1.2.2 |
ON NOMINALISM, PHENOMENALISM
AND THEIR ANTITHESES |
Real existence of predicates or universals, or of all
abstract entities, is denied by nominalists. They do not want
to commit themselves to (the existence of) 'something which all
gray objects have in common', or to any other attribute or
relation. According to another definition of nominalism they
admit solely the existence of particulars and eschew all
reference to nonindividuals such as classes or sets of particulars.
Also numbers as classes of classes do not exist, then, in
the nominalists' eyes. However, this respectable sobriety may
not preclude these same nominalists from recognizing as an individual
any sum of two arbitrary individuals, even if they have
no property in common whatsoever, or even if no general term is
applicable to both of them (while not being applicable to other
possible candidates). Should the legs of your body and the legs
of your table be individuals in such a nominalist 'calculus of
individuals', the sum of one of your legs and one of the table's
legs will be an individual as well.
Nominalists who do not want to recognize the existence of classes or sets
may thus in the end be less sober than nonnominalistic realists who
recognize a limited number of nonindividuals (such as classes of
particulars) as existents, or as values for predicate variables.
Our recognition that abstract entities 'exist' and can be
'things' in a world of their own, not in space and time, is a
kind of realism, and is contradictory to the form of nominalism
founded upon a rejection of all abstract entities. On the
other hand, we will not admit sets of particulars as existing
just because one can theoretically construct such sets. (If a
set exists, then it has one or more attributes, not only
nonattributive members.) Not admitting this kind of
nonindividuals, our system could be called "nominalistic" in this
respect. Perhaps conventionalism would better describe our
position, for we consider the calculus of classes or functions,
or set theory, a 'creation of the mind' good for a convenient
interpretation of the world. But then, we will not adopt the
kind of conventionalism according to which all formal and
scientific theories are nothing else than systems of linguistic
conventions. What complicates matters is that we need to make a
distinction between existence and thingness,
and that we will recognize the existence of relations, but not
their thingness when talking about physical reality (in what will
turn out to be the primary domain of discourse). Traditional
philosophical distinctions like the one between nominalism or
'antirealism' and (nonnominalistic) realism, may just not be
applicable anymore to novel conceptual structures, or at least
not unequivocally. Moreover, there are good reasons to avoid
terms like nominalistic and realistic altogether, because of
the dissimilar meanings they both have been given. The term
realistic is not only opposed to nominalistic or
antirealistic, and not only to unrealistic, but also to
particularistic, in which case it refers to a different
interpretation of phenomenalist systems.
The distinction between phenomenalist and physicalist systems is, then,
itself one in addition to the distinction between nominalism and
(logical, nonnominalistic) realism.
If physical entities such as objects or processes are chosen
as the basic units of an ontological, constructional system,
then it is called "physicalistic"; if phenomenal entities such
as qualia or presentations are chosen, then "phenomenalistic".
(A quale is a property which is considered an object of
experience rather than the physical entity itself which has
that poperty: it is the characteristic presented by that
physical entity.) In both systems the basic units are individuals
which can be perceived by one of the senses. Physicalists
and phenomenalists both claim epistemological priority for their
own basic units, but it has been demonstrated already by
theorists with less absolutist pretensions that it is hard
to understand what such would mean in the first place. Now, if
the basic units of a phenomenalist system are nonconcrete, qualitative
elements, then it has also been called "realistic";
if they are spatiotemporal particulars like phenomenal events,
then "particularistic". Those free from metaphysical, ontological
or epistemological absolutism have also made clear before,
that when choosing between these different systems, it is most
of all the way in which they deal with the relationship between
qualities and particulars which matters (and not so much the
metaphysical priority of either qualities or particulars).
A particularist theory is faced with a problem of abstraction: how
to obtain repeatable, abstract universals from concrete particulars.
But an (antiparticularistic) realist theory is faced with a problem of
concretion: how to obtain unrepeatable, concrete particulars from
abstract qualities.
Attention has already been drawn to the fact that every
perceptible object can be described in two different ways, namely
(1) by the parts it consists of --as such it is conceived
of as a whole of component parts-- and (2) by what it looks
like, what it does, how it relates to other objects, how it is
changing or not, and so forth. The latter thing is done by
mentioning the object's inherent characteristics, its actions,
its relations, and its changes or lack thereof. All these latter
elements of the object, whether 'essential' or accidental, are
attributes and relations, that is, predicates, of the thing in
question in a broad, but strict sense. By speaking of these
elements the object is conceived of as something that has a
collection of attributes, and a number of relations with other
objects. But while every object presents itself in this view as
a whole of component parts and/or by the attributes it has, the
parts themselves are in their turn also presented by their parts
and the attributes (and relations) they have. This pattern can
be pursued until those objects emerge which cannot be described
by mentioning their parts any longer, but only by referring to
their attributes (so far as the internal elements are concerned).
We are assuming, then, that there is an end to this process of
dissection -- an assumption necessary to arrive at the
basic elements of this ontological system: the attributes. On this
construction every object is a more or less complex system of
attributes, a structure of which the 'materials' (the minimal
basic units) are the attributes. (Relations exist between things
and are in this domain of discourse no material of any of the
discrete things they relate to one another.)
Altho an attribute is a
nonphysical entity --and therefore our system not
physicalistic-- it need not be a phenomenal entity either.
It may be or cause single phenomenal colors, sounds, tactual sensations,
and so on, but it may also be a mental or other nonphenomenal (and
nonphysical) entity.
Insofar as our system is phenomenalistic, however, it is also realistic
(in a nonparticularistic sense).
