Strictly speaking, 'doxastic relevance' is no relevance,
and amongst ourselves it is, indeed, superfluous to speak of
"nondoxastic relevance". Yet, in a wider or introductory
context this addition stresses our ontological and epistemological
position that something is not relevant merely because it is
believed to be relevant. We have seen in the Book of Instruments
how the use of the notion of relevance (especially the
phenomenological one) sometimes tends to be excessive and
doxastic, and we rather take a certainty for an uncertainty.
This is not to suggest, of course, that we ourselves can be sure
about what is relevant; actually no-one can. It does mean tho
that we think it very important to apply at least the criterion
of inconsistence, which does away with partial relevancy, the
criterion of the fake focus, which does away with fake focuses
of relevancy and the criterion of mere correlation, which does
away with pseudofactual relevancy. Even if we, or other people,
believe something to be relevant, it is not relevant if,
for example, the focus of relevance is fake. The doxastic
relevance is, then, irrelevance all the same. Nondoxastic
relevance and irrelevance, however, are there, whether believed
in or not.
The question of why the nondoxastic relevance we are concerned
about must be discriminational relevance, has only to
be answered for the role of relevance in the ground-world as our
principle of relevance is a nonpropositional one. Hence, it need
not be explained why relevance-relatedness or semantic relevance
are not interesting here. The fact that we are dealing with a
principle of relevance which, when interpreted, yields a norm,
makes it definitely impossible, too, that the relevance concerned
is statistical or causal, for statistical and causal
relevance are modal conditions, not factual ones. If something
ever should be statistically or causally relevant, then only as
a means to something else. The relevance is, then, merely
instrumental, not a perfective (or perfectively instrumental)
value, which it is supposed to be on the normative principle of
relevance.
The relevance of a nonpropositional principle of relevance
must be value-dependent relevance, and must be classified on
the basis of the type of fundament involved. To classify it on
the basis of the type of terminus involved would either leave
the principle of relevance without content, or would establish a
different value for which the relevance would merely be a means.
'Moral' or 'motivational relevance', for instance, leave a
principle of relevance devoid of any practical meaning even when
the fundament is given, because the question which remains then
is with respect to what moral value or with respect to
what kind of motivation should something be relevant?. On
the other hand, when the terminus is defined in denotative
terms, it is the value of the terminus which is given its
normative significance by the 'principle of relevance', not
relevance itself. For example, when something has to be socially
relevant (or 'of practical relevance to the interests of society
at large'), and the social goal is simultaneously defined as the
greatest happiness of the greatest number of citizens, it is
this particular happiness value which is made the subject of
the relevance principle. The relevance principle tho, holds
independently of the kind of terminus involved, that is,
independently of the focus of relevancy. Setting aside forms of
relevancy which are gradations of value-dependent relevancy on a
decision-theoretical level, such as 'topical' or 'primary'
and 'marginal' or 'minor relevance', the object of the relevance
principle is therefore nothing else than pragmatic or
discriminational relevance.
Pragmatic relevance in the ground-world is itself merely a
type of discriminational relevance, namely the relevance of
distinctions made by the speaker or writer between words,
phrases, statements, and so on. It is the relevance of a speech
act done where an alternative act could have been done, that is,
where a decision between different acts has been taken. It is in
this decision that one or more distinctions are, or have been
drawn. The very need of a principle of relevance was demonstrated
in this field in the first place. When we concluded --in the
Book of Instruments-- that even 'purely descriptive theories'
ought to be true and relevant, this true and this relevant
referred to statements made as part of a theory. But making a
statement is a speech act in which a person distinguishes
one class of things from another, and in which `e decides to
mention some things and not to mention other things. Since the
pragmatic relevance involved in speaking and writing is only a
special kind of discriminational relevance, and since discriminational
relevance is the most extensive form of relevance which
can be the subject of a nonpropositional relevance principle
which does not fix a focal determinant, it is this form we must
be dealing with when talking of such a principle.