In chapter 7 of this book we will discuss several types of normative-philosophical theories. We will then see that many ethical philosophers distinguish consequentialist from deontological (or other) theories. Consequentialism refers to the theory that the ultimate standard of what is right or wrong is a set of one or more nonmotivational values which are brought into being. (Nonmotivational is added here to distinguish them from values such as virtue, praiseworthiness and, perhaps, courage.) Good and evil are defined in terms of these basic values and an act is right, consequentialistically speaking, if, and only if, it (or a rule under which it falls) produces, or is intended to produce, at least as great a future balance of good over evil as any available alternative. Deontology, on the other hand, refers to a theory in which an action or rule is said to have certain intrinsic right-making characteristics other than the value it brings into existence, that is, other than its good or bad consequences. (Deontology, which derives from deon, that is, duty, is also defined in such a way that every ethical theory which is not exclusively consequentialistic would be 'deontological', but this definition is not justifiable, neither from a systematic nor from an etymological point of view.)

It is evident that a system of goal-dependent relevancy has a plainly consequentialistic structure, especially when the focus is a value such as the best quality of work to be done. Yet, this does not mean that the goal(s) aimed at could not be of a deontological nature, or be defined in deontological terms. For example, if a theorist argues that it is a 'morally relevant fact' that someone breaks 'er promise when 'e has entered into a (quasi-)contractual obligation to pay, the fact in question is relevant in respect of a 'deontic' goal, namely keeping one's promise. Moreover, the relevancy is then obviously of the moral type because of the very nature of the goal. As a deontic goal, an obligation or duty can be represented by a variable (with at least two values): 0 for not keeping one's duty, and 1 for keeping it, or vice versa. Some moral philosophers might argue that if at least one goal is of a deontic nature, the total complex of relation and goal(s) --the superstructure with the base on which it is erected-- is not (exclusively) consequentialistic. The hidden premise in this argument is that the descriptions of the actions deontic goals prescribe or forbid, are not themselves value-dependent.

The question is whether any description of reality can ever be given without that description being (believed to be) relevant to a certain purpose. Someone would have to maintain that this purpose is at least once of a deontic nature again. It may also be that the deontologist's description of a certain action or omission immediately depends on one or more nondeontic goals. Take, for example, a party which is going to be attended by a great number of people, and suppose that a particular person has been invited to that party just like so many other people. Now, suppose too that this person has told the host that 'e would come, but that 'e actually does not go, while having no overriding (deontological) excuse for not going. The fact is that because of the large number of people attending, the host will hardly miss 'im, or not at all. There are people --deontologists among them-- who would not call this "a case of breaking a promise", but they obviously refrain from describing the invited person's omission in this way, since 'er not doing what 'e said 'e would do does not harm anybody (or less than 'e would harm 'imself by going). This interest in the absence of any harm involved, or of minimizing the harm involved, is itself a consequentialist consideration tho. This may also be the very reason why in such a situation many people may not even expect someone to necessarily do what 'e says 'e will do.

Both analytical and phenomenological thinkers have emphasized how much the typification and description of reality depend on what the one who does the description believes to be relevant. The phenomenologist, for instance, may write that someone's 'types were formed in the main by others, predecessors or contemporaries' and that the sum-total of the various typifications constitutes a frame of reference in terms of which not only the sociocultural but also the physical world has to be interpreted'. This issue of the relevance of descriptions must play an even greater part with respect to evaluative concepts. One may, then, not simply presume anymore that, for example, (the duty) not to steal is an (ultimately) deontological goal because stealing is as value-laden as property is. As has been pointed out in the previous chapter, the crux of the matter is what kind of taking is, or should be described as, 'stealing'.

We shall not try to answer the question whether relevancy also ultimately remains a consequentialist construct, even in deontological theories. As we have not yet uttered a judgment on the goal(s) (moral or not) one ought to choose, the answer to that question is of little interest here. What is far more interesting at this place is whether relevance is automatically part of a normative system, if it is consequentialist (or insofar as it is consequentialist). One such system is utilitarianism in which the sole goal or ultimate value is (the maximization of) happiness and/or the minimization of unhappiness. Relevancy is, then, definitely not a (typically) utilitarian concept, since it requires in no way happiness or utility to be a focus of relevancy, let alone the only focus. But does a utilitarian system by itself require relevance or --to be more precise-- discriminational relevance with respect to happiness or utility? We will see that if the promotion of utility, or the likelihood of its occurrence is merely a question of the correlation between an action or rule and a change of conditions, a consequentialist system such as utilitarianism can collide with requirements of discriminational relevance even tho utility would be the sole focus of relevancy. This has indeed happened, and particularly violently so when utility is not regarded as the sole focus. To avoid it one must either add an independent principle of discriminational relevance to the system --to say nothing about one or more additional goals-- or incorporate this principle into the definition of consequentialism, especially into the meaning of bringing about or producing a value. The question why correlation is no proof of discriminational relevance in itself is one we will deal with in the next division.