You are on page 1of 22

De Gruyter

Chapter Title: ONTOLOGY

Book Title: Ontological Investigations


Book Subtitle: An Inquiry into the Categories of Nature, Man and Soceity
Book Author(s): Ingvar Johansson
Published by: De Gruyter. (2004)
Stable URL: https://www.jstor.org/stable/j.ctvbkk4h6.5

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

This book is licensed under a Creative Commons Attribution-NoDerivatives 4.0


International License (CC BY-ND 4.0). To view a copy of this license, visit
https://creativecommons.org/licenses/by-nd/4.0/.

De Gruyter is collaborating with JSTOR to digitize, preserve and extend access to Ontological
Investigations

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
1

ONTOLOGY

This book is a book about the world. I am concerned with


ontology, not merely with language. Many ontological treatises
concentrate largely on the nominalist-realism problem or try to pin
down in minute detail some ontological category. Such undertakings
are necessary, and will be met with below. Equally important,
however, is the need to develop a coherent system of all the most
abstract categories needed to give a true description of the world.
Whether we are nominalists or realists this should be regarded as
one of the central tasks of philosophy. At any rate it is the task of
this book.

1.1 T H E O R I E S O F CATEGORIES
My use of the word 'category' is intended to remind the reader of
Aristotle's famous book Categories. Originally, the Greek word
corresponding to the English 'category' meant 'predicate'. Pre-
dicates describe properties. A category, however, is not just any old
property. Properties can be arranged according to their degree of
abstraction. To take a simple example, the properties light red and
dark red, which seem to be concrete properties, are nevertheless
abstract in relation to all the different shades of colour which are
called 'light red' and 'dark red' respectively. But, of course, taken
together, light red and dark red constitute a more abstract property,
namely red. And this property in turn, together with properties like
blue and yellow, constitutes the property colour, which is more
abstract still. Climbing up the ladder of abstraction, we will at the
end reach the property of just being a quality. If this really is the
end, then we have found a category. In this manner we can also

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

find other categories. By means of predicates like 'chair', 'stone',


'man', 'horse', 'electron', it is possible to arrive at the category of
substance. Concepts like 'longer than' and 'father of' indicate in
the same way the category relation. Aristotle lists the following ten
categories:
Substance
Quantity
Quality
Relation
Place
Date
Posture
State
Action
Passivity 1
T h e difference between 'posture' and 'state', and between 'action'
and 'passivity', might be in need of some explanation. T o be sitting
and to be lying are instances of posture, while having shoes on and
being armed are instances of states. T h e former relate to the body
in itself, the latter take something more into account. To cut is an
instance of action, but being cut is an instance of passivity.
Aristotle's Categories is in my opinion more of a Kategorientafel
(table of categories) than a Kategorienlehre (theory of categories).
This book is intended to be a Kategorienlehre in the strict and narrow
sense, that is a realist theory of categories regarded as real aspects
of being. This is the view of Aristotle, and the view subscribed to
here. In a loose sense of the word, however, a Kategorienlehre can be
extracted from the works of every great philosopher. In the case of
a nominalist like David Hume the table of categories might be:

Impressions
Ideas
Principles of associations:
Resemblance
Contiguity
Causation 2

In order to get such a list from Kant, we simply disregard the


distinctions between the noumenal world, the world of experience,
and the realm of the transcendental. Both his 'forms of intuition'

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

and his 'categories of understanding' can then in a loose sense be


considered categories of the world. We then obtain the following
list:

Forms of intuition:
Space
Time
Categories of understanding:
Quantity
Unity
Plurality
Totality
Quality
Reality
Negation
Limitation
Relation
Inherence and subsistence (substance and property)
Causality and dependence (cause and effect)
Community (reciprocity between agent and patient)
Modality
Possibility - impossibility
Existence — non-existence
Necessity - contingency3

A comparison of the category systems of Aristotle and Kant is


illuminating first of all because it shows that a category system is
not necessarily confined to our most abstract concepts. It might, as
in the case of Kant, also include somewhat less abstract concepts.
For some of Aristotle's categories (substance, action, and passivity)
seem, in Kant's list, to have become subcategories of the category
of relation. Another interesting point is that two of Aristotle's
categories (posture and state) are quite absent from Kant's list.
This is pardy due to the fact that Kant may be said to make a very
sharp distinction between subject and object, a distinction which is
not to be found in Aristotle. Subjects act and are in some sense
free, in contradistinction to objects, which fall under a deterministic
natural scientific causality principle. The category of subject is not
listed by Kant, because according to him, subjects cannot be found
in the world of experience: here determinism reigns supreme.

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

Subjects exist, as K a n t sees it, only in the so-called n o u m e n a l


world, a world about which we can have no knowledge. N o t h i n g
like this is to be found in Aristotle, who in a sense regards all
objects as subjects. Every object obeys a teleological activity
principle.
But j u s t as some of Aristotle's categories are not regarded as
categories by K a n t , so also some of K a n t ' s categories are not
regarded as categories by Aristotle. Especially noteworthy is the
category of being or reality. According to Aristotle, categories are
aspects of being, and so being itself cannot be one category a m o n g
others. F r o m this we can see that a table of categories need not list
all the abstract concepts of the metaphysical system of which it is
part.
A system of categories may allow relations to exist between the
categories, (i.e., relations apart from the relation of simple
differences). K a n t , for instance, holds that there is a specific
relation hidden behind his triplets of subcategories. T h e first two
subcategories taken together are supposed to yield the third.
Totality, to take an example from the category of quality, is to be
regarded as constituted by a synthesis of unity a n d plurality. T h i s
line of thought is taken to its extreme in another famous category
system, t h a t of Hegel in whose Logik all the categories, subcategories,
a n d sub-subcategories unfold in a long chain, which m a y be
represented as follows:

Being (quality —» quantity —» measure) —»


Essence (essence as reflection into self—» appearance—» reality) —*
C o n c e p t (subjectivity —* objectivity —» the idea) 4

K a n t ' s categories, relation, substance, causality, a n d reciprocity, have


in Hegel's system become sub-subcätegories, that is, subcategories
of the subcategory of reality. The arrows are m e a n t to indicate that
the categories derive from each other in both a logical a n d a n
ontological sense. Hegel is a thoroughgoing idealist, so his
categories can merge with reality in a way that is not open to
Aristotle (and not even to K a n t ) . I t is this idealism which makes it
possible for him to treat concept as an ultimate category.
O n e way of obtaining a deeper u n d e r s t a n d i n g of theories of
categories is to look at Gilbert Ryle's concept o f ' c a t e g o r y - m i s t a k e ' .
His example, by now classical, is of a foreigner visiting Oxford,
w h o is shown the colleges, libraries, a n d so on. 5 W h e n the sight-

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

seeing is over the foreigner asks 'But where is the University?'. The
question has as its presupposition the idea that the University is
some kind of building among the others. The man does not
understand that the University is the functional totality of what he
has been shown. In effect he regards universities as instances of the
category of substance instead of instances of the category of
function. Another example of a category-mistake is the question
'What colour has the number five?'. Numbers are not instances of
things, and cannot have sensible qualities such as colours.
A category-mistake represents an absurdity or an a priori
falsehood rather than something which is a posteriori false. If you see
something as a category-mistake, you do not feel in need of
empirical evidence for your view. You just understand that
something is wrong. A category-mistake represents the unthinkable.
If you are to convince someone that he has made a category-
mistake, you must show him that he has to think in another way. A
theory of categories — as a by-product - draws the line around the
thinkable. Such a line, however, should not, many philosophers
notwithstanding (among them Ryle), be regarded as immutable. It
can change, and it has changed through history. In this respect it
behaves like science and common sense. (Such changeability is
discussed in chapter 16.)

1.2 C O N T E M P O R A R Y O N T O L O G I C A L P R O B L E M S
The traditional ontological problems, going back to the philoso-
phers of ancient Greece, are by no means resolved. This book can,
for instance, be seen as an attempt at a resuscitation of Aristotelian
thinking on ontology. But the problem which has been the driving
force leading up to the theory of categories to be presented, is of
very modern origin. In my opinion, the outstanding ontological
problem of today is how to situate society in nature. Modern
physics, together with the theory of evolution has, in an irreversible
way, undermined all idealist interpretations of nature. The
emergence of the social sciences, on the other hand, has given rise
to the idea that society is some kind of'second nature'. The latter
term is explicitly used by Marx, but implicitly presupposed by
most social scientists. As long as ontologists do not take the
problem of the relation between nature and society seriously, I
think we will have to live with an unsolved conflict between those

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

who regard man and society as something distinct from pure


nature and those who, for reasons of science, think that at bottom
everything is actually bundles of elementary particles.
I n order to get a better understanding of nature, man, and
society — and especially of the relations between them — I think we
have to rethink our most abstract categories. Now, of course, many
philosophers even in our time have proposed new categories, a n d
some of these (for example, tendency and intentionality) will have
a prominent part to play in this book. But what is usually missing
is the insight that the time is ripe not merely for modifying or
reintroducing some category, but for creating a new categorial
system.
Some philosophers have been interested principally in the social
sciences and have neglected the natural sciences, or have perhaps
used a preconception of the latter merely as a foil for discussing the
former. Others have taken things the other way round. In my
opinion we have simultaneously to revise both our view of nature
and our view of man and society. O n e has to try to grasp the real
categories needed by both the natural a n d the social sciences, even
if it looks like an undertaking which no m a n can perform alone. It
is today impossible to have a firm grasp even of the essential ideas
in all the sciences, but the attempt to create a new categorial
system must none the less be made.
Traditional theories of categories can be classified along at least
two different dimensions. They are, first, either atomistic or
holistic; and, second, they either contain or lack the category
subject. T h e thesis of ontological atomism is essentially the
proposition that facts are not existentially dependent on each other.
' E a c h item can be the case or not the case while everything else
remains the same', to paraphrase 1.21 of Wittgenstein's Tractatus.6
Ontological atomism is usually accompanied by an epistemological
presupposition to the effect that the understanding of the world as
a totality is secondary to the understanding of its parts. As the
parts can exist independently of each other, so they can be
understood independently of each other; and then it is as if the
understanding of the totality equals the sum of the understanding
of the parts. Holistic views, in contrast, maintain that any
particular item can be what it is only if all other items are what
they are. T h e part is existentially dependent on all other parts. It
cannot disappear or change without something else being changed.

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

Because of this, it is also impossible truly to understand a part


without at the same time grasping everything else, i.e. the totality.
Truth, to cite Hegel, is totality.
An atomistic ontology imposes some categories and outlaws
others. Now, of course, the same is true of holistic ontologies.
Different things become thinkable in different ontologies. The
opposition between atomistic and holistic ontologies is, however, as
has been said, not the only relevant opposition, especially not
when discussing man's place in and relation to both nature and
society. The other great opposition is that between traditions which
allow no place anywhere in science for the subject category, and
those which maintain that the subject category captures precisely
what is specific to the social sciences and humanities. On the one
hand we have the philosophers of the French enlightenment,
nineteenth-century positivism, logical positivism, behaviourism,
and structuralism; on the other we have the Romantic philosophers,
neo-Kantianism, existentialism, and hermeneutic philosophy. Kant
could not think of a subject as inhabiting the world of experience,
and, in consequence, he could find no place for the subject in
science either. Ryle could not think of a subject as a substance. The
hermeneutic philosophers cannot think of a world at all without
subjects.
That the subject category has become enigmatic is easy to
understand, even independently of calling in popular reductionist
views within specific sciences, such as behaviourism within
psychology and epiphenomenalism within neurology. Obviously,
we have two different intuitions tending in opposite directions. Our
presently available knowledge of history makes certain patterns
appear to us irresistibly from changes of fashion to wars and
revolutions. Such patterns cry out for explanation, and even for
deterministic explanation. History does not appear to modern man
to be merely a sum of chance events. At the same time, however,
we all seem to know that behind these patterns with their
multitude of actions there are people - people capable of free
action. It is difficult not to ascribe some kind of freedom to humans
making them differ in kind from at least inanimate nature. But the
subject category, which incorporates this peculiar freedom, seems
impossible to reconcile with the thought of completely lawlike
patterns in history.
The theory of categories to be presented here might be seen as

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

an attempt to surpass both the traditional opposition between


atomism and holism and the opposition between pro-subject and
anti-subject category systems. In neither case am I the first to try,
but I do not know of any philosopher who has taken as his task
finding a true via media between both oppositions at once. 7
I regard the ontological problems just mentioned as analogous to
the problem of motion viewed from the perspective of Zeno. Zeno's
paradoxes were intended to show that although we do see
movement, we cannot think it. And Zeno assumed what most
category systems presuppose: that what cannot be thought cannot
really exist. He thereby reached the conclusion that movement is
an illusion. One of his paradoxes, the so-called Race Course, can
be stated in the following manner. Assume that you intend to walk
from one place, A, to another, B, along a straight line between
them. In order to come to В you have to pass the midpoint
between A and B, but, in order to do this, you have first to pass the
midpoint between A and the aforementioned midpoint. A distance
can be halved infinitely many times. There is no end to the
division, because when you divide a distance, however small, you
will always get new distances. This means that you can construct
infinitely many 'midpoints' between A and B, which, in turn,
means that in order to be able to walk from A to В you have to be
able to traverse infinitely many intervals. According to Zeno,
however, this is impossible. In order to traverse infinitely many
intervals you need an infinitely long time, which, of course, means
that you will never arrive at B. It is impossible to think movement.
Zeno's conclusion is not inevitable. T o see this, one has only to
maintain that Zeno's reasoning power was insufficiently great and
that, in contradistinction to his views, it is possible to think
movement. But in order to have any force, such a claim also has to
show in some detail what the true thinking of movement is like.
This was not done until Newton and Leibniz constructed the
differential and integral calculus, about 2,000 years after Zeno. In
this kind of mathematics, it is possible to handle infinitely small
magnitudes; and Zeno's paradoxes deal with infinitely small
distances. With the differential and integral calculus it became
possible to think, without any contradiction, that a body at a
particular moment can be in a certain place, and, notwithstanding
this, also at exactly the same moment, have a certain velocity.
Moment here literally means point of time. Were it to mean a

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

period of time, however small, then a contradiction would actually


arise. A body cannot during a period of time be both in movement
and at a particular place. In order to be able to think movement
man had to be able to think new categories; a completely new
concept of point had to appear. A real effort was needed; as I said,
it took 2,000 years.
Newton and Leibniz taught us to think what we perceive. As I see it,
this is what we still have to learn with regard to the relations
between nature, man, and society.
From the point of view of the category system here put forward,
the emergence of modern physics and the decline of the
Aristotelian world view had, for the historical and the social
sciences, two unfortunate consequences. Modern physics, first,
removed the subject category from the natural sciences. In a long
tradition that begins with the French Enlightenment this has been
taken to imply that the subject category should be banished from
all the sciences. T o take seriously the category of the subject was
regarded as involving a kind of prescientific and anthropomorphic
way of thinking. Second, modern physics taught us that science, in
order to develop, sometimes has to go beyond the limits of common
sense; Aristotelian physics seems to be closer to the perceptions of
children than is the mechanistic world picture. As the psychologist
Jean Piaget has argued, children seem to pass through an
Aristotelian stage before they internalize our world view. 8
The criticism of common-sense conceptions, however, has been
carried to such extremes that very few philosophers have noted the
importance of a point made above: that scientific progress may
sometimes consist in learning to think what we perceive. This, in
spite of the fact that even the emergence of modern physics
contained an example of just this phenomenon, viz. the develop-
ment of the concept of movement.
Sometimes science and philosophy have to transcend common
sense, sometimes they have to defend it. There is, however, no
foundational transhistorical common sense. Post-Galilean differs
from pre-Galilean common sense. There is an interaction between
science, philosophy, and common sense, but very seldom does this
interaction cancel all conflicts between these perspectives. In
particular, there are conflicts today. I think today's common sense
embraces both a naive realism and the view that there are subjects,
views which are partly in conflict with today's science realistically

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

interpreted, but views which I shall later defend below (see


sections 13.7 and 15.2, respectively). One aim of this book is to
create an ontological system which does justice both to the subject
category and to naive realism without making science a mere
instrument of prediction. In other words, I am trying to effect a
specific interaction between today's philosophy, science, and
common sense. An interaction which I think takes us closer to
truth.
Now some words in anticipation. In order to understand man
and society we shall have to understand a category of intentionality,
but in order to understand intentionality it is necessary to contrast
it with a concept of tendency. This latter concept has been
neglected in modern philosophy because it has been wrongly
assumed that it is excluded from Newtonian physics. There is a
growing literature on both 'intentionality' and 'tendency' into
which my discussion will neatly fall.9 The concepts of intentionality
and tendency are necessary for understanding the concept of
action, but they are not sufficient. One of my main theses is that it
is not possible to obtain a philosophical understanding and
acceptance of our concept of action until we have rethought the
concept of the momentary instant, which plays such a prominent
role in the whole of modern physics. To anticipate once more, I
think the concept of the momentary is ontologically inevitable, and
I also believe that, for many reasons, it has come to swallow up
other equally necessary concepts referring to phenomena extended
in time. The latter kind of concept is necessary for understanding
action, but it is also necessary for a proper understanding of nature
and the natural sciences. In order to understand the part-whole
relation needed in the social sciences, we first have to understand
the kind of categorial part-whole relations presupposed by
physics.
When I talk about modern physics, I am using this expression in
the sense in which it contrasts with medieval or Aristotelian
physics. Thus 'modern physics' includes relativity theory and
quantum mechanics, though it is not confined to these. Speaking
about relativity theory and quantum mechanics might seem
frightening, especially in view of the long discussions of parts and
wholes in quantum mechanics. I should therefore say at once that
in spite of my occupation with the problem of parts and wholes, I
shall in this book totally neglect quantum mechanics and its

10

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

philosophical problems.10 My justification for this is my belief,


argued for later on, that not even the part—whole problem within
Newtonian physics has hitherto been correctly understood, which
means that, like the discussions within the philosophy of the social
sciences, the philosophical discussions related to quantum mechan-
ics take place against a background of false assumptions.
With regard to relativity theory, things are a little different. The
category system I put forward tries to rehabilitate a 'container
concept' of space and time. As many people, scientists and
philosophers alike, take relativity theory to imply the negation of
such a concept, I have to give some arguments for my contrary
view. However, in the first chapters a container conception of space
and time is merely presupposed, and those who accept such a
concept may simply skip those parts where relativity theory is
discussed (parts of chapter 10).

1.3 UNIVERSALS IN RE
The theory of categories to be put forward takes its departure from
a realist conception of universals. More specifically, it is the
Aristotelian conception of universals as universals in re. This
means, to quote David M. Armstrong, that 'Universals are nothing
without particulars. Particulars are nothing without universals'.11
Everything that exists in the world has both a particularity-aspect
and a universality-aspect, and there are no universals outside the
world. In contradistinction to Plato's transcendent realism, this is
an immanent realism.
The view that universals exist in particulars is, to say the least,
not the commonly accepted conception nowadays.12 In spite of
this, however, I do not intend to argue at length for my position.
There is, even in the perennial disputes within philosophy, a
certain sort of development which makes it possible to take a stand
on other people's shoulders. In the case at hand, I am going to rely
on Armstrong's revival of universals in re, that is to say, on the two
volumes of his Universals and Scientific Realism.
As Armstrong rightly points out, the opposite of realism about
universals ought really not to be labelled 'nominalism' but
'particularism'. Realism says that there are universals, particularism
denies this and says that there are only particulars. This denial is
compatible with different versions of what the seeming universality

11

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

consists in. Nominalism in the true sense is one such version.


Armstrong, however, follows tradition and retains the label
'nominalism' but distinguishes between five kinds of nominalism:
predicate nominalism (i.e., true nominalism), concept nominalism,
class nominalism, mereological nominalism, and resemblance
nominalism. I regard all these nominalisms as false, and refer to
Armstrong for detailed arguments. I shall merely make three
comments which I myself regard as fundamental.
First, if one is a materialist and thinks that there is a world
independently of all consciousness, then it seems incredible to
maintain that a property such as extension did not exist until man
created the corresponding predicate or concept. If there is a world
independently of man there must be universals independently of
man.
Second, if all the things which we believe have the property of
extension do not have such a universal property in common, but
are merely brought together by the general predicate or concept
'extension', then such a bringing together must rely on the relation
'falling under the predicate (or concept) extension'. This relation,
however, must then be a universal in its turn. And so the problem
has just moved one step. The nominalists have to face an infinite
regress here.
Third, many attempts to get rid of universals rely on a presumed
fundamental relation of similarity or resemblance. Many nominal-
ists try to define all universals by means of particulars and the
relation of resemblance. Extension, for instance, is then nothing
else than a class of particulars resembling one another. Such a view
contains at least two obvious difficulties. First, 'resemblance' seems
to refer to a universal, and so one has not got rid of all universals;
and if one universal is accepted one must have specific reasons not
to allow more than one universal. Second, resemblance must be
resemblance of something, and these somethings cannot be undiffer-
entiated particulars since then the resemblance is inexplicable.
Resemblance is resemblance of something in some respect - that is,
surely, in some universal aspect.
So much for ordinary nominalism. Since transcendent realism
seems not to be a live option today, I shall pass it by in silence; it
is, however, discussed and criticized in Armstrong's books. But
before taking the existence of at least some universals in re for
granted, yet another position with regard to universals should be

12

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

mentioned; a position which can be easily conflated both with the


kinds of nominalism earlier mentioned as well as with immanent
realism. The position might be called 'the property-instance* or
'the quality-instance point of view'. It is a particularist position,
but the particulars are not here regarded as bare or propertyless
particulars. The particulars are property-instances. 13
According to this last-mentioned view, two instances of exactly
the same colour hue do not have, as I maintain, a universal in
common. They are merely two numerically distinct property-
instances, which means that there is specific particularity without
universality. In immanent realism qualitative identity is as
fundamental as numerical identity, but in the 'quality-instance
point of view' the concept of 'qualitative identity' is a derived
notion. Now, the big problem here, as of course Armstrong has
noted, 14 is to explain how the particulars are to be classed and
sorted. How, in the case of the two colour hue instances can we
find ontological reasons for saying that they have exactly the same
colour hue? There seems to be something universal lurking in the
back of the quality-instance position too.
I want to stress that in one sense I am not in opposition to
'quality-instances'. On the contrary, the conception of universale as
existing in re imply a conception of quality-instances. But my
quality-instances have a universal aspect; they are not mere
particularity. This makes it possible for me to speak about both
universals and instances of universale.
So far, I have only subscribed to a 'minimum conception' of
immanent realism — that there are at least some universal properties,
namely the most specific properties possible. With some reser-
vations, this is Armstrong's position. I myself, however, shall later
on in several respects enlarge the number of different kinds of
universals which are to be found in the world (see below, pages 15,
33, 34).
Having outlined my view on universals, some words are needed
about the relation between universals on the one hand and terms in
our language on the other. Instances of universals can be referred to
by means of proper names, universals in their universality can be
expressed by means of general terms. However, the former fact
does not imply that all proper names refer to some definite
universal-instance, and the latter fact does not imply that all
general terms refer to some definite universal. Those general terms

13

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

that actually do refer to one universal I shall call 'realist general


terms', and those which do not so refer I shall call 'conventional
general terms'.
Let us now look at our ordinary general property terms and ask
ourselves which are realist and which are conventional. Take
'blue'. This term dpes obviously not refer to one universal, but to a
disjunction of universals, i.e. all the different blue hues. To say that
something is blue, is not to describe that thing in terms of one
universal but to say that the thing instantiates one of several
possible universals. T w o instances of the same universal are
qualitatively identical, but two blue things need not be qualitatively
identical. The same argument goes also for terms like 'dark blue',
'medium blue', 'light blue', 'turquoise', etc. It does not take much
time to realize that we actually lack realist general terms for the
colour hues; at least, if we do not bring in the physicists' talk of
colour hues as frequencies of electromagnetic waves. In the latter
case each frequency refers to one universal.
If universals exist in re, independently of man and language, it is
not at all an odd fact that there might be universals without any
corresponding realist general terms. Language has to satisfy many
pragmatic requirements which, from an ontological point of view,
are totally irrelevant. It is not strange that many of even our most
specific property terms are conventional and refer to a disjunction
of universals. Notwithstanding this, I hope the context will make it
clear when I am referring to one specific universal and when I am not.

1.4 D E T E R M I N A T E S AND D E T E R M I N A B L E S
Property concepts can be ordered into levels of abstraction. If there
were concepts referring to the most determinate colour hues some
of these would be subsumed under the concept 'dark blue', some
under the concept 'medium blue', etc. The concepts 'dark blue',
'medium blue', 'light blue', and 'turquoise', in turn, are subsumed
under the concept 'blue'; and 'blue' under 'colour hue'. Another
way of speaking of such abstraction hierarchies is to say that
'colour hue' is a determinable concept for all the subsumed concepts,
which then are determinates for 'colour hue'. This means that 'blue'
is a determinate with respect to 'colour hue' but a determinable
with respect to 'dark blue', 'medium blue', 'light blue', and
'turquoise', and their subsumed concepts.

14

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

The determinable-determinate distinction is only of fundamental


interest in ontology if it applies to universals themselves and not
mere terms, 15 All immanent realists agree that the most specific
properties possible are universals, but diverge on the question
whether there are any determinable universals. Armstrong, for
instance, is of the opinion that there are only determinate
universals, whereas I maintain that there are determinable
universals as well. This view, it must be stressed, does not at all
amount to saying that all general terms are realist terms, that is,
terms which refer to universals. I regard the term 'colour hue' as a
realist term, the terms 'blue', 'green', 'yellow', etc., as conventional
terms. In my opinion 'colour hue' does refer to a determinable
universal, but the latter terms do not. They refer to a disjunction of
determinate universals. Most conventional general terms hover, so
to speak, between realist ones.
I shall put forward some arguments for my thesis by considering
Armstrong's main counterarguments. He has two objections to the
view that 'the property redness has the property being coloured':
In the first place, it is not the way that we talk. We do not say
that redness is coloured. Particulars may be coloured, but redness,
we say, is a colour. Ordinary language does not attribute a
property to redness, but only membership of a class. And as we
have constantly emphasized in this work, membership of a class
does not automatically bestow a property upon the member. In
the same way, triangularity is not said to be shaped. Particulars
are shaped. Triangularity is a shape. It is a member of the class of
shapes. 16
I think Armstrong misrepresents the view that there are deter-
minable universals; at least his account does not fit the view I want
to defend. Redness (i.e., a determinate hue of redness) does not have
the property of being coloured. Both being red and being coloured are
properties which substances have (cf. chapter 3). Substances have
at one and the same time both determinate and determinable
property-instances. On the other hand, a denial that colour is a
property of redness does not imply that to be coloured is merely to
be a member of a class. The determinate-determinable relation is a
quite specific relation which is identical neither with 'being a
property of nor with 'being a member o f . I shall return to this
relation later (in chapter 9), but to get an idea of what is involved,

15

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

one can notice that if there are determinable universale then such
universale cannot be instantiated without some determinate being
instantiated, and likewise the determinates in question cannot
possibly be instantiated without the determinable being instantiated.
What is Armstrong's second objection to determinable universale?
In the second place, not only does redness resemble the other colours
in being a colour, it also differs from them in colour. Redness and
blueness differ as colours. The property in which they are
supposed to resemble each other is the very thing in which they
differ. This seems impossible. Things cannot differ in the respect
in which they are identical. The conclusion must be that the
resemblance of the colours is not a matter of their having a
common property.
Similarly, not only does triangularity resemble the other sorts of
shape, it also differs from them in shape. Triangularity and
circularity differ as shapes. So their resemblance cannot be
constituted by a common property. 17
The kernel of the argument is contained in the sentences, 'The
property in which they are supposed to resemble each other is the
very thing in which they differ. This seems impossible. Things
cannot differ in the respect in which they are identical.' Armstrong
regards the belief in determinable universals as contradictory, and
goes on to say that only absolute idealists, who do regard the world
as ultimately contradictory, can accept this notion of identity-in-
difference. But matters are almost the other way round. The
existence of determinable universals, first, does not entail a
contradictory reality; and, second, provides the best explanation of
the way we speak of similarities and dissimilarities in the world.
What I have said about determinate and determinable universals is
meant to imply that things cannot differ with respect to
determinates if they have qualitatively identical determinates, and
similarly that they cannot differ with respect to determinables if
they have qualitatively identical determinables. 'Things cannot
differ in the respect in which they are identical', as Armstrong
rightly says. But, of course, if the determinates are not identical,
then the things differ with respect to determinates. And this in no
way is in conflict with or contradicts the fact that the things
simultaneously instantiate the same determinable universal.
'Identity-in-difference', if understood correctly, just means identity

16

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

of determinable and difference of determinate, and no contradiction is


involved. Actually, the criticism presupposes what is to be proved,
namely that there are only determinate universals. Only then will
it become impossible to explain in what way a red thing and a blue
thing are identical. They are just different, and so this difference
must do the impossible and explain the presumed identity.
Another modern philosopher who subscribes to universals in re
but restricts his realism to determinates, is Brand Blanshard. He
has an objection which is of a different type from that of
Armstrong:
Not only is it hard to extract an identical genus from such
different species; one cannot even be sure of the species from
which one is supposed to extract it. According to the theory, blue
is blue, and remains identical through all the changing hues of
blue. But where does blue end as it passes through violet and
purple towards red on the one side, and through peacock blue
towards green on the other? If the concept of blue is definite,
there should be no doubt where it is present and where not, but
there is doubt on both sides. Indeed, one could quite well start
with some intermediate quality like purple and include under
one's new genus a set of qualities that had previously been
assigned to blue and red. Thus qualitative universals, if they
exist at all, will be so blurred as to melt away into each other. 18

The general term which Blanshard discusses, 'blue', should also


in my opinion be criticized in the way Blanshard criticizes it. 'Blue'
is a conventional term which refers to a disjunction of universals
capable of being ordered in an ordinal scale. The latter possibility
is a possibility independent of our language. But how the ordinal
scale is divided conceptually, that is, how the colour terms are
delimited, does not tell us anything about the world. The
subsumption of the determinate colours under different terms is
conventional. Blanshard's mistake lies somewhere else. It consists
in taking for granted that if 'blue' is a conventional general term,
then all the determinable terms 'above* it must be conventional
terms too.
Blanshard makes a false assumption. As a matter of fact his own
line of argumentation can be used to show this. According to
Blanshard, if 'blue' refers to a universal it should be 'definite, there
should be no doubt where it is present and where not'. This is not

17

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

true of'blue', but it really is true of'colour'. If'colour hue' is taken


in the phenomenological sense as perceived colour hue, then it is
impossible to find the kind of conventional ingredient which is
obvious with regard to colour terms like 'blue', 'green', etc. T h e
precise limit between blue and green is a limit within a language,
whereas the limit between colour and, for example, extension is a
limit in the world.
Things are, however, just a bit more complicated. For Blanshard
could reply in the following way. Colour hue ought to be regarded
not as a phenomenological property, but as a property in the non-
perceivable part of the world. And then 'colour hue' is obviously
not a term with definite limits. Like 'blue', 'colour hue' represents
a conventional delimitation of a scale. T h e theory of electromagnetic
radiation revealed 'colour hue' to be a conventional term. What
looked like a realist term was shown up, by the advance of science,
to be a conventional term. Both light and heat radiation as well as
radio waves are today regarded as electromagnetic waves of
different wavelengths between 0.4 and 0.8 μπι, heat radiation exists
between 0.8 μτη and 1 mm, and radio waves have longer
wavelengths. The corresponding general terms cut out relatively
conventional parts of the scale of wavelengths wherein all
electromagnetic radiation may be placed. T h e divisions made are
grounded in our experience of the radiation. As a division of the
radiation in itself, it is as conventional as our division of the visible
spectrum.
Such processes of 'conventionalization' occur now and then in
the history of science, and they perhaps deserve more attention
than they get from philosophers of science, but I shall not dwell on
them here. Instead I want to maintain that a process of
conventionalization can never be complete in the sense that all
determinable terms become conventional terms. The claim that
each particular determinable term can in principle turn out to be a
conventional term, is a quite different claim than that all
determinable terms whatsoever could be conventionalized. The
conventionalization of 'colour' presupposes that other determinable
terms are realist general terms. It presupposes, first, a non-conventional
distinction between primary and secondary qualities. T h e conven-
tional term 'colour' refers to a disjunction of primary determinates,
the different wavelengths of light, and so we need a new term
which refers to perceived colour hue. This distinction between
primary and secondary colour cannot be regarded as a distinction

18

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

merely in our language, it must refer to a distinction in the world.


Second, the traditional theory of electromagnetic radiation presup-
poses that electromagnetic phenomena are different in kind from
(for example) gravitational phenomena. The term 'electromagnetic
phenomena' must therefore be regarded as a realist term. If in the
future a field theory is established which unites electromagnetism
and gravitation, that is to say, if 'electromagnetic radiation', were
to be conventionalized then this theory would contain other
determinable terms which should be regarded as realist terms.
The arguments put forward in favour of the existence of
determinable universals supply us with two criteria for the
detection of such universals. First, determinable universals have to
have absolute limits; there must, metaphorically speaking, be real
gaps between different determinables. Between (for example)
volume, shape, and colour there are such gaps, but between blue
and green there are not. Second, determinables delimit their
determinates in such a way that a determinate cannot be
determinate to more than one determinable at a specific level. A
determinate red shade can only have colour as determinable, not
volume or shape. In this sense determinables constitute a kind of
lawlikeness. This is an important point, but it will not become fully
clear until chapter 9.
So far I have only mentioned three determinable universals:
volume, shape, and colour. These are the first determinables with
regard to their species infimae, the most specific determinates possible
of volume-determinates, shape-determinates, and colour-hue-
determinates. But they are not last determinables, the summa genera
or categories. Volume, shape and colour in turn are determinates
of the determinable property. And, as will be argued in chapter 3,
substances are determinates of the determinable quality; the latter
being one of the categories. Qualities as well as the other categories
and subcategories are, it has been claimed, universals in re. There
are absolute limits or gaps between the categories, and they appear
in a kind of law (the 'existential dependencies' discussed in
chapter 9).
Such a realist stand with regard to determinables and categories
means that I intend my category system to be a Kategorienlehre in
the full sense of this word. In the chapters to follow I try to speak
about the world, not - or not only - about our most abstract
concepts or about idealizations and purely fictitious entities. The
latter play a very important role in science, and even more in

19
This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGICAL INVESTIGATIONS

technology and social engineering, but not within ontology.


Ontology is concerned with what has real existence.
In the next chapter I shall introduce the category of ontological
level and explain one characteristic of the present theory of
categories, its irreductive materialism. The other categories,
subcategories, and distinctions needed will then be introduced and
argued for in what I hope is the right pedagogical order.
At the beginning of this chapter I made a distinction between a
Kategorienlehre (theory of categories) and a Kategorientafel (list of
categories). For some chapters it will look as if I were merely
presenting a Tafel, but later on the categories will be connected
with each other. In order to facilitate reading what follows, I shall
at once list the categories to be dealt with, along with some
subcategories. Here they are, in ontological order of dependence in
so far as this can be represented linearly:

Space-time

State of affairs

substance
Quality
property

External relation

Grounded relation

Inertia

Spontaneity

Tendency

presentational

fictional

Figure 1.1

20

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms
ONTOLOGY

This list must at once be supplemented by two comments. First,


there is also the 'category' of existential dependence (chapter 9). I
put 'category' here within scare-quotes because this is a special
kind of relation which connects the other categories, and so is not
really a category. Like 'exists', 'identity', and 'number', the
relation 'is existentially dependent upon' applies to all the
categories; it is transcategorial. Second, the crucial distinction
between exclusive and inclusive qualities (chapter 4) is not a
distinction between different categories either. This distinction
conceptualizes different relational properties which qualities may
have with regard to space and time.
With the help of these categories and 'categories', other less
fundamental categories can be analysed and defined, categories like
ontological level, action, function, pattern, pure gestalt, efficient
causality, machine, organism, subject, and nested intentionality.
And once these are available, it becomes possible to understand
what our intuitive distinctions between nature, man, and society
really amount to.

21

This content downloaded from 189.6.24.94 on Wed, 25 Sep 2019 23:57:49 UTC
All use subject to https://about.jstor.org/terms

You might also like