The fundament of relevancy is a distinction: a distinction which should not be made if it is irrelevant, but which may --both on the proscriptive and on the prescriptive account-- be made if it is relevant. The distinction is drawn on the basis of a factor which divides the world, or a domain of discourse, into two or more parts. Examples of such divisions are a 'world' consisting, or said to consist, of human and nonhuman beings; or a 'world' said to consist of men and women; or a predicative 'world' with mental health as distinct from mental disability. Sometimes, what is presented as 'one' division of the world may involve several factors of distinction, like when the world is said to consist of human beings, (nonhuman) animal beings, nonanimal living beings (or plants) and nonliving beings. All these differences in themselves, however, are not relevant or irrelevant: humans are there just being human, whereas nonhumans are there just being nonhuman; mental health is there just being mental health, whereas mental disability is there just being mental disability. It is the way in which a difference is taken up in a speech act, or in a nonlingual act, which is pertinent or impertinent. A distinction is not made until a difference is taken up (or, if imaginary, created) and members on one side of the divide are somehow treated differently from members on the other side of the divide. The difference in treatment may, then, be a difference between being mentioned and not being mentioned, between being allowed to enter a certain place and not being allowed to enter that place, between being worshiped and not being worshiped, and so on. Now, if the distinction made, or the difference in treatment, is irrelevant, the relevance principle requires that either all members be mentioned, or that no-one be; that either everyone be allowed to enter a certain place, or that no-one be; that either everyone be worshiped, or that no-one be.

Predicates like being-mentioned and being-allowed-to-enter are noncatenical. Being-worshiped, however, is a special kind of being-honored, and being-honored, being-dishonored and the neutral being-neither-honored-nor-dishonored are concatenated. Let us assume that a situation in which everyone would be worshiped (if possible at all) needs no explanation, but what does a situation in which no-one is worshiped look like from the honor-catenary point of view? We assume then, too, that to worship means to honor greatly, and that all beings which are catenal with respect to the honor-catena are also catenal with respect to the catena of being-honored, and vice versa. In the situation in which no-one is worshiped some honor-catenals might be dishonored, some might be honored, nor dishonored, and others might be honored but not so greatly that it would amount to worship. This implies, however, that in a situation in which no-one is worshiped, honor-catenals could still be treated in very different ways, but also these dissimilarities have to be relevant on the relevance principle. In other words: for catenated predicates it does not suffice to say that all things concerned should either have them or not have them, if any difference in treatment is irrelevant. For catenated predicates the precise formulation is that all things concerned should in such a case have the same catenated proper predicate. Hence, all of them should be worshiped or otherwise honored to the same degree, all of them should be dishonored to the same degree or all of them should be neither honored, nor dishonored.

It is plain, then, that the relevance principle does by no means prescribe that all honor-catenals should be worshiped, honored, dishonored or neither honored, nor dishonored; that is, it does not prescribe any particular predicate of the honor catena (or of the catena of being-honored). Yet, it apparently does prescribe a predicate of the original catena`s difference-catena, namely the neutrality of the honor-difference catena. In the event that a difference in honor-catenary treatment between two honor-catenals is irrelevant, both should have the neutral predicate of the honor-difference catena.

This leaves us with the question of what to think about noncatenical predicates. They do not admit of degrees and an entity either has them, or does not have them. If, and insofar as, these predicates are really noncatenical and do not have any bearing on predicates which are catenated, we need not bother about them from the standpoint of the relevance principle. But --as noted in the Book of Instruments-- all noncatenical predicates do appear to indirectly admit of degrees somehow, or to cause something which does admit of degrees. Even if a predicate being mentioned once would not be catenical in any way whatsoever in a domain in which everything is and can be mentioned only once or not at all, it may be construed as a catenated predicate as soon as there is one thing which is or can be mentioned twice. And again, the relevance principle will then not prescribe that something be mentioned not at all, or once, or twice or more times, but only that all things will be mentioned the same number of times when a difference in treatment is irrelevant.

A living being can be mentioned twice or many more times, but it can be killed only once and therefore, so it seems, we cannot in the same way devise a catena on the basis of the number of times that something can be killed. Killing and being-killed are therefore purely noncatenical --one would say. What the principle of relevance teaches on the intentional level is only that if one living being is killed on purpose and another one not, there should be a relevant difference between the two living beings. It does in no way teach, for example, that living beings must be killed, or for that matter, must not be killed; to conclude this, at least one other principle is needed. Altho killing itself is not catenical, it can have, of course, an enormous impact on the happiness-catenary state of sentient beings thru the suffering and anxiety it causes in the being (possibly) killed and those living with it. There ought to be a relevant difference between two (kinds of) living beings if one is made to suffer more than the other. The relevance principle thus requires in the first instance also neutralness with respect to the happiness-difference catena. (Note that in questions of killing the right to personhood plays a great part, but that there is no likeness between this metadoctrinal consideration and the doctrinal considerations we are involved in here. It is in the chapter on life and nonlife that we will deal with the subject of killing from both a doctrinal and a metadoctrinal perspective.)

In "Equal, unless .." (I.5.1.3) it has been argued that the only two systematic approaches to the burden-of-proof issue in matters of relevancy rest on an equal,unless- or on a different, unless-tenet. It has been demonstrated there that almost all theorists on issues like equality, justice and nondiscrimination speak of "departures from equality" and of "differences having to be justified". Hence, where the relevance principle implies that there should be difference-catenary neutralness when a distinction is irrelevant, and that there may be difference-catenary unneutralness where it is relevant, all these theorists maintain, catenically speaking, that it is the unneutrality in question which has to be justified, not the neutrality. If people were omniscient, it would not matter whether they opted for the equal, unless- or the different, unless-approach, just like equal, unless different and different, unless equal are tautologies which merely signify the same from a purely truth-conditional standpoint. Since people are not omniscient, however, the equal, unless tenet favors neutralness in general and difference-catenary neutralness in particular. A possible exception is the case of purely noncatenical predicates which do not have any bearing on a catenical one, but in such a case the relevance principle does not favor unneutralness either, whether interpreted with equal, unless or different, unless.

The equal, unless tenet favors neutralness, because only when we can be 'sure' that a distinction is relevant may we make it with the ensuing difference-catenary unneutralness. When we are in doubt, however, that a distinction is relevant, we are not allowed to make it. Yet, it could still be that the distinction is relevant nevertheless, and that we maintain neutralness where unneutralness would be allowed (on the proscriptive view) or even prescribed (on the prescriptive view). The equal, unless tenet is therefore neutralistic in a catenical sense in that it normatively puts neutralness above any degree of unneutralness.

When we choose the equal, unless tenet and reject the different, unless one, we do not take this decision because most people adhering to a systematic code have traditionally also implicitly or explicitly preferred the former to the latter one. This fact will only make it easier for us to pursue on our course without having to lose ourselves in an endless dialectic of arguments and counterarguments. Later it will turn out that we actually choose the equal, unless tenet as the basis of our interpretation of the principle of discriminational relevance because it is neutralistic to do so.