The question of how to establish or prove that a certain action is right or wrong is used as an argument against objectivism. Objectivists argue that normative propositions have an objective reference in the same way as factual propositions. Objectivism is also used as a synonym of epistemological realism, the theory that reality exists independently of the mind. Our own ontology is definitely realistic in the epistemological sense as we have assumed that there is a reality independent of the mind. It is not objectivistic in that we take norms or normative conditions to be objective entities or qualities such as planets, people or attributes, but in that we say that normative statements can be true or false, just like factual ones, independently of the person making the statement. (The belief that they are true or false is in itself cognitivism, but all objectivism is cognitivistic.) What we thus oppose is the view of subjectivists who claim that judgments such as stealing is wrong are neither true nor false (noncognitivist subjectivism) or else that the utterance only describes the psychological state of the person saying it (cognitivist subjectivism).

A third position between objectivism and subjectivism is the theory that values have a relational status, that they are 'neither exclusively a property of objects or acts nor exclusively created by human beings', and that they thus have both a subjective and an objective aspect. Yet, those who conceive of value in this way think of something like a diamond which is said to be only 'valuable to someone and in some capacity'. This sort of value, however, is instrumental. The relational status cannot hold for ultimate values of a universal nature. For example, if happiness is a value in itself the value relation does not concern the value of happiness as such, but an object which can indeed be valuable to a happiness-catenal in its capacity of something that makes happy or happier. The so-called 'value' of the object is then another type of value than the value happiness and, more importantly, it presupposes happiness as a value in itself. This third position must therefore still rest upon an objectivist premise with regard to at least one ultimate value.

The type of subjectivism which teaches that ethical judgments are neither true nor false but are 'merely expressive of the feelings of those who utter them and evocative of the feelings of those who listen to them' is emotivism. This type of ethical theory is not only subjectivistic but also 'noncognitive' (or 'nondescriptive'). Noncognitivists deny the 'possibility of proving, demonstrating or otherwise establishing that something is good or right' or that people should 'morally act in certain ways or refrain from acting in certain ways'. Some noncognitivists teach that ethical judgments are simply expressions of emotions 'much like ejaculations' (while not saying the same about factual judgments). Carried to an extreme, subjectivism and noncognitivism thus terminate in normative skepticism or degenerate into value-nihilism.

There are at least two reasons why we shall not embrace subjectivism and noncognitivism in general: firstly, because of the impersonal, normative assertions which can plausibly be made as discussed in section 3.2.3 (for a start they concern truth and relevance as values in themselves); and secondly, the fact that normative theorizing ultimately depends on one or more general hypotheses is not what distinguishes it from scientific, factual theorizing. Also science has its postulates, albeit primarily factual instead of normative ones. But for the noncognitivist or subjectivist who feels like desperately sticking to 'er position, we have a similar message as for intuitionists: if your judgment is merely an emotion or kind of wish, let your emotion or kind of wish be the same as our judgment; if not, we will have to talk about it, and then you will have to try to justify your judgment in a rational or valid way. It does not follow that a subjectivist, noncognitivist theory such as emotivism is, or has been, a worthless theory -- on the contrary. Even while holding the objectivist view that normative propositions are true or false and not (necessarily) about the psychology of a particular person, we can admit that uttering a certain normative proposition does express the speaker's attitude, and that the aim of the utterance is to evoke a similar attitude in the listener. What emotivists have also stressed is the difference between the conceptual or descriptive and the evaluative or emotive meanings of words (as explained in 3.3.2). This difference is most useful to distinguish normative judgments which are analytically true (like one should not steal and other intuitionist pet examples) from more informative ones.

Cognitivist theories are either factualist or nonfactualist. Factualist theories (called "definist" by others) claim that should or ought can be defined in terms of is, or that normative values can be derived from facts. They do not recognize the triadic sphericity of reality. (It might be interesting to investigate if and why the modal sphere remains a separate sphere nevertheless.) The best-known kind of factualist theory is (ethical) naturalism, according to which all ethical judgments are not only true or false but also entirely reducible to natural science. Nonfactualist theories (often called "nonnaturalistic") deny that words such as should and good are entirely definable in nonnormative terms. Altho nonnaturalism is used by some as a synonym of intuitionism, we shall understand by 'an intuitionist theory' a 'theory which professes that basic principles and value judgments are intuitive or self-evident'. With our strict separation of the normative from the factual and modal spheres, our own position is also a nonfactualist one, but it is at the same time nonintuitionistic in that we are not willing to depend on more intuition or on more, or less plausible, 'self-evident truths' than in natural science. Science is not primarily to provide us with crucial, factual information about the ground-world as in naturalist theories; science (or philosophy of science) is primarily to provide us with crucial, normative information about disciplinary thought itself.

Opponents of those who have attempted to define ought in terms of is have said that people commit the 'naturalist fallacy' by identifying a normative judgment with a factual one, or by arguing from premises of one logical type (descriptions) to conclusions of another logical type (prescriptions). This accusation has triggered off a whole series of arguments and counterarguments. In this debate neither party seems, or seemed, to realize that logics cannot prove the one or the other position to be the sole correct one, because before anyone is able to judge whether a factualist argument is valid or not in 'the' logical sense, the two parties will first have to agree on an ontological framework in which to express their nonnormative, factual and either their pseudonormative, factual or their 'truly' normative propositions. Every logical 'proof' of a factualist or antifactualist argument therefore begs the question. And not surprisingly, for in the end the two positions amount to exactly the same in practise. Leaving aside the modal sphere, the factualist calls all 'er judgments "factual". But this all is metaphysical --see 1.7.1--: where all judgments are factual, no judgment is. What happens is that the factualist must differentiate 'truly' nonnormative, factual and something like 'pseudonormative', factual judgments. This, however, will make none of the normative issues we raise and none of the normative statements we make less serious or more trivial. To be more clear about the nature of these issues and statements, they might as well be termed "normative" straight away, instead of something like "pseudonormatively factual".

To sum up: objectivists are (if consistent) always cognitivists, but these cognitivist objectivists may be either factualists or nonfactualists; subjectivists are either cognitivists or noncognitivists; cognitivist subjectivists are always nonfactualists. Our own position is objectivistic, cognitivistic and nonfactualistic, but insofar as this is a question of ontology our attitude towards this choice itself is instrumentalistic. For us ontology is a means, not an end in itself. On our account what is is ontologically entirely separate from what probably is/can be, and this again entirely separate from what should be. This does not mean that the belief in what is (not) and can(not) be is not very often determined by the belief in what should (not) be, and vice versa. The belief in what cannot be, for instance, is but too often a mere expression of the lack of belief in what should be.