Both descriptive and normative theories (and also speculative or statistical, modal ones) need be coherent in order to be valid, and both depend on a number of axioms which must be accepted without proof. But --it has been contended-- which descriptive theory is 'true' can be empirically ascertained: it is verified if it corresponds with reality (the facts); falsified if it does not correspond with reality. Normative, ethical or moral theories --it is then contended-- cannot be empirically verified or falsified. But this in itself is a trivial truth simply following from what empirical means: capable of being verified or falsified by observation or experiment. The underlying suggestion is tho, that there would be only one true descriptive or factual discipline, whereas all coherent, normative theories would or could be equally 'valid' however contradictory among themselves. In four steps it can be demonstrated that this kind of factualism is in complete error, and that every so-called 'purely descriptive' form of theorizing depends as much on norms or normative choices as an adequate normative discipline.

Firstly, let us agree that there would indeed be solely one true, factual theory corresponding with reality as a 'unified whole', and that all other coherent, factual or descriptive, theories, and all incoherent ones, are false. This, then, would be a fact and ... not more than that! (Strictly speaking, only the true theory's correspondence with reality is a fact, whereas in the case of false theories there is no fact, or the 'fact' is a privative one.) It is not a fact tho, that an utterance or a complex of utterances ought to be true or ought not to be false: that is a norm -- a norm of correspondence to be precise. Thus, as soon as someone assigns a higher value to a theory which has been verified because it has been verified, or which has never been falsified because it has never been falsified, 'e is involved in a normative act or an act expressive of (the belief in) a norm (if it is not to be a mere question of personal preference). The suggestion, however implicit, that a true theory or truth is superior to a false theory or falsehood requires or presupposes a normative principle of truth. Hence, those who stress the need of empirical verification or falsification, and who reject untrue, factual or descriptive theories must also reject any and every normative theory that does not recognize a principle of truth(fulness) even tho that normative theory might be coherent.

Secondly, no descriptive discipline could describe the whole of reality in all its details, and certainly covering law models do not do that. (Isolated factual conditions as such do not explain anything.) Every descriptive theory must confine attention to those aspects of reality it deems relevant. In itself it is simply a fact that some aspects or details are relevant, while others are not. But as soon as someone assigns a higher value to a theory or description which describes a relevant state of affairs than to an irrelevant theory or description, altho both theories or descriptions are true, 'e is involved in a normative act or an act expressive of (the belief in) a norm (if it is not to be a mere question of personal preference). The inherent suggestion is that a relevant description or relevance is superior to an irrelevant description or irrelevance, and this requires or presupposes a normative principle of relevance. Hence, those who demand a true theory or description to be relevant, and who reject irrelevant, factual theories or descriptions must also reject any and every normative theory that does not recognize a principle of relevance even tho that normative theory might be coherent and recognize a principle of truth.

Thirdly, something is only relevant with respect to a certain entity looked upon as a goal, objective or value to be aimed at (as we will see in the chapter on relevancy). To treat an entity (a certain point of a scale, or a certain predicate of a catena, say) as a goal or objective is a normative act or an act expressive of (the belief in) a norm (if it is not to be a mere question of personal preference again). This requires or presupposes a normative principle which determines what the one or more goals are (or should be) with respect to which something is relevant or not. Hence, those demanding a true theory to be relevant with respect to one or more particular goals, and who reject factual theories irrelevant to this goal, or these goals, must also reject any and every normative theory that does not recognize a principle furnishing at least this particular goal, or these particular goals, even tho that normative theory might be coherent and recognize a principle of truth and one of relevance. (An example of such a goal may be the happiness of all happiness-catenal beings.) Of course, it may be that the developer of a 'purely descriptive' discipline will object that choosing goals is not 'er task at all, and that 'e 'has to' leave that task to others. (Note that having to itself is normative.) Yet, without this choice 'e cannot describe anything in a relevant way, and leaving the choice to a particular other person or group of persons is itself a choice which is heavily value-laden. If the factualist actually leaves the choice to the state, 'e chooses the goals of the state; if 'e leaves the choice to private enterprise, 'e chooses the goals of private enterprise; and if 'e leaves the choice to a community of egalitarians, 'e chooses the goals of egalitarianism. Moreover, the factualist's work is not only of normative significance with regard to 'er explicit choice of objectives, or clients, or people under whose orders 'e decides to act, it is also of normative significance in 'er way of classifying, in 'er terminology, in the assessment of results, and so on and so forth.

Fourthly, of two true theories or descriptions which are both relevant with respect to the same goal(s), the simpler theory or description is superior to an unnecessarily complicated theory or description. It may not always be clear which theory or description is simpler in practise but simpler is to be understood as requiring a smaller number of primitive notions and axioms or principles. Applying this same principle of parsimony to normative theorizing means that a coherent normative theory which recognizes a principle of truth, a principle of relevance, and two or more principles furnishing its goals must be rejected too if another coherent normative theory which recognizes a principle of truth and a principle of relevance has only one principle furnishing the same or the same and even more acceptable goals. It does in no way follow from this what that goal should be, or that there is one ultimate goal. It does follow, however, that the original factualist claim that all coherent normative theories would or could be equally 'valid', independently of the existence and adoption of descriptive theories, cannot be sustained.

We have started with drawing a sharp dividing-line between the factual and the normative spheres, between fact and value. Subsequently we have discovered tho, how this distinction becomes blurred more and more when not considering facts and norms as objective givens in themselves, but when considering thought about facts and norms, that is, factual and normative beliefs or disciplines. Facts and norms themselves may be entirely independent of each other, in our theorizing on facts and norms, the normative principles of factual thought may not differ from those of normative thought at all. The question whether the normative principles of factual thought actually recognized or presupposed are really the same as those of normative thought is itself a factual issue we cannot deal with here. Even if ultimately the norms of a 'purely descriptive' discipline and of a normative discipline are necessarily the same, the facts need not be.