The question of whether all the correlations suggested between cognitions, affections and conations, and between whole attitudes, really exist; and if so, whether they are causal connections; and if so, which one of both relata is the cause and which one the effect, is itself a problem of empirical science. From a normative standpoint it is not necessary to postulate that all attitudes, and also practises, do indeed lend each other support, that when a person displays, for example, one kind of discriminatory attitude 'e will probably display the other kind as well. Yet, not postulating such correlations is something else than postulating that they do not exist. Even a utilitarian, or other person exclusively interested in the harmfulness or harmlessness of kinds and acts of discrimination must, strictly speaking, consider clusters of kinds and acts of discrimination which are correlated. Whether these 'clusters' are so small that they solely comprise one kind or act of discrimination, or whether one cluster comprises literally all kinds and acts of discrimination is, then, an empirical problem again.

Social scientists have admitted that the concepts of 'consonance' and 'attitudinal' or 'psychological consistence' are vague. A complete clarification of the meanings of these notions has not come forth yet. In spite of this, the researcher has to be able to denote a priori whether a relationship between, say, two cognitive elements is consonant or dissonant. And when speaking of "consistence", 'e has to make clear what it means that the components of individual or social attitudes, that is, cognitions, affects and behavioral tendencies, and also practises or actions themselves, cohere. These attitudinal components, practises and actions are not necessarily incoherent because they are different, even when pertaining to the same object or issue. Sentiments, beliefs, thoughts, and the practical realization of a person's beliefs and thoughts in actions, almost all admit of degrees; that is, almost all of them have intensities which may be equal or unequal. Nevertheless, psychic-social phenomena, such as emotions, volitions, convictions and actions of largely different intensities are still coherent when pertaining to the same object or issue so long as the directions in which they work are the same. Even when they do not pertain to exactly the same object or issue, but when the connection is logically possible, people speak of "logical coherence" without requiring that there be any necessary connection. Hence, with respect to a given set of cognitions, the logical criterion of coherence is basically reductive (or 'negative') in that it only eliminates those cognitions which cannot be logically combined with the other cognitions. It does not require, nor provide a common denominator for those cognitions tho. All it demands is that the combination of different thoughts, sentiments, tendencies and actions not be logically invalid. In this sense coherence is not more than the absence of truth-conditional incoherence.

For a structural foundation an attitudinal system requires a standard of coherence according to which the attitudinal components can be assigned to two or more different ground-world categories. Starting from such a 'positive' criterion, thoughts, sentiments and actions cohere when they belong to the same category, that is, have a specific common feature, while they do not cohere when belonging to different categories. The criterion to be suggested here for structural consistence is the norm of inclusivity itself. On the basis of this norm, ground-world thoughts, beliefs, sentiments, behavioral tendencies and actions are either inclusive, when in accordance with it, or exclusive, when not in accordance with it. It is, then, either because of their consistent inclusiveness (or 'openness') or because of their consistent exclusiveness (or 'closedness') that normatively significant ground-world beliefs show structural consistence; and it is, then, because of the inconsistent combination of inclusive and exclusive beliefs that they are structurally nonconsistent, even tho they may be logically consistent.

With a structural criterion like (the norm of) inclusivity it becomes possible to determine what kind of belief or attitude belongs to the same category as another kind of belief or attitude. (What is meant exactly by kind will have to be pointed out by means of the classification system already mentioned. Such a system is also a prerequisite for any adequate scientific research with respect to attitudinal consonance or consistence.)

The norm of inclusivity holds independently of any factual-modal condition, independently of any empirical fact or correlation. Yet, if the interplay between especially attitudes and practises which are discriminatory could, indeed, be proved to exist, it would for many people (save thoroughly monistic utilitarians) probably be easier to accept that the harmlessness or harmfulness of particular types of such attitudes and practises is not of immediate import. It would, then, need no further explanation that with the endorsement of the requirement of discriminational relevance all forms of making nonrelevant distinctions become equally condemnatory on principle; that is to say, on this principle.