so.”

I do not think that it is an exaggeration nor a misrepresentation of the theistic position that all believers regard their god or gods as a source, at least, of moral teaching and guidance. Many go further and state that their deities are the only source of solid and objective morality at all. The mantra is, “If there is no god, there is no foundation for objective moral values.” Such moral guidance as is forthcoming from gods may take the form of precepts in scriptures, vicarious communication through their representatives on Earth, or a conscience urging us towards the good implanted directly into our bodies. This godly morality applies to all people at all times and must not be contravened. It is thus dubbed ‘0bjective’. This is a very different definition of objective morality than the one offered above: No rational or dispassionate person can reject these ethical statements because the guy in charge said so. Rather this divine command theory eerily resembles subjective morality.

In Plato’s dialogue Euthyphro, Socrates confronts the eponymous character over his intention to prosecute his own father for manslaughter. Making the not unreasonable assumption that such a man must be confident in his own piety, Socrates performs his usual duty of relentlessly questioning his interlocutor into submission. His challenge is to request that Euthyphro define piety, which forces this hapless fellow onto the horns of a particularly ill-tempered dilemma. “Is the pious loved by the gods because it is pious? Or is it pious because it is loved by the gods?” If piety is something independent and is only communicated by the gods, deities become irrelevant middle-men between humans and holiness. If piety is simply that which is demanded by the gods, it is subject to the idiosyncrasies of of the mighty and holds no deeper meaning than that. The gods are rendered either irrelevant to piety, or piety is rendered meaningless by the gods. The same principle is exactly applicable to morality. Is the good loved by the gods because it is good? Or is it good because it is loved by the gods?

There are those among the unbelievers that entirely agree with this line of reasoning, and therefore hold that morality is subjective and cannot be defended in an objective manner. In the worst of cases they lapse into moral relativism, a position as repugnant as it is self-contradictory (A moral principle is good if the society deems it so, except for the moral principle of relativism, which holds among all societies). However, there is a way to construct an objective morality that does not depend upon the divine, but only upon those minds that seem to clearly be present: the minds of conscious creatures. On this view, objectivity is not a nebulous, universe-spanning concept, but applies only to a limited sphere, namely those entities which are capable of suffering. If we grant the uncontroversial premise that the best possible flourishing is better than the worst possible suffering, we have all the foundation we need to build an objective moral system. To argue that it is not good or moral to promote flourishing and avoid suffering is, to me, to lose sight of any reasonable concept of ‘good’. In any given situation concerning a human, there will be options as to how to proceed, and among those options there will be some which will cause suffering and some which will cause flourishing. This is simply factual. I can demonstrate that drinking brake fluid will cause me to suffer. Thus, it is objectively true that to drink brake fluid would immoral, insofar as it causes nothing but suffering. Because there is tremendous diversity and uncanny similarity among all conscious creatures on Earth – chauvinistically singled out because we do not as yet have any proof of extraterrestrial life – there will be no simple, sweeping answers to moral questions. But we will always be faced with paths which are more likely to cause suffering and paths which are less likely. The objectivity lies in the demonstrable conclusions and the ability to apply such a method in all cases, if not in practice then certainly in principle.

I once wrote an essay entitled, “Why Moore’s “Naturalistic Fallacy” is Neither Naturalistic Nor Fallacious”. While I hold only a tenuous grip on the conclusions that I reached in that particular composition, while still thinking I was mostly correct, I can report that it is very satisfying as a critic to find that your target fails on multiple levels. This may seem childish, but it is undeniably true in my own case. In the case of objective morality as based in gods’ existence, we find that not only is such morality not objective, it is not moral either. It is based on what some being or beings think, and is thereby subjective. There is no guarantee, and plenty of evidence to the contrary, that the thoughts of such beings as they are reported to us will be beneficial or in any sense good. (Should anyone doubt this, please consult any given holy book.) Rather this theistic morality seems an impediment to building the better and more comprehensive ethical framework towards which we are still struggling. Cast off the notion of divine parental permission, focus our energies on the eradication of suffering and the promotion of flourishing, and it seems clear that we can accomplish yet greater feats of compassion than we have thus far.

[Addendum: The ideas for a godless objective morality are not entirely my own, but have been heavily influenced and inspired by Sam Harris’ The Moral Landscape: How Science Can Determine Human Values. I highly recommend checking it out, as it makes a powerful case for the fusion of ethics and science, and lays out a plan for putting an end to quibbling over moral dilemmas – thus putting me out of a job…]

1 Comment

Filed under Atheism, Ethics, Philosophy, Religion