Phenomenology and Ontology PDF

Download as pdf or txt
Download as pdf or txt
You are on page 1of 220

Phenomenology and Ontology

PHAENOMENOLOGICA
COLLECTION PUBLIEE SOUS LE PATRONAGE DES CENTRES
D'ARCHIVES-HUSSERL

37

J. N. MOHANTY

Phenomenology and Ontology

Comite de redaction de la collection;


President: H. L. Van Breda (Lou vain) :
Members: M. Farber (Buffalo), E. Fink (Fribourg en Brisgau),
A. Gurwitsch (New York), J. Hyppolitet (Paris), L. Landgrebe (Cologne),
M. Merleau-Pontyt (Paris), P. Ricoeur (Paris),
K. H. Volkmann-Schluck (Cologne), J. Wahl (Paris);
Secretaire: J. Taminiaux, (Lou vain) .
]. N. MOHANTY

Phenomenology and Ontology

•:
.~
. .

MARTINUS NI]HOFF / DEN HAAG / I970


© I970 by Martinus Nijhoft, The Hague, Netherlands.
Soflcover reprint ofthe hardcover 1st edition 1970
All rights reserved, including the right to translate or to
reproduce this book or parts thereof in any form.

ISBN-13: 978-94-010-3254-4 e-ISBN-13: 978-94-010-3252-0


DOl: 10.1007/978-94-010-3252-0
PREFACE

Most of the essays that follow have originally appeared in


philosophical journals, Indian and Western. They are reprinted
here with the hope that in spite of the wide variety of topics
with which they deal there is nevertheless a certain unity of
treatment. A few major ideas and distinctions run through all
the essays: I need not further single them out here.
For permission to reprint, I have to thank the editors of the
journals and books in which the essays originally appeared.
My former pupil Miss Manjusree Ray has been kind enough to
help me in preparing the book for the press.

J. N. MOHANTY

May, I968, Calcutta


CONTENTS

Preface v
Part One

I. Modes of Givenness 3
II. The Given 12
III. Thought and Action 22
IV. Meaning and Truth-I 30
V. Meaning and Truth-II 50
VI. Language and Reality 60
VII. On Reference 72
VIII. Remarks on the Content Theory 84
IX. Phenomenology and Ontology 92

Part Two
X. A Note on Modern Nominalism I07
XI. A recent Criticism of the Foundations of Nicolai
Hartmann's Ontology II5
XII. Remarks on Nicolai Hartmann's Modal Doctrine 129
XIII. The 'Object' in Edmund Husserl's Phenomenology 138
XIV. Individual Fact and Essence in Edmund Husserl's
Philosophy 152
XV. Gilbert Ryle's Criticisms of the Concept of Conscious-
ness 163
XVI. On G. E. Moore's Defence of Common sense 170

Part Three
XVII. Reflections on the N yaya theory of A vayavipratyak$a 183
XVIII. Nyaya Theory of Doubt 198
PART ONE
CHAPTER I

MODES OF GIVENNESS*

The essence as well as the strength of positivism lies in its


emphasis on the given as opposed to the constructed. However,
in its zeal in favour of the given, positivism - in some of its
modern forms - has taken recourse to constructions and reduc-
tions. The purpose of this paper is to argue out the case for a po-
sitivism which refuses to construct and to reduce and which, in-
stead of dogmatically clinging to one favoured mode of givenness,
recognises a hierarchy of such modes. This type of positivism may
be named 'higher positivism'; it is opposed to 'lower' positivism
which takes recourse to constructions and reductions and which
admits some sort of sensuous experience as the only mode of
givenness.
One may profitably be inspired by analogies, and nothing
more than analogies, from otherfields of metaphysical philosophy.
Philosophers have sought to distinguish between lower and high-
er naturalism. Seeking to draw this distinction, the late Pro-
fessor Pringle-Pattison in his Gifford Lectures on The Idea of
God laid down that the principle of continuity is not incompatible
with emergence of real difference. Whereas lower naturalism
reduces all phenomena to a dead level of uniformity of the
material order, higher naturalism admits continuity of process
without losing sight of real qualitative difference where the same
appears. This latter insight had led, in the theory of evolution,
to emergentism as opposed to the repetitive theory.
Although the area of the problem with which the present paper
is concerned is quite unrelated to the above, one can neverthe-
less draw inspiration from it. It is necessary that the truth
• Presented to Professor Josef Konig, Gottingen, on his sixtieth birthday, and
later published in Archiv jur Philosophie, Stuttgart, I958.
4 MODES OF GIVENNESS

underlying positivism be recognised without at the same time


losing sight of the aberrations it is suffering from. Positivism, to
be revitalised, must be made more comprehensive, less dogmatic
in its persuasion and more open to novelties. The attachment to
the given has to be retained while the prejudice in favour of the
sensuously given has to be overcome.

2. Starting with the sense-data as the only given elements in


experience, positivism has been struggling with the problem of
things on the one hand and that of universals on the other. Each
of these problems is two-faced: it is linguistic and ontological in
one. The ontological problem of the constitution of things as also
of the reality of the universals is inextricably blended with the
linguistic problem of ascertaining the meanings of the corres-
ponding words of the language. Modern positivism has here
favoured the reduction of things (and of universals) into sense-
data. 'Reduction into' is only the reverse side of 'construction out
of'. Things are constructions. Whether this finding has to be
interpreted ontologically or linguistically is difficult to decide.
No doubt, the linguistic issue dominates. We are given the hope
that all thing-words could be eliminated, yielding place to names
of sense-particulars. This involves a rule of translation. Under-
lying all this is the idea of an ideal language in which every word
shall be a 'logically' proper name designating immediately
given data of acquaintance. But the entire significance of the
theory cannot be limited to the linguistic issue: the ontological
bearing is undeniably there. Things are nothing but families of
sense-data. The epistemological basis is also only too obvious:
sensation is the only mode of givenness, the rest is construction,
arrangement and interpretation. Or, perhaps better, nothing
else is there. It is not surprising therefore that positivism should
thus be an ally of constructionism and formalism, for these latter
are supposed to yield the tool with which what positivism fails
to admit has to be made good. Things and universals have to be
constructed formally out of the material supplied by the senses.
Russell makes use of the principle of Occam's razor to defend
his constructionism. This means also a reformulation of the prin-
ciple. The principle now runs: whenever possible, logical construc-
tions are to be substituted for inferred entities. Reality of the
MODES OF GIVENNESS 5
so-called thing could only be established, if at all, by an inference
to the effect that sensible qualities require either a support
wherein they inhere, or a cause whose effects they are. Such
inference from the known to the unknown is here sought to be
rejected in favour of constructing the unknown in terms of
the known. Instead of an unknown support or cause of the
sensible appearances we have a 'family' of them.
Difficulties of this theory of construction have been pointed
out by Stebbing1 and Black 2 amongst others. Stebbing has in-
sisted that on this theory every proposition about a material
object, say, this table, is reducible to a number of hypothetical
facts about my future experience of the sort "If I lean on this
table, it won't rise into the air, and so on". As a consequence,
every categorical statement about things is in fact hypothetical,
involving a number of verifiable predictions. This is however
untenable for the reason that though when I know I perceive
this table I also know certain hypothetical facts, yet the two
knowledges are not for that reason identical. Black has rightly
drawn our attention to the fact that Russell is perhaps playing
on an equivocation with regard to the term 'meaning'. Either,
the thing-words (that are not, according to Russell, logically
proper names) do mean something for us if we are not acquainted
with them in Russell's sense, or a new sense of 'meaning' is
introduced. In the former case, Russell is accepting the common
usage of the term; in the latter case, his argument could be shown
to involve petitio principii. Further considering the issue from the
linguistic point of view, the symbols that are claimed to be com-
pletely translatable on the lines proposed by the theory must be
demonstrated to be thoroughly dispensable which they, in fact,
are not. Complete translation of the symbols in question is not
possible. Ontologically, we may insist that the unity of a thing
is totally different from the unity of a family, a series or a nexus
of data. Epistemologically, it could be added that things are
given in a way radically different from that in which sense-data
are given.
The application of Occam's razor as suggested by this theory

1 L. s. Stebbing, "Logical Positivism and Analysis," Proceedings of the British


Academy, Vol. XIX, London.
2 M. Black, Language and Philosophy, Ithaca, Cornell University Press, I949, Ch. V.
6 MODES OF GIVENNESS

is tempting because the razor is claimed to be rescuing us from a


hazardous inference. The danger is imaginary, the rescue is
therefore superfluous. The things are not inferred; they are given
as such, prior to their analysis into the data. The application of
Occam's razor is carried out against a false enemy. What is given
as such needs no replacement.
The fact is that things are not given in that theoretical attitude
in which the sense-data are given. Taking therefore the attitude
of reflection if I approach the things, they fall asunder into those
sensory data that now lay claim to be the absolute firsts; the
things either recede to the background as unknown inferred en-
tities or else they dissolve into the rarefied logical constructions.
The peculiar mode of givenness of things is missed thereby.
The sense-data approach to things is destined to end in failure.
Things as such are given in a pre-theoretical attitude. They
qua things are given in practical relationship with the environ-
ment through affective-emotional experiences, through action
and practice. Action is a mode of disclosure. It discloses reality
as things. The idea of things is inseparable from the idea of prac-
tice. In fact, the world of things is a world of practical relation-
ships, actual and possible, a horizon of actual and possible activity.
On the other hand, try to understand thing-perception in terms
of the theoretical subject, and one is led to that insoluble impasse
which ends in the very denial of things and which, not even
satisfied with this, goes to the extent of denying that things
are at all given.
Theory should know its limits. It should recognise what lies
beyond its frontiers. The primacy and irreducibility of the dis-
tinction between the theoretical and the practica1 3 has to be
recognised. This recognition saves philosophy from much in-
tellectual embarrassment. What however is most often lost
sight of is that practice is no less a mode of disclosure than theory.
It is usual to see in practice a process of making and remaking,
modifying, changing, planning, calculating, foreseeing, utilising
etc .. But nevertheless through all these the world of things as
well as the thinghood of things comes to be revealed. To a purely

3 This distinction was brought home to the present author by the lectures of
Professor ]. Konig at Gottingen during the years 1953-54. The present use of it
however is the present author's own.
MODES OF GIVENNESS 7
theoretical subject there would be no things, but only sense-data,
essences, and the subjective. Given the thing, it can of course be
analysed, through subsequent reflection, into sense-data; the
temptation therefore to recompose that original unity out of
these data is natural for the theoretician. One has only to re-
member that, to avoid this temptation, that which is presupposed
in analysis cannot be got back through analysis. The unity of the
thing as revealed in its original mode of givenness through
practice is radically different from that unity - be it that of
'family', series, or of nexus which analysis seeks to substitute for
it.

3. The theoretical attitude - presupposing, and supervening


upon, the practical - moves through three stages: intuition,
thought and introspection. Reflection upon things as given through
practice discloses sense-data. Reflection upon the data yields
essences. Introspection is a turning away from the objective
attitude of the preceding stages. Each of these by itself is a
separate mode of disclosure, bringing into givenness the sense-
data, essences and the subjective respectively.

3. (r) That the sense-data are given is what lower positivism


starts with. It only errs by taking these data as the absolute
firsts: it misses that pre-theoretical mode of givenness which as
we have insisted, philosophy - in spite of its theoretical nature -
should recognise and seek to incorporate into itself. The sense-
data are given in the most primitive form of theoretical reflection.

3. (2) We pass on next to the controversial question about


'essences'. These are given in what is rather equivocally called
'thought'. Since Kant it is usual to insist on the constructive and
interpretative function of thought at the cost of its intuitive func-
tion. What is given - according to this modern view - is the
particular, the sense-impressions which thought only welds into
judgments. On this view - expressed through Kant's celebrated
definition of understanding as the faculty of judgment - thought
is the subjective source of a priori forms. The a priori is held to be
subjective and formal, any objective and material a priori being
ruled out of possibility. Thought becomes wholly discursive, an
8 MODES OF GIVENNESS

endless process of mediation. Within the limits of sense-ex perience


it yields constructed appearances; outside of this limit it creates
illusions. Nowhere does it have the capacity of presenting or
disclosing reality. No wonder that Kant should be reckoned as
the father of what we have called 'lower positivism', according
to which all other propositions except those that are empirically
verifiable and those that are analytic i.e. tautologies are meaning-
less. Kant, no doubt, made the synthetic a priori the cornerstone
of his own philosophy, but his synthetic a priori remained a
composite of alien elements, material from sense and form from
thought - thought having no matter of its own. It was then
easy to deny the possibility of this union and to split up knowledge
into the sensuous and the analytical-tautological!
As against this conception of thought, we advocate a theory
that claims affiliation to the classical Platonic-Aristotelian-
Nyaya theory of thinking. Thought on this theory is basically a
mode of disclosure. Thought discloses ideas, essences, idealities.
Thought thereby gains an ontological significance. It reveals the
ideal factors and structures inherent in real events and processes.
It discloses being as essence, to use Hegel's mode of expression.

3. (3) What sense-intuition reveals as accidental and fleeting


sense-data, thought transforms into essences, nontemporal and
unchanging. Already in sense-intuition, as a mode of theoretical
consciousness i.e. as divorced from practice, there is freedom from
real existence; sense-intuition can be studied in its essence
through phantasy and imagination (Husserl). If ordinary sense-
perception pretends to give an existent real thing with its qua-
lities, it is because such perception is an admixture of theore-
tical consciousness with practice; the possibility however of
liberating the theoretical from the practical is there. Just as
naive realism commits the error of taking the things as such as
given in theoretical sense-intuition, so does the critical realism
of George Santayana and others commit the opposite error of
taking the essences alone as given. It is not the essences that are
the data of sense-intuition; the essences are data of thought
while sense-intuition apprehends the accidental and particular
sense qualities. It is these latter qualities which thought now
reveals as essences but which were not given as essences in sense-
intuition.
MODES OF GIVENNESS 9
There is a type of philosophical logic according to which if A is
given in one mode of experience but if subsequently A manifests
itself as B, then A is an appearance of B so that it js really B
which formerly appeared as A! In the present case since the data
of intuition reveal themselves to thought as essences, it is really
the essences that - according to this logic - were given to sense-
intuition. A thoroughgoing application of this logic would lead to
a reduction "from above" just as 'lower' positivism is reduction
"from below". For if we apply this logic systematically we shall
be led to the metaphysical thesis that it is really the essences that
appeared as sense-data, and so on. Ultimately, what are given in
the lower modes of experience would be in that case false appea-
rances of what is given in the highest. This is a metaphysical
thesis, for it appeals to a distinction between what is given and
what is really given. And yet it is not a purely metaphysical
distinction; it is a metaphysical distinction which yet pretends
to be phenomenological. For, instead of distinguishing between
what is given and what is, it seeks to distinguish between what
is apparently given and what is really given. This latter dis-
tinction between the apparent and the real is untenable. What
is given is also really given. Phenomenology must avoid both
types of reduction: "from above" as well as "from below";
it must satisfy itself with distinguishing between levels of given-
ness and exhibiting, where possible, the transition from one to
the other.

4. A totally new realm opens out with introspection: the


realm of the subjective. But even here one has to be cautious not
to miss the subtle phenomenological stratifications. One at once
faces the thing as apprehended, the sense data as perceived, the
essences as thought. Being is apprehended as mental content.
As mental contents, things, sense-data and essences are all alike.
Even here one is apt to mistake these data of introspection as
what really appeared as objective. In 3. (3) we have discarded this
logic. Besides, the mental content is not wholly subjective. It
stands mid-way between the subj ective and the objective. The thing
as perceived is the thing itself; but it is also not the thing itself.
It is and yet is not that. Hence the peculiarly evasive character
of the mental content, tending to be and yet refusing to get
IO MODES OF GIVENNESS

identified with the object. Introspection of this primary order is


not purely subjective; it is sUbjective-objective.

4. (I) Introspection par excellence, that is to say, the mode of


givenness of the genuinely subjective begins with the disclosure of
the mental functions, as contrasted with the mental states or
contents. From the thing as apprehended, the sense-data as
perceived, the essences as thought one passes on to the func-
tions of apprehending, perceiving, sensing, imagining and thin-
king. This would yield a phenomenology of the noetic functions,
the best known example of which is found in Kant's Kritik der
reinen Vernunft.

4. (2) K. C. Bhattacharyya distinguishes between the above


two modes of introspection in the following manner: "The dis-
tinguishing of knownness etc. as an abstraction from the object is
what is called psychological introspection. The distinguishing of the
subjective function of knowing etc. as other than this abstrac-
tion of objectivity may also be called introspection and it may
be with greater right". In the latter kind of introspection, "the
functions represent the modes of freedom from the correspon-
ding psychological abstractions". 4
Without claiming to be a correct interpreter of Bhattacharyya' s
rather difficult text, we may distinguish between the two kinds of
introspection thus: the one discloses the object as known, as
apprehended, as perceived, as imagined, as thought etc., while
the other discloses the corresponding functions which on their
part represent "the modes of freedom" from the objective cha-
racter of the first kind of introspection. The two perhaps cor-
respond to the two levels of H usserlian phenomenology: the
noetic and the noematic.

5. We have in the above outlined five modes of givenness;


what are disclosed through these modes are things, sense-data,
essences, mental contents and mental functions. These five modes
fall into two principal groups: the theoretical and the practical.
Things are given practica.lly; the rest is given theoretically.
4 K. C. Bhattacharyya, The Subject as Freedom, The Indian Institute of Philosophy,
Amalner, India, 1930, p. 30.
MODES OF GIVEN NESS II

Amongst these latter, the contents and the functions are said to
be given in introspection. The functions alone however are data
of introspection in the true sense.
These five however do not exhaust all that we are acquainted
with. For there is that immediate self-consciousness which is
neither practical nor theoretical, which in fact - so far as its
own mode of givenness is concerned - transcends that distinction
and yet accompanies both. Each of the other modes of con-
sciousness we have mentioned is an awareness at ... ; self-con-
sciousness however which accompanies all our awareness is not
itself an awareness of ... It would be wrong to hold that this
self-consciousness is an introspective datum. It is an 'enjoyment',
not a 'contemplation' - to use a distinction familiarised by Sa-
muel Alexander.
We have thus three levels of the subjective: the mental con-
tents, the mental functions and self-consciousness. The first has
a curious status, being also interpretable as objective or at least
not distinguishable from the object. The functions are ontolo-
gically subjective but epistemologically still objective, for they
are 'contemplated', i.e., are made objects of awareness; their
awareness is awareness at ... Self-consciousness is both ontolo-
cally and epistemologically subjective; it is not an awareness at ..
CHAPTER II

THE GIVEN*

The purpose of this paper is to suggest a satisfactory account


of the chief modes of givenness. From this point of view, con-
temporary philosophy seems to me to be in a highly unsatis-
factory state: it either totally denies that there is at all any-
thing given or, if it accepts the given, restricts the same to one
favoured kind of objects. Those who deny that there is at all
anything given do so on the supposed ground that all human
knowledge is interpretative, though they differ amongst them-
selves as to the source and the nature of such interpretations:
the source may be either the a priori constitution of the human
mind (Kant), or a metaphysical system (Hegel), or the rules of
the language one uses (Wittgenstein). There are on the other
hand the sense-datum philosophers who base their philosophy on
the notion of the given, but recognise only one mode of givenness,
i.e., sensation, and only one kind of objects that are given, i.e.,
the sense-data. This has led to endless difficulties as to how we
come to perceive physical objects. As against this philosophy,
some belonging to the first sort of persuasion have formulated
the Alternate Language Thesis, according to which since nothing
is absolutely given, what has to be regarded as given depends
upon the linguistic decision that one makes, so that each such
decision commits one to a particular conception of what is given.
In short, nothing according to these last mentioned philosophers
is given, but only anything may be taken as given.
It would seem that both these groups of philosophers are mis-
* Read before the Delhi Philosophical Colloquium, I96z, and subsequently printed
in the Proceedings of the Colloquium, India International Centre, New Delhi, I964.
THE GIVEN 13
taken. The first group of philosophers proceed to construct
conceptual or linguistic systems whose basis remains arbitrary.
The philosophers of the second group are dogmatic in holding on
to one favoured type, and are inevitably led to reductionism,
i.e., to the programme of reducing all other entities to that one
type. The errors of both may be traced to a basic misconception
about what a philosopher could and should do in such an en-
quiry as this. The real task of the philosopher here is not to
find out which things are given and which not, nor even to re-
commend which things should be taken as given, but to bring
out or make explicit, by analysis and phenomenological des-
cription, in the case of those that would ordinarily be regarded
as given, the precise meaning, implications and the modes of
their givenness. That things and persons are given in unreflective
life is beyond doubt, so much so that the word 'given' has its
primary application precisely in those cases. If this application
be rejected and the word suitably defined so as to exclude from
the range of its application whatever one wishes to be excluded,
the word would be transformed into a philosophical jargon and
in that case any philosophical decision would be but an analytic
consequence of one's initial linguistic proposal and would there-
fore be trivial. There is likewise a sense in which it is meaningful
to say that we see colours, hear sounds or taste tastes, and it is
again the task of philosophy to explicate this sense. It is only
after philosophy has explicated the modes of givenness in all
these different cases where we would ordinarily say something is
given, that the final speculative task of assigning a generic mea-
ning to 'givenness' could be attempted.

II

We may profitably begin with pointing out, as a sequel to what


has been said above, what 'to be given' surely does not mean. In
the first place, to be given is not the same as being analytically
the simplest. This has been one of the principal sources of many
philosophical errors. It has been wrongly supposed that what is
complex, structured and dependent could not have been given,
so that it is only the simple, not further analysable quale that is
I4 THE GIVEN

given, the rest being only either inferred or constructed. This


error is partly due to a faulty psychology, and partly due to a
wrong conception of the function of analysis. In the second
place, we have to get rid of the only too commonly held erro-
neous opinion that to be given is to be the content of one's
immediate experience. The notions of 'content' and 'immediacy'
involved are hopelessly ambiguous. Most of those who hold
this view surely mean by 'content' what is a real constituent of
one's experiencing so that to be given would amount to be en-
joyed or felt in a sense analogous to that in which one enjoys
one's pleasure or pain. They however fail to see that the modes
of consciousness to which anything is given are all self-trans-
cending, so that to be given to a consciousness is not the same as
being a constituent of it or possessed by it. The given may well
fall beyond the consciousness. It has again been held that to be
given is to be passively witnessed, contemplated and received -
a notion that may be traced back to Kant's characterisation of
the faculty of intuition as receptive. This is a typically philo-
sopher's error, betraying the philosopher's prejudice in favour of
contemplation as against action. It is forgotten that even our
most elementary perceptions involve bodily movements and
adjustments, not to speak of the more complicated processes of
discovery. As we shall argue in this paper, things and persons
which confront us (and are certainly given, if the word 'given' is
to have any sense at all) - are given not to passive contemplation
but to an acting, manipulating, planning, appreciating and eva-
luating consciousness. The epistemologist's love of certainty
has often led him to suppose further that the given must be
indubitable. The sense datum philosophers have thus argued that
since we may always be in error about physical objects and can
never err in our reports about sense data, it is only the latter
but not certainly the former that could be said to be given. If
to be indubitable means to be free from the very possibility of
doubt, it is incomprehensible why reports about sense data are
so, for certainly the person who reports "this is green" might
very well come to doubt, in a Cartesian mood. "Was that really
green, or did I only think I saw a green patch?" At the moment
when one sees the green patch one may of course be free from
doubt and have a feeling of certainty, but surely it is in no greater
THE GIVEN IS
degree so than at the moment one perceives a physical object.
Certainly, this certainty does not as such exclude the possibility
of subsequent doubt, and surely the fact that one may sub-
sequently doubt does not rule out the possibility that the thing
that claims to be given is really given. We may be in error in
thinking that a certain physical object is given in a particular
case, but we can be so in error only because there are cases of
physical objects being truly given.
There is left yet another misconception to be removed. It
is often supposed that the given must be something self-complete,
depending upon nothing and leading up to nothing else. More
precisely, the experience of having something given to one's
consciousness must be a self-sufficient experience, an absolute
first presupposing no prior experience and not by itself leading up
to any. This variety of atomism vitiates not alone the concep-
tions of the sense datum philosophers, but also the conceptions
of the physical object philosophers. A physical object is supposed
to be given as a self-complete object, as much an atom as a sen-
sory quale, though only complex and structured. Both the sense
datum atomism and the physical object atomism have to be
rejected if we are to have a satisfactory philosophy of the given.

III

I have made it clear in the first section that for me it is un-


questionable that things and persons are given and that the task
of philosophy in this connection could only be to explicate the
meanings, and the modes of such givenness. I may now proceed
to this task in a rather brief and schematic manner within the
limits of this short paper. My main contention is, that physical
objects and persons are given in a mode that may be called the
practical mode. What I mean thereby is that they are primarily
given not through passive contemplation but through actual and
anticipated practical relationships. I am not denying that we
perceive physical objects, but only contending that our per-
ception of physical objects is far from being the contemplative
affair that most philosophers would make it out to be.
With regard to the mode of givenness of physical objects, I
I6 THE GIVEN

would like to concentrate on two points. The first of these may be


developed by way of answering a common objection to the claim
that a physical object is at all given. It is often objected that
when I think I am seeing the yonder tree what I really see is -
if not a patch of variegated colours - certainly only a part of the
tree, this side of the trunk for example. It is surely undeniable
that many of the parts of the tree remain hidden from my view,
and if that be so how could it be maintained that I perceive the
tree at all? In view of the way I have above formulated my task,
I could state the problem posed by the objection in the following
manner: in what sense, could the tree be said to be given, when
some of its parts are surely hidden from my view? Or, how should
my awareness of the tree be related to my awareness of its parts
such that the former awareness should present the tree as a
whole even when some parts of it are not presented to me? It
is to be noted that I am speaking a sUbjective language, for I am
interested not in the tree and its relation to its constituent parts
but in the relation of my awareness of the tree and my aware-
ness of its parts.
It should be pointed out at once that it is not necessary for a
thing to be given that all its parts should also be given, for if that
were so then only simple qualia and relations could be given and
nothing else, for every physical object has its constituent parts,
and if one takes physics seriously, not all the constituents could
after all be given. If we accept this condition, we would only be
drifting towards that mistaken identification of the given with
the simple which has already been rejected.
Phenomenologically, the following seem to be undeniable: the
indispensable condition for a physical object to be given is, that
some of its parts be given. This condition is fulfilled, when the
parts that are given are adequate to awaken a consciousness
intending the object asa whole. Therecannodoubtbe a thought of
a physical object, and that thought also intends the object. But
the consciousness to which the object is given exhibits an in-
tention which is founded upon, but goes beyond, the appre-
hensions of its parts. In fact, in such a case, the intention of the
object as a unity forms the principal intention which comprehends
within it the awarenesses of the parts, and this holds good,
notwithstanding the fact that the latter awarenesses form the
THE GIVEN 17
indispensable condition of the former. A physical object is thus
a unity that is more than the sum of its parts: that would be the
rendering of the above in the material mode of speech.
Furthermore, apprehension of a physical object may give
rise to further intentions in which a larger unity is presented,
and this continuous transition from founding intentions
to founded intentions, presenting larger and larger unities,
is only interrupted by incursions of fresh interests de-
manding a total turn of the attention to other directions or to
other pursuits. What I have been driving at is this: a physical
object is given as the object of a dominating intention that per-
meates and comprehends the awareness of its parts, and this
object again leads up to further intentions, in which still larger
unities are presented. As said in the previous section, a physical
object is not given as an atom; it always emerges from within an
environment and also leads us beyond itself to other objects and
relationships. This was the first point that I wished to emphasise
in connection with our perception of physical objects.
The second point is this, the physical objects as a 'stubborn
matter of fact' (Whitehead) with its "Harte des Realen" (Nicolai
Hartmann) is given primarily through practical relationships.
Heidegger's phrase "Zuhandensein" is apt, except for the fact
that it suggests as if all physical objects were objects for use.
There is no doubt however that it is through action that we
recognise the physical object as such, as something which would
offer us resistance and so needs manipulation, which has its own
laws independent of our plans and so has to be mastered, and so
on. The more we take up a contemplative attitude towards it,
the more does its unity disintegrate into its component abstrac-
tions, qualia and relations, patterns and structures: the unity of
the object gets lost. It can at best be theoretically represented as
a logical construction, or perhaps as a mere jafon de parler.
It is also possible to cultivate another sort of contemplative
attitude towards the physical object, i.e., the aesthetic.
In such an attitude, one is interested in an aesthetic enjoyment of
the essence, the rasa, of an object which, e.g., the painter seeks
to convey in painting a still life. Van Gogh's pair of shoes re-
presents a universal essence, but not any particular pair of shoes
with its unique particularity. It is quite possible in such an at-
I8 THE GIVEN

titude that the sense-data are "abstracted"; the pure quale is an


"aesthetic abstraction" (K. C. Bhattacharyya). The physical
object is given in the primary unreflective practical relationship;
the sense-data are abstractions in the aesthetic-contemplative
attitude and curiously enough the analyst's favourites. In
what sense, and in what mode they also may be said to be given
is a further question beyond the scope of this paper. The alter-
nate Language thesis which speaks of the sense-datum language
and the physical-object language as two alternate languages
fails to recognise the undeniable primacy of the physical object
language.

IV

Practice may be, according to a distinction made by Kant in


his Introduction to the Critique of Judgment, either technically-
practical or morally practical. The physical object is disclosed
through action which intends to use it as means to an end or
which meets resistance from it. The person is disclosed through
moral relationships whose essence, as Kant rightly saw, lay in
acknowledging the other as an end in itself. Just as the physical
object disintegrates into abstractions when made a subject of
contemplation, so does the person tend to degenerate into a thing
as the moral attitude is replaced by the pragmatic, as the end in
itself comes more and more to be treated as a means to an end.
Happily though, the person never becomes quite a thing.
As we have hinted at before, the two conflicting theories of
outer perception could be reconciled (not after the manner of the
Alternate Language Thesis, but) by regarding the physical objects
and the sense-data as being given respectively in two different
attitudes, the practical and the theoretical. Similarly we could
attempt a reconciliation also of the two dominating theories
about the nature of the self. According to one theory, the self
is truly a person, a concrete existent being who is 'in the world'
as much as 'with others'. According to the other theory, the self
is a pure epistemological subject, a passive witness of the world
and the others, detached and contemplative. My contention
here is that the person and the subject are given in two different
THE GIVEN

attitudes: the former in living, moral, evaluating relationship


with others, the latter through contemplation on the implications
of the attitude of pure understanding.
In order to bring home the principle on which the suggested
reconciliation has to be carried out, I would begin by introducing
a distinction between what I would call 'understanding' and
'knowing'. Both that there is a distinction and the point of the
distinction would be clear from the possibility that one may
simply understand the sentence "s is p" without knowing that
S is p. Knowing that S is p however presupposes an understanding
of the sentence "s is p", knowing implies that the fact known
should be given, and not merely thought of. The distinction no
doubt is an ideal distinction, for there are cases where the dis-
tinction is blurred. The more contemplative a knowing is, the
more does it approximate towards the attitude of understanding.
The more does knowledge succeed in making unique reference,
the more does it approximate towards the ideal of knowing.
And the knowledge of the unique individual - whether of a phy-
sical object, or of a person - is far from being a contemplative
affair: this exactly is the paradigm case of knowing.
Now, one may reflect on the attitude of pure understanding.
Such reflection may be a reflection upon the objects of pure
understanding: or, it may be a reflection upon the subjective
aspect of it. The former kind of reflection leads to the discovery
of abstract entities like meanings; the latter reflection leads to
the discovery of the detached, contemplative, witness self as the
who of understanding and as distinguished from the person who
is the agent of concrete knowing. As the objects of pure under-
standing are discovered to be impersonal and outside the con-
text of practical-emotional-volitional relationship, so also the
who of this attitude is progressively disclosed to be a passive
witness and not an active person.
Further, I recognise myself as a person in so far as I recog-
nise the Others as persons as well, so that a plurality of persons -
each an end in itself is a not further reducible phenomenological
datum in this attitude: the others remain Others to me, and yet
ends in themselves.
But I cannot, in the sort of theoretic contemplation men-
tioned above, realise the others as subjects as I would in my own
20 THE GIVEN

case. The enjoyment of subjectivity is thus primarily egoistic,


whereas the enjoyment of personality is primarily social. We
need not discuss here how, in the subjective attitude, the initial
ego-centricity may be transcended.
However, the person and the subject are given in two different
attitudes. Different spiritual ideals, even different religious
faiths, may be traced to these two attitudes. But a discussion of
these also falls beyond the scope of the present paper.

v
I have briefly sketched two radically different modes of givenness,
the practical and the theoretical. In the former, the external world
of physical objects and the community of persons are given:
there are correspondingly two subdivisions of the practical mode
for whom I have suggested the Kantian designations 'technically-
practical' and 'morally practical'. In the latter, sense-data,
abstract entities, and the passive witness Self are given, and
there would be corresponding subdivisions of the theoretical
mode. The resulting scheme has no claim to exhaustiveness. The
main distinction between the practical and the theoretical
however may claim to be a useful and illuminating explanatory
principle. It would however lose its value if attempts are made to
bridge the gulf separating the two. Using a metaphor, I should
say there is no continuous passage from the one to the other: a
"jump", a radical change of attitude, is indispensable.
This leads me to formulate what I consider to be a cardinal
principle of all phenomenological philosophy. I shall call this
principle, "the principle of phenomenological discontinuity".
It states that phenomenology should recognise, wherever it
comes across, radical discontinuities amongst phenomena and
should not seek to blur the distinctions out of a sheer system-
building interest.
Keeping this in mind, we may say that no intelligible relation
can be formulated between (r) physical objects and sense-data
and (2) the person and the subject. With this, many traditional
philosophic discussions are rendered pointless, for the things
which have been sought to be related are given in two totally
THE GIVEN 2I

heterogeneous modes. They cannot be given "together", and hen-


ce cannot be related.
There is surely a difference between the two cases. The sense-
data, though "aesthetic abstractions" from the physical object,
are yet appearances of it; they are its. The subject is not in a like
manner an appearance of the person. Nor is the subject either a
real constituent or substratum of the person. The person is also
a subject. It is the same 'I' who is both. The T is thus ambi-
guous.
The wider and interesting task of ascertaining a generic mea-
ning of givenness would not be attempted in this paper.
CHAPTER III

THOUGHT AND ACTION*

I. Two questions, different from each other, are suggested by


the title of the present symposium. First, how is the distinction
between thought and action to be drawn? Next, how are the two
related, as they certainly are? A satisfactory discussion of the
second question requires that we have already some clear idea
of the answer to the first question. We therefore have to start
with the first.

2. Before proceeding with this, let us however take note of


some ambiguities in the use of the term 'action' and clarify our
position with regard to them. There is, to start with, a dis-
tinction between moral action and non-moral action, and any
satisfactory philosophy of action must take this distinction into
account. Further, we are to exclude all non-human actions,
examples of which are bodily (physiological) actions and reactions
as well as natural processes (those implied, e.g., in such state-
ments as "the river erodes the bank" etc.). Both moral and non-
moral actions are here contrasted with non-human actions. We
shall consider only such actions as are accompanied by some-
one's consciousness of being the doer.

2.I. The distinction between moral and non-moral actions is


not however exhaustive, for there are other types of human action
that do not fall within either of the two. There are for example
religious and aesthetic (creative and appreciative) activities.
Karl Jaspers suggests a classification of human actions that ap-

* Presented at a symposium on the occasion of the 32nd session of the Indian


Philosophical Congress, Srinagar, 1957. Originally published in the Proceedings of
the Congress.
THOUGHT AND ACTION 23
pears to be sufficiently comprehensive - the classification into
those that are conditioned and those that are unconditioned.
Even if it may be difficult to justify the names given to these two
groups of human actions, it seems, in any case, to be obvious that
there subsists a radical difference between the two groups of
actions. They are not merely different kinds of actions; they are
also as actions different. Unconditioned are the moral, aesthetic
and religious actions and, may be, many more.

3. The distinction between thought and action is sought to be


obliterated in various ways. One of these is only too common.
Thought, it is said, is nothing but thinking, and thinking is a
mental action, and so generically the same as other kinds of
action. First, this view misses the essence of thinking as such,
for this essence lies not in thought's being a temporal mental
act, but in something else which we have to look for. Secondly,
for the purpose of specifying the topic of this symposium, we
prefer to exclude from consideration that diluted sense in which
thought is an action.

3.1. The pragmatist's contention is different. He is not insis-


ting on the triviality that thinking is a mental act, but is rather
pointing towards the practical character of all so-called theoreti-
cal activity. All so-called theory serves the purpose of practice.
The distinction between the theoretical and the practical is a
distinction that falls within the practical. Thought is practical
in a twofold sense; it arises out of a practical situation, and again
it is a tool, an instrument, to get over that situation.
Much of it may be granted. It may be granted that much of
our thinking in daily life as well as in the sciences arises out of
some practical situation, and further that much of it aims at
solving some practical problem. But at the same time we are also
to grant that much of our thinking, in fact, the so-called higher,
abstract or pure thinking is free from the urgency of any prac-
tical situation. Certainly, such pure thinking is also involved in
problems and seeks to get out of them; but all problems are not
practical. There are theoretical problems and theoretical situ-
ations, the like of which concerns us when we philosophise.
But we may even say more. Even in those cases where we
24 THOUGHT AND ACTION

think out of, or rather in, some practical situation, the essence
of thinking is radically different from that of practice or action.
The fact that thought can, and in fact does, attain to freedom
from practical relevance proves its essential freedom. Thought
and action are essentially distinct. That they are nevertheless
found in some cases to be associated with each other has to be
explained otherwise than by any sort of false reductionism.
3.2. Heidegger goes beyond the pragmatists, although it seems
at first as if he is giving nothing other than a pragmatist account
of thinking. Both intuition and thought are, for Heidegger,l
derivative from that primitive understanding which characterises
human existence. This primitive understanding (which is not a
kind of knowledge, but a fundamentalfactor of human existence 2)
is anticipatory and projective in character. Heidegger rejects
the conception of a pure subject. Human existence again is not
a present fact or thing, but an everpresent possibility of existing.
This possibility involves anticipation of, and projection into, the
future. This "being ahead of oneself", this anticipation and pro-
jection, constitutes that primitive understanding which has
always a feeling-tone, as it were. 3 The cool and contemplative
theoretical thinking (leading to science and philosophy) as well
as practical action, - both are derivative from this primitive
human existing.
Heidegger is not deriving theory from practice, nor is he plea-
ding for the primacy of practice over theory.4 Here he avoids the
error of the pragmatists. He is only insisting on the primacy of
human existence. Thinking (scientific or philosophic) is, for
Heidegger, a mode of existence. 5 It is therefore derivative from
the basic existential categories. While we need not, in our pre-
sent context, pursue this philosophy further, we may safely say
that Heidegger is dissatisfied with the dichotomy of theory and
practice and that he seeks to overcome this dualism not by re-
ducing the one to the other but by deriving them from a higher,
1 Martin Heidegger, Sein und Zeit, Max Niemeyer, Tiibingen, 7th edn. I953, p. I47.
2 Martin Heidegger, Kant und das Problem der Metaphysik, Vittorio Klostermann,
Frankfurt, 2nd edn. I95I, p. 2IO.
3 Sein und Zeit, p. I42: "Verstehen ist immer gestimmtes".
4 Ibid., p. I93: "Das Phiinomen driickt daher keineswegs einen Vorrang des 'prak-
tischen' Verhaltens vor dem theoretischen aus ... 'Theorie' und 'Praxis' sind Seins-
moglichkeiten eines Seienden, dessen Sein als Sorge bestimmt werden muss."
5 Ibid., p. 357.
THOUGHT AND ACTION 25
a more original unity. This too is reductionism, but this type of
reductionism is philosophically more satisfying, since it has the
courage to look squarely at the initial dichotomy. It shall be our
attempt in the present paper to explore some means of over-
coming the dualism between thought and action, but we could
do that only by a phenomenological receptivity to the facts as
they present themselves to us, and not by trying rival meta-
physical hypotheses.

4. It is often said that thought is rational and specifically


human, while action is irrational. This opinion is so erroneous
that it cannot even be adequately formulated, and although it
may be considered superfluous to mention such an opinion,
there are not a few sensible persons who think along this line.
The truth however is the contrary. For just as man alone can
think, so also man alone can act 6 . Both thought and action ori-
ginate in the essential constitution of man. It is a sheer mis-
understanding of the true nature of human action to take it as
nothing but a succession of real spatio-temporal events. It is as
much a mistake to take human existence as a natural fact as to
interpret his actions as natural processes. Man is as much a
thinking being as an acting one. While on the one hand, he is a
subject, he is also on the other a person.

4-1. We have therefore to seek for the distinction between


thought and action elsewhere. In this connection, two funda-
mental questions may be asked: first, who thinks and who acts?
Secondly, what is given in thought, and what is given in action?
An answer to the first question has been suggested above.
The who of thinking is the subject. The who of action is the per-
son. The relation between thought and action may thereby be
illuminated through the question: how are the subject and the
person related to each other?
With regard to thought, it is often rightly said that it is no
private affair of an individual, but an objective process (thin-
king) or an objective product (thought). In thinking, I am no
more this or that individual person but a universal subject,
capable of stepping out of the privacy of my individuality and
6 Martin Heidegger, Was heisst Denken?, Max Niemeyer, Tiibingen, 1954, p. 51
THOUGHT AND ACTION

participating in what is common to all. Hence the essential


communicability of thought, as contrasted with the utter pri-
vacy of sensations, feelings, etc. The who of thinking is therefore
not this or that individual, but a subject that is universal.
Whereas it is the universal subject who thinks, it is the in-
dividual person who acts. It is not only the who of action that is
individual, but all action takes place within a determinate situ-
ation. The person in fact is not a thing or substance who acts but
could have ceased from acting. On the contrary, the person is
essentially constituted by his actions; he is what Max Scheler
named an 'act-centre'. Every action affects or modifies the per-
sonality of the person, though not all in the same degree. And,
conversely, no action is intelligible without reference to the in-
dividual person who acts and the determinate situation within
which the action takes place. Further, an action requires not
merely the doer and the situation but also other persons.

4.2. In thus drawing the distinction we are no doubt descri-


bing ideal limiting cases, whereas in fact we have to accept com-
promise. That is to say, the who of thinking, the subject, does
not always attain to complete universality. In such cases, one
thinks no doubt, but thinking is motivated, as one says. Action
determines thinking. The person pretends to think. The ideal of
thought however is to eliminate the person, so that the subject
is reduced to a pure zero, a passive witness, 'Zuschauer'.
To describe limiting cases is however no fault, for it is only
through this that one can raise essential distinctions into clear
relief. Factual interweaving need not prejudice us against es-
sential distinctions.

4.3. Whereas thought may in fact be practically motivated but


is essentially (or, ideally aims at being) universal, action is in
fact individual but may ideally aim at universality (as in moral
action).
The universality which is thought's and the universality which
action may aim at are radically distinct from each other. The
subject who thinks is universal inasmuch as it is zero, empty of
contents, a bare witness. The person who acts aims at univer-
sality by enriching himself in content, by stripping off elements
THOUGHT AND ACTION 27
that jar, by attaining to inner self-consistency as well as outer
harmony (with other persons etc.).

5. The second of the two questions suggested above is: what


is given in thought, and what is given in action? The entire idea
of givenness may be challenged in the present context. It may be
said that nothing is given in thought, for thought only constructs.
Similarly, it may be argued that nothing is given in action; that
action only brings about some change, modifies the state of things.
As against such contentions we shall here insist on the possi-
bility of considering both thought and action as two different
modes of disclosure. Both are modes of givenness in spite of the
great diversity subsisting between them.

5.I. That thought is a mode of givenness has been denied by


those who insist on the constructive and interpretative function
of thinking. What is given is the particular, the sense-impression;
thought imposes its conceptual scheme on the impressions, in-
terprets them, and welds them into judgments. This is the view of
thinking in vogue since Kant. It makes thought subjective;
what is universal and necessary in our thinking is explained as
universal form, subjective though. The a priori is held to be
both subjective and formal, any objective and material a priori
being ruled out ab initio. Thought becomes wholly discursive,
an endless process of mediation. Within the limits of sense-ex-
perience it yields constructed appearance; outside of this limit
it creates illusions; nowhere does it have the capacity of giving
reality.
As against this we advocate here a theory that claims affilia-
tion to the classical Platonic-Aristotelian theory of thinking.
Thought on this theory gives Ideas, Concepts, Eidos, Essences.
Thought does not construct them, but recognises them. Thought
gains thereby an ontological significance. It reveals the ideal fac-
tors and structures inherent in real events and processes. Thought
discloses Being as Essence.

5.2. That however action could be considered as a mode of


disclosure is less often recognised. It is one of the services of
28 THOUGHT AND ACTION

Martin Heidegger to have insisted on this aspect of action. 7


Both Max Scheler and Nicolai Hartmann have also recognised
some sort of cognitive value of actions.
Max Scheler8 brings this out in the following way: Thought
gives the essential 'what'; sense-perception gives the accidental
(the here-now) 'what'; action gives the real 'that'. It is through
action and the accompanying feelings of concern and care that
we get involved in real situation; in trying to modify this real
situation we experience resistance and obstruction; in trying to
overcome such resistance and obstruction we are presented with
real existence. Thought cannot reach this existence; its intuitions
are confined to the sphere of ideality. Nor does sense-perception
originally give real existence; it presents only the accidental
qualities.
Thought can dissociate itself from real existence. Sense-
perception can be studied in its essence through phantasies and
imagination (Husserl). But action is so tied to the real situation
that to study it in isolation is to miss it altogether.

5.3. We have referred in 2.I to various kinds of actions. It


may be pointed out that each of these kinds of actions has an
essential reference to real existence and presents real existence in
some aspect or other. The real world is not exhausted by things
like tables, chairs, houses and trees, but consists of such diverse
entities as the human person, societies of persons etc .... Action
brings us face to face with their existence; we can contemplate
only their essences dissociated from their existence.

6. We have tried to bring out the essential distinction between


thought and action in two ways: (a) The subject thinks, whereas
the person acts. The former in its purity is universal, because it
is empty of all contents, a passive witness; the latter is an in-
dividual that aims at universality by enriching and harmonising
its own individuality. (b) Thought gives essences, Ideas, Eidos,
the ideal structures inhering in the constitution of the real;
thinking is free when such idealities are inspected in themselves,

7 John Wild, The Challenge of Existentialism, University of Indiana Press, 1955,


P·9 8 .
8 Max Scheler, Die Stellung des Menschen im Kosmos, Miinchen, 1947, p. II.
THOUGHT AND ACTION 29
bound when they are sought to be discovered within reality.
In the latter case, action helps thinking. Action, on the other
hand, discloses real existence in all its diversity.
That thought and action, essentially distinct though, are yet
factually interwoven may be explained by two reasons amongst
others: first, the subject and the person, and secondly, essences
and reality, are factually interwoven. The problem of the relation
between thought and action may therefore be pushed back to
two more original problems: the problems of the relations be-
twen the subject and the person, and between the essences
and reality.9 We need not carry out this immense task in this
paper. It is enough if a line of approach to the topic of our sym-
posium has been suggested.

7. Another line of approach which again could not be worked


out here is a study of the time-structures of the two, thought and
action. Thought, it could be shown, is backward-looking. Al-
though the essences that are given in thought are timeless,
nevertheless thought discloses them only in what has been, in
what is past. It is significant that the German word "Wesen"
(=essence) is connected with the past perfect of "sein" (=to be),
i.e. "gewesen". Hegel makes this clearer when he defines "es-
sence" as "das vergangene, aber zeitlos vergangene Sein".10 The
timeless essence is recognised, recollected (Plato's amnesia).
Action on the other hand, it could be argued, is anticipatory; it
is forward-looking. It plans and calculates. It seeks to get out of
the present, to build the future. The pragmatists read into thought
what holds good of action. This they do when they take our ideas
as tools, as plans of action, as programmatic and methodical in
significance. Here Hegel is rjght, and the pragmatists wrong.

9 The former is ably dealt with in Nicolai Hartmann's Das Problem des geistigen
Seins, Berlin and Leipzig, 1933; the latter is discussed in ample detail in the present
author's Nicolai Hartmann and A. N. Whitehead: A Study in Recent Platonism, Cal-
cutta, Progressive Publishers, 1957.
10 Hegel, Die Wissenschaft der Logik, Phil. Bibliothek edn., 1951, Vol. II, p. 3.
CHAPTER IV

MEANING AND TRUTH-I*

I. I have chosen as the theme of my address two questions


which to my mind owe their importance in philosophy to the
fact that they place us exactly on the borderline between phi-
losophy of language, theory of knowledge and ontology. These
two questions roughly formulated are: what is the nature of the
meaning of expressions? How is the notion of meaning related to
that of truth? The precise implications of these two questions
however could emerge only as we proceed with our discussions.
I am aware that in so formulating the two questions, especially
the first of them, I am running counter to the prevalent philo-
sophical orthodoxy. One of my main purposes shall be to in-
sist, as against this orthodoxy, that there are genuine philosophi-
cal problems about the nature of meaning and about the nature
of truth. The problem of meaning has been sought to be dissol-
ved by reducing that notion either to that of synonymity (Quine),
or to that of truth-conditions (Carnap) or to that of use (Wittgen-
stein). Likewise, the philosophical problem of truth has been re-
garded as superfluous on the supposed ground that 'is true' is
no genuine predicate at all (Ramsey and Ayer)l. Postponing my
defense of the notion of meaning for the present, I would like to
examine very briefly the attempt to eliminate the problem of
truth.

I.I. It has been argued that since (( 'S is p' is true" is equi-
valent to'S is p', the predicate 'is true' is superfluous and has at
best - as Geach says2 - the function of cancelling the quotes. (It
* Presidential Address, Logic and Metaphysics Section, 36th session of the Indian
Philosophical Congress, Santiniketan, I96I [Proceedings, pp. 27-47.]
1 A. J. Ayer, "The Criterion of Truth" in Macdonald (ed.), Philosophy and Analysis.
2 P. T. Geach, Mental Acts, London, Routledge & Kegan Paul, p. 96.
MEANING AND TRUTH-I 31
is at once obvious that this argument closely resembles Russell's
elimination of the philosophical problem of existence). As against
this one may argue that though the above equivalence holds
good, it does not follow that the predicate 'is true' is redundant.
The redundancy is wrongly supposed to follow from the equi-
valence partly owing to the failure to distinguish between the
different senses in which the connective 'is' is used in the type-
sentence'S is p' as it occurs on the left-hand side of the equivalen-
ce and as it occurs on the right-hand side. As a consequence, it
has been wrongly supposed that the'S is p' remains unaffected
by the addition of the predicate 'is true' and that it re-appears on
right-hand side unaffected, though only without the quotes.
The error in the argument however would immediately be clear
if we bear in mind that the 'is' in'S is p' is not quite the same on
both the sides of the equivalence. On the left hand side expression
the 'is' is a predicative 'is', whereas on the right hand side ex-
pression it is an existential 'is', so that what the equivalence really
amounts to is this: saying that the predicative assertion'S is p'
is true is the same as saying that S is in fact p. In that case, the
predicate 'is true' far from being redundant fulfils the function
of transforming the predicative 'is' into an existential 'is'. Hence
the problem of truth is genuine, at least in the sense that it
survives the above discussed attempt at its elimination, though as
we shall show later on, there is another important sense in which
the problem may be replaced by the problem of falsity.

2. Now since one of the questions that I have placed before me


concerns the relation between the concept of meaning and that
of truth, I should right at this stage begin by clearly separating
the two concepts from each other, and in so doing I shall of course
be rejecting two diametrically opposed ways of obliterating that
distinction. I find no better instance of the tendency to reduce the
notion of truth to that of meaning than the following passage
from Harold H. Joachim's The Nature of Truth (Oxford, 2nd,
edn., 1939). "Anything is true", writes Joachim 3 , "which can
be conceived. . . Conceivability is the essential nature of truth."
Explaining what is meant by conceivability, Joachim goes on to
say: "And to be 'conceivable' means to be a 'significant whole',
3 H. M. Joachim, The Nature at Truth, Oxford, 2nd edn., 1939, p. 66.
32 MEANING AND TRUTH-I

or a whole possessed of meaning for thought." And with this the


reducing of the concept of truth to that of meaning is complete.
At the other extreme, we have the more recent attempt to reduce
the concept of meaning to that of truth-conditions, which if
successful, would go a long way towards eliminating the philo-
sophical problem of meaning. We find a clear enunciation of this
principle in Carnap's Introduction to Symbolic Logic where we
are told that "a knowledge of the truth-conditions of a sentence
is identical with an understanding of its meaning". 4 By the
truth-conditions of a sentence are meant the circumstances under
which a sentence can be used to make a true statement.
Joachim's theory may safely be set aside on the ground that
the false is as much conceivable as the true, and if the theory may
still seek to maintain itself it must be by giving to the notions of
meaning and truth connotations far beyond what their ordinary
uses imply. The other reduction, namely the reduction of mea-
ning to truth-conditions is nearer the mark, for it rightly sees
the close relation that subsists between the two concepts but
errs if it aims at cleanly identifying them.
There are at least two forms of this theory differing in the
degree of intimacy that is sought to be established between the
two notions. There is on the one hand the Russell-Wittgenstein
form of the theory which, rejecting the Fregean distinction
between sense and reference, identifies 'having a sense' with
'being true-or-false'. Then there is the Frege-Strawson form of the
theory which seeks to retain the distinction between sense
and reference, and therefore between meaning and truth, but
has curiously enough ended up with the same, though rather dif-
fident, identification of the meaning of a sentence with the cir-
cumstances in which the sentence could be used to make a state-
ment. 5
The following remarks should, I believe, be sufficient to bring
out the weakness of this theory in either of its forms: in the first
place, expressions that are not sentences are also meaningful and
the theory that identifies meaning with truth-conditions evi-
dently does not apply to such expressions. Nor does the theory

4 R. Carnap, Introduction to Symbolic Logic, New York, 1958, p. 15.


6 G. E. M. Anscombe, An Introduction to Wittgenstein's Tractatus, London, Hut·
chinson, 1959, pp. 59-60.
MEANING AND TRUTH-I 33
apply to sentences other than the indicative ones. Further, even
with regard to the indicative sentences the theory does not apply
to sentences other than those that are effectively decidable either
as true or false. 6 And finally even with regard to effectively
decidable indicative sentences a distinction has to be drawn
between meaning and truth-conditions, for in order to be able to
decide what the truth-conditions of a given sentence are one
must' first be able to understand what the sentence means. 7 It

must however be said in all fairness to the theory that it does not
exactly identify 'having a meaning' with 'being true-or-false',
and the notion of 'true-or-false' with the alternation involved in
it clearly points to the fact that the notion of meaning is logi-
cally prior to that of truth-value or even to that of truth-con-
dition. Phenomenologically speaking, the best evidence for this
independence is provided by the fact that one might very well
install oneself in the attitude of mere understanding without
entertaining the question of truth or falsehood.
There is still another consideration which ought conclusively
to prove that the two notions have to be kept separate. We may
ask: of what is meaningfulness predicated, and of what is truth
or falsity predicated? Of course, it is a linguistic expression, a
word, or a sentence, of which alone we could say that it is either
meaningful or meaningless. Those who wish to identify meaning
with truth-condition also maintain - as a necessary corollary of
their contention - that truth or falsity are semantical and me-
talinguistic predicates of sentences in the object-language. But I
wish to maintain as against this that truth and falsity are not
predicates of sentences: so far I agree with Strawson,8 though I
would differ from him in holding that they are descriptive pre-
dicates. But predicates of what? Once we reject the claim of
sentences, there are only two other claimants left: the propo-
sitions expressed by the sentences, or the knowledge or iiiiinas
as understood in Indian philosophy. Propositions as subsistent
entities - what else are they if we at all admit them as distin-
guished from linguistic expressions on the one hand and the men-

6 Cpo M. Dummet, "Truth" in Proceedings Aristotelian Society, 1958-9, pp. 141-162.


7 W. H. Werkmeister, The Basis and Structure of Knowledge, p. 44; Dummet,
loco cit., pp. 148-9.
8 P. F. Strawson, "Truth" in Macdonald, loco cit., especially pp. 262-3.
34 MEANING AND TRUTH-I

tal acts or attitudes on the other? - are not the sort of things
that could be said to be true or false. Consequently, we are left
with the knowledges and these alone are the proper subjects for
the predication of truth or falsity. We have thereby no doubt
left the conception of knowledge or iiiiina vague, but our own
immediate purpose in hand would have been served if we can
emphasise that truth or falsity on the one hand and meaning-
fulness on the other are predicated of quite different sorts of
things - which corroborates our conclusion that the two notions
are really distinct.

3. Before I come to deal with the notion of meaning in parti-


cular, a few words on the philosophical method adopted here
would not be out of place. For in matters like those with which
we are at present concerned it is most necessary to start with a
correct perspective in one's mind. This is particularly so with regard
to the concept of meaning. Using Carnap's well-known distinc-
tion, we can say that both the material mode of speech and the
formal mode of speech are one-sided extremes. Dogmaticontology
is as much naive as the modern decision in favour of the formal
mode of speech is. And it must be remembered that the dis-
trust of dogmatic ontology is nothing new: it is in fact as old as
Kant (whose entire philosophy in the First Critique may be
regarded, as Pichler and, following Pichler, Wein have shown, as
an attempt to translate the categories of the old dogmatic on-
tology into the language of transcendental idealism).9 Only
whereas Kant sought to transform ontology into transcendental
logic, the moderns have tried to transform it into the formal
mode of speech. As a corrective against the naivety of the ma-
terial mode of speech, all this is indeed valuable. But if made
into an absolute philosophical method the resulting 'lingua-
centric predicament' becomes, if less naive, certainly more ar-
bitrary. Faced with such a situation, it seems to me that we are
in need of a perspective - should I say, a mode of speech? -
which would be able to integrate the two opposed and naive
9 H. Pichler, Ueber Christian Woliis Ontologie, Leipzig, 1910; H. Wein, Zugang
Bur philosophischen Kosmologie, Mtmchen, 1954. For an account of these see Mohanty,
"The Principles of Modern Kant-Interpretation in Germany" in Krishna Chandra
Bhattacharyya Memorial Volume, Indian Institute of Philosophy, Amalner, pp. 104-
119·
MEANING AND TRUTH-I 35
extremes, and do so not in the sense of combining them into a
unity but in the sense of providing the basis on which both could
stand. Such a mode I find in the phenomenological mode which
corrects the naivety of the ontological assertion and provides a
check to the arbitrariness of linguistic decision, i.e. to the con-
ventionalism inseparable from the formal mode of speech. By the
phenomenological mode I mean the appeal to the mode of given-
ness instead of to the mode of existence or the linguistic usage.
I believe that all these three may be regarded as three different
approaches to any philosophical problem, but I also believe that
of these three the phenomenological one possesses an epistemic
priority. The three in their unity constitute the philosophical
method.

3.1. Let us see in brief if the above methodological remarks


prove fruitful in some other fields before we could confidently
carry them over to the problems of meaning and truth. It is well
known that many recent thinkers have in their attempts to
avoid the ontology either of physical objects or of sense-data
come to re-interpret the two traditional theories of perception,
the physical object theory and the sense-datum theory, as but
two alternate linguistic devices each of which could profitably
be adopted, the decision being a matter of convenience and other
extra-theoretical considerations. Now, that an ontological theory
has a linguistic correlate is an important finding. But to leave
the matter at that and to rest content with an ultimate alter-
nation seems to be an escape rather than a solution of the pro-
blem of perception. Under such circumstances, the phenomeno-
logical method may provide us with the clue: it can help us to
reconcile the two theories by exhibiting them not as two alter-
natives between which one is free to choose but as representing
two different but successive modes of givenness, the physical
object being given in the primary, unreflective and practical
mode, the sense data on the other hand being given in subse-
quent, reflective and contemplative mode. Detailed explication
of these two modes could rehabilitate the two theories to their
respective rightful places by freeing them from the arbitrariness
of linguistic conventionalism. A similar way out may be suggested
from the traditional impasse between nominalism and realism
MEANING AND TRUTH-I

and the modern alternation between the two corresponding


languages. In other words, it can be shown that particulars and
universals are given in two different modes: the former in sense-
perception which is primarily action-oriented in character, the
latter in thought. Since it is not part of our present purpose to
explore these problems, it is enough to have hinted at the re-
levance of our methodological remarks, and we may now pass
on to a consideration of the question of meaning.

II

4- In no other area of philosophical thinking is the modern


philosopher more suspicious of the temptation to ontologise than
in discussions about meaning. The main issue here may be for-
mulated as one between Platonism and Anti-Platonism and our
need is to find a stable via media. Certain forms of Anti-Pla-
tonism may safely be regarded as dead. I have especially the so-
called image theory in my mind, and Wittgenstein's tirades
against it in the Philosophical Investigations, just as they are,
seem however like flogging a dead horse. Amongst the Anti-
Platonist theories, I would consider only three varieties: the
theory of meaning as a dispositional property of signs (C. L.
Stevenson), Operationalism (the latter Wittgensteinian school);
and the meta-linguistic relativism (B. L. Whorf and Hans Lipps).
C. L. Stevenson in his Ethics and Language (1944) attempts to
avoid the unstability of psychologism without at the same time
taking meanings as any sort of entities. Stevenson rightly sees
that the problem is to find the constant meaning amidst psy-
chological flux. Psychological reactions to an expression (as to,
say, coffee) fluctuate. But one speaks of a disposition or power
that is relatively stable. In order to get at the relative identity of
meaning, Stevenson undertakes an analysis of the concept of
disposition or power. This does not mean that disposition or
power is some kind of entity that exists over and above its tan-
gible manifestations. Neither is the disposition the cause of the
psychological responses (that would be like saying man's rational
faculty is the cause of his reasoning activities). Avoiding such
erroneous solutions, Stevenson attempts a definition of dispo-
MEANING AND TRUTH-I 37
sition in use. This involves detailed complications which we need
not examine for our present purpose. Granted that such a defi-
nition is possible it becomes relatively easy to say what meaning
is. Meaning in that case becomes a dispositional property of the
sign. This dispositional property is relatively stable, while the
actual psychological effects vary.
The hypothesis of potency, power or sakti - whether in the
metaphysical or in the positivistic sense - is properly called for to
explain only the relation between a word and its meaning. This is
how the Mimamsakas and the Naiyayikas introduce the notion of
sakti. But to reduce meaning itself to this potency is not phe-
nomenologically justifiable. Potency or power, on any interpre-
tation of those words, is not felt as being given. It is either in-
ferred as a metaphysical entity or, as with Stevenson, only defi-
ned in use. In the latter sense, it may be said to be a logical con-
struction in Russell's sense, but in any case it is not what is felt
to be given as such in experience of understanding an expression.
Meaning however is felt to be given. What is felt to be given can-
not be reduced to what is only either inferred or constructed.

4.2. B. L. Whorf's paper on "Language, Mind and Reality"


draws our attention to the supposed fact that the fixation of
objective meanings is a process that falsifies the original nature of
language. According to Whorf, language has two aspects: the
'patternment-aspect' and the name-giving aspect. In the latter
aspect, language gives names to parts of a whole, isolates them
and fixes them as self-subsisting entities. In the former aspect,
language is not concerned with 'names' and 'forms', but with
pure patternments. The reference of words that dominates lan-
guage in its name-giving aspect is at a minimum in its higher, the
patternment aspect. In this higher aspect language is algebraic in
nature, the symbols are variables. This algebraic nature persists
even in the usual lexical language where "sentences, not words
are the essences of speech just as equations and functions, not
bare numbers are the real meat of mathematics." 11
From this follows that word-meanings are not self-subsisting
but are isolated artificially from the living patternment-aspect.
10 Included in B. L. Whorf, Language, Thought and Reality, New York, 1956.
11 Whorf, loco cit.
MEANING AND TRUTH-I

The identity of meanings is an illusion fostered by the lower mind,


the name-giving aspect.
From an allied but different point of view Hans Lipps in his
Untersuchungen zu einer Hermeneutischen Logik (Frankfurt am
Main, I938) offers a forceful statement of the view that there are
no identical word meanings, that the so-called invariable ob-
jective meanings are only abstractions cut off from the living
linguistic meaning which is determined in each case by the actual
conversation and the action and reaction between the speaker
and the hearer. That is the living context from which a word
could only be artificially cut off. Apart from such artificial se-
paration there are no identical meanings.
In the above arguments there are several points that need be
considered separately. In the first place there is the fact that
most words change their meanings under circumstances. Besides
there are essentially ambiguous expressions including the so-
called 'indexical expressions' (C. S. Peirce). Secondly, language
consists in actual speaking and hearing and meanings arise only
in this context of speaker-hearer relationship. Thirdly, the inner
reality oflanguage consists, as Whorf puts it, in a 'pure pattern-
ment' and certainly not in that name-giving aspect which comes
to the forefront due to the limitations of what he calls the lower
mind.
In reply we may point out the following: the supposed ambigui-
ty and fluctuations in meaning may be shown to be rather due to
the imperfections of our system of symbols than to the non-
availability of identical meaning-contents. For, though an ap-
parently identical expression may convey different meanings in
different contexts it is theoretically possible to take hold of each
such meaning in each case and tie it to a fixed expression. The
fluctuations will then be seen not as fluctuations of meanings
themselves but of the use of an apparently identical expression.
Without entering into the distinction between lower mind and
the higher mind, we might admit that in its purely formal aspect
language does exhibit a "patternment" aspect and that from this
standpoint the purely material meaning-contents are of no rele-
vance. But we are presently concerned exactly with these mea-
ning-contents on whose nature the patternment aspect throws
but little light. The second of the above facts places before a
MEANING AND TRUTH-I 39
Platonic theory of meaning an almost insurmountable difficulty.
The living context of conversation or speech may be viewed either
from the side of the hearer or from the side of the speaker. Much
discussion in Indian philosophy in the context of the sabda pra-
mana takes the standpoint of the hearer. But it may argued
that the speaker's standpoint is more fundamental. The expres-
sion as spoken may be regarded as the basic phenomenon, so
that the expression as heard my be reduced to the former. The
hearer understands it as if he were speaking it. The speaker may
hear his own expression, but that is a subsidiary phenomenon.
If that be so, the apparent relativity implied in the speaker-
hearer relationship may be dispensed with. Spoken meaning or,
more correctly, meaning belonging to an expression as spoken
may be, since it is the basic phenomenon, isolated from the act
of speaking as its intended correlate. Such contents then may be
regarded as identical entities.

4.3. Another way of retaining the objectivity of meanings


without getting involved in Platonism of any kind is that of the
later Wittgenstein and the Operationalists according to whom
the meaning of an expression is the same as the rules governing
its use in a language. We know the meaning of this theory if
we know these rules, convention etc., or if we know how to use
the expression. This view has the merit that it can explain the
objectivity of meanings, for the rules and conventions are objec-
tive features of the particular language game under consideration.
But what happens when we grasp the meaning of a word or of a
sentence in a flash as it were? Surely it is not the use that we so
grasp. Nor is it a set of rules and conventions. Wittgenstein
rightly emphasises that what is present before the mind is not a
mental picture, for a picture does not force upon us any parti-
cular application of the expression under consideration. And yet
there is something in the expression qua expression and something
which the mind grasps that makes possible an understanding of
it prior to all applicaion, that in fact in a way predetermines
which applications of it are right and which not. Right applica-
tion may be a criterion of understanding. But should we say in
face of the above difficulties that though understanding does not
amount to actual application it however does amount to appli-
MEANING AND TRUTH-I

cability? The switch- over from actual application to applica-


bility is similar to the switch- over from actual verification to
verifiability. What these two cases teach us is that a recourse to
the language of possibility is indeed inevitable. But the posi-
tivists are not able to give a satisfactory account of the possi-
bility and the disposition-words; in doing this they must in-
evitably come back to actual application (or, verification) thereby
completing the circle. It is therefore time that the now fashion-
able opinion that 'know' and 'understand' are capacity-words
should come to be suspected and should be made to suffer the
same fate as the one- time fashionable opinion that categorical
material object statements are reducible to hypothetical sense
datum statements. We have to face the grim reality that though
there is an important sense in which they are dispositional, yet
the really fundamental sense of these words is episodic. In the
case of understanding or meaningfully using an expression this
must be an intellectual awareness or an intention whose in-
tended correlate is what we might call meaning in the substan-
tive. Wittgenstein says that an intention "is imbedded in its
situation, in human customs and institutions".1 2 If there were no
rules of playing chess, Wittgenstein argues, I could not even in-
tend to play the game of chess. Wittgenstein's arguments prove
only this much - and this can hardly be denied - that certain
objective circumstances must be given in order that I could even
intend in a certain way. The fact that a conventional system of
signs with rules of operation must be given in order that I could
even intend constructing a sentence does not decide the issue
under consideration. This is not a sufficient reason for holding
that the linguistic expression is nothing more than a merely
physical sign or that our understanding of it is nothing other
than the capacity to operate with it in accordance with custom-
bred conventions. Our convention would be that given such a
set of signs and rules of operation developed through custom,
such a set would not amount to language - nor would the opera-
tion amount to an understanding - unless the said intellectual
act supervened.

s. W. v. O. Quine has in recent times made the most notable


12 L. Wittgenstein, Philosophical Investigations, Oxford, I954, p. Io8e.
MEANING AND TRUTH-I

attempt in the direction of avoiding Platonism in theory


of meaning. With regard to Platonism in general he has, as is
well known, suggested that class-names or functions should be
prohibited from appearing within the quantifier. Such quanti-
fication should be disallowed, for it would amount to an onto-
logical commitment regarding the 'existence' of abstract en-
tities. This particular device meant to limit the 'ontological
commitment' of a particular logic need not interest us at present.
What we are rather interested in is his suggestion that the pro-
blem of meaning, as distinguished from the problem of reference,
should be concerned only with one question, namely the question
of synonymity. In Quine'S opinion, "a felt need for meant en-
tities may derive from an earlier failure to appreciate that mea-
ning and reference are distinct." The hypostatisation of meanings
is based on the confusion between meaning and reference. If
the distinction between meaning and reference be borne in mind,
the question "What sort of entities are meanings?" becomes
senseless. Search for entities would then come under theory of
reference, while theory of meaning would be left only with one
question, how and under what conditions is synonymity possible?
There would of course remain the allied notion of analyticity.
Platonism would then be entirely ruled out of theory of meaning.
I must confess I do not share the happy optimism of Quine.
His diagnosis is not beyond doubt. For both Frege and Husserl
have been charged of Platonism even though both of them
draw the distinction between sense and reference. The truth
seems to be the reverse. The distinction between sense and re-
ference is just what the Platonist can most rely upon. By exclu-
ding from consideration the objects referred to, attention may be
fixed on meaning as such and now that the meaning cannot be
identified with the objects referred to it has to be accorded a
distinct ontological status. The problem of synonymity is only
another aspect of the problem of identity of meanings. The
question which is an empirical one enquires into the criteria by
which we could regard two expressions as having the same mea-
ning; but the whole discussion, so much fashionable now, pre-
supposes that the two expressions under consideration have an
identical meaning.
42 MEANING AND TRUTH-I

6. Our examination of the various anti-Platonistic theories


has led us to the conclusion that Platonism in theory of meaning
is founded on an incontrovertible phenomenological basis, that
there is an important sense in which meanings retain an iden-
tity amidst fluctuations of situational and conventional con-
texts and an objectivity in the midst of varying subjective in-
tentions and responses. This lends to the Platonistic hypostati-
sation of meanings its point of advantage over against all forms
of psychologism and all forms of conventional and operational
relativism. But we can accept such a Platonism only as subject
to certain provisions which may now be stated as follows:

6.1. In the first place though meanings by virtue of their


identity in discourse and their objectivity could be called uni-
versals, yet not all meanings are universals. This is especially
so in the case of the meanings of the so-called ego-centric parti-
culars like T and 'This'. These expressions make possible re-
ference to unique particulars, and there is something about
their sense that refuses to be idealised.

6.2. It has often been pointed out that Platonism in theory of


meaning involves the thesis that in understanding an expression,
we inspect impalpable meanings. This however is not true. It
should be remembered that we have to examine cases of purely
symbolic thinking and understanding, not amounting to know-
ledge. Thinking or understanding as constituent of a knowledge
is of course not directed towards the meanings themselves un-
less in a subsequent act of reflection. But the same holds good
also of purely symbolic thinking and understanding. Expressions
have not only meaning but also reference, and the intended
reference characterises even those cases where the psycholo-
gical intention to refer has patently been excluded. The function
of the meanings is to make reference possible and also to
give the reference its determinate character, while they
themselves are not the sort of things that could be inspected in
themselves: this may be expressed metaphorically by saying
that there is a transparency about them so that they serve as the
media for reference. Or, if you like, the meanings are at the same
time and in the same act given only in the sense of being intended.
MEANING AND TRUTH-I 43
But then it could indeed be shown that there is no other gene-
ralised sense of being given other than being intended.
To the above it might be objected, following Strawson, that
reference is not a function of expressions qua expressions, that
expressions as such have only meaning and that it is only a ge-
nuine use of them that refers. If in the case of a merely symbolic
thinking I am not genuinely using an expression, i.e. if I am not
using it to make a statement, I am not referring at all, in which
case I could then be concerned with meanings not as transparent
media for reference but as opaque entities for inspection. But
such a theory of reference is not tenable, for Strawson's theory
involves as I have shown elsewhere a hopeless circularity.1 3
This circularity shows that reference like meaning has to be
ascribed to expressions qua expressions. This however need not
lead us to the acceptance either of the Meinongian ghosts or of
Russell's dissection of them. Let us for the moment suppress our
interest in ontology. In that case we shall see that every use of
expressions refers, though in some cases, owing to reasons which
may be either a priori or empirical, the reference may remain a
merely intended reference, incapable of being fulfilled. Now in
the case of merely symbolic thinking or understanding, I may have
absolutely no interest in the fulfilment of the intention: my
interest may be merely to grasp the intended reference through
the intended meaning.

6.3. As has been emphasised in paragraph 3, the objectivity


and the ideality of meanings has to be understood against the
light of both a linguistic convention and a subjective intention.
A linguistic convention, as Wittgenstein rightly saw, is a neces-
sary condition of there being a meaning-intention at all, which
however by no means implies that the meaning so intended is
produced by the convention. A more intimate relation however
subsists between the ontological and the phenomenological mo-
des, between the objective meaning and the subjective intention.
In fact, the "I mean" and the "It means" are correlative aspects
and complementary descriptions of the same unitary phenome-
non. The formerfrees the latterfrom the charge of hypostatisa tion,
13 J. N. Mohanty, Edmund Husserl's Theory of Meaning, The Hague, 1964, pp.
21-23·
44 MEANING AND TRUTH-I

while the latter rescues the former from the charge of relati-
vism. To reduce the one to the other for the sake of metaphy-
sical simplicity would amount to a distortion of the phenome-
non.

6.4. But there would still arise no doubt the transcendental


question: what is the a priori condition of the possibility of
intending universal and objective meanings? How is it that
meanings though they arise out of subjective intentions yet
emerge as objective, overindividual, impersonal entities? This
transcendental question is parallel in its structure to Kant's
problem in his transcendental deduction of the categories. Kant
asked, how are the categories though subjective in origin yet
objectively valid? He was concerned with the question of syn-
thetic a priori truth. Our transcendental question is logically
prior and concerns the possibility of universal, objective and
impersonal meanings. However, into the solution of this im-
portant question we cannot enter here.

III

7. It is now left for us to connect the two notions of meaning


and truth. The proper line in which the connection has to be
sought has already been indicated in our discussions. It is only in
the case of indicative and effectively decidable statements that
the question of truth and falsity arises, and it is only in these
cases that the relation has to be investigated. But again as we
have already said truth and falsity primarily characterise not
sentences, not propositions, but strictly speaking knowledges. We
call a sentence or a proposition or a statement true only in so
far as it is regarded as a constituent of a knowledge of some per-
son or other.

7.I. Modern philosophers are as much suspicious of the word


'knowledge' in the entitative sense or of the verb 'knowing' in
the episodic sense as of anything else. But they have only suc-
ceeded in analysing it into dispositional language whose inter-
pretation is no easier than that of names of psychic entities or epi-
MEANING AND TRUTH-I 45
sodes. Indian philosophy also recognises a dispositional sense of
the word 'knowledge', and in this sense 'to know' of course means
to have the right samskara about a matter. But there is an im-
portant sense in which 'to know' describes something that has
happened, an occurrence. Perhaps, saying that it is an achieve-
ment-verb is nearer the mark. As a more concrete account we
uphold the conception of 1'nana as an occurrent which is propo-
sitional in nature, and from which the corresponding proposition
is an abstraction. 14 Hence when a statement is said to be true it
is so only as an abstraction from a knowledge. But when it is
found to be false its claim to be knowledge is rejected: it remains
however a statement though still more than a mere sentence.
Though more than a mere sentence it is still a sentence and is so
far meaningful apart from being referential. Meaning is a pre-
supposition of knowledge for it is only through meaning that
reference is possible and in knowledge the referential aspect
predominates.
Thus one could say that in any concrete case of knowledge
there are three strata, one built as it were upon the other. There
is, as logically the most primitive, the sentence as a meaningful
expression and with a merely intended reference. Built upon it
as the basis we have the statement which qua statement might
express mere belief, i.e. may be true or false. The final stratum
is the knowledge itself as whose vehicle the statement is
true and not false, though the knowledge itself is definable
only in terms of truth: the very possibility of its being excluded
by its very definition.
Our above analysis accords its rightful place to each of the
two principal theories of truth in Indian philosophy, the theories
of svata/:tpramanya and parata/:tpramanya. The former is right in
so far as truth is an intrinsic property of knowledge i.e. in so far
as they are definable in terms of each other. The latter is right in
so far as truth is an extrinsic property of mere statements, i.e.
in so far as the latter may be true or false.
This also shows an element of truth in the contemporary denial
of the philosophical problem of truth. For truth being a defini-
tional property of knowledge, the problem of truth is the same as
14 For further discussion on this, J. N. Mohanty, Gangesa's Theory of Truth,
Santiniketan, I966.
MEANING AND TRUTH-I

the problem of knowledge, and in this sense no criterion of truth


is called for, though a criterion of falsity is necessary. But with
regard to a mere statement the problem is a genuine one of de-
ciding whether it is true or false.

8. The noematic contrast between meaning and truth roughly


corresponds to the noetic distinction between understanding and
knowing. Understanding a sentence of course as such amounts to
knowing its meaning, but we would however prefer to call this
sort of knowing 'understanding', for meanings are, as said before,
merely the transparent media for reference. In this sense then to
understand a sentence'S is p' is also to understand what kind of
facts it refers to, but this does not as such amount to knowing
that S is p, i.e. to knowing the fact it is used to refer to. But
knowing that S is p presupposes an understanding of the sen-
tence 'S is p'. Understanding the expression which states the
fact is presupposed by knowing the fact so stated. Knowing
implies a contact - speaking of contact metaphorically though -
with the fact referred to by the expression, or in Husserl's lan-
guage a fulfilment of the intention imbedded in the expression.
This is not to say anything regarding the nature of the 'ful-
filling' experience. There is in other words a fundamental dif-
ference between the way expressions refer and the way knowledge
refers. The reference of an expression is a presupposition of the
reference of the knowledge which has that expression for its
vehicle but which nevertheless exhibits entirely new pheno-
menological features.

8.1. There are two extreme cases where this fundamental dis-
tinction appears to be blurred. In the first place, there are
proper names - perhaps Russell's 'logically proper names', the
'this' and the 'I' - in whose case understanding amounts to
knowing. There are, on the other, abstract sciences like mathe-
matics and logic in whose case again understanding seems to be
the same as knowing. In both cases, the distinction between
meaning and reference seems to get blurred: in the former the
meaning is assimilated into reference, in the latter the reference
into the meaning. But closer examination would reveal that even
in these cases understanding may not quite amount to knowing.
MEANING AND TRUTH-I 47
In the former case, the "this" has still a theoretical meaning
which might be grasped without an identification of the unique
reference involved. In the latter case, we might argue - following
the intuitionists - that understanding the expression 'the first
prime number between 2000 and 3000' need not imply a know-
ledge of that number: it can be said to amount to knowledge
only when the number can be constructed in intuition. The only
difference between the two cases seems to be that in the former
the passage from understanding to knowing involves a transition
from the theoretical attitude to the practical, whereas in the
latter the passage is within theoretical consciousness. The really
pertinent question therefore concerns the former, and in a way
all concrete knowing. Since - as has been said by implication -
the gulf between the two attitudes is bridged by intuition and
since in all cases of concrete knowing the intuition would have to
be such as to involve a unique reference, the question arises,
how is such unique reference possible? This question must not be
confused with the quite different question, how is extra-lin-
guistic reference of language possible? to which E. W. Hall has
devoted two excellent papers in M ind.l 5 It seems to me how-
ever that extralinguistic reference of language is not a real
problem, for such reference is an essential character of a language
qua language. What however is a problem is, how it possible
to make unique reference? It must be emphasised that within
language purely unique reference is not possible, for even Russell's
'ego-centric particulars' or Hall's 'empirical ties' have an aspect
of meaning distinguishable from that to which they may be used
to refer uniquely. Unique reference presupposes a transcendence
of the attitude of pure understanding. The unique individual must
be known. And the knowledge of the unique individual is far from
being a contemplative perception. It must be recognised as a
radically different mode of knowing, intimately associated with
and built upon interests, dispositions, evaluations and practical
manipulations. For, the more contemplative a knowing is, the
more it approximates towards the attitude of understanding.
The more does it succeed in making unique reference possible
through a practically oriented approach, the more does it be-
15 E. W. Hall, "The Extra-linguistic Reference of Language" in Mind, 1943, pp.
230-4 6 ; 194-4, pp. 25-47.
MEANING AND TRUTH-I

come concrete knowing. Scientific knowledge in its escape from


unique reference seeks to transform itself into scientific under-
standing. It is not for nothing that perception of the physical
object and recognition of the other persons, the Thou (but not
merely the other mind), have frequently been regarded as the
standard type of knowledge. And in both cases, the knower or
the subject is not a passive onlooker but an active agent.1 6

8.2. Thus we are led to the conclusion that though, as maintain-


ed above, knowledge and truth are definable in terms of each other,
yet the confirmation of a belief requires, in cases of relatively
more concrete knowing, a pragmatic test whose function cannot
be replaced by the merely theoretical consistency or coherence.
The close interrelation of concrete knowing with the practical
attitude more clearly reveals the gulf that divides it from the
attitude of pure understanding.

8.3. Both understanding and knowing however are objective


modes of consciousness in so far as both are consciousness of ...
Meaning as well as truth are, in Husserl's language, 'mundane'
notions inasmuch as they presuppose the conception of a world
of objects. That they do so is brought out by the fact that phi-
losophizing about them inevitably exhibits an ontologising ten-
dency. They are therefore to be distinguished from the idea of
pure consciousness which is not a consciousness of . . . ,and which
therefore represents the idea of the limiting point of our turning
away from all ontology, empirical or a priori. The notion of
such a pure consciousness is not only beyond ontology but also
beyond phenomenology, for if phenomenology is concerned with
modes of givenness one can always ask what is so given, from
which it follows that phenomenology, though the more funda-
mental of the two, has a correlative ontology. But representing
as it does the very negation of the ontological attitude, pure
consciousness that is not a consciousness of ... cannot be even a
phenomenological notion and can at best be postulated only
by an act of faith.

16 For a similar viewpoint, see S. Hampshire, Thought and Action, London, Chatto
&Windus, 1959, especially pp. 47-53.
MEANING AND TRUTH-I 49
For phenomenology, the intentional consciousness i.e. the
consciousness of ... is the ultimate notion. This intentionality of
consciousness develops into two modes: the extra-linguistic
reference of expressions through meanings, and the objective
reference of knowledge, the former being the presupposition of
the latter.
CHAPTER V

MEANING AND TRUTH-II*

The problem of meaning and truth is a vast and widely rami-


fied one, and for our present purpose it is essential to make sure,
at the beginning, what exactly is the problem we would be
dealing with. The central problem, as I see it, is one of distin-
guishing between, and correlating, the concepts of meaning and
truth. In this context, many other questions otherwise interes-
ting though especially the more familiar ones regarding the
criterion of meaningfulness and the criterion of truth become of
secondary significance. What I propose to do in this paper is to
ascertain, in the first place, the precise nature of these concepts;
then, to bring out certain essential differences between them;
and finally, to throw some light on the way they function to-
gether in the total structure of human knowledge. Any enter-
prise of this nature ought to set out, if it is not to be dogmatic, by
considering the many fundamental objections that modern phi-
losophy and semantics have to offer. There is, for example, the
most fundamental objection that there are no philosophical pro-
blems of meaning and truth at all. This objection in its various
forms has been considered by me elsewhere,l so that I would
prefer not to return to it now. One has also to make initially
clear one's attitude to the linguistic philosophers, especially
to those who seek to replace the concept of meaning by the
concept of use, and those who make 'truth' a meta-linguistic
predicate of sentences in the object-language. On both these
contentions I shall have occasions in this paper to say a few

* Read at the first regional Seminar of the Centre of Advanced Study in Philosophy,
Visva·Bharati University, Santiniketan in April, 1964; and published in the Visva·
Bharati Journal 0/ Philosophy, Vol. I, 1964, No. 1,9-14.
1 See Meaning and Truth·I of this book.
MEANING AND TRUTH-II 5I
words. For the present I would rather begin by attempting pre-
cise formulations of the two key concepts of this paper.

I
It is of course obvious that "What does the word 'cat' mean?"
is not a philosophical question. Nor is "What does the sentence
'The cat is on the mat' mean?" a philosophical question. One
answers these questions if one knows the English language, and
one expects no special philosophical ability for this purpose.
What, then, is philosophical about the question "What is mea-
ning?"? It has also to be admitted that there is no meaning as
such and that meaning is always meaning of ... , so that the
philosophical question "What is meaning?" should really be
"What is (the) meaning of ... ?". It is indeed difficult to see why
"What is the meaning of 'cat'?" is not, but "What is the mea-
ning of ... ?" is a philosophical question. Asking "What is the mea-
ning of ... ?" is, of course, asking the meaning of an expression.
What, then, are we asking when we are asking philosophically
the meaning of an expression? The peculiarity of the situation is
that as soon as we specify the expression whose meaning we are
asking for the question ceases to be philosophical. This consi-
deration shows that the two questions - "What is the meaning of
'cat'?" and "What is the meaning of an expression?" - do not
have the same logical nature. Nor can it be said that the former
alone is a genuine question whereas the latter is a mere dummy
question, in the same way as a mere proposition functional is not
a proposition but only a dummy.
We may say that in asking "What is the meaning of an ex-
pression?" we are asking about what constitutes an expression
qua expression. Understood in this sense, the question may be
construed as being about the meaning of 'an expression qua ex-
pression'. We are indeed asking about the essential and con-
stitutive functions of expressions. We are not expecting to be
given a meaning in the same manner in which one gives the
meaning of 'cat'.
The philosophical problem, therefore, is to find out the es-
sential functions of expressions qua expressions and to enquire
into the conditions of their possibility.
52 MEANING AND TRUTH-II

Expressions are, of course, used to refer, or to perform many


other jobs, but they can be used because they are already at
hand as expressions, i.e., as meaningful. Their meaningfulness is
not an epiphenomenon of that use, but their usability rests on
their meaningfulness. Use is no doubt intimately connected with
the determinate job they perform in determinate situations.
Use, and all its attendant pragmatic factors, for example, make
possible the uniqueness of reference. An expression as such, quite
apart from its use, remains at the level of generality. Even if it
intends (or tends) to refer, such reference does not quite become
fully determinate. What constitutes, then, the essence of ex-
pressions qua expressions is the generality of meaning.
It is sometimes said that an expression expresses its meaning,
as if it were an expression prior to its expressing the meaning.
This, however, is false. An expression is in inseparable unity
with its meaning, so much so that in meaningful discourse they
are not even distinguished. A word as a mere physical event
(or object) is, of course, different from the meaning it is used to
communicate. But an expression qua expression is not to be
identified with that merely physical event (or object). The dif-
ference would be clear if only one contrasts the awareness of a
word as a physical event and the apprehension (or use) of it
meaningfully. In the latter attitude the outer sensory perception
(that constitutes the former) still remains but it enters into an
inalienable unity with an intellectual non-sensory act of a ra-
dically different kind. Such an inalienable unity is not merely
associational. In a merely associational unity the associated
elements are nevertheless distinguished. In the unity of a word
with its meaning there is a non-apprehension of distinction or
bhedagraha.
Philosophers err here when they start with bare words considered
as mere physical events (or objects) and seek to derive their
subsequent meaningfulness as a function of their use according
to rules. There is no doubt a sense in which words become mea-
ningful through such a process. Use and association determine
which symbols are to be tagged on to which meanings, but they
neither manufacture the meanings themselves nor do they suffice
for a post-mortem analysis of the sort of unity that comes to
obtain in effect between the word and its meaning.
MEANING AND TRUTH-II 53
Regarding the constitution of expressions qua expressions,
then, we find the following aspects which are welded as it were
into an organic unity:
(a) From the linguistic point of view, an expression belongs to
a language system, i.e., to a system of other expressions with
which it is related by conventional rules governing their use,
rules of formation and transformation. These rules determine
the linguistic meaning of an expression, which is what we seem
to be giving when we give its meaning in words.
(b) The linguistic meaning, however, is not meaning proper.
The real contentual (inhaltliche) meaning which is prior to lin-
guistic meaning and is the real determinant of linguistic synony-
mity is the general sense by virtue of which we succeed (i) in
breaking through linguistic relativity as evidenced from success
in translation from one language to another, and (ii) in trans-
cending the linguacentric predicament by opening the door to
the extra-linguistic world.
(c) It is this contentual meaning which makes reference pos-
sible and determines the mode of reference. Language provides
us with demonstrative symbols and mechanisms like definite
descriptions for the purpose. The referring function is essential to
expressions, though (i) the intention to refer may not be fulfilled in
experience, and (ii) the reference, in any case, must have to stop
short of being unique unless pragmatic factors are taken into
consideration.
(d) Subjectively, the apprehension, or understanding of ex-
pressions qua expressions amounts (i) in relation to (a) above, to
the ability to use them in accordance with the rules of the lan-
guage system concerned, and (ii) , in relation to (b) above, to a
grasping of the meaning.
Of these four, (c) is a derivative function of (b). Reference is
made possible through meaning. (Even mere ostentation cannot
refer, unless it is tagged on to a system of meaningful discourse.)
In that case (a), (b) and (d) are the three main aspects of an
expression qua expression. These are the linguistic, the conten-
tual (or intensional), and the subjective aspects. It seems to me
that three possibilities are open before us. We may start with
making the first, i.e., the linguistic aspect basic and reduce the
others to it. The success, however, of such a venture is likely to be
54 MEANING AND TRUTH-II

deceptive, for the language (the word, the sentence, etc.) which
it takes as basic is already charged with meaning and is not a
mere physical event. We may likewise take the intension, the
meaning, to be basic and reduce the others to it. Expressions
become accidental vehicles of eternal meanings, and the sub-
jective acts become the means of grasping them. The danger is
that thereby we hypostatise meanings and cut off their moorings
in language on the one hand, and in the subjective, on the other.
Subjectivism is a third possibility - reducing meaning and lan-
guage to their essential origin in the subjective acts of spon-
taneous, interpretative, creative thinking. Its risk lies in a pos-
sible blindness to the aspect of receptivity, i.e., to the objective
restraint that is experienced in operating with a given linguistic
system and in the apprehension of meanings. Thus, all the three
systems would be one-sided and would err at some point. In
order to grasp the essential structure of expressions qua ex-
pressions we have to fix our attention on the total nexus of
phenomena in their inalienable unity. The three aspects supple-
ment one another.
The doctrine of the ideality of meanings has to be understood
in this enlarged perspective, if we are not to be guilty of hypo-
statisation. The ideality is not an original one. Meanings are not
self-existent. They do not constitute an ontological region of
their own. Their ideality is derivative. They presuppose both a
linguistic system and the spontaneity of the subjective act of
thinking. At the same time such is the very nature of human
subjectivity that what it generates it also receives as passively
pre-given. This gives rise to the transcendental question - how is
it possible for a real personal subject to generate meanings that
are ideal, impersonal and objective?
The ideality of meanings is purely phenomenological. Of them
it holds good that they are as they are given. They are in that
sense pure phenomena. They do not point to a noumenon whose
phenomena they are. Understanding captures them - if it does this
at all- in toto, not in aspects or perspectives. In this last respect
they are like sense-data. But they are unlike sense-data in so
far as the sense-data point to the perceptual object as their
substratum which is presented through them. Meanings, no
doubt, make reference to reality possible, but this reference is
MEANING AND TRUTH-II 55
not to be construed as the reference of a phenomenon to its
noumenon. The analogy, often stressed, between perceptual
apprehension of sensible objects and intellectual apprehension of
meanings breaks down here.

II

One way of determining the nature of the concept of truth is


to ask: of what is 'true' predicated? This, as we know, is answered
in many different ways. I would outright reject the suggestion
that it is predicated of sentences: I think Strawson's arguments
against this view are decisive. The other claimants for this po-
sition are: propositions (Russell), statements (Austin), beliefs or
judgments, and knowledge.
We may be on the way towards a resolution of this controversy
if we start by recognising that each of these is in fact called true.
The next step would be to see that this fact that so many dif-
ferent things are called true is not an instance of the unrelia-
bility of ordinary language, but rather points to the many dif-
ferent dimensions - not unrelated to each other - of the concept
of truth. It would be worthwhile, therefore, to distinguish be-
tween some of these dimensions (or usages, if you like), and to
single out that one with which it would be fruitful to be con-
cerned in the present context.
Beliefs or judgments are either affirmations or denials. By
'proposition' we mean the content of the belief, that which is
affirmed or denied. The distinction may be brought out also
with the help of the fact that an identically same proposition
may be affirmed, denied, doubted, supposed, or even merely
entertained. When one affirms or denies a proposition and ex-
presses the affirmation or denial in a sentence, spoken or written,
one is said to make a statement. Making a statement is a pu-
blicly observable event, but for that reason one should not count
it as the only thing that is besides the sentence with which the
statement is made. A statement expresses one's affirmation or
denial. The latter are about a proposition. A proposition may
also be said to be the contentual meaning of a sentence, that which
in the last resort accounts for the synonymity of two or more
sentences.
MEANING AND TRUTH-II

Keeping these distinctions in mind we may now proceed to


exhibit the different levels at which the problem of truth makes
itself felt:
I. When one speaks of the truth of judgments one refers
chiefly to the truth (or falsity) of one's beliefs, iIi other words of
one's affirmations or denials. Affirmation is taking to be true;
denial is taking to be false. Yet an affirmation may be true or
false, so also may be a denial. This shows that truth (or falsity)
of beliefs or judgments presupposes another sense of being true
(or false). In this latter sense, truth (or falsity) belongs to
propositions which are affirmed or denied.

2. In this second sense, there are true propositions (e.g., the


propositions meant by the sentences "2+2=4" and "Snow is
white") as well as false propositions (e.g., propositions meant by
the sentences "2+2=5" and "Snow is black"). A true propo-
sition deserves to be affirmed, a false proposition deserves to be
denied. Designating a true belief by h and a false belief by h,
and a true proposition by t2 and a false proposition by f 2, we
may then say:
h consists in (a) taking t2 as t2
or (b) taking f2 as f 2, whereas
f1 consists in (a) taking t2 as f2
or (b) taking f2 as t2.
It is to be noted that fz as such does not amount to error. What is
erroneous is the affirmation of f2 or denial of t2. Again, it is
important to bear in mind that a proposition, t2 or f 2, need not be
affirmed or denied: it may be, as said before, doubted, presumed,
assumed, supposed, and finally merely entertained.

3. A statement is true only in a sense that is derivative from


the sense in which a belief is so, for in making a statement one
expresses the belief in a publicly observable manner. A true
statement may, then, be designated t'1 and a false statement f'1.
Thus, til presupposes t1 and h presupposes t2 and f 2.

4. In all the above senses truth is opposed to falsity. This


opposition is essential to these meanings of 'truth'. There are,
however, two other senses in which 'truth' is used without exactly
MEANING AND TRUTH-II 57
entailing such a contrast. In one of these senses truth belongs to a
knowledge. Whereas a belief, a proposition or a statement may be
either true or false, knowledge is only true and cannot be false.
A false knowledge is a contradiction in terms. Truth is an essen-
tial or a definitional property of knowledge. Thus the truth of
knowledge has no meaningful contrast with falsity. When a know-
ledge, or what claimed to be so, is seen to be false it ceases to be
knowledge. Its pretension is exposed.

5. The objective correlate of knowledge is reality, and that


also is often called truth. In this sense, again, truth has no mea-
ningful opposite. There is no falsity in reality. The quasi-objec-
tivity of falsity ends with propositions and does not extend
beyond it into the sphere of reality.
From amongst these various meanings of 'truth' it is ne-
cessary to single out that which is its primary sense. I wish to
argue that in its primary sense truth pertains to knowledge.
The fifth meaning of truth may be left out of account, for truth
is predicated of reality only as a transferred epithet, reality
being the objective correlate of knowledge that is true. It cannot
be said that there is a circularity involved here, for knowledge
is knowledge inasmuch as it is of reality. There is no circularity,
for knowledge is not true because reality is the truth, but on the
contrary reality is called the truth because it is the objective
correlate of knowledge that is true. It is also possible to argue
that the sense (r) is derived from the sense (4), for in the first
place belief may be regarded as falling short of knowledge and as
being true only in so far as its propositional content agrees with
that of knowledge. The ideal is knowledge where the true propo-
sition is not only judged (believed, etc.) to be true but insight-
fully judged to be so, i.e., where the intended meaning is realised,
fulfilled - if you like, verified - in intuitive evidence. Belief, even
when true, falls short of this, for it terminates in the proposition.
A true proposition is the meaning-content of both a true belief
and a knowledge, but knowledge does not terminate in it; know-
ledge is not of the proposition but of the reality whose intuition
fulfils or verifies the meaning-intention. Thus the truth of a
true proposition, and that of a true belief, derive their truth-
character from the truth of the knowledge into which they might
58 MEANING AND TRUTH-II

possibly enter and in which they are to be fulfilled. The primacy


of the truth of knowledge is also suggested by the interesting
fact that it alone - leaving aside the sense (5) - is absolute in the
sense of having no opposite in the strict sense. I believe that this is
the basis of the Mimamsa theory of svataltpriimiinya, the Nyaya
theory of parataltpriimiinya being based on an assimilation of the
concept of knowledge to that of belief which mayor may not be
true.
We may sum up the very complicated situation thus:
(I) True belief presupposes a true proposition. Hence the
truth of the former presupposes the truth of the latter.
(2) A true proposition is the content of a (true) knowledge.
Truth of the former presupposes that of the latter.
(3) Truth of knowledge on its own part does not depend upon
the truth of its propositional content but consists in the fact
that in knowledge reality itself is apprehended.

III

Now that the concepts of meaning and truth have been iden-
tified and located we may proceed to investigate the relation
that obtains between them. In any concrete knowledge situation,
such as when I know an object while at the same time either
naming or referring to it by an expression (as when I name what I
perceive as "This man here before me"), it should be obvious that
meaning and truth are together in an inalienable unity.
Meaning belongs to expressions and is grasped through an act
of understanding, howsoever such an act may be interpreted.
Truth belongs to knowledge. The difference and yet the inter-
relation between them is exactly what concerns us here. It must
be clear that the two represent two different attitudes. In the
attitude of mere understanding I merely apprehend the meaning,
and in so far as the meaning determines the mode of reference I
also apprehend the type of reference it is capable of being used
to make. But all this does not amount to knowledge. When, on
the other hand, I know, the general reference is made determinate
through intuition or experience; the object, and not the meaning,
is what is given as it is really in itself. In this sense, knowledge of
MEANING AND TRUTH-II 59
A as A presupposes understanding of the meaning of 'A'. In
this sense, meaning is presupposed by truth, but truth goes
beyond meaning. Whether truth can be had except through the
via media of meaning would depend upon whether or not we
have a mode of knowledge which is not linguistic. It seems,
however, to me - to quote Merleau-Ponty - that we are 'con-
demned to meaning'. The reference of knowledge to its object
presupposes the reference of expressions. The latter reference,
however, remains indeterminate. We can think of things that
we do not know, and can understand possibilities that exceed the
scope of actual knowledge.
Formal logic may be construed as a logic of meanings. It deals
with two levels of problems: first, the problem of demarcating
meaningful from meaningless sequences of meanings, and, next,
the problem of the relations of compatibility, incompatibility
and entailment amongst meanings. Accordingly, formal logic
falls into two strata: a logical grammar and a logic of non-con-
tradiction.
An incidental point needs to be emphasised. It concerns the
distinction between analytic and synthetic truths. From the
point of view of a theory of meaning, such a distinction is rele-
vant: there are true propositions which are true by virtue of
their component meanings, and there are those whose truth is
left undetermined by the relations obtaining among the com-
ponent meanings.
In a logic of knowledge, on the other hand, such a distinction
becomes pointless. For all knowledges are so far alike that their
truth consists in the apprehension of their objects as they really
are. It is only when the propositional content is abstracted from
the knowledge and is treated as a self-subsistent meaning that
the question of its analyticity or otherwise becomes significant.
This, I guess, would partly explain the absence of this distinction
in Indian logic which is a logic of knowledge rather than of
meaning.
CHAPTER VI

LANGUAGE & REALITY*

Of the many problems that come under the title of this sym-
posium, two stand out as the most important. These are: 'How is
language related to reality?' and 'Is language a suitable medium
for knowing reality?'. This paper shall have something to say
about each of these problems. However, each of these questions
reveals an ambiguity that is due to the ambiguity of the word
'reality'. By 'reality' is sometimes meant real things, events,
facts and persons which go to constitute what we in common
parlance call the real world. But 'reality' is also sometimes,
especially in metaphysical discourse, taken to mean ultimate
or metaphysical reality in which case it denotes something that
stands behind and beyond the world of things and persons which
is but its appearance. One who asks the first question, namely
'What is the relation between language and reality?' may be
asking 'How is language related to the real things and persons
which constitute the real world?', or he may be asking 'How is
language related to the ultimate metaphysical reality, to the
Absolute or Brahman ?'. Similarly, the second question may mean
either the same as 'Is language a suitable medium for knowing
the nature of the empirical world of things and persons?' or the
same as 'Is language a suitable medium for apprehending the
nature of the ultimate metaphysical reality?'.
Now of these two sets of questions it seems extremely dif-
ficult to tackle those concerning ultimate metaphysical reality.
In order to be able even to make a start with them one should
have a conception of that reality, which raises a metaphysical
problem of the highest order and importance. The term 'meta-
* Read at the Centre of Advanced Study in Philosophy, Banaras Hindu University
in November, 1965.
LANGUAGE & REALITY 6r

physical reality' which is not one in ordinary use has to be suit-


ably defined, and it seems our approach to the questions would
largely depend upon how we define this key term. But there
are at least two difficulties here. In the first place, how can we
discuss the relation between language and reality, if giving a
meaning to 'reality' is a matter of linguistic decision? What one
needs in order to gain a vantage point for commencing the dis-
cussion, is to get beyond language to some non-linguistic ex-
perience so that it may be worthwhile attempting to correlate
that to language. The conception of a metaphysical reality does
not provide me with such a vantage point. Secondly, dealing
with metaphysical reality, I am afraid, any answer I could
offer would be trivial in the sense of following from the definition
of 'reality' with which I start. I might so define 'reality' that
the real is ineffable, and then to contend that language could
have no meaningful relation to it, or that it is inaccessible to
language would be an analytic consequence of that initial de-
finition. It is because of such suspicions that I shall refrain from
dealing with this aspect of the problem.
Turning then to the world of things and persons, to the real
world as we call it, and its relation to language we find our task no
easier. The most difficult and also the most important step in
philosophical thinking is a clear formulation of the problem;
nowhere is this a more difficult task than in the present case. It
was pointed out a little while ago that in dealing with meta-
physical reality we run the risk of arriving at a conclusion which
is but an analytic consequence of our initial definition of 'reality'.
But do we not run the same risk even when we are dealing with
the real world of things and persons? Consider, for example, the
thesis advocated by many that the world consists in facts (not
things and persons). Now what are facts? They are roughly
definable only with reference to sentences. A fact is what makes
a sentence true. And now to contend further that there is a
relation of correspondence, accordance or picturing between
language and reality is but an analytic consequence of the way
the thesis looks upon reality as consisting in facts. Certainly if
you look upon the world to consist not in facts but events, things
and persons, you cannot contend that language pictures them,
it can at best describe them.
62 LANGUAGE & REALITY

It seems to me therefore that the true problem is the relation


of language not to reality but to our experience of the real.
We call it formulation (I). What we are to call 'real' may to some
extent depend upon a linguistic context, not so however our
experience of the real. It may however be argued against this
contention that there is no experience of the real that does not
presuppose a linguistic framework, that there is no merely given
that is not conceptualised and therefore not subject to the
interpretations flowing from a linguistic system. The crucial
question then is, is there a non-linguistic apprehension of the
real? I say 'crucial', for upon an affirmative answer to this
question depends the meaningfulness of the very problem we are
out to discuss.
The contention that what is to be regarded as given depends
upon one's linguistic context is closely connected with two other
theses: first, that the problem of existence is a linguistic problem,
and secondly, that proper names are not indispensable for a
language. 1 Taken together they liquidate ontology, or rather
assimilate it to a philosophy of language. An assimilation of the
one to the other would not do: that would mean a liquidation of
our problem.
I have said above that the true problem is the relation of
language to our experience of the real. At this point it may be
asked, does not language refer to the real, and if so is it not worth
while investigating the nature of this referential relation? To
this I would reply that our problem is in one sense concerned
with the referential relation, in another sense it is not. Let me
explain my contention.
I t has been well established by modern researches in theory of
meaning (Frege, Husser!, Quine) that one of the essential func-
tions of language is to refer. This function of referring is different
from the other function of meaning. It has been held that the
meaning function is an immanent function of language in the
sense that it derives from the rules of formation and transfor-
mation of a language, while the referring function relates lan-

1 Cpo R. Carnap, Empiricism, Semantics, and Ontology included in Meaning and


Necessity, Enlarged edition, I9S6; and W. V. O. Quine, Word and Object, M.LT.,
I960, especially 37.
LANGUAGE & REALITY

guage to extra-linguistic reality. Whether meaning (even as dis-


tinguished from reference) is purely linguistic or whether one
can speak reasonably of a non-linguistic meaning is a question
into which I need not investigate at present 2 • With regard to
the function of reference however it ought to be admitted on all
hands that language refers to a real beyond, so that it would not
do to reduce even this into an intra-linguistic relationship3.
Assuming all this to be well-founded conclusions, we may still
deny that the semantics of reference is all that concerns us when
we are concerned with the relation of language to reality. The
following consideration will bear this out. The reference that is
intrinsic to language and that constitutes an essential function
of language qua language is merely intended reference, and does
not by itself carry a guarantee that the reference must find a
foothold in reality. The reference does not provide language
with a point where it can be hooked on to reality. In believing
that it does, Russell was wrong. Russell was wrong in thinking
that the sense or meaning of a name is its designatum, as also in
thinking that meaningful use of names carries with it the guaran-
tee that there must be in reality things which bear those names.
Wittgenstein - as Anscombe has rightly pointed out - gave co-
gent reasons for rejecting this view of Russell,4 and these need
not be repeated here. Language qua language refers (and need
not picture or even designate, for referring is not always naming),
but whether the intended reference is fulfilled in reality, i.e.
whether there exists in reality the thing that is being referred to
is quite another question. Thus the relation of language to reality
is not established through the referring function which belongs
to the former. If the question is, how is language hooked on to
reality, and if reference fails to do this job, we may further
ask, 'What completes the unfulfilled task, and finally nails
language to reality?'
My purpose so far has been to formulate the aspect of the
problem which seems to me to be of the utmost importance. The
vague question 'What is the relation between language and
2 For further elaboration, see my "Meaning and Truth", Visva Bharati Journal
of Philosophy, Vol. I, NO.1 (reprinted in this volume).
3 Cpo my Edmund Husserl's Theory of Meaning, The Hague, 1964, ch. II.
4 G. E. M. Anscombe, An Introduction to Wittgenstein's Tractatus, London, 1959,
PP.45-46.
LANGUAGE & REALITY

reality?' has now been refined and narrowed down to 'What


besides and in addition to the function of referring intrinsic to
language qua language hooks language to reality? ... (za).
But we should not forget that there is also a reverse side of
this question and that may be formulated as 'What makes reality
a referent of language?' ... (zb). Whereas (za) is formulated from
the side of language and asks how it is tied to reality, (zb) is
formulated from the side of reality and asks how reality comes
to be the referrent of language. It should be noted that the
second is the more fundamental question. It enquires into the
very possibility of language. It is in fact a part of the question
'How is language qua language possible?'
Taking up now the question (za) along with the earlier pro-
posed question (1), we may formulate the resulting problem as:
'How is the intended reference of a linguistic expression com-
pleted, fulfilled and integrated into a complete experience of
the real?' (3). Or, what amounts to the same, 'How is the gap
between linguistic reference and bodily apprehension of the
real overcome, so that there comes about, so to say, a perceptual
judgment, a sort of felt non-distinction between language and
reality, between the name and the named, the sentence and the
fact?' (3').
In (3') I have spoken of several things which need to be com-
mented upon. In the first place, it has been suggested that there
is a gap between linguistic reference and bodily apprehension of
the real. It is necessary to be aware of the full implications of
this. Language, of course, qua language refers, and the reference
is to a reality beyond. But, besides the fact made out earlier that
the reference is only intended reference, there is also the sup-
porting consideration that understanding this reference is not
the same as knowing the thing being referred to. Talking about
an object or understanding the talk about it does not amount to
apprehending that object excepting as the object that is being
referred to. The existence of the gap may be further exhibited
thus: linguistic reference is always through the medium of mea-
ning. Language does not contain an unmeaning mark. Meaning
has an unavoidable aspect of universality and Uberzeitlichkeit.
The real however when it is given as real is given as something
unique, as a this-there. Even the indexical expressions which of
LANG U AGE & REALITY 65
all are the most completely referring have an aspect of meaning
as distinguished from reference, and do their job of referring only
through meaning what they mean. There are no logically proper
names in Russell's sense. Now this condition that linguistic
reference is determined by the medium of meaning also
imposes on it certain limitations. It does not achieve that
absolute determinateness and uniqueness which belong to the
real referent. This gap has to be filled in. That the gap is in fact
overcome is a conditio sine qua non of the possibility of our lin-
guistic identification of the reals. But how does that become pos-
sible? The question then is not, 'How is it that language comes to
refer beyond itself?' (for this reference is a consti tu ti ve aspect of
language qua language), but 'How is it possible to make lin-
guistic identification of a unique real ?'.
The existence of the gap may be denied by saying that there
is in fact an inseparability between language and reality. This
latter contention may be supported by an appeal to the alleged
fact of sahopalambha, in other words to the fact that a real and
its name are never experienced separately. This fact may be
taken to entail that the two are inseparable and so non-distinct.
If this conclusion holds good then it would further follow that
there can be no non-linguistic apprehension of the real, for the
real qua real is, on this view, a designatum of such and such ex-
pression. Such a point of view will also liquidate the problem as
formulated in (zb), for if to be real is to be the referent of an
expression it is pointless to ask how does reality come to be
referred to by language. It is pointless then to ask, how is lan-
guage possible. To vindicate the possibility of the questions we
have asked it is necessary therefore to expose the hollowness of
this theory.
As regards the alleged fact of sahopalambha I think it to be
based on several confusions. In the first place, there is a con-
fusion between mentioning an expression and using it. A name
may be mentioned without referring to the named object: there
is then no sahopalambha. Secondly, even when a name is being
used it is one thing to use a name to refer to the named and
another to know the named. Mere use of the name does not carry
a guarantee that there is an apprehension of the real named
excepting as so named. Even if we leave open the question whether
66 LANGUAGE & REALITY

the named can be apprehended without apprehending its name it


seems indisputable that a name can be apprehended without
there being an apprehension of the named.
Although - as argued above - language and reality are sepa-
rated by a cleavage, yet as our formulation (3') implies, there is
in perceptual judgment a sort of felt non-distinction between the
two. A sort of identification is achieved in consciousness. There is
no ontological identity. This non-distinction cannot be explained
as a mere associative synthesis, for in association the associated
elements are clearly distinguished.
The problem then concerns briefly the relation of language to
perception, for it is in perception that real things and persons are
bodily presented. It seems to me that this is a far more important
problem than the much discussed one concerning the relation of
language to thought. For it is perception - and not thought -
which gives us a sense of reality of its object. The fact that per-
ception is sometimes delusive does not go against our contention.
No wonder that those who believe in a metaphysical reality and
hold both that it is knowable and yet not knowable by thought
hold the view that thought leads to a higher, intellectual intuition
or perception. The analogy with perception is quite understandable.
But the problem of the relation of language to perception of
reality itself presents not one question but a number of them.
The non-distinction to which reference was made a little while
ago may be studied as a static unity achieved in savikalpa per-
ception, or as the limiting point of a dynamic process through
which the gap is progressively overcome. Consider the simple
perceptual judgment 'This is a pen' in which the linguistic
component with its meaning and referential aspects stands non-
distinguished from the bodily apprehension of a real pen before
me. It is necessary to bear in mind that the apprehension itself
is not linguistic, as it is sometimes erroneously supposed to be.
Even on the Nyaya theory what distinguishes savikalpa per-
ception from the nirvikalPa is the presence of certain epistemic
contents like vise$yatri and prakarata and also the fact that the
latter cannot be (avyapadesya) while the former can be expressed
in language. 5 It however stands in a close relation with language,
5 Even on the older view which regarded savikalpa as in some sense linguistic,
the padaviicyatva is said to appear only as upanita. Thus, "visi$taifiiine tattatpada·
LANGUAGE & REALITY

and I do not think the N yaya is very clear as to the nature of


this relationship.
Perhaps it is possible to throw some light on this mystery by
studying the nature of the graduated synthesis that leads up to
the felt non-distinction. For this purpose let us make a further
change in attitude. We transformed the original question 'What
is the relation of language to reality?'. This transformation in-
volved a change of attitude from the objective and ontological
to the subjective and phenomenological. It was sought to be
justified on the ground that the real is what we confront in
perception. It is necessary, in order to complete the transfor-
mation, to make a similar change in our attitude also with re-
gard to the other term of the relation, namely 'language'. Let us
ask the question not about language but about our experience
of it. Once we do so then the entire problem gets transformed into
the subjective mode. We ask, namely, 'How is the experience of
language related to the perceptual apprehension of reality?'.
In other words, we ask, 'How is it that in perceptual judgment
language and reality both come to achieve a sort of felt non-
distinction?' (4).
But how is language experienced? What is the specific and ori-
ginal mode in which language qua language is given? Language
qua language is certainly not given in outer perception alone.
Of course, the speech is to be heard, or the written words are to
be seen. But auditory and visual perception, as the case may be,
does not constitute the mode of givenness of language qua
language. If it is the function of meaning that transforms the
merely physical shape or sound into an expression, then to appre-
hend language as such is to understand it or to use it meaning-
fully. The outer perception provides the basis on which is built
up the higher intellectual act of understanding.
Our question (4) may then - in view of the fact that we are to
examine graduated synthesis and not the static unity - be re-
stated thus: 'How do we pass on from a bare understanding of
an expression to a bodily identification of the real referred to
byit?'·(5)
Strawson refers to a class of expressions which may be re-
vacyatvamupanUam bhasate iti pracinamatanusaritvat" (Dinakari on Muktiivali on
Kiirikii 51).
68 LANGUAGE & REALITY

garded as capable of identifying uniquely the things referred to


by them. These are what he calls 'pure individuating descriptions'. 6
An example is 'The first dog to be born at sea'. Now if such
expressions can do what they are taken to be capable of doing,
then an answer to (5) may be found at least in some cases, on a
purely linguistic level. For, in that case an understanding of a
pure individuating description would be sufficient to lead us
to an identification. But this again is not the case. First, how
am I to make sure that there is one and only one individual
satisfying the description? There is, in Strawson's words, the
possibility of a massive duplication. Secondly, even if I know that
there is an individual satisfying the description, knowing this
does not amount to knowing the individual itself. What is re-
quired is a bodily identification of the uniquely real. And this
is what language by itself would seem surely incapable of doing.
Taken as an objective expression there can in fact be no non-
distinction between linguistic expression and a uniquely real
referent. The objective expression by virtue of its generality of
meaning overflows its referent on any given occasion. It is
only the subjective speech which achieves the non-distinction,
which is felt as non-distinct from the fact being referred to.
Perhaps it is here that we may look for an answer.
Identification requires a framework in which to locate, and a
field from which to pick out. Mere non-linguistic experience
cannot by itself identify. It offers a field from which to choose.
But the perceptual field is always relatively vague; it has a
fringe and a core of indeterminateness. Language confers deter-
minateness. Linguistic description narrows down the field. Speech
as distinguished from objective language, offers not only a frame-
work but also an absolute point of reference namely the speaker
himself. Unless there were a speaker the supposed identification
of language with reality being referred to would not have taken
place. It is the speaker who uses language to identify: the result
is a non-distinction in consciousness.
It should follow that the non-distinction is more where speech
confers more of determinateness in the perceptual field. Con-
trast the perceptual judgment 'This is red' with 'This is magnificent' .

6 P. F. Strawson, Individuals, London, I96I, p. 26.


LANGUAGE & REALITY 69
Though in the former case there is a felt non-distinction between
the linguistic expression and the fact being referred to, yet it is
also the case that the fact of this being red is a fact indepen-
dent of my or any one else's saying so. But in the latter case the fact
of this being magnificent is a fact only in so far as it is being said
so by some speaker or other. Here in a sense speech makes the
fact so. Josef Konig, following Brentano, distinguishes between
'determining' and 'modifying' predicates. 7 I am not sure if an
absolute line of distinction can be drawn between the two kinds
of predicates. It may be difficult to decide whether a given
predicate is determining or modifying. The fact remains however
that the non-distinction between speech and apprehension of
reality may be achieved in degrees, depending on the degree of
indeterminateness of the perceptual field and the degree of
determinateness conferred by speech on it.
Communication is not the only purpose of speech. That no
doubt is one of its chief uses. Communication is possible be-
cause of the generality of meanings, because speech, in spite of
its dependence on the speaker, conveys meanings that are im-
personal. But it would seem language has another important and
perhaps more basic function which consists in locating and iden-
tifying reals. It is one thing to have an experience. But if we are
to identify the thing experienced and give it a place in a scheme,
language is an indispensable aid. In this sense language is like a
map. One may know a place, and yet cannot locate it except
with the help of a map. Language provides us with a framework
in which and in whose terms to locate and identify a real. It is no
wonder that sometimes a map is felt as non-distinct from that
which is mapped.
By saying all this I do not intend to suggest that language
is a picture of reality in any sense. By emphasising the fact that
language provides us with a framework I wish to bring to the
forefront the aspect of it as a pattern as against the mere voca-
bulary aspect. This I believe is in conformity with the findings
of such linguists as Humboldt, Whorf and Chomsky.8
Logical atomism believed that both language and reality

7Josef Konig, Sein und Denken, Halle, 1937.


8A. V. Humboldt, Menschliche Sprachbau; B. L. Whorf, Language, Thought and
Reality, 1956; and N. Chomsky, "Current Issues in Linguistic Theory" in Fodor and
Katz, The Structure of Language, Prentice-Hall, 1964.
70 LANGUAGE & REALITY

could be analysed down to the simple, not further analysable


elements, and it further held the thesis that there is "some real
and non-conventional one-one picturing relation between the
composition of the expression and that of the fact". 9 This the-
ory fortunately is a dead horse and it would not be worth while
flogging it any more. Few remarks in this connection may
however be made here.
Meaning of expressions may be simple or compound. But the
simplicity or complexity of an expression does not show whether
its meaning is simple or compound. Further, we are in no better
position in trying to infer the simplicity or complexity of the
object being referred to from the simplicity or complexity of the
meanings through which reference is being made. For example,
'simple object' which is a compound expression and has a com-
pound meaning refers to a simple object. Nor is there always a
one-to-one relationship between a compound meaning and the
compound object referred to through it. Bolzano gives the example
of 'The country without a hill'. In fact the entire talk about
simplicity or complexity is equivocal. It may mean simplicity
or complexity of meanings, or it may mean simplicity or complex-
ity of our consciousness of meanings. It may be that a simple
meaning is apprehended in consciousness through many phases
and stages. Neither expressions nor meanings therefore picture.
Expressions refer and meanings make reference possible by im-
parting to it a certain determinateness. 1o From what has been
said it follows that language is neither the same as reality nor a
picture of it. Nor is our consciousness, or even our experience of
reality necessarily linguistic. But language with both its voca-
bulary and patternment aspects is a necessary, though not suf-
ficient,n precondition of our identifying a real. It in fact pro-
vides us with a coordinate system in which to locate the given,
so that the real is apprehended as having such and such linguistic
designation. This is to know par excellence. In this sense of 'know',
to know A is to know it as A which entails, amongst other things,

9 G. Ryle, "Systematically Misleading Expressions", Proceedings of the Aristotelian


Society. 1931-32.
10 For further elaboration, see my Edmund Husserl's Theory of Meaning, Ch. V.
11 The problem has been further developed in my "On Reference" in Argumen-
tationen, Festschrift fur Josef Konig, Gottingen, 1964. (Reprinted in this volume)
LANGUAGE & REALITY 7I
knowing it as having the name A. However, knowing a thing
as having the name A is not the same as knowing an A.

Heidegger says that the essential function of language is to


manifest Being to man; it opens up Being for man,12 Without
taking a stand with regard to Heidegger's peculiar conceptions
of Being and openness, we may still regard our thesis as a partial
substantiation of his thesis though on a more modest and com-
mon sense level.

If now we turn to the question 2 (b) and ask, 'How is it that


reality comes to be a referent of language?' and thereby ask,
'How is language qua language possible ?', we seem to abandon
the field of phenomenology to which we had confined ourselves
so long and to have entered the precincts of speculative meta-
physics. On so large a question, which should not be confused
with a genesis of language, we may make only a few, very un-
satisfactory, remarks. In the first place, it may be said that the
referentiality of language presupposes the intentionality of
consciousness. The latter does not depend upon the former but
makes it possible. In fact, linguistic reference is one mode in
which the intentionality of consciousness is objectified. But to
the question why after all a real should come to have desig-
nation this is not the answer. An answer, if there be any, may be
sought in either of two directions. We may either find it in some
need of reality to be made articulate, or we may find it in the
nature of human reason i.e. in man's specific mode of relation-
ship to the real. The former shall lead us to a sort of mysticism
of the later Heidegger, the latter to a more truly existential
philosophy. To choose between these two possible answers, or
even to show if they can be synthesised would be going beyond
the limits of the present paper.

12 M. Heidegger, Ein/uhTung in die Metaphysik, Tiibingen, 1953, p. 131; and his


letter ObeT den Humanismus, Bern, 2nd edn. 1954.
CHAPTER VII

ON REFERENCE*

I. The problem of reference, and that of meaning, first come


into clear relief and exhibit their problematic character, only
when one has already drawn an initial distinction between mea-
ning and reference. If one fails to make this distinction and there-
by implicitly identifies the two, both the problems get watered
down to how words or expressions could after all stand for some-
thing, - a question to which conventionalism provides a fairly
convincing answer. The problem however is much deeper, and
for an appreciation of it the Frege-Husserlian distinction be-
tween meaning and reference is an indispensable starting point.

I.I. There are many ways of introducing this distinction. The


best seems to be by way of showing (a) that two expressions may
have different meanings and yet an identical reference; (b) that
an expression while retaining its meaning unchanged refers now
to one object, now to another; and (c) that an expression does
not suffer any diminution or damage in its meaning from the
fact that its reference fails. Examples for (a) are 'The morning
star' and 'The evening star'; for (b) are the general names like
'man'; and for (c) such cases as 'The present King of France'.
The most plausible explanation of such possibilities is provided
by the assumption that the meaning of an expression is different
from that which it refers to. Once the point of this distinction is
seen, one also begins to realise the reality of the problem about
meaning: 'What then is the meaning of an expression, if it is not
the thing being referred to?'.

* Published in Delius and Patzig (ed.), Argumentationen, Festschrift fur Josef


Konig, Gottingen, Vandenhoeck & Ruprecht, 1964, pp. 159-169.
ON REFERENCE 73
1.2. The problem of reference cannot however be felt with its
full force and with all its implications even at this stage. The
question about meaning has to be pursued a few steps further
before we could be in a position to formulate the problem of
reference which is to be the chief concern of this paper.
There are several answers to the question 'What is meaning?'
that suggest themselves to us immediately. One of these has
been explicitly rejected by Frege on good grounds, and has fallen
into disrepute; another continues to enjoy unabated reputation.
The first is the theory that the meaning of an expression is
nothing but an image, against which Frege says: "Die Vor-
stellung ist subjektiv: die Vorstellung des Einen ist nicht die des
A ndern. . .. Die V orstellung unterscheidet sich dadurch wesent-
lich von dem Sinne eines Zeichens, welcher gemeinsames Eigen-
thum von V ielen sein kann und also nicht Theil oder Modus der
Einzelseele ist."l An image is private, subjective, i.e. in some-
body's mind and therefore is not communicable or shareable.
The meaning of an expression however is in an important sense
public, objective, not a real constituent of anyone's mental
history; it is communicable and shareable. Meaning therefore
cannot be an image.
The other theory holds that since meaning has been distin-
guished from reference, the question 'What is meaning?' does
not intend us to produce an entity, be it subjective (an image) or
objective (a subsistent entity). All questions about entities
being transferred to the theory of reference, meaning as distin-
guished from reference could only be linguistic (sprachliche) mean-
ing. This theory finds support in the undeniable fact that when
asked to give the meaning of an expression one produces not a
non-linguistic entity but another synonymous expression. Mean-
ing therefore is linguistic. The problem of meaning is really
one of synonymity.
As against this, I can do no better than quote the following
passage from C. I. Lewis:
, 'We must express meanings by the use of words; but if meaning
altogether should end in words, then words altogether should

1 G. Frege, "Dber Sinn und Bedeutung"., Zt. /. Philos. Kritik, NF :roo (:r892), p.
29. Now also in G. Frege, Funktion, Begriff, Bedeutung, G6ttingen, 1962, p. 42.
74 ON REFERENCE

express nothing. The 'language system' as a whole would 'have no


interpretation' and there would be no such fact as the meaning
of language." 2
The idea of linguistic meaning is unobjectionable, but it cannot
be self-explanatory. It ultimately does refer back to the ma-
terial (sachliche) meaning. This is clear from the recent attempts
of account for synonymity on a purely linguistic and conven-
tionallevel. Convention of course does account for the fact that
any two expressions should have been used as synonymous, but
that they are so means only that they have been made to ex-
press the same meaning.
The meaning cannot also be the same as the use of the ex-
pression in a language game. As Wittgenstein himself rightly
realised 3 , the concept of use is the concept of something extended
in time, while there is a sense in which apprehending the meaning
of an expression is not an act that is extended in time but an in-
stantaneous apprehension. And it is this sense of meaning with
which we are concerned in the present context.
I.3. Any acceptable theory of meaning, it seems to us, should
be able to satisfy two criteria: in the first place, it must be able to
account for that universality, objectivity and identity of meanings
which are the pre-suppositions of all communication of thought, and
therefore of all linguistic behaviour. What Husserl calls the ideal-
objectivity of meanings is an undeniable phenomenological
datum. Those who deny it do so only under the pressure of their
preconceived metaphysical prejudices. Maybe this is only one
aspect of the phenomena. However, into the question how far
and in what manner the phenomenon of the ideal-objectivity
of meanings could be harmonised with other aspects of the total
phenomenological situation we need not, and cannot enter here 4 .
The second criterion to be fulfilled by a satisfactory theory of
meaning follows from the fact that meaning also serves as the
medium for reference. The possibility as well as the limitations
of this function have to be exhibited. It is to this aspect of the
problem that we shall now turn our attention.
2 c. 1. Lewis, An Analysis of Knowledge and Valuation, Illinois, 1946, p. 140.
3 L. Wittgenstein, Philosophical Investigations, Oxford, 1953, p. 53.
See ]. N. Mohanty, "Meaning and Truth", Presidential Address, Proc. Ind.
4
Phi/os. Congress, 36th Session (1961), Santiniketan, pp. 27-47. [Reprinted in this
volume]
ON REFERENCE 75
I.4. The problem of reference is falsified when we regard re-
ference as an intra-linguistic function as does modern semantics
when it treats names as designating other expressions. This is
not to cast reflection upon the value of semantics as a science,
but this clearly implies a serious limitation of its philosophical
significance. For the problem of reference is, how in the first
place within linguistic discourse we succeed in making extra-
linguistic reference; and secondly, how does it come about that
we also succeed very often in making unique extralinguistic
reference although the medium of such reference is provided by
the general ideal-objective meanings.

2. One of the contentions of the present paper is that ex-


pressions that are names possess, as such, the referring function.
Connected with this is the contention that in a very important
sense the problem of extralinguistic reference is illusory inas-
much as such reference is an essential function of language qua
language. Language would not be language and would not serve
one of the chief purposes it is meant for if it did not refer to ex-
tralinguistic objects. The situation is analogous to - and in fact
not unconnected with - the problem concerning the possibility
of our consciousness of transcendent objects. How can conscious-
ness, it is asked, apprehend objects that are beyond it? Here
again part of the sting of the problem is gone once we realise that
such self-transcending reference belongs to the very essence of
consciousness, and asking how consciousness could after all
perform this miracle is the same as asking how consciousness
could be consciousness. That language should succeed in making
extra-linguistic reference presents a similar situation, and phi-
losophy could hope to achieve only limited results, if any, here,
for the ultimate character of this fact has to be first and foremost
acknowledged.
The fact remains nevertheless that such extra-linguistic re-
ference is effected through the medium of ideal-objective mea-
nings, and this perhaps is in need of further explanation. How-
ever we may now turn to a possible objection to our basic con-
tention.

2.I. We have contended in the above paragraph that names


ON REFERENCE

possess as such a referring function. Against this it may be ob-


jected that reference is not a function of expressions qua ex-
pressions, for expressions as such have only meaning while it is
only a genuine use of them that refers. The purpose of this ob-
jection, as is clear from Strawson's celebrated essay from which
it is derived,5 is to steer clear of both the postulation of Meinong-
ian entities and the Russellian analysis of referring expressions
with a view to avoiding the Meinongian ghosts. As is well known,
on the Meinongian thesis, such a sentence as "The present King
of France is wise" could be meaningful only if there is a present
King of France, and if there exists no such person there should
subsist such a one; whereas on the Russellian analysis the sentence
should be further analysed so that the apparent subject is seen
not to be the logical subject at all. Both the theories assume,
according to Strawson, that a referring expression to be meaning-
ful must stand for or refer to something, against which Straw-
son suggests that not reference but only meaning is a function
of expressions. An expression however may be used to refer.
Such uses of it as do not refer are cases of spurious use. Genuine
uses alone refer.
There is no doubt that both the solutions that Strawson criti-
cises wrongly assume that meaning is the same as reference. They
could therefore be avoided by distinguishing between the two func-
tions of expressions. It is not however necessary to go all the way
in Strawson's company. For Strawson's complete separation of
the two functions by assigning the one to expressions as such,
and the other to their uses is possibly due to his assumption
that an expression can only then be said to refer when there
actually exists something which is being referred to. This as-
sumption however may be questioned. Before coming to it,
however, let us in brief point out what seems to us to be a vi-
cious circularity involved in the solution offered.
The circularity may be brought out in the following manner:
How to know that a certain use of an expression does not refer?
The reply would be, when the use is not genuine. But how after
all can one be sure that the use is a spurious one? Because it
does not refer at all, would be the reply.
Now, both Russell and Strawson may be said to assume that
5 P. F. Strawson, "On Referring", Mind, July 1950, pp. 320-344.
ON REFERENCE 77
an expression (or its use) refers only when there is something to
which it refers! In Russell's theory an expression 'A', and in
Strawson's theory its use, refers only then when there actually is
the A. When there is no such A, Russell would analyse 'A' in the
manner outlined in his Theory of Descriptions, while Strawson
would say it is not a genuine use of 'A'. I wish to suggest that
this need not be so. Russell explicitly, and Strawson implicitly,
identify genuinely referring expressions with 'logically' proper
names, for is it not so that if I have or use a logically proper name,
the something which I so name must exist? To refer is certainly
not to designate.

2.3. As against this prevalent attitude, I wish to suggest that


reference in the sense in which it is a constitutive function of
expressions is nothing but intended reference. Saying that ex-
pressions qua expressions refer means only that they contain an
intention which as it were is directed towards the extra-linguis-
tic world. But that there is such a directedness or an intentio-
nality does not imply that there is a point in the extralinguistic
world where the intention terminates. Whether there is such a
point or not, whether in any particular case the intention is fulfilled
or not or can possibly be fulfilled, is an empirical question,
and therefore falls outside the scope of philosophy. It is also an
empirical question to ask what are the causes of the fact that
there are such and such intentions and no others in a given lan-
guage: part of the reason for it may be found in the structure of
the language and part in other contingent features. But what
belongs a priori to the essence of a linguistic expression as such is
the fact that there is such an intentionality.
If the above contention is true, then it follows that the pos-
sibility of a name or of using it in a language does not imply that
there exists in the actual world something which is so named. On
the other hand, the expression is made or constituted by an in-
tention just as its use makes a conscious claim upon the world
that the intention is fulfilled in a definite spatio-temporal situ-
ation. Language or the use of it does not contain within itself the
guarantee about what is or is not.

3. The reference which characterises linguistic expressions as


ON REFERENCE

such enters into, and is further reinforced in the different con-


scious states that make use of linguistic expressions. It shall be
our purpose in this section to attempt a brief sketch of the sub-
jective phenomenology of reference in so far as it characterises
the various modes of consciousness.

3.I. Reference characterises consciousness as such in the sense


that all consciousness is of something, or intends something.
Definite modes of consciousness have definite modes of intending
their objects. If we disregard the various modes and even ab-
stract from any determinate form of consciousness we reach the
idea of pure consciousness, of consciousness as such. To such pure
consciousness we can assign only an indeterminate reference to
something, to what, using Kant' language, could be called ob-
ject-in-general. Compared to this the reference of an expression
is more determinate, for each expression intends a determinate
possibility. However, subjectively speaking, the intentionality
of pure consciousness constitutes the most primary level which
makes possible all higher forms of reference including that of an
expression. This only shows that an expression considered as such
is not a merely physical event, but is the product of - in a rather
peculiar sense of 'production' - conscious intentionality: the
sense of this could be partly explicated by pointing out that un-
less there were the intentionality of consciousness expressions
qua expressions would not be possible.

3.2. Presupposing the referentiality of expressions qua ex-


pressions, there are three distinct modes of consciousness with
their distinct modes of referential character. We distinguish here
between (i) understanding the sentence'S is p', (ii) thinking of
the fact that S is p, and (iii) knowing that S is p. To be noted in
this connection are the following:
(a) Knowing that S is p presupposes an understanding of the
sentence'S is p', but not vice versa. One may quite well under-
stand the sentence'S is p' without knowing that S is p.
(b) Similarly, knowing that S is p involves a thought of S
being p, but not vice versa. The latter thought does not as such
amount to the former knowledge.
(c) The distinction and the relation between (i) and (ii) is
ON REFERENCE 79
more subtle, and is likely to be overlooked. There is however
sufficient linguistic and introspective evidence for drawing the
distinction between (i) and (ii). I may be interested only in get-
ting the meaning of a sentence (as for example when I am interes-
ted in testing my knowledge of a language I have newly learnt)
without getting involved in a thought about the fact being re-
ferred to. We may not therefore be unjustified in suggesting
that there can be bare understanding of'S is p' not amounting
to a thought of the fact that S is p, though any such thought
presupposes the former understanding. We have thus a series of
distincts (in Croce's sense of the term) such that (i) does not pre-
suppose (ii) and (iii), and (ii) does not presuppose (iii), whereas
(iii) presupposes both (i) and (ii) , and (ii) presupposes (i).

3.3. These distinctions may be sought to be denied on the


grounds, (a) that an understanding of a sentence presupposes -
contrary to what has been said in the above paragraph - think-
ing of its meaning as also a knowledge of it; and (b) that think-
ing of anything presupposes a knowledge of it for the obvious
reason that one cannot think of something one does not know.
As against (a) we may point out that what we have contended
is only that an understanding of'S is p' does not presuppose a
thought of the fact that S is p. Thinking of its meaning may be
involved in so far as I do not rest merely in understanding but
proceed further to analyse and correlate it to other sentences or
meanings. Nevertheless, the point made out in (c) in para 3.2
stands. It may be further pointed out that the objection that
understanding involves knowing the meaning rests upon a cer-
tain ambiguity of the word 'know'. According to the usage here
being followed, 'knowing' the meaning of an expression is the
same as understanding the expression, which renders that part of
the objection (a) trivial. What is more is that the usage which
permits us to speak of knowing the meaning of an expression
distorts, in a very important manner, the true nature of the
phenomenon. Knowing is either knowing an X, where X is not
a that-clause, or knowing that something is the case. In both
cases, knowledge has an object, i.e. something (a thing, a person
or a fact) that is being known. Metaphorically speaking, it must
be something in whom the intentionality of knowledge terminates.
80 ON REFERENCE

The same however cannot be said of meanings. For when we


understand an expression, its meaning is not an object in the
above sense. The function of the meaning is to make reference
possible, and to give the latter its determinate character. But
by itself it is not the sort of thing that could be inspected.
Again speaking metaphorically, the meaning of an expression
provides a sort of transparent medium for reference. The in-
tention does not terminate in it, but rather goes through it
towards the referred entity. Saying that meanings are known
therefore is using the word 'know' in a rather improper sense.
A similar equivocation lingers also in the objection (b). De-
fining 'knowing' in the strict sense such that knowing that S
is p entails that'S is p' is true, we could easily avoid this point
as well.

3-4- The distinctions introduced in 3.2 and defended against


some possible objections in 3.3 may now be further explicated in
the following manner: -
(a) Understanding the sentence'S is p' involves:
(i) knowing the meaning of'S is p' (in the improper sense of
'knowing' explained in 3.3) ;
(ii) knowing the type of entity being referred to (in a sense of
'knowing' that is less improper than that in (i) above);
(iii) a predominance of the linguistic meaning (as opposed to the
material meaning).
(iv) Added to all these, it must be mentioned, the whole act of
pure understanding (which falls short of thinking and there-
fore of knowledge) terminates in awakening the simple
meaning-intention and does not lead to a further attempt
at its fulfilment. The referential aspect is thereby thwarted
and remains restricted to being the minimum accompani-
ment of the meaning-intention, i.e. to the broad type.
(b) While understanding terminates in the meaning-intention,
the reference being an unwanted but unavoidable accompani-
ment, there is in the act of thinking ('I am thinking of .... ')
a conscious attempt to go beyond the intention, to sharpen the
reference, to reach the material meaning. This partly accounts
for the introspective datum that thinking is a more active, self-
conscious process than is the mere understanding of another's
ON REFERENCE

utterance or writing. Reference comes to the forefront, becomes


self-conscious, though still not completely realised or fulfilled.
The object of thought not amounting to knowledge is the pos-
sibility of S being p, or the ideal construction S-being-p.
(c) In knowledge, taken in the strict sense, the meaning-
intention is fulfilled, the material meaning finally swallows up
the linguistic, the reference becomes determinate and actual
and no more remains at the level of the type and the possible.
The meaning aspect is taken up and subordinated to the re-
ferential. I do not any more stop with the medium, but succeed
in seeing the object through it such that the medium is not itself
seen at all, thus completely excluding even the possibility of
that improper sense of knowing in which we speak of knowing
the meaning.
Thus the two aspects of meaning and reference are present in
varying degrees of determinateness exhibited in a rather graded
form. In the act of pure understanding meaning dominates and
reference is suppressed to the minimum; in thinking there is a
more or less even balance: in knowing meaning is taken up into
the reference.

4. Knowledge, in accordance with the above analysis, may be


regarded as the one end of a scale whose other end is represented
by pure symbolic understanding and at whose middle stands
thinking of ... The tendency which finds itself realized in know-
ledge is the demand of reference to be actual and more and more
determinate and the necessity of meaning to shade off its opa-
city and transform itself into what it should be, i.e. a transparent
medium for reference.
It is noteworthy however that of all those cases where we do
speak of knowing two stand out preeminently. These are our
perceptual knowledge of physical objects and of other persons.
The reason why these are regarded almost as paradigm cases of
knowledge seems to be that in these cases reference is not only
actual and determinate but also unique. Here knowledge is of
a unique individual, i.e. of this object or of this person here be-
fore me. One of the problems we had proposed for ourselves in
para I.4 was: 'How does it come about that we also often suc-
ceed in making unique extralinguistic reference, although the
82 ON REFERENCE

medium of such reference is provided by the general ideal-


objective meanings?" Now we are in a position to say a few words
about this extremely intriguing problem.

4.1. It has been pointed out in para. 2 that according to one of


the basic contentions of this paper expressions qua expressions
refer, so that the simple question of extralinguistic reference is
in one sense no problem at all. This does not however amount to
saying that the question of unique extralinguistic reference is
likewise no problem at all. On the contrary, it is a serious pro-
blem, the more so because of the fact that the medium of reference
remains the realm of general ideal-objective meanings. But
again to solve this problem is really in effect to dissolve it, for
the solution could only consist in showing that within a purely
linguistic discourse unique reference is not possible. The fact that
we nevertheless do refer to unique individuals is to be accounted
for by asking, what else supervenes here and gives language the
power (or a mere semblance of the power) which it does not really
possess?

4.2. The problem may be further precipitated by formulating


it in the form of a paradox: Reference in order to be unique
requires a total elimination of the medium of meaning. And yet
on the other hand without such a medium no reference - and
therefore also no unique reference - is possible. The paradox may
be sought to be resolved by pointing out that mere ostension
does succeed in making unique reference. This however, as
Wittgenstein has rightly shown6 , is not true. Ostension to do
its job presupposes a context of linguistic discourse.

4. 3· For the same reason as is contained in the first proposi-


tion of the paradox, the so-called 'ego-centric particulars'
(Russell) or the 'empirical ties' (Hall) fail to rise up to the oc-
casion; and yet it is these words - 'this', 'here', 'now', 'I' - if
any, that are the likely candidates within linguistic discourse
most suited for the task. They no doubt specify and determine,
but complete uniqueness is not achieved by them. The reason

6 L. Wittgenstein, loco cit., pp. I3e-I6e.


ON REFERENCE

for this is that they also have an aspect of meaning whose gene-
rality transcends the unique referent.

4+ The problem which almost seems hopeless may be


solved on the following lines: -
The total situation in which unique reference takes place may
be exhibited as having a stratified structure with these layers:
(i) Linguistic discourse, through the medium of meaning,
makes a general reference which delimits the field within which
the object of reference is to be located.
(ii) The ego-centric particulars like 'this' effect a further
narrowing down of the field, though as pointed out in 4.3, they
also fail to achieve unique reference all by themselves.
(iii) What is left over, the unfulfilled part of the job is com-
pleted not by anything else within discourse - not even by mere
ostension, for that also leaves room for ambiguity as shown
above - but by an active participation and communion with the
object. This last step, and indeed the most important step from
the present point of view, involves a transcendence of the purely
theoretical-cognitive attitude.

4-5. The key points in the above analysis may be stated thus:
in the first place, linguistic discourse appears to succeed in
making unique reference only because it is not self-sufficient,
but is tagged on to a non-linguistic, non-theoretical participation
in, or communion with the world of objects and persons. Second-
ly, this latter participation or communion would have been
equally ineffective, though in an opposite sense, were it not
for the fact that it finds itself within a linguistic discourse.
Finally every individual object or person presents itself from with-
in a field which again is possible only within the world. Sub-
jectively speaking, a non-thetic, non-linguistic consciousness of
the world is the background from which language serves to cut
off a field which is further narrowed down by the linguistic
mechanism till the last gap is bridged by a thetic but practical
relationship. Uniqueness is the objective counterpart of this
practical relationship, and the unique reference of language,
even of the' this', is derived from it.
CHAPTER VIII

REMARKS ON THE CONTENT THEORY*

In this paper I wish to clarify certain points which I consider


to be muddles in connection with the content theory. By 'con-
tent theory' I mean the philosophical contention that knowledge
involves three factors: a subject, an object and a content.
Amongst those who subscribe to such an analysis of knowledge there
is a great deal of divergence of opinion regarding the more
specific problems about the nature of each of the three factors
and the exact relation between any two of them. I do not wish to
take up these issues in this paper. I only wish to draw attention
to certain mistakes which, in my opinion, have characterised the
entire discussion for and against the content theory. Once these
mistakes are set aside, it would be possible - so it seems to me -
to rehabilitate the content theory on a sounder basis than before.

In the first place, it would be profitable to state an argument


which is most commonly advanced in favour of the content theory,
viz. the argument from error and other allied phenomena. It is ar-
gued that a two-factor analysis which suits cases of right know-
ledge fairly well does not, however, apply to cases of error and
illusion where there are obviously three factors involved. If,
therefore, we start from cases of right knowledge and, taking up a
two-factor analysis which they suggest, seek to extend it to
cases of error, we are sure to be disappointed. What we should do,
then, is to start with cases of error, take up a three-factor ana-
* Published in the Visva Bharati Journal of Philosophy, Vol. I, NO.2, :£965, pp.
38-42.
REMARKS ON THE CONTENT THEORY 85
lysis suggested by them and seek to extend it to cases of right
knowledge. Success awaits us here, for it is possible to contend
that in cases of right knowledge as well the content and the
object are distinct but not distinguished as they agree with one
another. Hence if we are to have a theory of knowledge which is
to apply to all cases of knowledge, true or erroneous, this can
only be some form of the content theory.
What I wish to dispute now is the importance accorded to this
argument. It seems to me that it is not only not necessary in a
theory of knowledge to start from the phenomena of error and
illusion, it is positively misleading to do so. It is misleading in
so far as in error and illusion the object is not known at all, and
this might quite understandably raise the suspicion that the
same holds good even in cases of the so-called right knowledge.
The point of distinction between right knowledge and error,
and the further consequence that the former alone deserves to be
called knowledge, are likely to be lost sight of if a theory of
knowledge starts from the latter and by analogy with it seeks to
explain the former. For, admitting that there is a content in
cases of error and illusion, it may still be argued that the same
may not be true of right knowledge whose precise point of dis-
tinction may be said to lie in this that in it the object is known
as it is. If, then, the content theory needs a supporting argument,
such argument should take into consideration the very pheno-
menon of right knowledge.
No knowledge is merely the intentional reference of conscious-
ness to any object, or to an indefinite object-in-general. Con-
sciousness, of course, must be of some object or other, but this
indefinite reference is not adequate to constitute a knowledge.
Knowledge is of a definite object, and this definiteness is a
constitutive character of knowledge. It is, of course, true
that knowledge does not confer definiteness on its object.
It rather derives its definiteness from its object, it is
vi$ayanirupya. Not only is a pot different from a cloth, but the
knowledge of a pot is different from that of a cloth. This latter
difference is derived from but not wholly reducible to the former
difference. This definiteness of a knowledge, as the more in it
than bare consciousness, is the epistemic content. And its pre-
sence is revealed only to reflection on the knowledge whose
86 REMARKS ON THE CONTENT THEORY

content it is. For, in reflection a knowledge is recognised not


only as a knowledge but also as this knowledge, as having such
and such structure. The very fact that a knowledge is capable of
logical analysis shows that what is analysed is the content, not
pure consciousness nor the object known.
Once the presence of the content is thus discovered, the fur-
ther question as to how this is compatible with knowledge of the
object may be taken up.

II

In most discussions of the content theory it is assumed that


such a theory entails some sort of representationism. The ad-
vocates of the theory try to show that direct knowledge is not
possible. They look upon the naive or direct realist as their chief
adversary. Even the critics of the theory have drawn attention
to the ruinous consequences of its representationist character
and argued that if the content always stands between the sub-
ject and the object then the object is perhaps never known at
all, so that the knowledge intention should be deemed
as terminating at the content and never going beyond it. As a
result one may be led even to doubt or deny the very existence
of an object other than the content. But then, if there is nothing
other, the content is but the object, and there is no point any
more in the content theory.
Now it seems to me that the issue between a two-factor and a
three-factor analysis of knowledge should not be muddled up with
the issue whether knowledge is direct or indirect. The former
issue is relatively clear and well-defined; the latter is not so,
owing to the fact that there is hardly any clear idea as to what
the words 'direct' and 'indirect' are to mean in the present con-
text. It is not true - as I shall try to argue presently - that ad-
mission of a content of knowledge entails that knowledge of the
object is indirect. The belief that it does so entail has been one of
the great stumbling blocks on the way to a clarification of the
knowledge situation. This erroneous belief, viz. that admission of
a content of knowledge entails that knowledge is indirect, rests,
on its part, upon (a) a mistaken notion of the nature of content
REMARKS ON THE CONTENT THEORY 87
and its role in a knowledge situation and (b) a mistaken notion
about what is meant by 'direct apprehension'.
(a) It is supposed that if there is at all to be content of know-
ledge then this content, by its very nature, would stand between the
subject and the object, and owing to this intermediary position
that it would occupy it would not permit the object to be known.
It would virtually arrogate to itself the status of the object. At
best it would be but a poor representative, perpetually under
the shadow of suspicion. Such a theory would no doubt give a
plausible account of error, but at what great cost! It would lay
every knowledge under the suspicion of being erroneous; and
in so far as the object would never be known at all, there would
be no true knowledge anywhere.
This apprehension is no doubt genuine if it is directed against
the usual conceptions of the nature of the content. The notion
of content, however, need not be so formulated. It is necessary at
this stage to bear in mind what should be the proper task of a
theory of knowledge. A theory has to be true to phenomena;
and where there is conflict between a theory and the relevant
phenomena, it is the former, and not the latter, that should be
discarded. It is not, therefore, within the rights of a theory of
knowledge to tell us (far less, to prescribe) what we know and
what we do not know. Thus, when some philosophers tell us
that we do not perceive physical objects, like houses, trees and
tables, but only infer or construct them, they are going beyond
the limits of what they are legitimately entitled to say. The
philosopher's contention should not be incompatible with our
ordinary belief that we perceive physical objects. That we know
objects (if we know anything at all) is the fact from which alone
we could take our start, which we cannot dispute, and with which,
therefore, our subsequent philosophical contentions should not
come into conflict. All that we have to do is to analyse such know-
ledge (but not to dispute it), to discover any new hitherto un-
noticed phenomenon implicit in them, to lay bare the presup-
positions and implications, if any, of such knowledge and fi-
nally, by viewing it in the light of other types of knowledge, to
interpret and evaluate it. This being the case, our conception of
content should be such that it does not conflict with the phe-
nomenon of our knowledge of objects. There are in fact two phe-
88 REMARKS ON THE CONTENT THEORY

nomena to be harmonised. There is, on the one hand, the un-


deniable phenomenon (testified by unreflective consciousness)
that we know objects, and there is, on the other hand, the fact,
discovered by reflective consciousness, that all knowledge has a
determinate content. To deny the content in the way the naive
realist does it would conflict with the latter, to deny the object
by reducing it to a system of contents - in whatever manner
'system' may be understood - would conflict with the former.
For, to harmonise the two phenomena is also to maintain the
distinction there is between the two. Where the distinction is
obliterated, the gap bridged, there one of them is just reduced to
the other. Moreover, if the object were a system of contents,
actual or possible - for it could not be a system of actual contents
alone - such a system could not be given, it could only be con-
ceived or anticipated: the object would, in that case, be a Kant-
ian Idea, not a knowable, and certainly not what is known. To
accept the distinction between object and content and yet to set
up the latter as what alone is given - as what intervenes, stands
in between, or even as what represents - is to cast doubt on the
veracity of the first phenomenon which should rather be our
undisputed starting point.
The only way such harmonisation could be effected is to ad-
mit that the content serves as a transparent medium for reference.
I would like to contrast this notion of content with the one that
leads to scepticism, by calling the latter a conception of the
content as opaque, and the former the conception of it as trans-
parent. What we need in a theory of knowledge is the con-
ception of a transparent content. It is to the task of elucidating
this conception that we may now turn.
Consider the case of making a meaningful statement, or hear-
ing one being made. The sentence used to make the statement has
a meaning which is the proposition expressed by it. However,
the speaker's primary intention is to refer to the fact it is being
used to state, and not to convey the meaning, though at the same
time the meaning, i.e., the proposition, functions as the medium
which makes the reference possible. The hearer, if his attitude
is a reflective one of understanding the meaning, may simply
entertain the proposition; but in the primary cognitive attitude
the hearer knows or attends to the fact that is being stated,
REMARKS ON THE CONTENT THEORY 89
though understanding the proposition makes this possible.
Here the meaning-content, i.e. the proposition, does not thwart
objective reference, but renders the latter possible as its in-
dispensable condition.
The same, I would suggest, holds good of all knowledge-
situation. The contents are discovered through reflection; but in
the unreflective cognitive attitude they are nevertheless there,
not as superfluous appendages, but as the necessary media of
objective reference. They do not hinder the object from being
known by arrogating to themselves its status, for their very
function is to make knowledge of object possible. It is to be
noted that the conception of content as transparent leaves room
for further determination of its status. There remain open such
questions as - Is the content an additional entity, or is it the
object as known (in the Nyaya terminology, is the vi:jayatii an
atirikta-padiirtha), etc.?
(b) It may nevertheless be felt that though the concept of
transparent content makes room for knowledge of the object it
does not leave room for direct knowledge and so cannot avoid
some form of representationism. This charge can be avoided only
by way of clarifying some misconceptions about what is meant
by 'direct knowledge'. It seems to me that the contrast between
direct knowledge and indirect knowledge should better be re-
placed by that between perceptual knowledge (pratyak:ja) and
non-perceptual knowledge (parok:ja). In that case, the issue would
be - can objects be perceptually known, or are they always
to be inferred? This issue may be settled relatively easily, e.g.
by appealing to the principle, as Vatsyayana does in his defence
of avayavipratyak:ja, that inference presupposes perception, so
that if there is no perception of the mark there would after all
be no inference either. 1
If, on the other hand, the issue is formulated in terms of direct
knowledge vis-a-vis indirect knowledge, a settlement of the issue
is hindered by the difficulty of clarifying the meaning of the key
words involved. If 'direct knowledge' means knowledge of an
object without mediation by a content or idea, then the conten-
tion that knowledge according to the content theory has to be
1 Cpo J. N. Mohanty, "The Nyaya Theory of Avayavipratyak$a" in The Journal of
the Indian Academy of Philosophy, Vol. II. [Reprinted in this volume}.
90 REMARKS ON THE CONTENT THEORY

indirect would be but a naked tautology. The issue, then, is -


is there a content or not; and it is advisable not to state it in the
misleading language of direct or indirect knowledge. If by 'di-
rect knowledge' be meant knowledge which is not mediated by
anything whatsoever, there is no such knowledge, for the medi-
ation or the instrumentality of sense-organs at least is un-
avoidable. It may be conceded that perception contains a core of
direct knowledge, viz. the reception of sense-data, while around
this core there cluster inferred or associational elements. But
against this contention we may say two things. The less im-
portant of the two is that such a position entails - as is explicitly
held by most of the Western philosophers - that perception
contains an element of inference or construction, which, to my
mind, amounts to denying the very perceptuality (pratyak$atva)
of perception. The other, and the more weighty, point is that this
pin-pointing of the directly given to sense-data is based on a very
dubious conception of the given. 2
To be directly known is to be perceived. What is directly
known is not inferred. In this sense of "direct", admission of a
content of knowledge is not incompatible with direct knowledge
of the object, provided we correctly formulate the nature of the
content. However, it is not necessary that the object should be
always directly known. It may be inferred, and in so far as it is
inferred it is not directly known, though known all the same. The
central point is that the content does not hinder the object from
being known. It is also necessary to bear in mind that where an
object is indirectly known, say inferred, it is nonetheless known,
and it is not the case what is known is the content. However,
there is no denying the primacy of perceptual knowledge. The
object that is inferred could be perceived. One could say, it
would have been best if it were actually perceived. Perception,
therefore, is the paradigm case of knowledge.

III
We have shown earlier that that in knowledge which makes it
amenable to logical analysis is the content. Now this is likely to
2 Cpo J. N. Mohanty, "The Given" in the Proceedings 0/ tke Delhi Philosophical
Colloquium, Indian International Centre, Delhi, 1964. [Reprinted in this volume].
REMARKS ON THE CONTENT THEORY

throw light on the nature of the content. It at least tells us what


the content is not. The contents in the case of perception are not
sense-data, for sense-data are not the sort of things that are
capable of logical analysis. Likewise the content cannot be of the
nature of sensation or image. For, sensation, sense-data and image
are such that if they were to stand between the subject and the
object they would themselves be the objects apprehended and
would not be the sort of transparent medium for reference which
any content qua content should be if it is to fulfil its true epistemic
function. The content, then, would be of the nature of meaning. It
would rather be that from which the proposition regarded as the
meaning of an indicative sentence is an abstraction. A logic of such
cognitive contents would, then, be in a certain sense prior to a logic
of propositions.
CHAPTER IX

PHENOMENOLOG Y AND ONTOLOGY

I. PHILOSOPHICAL METHOD

(a) Puzzles: "Zuriick zu den Sachen", "Back to the facts" was


Husserl's epoch-making slogan. Though I have long since been
captivated by this slogan, I have never ceased to be puzzled by
it. Husserl's programme attracted me because of his rejection of
speculative metaphysics. Philosophy, it seems to me, must in
some sense be descriptive. But what should it describe? What
sorts of facts are those that philosophy should describe? How are
these facts different from the facts described by ordinary per-
ceptual statements as also by the descriptive statements of
the sciences? Further, is pure description at all possible? Pure
description, it would seem, would be possible only if it were also
possible for something to be given without the least admixture
of interpretation. But is not all that seems to be given shot
through with interpretations? More particularly, is not all given-
ness subject to the mould imposed by language? Is there, po-
sitively asking, any mode of direct i.e. non-linguistic access to
the facts concerned? Finally, how is the ancient and age-old
conception of philosophy as explanatory and interpretative in
function to be accommodated? In other words, should philo-
sophy abandon its age-old function which, as it were, has come
to constitute its defining character and assume a totally new role,
and if so would not the new science forfeit its claim to be called
'philosophy' ?
(b) Hegelians like Bradley and Blanshard, speculative meta-
physicians like Whitehead, existentialists like Kierkegaard and
Heidegger, and linguistic philosophers like Ryle, Austin and
Goodman - all agree in condemning the idea of the unadul-
PHENOMENOLOGY AND ONTOLOGY 93
terated given, though they might differ amongst themselves
as to the nature and the source of the interpretations involved:
the interpretations are tagged on either to an ontological scheme,
or to a conceptual framework, or to a language system. White-
head expresses the point most forcefully thus: "If we desire a
record of uninterpreted experience, we must ask a stone to re-
cord its autobiography".l Heidegger writes in the same vein:
Alles vor-priidikative schlichte Sehen des Zuhandenen ist an ihm
selbst schon verstehend-auslegend".2 The most primitive 'seeing'
is 'interpretative'. It is not therefore surprising that Heidegger
should have subscribed to the apparently paradoxical view that
the phenomenological method of description is really inter-
pretative. 3 Similarly, for the Hegelian there is no absolute im-
mediacy. Immediacy is either a merely logical presupposition
or the consummating goal towards which experience and know-
ledge are moving and in which they would but commit suicide:
in neither case is it capable of presenting data for phenomenolo-
gical description. For the later Wittgenstein, even os tension
to be meaningful should presuppose a language system. And
curiously enough, all of them - excluding the extreme Hegelians
alone - recognise the value of description as a philosophical method.
Whitehead speaks of his metaphysics as being descriptive in
character. 4 So does Heidegger also. Wittgenstein writes: "Phi-
losophy may in no way interfere with the actual use of language;
it can in the end only describe it". 5
The contradiction can be resolved, so it seems to me, by
recognising that what the phenomenological method of des-
cription stands for is not the primacy of intuition over under-
standing. It does not aim at getting hold of the primitive im-
mediacy alone, although it seeks to do that also. The opposition
between intuition and understanding, between the given and the
interpretations is a false one. Reflection and understanding
lead to fresh intuitions. And intuitions may serve to lay bare in-
terpretations that have come to be sedimented in the structure
of phenomena.
1 A. N. Whitehead, Process and Reality, New York edition, p. 22.
2 M. Heidegger, Sein und Zeit, Tiibingen, 7th edition 1953, p. 149.
3 Ibid., p. 150.
4 A. N. Whitehead, loco cit., p. 19.
5 L. Wittgenstein, Philosophical Investigations, Oxford, 1953, p. 49.
94 PHENOMENOLOGY AND ONTOLOGY

(c) The idea of a philosophical system, understood in the sense


of traditional speculative metaphysics, is obviously incompa-
tible with such a conception of philosophical method. For a
phenomenological philosophy should neither deduce nor reduce.
It should avoid, as Nicolai Hartmann constantly reminded us,
both kinds of reduction, "nach Unten" nach "nach Oben". There
is however another sense in which there is room for system in
such a philosophy. Two principles seem to me to be illuminating,
and both I owe again to Nicolai Hartmann (though not in the
form in which I here make use of them). There should be, in the
first place, a minimum of speculative metaphysics. Pure des-
cription is an ideal after which we could only strive, and in this
striving our heuristic principle should be to make as few spe-
culative-deductive commitments as possible. In the second place,
the system should be an 'open' rather than a 'closed' one. A typi-
cal example of what is meant by a closed system is the Hegelian
philosophy in which there is an explanation for every phenome-
non in the sense that phenomena are deductively linked up
with each other so as to lead up to the highest Idea. Now by an
'open' system I mean one in which phenomena 'announce' their
own status or 'top os' and, when they are conceptually placed
side by side, there are found to be missing links or gaps that can
only be conceptually filled in, but need not be. The Advaita
Vedanta is a good example of such an open system, though the
later exponents of the system no doubt sought to close the gaps
arbitrarily. Gaps are to be filled in, as far as possible, by fresh
phenomenological data (this is what Sri Aurobindo seems to be
doing in his peculiar doctrine of grades of consciousness), failing
which they are themselves to be recognised as irreducible.
(d) What are those phenomena which philosophy should des-
cribe? It was a common doctrine of the phenomenologists that
these are nothing other than essences and essential structures.
The full implication however was not as clearly brought out as it
should have been. There is no doubt that the different aspects of
the situation were seen in different contexts by different thin-
kers; what is still lacking is a perception of the essential unity
and interrelation of these aspects. There were inevitably three
different groups of thinkers: the ontologists (Scheler and Hart-
mann), the subjectivists (the later Husserl, Natorp) and the
PHENOMENOLOGY AND ONTOLOGY 95
linguistic analysts (Hans Lipps, Gilbert Ryle). It has not yet
however been sufficiently realised that these three groups were
concerned with three different aspects of the proper object of phi-
losophical thought. I would now proceed to explicate the above
statement.
Critical philosophers since Kant have from time to time sus-
peted naive ontology, and have sought to reduce ontology
either to transcendental logic (Kant) or to some kind of meta-
linguistic commitment. Quine, for example, following Russell,
has reduced ontological commitments to the range of values that
are permissible for the variables in the existential quantifier.
But more interesting is Carnap's attempt to translate statements
in the material mode to those in the formal mode. Now we may
broadly distinguish between two different ways of replacing the
material mode of speech: the formal mode of Carnap and the
phenomenological mode. Whereas the former considers the way
the corresponding words are used, the latter considers the mode
in which the entity under consideration is given. Thus while the
ontological or the material mode speaks of entities and the formal
mode of the logical behaviour of the corresponding words, the
phenomenological mode considers the mode of givenness of that
entity. Now the phenomena that a phenomenological philosophy
has to describe are such that with regard to them the ontolo-
gical mode of speech is inseparable from the formal and the
phenomenological modes. This may be clarified with regard to
two statements which I quote below: one from the K. C. Bhat-
tacharyya and the other from Edmund HusserI.
K. C. Bhattacharyya writes:
"Speakability is a contingent character of the content of
empirical thought, but it is a necessary character of the
content of pure philosophic thought."6
Referring to the points of distinction between the phenomenolo-
gical and the ontological attitudes, HusserI writes:
"Das sind kardinale Unterschiede, dienur Verallgemeinerungen
des einfachen Unterschiedes sind, dass Bedeutungen setzen und
Gegenstiinde setzen zweierlei ist .. "7
Both Bhattacharyya and HusserI are thus drawing attention
6 K. c. Bhattacharyya, Studies in Philosophy, Vol. II, Calcutta, 1958, pp. 102-3.
7 Ideen, Vol. III, Martinus Nijhoff, (Husserliana), The Hague, 1950, pp. 88-89.
96 PHENOMENOLOGY AND ONTOLOGY

to the fact that the essences which phenomenology has to des-


cribe are but meanings, and thus any ontological assertion about
them cannot but have an equivalent in the formal mode. Like-
wise, such an essence has its own mode of givenness which is not
just accidental to its own mode of being. Far from it, the mode
of givenness and the mode of being are but inseparable poles of
the same phenomenon. This equivalence however does not hold
good of empirical facts of whom, as Bhattacharyya rightly sees,
speakability is - and I should add, givability - is only a con-
tingent character.
Thus those philosophers who deny to philosophy any factual
content and assign to it the task of exploring the logical behav-
iour of words are partly right, as also are those who take up the
programme of bringing to light the subjective constitution of the
objective essences. The former are right in so far as the objects of
philosophical description are, unlike empirical facts, revealed
through meaning analysis. They err however in thinking that
philosophical statements are for that reason of no ontological
significance, and also in concentrating on a therapeutic use of
their linguistic wisdom. We contend on the other hand that
ontology is possible through linguistic analysis, though such
ontology would be very different from naive dogmatic ontology.
The same is also true of the programme of constitution-analysis.
Both meaning-analysis and constitution-analysis are the gate-
ways to a critical ontology, and the three in their unity con-
stitute the integral phenomenological method.
The reason why linguistic analysis has been regarded as on-
tologically unfruitful and why questions of language and ques-
tions of fact have been sharply sundered seems to be a mistaken
notion about the nature of language which on its part derives
from a mistaken theory of meaning.

II. MEANING AND REFERENCE

Meaning has wrongly been taken to consist in use. After the


early empiricists' image theory and the latter-day positivists'
verifiability theory were rightly abandoned, the modern analysts
have taken to the view that knowing the meaning of an expres-
PHENOMENOLOGY AND ONTOLOGY 97
sion is the same as knowing how to use it in accordance with the
rules and conventions of the language under consideration. Such a
theory of meaning has, in the eyes of its propounders, two dis-
tinct advantages. On the one hand, it renders it impossible to
speak any longer of meanings as entities. On the other hand,
it avoids the subjectivism of the image theory and the veri-
fiability theory, for the rules and conventions that confer
meaning are far from being sUbjective. Now if this is what con-
stitutes meaning, and if all extra-linguistic facts, objects, events
(whichever language one may prefer) are objects of reference, then
analysis of the meaning of an expression should bring to light
its behaviour within the language concerned and cannot by
any means provide us with a basis on which to build an ontology.
Now, such a conventionalist and operationalist theory of
meaning is inadequate for the following reasons: It cannot ac-
count for what happens when we grasp the meaning of a word
or of a sentence in a flash as it were. It would not do to say that
what we grasp is a set of rules and conventions for reasons of which
Wittgenstein himself is aware. s It is not the bare possibility of
use, for what we grasp is not this possibility itself but rather
that which makes its use possible. Wittgenstein is right9 that
certain objective circumstances, human customs and institutions
in the present case, a conventional system of signs with rules of
operation - must be given in order that I could intend a meaning.
But this does not imply that a linguistic expression is nothing
but a physical sign or that our understanding of it is nothing
other than the capacity to operate with it in accordance with
rules. We contend on the other hand that given a set of signs
with rules of operation, the set would not amount to language -
nor would the operation amount to an understanding - unless an
intellectual act or meaning intention supervened. Now the mean-
ing in the substantive, the much abused abstract entity, is a
phenomenological datum not to be liquidated by any theory,
though its platonic character would be less appalling if only we
remember that it is but the intentional correlate - as HusserI
would say - of the meaning intending act.

8L. Wittgenstein, lac. cit., p. 53.


9Cpo J. N. Mohanty, Edmund Husserl's Theory at Meaning, Martinus Nijhoff, The
Hague, 1964, p. 40.
98 PHENOMENOLOGY AND ONTOLOGY

In other words, it must be recognised that meanings


have an identity and objectivity about them, which all forms of
psychologism and conventionalism threaten to destroy. At the
same time the 'It means' has its necessary correlate in the 'I
mean', and both have a necessary reference to 'in the language
L'. Platonism in theory of meaning must therefore be so inter-
preted as to be the fulfilment of both subjectivism and formalism.
The relation between the 'I mean' and the 'It means' is especially
intimate: using an analogy which I have elsewhere employed10 ,
one could say just as in modern physics it has come to be recognised
that the wave theory and the corpuscular theory of elementary
particles are not rival hypotheses but are really complementary
descriptions, so in theory of meaning it has got to be recognised
that platonism and anti-platonism, the ontological and the sub-
jective approaches, are complementary, not rival and not even
alternate, ways of describing the same unitary phenomenon.
The ontological hypostatisation has to be supplemented by bear-
ing the subjective and the linguistic "backgrounds" in mind,
just as the subjective and the linguistic relativism has to be over-
come by making it subservient to a recognition of the ideality
and the objectivity of meanings.
To admit the element of truth that is there in a platonistic
theory of meaning is not however to betray one's failure to dis-
tinguish between meaning and reference, for it is precisely this
distinction - kept in mind by both Frege and Husserl - that
provides the basis for platonism. It cannot also be said that
platonism here implies the unpalatable thesis that in understan-
ding the meaning of an expression we inspect impalpable ab-
stractions. It is particularly against this last accusation that a
few words of clarification are needed.
Understanding an expression'S is p' mayor may not be a
constituent of the knowledge that S is p. When it is a consti-
tuent of the knowledge the intention is of course directed to-
wards the fact referred to, and not towards the meanings them-
selves. But even when the understanding is a mere understanding
not amounting to knowledge, we do not inspect meanings, for
meanings are just not the sort of entities which could be in-
spected. They may perhaps be more appropriately described as
10 Ibid., p. 75.
PHENOMENOLOGY AND ONTOLOGY 99
transparent media through which the intention is directed to-
wards the object of reference. They are not intended in the same
sense in which a fact being referred to is: the latter is the ob-
ject intended, the meaning is the mode of so intending. Our
theory that even a mere understanding of an expression not
amounting to knowledge is also to refer implies that reference,
like sense or meaning, belongs to expressions qua expressions,
and it should be obvious that with this we are rejecting the view
of those philosophers who hold that expressions have only mea-
ning while it is only a genuine use of them that refers. The Meinong-
ian ghosts could be avoided if only we bear in mind that not
all intended references are capable of being fulfilled.
A distinction between understanding and knowing may be
recommended on the basis of the above considerations. It fol-
lows from the above that to understand a sentence "s is p" is
also to understand what kind of facts it refers to, but this does
not - unless the understanding has entered into a knowledge as a
constituent thereof - amount to knowing the fact. Knowing that
S is p however presupposes an understanding of the sentence
"s is p". Knowledge in this sense implies a unique identification,
in Husserl's language a fulfilment. This is not more than sug-
gesting that the two, understanding with its generality and know-
ing with its unique identification, represent two poles between
which human cognitive endeavour oscillates. Scientific know-
ledge, in trying to get away from unique reference to generality,
aims at pure understanding. Perceptions of physical objects and
of other persons are, not without reason, regarded as the stan-
dard types of knowing, for they excel just in the unique iden-
tification achieved.

III. THE THEORETICAL AND THE PRACTICAL MODES OF


GIVENNESS

How is such unique reference possible? As I have said else-


where, the extra-linguistic reference of language presents no
problem at all, for such reference is an essential character of
language qua language. The really pertinent problem is, how is
it possible to make unique reference? Most philosophers who have
IOO PHENOMENOLOGY AND ONTOLOGY

dealt with this problem fall back upon such words as 'this',
here', 'now'. But the mere fact that even these words have an
aspect of sense distinguishable from their reference makes it
possible that one can understand them without knowing what
precisely is being uniquely referred to. We have to recognise
that a knowledge of the unique individual is far from being a
contemplative affair. It is necessary to recognise the distinction
between two basic modes of givenness: the theoretical and the
practical. The unique particular (physical object) and the unique
individual (person) are identified not through theoretic con-
templation but through practical relationships.
In theory of perception this distinction saves us from much
embarrassment. It is well known how the vexed question of the
perception of physical objects has been treated in contemporary
philosophy. Starting with the sense-data, the qualities or the
essences as the only given elements, philosophers have found
themselves compelled to regard the physical object either as
totally unknown (but believed on 'animal faith') or as
merely inferred or even as a construction out of those elements.
N one of these consequences has proved acceptable for one reason
amongst others that they all go against that primary evidence
with which things are given in unreflective experience. Things
are not given in that theoretical attitude in which one discovers
sense-data or essences. Philosophers, as Whitehead rightly saw,
have erred in regarding perception as a contemplative experience,
and the percipient person as a passive epistemological subject
which he is not. In the reflective and contemplative attitude,
the stubborn brute physical objects recede to the background
and in thejr place we are confronted with sense-data and essences.
This, and not the much discussed argument from illusion, is the
real source of the sense datum approach, which is destined to failure
as it surreptitiously seeks to replace the data of one mode by
those of another. It is not true to contend that the sense datum
theory is only an alternate linguistic recommendation, the
physical object language being another such possibility. The
truth seems to be that they are not co-ordinate possibilities
but are rooted in two successive modes of disclosure, the phy-
sical object being given in the primary, unreflective practical
mode and the sense-data in a subsequent reflective and theo-
PHENOMENOLOGY AND ONTOLOGY IOI

retical mode. Each theory, or each language if you like, has thus
its own justification. Given the physical object, it can of course
be analysed into sense-data and essences. 11 But the theoretician's
attempt to recompose that original pre-theoretical unity out
of these data is bound to fail.
As with the problem of our perception of physical objects, so
also with the problem of our knowledge of other persons. Here
again it is an erroneous procedure to begin with my awareness of
myself and then to search for some means (inference, empathy or
Einfuhlung) by which I could reach, apprehend or realise other
selves. It is again through a system of practical relationships
that I discover others as well as myself, in fact both together and
as inseparable. This is the element of truth in Heidegger's con-
tention that the person is, in his essential structure, 'with others'.
As the practical relationships reveal the person-with-other-
persons, so does theoretical reflection lead through various
stages to the transcendental subjectivity of Kant and Husserl.
Husserl was right when he insisted that even the transcen-
dental subjectivity is intentional, so that the Advaita conception
of a pure non-intentional consciousness is beyond phenomenolo-
gy. Such a purely non-intentional consciousness is not given at all
in any of the modes of givenness, theoretical or practical. The
notion of pure consciousness represents the limiting point of
our turning away from the ontological attitude, and since the
phenomenological mode has always its correlative ontology,
the supposed pure non-intentional consciousness cannot even
be a phenomenological datum. It can therefore be postulated
only by an act of faith.
The practical mode thus reveals on the one hand the world of
things and on the other a community of persons, the 'It' and the
'Thou'. Corresponding to this we may draw the distinction,
following Kant,12 between the technically practical and the
morally practical. Both present real existence and conceptually

11 Professor J. N. Chubb (Bombay) tells me that late K. C. Bhattacharyya in an


unpublished letter to him expresses the view that the sense-data are results of "aesthe-
tic abstraction". Professor Chubb also recalls a remark once made by the late Pro-
fessor R. G_ Collingwood to him that no intelligible relation could be formulated
between physical objects and sense-data since they belonged to two different levels.
12 1. Kant, Critique 01 Aesthetic Judgement, E. Tr. by J. C. Meredith, Oxford, 19II,
Introduction, p. 9.
102 PHENOMENOLOGY AND ONTOLOGY

irreducible unities which, when sought to be grasped by theo-


retical consciousness, disintegrate into endless series.

IV. A PHENOMENOLOGICAL DISCONTINUITY

Philosophy is generally taken to involve reflection. The sug-


gestion then that the unreflective practical attitude, and the
corresponding mode of disclosure, be accorded the recognition
that we have given to it may seem to involve a paradox. How can
reflection accept the unreflective orientation and its disclosure?
And should not philosophy rather challenge the naivity of that
primitive attitude, and then shatter it into bits in order to re-
construct it in accordance with reflective categories? Or at
least should not philosophy 'transcend' that primitive naivety?
Or, if one is in sympathy with that primitive naivety, should not
one, as a philosopher, undertake the task of proving or deduc-
tively demonstrating - as G. E. Moore sought to do - the beliefs
ascribed to it?
Now as is well known, the main point of distinction between a
phenomenological philosophy and a deductive metaphysics is
that the former "describes" whereas the latter "deduces",
"proves" or "explains". Accordingly, a phenomenological
philosopher has to face a paradox, a paradox that is involved in
his very method. On the one hand he has to install himself in an
attitude, be it of unreflective naivety or of reflective contem-
plation. At the same time, he has to transcend it to be its passive
witness (siik$in), to suspend belief as Husserl would say in order
to be able to describe. This paradox cannot be resolved, and has
to be accepted: this simultaneous participation and transcen-
dence - which in fact provides the key to a phenomenological
philosophy. The philosopher therefore need not accept the be-
liefs of unreflective attitude just as he need not also reject them.
Achieving the needed transcendence, his job is to tell the tale.
He is not to be a partisan but an impartial spectator. Moore,
when he undertook to defend common sense as against specu-
lative philosophers, committed a double error: he ascribed to
the pre-reflective attitude beliefs which are themselves only
other philosophical theories, and he sought to demonstrate what
PHENOMENOLOGY AND ONTOLOGY 103

has a non-demonstrative certainty. He failed thereby to ex-


hibit the nature and the source of that certainty.13
A phenomenological philosophy has to accept a radical phe-
nomenological discontinuity, and any attempt to overcome it is
to be suspected as originating from a too hasty desire for achiev-
ing metaphysical simplicity. One of the beauties of the Ad-
vaita Vedanta is that it accepts such a discontinuity between
the vyavaharika (the empirical) and the paramarthika (the trans-
cendental), and considers any relationship between them as
being logically indescribable. Our distinction between the prac-
tical and the theoretical may be regarded as a pale reflection
of that spiritual philosophy on the level of secular philosophi-
sing.

18 J. N. Mohanty, "On Moore's Defence of Common Sense", Indian Journal 01


PhilosoPhy, II, 1960, NO.4, 1-10. [Reprinted in this volume].
PART TWO
CHAPTER X

A NOTE ON MODERN NOMINALISM*

The purpose of this note is to draw attention to a most per-


plexing characteristic of some modern nominalistic theories,
especially as we find them in the writings of W. V. O. Quine and
Goodman.
Before coming to the main point of this note, it would not
perhaps be out of order if I recall that I have elsewhere! drawn
attention to some distinguishing features of modern platonism
(using that much-abused term in a very wide sense) as we find
it in German phenomenology as well as in Whitehead, Russell
(in some phases of his philosophical career) and Santayana. Two
such features may be mentioned straightaway: first, platonism
has ceased to be a speculative and metaphysical doctrine and
has been given a phenomenological, descriptive basis; secondly,
the platonic entities are no more taken to be universals in the
traditional sense. Whitehead, Santayana and the German phe-
nomenologists refuse to identify their platonic entities with
universals. On the contrary, they speak of ideal singularities
and of fully determinate essences. I cannot here substantiate
these two characterisations. But let me conclude this digression
by remarking that in these modern forms of platonism the dis-
tinction between individuals and universals has been blurred
and that between the real and the ideal has been regarded as fun-
damental.

* First published in the Proceedings of the Indian Philosophical Congress, Cut tack,
1959·
1 Cpo my Nicolai Hartmann and A. N. Whitehead: A Study in Recent Platonism,
Calcutta, 1957.
I08 A NOTE ON MODERN NOMINALISM

Turning now to the type of nominalism developed by Quine


and Goodman, we find another most interesting, and yet per-
plexing, manner in which the same distinction between indivi-
duals and universals is sought to be blurred. What we get at the
end is so very different from what has traditionally been called
'nominalism' that we are left wondering what has in the long
run been achieved. One even wonders if what has been achieved
is not, in some sense, a "meeting of extremes", of the extremes
of Universalism and nominalism.
I should now hasten to explicate and substantiate this last
remark. Universalism is not the same as platonism. For, what
has come to be called 'platonism' consists in the admission of
'abstract' entities into one's ontological scheme; but this admis-
sion does not necessarily entail an expulsion of concrete parti-
culars from that scheme. Traditional platonic realism is, in fact,
dualistic. Universalism is the ontological theory which denies
particulars and reduces them to universals; if a universalist is
a linguistic philosopher, he may content himself with a substi-
tution of names of particulars by names of universals. Strictly
speaking, nominalism is the exact antithesis, not of platonism,
but of universalism. For, just as universalism admits only uni-
versals into ontology (or 'names' of universals into the ideal
language), so does nominalism admit only concrete particulars;
both are monistic. Now, the type of nominalism developed by
Quine and Goodman is, in one of its major aspects at least, not
far removed from what has been called here universalism. Quine,
for example, considers the whole category of 'names' as super-
fluous and would permit reference to particulars only indirectly
over variables within the quantifier. So far as Goodman is con-
cerned, the basic elements of his ontology are identically repeatable
qualia that are as much individuals as universals, the concrete
particulars like tables and chairs being constructions out of
such elements. Before considering these two points in some more
detail, let me state the main motives, principles and devices of
this type of nominalism.
A NOTE ON MODERN NOMINALISM rog

II

Modern nominalism springs from two motives: there is, to


start with, a distrust of abstract entities, a distrust that, Quine
and Goodman confess, "is based on a philosophical intuition that
cannot be justified by appealing to anything more ultimate".2
One may conjecture that this distrust and this refusal are streng-
thened by the refusal of 'common sense' to admit such absurd
and fictitious entities as 'the present king of France', 'the round
square', etc. It would, however, be too hasty to invoke common
sense for supporting our refusal to admit such entities as red-
ness or circularity. In fact, one cannot agree with the tendency
to huddle together such entities as redness and circularity with
such others as the present king of France and round square, to
label them all as abstract, and then to banish them together
from ontology. This procedure is unjust; common sense rebels
against the admission of fictitious entities, whereas it is not com-
mon sense but a certain philosophical theory that rebels against
the admission of redness and circularity. In the former case, we
may call to our aid an intuition that is not further justifiable;3
in the latter case, what moves the nominalist is not an intuition
but an article of faith. 4
The other motive, clearly stated by Goodman,5 is the logical
incomprehensibility of classes (an incomprehensibility which, as
Dummet has pointed out,6 had puzzled also McTaggart). There
may be two classes, A and B, which are as classes different and
yet have no difference in their contents.
It is not the purpose of this note to answer this difficulty.
2 W. V. Quine and N. Goodman, "Steps towards a constructive Nominalism",
Journal of Symbolic Logic, 1947, NO.4, p. 105.
3 But even amongst entities banished as 'fictitious' one should be able to detect
valuable distinctions. The case with 'the present king of France' is different from that
with 'round square'. Strawson has convincingly shown that the former expression
is not meaningless, even if a certain use of it may have no reference.
4 Cpo A. Church: "The extreme demand for a simple prohibition of abstract en-
tities under all circumstances perhaps arises from a desire to maintain the connection
between theory and observation. But the preference of (say) seeing over understan-
ding as a method of observation seems to me capricious". ("The need for Abstract
Entities in Semantic Analysis "in Proceedings of the American Academy of Arts and
Sciences, Vol. 80, NO.1, p. 104).
5 N. Goodman, The Structure ot Appearance, Cambridge, Mass., 1951, p. 108.
6 Mind, r955, pp. lOI-r09.
IIO A NOTE ON MODERN NOMINALISM

Wang has complained that Goodman does not make explicit


what is meant by 'content'. 7 But quite apart from that, a platon-
ist of the modern type may not undertake a defence of extension-
al classes; he may agree that classes are constructions; he may
still insist that the basic entities are fully determinate individuals,
abstract or concrete, - the concrete individuals being unrepeat-
able and spatio-temporally unique, and the abstract individuals
being identically repeatable and qualitatively unique.
I am aware that this would not be far from the position ac-
cepted by Goodman himself. Qualitative ("abstract") particles
of experience and spatio-temporally bounded ("concrete")
particles are equally acceptable basic elements for him.8 Good-
man's system is not 'particularistic' like Carnap's. His basic
elements are abstract. He is a nominalist in whose system the
basic elements are abstract identically repeatable qualia. Such a
nominalism is welcome to the platonists, for was not the point at
issue between traditional realism and traditional nominalism
the question if there are or are not identically repeatable common
characters ?9
Goodman no doubt calls his basic elements individuals. But
his concept of the individual defies any attempt to assimilate it
to any of our common or philosophical notions of it. I would
only refer to the two interesting criticisms of this notion by Lowe
and Wang.1 0

III

Quine, I suppose, is more thoroughgoing in his nominalism.


It seems to me that he would not subscribe to the ontology of
qualities, for such an ontology takes predicates as names and is
implicitly committed to using variables within a quantifier to
refer to universals by "binding predicate letters", a procedure
against which Quine's campaign is primarily directed.

7 The Philosophical Review, 1953, pp. 413-4.


8 Journal of Symbolic Logic, 1947, NO.4, p. 105.
9 Cpo H. H. Price, Thinking and Experience, London, 1953.
10 V. Lowe, "Goodman's Concept of an Individual", The Philosophical Review,
1953. pp. II7-126; and H. Wang. "What is an Individual?", The Philosophical
Review, 1953, pp. 413-420.
A NOTE ON MODERN NOMINALISM III

"To be is to be the value of a variable".l1 Names alone are


substituents of variables ;12 the named entities are the values. It
follows that whatever is designated is or, better perhaps, is to be
the designatum of a name. Predicates are not names. But, why?
To say they are not names because they are not substituents of
variables or that they do not designate any entity is to state
that very thing in another language. In reply, we find that a
word to be a name must fulfil a condition, namely, the condition
that "existential generalisation" with respect to that word
must be possible.1 3 This means as is well-known that whereas
the singular existential statement "Calcutta is a city" entitles me
to infer the general existential statement "(3x) (x is a city)"
or "there is something which is a city", the statement "I7 is a
prime number" does not permit a similar generalisation of the
form" (3x) (x is a prime number)".
Against this argument and this methodological device, let me
state my objections as follows:
Firstly, the argument, if intended to prove that predicates are
not names, hopelessly fails, because it is incurably vitiated by
petitio principii. It assumes what it intends to prove, and it
conceals that assumption under the mystifying garb of logical
symbolism. But let me confess, I am not sure if it is intended by
its authors as a proof. Perhaps it is only an explication of the
ontological position they advocate; or, perhaps, it is only a pre-
scription of how predicate words should be used.
But, in any case - and this is my second point - it cannot be
said that this rule regarding existential generalisation conforms
to our ordinary use of predicate-expressions. For, as Warnock
has shown,14 we may, and also we do, use statements of the form
"there is something which is .... " with regard to predicate
words or the corresponding singular abstract nouns. So the mere
use, or absence of such use, by itself does not decide any issue,
nor can the ontological issue be decided by banning such use.
Thirdly, (here again I am indebted to Warnock) - Quine's

11 w. v. O. Quine, "Designation and Existence", Journal of Philosophy, Dec.,


1939, p. 708.
12 Ibid., p. 707.
13 Ibid., p. 706.
14 G. J. Warnock, "Metaphysics in Logic", Proceedings of the Aristotelian Society,
1950-51, pp. 197-222.
II2 A NOTE ON MODERN NOMINALISM

insistence on the ontological importance of the existential gene-


ralisation "embodies the presupposition that the question wheth-
er there are or are not abstract entities is just like the question
there is or is not a city called 'Leeds'." (Warnock). Whatever
else the platonists may have said, they have certainly laid stress
on the radical difference between the way in which particulars
exist and the way in which universals or Ideas exist. To reduce
both forms of existence to a common logical scheme - 'to be is to
be the value of a variable' - is not to do justice to what the pla-
tonist wanted to say. The word 'entity' deludes.

IV

To make clear what possibly could have been meant by the


platonist - I am not claiming that any platonist has actually
said this - let me draw upon some of the ideas of people who are
either nominalists or nearer the nominalists' camp. In his paper
"Form and Existence",15 Geach has sought to bring out the na-
ture of Form according to Aquinas, and has contrasted it with
the nature of Form according to Plato. I am not at present in-
terested in the question whether the contrast as developed by
Geach is historically true, that is to say, whether Plato and
Aquinas held the opinions ascribed to them in that paper. What
interests me is the possibility that the conception of Form may
be formulated in two different ways. For Aquinas, as also for
Frege -so we learn from Geach - what holds good of, or can meaning-
fully be said of, an individual does not hold good of, or cannot
meaningfully be said of, a Form. For example, manyness cannot
be meaningfully predicated of an individual; it is only Forms
that are repeatable and capable of manyness. An object is never
repeated; a Form alone may be repeated. If we bear this in
mind, the common platonic doctrine that a Form is a single
entity over against its many instances is to be rejected. So far
as this part of the argument is concerned, it seems to me that
there is something deceptive about the concepts of repeata-
bility', 'manyness' and 'oneness'. For it could be replied that a
Form is repeatable just because it is one, and its identity pre-
cisely is such as to make room for this repeatability; that no-
A NOTE ON MODERN NOMINALISM II3
thing is repeatable unless it retains its identity through repe-
titions. Similarly, an individual is one in the sense that its iden-
tity does not permit repetition; it does not admit of manyness in
the same sense in which a Form does. The two formulations,
therefore, the one ascribed to Plato and the other to Aquinas and
Frege, are not incompatible with each other. What interests me
here especially is the next step in Geach's argument as he goes on
to develop the conception of Form as a function. What consti-
tutes an example of a Form is, according to Geach, not 'wisdom'
simply, not 'the wisdom of Socrates', but 'the wisdom of. .. ' One
has then to admit into the very constitution of a Form a certain
openness, a certain indeterminateness, demanding determination.
Now, it seems to me that even if the platonist continues to think
of 'wisdom' as an entity, he may nevertheless make room for the
recognition that it is a peculiar entity of such a kind that it has
an indeterminateness about it, an indeterminateness that claims
determination.
When Quine tells US 16 that abstract terms are syncategore-
matic or when he prefers 17 to understand attributes as "open
sentences" rather than as propositional functions, he has the
same peculiar character of abstract entities in his mind. And pla-
tonism may quite well admit this element into its notion of
abstract entities. A similar suggestion, though couched in the
language of old-day philosophizing, is found in Stout's idea of
universal as a "distributive unity of particulars" .18 The uni-
versal is a unity, a unity which, as Stout emphasizes, is irre-
ducible to any of the other types of unity including the unity of a
substance. To this peculiar type of unity belongs a distributive
reference, an indeterminateness demanding determination or
individualisation. I should here add that I have already spoken
of two kinds of determinateness (and of indeterminateness).
The modern platonist speaks of determinate essences; but, again,
as I have just remarked, even the most determinate essence has
an indeterminateness (in quite another sense). 'Wisdom' is

15 Proceedings ot the Aristotelian Society, 1954-55, pp. 251-272.


16 Journal ot Philosophy, 1939, p. 708.
17 w. v. O. Quine, "On Frege's Way out", Mind, 1955, pp. 145-159.
18 G. F. Stout, "On the Nature of Universals", Proceedings ot the British Academy,
Vol. x.
114 A NOTE ON MODERN NOMINALISM

always 'wisdom of .... '; 'this specific shade of red' is also 'this
specific shade of red of .... '
Nothing that Quine has said serves the purnpose of demolishing
platonism.
I would conclude this note by referring again to that peculiar,
and often perplexing, blurring of the distinction between indivi-
duals and universals that we find not only in modern platonism,
but also in modern nominalism. Also, we have seen in part how
the position advocated by Goodman is a "meeting of extremes",
universalism and nominalism. I would now urge that the same
strain is also there in Quine's thinking. But for substantiating this
point, I can do no better than quote from Strawson's "Singular
Terms, Ontology and Identity" .19 Strawson rightly detects in
Quine, on the one hand, "a professed nominalism, the acknow-
ledged preference for an ontology of concrete particulars" and,
on the other, an "unconscious drive towards platonism, showing
itself in the consequences - not, of course, envisaged by Quine
himself - of the attempt to discard singular terms as fundamen-
tally superfluous" .
This perplexing situation, in which the controversy between
realism and nominalism finds itself today, demands that some
fundamentally new way of establishing the difference between
individuals and universals be found out. But to do that is not
the purpose of this note; to have drawn attention to the situ-
ation is enough for my purpose.

19 Mind, 1956, pp. 433-454.


CHAPTER XI

A RECENT CRITICISM OF THE


FOUNDATIONS OF
NICOLAI HARTMANN'S ONTOLOGY*

A system of thought, like a house with brick and mortar, has


a foundation and a superstructure. It is the superstructure which
attracts more attention, applause or censure. The foundation
remains unnoticed. Yet it is the foundation which hides the
secret sources of nourishment of the entire structure. This
foundation consists of a group of basic concepts and assump-
tions which the thinker brings into play. The greatness of a
thinker lies in the originality and strength of these concepts and
assumptions. The mediocre build on nothing new.
Most of the attention which has recently been paid to Hart-
mann's ontology has been concentrated on the various branches
of its widely ramified superstructure. The comparative neglect
of the fundamentals has led either to an exaggerated applause
or to an exaggerated censure.1 People have found in his philo-
sophy the most comprehensive 'system' of categories. Others
confess complete lack of interest and complain of his superficial
philosophical genius. Both these attitudes are born out of ex-
clusive attention to the superstructure of his thought. A critical
study of the foundation shows that Hartmann can stimulate
genuine philosophical analysis, and yet that his 'system' is not
the best part of what he has left for us.
* First published in the Journal of the Department of Letters, University of Cal-
cutta, I, 1957, 1-12.
1 This neglect may be noticed in the comparatively lesser attention paid to N.
Hartmann's best philosophical work Moglichkeit und Wirklichkeit, Berlin, 1939.
For example, a volume exclusively devoted to his work (Nicolai Hartmann, der
Denker und sein Werk, Gottingen, 1952) contains no study of his modal doctrine.
Yet it is the modal doctrine which, in Hartmann's case, constitutes the basis of the
entire thought-structure.
II6 FOUNDATIONS OF HARTMANN'S ONTOLOGY

That the modal doctrine constitutes the innermost basis of


Hartmann's thought has been most successfully demonstrated
by Josef Konig, Hartmann's successor at Gottingen. Konig has
attempted to canalise his own critical study of Hartmann's
ontology in a novel creative line. It is this Hartmann-Konig
controversy which the present paper seeks to present. The con-
troversy would throw light not only on the basis of Hartmann's
ontology, but on the basis of ontology in general.

II

How to get at the basis of Hartmann's ontology? Let us sketch


the superstructure, and then descend into the depths of the
foundation.
Following the Aristotelian tradition, Hartmann takes on-
tology as the science of beings as beings. Ontology is concer-
ned with what first makes beings beings. The word "Sein"
gives rise to the illusion, as if there is some entity or attribute
corresponding to it, something over and above, may be, under-
lying or pervading the various beings. Hartmann rejects this
thought. A science of beings as beings is not a science of any such
entity or attribute as Sein. On the other hand, it can only be a
science which lays bare the various spheres of being along with
their general and special categories and inter-categorial (hence,
inter-sphere) relations. Hence, ontology becomes a doctrine of
categories, a "Kategorienlehre".
There are two primary spheres of being: the real and the ideal.
The real consists of the chain of temporal events. The structure
of the real sphere is a stratification of various levels: the ma-
terial, vital, psychical and spiritual. The stratification consists in
the relation of "founding". The higher level is "founded" on the
lower. The lower provides the basis for the higher. The real
sphere has its general categories, those which determine the
entire sphere, irrespective of the differences of strata. Such
categories are, for example, the modal categories. But each
stratum of reality has also its own special categories. The re-
lation in which two levels of reality stand to each other is con-
cretely illustrated in the relation in which the categories of the
FOUNDATIONS OF HARTMANN'S ONTOLOGY II7

two stand to each other. The inter-categorial relations thereby


gain a new significance.
The ideal sphere consists of such elements as the essences
(compare Whitehead's 'eternal objects', Santayana's 'essences')
and the mathematical entities on the one hand, and the values
on the other. Hartmann's uncompromising realism rejects any
attempt at sUbjectivising. The ideal sphere has its own general
and special categories, just as there are also such categories as
are common to the two primary spheres: the real and the ideal.
Besides the two primary spheres, there are two secondary
spheres of being - the spheres of 'logic' and 'knowledge'. These are
mid-way spheres inasmuch as they share the categories of both
the primary spheres. (Compare Whitehead's 'hybrid' entities.)
To keep these primary and secondary spheres along with
their general and special categories before the mind, in their dis-
tinctions as well as in their interrelations, is essential for an un-
derstanding of Hartmann's ontology. Hartmann displays great
acumen in drawing these distinctions and in keeping clearly
apart what he considers to be distinct. Through these distinc-
tions, he claims to have the clue in hand for avoiding many of
the errors of the traditional ontologies.
The key to this entire discussion lies in the formulation of
the nature of the ideal sphere. In setting aside what he calls the
errors of tradition, Hartmann shows here his capacity at its
best. We shall do best to catalogue the errors which Hartmann
rejects and then pass on from this negative consideration to the
more positive aspect of the situation.

I. The distinction between the ideal sphere and the real sphere
is not the same as the distinction between Form and Matter.
Neither is the ideal sphere a realm of mere Forms, nor is the real
that of mere Matter. The idealities are also material in character
and the real is also formed content. The Form-Matter distinction
thus reappears within each of the two primary spheres and so
cannot be identified with the distinction between the spheres.

2. The real-ideal distinction is not identical with the distinc-


tion between 'concretum' and 'category'. The ideal sphere does
not determine the real as the categories determine the 'concre-
118 FOUNDATIONS OF HARTMANN'S ONTOLOGY

tum'. On the other hand, within each sphere, the 'concretum'-


category distinction reappears. There are real categories (i.e.,
categories of the real world) as well as ideal categories (i.e.,
categories of the ideal sphere). Further, the idealities possess a
primary mode of being, being for themselves ("Fur-sich-sein") ,
whereas the categories have no independent mode of being, but
are only in the 'concretum' which they determine or constitute.

3. Similarly, it is also wrong to identify the ideal with the


a priori and the real with the a posteriori. The distinction between
the a priori and the a posteriori is a gnoseological distinction, a
distinction between two modes of cognition, whereas we are
here concerned with an ontological distinction, a distinction
between modes of being. While the two pairs cannot be identified,
neither can we claim an exact correspondence between the two.
For there can be a priori knowledge both of the real world
(as in mathematical physics) and of the ideal, whereas it is
true that the real alone is apprehended a posteriori.

4. More plausible, but not any more tenable, is the identifica-


tion of the real with Dasein (existence, or 'that') and of the ideal
with the Sosein ('what'). The distinction between Dasein and
Sosein is a distinction between the two aspects of being and
should be kept apart from that between the spheres. The relation
between the aspects is conjunctive (that is to say, any particu-
lar being has both Dasein and Sosein) whereas the relation be-
tween the spheres is disjunctive (that is to say, any particular
being is either real or ideal). Ideal being, like the real, has its
own Dasein and its own Sosein.

5. There is another tradition, wrongly attributed to Plato but


first explicitly championed since Leibniz, according to which
the ideal sphere is a realm of mere possibilities whereas the real
is the world of actualities. To this the tradition adds as self-
evident that the actual is a selection from the many possibilities,
so that the real can be seen as an actualisation from amongst the
ideal. This tradition arose historically, as Faust shows 2 , in con-
nection with the theological doctrine of creation and later on
2 A. Faust, Der Mogtichkeitsgedanke, Heidelberg, 1931, 2 volumes.
FOUNDATIONS OF HARTMANN'S ONTOLOGY IIg

gained plausibility from the rise of abstract formal logic and of


alternative systems of logic and geometry. Hartmann devotes
much of his analysis to the task of demolishing this tradition and
as we have said, one of the principal results of his modal analysis
is that he has provided us with an alternative conceptual scheme
which demonstrates the illusory character of the self-evidence
which this tradition claims on its own behalf. A brief summary of
this modal analysis is of central importance for our purpose: -
Each sphere, according to Hartmann, has its own modal
category. The real sphere has its own modes of possibility,
actuality and necessity, just as the ideal sphere has its own modes.
Meaning by 'real-possibility' the mode of possibility in the real
world and by 'ideal-possibility' the mode of possibility in the
ideal sphere, we could say that the two are radically different
modes. By saying that they are 'radically different', we mean
that real-possibility and ideal-possibility are not species of a
common genus 'possibility', as their names illusorily suggest.
On the other hand, they are not only different possibilities, but
are, as possibilities, different. 'Real-possibility' has meaning and
significance only in connection with a real being, just as 'ideal-
possibility' is meaningful only in connection with an ideal being.
Only a real being can have real-possibility, whereas an ideal
being alone can have ideal-possibility. The suggestion of a com-
mon genus is illusory, so that the term 'possibility' becomes
devoid of any meaning of its own.
The same can be said of 'real-actuality' and 'ideal-actuality',
as well as of 'real-necessity' and 'ideal-necessity' (by construing
these terms as in the case of the pair in the above paragraph).
We are at present concerned directly with the modes of actu-
ality and possibility. 'Real-possibility' is to be defined as the com-
pletion of the series of conditions necessary for generating a
real event. A real event X is then and only then possible (that
is to say, 'real-possible') when the series of conditions required
for X's coming into being is complete, i.e. when the completed
series is given. 'Ideal-possibility' consists in compatibility, lo-
gical non-contradiction being a special case under it.
With this new conceptual scheme in hand, we readily begin to
see through the illusion which lends support to the tradition. It is
wrong to say that the ideal sphere is a realm of possibilities out of
I20 FOUNDATIONS OF HARTMANN'S ONTOLOGY

which the real is a selected actualisation. The ideal sphere has its
own modes of possibility and actuality, just as the real world has
its own. There are the modes of 'ideal-actuality' and 'real-possi-
bility' which the tradition overlooks. And, the real-actual is an
actualisation of 'real-possibility', but not of 'ideal-possibility'.
N either in the real world nor in the ideal, is there mere possi-
bility, that is to say, possibility which is not actualised. In the
real world, an event is possible only when the series of con-
ditions is completely given. But when this series is complete, the
event is also actual. That is to say, what is 'real-possible' is si-
multaneously, at that very instant, also 'real-actual'. Similarly,
in the ideal sphere actuality being only a secondary mode, a
mere shadow of the mode of possibility, what is 'ideal-possible' is
already 'ideal-actual'. As such, the usual notion that the range of
possibility is wider than that of actuality is false. It does not
hold good of either of the primary spheres of being.
By rejecting the above errors of tradition, Hartmann gives a new
form to the two-world theory which ever since Plato has been a
recurrent philosophical motive. This novelty may be stated with
regard to the problem of the relation between the two spheres.
The alternatives in terms of which the tradition of philosophy
has formulated its answer are the following: -
Either, the relation is one of determination, the precise na-
ture of the determination being conceived mainly in two ways.
The ideal sphere may determine the real world as the categories
determine the concretum they constitute. Or, the ideal sphere
may be the telos, the perfection towards which the real aspires.
Both forms of determination - whether categorial or teleolo-
gical - are rejected by Hartmann as being distortions of the si-
tuation. The identification of idealities with categories has al-
ready been shown to be erroneous. Teleological determination a-
gain is out of place in a purely ontological situation. The ideal
and the real are two primary spheres of being and there is no
degree of being, no scale of perfection in order of being. Not only
is the ideal sphere not more perfect than the real, but one can
rather say the reverse, if one can at all speak of more or less
perfection. Hartmann does suggest a reversal of the usual judg-
ment of value by demonstrating the superiority (?) of the real
world over the ideal.
FOUNDATIONS OF HARTMANN'S ONTOLOGY I2I

Or, the real may be conceived as being a selection from the


ideal. This again may be supplemented by the theological appeal
to a God who chooses the best of all possible worlds. The above
modal doctrine has already exposed the fallacy underlying this.
With this, room is left only for a phenomenological approach.
To talk of determination - causal, categorial or teleological -
is to indulge in speculation. Phenomenologically, we have the
following before us: - firstly, we have the Fiir-sich-sein of the
two realms of being. Secondly, we have also a certain degree of
inter-weaving of the two realms. This is illustrated, for example,
in the possibility of mathematical physics, in general, of a priori
knowledge of the real world.
Allowing for the least measure of speculation, how are we
then to formulate the relation between the spheres? Hartmann
attempts this in his own doctrine of partial categorial identity.
The two realms, being autonomous and primary, have their own
categories. What we have said to be the phenomenon of inter-
weaving of the two realms is nothing but a partial identity be-
tween the categories of the two.
To ask further about the rationale of this identity, as to its
'why' and 'how', is for Hartmann, to ask those ultimate meta-
physical questions which offer a limit to solvability. Such ques-
tions point beyond phenomenology, but cannot be themselves
answered.

III
The significance of all this for ontology is great. The separa-
tion between the two spheres is now complete. The autonomy
and independence of each from the other is established. With the
radical separation between the modalities of the spheres, the
crux of the situation is reached. The modalities are the most
fundamental categories. They are the categories which along
with the inter-modal relations bring into concrete relief the mode
of being (Seinsweise) of each sphere. With demonstration of the
radical difference between the modal categories of the two spheres,
absolute separation between the spheres is set on a sure footing.
This is further strengthened by rejection of the possibility of
any 'influence' or 'determination' of the one by the other.
1ZZ FOUNDATIONS OF HARTMANN'S ONTOLOGY

Ontology, therefore, divides itself into two special branches:


an ontology of the real world and an ontology of the ideal sphere.
Let us call them for the sake of convenience - without however
attributing the terminology to Hartmann himself - 'Realontologie'
and 'Idealontologie' respectively. Since the two spheres are now
separated radically from each other, each has its own mode of
being, its own categories, each of the two special ontologies must
be autonomous. The categories and structures of the real world
can be described without reference to the categories and struc-
tures of the ideal sphere. Hartmann thus aims in the first two
volumes of his ontology, at laying the foundation of an autono-
mous Realontologie. More and more, it is the real world which
comes to be the principal theme of his philosophizing. The dignity
of the real world forces itself upon him in contrast with the sha-
dowy airiness of the idealities. After the general considerations
of the first two volumes have laid the foundations, the special
Kategorienlehre in the last two volumes is devoted entirely to the
real world. In all this, Hartmann's motive is to develop an auto-
nomous Realontologie, an ontology of the real, independent of,
and without reference to, the ideal considerations.
Is such an autonomous Realontologie at all possible? This
is a question which is of central importance for ontology and for
philosophizing in general.

IV

In Archiv fur Philosophie (1948), Josef Konig (then of Hamburg)


published a paper entitled "Uber einen neuen ontologischen Be-
weis des Satzes von der Notwendigkeit alles Geschehens", which is
meant to be a criticism of a very special doctrine of Hartmann's
ontology, but which lays down the principle for questioning the
very foundations and possibility of an autonomous Realontologie.
The special doctrine which Konig's paper seeks to examine
is the doctrine of thoroughgoing necessity in the real world.
The modal analysis which keeps apart the modalities of each
sphere makes the distinction between "ideal-necessity" and "real
necessity". "Ideal-necessity" is the necessity with which one
essence includes or excludes another, a mathematical conclusion
FOUNDATIONS OF HARTMANN'S ONTOLOGY I23

'follows' from its premise, or with which a logical system hangs


together. "Real-necessity", on the other hand, is the necessity
with which the chain of real events is constituted, so that nothing
in the chain is fortuitous or accidental.
The thoroughgoing necessity which characterises the real
world is not to be identified with any particular type of necessity,
either causal or teleological. To say that the real world has thor-
oughgoing necessity does not mean that there is strict causal
determination all through the real process. The real world has a
stratified structure and it may be reasonable to suppose that
the various strata exhibit differing types of necessity. But all
through the various strata there runs a common thread, a thor-
oughgoing determination which allows no accident, a Real-
zusammenhang which makes of real events a continuum-like
interconnectedness.
Hume missed this basic Realzusammenhang. Hume's inquiry
failed, because he was searching for the wrong thing at the wrong
place. He was searching for "ideal-necessity" in the real process,
- a search which, by its very nature, is doomed to failure. 3
Konig's criticism was directed against this doctrine of Real-
zusammenhang. The starting point of this is an examination of
what Hartmann means by a real being. A real being of Hart-
mann is a real event as happening here and now. It is only with
regard to such a real event as happening here and now that the
idea of real-necessity meaningfully holds good. Precisely this is
what Konig contends to be absurd.
The idea of necessary connection between A and B implies the
idea 'if A, then B'. This latter implies the repeatability of both
A and B. But if A and B be real events of Hartmann's conception,
they cannot by their very nature be meaningfully thought as
repeatable. Since, as said before, the real event of Hartmann is
thought of as happening here and now, which means that the
idea of A's existing here and now is included within the very
idea of A's being a real event, the idea of repeatability and hence
also the idea of necessity cannot be meaningfully predicated of A.
That his own criticism of Hartmann's doctrine of Real-
necessity is but an exemplification of a more general principle
of criticism is suggested by Konig himself. This at once brings
3 Cpo Whitehead's criticism of Hume.
I24 FOUNDATIONS OF HARTMANN'S ONTOLOGY

out, according to Konig, the novelty as well as the absurdity of


Hartmann's ontology.
Traditional ontology since Aristotle had as its subject matter
for theorising the scale of being from the summum genus down to
the last species which is also a genus. This means that the par-
ticular, the individual was always excluded from the subject
matter of theory. The particular as such was recognised to be a
limit to theory.
The novelty of Hartmann's ontology lies in an attempt to
make this particular as such the subject matter of theory. Hart-
mann's real being is a particular taken as such, i.e., with its
transient particularity. This novelty is also the highly paradoxi-
cal character of Hartmannian Realontologie.
Konig seeks to clarify this situation with the help of a dis-
tinction whose importance for logic and ontology he has been
emphasizing through his published papers and university lec-
tures. This is the distinction between 'theoretical this' and 'pract-
ical this'.
The 'this' may mean 'such and such', or it may mean the
'this-there' (identified ostensively). In the former case, 'this'
also means 'this sort of'. In the latter case, the 'this' implies a
'pointing out', - an act which involves somebody for whom the
'pointing out' is meant. For these two kinds of 'this', Konig
chooses the terms 'theoretical this' and 'practical this'. If A be a
real being of Hartmann, A is a 'practical this'. By its very nature,
A can only be pointed out as 'this-there'. Of such an entity, no
theoretical statement can meaningfully be made. 4
The principle implied in Konig's above-mentioned paper
(whose explicit intention was to question Hartmann's doctrine
of real-necessity) has been developed by the present author in
his Gottingen thesis entitled "An inquiry into the problem of
ideal being in the philosophies of Nicolai Hartmann and A. N.
Whitehead"5 with a view to extend its scope to cover the entire

4 The corresponding distinction in logic is between 'theoretical sentences' and


'practical sentences'. This distinction was laid down by Konig in his Vorlesung on
'Theoretische und praktische Siitze' during the summer semester, 1953, and the winter
semester, 1953/54 at the Gottingen University.
5 Now published under the title Nicolai Hartmann and Alfred North Whitehead:
A study in recent Platonism, Calcutta, 1957.
FOUNDATIONS OF HARTMANN'S ONTOLOGY I25

basis of Hartmann's ontology. Thereby, the very task of ontology


is questioned from a fresh point of view.
We can interpret Konig's suggestion in our own language
thus: - Absolute separation between the two primary spheres of
being leads to an attempt to build an autonomous ontology of
the real world. The real, when it is robbed of all ideal elements
and structures, is nothing but the transient particular, the mere
'this-there', Konig's 'practical this', Hartmann's real being.
The elements and structures in the real world which are essen-
tially repeatable and recognisable are all ideal. In the stream of
real events, taken as real happenings, nothing is repeatable.
Of such events, no theory can be made. And yet Hartmann's
concepts of 'real-possibility' and 'real-necessity' are elements of
such a theory which Hartmann seeks to build up.
This is supported by a critical examination of Hartmann's
philosophy of ideal being and the way the two spheres have been
kept asunder. It becomes clear that the way the spheres have
been kept apart depends upon and, in its turn, influences Hart-
mann's formulation of the nature of ideal being.

v
Comparison with the cosmology of A. N. Whitehead affords us
with an important case of the principle involved here. White-
head's distinction between actual entities and eternal objects is a
parallel to Hartmann's two sphere theory. But also like Hart-
mann, Whitehead recognises two secondary spheres of 'hybrid'
entities: propositions (corresponding to Hartmann's sphere of
logic) and feeling (which includes Hartmann's sphere of know-
ledge). Thus the ontology and the cosmology present the same
external pattern. But their inner motives and executions show
great differences and these differences illustrate the principle of
criticism suggested in the above section.
To bring this out, the course of development of Whitehead's
platonism must be mentioned. The early works on natural phi-
losophy had introduced the distinction between 'events' and
'objects'. 'Objects' were those elements in nature which were
'recognisable', which are the same. 'Events' on the other hand
I26 FOUNDATIONS OF HARTMANN'S ONTOLOGY

were the passing, transient, spatio-temporal factors. White-


head's account of 'objects', which cannot be elaborated here,
shows that in this early formulation, the separation between
'objects' and 'events' was absolute. 'Events' were what 'objects'
were not; 'objects' were what 'events' were not. When one re-
views the different classes of 'objects' ('sense-objects', 'percep-
tual objects', 'scientific objects') which Whitehead admitted,
one realises that the 'events' were nothing but the passing 'point-
instants', so that every other content of our experience was
itself an 'object' or composed of factors which were themselves
'objects'.
In the later cosmology, the nature of this distinction under-
goes a significant change. The category of 'events' is now re-
placed and enriched by the category of 'actual entity'. The
'objects' are raised to the dignity of being 'eternal objects'.
Whereas in the earlier natural philosophy, 'events' were only
'thin slabs of duration', 'point-instants', now the actual entities
are recognised to be enduring, to have all the richness of content
within themselves. The 'perceptual objects' and the 'scientific
objects' of the former phase are no more now eternal objects,
but actual entities. The only eternal objects are the so-called
'sense objects' (the eternal objects of the 'subjective species')
and the mathematical entities (the eternal objects of the 'ob-
j ective species').
While thus the two categories underwent a change, what is
directly relevant for us is the following: -
Whereas the 'events' and 'objects' mutually excluded each
other so that the 'events' were what the 'objects' were not and
vice versa, now the actual entities and eternal objects stand in
organic relation with each other. The actual entities are what
they are because of the eternal objects which have found 'in-
gression' in them, so that no description of the actual entities is
possible without referring to the constituent eternal objects.
Hartmann's distinction between real being and ideal being is as
radical, as absolute as Whitehead's early distinction between
events and objects. We have seen that this way of keeping the
two spheres asunder makes it impossible to theorise about the
real sphere. In spite of his two entity theory, Whitehead's later
cosmology avoids this error. Whitehead realises that any des-
FOUNDATIONS OF HARTMANN'S ONTOLOGY 127

cription of the actualities must necessarily take into considera-


tion the eternal objects ingredient in them. Whitehead's cos-
mology therefore, though eminently concerned with the real
actual world, is no autonomous 'Realontologie' in the above
specified sense. 6
The real as such, as merely real, as the stubborn matter of
fact, "die Harte des Realen" is not a subject matter of theory.
Theory always transcends this stubborn matter of fact and, in its
attempt to come back to it, always falls short of the ideal. Actu-
ality, in Whitehead's language, must be described in terms of the
eternal objects and yet cannot be exhausted by them. This is the
significance of Whitehead's doctrine of "infinite associate hier-
archyof eternal objects", developed in the chapter on "Abstrac-
tion" in Science and the Modern Wodd.
The real as such, as merely real, as the stubborn matter of
fact is reached through practical relationship, vital-emotional-
organic situations. Both Whitehead and Hartmann recognise
this. That this also imposes a limit upon the scope of theory is not
recognised by Hartmann.
A significant suggestion is made by Whitehead when he tells
us that 'theory' is a 'hybrid' entity. That the ontological 'Fra-
gestellung' can and should be extended to ontology itself is not
seen by Hartmann. Hartmann's blind faith in the capacity of
intentio recta covers up this vision. Direct access to reality is
possible only through vital-emotional-organic relationship.
In theory, language intervenes. It is interesting for this purpose
that Whitehead does not distinguish between 'proposition' and
'theory', and hurls them together as constituting one class of
'hybrid' entity. Hybrid entities are constituted of both eternal
objects and actual entities. Hartmann recognises this of logic,
but not of ontology.
Theorising about actuality must involve reference to possibil-
ities. Theory moves in the realm of meaningful possibilities. The
so-called ideal-possibility is the only theoretical possibility. The
so-called real-possibility of Hartmann is no meaningful concept

6 The Process and Reality also makes a distinction between 'general' possibilities
and 'real' possibility, but the latter is not an independent and autonomous mode,
but only a limitation of the former.
Iz8 FOUNDATIONS OF HARTMANN'S ONTOLOGY

within the realm of theory. It may be designated practical possi-


bility 7, - a concept which is in need of further elaboration and
specification.

7 This suggestion was given to the author by Josef Konig inthe course of a private
conversation, although the responsibility of using it in the present form is the pre-
sent author's own.
CHAPTER XII

REMARKS ON NICOLAI HARTMANN'S


MODAL DOCTRINE*

In his Moglichkeit und Wirklichkeit (Berlin I937, 2nd ed.


I949), Hartmann gives us an ontological theory of the modes. He
starts from a distinction between the modes of the various sphe-
res of being, primary and secondary. The two primary spheres of
being, according to his ontology, are the real and the ideal.
The two secondary spheres are those of logic and knowledge.
The modes of the real world are accordingly contrasted with
those of the ideal realm; the modes of the realm of logic are
again different from those of knowledge. The modal doctrine is
thereby divided into four parts. But there must be also a part
on the relations between these different spheres.
Traditional discussion of the problem of modality did not see
clearly through these distinctions. This gives to Hartmann's
treatment of the problem its originality. Further, these modes of
the various spheres are distinguished from the naive day to day
consciousness of modality.
The ontological point of view requires specificatjon. For this
purpose, we are to distinguish between three different approaches
to the problem of modality:
First, it is possible to consider the modalities as criteria for
classifying all objects in the three groups, those that are merely
possible, those that are both possible and actual, and those that
are possible, actual and also necessary.
Secondly, it is possible to consider the modes as if they were
different stages of a process. Thus, it may be said that a thing

* First published in Kant-Studien, 54, 1963. pp. 181-187.


130 REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE

first becomes possible, then is made actual, and further mayor


may not be necessary. The process however may not be carried
to the end; what is possible may never be actualised.
Thirdly, the modes may be taken neither as criteria nor as
stages of a process, but as the constituent aspects of the existent
or the subsistent, as the case may be. This is the point of view
which we may call the critical point of view, because we may
trace it to Kant. Kant starts from the given object of experience
and then asks how the same is possible, actual and necessary.
Hartmann rejects the first two approaches. Modes are for him
neither criteria nor stages, but the most primary characteristics
of the being of anything. As such, given an object of experience,
we can ask: what makes it possible? What makes it actual? What
makes it necessary? Thus in an important sense, Hartmann's
treatment of the problem is similar to Kant's, even though
Kant's own solutions are rejected by Hartmann. For Kant,
the given is possible when considered in relation to its form and
actual when considered in relation to its matter. Hartmann
finds this not only inadequate but also misleading: to this how-
ever we shall turn later on.
The second approach is attributed to Aristotle. Both the
first and the second approaches attribute to the merely possible
which is not tor has not yet become actual' a sort of ghostly
existence - a position in between being and non-being. Aristot-
le's doctrine of dynamis and energia is further criticised as an
illegitimate extension of the categories of the sphere of organic
being to the entire domain of being. Further, if a prior stage of
mere possibility is admitted, the question arises as to what must
be added to it in order to render it actual. Kant had shown that
any answer to this question is absurd. For that which must be
so added, argued Kant, must be other than the possible, that
is to say, must be impossible. 1
As such, we come back to the critical formulation of the question.
This is one of the points where we begin to see the influence of
Kant on Hartmann's ontology which claims the name of critical
ontology.

1 I. Kant, Kritik der reinen Vernuntt, Kant Werke III, p. 206, ed. by Cassirer.
REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE I3I

II

Hartmann's entire discussion of the mode of possibility sets


out from a distinction between disjunctive and indifferent possi-
bility. There are two kinds of possibility. The one always implies its
contradictory: that is to say, the possibility of A always implies
the possibility of non-A. If it is possible that it rains, it is also
possible - we could even say that the first possibility implies this -
that it does not rain. Or, the possibility of the being of A implies
the possibility of the non-being of A. Hartmann calls this kind of
possibility disjunctive, because it hides a disjunction within it.
The other kind of possibility is called indifferent because from the
possibility of A nothing can be said about the possibility of non-A.
Thus the indifferent possibility does not completely exclude its
contradictory, but neither does it necessarily contain the latter
within itself.
Now, when the possible becomes actual, it does not cease to
be possible. In the language of the critical point of view, we can
say that an actual A also contains its own possibility. That is to
say, possibility and actuality are not mutually exclusive modes
but go together. An A which is given in experience is at the same
time actual and possible, actual from one point of view and possi-
ble from another.
At this stage, the question arises: which of the two kinds of
possibility above enumerated is so compatible with actuality?
Hartmann shows that the so-called disjunctive possibility cannot
remain contained in an actual A. That is to say, when A is actual,
it can contain only its own possibility but not that of its con-
tradictory. The actuality of A is incompatible with the possi-
bility of non-A. The so-called disjunctive possibility is not there-
fore admissible into a critical philosophy of modality. On the
other hand, it is connected with the view of possibility as a
state, in which case the modes exclude each other and no question
of compatibility arises.
At least in the real world the reis no disjunctive possibility.
At least in the real world the actuality of A contains the possi-
bility of A but not that of non-A. In the real world therefore the
so-called disjunctive possibility is split up into two different modes:
I32 REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE

the possibility of A and the possibility of non-A. These are, in


the real world, two different modes and not members of an in-
separable disjunction.
With this splitting up of the so-called disjunctive possibility
into two different modes, the ground is prepared for Hartmann's
most important doctrine in the Moglichkeit und Wirklichkeit,
i.e., the doctrine of "Real-moglichkeit". This doctrine finds its best
expression - for Hartmann formulates it in diverse ways and
from diverse points of view - in his short formulation of the so-
called "Realgesetz der Moglichkeit": "Was real moglich ist,
das ist auch real wirklich". What is possible is also actual: this
holds good of the real world 2 • This law, according to Hartmann,
is "die kurze Formel fur einen umstiindlichen Revolutionsprozess
im philosophischen Denken". "Man kann dies en Prozess als die
A ustreibung der Gespenster aus dem Weltbilde des M ens chen
bezeichnen."3 This ghost is the merely possible in the real world.
The real world does not contain mere possibilities that are not
also actualised. On the other hand the real world is such that here
possibility is only a structural element of the mode of being of
an actual.
This doctrine of Hartmann, revolutionary as it is, has been
variously criticised. We shall however concentrate on under-
standing what exactly Hartmann aims at. The conviction which
he carries in his exposition of the doctrine is certainly based on
some bit of truth. We shall aim at finding this out. For this
purpose, the so-called "Spaltungsgesetz" puts us in a favourable
vantage position.
The above mentioned "Realgesetz der Moglichkeit" is proved
(or, rather demonstrated) by Hartmann in two different ways.
The one is called the formal proof, the other is called the material
proof.
The formal proof starts from a law which Hartmann claims to
be self-evident. This is the so-called "SPaltungsgesetz". His task
consists in demonstrating that the "Realgesetz der Moglichkeit"
follows analytically from this self-evident "Spaltungsgesetz".
Hence the proof is called formal.
The "SPaltungsgesetz" runs thus: "Was real moglich ist,
2 N. Hartmann, Moglichkeit und Wirklichkeit, Berlin, 2nd edn. 1949, ch. 21.
3 Ibid., p. 176.
REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE 133

dessen Nichtsein ist real nicht moglich". If something is possible


in the real world, its non-being is not possible in the real world.
The so-called disjunctive possibility being split up, the possibility
of A and the possibility of non-A are now two distinct, mutually
incompatible, modes. Though this "SPaUungsgesetz" is claimed to
be self-evident, nevertheless it requires to be clarified. This is
done by Hartmann in the following way: actuality presupposes
possibility; "presupposes" means here "contains". Only the
possibility of A can be contained in the actuality of A; the
possibility of A's non-being cannot be contained in A's actuality,
for if that were so then what is already actual could also be non-
actual which is impossible.
Once the "SPaltungsgesetz" is accepted as valid, Hartmann
claims, the "Realgesetz der Moglichkeit" follows by formal im-
plication. Let us therefore examine the meaning and the vali-
dity of the "SPaltungsgesetz":
The argument is this: because the possibility of A is contained
in the actuality of A, the possibility of non-A is excluded there-
from. Is this an immediate inference? No. The inference is rather
indirect, mediated by another premise: the possibility of non-A
implies the non-actuality of A. Since the actuality of A and the
non-actuality of A exclude each other evidently, the possibility
of A and the possibility of non-A also must exclude each other.
Is this new premise, i.e., the proposition that the possibility of
non-A implies the nonactuality of A self-evident? Hartmann does
not demonstrate this premise. He takes it as self-evident. On the
other hand, the absurdity which he brings out in the supposed
coexistence of positive and negative possibilities in the actuality
of A depends upon the truth of this premise.
We may suggest a vindication of this proposition. It may be
argued that this follows from the very nature of the critical
approach to the problem of modality. The critical approach as
we have seen, considers possibility not as a phase preceding
actuality but as a structural element in the mode of being of
anything. If that be so, one cannot meaningfully speak of the
possibility of non-A unless non-A is actual. And when non-A is
actual, A is certainly non-actual. From the critical point of
view, therefore, the possibility of non-A implies the non-actu-
ality of A. Hartmann's "SPaltungsgesetz" follows.
134 REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE

III

In the face of the above results, the following questions may


further be raised:

Granted that Hartmann's 'real-possibility' is a defined term,


1.
we may still ask - as we are entitled to ask in case of every de-
fined term - if there is really anything corresponding to it.
This question may be answered affirmatively. The ontological
situation corresponding to this term has been brought to light.

2. But we also ask: is this term in accordance with the meaning of


the term 'possibility' as used in ordinary language? This is a
difficult question. And it is interesting to find that Hartmann
recognizes it. The common usage of language is not fixed enough.
Hartmann would suggest a double answer: first, he would say
that the common language has disjunctive possibility as its
meaning. But, secondly, Hartmann would suggest that his own
use of the term is not totally strange to common usage, that we
do often use the term in his sense without being sufficiently
conscious of the fact that that would involve a radical change in
the philosophical point of view. That Hartmann tries to establish
this harmony with common usage gives to his doctrine a large
degree of plausibility.
Hartmann asks: "W as eigentlich meint man, wenn man sagt
'Was nicht ist, das kann doch sein', oder ,es ist vieles moglich,
was nicht wirklich ist'? M eint man denn im Ernst, dass dieses
,Mogliche' auch zur Zeit schon wirklich sein ,konnte'? Oder auch
nur, dass es unter beliebigen kunftigen Umstiinden wirklich werden
,k6nnte'? Man weiss ja doch sehr gut, dass dem nicht so ist. Man
weiss, dass es nicht wirklich werden kann, ohne dass noch mancherlei
V oraussetzungen dafur sich verwirklichten. Man weiss dass eine
ganze Kette von Bedingungen erfullt sein muss, bevor es wirklich
werden kann. Und wenn man sich die Sache zum Ziel macht und
sie aus eigener Kraft verwirklichen will, so weiss man nur zu wohl,
dass man die noch fehlenden Bedingungen erst beschaffen muss"4

4 Ibid., p. 155.
REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE 135

In this paragraph, Hartmann tries to show that his doctrine


is already contained in the popular use of language. Let us exam-
ine what this means for the understanding of Hartmann's
doctrine and its validity as a modal doctrine.
It seems that here Hartmann plays on a certain ambiguity of
language. Noone denies that the possible cannot be actualized
till the necessary conditions are all there. But does this mean that
the so-called possible is in fact impossible so long as the necessary
conditions fail?
The really decisive question which one must ask is whether
possibility as such is possibility of being actual. Is "Moglich-
keit" = "Wirklich-werden-konnen" ? To be actual is to be actual at
a definite time and place, so that the "Wirklich-werden-konnen"
really breaks up into as great a number of "Hier-und-jetzt-wirk-
lich-werden-konnens" as there are here-and-nows in the process.
In Hartmann's modal theory, possibility is a relational mode and
actuality is the fundamental mode: since the relational mode
presupposes the fundamental, possibility as such is also possi-
bility of being actualized. Something can be actualized only when
the necessary conditions are all there. It follows that something
is possible only when the conditions are all there ("is possible" =
"can be actualized").
We can see that in our common language there is this ambi-
guity. Sometimes we say, "It has not yet been possible on my
part to arrange the affair". In this case, we really mean that the
arrangement of the affair has not yet been actualized by me.
Hartmann's theory conforms to this ambiguous use.
Let us consider some other possibility-sentences:
(a) "It is possible that it rains to-morrow": Hartmann may
say that this statement conveys not ontological possibility but
epistemological possibility, and is due to ignorance or partial
ignorance on our part of the actual conditions.
(b) "Wood may burn" or "Poison may kill": Hartmann may
say wood may burn only under certain conditions, and poison
can kill only under certain conditions. No one denies this. These
conditions however may be included within the judgment of
possibility without implying the actuality of those circumstances.
We could then say: "Wood may burn under circumstances x, y .. ".
Weare not speaking of any particular actual piece of wood.
136 REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE

Hartmann would then say that the judgments express essential-


possibility, Wesensmoglichkeit, and not real-possibility.
(c) What about such an event as the death of a man X to-day
at this moment? Hartmann would say that this death has been
made possible by a completed chain of conditions. Before this
moment of his actual death, his death was really impossible.
It is clear that in this case "has been made possible" is a phrase
which is also used in language but is supposed to mean "has been
made actual". Hartmann js playjng on the ambiguity.
If that be so, can we further say: (a) while the chain of con-
ditions have brought about, i.e., actualized the death of X today
at this moment, (b) the possibility of his death was always there
(imbedded in the essence of X's being a man)? The absence of
the necessary conditions did not make it possible only in the
sense that it could not be brought about. It is clear that (b) would
be called by Hartmann Wesensmoglichkeit. Hartmann agrees that
this Wesensmoglichkeit underlies real-possibility. But where, in
common usage, is the idea of real-possibility implied if not in the
above mentioned ambiguity?

3. We may now pass on to a further question. It has been


said above that Hartmann's theory of real-possibility is not a
purely modal theory. What does this mean? What is a purely
modal theory?
We should for this purpose clearly separate two questions.
The one is the empirical question, if and when something will or
can happen. This question may variously be answered. The talk
of chance or of probability may be explained as being due to
imperfect knowledge of the real situation. Whether there is an
objective probability or objective indetermination arises here
and is to be discussed. The predictions of the sciences are on this
empirical basis. They answer questions as to whether and when a
certain anticipated event will happen. Similarly, when a certain
event happens, we can consider the conditions which have actu-
alized it.
On the other hand, there is the purely modal question, whether
something is possible or not. Or, to state it in the language of
critical philosophy, in what sense something given in experience
is to be regarded as possible?
REMARKS ON NICOLAI HARTMANN'S MODAL DOCTRINE 137

The answer to the empirical question is independent of the


modal enquiry. The working scientist who is deciding such
empirical questions need not have a modal theory to work with.
On the other hand, granted that something satisfies the require-
ments of modal possibility, the question arises as to whether it is
going to be actualized, and if so, under what conditions.
Hartmann's doctrine of real-possibility is an empirical theory.
Hartmann would probably say that to the above two questions
corresponds his own distinction between real modes and ideal
modes. But, a purely modal analysis does not require Hartmann's
division of modal enquiry according to the ontological spheres.
The modes must be common to the spheres. A purely modal
theory must demonstrate the unity between the spheres and be a
unified one. We are to ask: in what sense is it meaningful to
speak of the possibility, actuality and necessity of anything -
without consideration of whether this something is a real event
or an ideal entity.
We thus see that Hartmann's own doctrine of real-possibility
still leaves room for a more comprehensive, more truly modal
doctrine of possibility. Hartmann's theory of real modes serves
the purpose of giving us a pictur.e of the inner structure of the
real world. But it does not satisfy the needs of a genuine modal
analysis.
CHAPTER XIII

THE 'OBJECT' IN
EDMUND HUSSERL'S PHENOMENOLOGY*

Phenomenological enquiry concerns two main questions. First,


there is the enquiry into the nature ofpu re subjectivity and the
problems bearing on it. Second, there is the problem of the 'con-
stitution' of objectivity "as referred to its subjective source."l
In a paper on "Husserl's Phenomenology and Indian Idea-
lism",2 we had the first question principally in mind. Our inter-
est there was to examine the consistency and success with which
Husserl found room in his philosophical system for the idea of
pure sUbjectivity. Husserl's treatment of the second problem
received only a secondary consideration i.e., only so far as it
corroborated the main thesis of that essay. We were considering
how Husserl was struggling - though in vain - to free himself
from the shackles of objectivity.
But just as at times Husserl would have the courage to make
extreme assertions in favour of the single principle of subjecti-
vity, so also at other times he would explicitly refuse to limit pheno-
menological study to the investigation of pure consciousness
alone. More Hegelian at such moments, he would ascribe to
phenomenological enquiry as its main task the study of the re-
lation of the principle of objectivity to its source in the subjective
realm. Phenomenology is not merely a study of immanent for-
mations of the pure consciousness, but also of the intentionality
of such conscious formations and conversely of the 'constitution'
of the objectivities corresponding to such intentional experien-
ces.

* First published in Philosophy and Phenomenological Research, XIV, 1954,


pp. 343-353·
1 E. Husseri, Ideas, E. Tr. by Boyce Gibson, (paperback edition), p. 234.
2 In Philosophical Quarterly, Amainer, India, Vol. XXIV, NO.3.
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY I39

In any case, as our previous study has emphasized, there lurks


a fundamental contradiction in this way of looking at the pro-
blema. On the one hand, there is the recognition of conscious-
ness as constituting a self-contained and self-sufficient realm by
itself. On the other hand, intentionality is said to be a necessary
and universal feature of all consciousness; every consciousness
must be ot something. This something may, of course, be some
other experiential process, as in the case of what HusserI calls
'immanent experiences'. But it is, in most cases, some ob-
j ect. How a self-contained system can get associ a ted with an 0 bj ect
is thus a problem of major importance in this context and we
have seen how HusserI's solution or for that matter any consis-
tent solution on such premises - tends towards an extreme variety
of Idealism on the lines of the Vedanta of Sankara.
But we should be true to HusserI also. If he does evince
any such tendency, our dealings with him must yet be true to the
basic ideas of this thought. And even if we did follow him on the
lines of this particular tendency of his thought, we are yet ob-
liged to follow him right through the basic concepts guiding his
philosophic excursions. And that we propose to do in the pre-
sent essay.

II

HusserI's treatment of objectivity starts basically from the


recognition of intentionality as the principal feature of conscious
life. This doctrine of intentionality had been handed down through
Brentano. The distinguishing feature of conscious life had
been recognized as lying in the fact that all consciousness
is of some object. Recognition of this feature opens up, for
HusserI, a great variety of complex problems of phenomenolo-
gical importance. What is the meaning of this 'of' relation, when
it is said that consciousness is always 'of' some object? If con-
sciousness constitutes a realm of its own, how does it come to be
related to the object? Further, what meaning are we to attach
to the 'act'-hood of conscious acts, when we bring under this
general name a great variety of processes from presentations
3 I do not any longer think this criticism to be correct - J. N. M.
I40 'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY

and judgments to willing and desiring? The problem of 'meaning'


creeps in when we are asked how this reference to object is to be
intelligible.
The straightforward doctrine of intentionality is transformed
phenomenologically by Husserl into the more complex doctrine
of noetic-noematic structure of experience. 'Noesis' is a general
name for all objectifying acts; 'noema' for all objects of such
acts. The two terms respectively cut across a great variety of
'acts' and a correspondingly great variety of 'objects' of these
'acts'.
The problem of objectivity is, however, the other side of the
intentionality situation. If all conscious acts refer to some object,
it is no less true that all objects are 'constituted' by some ob-
jectifying act or other. This, as we shall see, gives a predominance
to the act over the object, to the noesis over the noema. The re-
lation is not merely one of equal partnership.
Helmut Kuhn 4 finds two arguments which exhibit this pri-
macy of the noesis over the noema in Husserl's philosophy.
First, analysis of 'act' shows that the act as a whole does not be-
come visible to us till we free ourselves from absorption in the
object as independently given; we are, instead, to look at the
object as 'intended' by the act, as an element in the act, as the
intentional end or telos of the act. The second argument carries
even more weight. The very objectivity of the object cannot be
defined except in terms of the objectifying act. How this is so
we shall see later on. The two lines of argument, however, point
in the same direction, i.e., towards establishing the supremacy
of the objectivating striving over the object, as ready-made.
Recognition of this primacy gives to Husserl's thought its
basic idealistic character.' But the distinctive nature of this
idealism will be clearer if we enquire into the exact nature of
this objectivating activity. The most important question in this
context is this: What is precisely meant when it is said that the
object is 'constituted' by the act? Further, what does Husserl
mean by 'objectivity', and how is it that this objectivity is not
definable, as we saw, except in terms of the objectivating act?

4 H. Kuhn, "Concept of Horizon" in Philosophical Essays in Memory of Edmund


Husserl, edited by M. Farber, University of Buffalo Publication.
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY I4I

III

To come to the second question first, we are to see in what


'objectivity' consists and what we mean by 'object'. Husserl's
reply is developed on the following lines: There is one consider-
ation which is clear to us prima facie. The same object can be the
object of perception, of judgment, of inference, of love or hate,
of desire or aversion. Thus there is a certain independence from
the varying modes of apprehension. This independence, how-
ever, is but the negative aspect of a positive character. Posi-
tively, an object possesses an identity amidst all variations of
acts. The same object appears as the telos of various intentional
processes. It is around this problem of identity that our consider-
ation of objectivity shall mainly center.
Where is the place of this identity in the general noesis-noema
doctrine? The noetic-noematic structure of experience implies a
strict parallelism between the two aspects; every noesis has its
corresponding noema. Thus for every conscious (or noetic) act,
there is a different noema. What then is the meaning of saying
that an identical object persists as the objective of varying acts?
Every noema has a nucleus, its own 'noematic' meaning,
through which it gets related to the object. Husser!, in this con-
text, introduces elaborate considerations regarding the problems
of 'meaning' and 'content'. This is not the place to go into those
details. We need only to remember that this noematic nucleus
is the central core around which gather all subsequent phases.
Now the different noemata of the varying acts, no doubt, possess
different nuclei. But, Husserl goes on to tell us, these various
nuclei "close up together in an identical unity, a unity in which
the 'something', the determinable which lies concealed in every
nucleus, is grasped as self-identical."5
How do we grasp this self-identity? Is it really given in intui-
tion? If all that we get is a series of perspective variations,
where do we get the identical something, the point of unification?
Husser! suggests that this problem of identity needs a solution
on a transcendental level. The problem of identity is, according to
Husserl, a problem of Reason; and its solution demands phe-
5 Ideas, p. 338.
142 'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY

nomenological clarification. Hussed does not leave the solution


where Hume left it by denying any identity; nor does he assert
such identity naively in the manner of the realist. The pheno-
menological solution proceeds on lines akin to Kant's. "How in
the spirit of phenomenological science we are to describe, 'noe-
tically and noematically', all the connections of consciousness
which render necessary a plain object (and this in common speech
always means a real object) precisely in its character as real"?6
Phenomenological clarification considers objects not as in-
dependently given but as intentional correlates of conscious
experiences. Now, intentional objects may be either themselves
other experiences, or what are called 'things'. Hence the dis-
tinction, which is discussed by Hussed in great detail, between
immanent and transcendent perception. In the former case,
the object is an immanent constituent of the process; in the latter,
the object and the conscious process do not constitute a natural
unity. The distinction, however, which is relevant to our present
purpose is this: in the case of immanent perception the object is
given completely and absolutely, while the object of transcen-
dent perception is capable only of perspective variations.
In the latter case, we can have only a series of variable per-
spectives; and this series is infinitely determinable in infinitely
various directions. But however far we extend these series in all
possible directions, a margin of indeterminacy is still left over,
so that in this case the object can never be completely given.
The problem of identity is therefore to be solved in two dif-
ferent ways in the cases of the two different types of intentional
experience. In both cases, we are to keep in mind, Hussed
makes use of the same criterion. True Being, in phenomenolo-
gical analysis, reduces itself to adequate givenness (Ideas, pp.
365f). This adequacy, however, is to be differently conceived in
the two different cases. In the case of immanent experiences,
adequacy means complete givenness of the object itself, with
all its possible and actual determinations; there is not the least
shade of indeterminacy left over. The object is just that which is
grasped.
In the case of transcendent perception, the solution is truly a
problem of Reason. Since we have never here complete given-
6 Ideas, p. 348.
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY 143

ness of the object, in its full determination, since a margin of


determinable indeterminacy is always left over, a principle of
Reason is invoked to make up for the deficiency.
Husserl's solution comes up to this. Though complete givenness
of the object itself is not possible in this case, Reason neverthe-
less prescribes an idea of complete givenness as an a priori deter-
mination of the continuum of appearances. 7 What is given is
this Idea. The Idea of an infinity is not itself an infinity; "the
insight that this infinity is intrinsically incapable of being given
does not exclude but rather demands the transparent givenness
of the Idea of this infinity".8 Thus though the objective factor
itself is not given, what is given is an Idea of this objective
factor; and Idea serves as "an a priori rule for the well-ordered
infinities of inadequate experiences".9
We are to recall here how closely the phenomenological point
of view resembles Kant's account of our apprehension of object.
Husserl is explicit in his indebtedness to Kant on this point.
The reference to Idea is a Kantian heritage. The term 'Idea' is
said to have been used in the Kantian sense.1 0 "I understand by
Idea," says Kant, 11 "a necessary concept of reason to which no
corresponding object can be given in sense-experience." The
completed series of perspective variations, in other words, the
full determination of the objectivity cannot be given to intuition.
Hence the Idea of such a completed series serves as an a priori
rule to make the incomplete givenness adequate.
But it is clear that the status Kant accords to the Ideas of
Reason is different from what Husserl confers on them. Kant
distinguishes between the concepts of understanding and the
Ideas of Reason. The former constitute objective knowledge;
the latter yield only pseudo-knowledge. Knowledge requires a
basis in sense-intuition. Ideas refer to transcendent things which
can never be given in intuition. Objective knowledge starts from
representations and is completed by synthesis of a concept.

? Ideas, p. 366.
8 Ideas, p. 367.
9 Ideas, p. 367.
10 Ideas, p. 366.
11 1. Kant, Kritik der reinen Vernun/t, KW III, edited by Cassirer, p. 264: "Ich
verstehe unter der Idee einen notwendigen Vernunftbegriff, dem kein kongruierender
Gegenstand in den Sinnen gegeben werden kann".
144 'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY

In the Deduction of the first edition, Kant gives an analysis of


our awareness of object. Apprehension of the manifold is fol-
lowed by reproduction in imagination and is consummated in
Recognition through concept. It is easy to see that the purpose
this concept of an object serves here in Kant's Deduction is
exactly what the Idea yields in Husserl's philosophy. Since for
Husserl sense is not the only source of intuition, and conse-
quently since he considers rational explication as equally in-
tuitive, he does not distinguish between the concepts and Ideas
in the Kantian manner. This is also corroborated by Husserl's
recognition that Kant is treading on phenomenological ground in
the first edition Deduction, though he is also accused of having
retraced his steps in the second edition Deduction.1 2
We have seen that corresponding to the Kantian Idea of
Reason, no object can be given in sense-experience. Husserl's
use of it is not unwarranted; for though in apprehension of ob-
jectivity, experience does provide a series of perspective varia-
tions, yet the completed series, the object in its full determination
is never given. An Idea of this completed series is therefore an
Idea, whose object cannot be given. The Idea, therefore, serves a
purpose, which may be, not illegitimately, interpreted as not
being un-Kantian in spirit. And yet there is no question of this
Idea's yielding here a dialectical illusion, for a specific sense-
core is already there. Further like the Kantian Idea, here also
the Idea of the object does prescribe an a priori rule for the com-
pletion of a series, which is infinite, and which is, in principle,
indeterminable by any empirical experimentation.

IV

We have, thus far, considered in some detail the second of


the two questions that we placed before ourselves. Now we may
revert to the first question. Husserl speaks of the 'constitution'
of objectivity, of the object-world, as being 'constituted' by its
subjective 'sources'. What could he mean by these phrases,
'constitution' and 'sources'? They are certainly important,

12 Ideas, p. I66.
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY 145

inasmuch as they concern the metaphysical status of the object-


world.
The reference to 'constitution' is not to be interpreted onto-
logically; that is, it contains no implication of 'creation'. The
object is not 'created' by some subjective process; nor do cer-
tain subjective processes, by working together, result in an ob-
jective formation. The idea of such 'creation' does not strictly
belong on the philosophical level. Kantian analysis of knowledge
is often wrongly interpreted by people unaccustomed to philoso-
phic reflection as an account of the various stages by which
subjective processes conjoin to terminate in a unity, which we
call the object. That is fundamentally a non-philosophical ap-
proach. Philosophical reflection proceeds on a level which is
only indirectly concerned with temporal sequence of processes.
And 'creation' is a matter of temporal attainment through pro-
cess. Philosophic analysis is a reflective analysis. Both Kant and
Husserl distinguish between true philosophic reflection and other
pseudo-philosophic reflections. The latter include psycholo-
gizings and historicizings. Both name true philosophic reflection
'transcendental reflection'; Husserl calls it also 'phenomenolo-
gical reflection'. It is from this higher point of view that we are
to consider the present question.
Phenomenological analysis is also functional analysis. The
point of view of 'function', Husserl tells us, is the essence of
the phenomenological standpoint. 13 Indeed, all epistemological
analysis is functional. It is the correct philosophical standpoint.
Kantian analysis is more so. Functional analysis of object is not
an analysis on the same objective level. Psychological or his-
torical analysis moves on the same objective plane, and hence
does not completely satisfy our reason, which aims at grasping
completely the principle of objectivity. Functional analysis is
regressive in the sense that it leads us back to more fundamental
principles; it is also transcendental in the sense that it takes us
back to principles which account for the very possibility
of the datum.
Now, 'constitution' of objectivity is a functional phrase and
is to be understood accordingly. In Husserl's view, the greatest
problems are the functional problems or those of the 'constituting
13 Ideas, pp. 230f.
146 'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY

of the objective field of consciousness'. "By 'function'," Kant


says, "I mean the unity of the act of bringing various represen-
tations under one common representation."14 The Kantian
account of objectivity is an account of how objectivity resolves
itself into several functions of the transcendental subject. The
categories are such functions, inasmuch as they can be under-
stood as the modes in which the transcendental subject 'acts'.
To call these the modes of 'activity' of the transcendental subject
does not carry the sense of any temporal process in which the
transcendental subject issues forth. They are called 'modes of
activity', 'modes of synthesis', etc., because they do not stand
for static faculties. Kant is not speaking about psychological
faculties, but is undertaking a transcendental reflection, the
next outcome of which is a perception that the principle of ob-
jectivity is no ultimate principle, which can be conceived as
adequate to itself, but is resolvable into certain basic functions
of a higher unity. Husserl's point of view is similar, though he
does not sketch it in such detail as we find in Kant. Objectivity
is 'constituted' within the field of pure consciousness, inasmuch
as the former represents a 'function' of the latter.
It is interesting to see that this functional point of view is
not really foreign to the doctrine of intentionality which forms
the basis of Husserl's philosophy. The doctrine of intentionality
recognizes conscious 'acts' as being directed towards some 'ob-
jective' reference. Now, to call these 'acts' is not to introduce any
sense of temporality. Here 'act' means nothing but 'function'.
Even Meinong means by 'act' only such an element in experience
as exhibits a variability independent of the reference to a given
object. Thus, the 'act' involved in an idea is the function of pre-
senting, the 'act' involved in a judgment is the function of judg-
ing, and so on.1 5
It is true that Husserl does not go so far with Kant. He
believes Kant to have abandoned the true phenomenological
point of view in his second edition Deduction. But it is really in
the second edition Deduction that the Kantian analysis comes to
a maturer grasp of the functional point of view. The first edi-

141. Kant, Critique of Pure Reason, A 68-B 93 (Kemp Smith's translation).


15 J. N. Findlay, Meinong's Theory of Objects, Oxford, 1933. p. 25.
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY 147

tion sketch of what may be called a transcendental psychology is


a prelude to this more mature undertaking.1 6

v
In this section, we are to consider the philosophical signi-
ficance of what Husserl calls 'phenomenological reduction'.
Phenomenological reduction 'brackets' the entire natural world.
Things of the outer world and events of the psychological world -
all go together. All objective formations, aesthetic, cultural,
social- suffer the same fate. The phenomenologist is relentless in
the use of his weapon, till he attains to pure immanent experience.
The least shade of transcendence is repugnant to this motive.
We need to examine the arguments that underlie Husserl's
censure of the objective world. Are they new or are they only
old wine put into a new bottle? Novelty in philosophical argu-
ment is a rare treat; and what passes for novelty is, in most cases,
not so in fact. There are two main arguments in Husserl's sketch
of phenomenological reduction in his I deen; and with regard to
both of them, doubt arises if they are but modern versions of the
same old idealistic arguments, that had drawn Kant's attention.

I. First, the absolute being of the immanent is contrasted with


the contingency of the transcendent. Being as Consciousness is
presented to us absolutely in immanent perception, while Being
as thing is presented in transcendent perception only through
perspective variations.
It may be that the argument contains something new; but it
certainly contains a restatement of the old idealistic faith that
the inner world is knowable with far greater certainty than the
outer. This is one of the fundamental convictions of the idealist,
who has, with this faith, always looked inward for knowledge
and enlightenment. Kant, in his Refutation of Idealism, challen-
ged this postulate to the extent of making inner knowledge de-
pendent on the outer. Kant's argument is well-known. Our so-
called self-knowledge presupposes knowledge of something outer.
16 Again, I do not think the interpretation of 'constitution' given here to be ade-
quate: it may even mislead. - J. N. M.
I48 'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY

Even apart from that, it is strange how Husserl seeks to cen-


sure the natural world by establishing the contrast between im-
manent and transcendent perception. Husserl has rejected
psychologism simply because he had the courage to see that inner
events are no less natural or objective than are the things or
events of the outer world. The objects of immanent perception
are no less natural or objective.

2. The second argument follows from the first and has more
obvious affiliation with the idealistic thought of the past. Here is
a revival of the Cartesian doubt, may be, as its author claims,
with profounder implications. Phenomenological reduction starts
with something like the Cartesian doubt, but it is not exactly that.
Nor is it a negation of the starting point. Midway between thesis
and antithesis is the neutral. Phenomenological reduction 'brac-
kets' the thesis, sets it 'out of action', 'disconnects' it, makes no
use of it, suspends belief in it without yet generating disbelief.
Since we are not to pass straight off into the antithesis of the
naturalistic thesis, we are not required to doubt all Being. To
doubt the Being of anything, Husserl recognizes, is a self-con-
tradiction. I can doubt any object of awareness only in respect
of its being actually there. This amended form of doubt necessi-
tates not a negation of Being, but a suspension of the natura-
listic thesis.
The whole philosophy of doubt, of which the present is a
modern version, is based on the faith, examined above, that inner
experience alone yields certain and indubitable knowledge. 17
Kant's Refutation of Idealism is a direct challenge to this belief.
But there is quite another way in which the very fundamentals
of Kantianism throw out a challenge to this type of philosophi-
zing. According to Husserl, as according to most idealists, no
awareness of objects gives anything absolute. That may be cor-
rect. But from this to pass over to the contention that the non-

17 "To every stream of experience, and to every Ego as such, there belongs ,in prin·
ciple the possibility of securing this selfevidence: each of us bears in himself the war-
rant of his absolute existence (Daseins) as a fundamental possibility ... That which
floats before the mind may be a mere fiction; the floating itself, the fiction-producing
consciousness, is not itself imagined, and the possibility of a perceiving reflection
which lays hold on absolute existence belongs to its essence as it does to every experi-
ence. (Ideas, p. 130). Is this not Cartesian?
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY I49

existence of the objective world is always thinkable, and then to


use this latter possibility as a weapon for idealistic censure of
the outer world, involves an unjust philosophic procedure. Kant
makes out a strong case against such a procedure. Here is a con-
fusion, a very subtle confusion indeed, Kant points out, between
the logical possibility of the concept and the transcendental
possibility of things. "For to say that the non-being of a thing
does not contradict itself, is a lame appeal to a logical condition,
which though necessary to the concept, is very far from being
sufficient for real possibility. I can remove in thought every ex-
isting substance without contradicting myself, but I cannot in-
fer from this their objective contingency in existence, that is,
that their non-existence is possible",18
The logical basis of the controversy gets a clearer relief as we
probe a little deeper. Eidetic study of empirical connections
reveals, Husserl goes on, that" 'the real world', as it is called, the
correlate of our factual experience, then presents itself as a special
case of various possible worlds and non-worlds, which, on their side,
are no other than correlates of the essentially possible variations of
the idea 'empirical consciousness'."19 The actual world is the
realization of one amongst infinite possibilities. So also with
actual experiences: All actual experience refers beyond itself to
possible experiences, which themselves again point to new possi-
ble experiences, and so on ad infinitum. 20 The existence of this
world is accidental. Not only is it possible that this world were
non-existent, but Husserl even goes so far as to assert that there
might not be a world or anything of that sort. The possibilities
might never be realized owing to irreconcilable oppositions.
This whole logic of possibility is opposed to Kantianism.
For Kant this logic of possibility struggles under an illusion. The
question "whether my perceptions can belong, in their general
connection, to more than one possible experience," whether other
forms of intuition than space and time, other forms of under-
standing than the discursive categories are possible is inadmissi-
ble. Kant is not concerned with mere logical possibility, but
with real possibility, and the latter belongs to the realm of possi-

18 Critique 01 Pure Reason, A 244 = B 302 (Kemp Smith's translation).


19 Ideas, P.134.
20 Ideas, p. 135.
150 'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY

ble experience. Possible experience for Kant denotes not ab-


stract conceptual possibility, but real possibility under the uni-
versal conditions of experience; Kant is concerned with the actual
and the valid, not with the possible.
Kant has another weighty argument against this logic of
possibility, which sees in the actual an accidental formation out
of many possibilities. On such a view, something must be added
to the possible to constitute the actual, "But this alleged process
of adding to the possible I refuse to follow. For that which would
have to be added to the possible, over and above the possible,
would be impossible. "21
Thus Husserl's censure of the objective world for the purpose
of effecting the phenomenological reduction is based on quite
dubitable postulates, which cannot afford a stable foundation
for the idealism which his system in the end aims at. What
Husserl calls 'transcendental reflection' is a surer basis than the
method of 'phenomenological reduction.' 22

VI

If the natural world is 'bracketed,' if all transcendence is


rejected, how is it that phenomenology has yet to do with a
study of the 'constitution' of the objective realm? If nothing but
pure immanent experience be relevant, real and not intentional
analysis of such experience should have been the only subject
matter of phenomenological study; but that is not so, and we
have seen in some detail how phenomenology is yet directly con-
cerned with the explication of the principle of objectivity.
The reason is important and with a consideration of this point,
we will be concluding the present study. HusserI states the prin-
ciple thus: "This disconnection (i.e., the phenomenological re-
duction) has also the character of a change of indicator which
alters the value of that to which the indicator refers, but if this
change of indicator be reckoned in, that whose value it serves to
alter is thereby reinstated within the phenomenological sphere. "23
21 Critique of Pure Reason, A 231, B 284 (Kemp Smith's translation).
22 Though I think some of the above criticisms to be still valid, I do not think
that they could justify a final disposal of the Reduction. - J. N. M.
23 Ideas, p. 194.
'OBJECT' IN EDMUND HUSSERL'S PHENOMENOLOGY 15I

What is disconnected is the naturalist's thesis. Phenomenology


is opposed to the naive assertion of thing-transcendence. But such
transcendence is also a phenomenon to be explained, never to be
explained away. To undertake this responsibility, the phenomen-
ologist admits a concession, a very significant concession in any
case. From the higher point of view, from the standpoint of
transcendental reflection, the same 'content', which as the sub-
ject matter of naturalist thesis had received censure, is admissible
for consideration, though only under a 'change of signature'.
This 'change of signature' is the effect of rising to the new level of
reflection.
Phenomenological reflection, as we know, considers things as
'intended', as intentional correlates of experiences. Under the
'change of signature', therefore, all transcendents must be "re_
presented in the phenomenological spheres by the whole nexus of
corresponding meanings and positions" "within the limits of
real and possible consciousness ... In other words, all objecti-
vities must be traced to their "constituting in the subjective
sources. "24

24 Ideas, p. 345-46.
CHAPTER XIV

INDIVIDUAL FACT AND ESSENCE IN


EDMUND HUSSERL'S PHILOSOPHY*

I. The purpose of this paper is not, as its title is likely to


suggest, to reopen the discussion on a distinction familiarized by
Husserl. What I intend doing is to examine critically some of
Husserl's attempts to bridge that gulf between fact and essence
with which he himself starts. For the purpose of this discussion
I shall single out two notions of Husserl and examine whether
with the help of these two we may not succeed in connecting
facts with essences in such a way as to overcome the original
distance. These are the notions of 'individual essence'l and 'an
individual fact's own essence'.2

2. Before we take up these key notions, it is necessary, how-


ever, to prepare the ground by introducing certain well-known
ideas of Husserl that are relevant in the present context:
(a) To begin with, it may be mentioned that notwithstanding
the radical distinction between the two realms of being, facts, and
essences, Husserl also evinces an awareness of a certain parallel-
ism between them. Just as what is given in empirical perception
is an individual, so what is given in essential intuition is an essen-
ce. Husserl even cautions us against treating the above as a mere-
ly external analogy; what we have here is rather a "radikale
Gemeinsamkeit."3 In both cases the terms 'perception' or 'in-
tuition' and the correlative term 'object' are used with equal

* First published in PhilosoPhy and Phenomenological Research, XIX, I959,


222-230.
1 Ideen, Vol. I, p. 75. The references to Husserl's works are to the new Husserliana
editions, except in the case of Erjahrung und Urteil, the Logische Untersuchungen,
and the Formale und Transzendentale Logik.
2 Ibid., p. 35.
3 Ibid., p. I4.
INDIVIDUAL FACT AND ESSENCE IN HUSSERL IS3

justification. In empirical perception what is revealed is the


individual spatio-temporal fact; so is an essence revealed, given,
'bodily' presented in eidetic perception. Eidetic perception is
also an original mode of perception in the sense that it has its
own specific type of objects that are primarily given through it.
The parallelism holds good between two modes of original
givenness. Essences are given in thought, that being the only
way they are given. Husserl is thereby treating thought as a
mode of disclosure, and not as mode of construction or analysis
as other philosophers have taken it to be. In this sense thought is
intuitive, but it is not so in any other mystical sense. 'Intuition'
is nothing but a generic name for all modes of original givenness.
Thought being one such mode may be said to be intuitive.
What Husserl calls "W esensschau" is not a mystical vision of
essences, but it is nothing but thought purified. Thought, though
it discloses essences, is ordinarily tied to sense perception and
practice. Ordinary thought discloses essences, but only as they
are in facts. To attain to a pure awareness of essences is to render
thought free from this naturalistic attachment. One way of
doing this, much emphasized by Husserl, consists in varying the
factual circumstances, in taking recourse to fantasy, etc., through
which the identity of the essence as well as its indifference to
fact comes into clearer relief.
Husserl no doubt courts the risk of overemphasizing the
"radikale Gemeinsamkeit" and thereby is partly responsible for
that bugbear of "Wesensschau". Amongst the limitations of this
parallelism, let us bear in mind one: awareness of the essence is
founded on a prior perception of the individual fact while em-
pirical perception on its own part is not so founded, it being the
absolute first.4
(b) Facts presuppose essences in a twofold way. The individual
fact is accidental; it could have been otherwise. This acciden-
tality is relative to an essential necessity. It is only from the
point of view of essences that the individual fact could be judged
as accidental. Otherwise, considered in its factual interconnec-
tions the fact is determined and thoroughly necessitated. What
more this relativity means and entails Husserl does not make
clear in § 2 of the Ideen, Vol.I, where it is introduced. The argu-
4 Josef Konig, Der Begri// der Intuition, Halle, 1926, p. 303.
154 INDIVIDUAL FACT AND ESSENCE IN HUSSERL

ment as it stands seems to be circular in nature: a fact is judged


accidental only when considered from the point of view of the
universal, there being nothing in the fact taken by itself to be-
tray its accidentality. A fact by itself does not reveal its own
accidentality, which could point to that essential necessity of
which Husserl speaks. On the other hand, it is the awareness of
essential necessity which reveals the fact as accidental. 5
(c) The relation between facts and essences however is said to
be more intimate and ontologically founded. Every individual
fact has its own essence,6 just as again every essence has its own
range of possible individual facts. 7 Leaving aside the latter point
for the present, let us turn to the former. An individual fact is
not only a mere 'this-there', a 'dies-da', but also a so-and-so. 8
There are essential predicates that hold good of it. It has no
doubt accidental properties, but these accidental properties
presuppose the essential ones.
Considered from this point of view the word 'essence' is now
taken to mean "das im selbsteigenen Sein eines Individuum als
sein Was Vorfindliche."9 Every 'what' or content of an individual
is an essence. This seems to be the familiar doctrine that the
qualities or features of a thing are all universals. There is how-
ever a difference between this old doctrine and what Husserl
intends telling us here. Husserl is not telling us that what we
perceive in an individual fact are essences, as they are according
to critical realists like Santayana. If that were Husserl's opinion,
that is to say, if he held that every 'what' of an individual fact as
given in the mode in which the individual fact itself is given
were an universal or an essence, then it would be difficult to
explain why he should have followed up the above account with
the statement: "Jedes solches Was kann aber 'in Idee gesetzt'
werden."lO Explaining this last statement, Husserl continues:
"Erfahrende oder individuelle A nschauung kann in Wesenschauung
(Ideation) umgewandelt werden."ll It is only when the empirical
5 Nicolai Hartmann seeks to overcome this kind of ontology by his own modal
analysis laid down in the work Moglichkeit und Wirktichkeit, Berlin, 1949.
6 Ideen, Vol. I, p. 12.
7 Ibid., p. 21.
8 Ibid., p. 13.
9 Ibid., p. 13.
10 Ibid., p. 13
11 Ibid., p. 13.
INDIVIDA UL FACT AND ESSENCE IN HUSSERL ISS
perception is 'transformed' into eidetic perception that the con-
tent of the former is also 'transformed' into an essence. This is
far from saying that the essence itself was given as the content
of empirical perception. In Beilage V to the new edition of the
Ideen, VOl.I, Husser! notes that the essence lies implicitly in the
individual fact.1 2 This notion of implicit existence is phenomen-
ologically questionable, for the distinction between what is given
and what is really given is a metaphysical distinction pretending
to be phenomenological. It is therefore phenomenologically
unsound to say that the essences were really given in sense per-
ception, although they were not then recognized to be so given.
On the other hand, what we are phenomenologically justified in
maintaining is that every content of empirical perception can be
'transformed' into a corresponding essence. It is only then that
"Das Erschaute ist . .. das entsprechende reine Wesen oder Eidos. "13
(d) Before we take up the key notion to be examined in this
paper, let us try to understand the only definition of individual
fact which Husser! suggests in the Ideen. The definition however
has to be developed through stages for it involves certain other
distinctions.
i. Distinction is made between independent and dependent
essences. "Dependent objects are objects of whom the essential
law holds good that when they at all exist they exist only as parts
of a more comprehensive whole of a definite correlative kind."14
The color of this paper is a dependent part, for it is not only
factually part of a whole, but even essentially can exist only
as part of a whole. "A part is dependent, if it belongs not mere-
ly to a factual but also to an essentially necessary ('idealgesetz-
lichem') connectedness."15 On the other hand, if there is a part
that has only a factual membership of a whole, but essentially
need not be so, then it is independent. This distinction holds
good of essences and is an ontological distinction. 16 Dependent
contents are called abstract; independent ones are called con-
crete.1 7

12 Ibid., p. 387.
13 Ibid., p. 13.
14 Logische Untersuchungen, Vol. II, Part I, p. 240.
15 Ibid., p. 2S1.
16 Ibid., p. 248'
17 Ibid., p. 248; also compare Ideen, Vol. I, p. IS.
IS6 INDIVIDUAL FACT AND ESSENCE IN HUSSERL

ii. Another distinction is required before the 'Individuum' can


be defined. This is the distinction between formal essence and
material essence.1 8 Formal essences are those that form the subject
matter of formal logic: concepts like 'one', 'many', 'whole',
'part', etc. The highest regional essence of the sphere of formal es-
sences is 'object in general'. Formal essences are said to be empty,
while material essences are essences in the strict sense. 19
iii. HusserI now defines an 'Individuum' as a 'this-there' whose
material essence is a concrete one. 20 An individual is not a bare
'this-there', but is a so-and-so-constituted one. Its 'what' again
can be 'transformed', in eidetic intuition, into a complex material
essence. This complex essence in the case of an individual is an
independent, hence concrete, essence. One immediately sees that
this definition holds good only of things and not of determinate
qualities. The shade of green here before me as I write this is
a 'this-there' whose material essence is a 'lowest difference', an
'eidetic singularity', which however is not an independent or
concrete essence in the sense defined above, for a shade of color
must with a priori necessity be coupled with an extended surface.

3. We may now take up the notion of 'an individual fact's own


essence' and critically examine whether it is to be retained. It is
no doubt one of HusserI's well-known doctrines that every indivi-
dual fact is an 'index' for an 'essential structure'. Corresponding -
using 'correspondence' in a rather odd sense - to every individual
fact there is a definite structure of essences. This means that a
group of essences of varying orders of generality - eidetic singu-
larities, species and genera - arranged in a definite pattern of
combination and subsumption are 'in' that individual fact.
This structured whole is a complex essence.
HusserI however goes further than saying this. He speaks of
an individual fact's own essence, seemingly meaning thereby that
each individual has its own essence which a priori cannot be the
essence of another individual. Taking up suggestions of this sort,
Jean Hering 21 distinguished between the individual object (fact),
18 L. U., Vol. Ill, PP.252 ff.
19 Ideen, Vol. I, p. 27.
20 Ibid., p. 36.
21 Jean Hering, jakrbuck fUr Pkilosopkie und pkanomenologiscke Forsckung,
Vol. IV, pp. 496-543.
INDIVIDUAL FACT AND ESSENCE IN HUSSERL I57
its essence, the 'Wesenheit'; and the 'Idee'. Leaving aside for the
present Hering's distinctions between 'Wesen', 'Wesenheit', and
the 'Idee' - which by the way could be given up without setting
up serious difficulties in clarifying Husserl's texts (the single
word 'Wesen' would do!) -let us consider what Hering calls the
'individual object's own essence." Such an essence, according to
Hering, is the essence of something, "und zwar Wesen von diesem
und keinem anderen Etwas" (i.e., is the essence of this something,
not of any other thing).22 This essence, Hering continues, is the
'Sosein' ('what') of the object, taken in the complete fullness of
its constitution (" Sosein des in der ganzen Fiille seiner Konsti-
tution genommen Objektes").
Now it is true that the palm tree that I see through my win-
dow as I write this 'has' a complex essence 'corresponding' to its
own 'what'. But in what sense is it meaningfully possible to
maintain that the complex essence concerned is the essence of
this palm tree, and not of any other possible object? In order to
be able to say so, it must be possible to demonstrate that the
essence in question contains within itself the principle of indivi-
duation of facts, so that this fact and no other can possibly have
this essence as its own. If this were so, the very distinction be-
tween facts and essences would stand obliterated; the world of
facts could in that case be deductively derived from the world
of essences. This would amount to a denial that facts are facts;
they would, in fact, be individuated essences. And yet how could
essences determine without 'residue' the this? Why should it be
a priori impossible that the same complex essence be 'illustrated',
'realized', 'ingredient', etc., in that also? Besides, to maintain
that the entire individual fact could be 'transformed' into an
essence that is only this individual's and none other's would go
against phenomenological evidence: individual facts are given
in an attitude totally different from that in which essences are
given.

3.I It is in Erjahrung und Urteil 23 that Husserl attempts a


fruitful discussion of the problem of individuation as well as of the
question of 'this individual's own essence'. Here Husserl clearly
22 Ibid., p. 497.
23 Erfahrung und Urteil, pp. 429-430; compare also Beilage I.
ISB INDIVIDUAL FACT AND ESSENCE IN HUSSERL

shows his recognition of the truth that an individual fact as such


cannot be 'transformed' into an essence that would be its own
essence. 'The same' object, no doubt, can be empirically given;
it can also be given in fantasy. Cannot an essence of 'this same'
object, and of this object alone, be brought into evidence through
the process of imaginative variation? Yes, Husserl would say.
But what is 'the same object'? By 'the same object' is not
meant, Husserl cautions us, simply the object as such ("der
Gegenstand schlechthin") , for when one speaks of the simple
object one means it as actually given here and now. What alone
can identically be given in experience as well as in fantasy as
'the same object' is "a content that can be identically separated
for intuition" (" als identisch herausschaubaren Gehalt") , the
"full meaning" ("voller Sinn") that has now the character of
actuality, and now the character of "fantasized", etc. This con-
tent is "der noematischer Wesensbestand" which is identically the
same in full experience as well as in quasi-empirical fantasy.
This is the "individual essence" of the object concerned. 24
This is nothing but what is common to the intentional correlates
of the acts of perceiving, imagining, remembering, and calling up
in fantasy this very palm tree here before me.
In order however that the essence thus gotten hold of may be
this individual's own essence we must have in the essence the
correlates of those elements that impart to this object its uni-
queness. The source of this uniqueness, Husserl is well aware, is
the unique temporal character of the 'now' .25 The essence of this
individual as such must therefore contain the noematic correlate
of this 'now'. Husserl intends no doubt to show that this is
possible. For example, he says: "Die Zeitdauer ist hier aber ein
identisches Wesen, so gut wie die F arbung . .... "26 But this is
doomed to failure. Husserl himself has elsewhere emphasized
the truth that the idealities are not temporally individuated. 27
The uniqueness of the this-now refuses to be absorbed into or
derived from an essence, however articulated and structured that
essence may be. This is another aspect of the truth that 'in-

24 Ibid., pp. 460-461.


25 Ibid., p. 4 64.
26 Ibid., p. 461.
27 Formale und Transzendentale Logik, p. 139.
INDIVIDUAL FACT AND ESSENCE IN HUSSERL I59

dexical' expressions or 'ego-centric particulars' are irreducible. 28


We must, therefore, conclude that though corresponding to an
individual there is always discoverable an essence, simple or
complex, such an essence cannot in any sense be meaningfully
spoken of as this individual's own essence.

3.2. This does not preclude us however from recognizing the


phenomenological truth in speaking of 'individual essences' in
quite another sense. Not all essences, Husserl saw rightly, are as
such universals that are instantiated in facts. There are also es-
sences, the 'lowest eidetic differences,' that are not instantiated,
but are themselves identically present in those facts which they
characterize. These are what Husserl calls the 'eidetic singulari-
ties'. These, it is obvious, fail to bridge the gulf between facts
and essences in the same way in which the notion of 'this in-
dividual's own essence' pretends to do.

4. Does the notion of the 'extension' of an essence serve this


purpose any better? Just as every individual fact has its essence,
we are told, so has every essence its range of possible individuals. 29
I t is obvious that to ascribe to an essence qua essence a range of
possible individuals is not to curtail that freedom from the realm
of facts which belongs to the essences as such. An essence qua
essence cannot have a purely factual extension 30 ; if it has an
extension, that extension can only consist in possible facts.
Possible facts are not facts qua facts. Nor are they the given
facts, only considered not as facts but as instances of the essence
concerned, for the idea of 'possible facts' is logically prior to that
of given facts.
Husserl in fact distinguishes between the 'eidetic' extension
and the 'empirical' extension of an essence. 31 The eidetic exten-
sion of an essence is constituted by the specific eidetic differences
coming under it; members of such an extension are themselves
essences. To the eidetic extension of 'color', of course, belong
essences like 'red', 'blue', etc., and also in the long run eidetic

28 Compare Bar-Hillel, "Indexical Expressions", Mind, 1954.


29 [deen, Vol. I, p. 21.
30 Erjahrung und Urteil, p. 426.
31 [deen, Vol. I, p. 33.
160 INDIVIDUAL FACT AND ESSENCE IN HUSSERL

singularities, i.e., lowest specific differences like 'milk-whiteness'.


It follows that only material essences that are not themselves
eidetic singularities have such an extension.
To be contrasted with this 'eidetic' extension is the 'empirical'
extension, consisting not of actual individuals but, as we saw
above, of possible individuals, of 'einen Gesamtinbegritt von
moglichen Diesheiten'. As has been emphasized, this extension
is not empirical in the strict sense, for its definition involves the
concept of possibility which is not an empirical concept. This
notion therefore fails to annul the distance between the two
spheres.

5. Connected with the above is the distinction between'eidetic'


judgments and judgments about 'eidos'.32 Essences may be
intuitively apprehended without yet being the objects of ('Ge-
genstiinden-W oruber') such apprehension. This is so in the case
of judgments that have for their objects not the Ideas themselves,
but the totality of possible individuals constituting the 'exten-
sion' of the essence concerned. Thus, for example, when in ge-
ometry we say something about 'all triangles' or, in general, when
we form a universal judgment, we are of course having an eidetic
judgment, but are not judging directly about the eidos concerned.
The possibility however of transforming an eidetic judgment
into a judgment about eidos is always there. The eidetic judgment
- to take Husserl's example - 'A color in general is different from
a tone in general' may be transformed, through a changed at-
titude, into the judgment, 'The essence color is a different es-
sence than the essence tone'. Similarly every judgment about an
eidos may be transformed into an eidetic judgment, i.e., into a
universal judgment about the possible individuals of this essence.
In no case, however, is the existence of these individuals pre-
supposed, so that the eidetic judgments never come down to the
level of the empirical.
This distinction draws attention to the truth that universality
is a derivative characteristic of the essences. The forms 'A tri-
angle' or 'All triangles' are derivative from the original essence
'Triangle in general'. 33 A species or an essence considered in
32 Ibid., p. 5.
33 L. U., Vol. II, Part I.
INDIVIDUAL FACT AND ESSENCE IN HUSSERL I6I

itself is as much singular as universal; only the aspect of spatio-


temporal individuation is absent.
In fact one can go further and distinguish between three kinds
of eidetic knowledge: a. Knowledge of the 'eidos' itself as the
object of such knowledge as the 'Gegenstand-Woriiber'; b. Per-
ception of the 'eidos' in a single real instance. (Husserl recog-
nizes this in I deen, Vol. I, p. I8, as an eidetic perception in a mo-
dified sense. In this case an individual is given only as instance or
individuation of an essence); c. Perception of the unending
totality of possible individuals constituting the extension of the
essence. Both band c give rise to eidetic judgments. Only a
gives rise to judgments about the eidos.

6. The arguments thus far may be summed up now. Although


it may be true that every fact is an 'index' for an essential struc-
ture, yet this only means that the content of empirical perception
'may be' transformed into the content of eidetic perception.
Although an individuum may be defined, as Husserl has sugges-
ted, as a 'this-there' with a concrete material essence, two points
must be borne in mind. First, the 'with' in the above definition
should be understood in a sense totally different from that in
which one understands such a sentence as 'This is a tree with
green leaves'. The 'with' occurring in the definition does not de-
note real possession. Next, the 'this-there' itself cannot be taken
to have another essence 'corresponding' to it. The temporal
uniqueness of the real individual cannot be derived from the
essences. Husserl in an exceedingly difficult paragraph 34 dis-
tinguishes the 'this-there' from the 'last material essence', brin-
ging them both under the common title 'categories of substra-
tum'. Every 'this-there' has no doubt its essence in the sense
explained above, but the 'this-there' itself is a category of the
real world and not of the realm of essences. We fail to bridge the
gulf between facts and essences, between the real and the ideal.
The phenomenological discontinuity remains and has to be re-
cognized. This discontinuity sets a limit to all deductive meta-
physics. By 'deductive metaphysics' is here meant all those
philosophies that attempt to deduce the world of real spatio-
temporal facts from non-temporal essences.
34 Ideen, Vol. I, p. 35.
162 INDIVIDUAL FACT AND ESSENCE IN HUSSERL

6.1. Husserl has warned, as is well-known, against the re-


duction of the idealities into real factors. This reduction how-
ever is as much unwarranted as the opposite deduction; both
fail to recognize that phenomenological discontinuity which
separates the two realms. By insisting on this discontinuity we
are not indulging in another metaphysical dualism. The dualism
or the discontinuity is rather, more strictly speaking, between the
two corresponding modes of givenness, or it is a discontinuity be-
tween the corresponding 'evidences'.
Things are given in the 'naturalistic attitude'. The 'natu-
ralistic attitude' is the practical attitude. The mode of givenness
of things is the practical mode. There is no 'naturalistic' con-
templation of the world of things: that would be the height of
the intellectualistic prejudice. Perception of things is not a the-
oretical act; it is inseparably bound up with practice. The world
of things is the horizon of actual and possible practice.

6.2. Here in this connection, Husserl's later emphasis on the


pre-reflective, pre scientific 'Lebenswelt' as the implicit presuppo-
sition of reflective thought is significant. It also reminds us, as
Fritz Kaufmann has pointed out in his review of Problems
Actuels de la Phenomenologie in The Philosophical Review (1954),
of Dewey's world of 'direct experience'. Husserl rightly raises
interesting questions about the possibility of a science of the
'Lebenswelt' as well as about the a priori structure, if any, of
this 'UmweU'.35 Here is besides a point where the researches of
the later Husserl merge into those of pragmatists like Dewey, and
existentialists like Heidegger. Besides from this point of view
there is the need of a fresh understanding and evaluation of
Husserl's 'Konstitutionanalyse' which smacks of deductive
metaphysics, but which has to be exhibited in its phenomeno-
logical harmlessness, i.e., in such a way as not to conflict with
that phenomenological discontinuity which has been empha-
sized in this paper. These are programs that are far beyond the
scope of the present paper.

35 Die Krisis der europiiischen Wissenscha/ten (Husserliana, Vol. VI), pp. 28, 36.
CHAPTER XV

GILBERT RYLE'S CRITICISMS OF


THE CONCEPT OF CONSCIOUSNESS*

I. In Chapter VI sec. (2) of his The Concept of Mind, (London,


I949) Gilbert Ryle criticises a concept of consciousness which in
my opinion is one of the most important of all the different con-
cepts of consciousness to be found in the different schools of
philosphy. This is the concept of consciousness as self-intimating,
self-revealing, self-luminous etc .. By making this supposedly
distinguishing feature of consciousness his main target for cri-
ticism, Ryle has done the service of drawing attention to the
right point, even if by way of criticism. For Western philosophy
as yet had only one clear formulation of a positive distinguish-
ing feature of consciousness: this is what Brentano and Husser!
called 'intentionality'. The self-luminousness theory however,
implicit though in much of traditional western philosophy, has
never come to the forefront except perhaps in Kant's notion of
the 'I think' which according to Kant must necessarily be able to
accompany all our representations and in Samuel Alexander's
notion of 'enjoyment' as distinguished from 'contemplation'.
In Indian philosophy, on the other hand, the schools of Mimamsa
and Vedanta have made the notion of self-luminousness (sva-
yamprakasatva) of consciousness the cornerstone of their epis-
temology and metaphysics. Considering the eight pages of
Ryle's arguments from the point of view of this Indian tradition,
his arguments appear to me to be a curious mixture of insight
and misunderstanding to both of which this paper seeks to draw
attention.

2. To start with, Ryle's emphasis on the analogy of light seems


to hold good also of the Indian discussions of this theme. Just as
* First published in The Visva Bharati Journal 01 Philosophy, III, Ig66-67.
THE CONCEPT OF CONSCIOUSNESS

light reveals other objects while revealing itself, so does con-


sciousness reveal itself while revealing objects other than itself.
If another lamp were required for the revelation of light, that
would surely lead to an infinite regress. So also, it is argued, if
consciousness were not self-luminous but needed something
else in order to be revealed, that something else would need other
revealing agencies and thus an infinite number of revealing
agencies would have to be postulated. As in the case of light, so
also in the case of consciousness, the infinite regress can be stop-
ped by admitting some one member as self-luminous, as needing
no revealing agency other than itself.
While the analogy of light is so commonplace, it is far from
the truth to assert with Ryle that the myth of consciousness is
"a piece of para-optics".l The Indian philosophers were well
aware that while speaking of self-luminousness they were not
doing anything more than making use of a highly appropriate
analogy. "Consciousness", says Ryle, "was imported to play in
the mental world the part played by light in the mechanical
world".2 This division of labour gives a false impression of the
real nature of the point at stake. The analogy fails in the long
run, for light itself in order to be revealed requires the self-
luminousness of consciousness. The self-luminousness of con-
sciousness is admitted not merely to explain the way we appre-
hend the episodes enacted in the second theatre called mind, but
also to explain how anything at all could be known. The dis-
tinction between the physical world and the mental world is not
essential to this notion of consciousness. Ryle's statement of the
notion of self-luminousness owes its inaccuracies to one reason
amongst others that he introduces and considers it in the con-
text of a two-world theory which however is not an essential
background for the notion. I am not of course denying that many
of those who uphold the self-luminousness theory have also
upheld the two-world theory.

3. I will now present the self-luminousness theory in the con-


text of Indian philosophical tradition and within such limits as in
my opinion are necessary for my present purpose. Then I will
1 The Concept of Mind, p. 159.
2 Ibid., p. 159.
THE CONCEPT OF CONSCIOUSNESS 165
consider the criticisms of Ryle. This will help in clarifying the
points I propose to make against Ryle.
The traditional Indian philosophers asked a question which
may be formulated as: "How do I know that I know something?"
Supposing I know that S is p, the object of my knowledge is the
fact that S is p. But how do I know that I know that S is p?
There are two groups of answers to this question. There are some,
the Nayiyayikas, who hold that if Kl is the knowledge of an
object 0, Kl is known only by becoming the object of another
knowledge K2. K2 however need not be known, but can be known
if so desired and if circumstances permit. When however it is
known, it becomes the object of Ka. This amounts to saying that
neither Kl nor K2 nor any other knowledge is known by itself.
If Kl is the primary knowledge (seeing, hearing etc.), K 2, Ka etc.
are introspections. A knowledge is the knowledge of its own ob-
ject and not of any other object. K I , by hypothesis, is knowledge
of 0. Hence, argues the Nayiyayika, only 0, and not KI itself, is
known by K I . KI therefore is not self-intimating. No knowledge
can be so for the same reason.
To the above argument of the Nayiyayika, those who defend
the self-luminousness theory reply as follows. To suppose that KI
can be known only by K2 and K2 by Ka would lead to an infinite
regress. My awareness of my knowledge would then depend upon
the completion of an infinite series. An infinite number of know-
ledges has to be postulated. In that case, I would never come to
know that I know. We must therefore suppose that the series
must somewhere have an end. Wherever we agree to close the
series, the last member must be known without being the object
of another knowledge, i.e., it must be self-luminous. If such a
self-luminous knowledge has at all to be admitted, why not say
that KI itself is so?

4. Under such circumstances we cannot but wonder as to the real


target of one of Ryle's major arguments. In this argument, Ryle
insists on the endless regress which the theory he is attacking
involves i.e. on the "infinite numbers of onion-skins of conscious-
ness" that the theory under consideration has to postulate. But
what precisely is the theory against which this criticism holds
good? Not the self-luminousness theory, for this theory has just
166 THE CONCEPT OF CONSCIOUSNESS

the special advantage that it dispenses with the supposed endless


regress. And I have in the above paragraph drawn attention to the
fact that the Indian defenders of the self-luminous theory attack
the Nayiyayika with the same weapon with which Ryle seeks to
deal his last death-blow at the self-luminousness theory.
To decide what precisely is the theory against which Ryle's
last argument holds good, let us see how he proves the infinite
regress. In his example, Kl is an inferential knowledge of O.
K2 is the apprehension of K 1 , i.e. of the inferring. K2 to be known
must require K3 and so on ad infinitum. The theory which Ryle
here criticises differs from the Nayiyayika's theory in two re-
spects, and these differences place the theory mid-way between
the Nayiyayika's and the Vedantist's theories. Like the Nayi-
yayika, the theory criticised by Ryle holds that Kl is known by
K2 andK 2 by K 3 . But unlike the Nayiyayika, the theory that is
being criticised by Ryle takes the higher order knowings in a non-
dispositional sense. Whereas all that the Nayiyayika claims is
that K2 can be, if so desired and if circumstances permit, known
by K3, this theory holds that K2 is necessarily made the object
of K 3 • Further, this theory differs from the Nayiyayika's in the
further point that whereas for the Nayiyayika K 1 , K 2 , K3 are
succeeding cognitions, for this theory they are simultaneously
imbedded in any and every mental state. For the Indian self-
luminousness theory, on the other hand, Kl is actually known
(in a non-dispositional sense) but not by K2 but by itself. K2, K 3 ,
etc. are uncalled for.
Thus we find that the infinite regress does not vitiate the
Indian formulation of the self-luminousness theory, for the very
point about the Indian theory is the assertion that Kl is known
without being the object of another knowledge, K 2 .
It is also interesting to see that the Nayiyayika by making
the higher order knowings from K2 onwards merely possible cog-
nitions has a good argument by which he could avoid the charge
of infinite regress. For the Nayiyayika will argue that Kl alone
is required for 0 to be known, and that Kl is invariably followed
by K2 while K2 need not be known unless there is a special
desire for it to be known, that when it is known it becomes
the object of K3 and that the same is true of K 3 • There is
therefore no infinite regress.
THE CONCEPT OF CONSCIOUSNESS

The Vedantist of course would try to show that an unknown


knowledge is a self-contradiction and that if K2 is unknown
there cannot be any desire on one's part to know it.
lt is rather Ryle's version of the self-luminousness theory
(which is the Nayiyayika's theory with the differences that the
higher order knowings are not taken to be merely possible and
that Kl, K2, K a, etc. are taken not to be successive but to be
simultaneous) that lies flagrantly open to the charge Ryle urges
against it. The self-luminousness theory as formulated by the
Indian philosophers however escapes Ryle's death-blow.
Ryle himself is not totally unaware of this truth. For he sees a
that the self-luminousness theory does not imply that there are
two acts of knowing either as synchronous performances or as
somehow indissolubly welded together. If this be so, one fails to
understand why he ascribes to that theory the view that there is,
for example, an apprehension of inferring over and above the
inferring of which it is the apprehension.

6. Another argument which Ryle advances against the self-


luminousness theory is admitted by him to be merely persuasive.
Nevertheless, I will try to find out a reply. The question is, how
do we, as ordinary men, vindicate our assertions of fact. Ryle
argues that we never appeal in our vindications to 'immediate
awareness' or to any 'direct deliverance of consciousness'. We
would rather support our statements of fact by saying that we see,
hear, smell or taste so and so. Asked if one really knows some-
thing, one never replies, Ryle argues, "Oh yes, certainly I do, for
I am conscious of doing so" .
I imagine the following conversation:
Mr A. - "Look here, there is a bird's nest in the tree!"
Mr B. - "How do you know that?"
Mr A. - "Well, I see it."
Mr. B unless he has been made sophisticated by study of
philosophy would not normally ask "How do you know that you
are seeing?". But even an unsophisticated person may ask other
questions: "Are you sure you are really seeing one?", "Are you
sure you are really seeing a bird's nest, or are you seeing some-

3 Ibid., pp. 159--9.


r68 THE CONCEPT OF CONSCIOUSNESS

thing else up there ?", "Are you sure you are seeing one or are you
imagining one?" etc.
The second group of questions is of course answered, and the
answer vindicated, without appealing to the self-luminousness of
consciousness. Further, Mr. A need not - now I am answering
Ryle - vindicate his assertion of the fact that there is a bird's
nest up there by appealing to any direct deliverance of conscious-
ness. He says, and need say even on the self-luminousness theory,
that he sees one. But the sophisticated question "How do you
know that you know?" has to be answered, if we are to avoid an
infinite regress, by admitting at some point of the answer that
something is known without being an object of another knowled-
ge. The self-luminousness theory says that my knowing that I
am seeing is not another act synchronous with seeing or indis-
solubly welded with it (Ryle, to be fair, sees thjs), but that my
knowing 0 and my knowing that I am knowing 0 are one and the
same act. The Vedantist is well aware that when the self-lu-
minous consciousness is said to be 'known' by itself, this word
'known' is used in a pickwickian sense and not in the same sense
in which one says of a proposition that it is known. That which
makes possible all knowledge cannot itself be an object of know-
ledge: this was also the point Kant wanted to make out against
the rationalist psychology of his time.

7. Of all Ryle's arguments, it remains to consider the one


appealing to facts like self-delusion. Although the self-lumi-
nousness theory does not claim to provide, as has been pointed
out, an answer to the question "Are you really seeing one or
are you imagining you are seeing one?", yet - it may be argued-
does not the very possibility of this question cut at the roots of
the theory? For, how, if consciousness be self-intimating, can I
be at all deceived about my mental states? How can I at all
entertain a doubt about the mental state I am experiencing?
How can I ask myself "Am I really seeing or am I imagining I am
seeing?"? "If consciousness was what it is described as being",
writes Ryle,4 "it would be logically impossible for such failures
and mistakes in recognition to take place" .
Replying, let me at once say what I have said before, namely
4 Ibid., p. 162.
THE CONCEPT OF CONSCIOUSNESS 169
that self-luminousness is not a property of mental states (ad-
mitting, provisionally though, the two-world theory) just as it
is not a property of non-mental objects. It is the property of
awareness, be it the awareness of a physical object or of a mental
state, be it awareness of the object seen or of my seeing of it.
Awareness itself is not a mental state.
If mental states and non-mental objects are so far on a par,
cases of mistaking one physical object for another (e.g., a rope for
a snake) and cases of mistaking one mental state for another
(e.g. an imagining for a seeing) have equal relevance for the issue
at stake. But none of these two groups of cases disproves the self-
luminousness theory, for this theory accounts not for the truth
of any knowledge but for the fact that I am not only aware of the
object known but at the same time am aware of my own aware-
ness, or using Samuel Alexander's phraseology, for the fact that
I 'enjoy' my own awareness without making the latter the object
of another awareness. False awareness is also an 'awareness of ... '
and this awareness is as much 'enjoyed' as correct apprehension
is. Only, the theory would add, nothing - neither physical nor
mental- can be known unless this self-luminous awareness were
there, that is to say, if consciousness were not self-luminous.
For the same reason, unconscious mental states (to which
Freud has drawn attention) present as little difficulty as un-
known physical objects.
I should add at the end that the purpose of this paper is not
to maintain that the self-luminousness theory as stated here is
philosophically invulnerable. My purpose has been to show that
Ryle's criticisms do not succeed in exploding this notion.
CHAPTER XVI

ON G. E. MOORE'S DEFENCE OF
COMMON SENSE*

I. In this paper I shall critically discuss G. E. Moore's Defence


of Common Sense! with the purpose of showing that Moore's
idea of defending common sense was entirely mistaken. This
mistake is based, as I shall try to show, firstly on a mistaken
notion about the nature of common sense and, secondly, on a
mistaken notion about the relation between common sense
beliefs and philosophy. Nothing that I say in this paper regarding
the value and the validity of Moore's Defence of Common Sense
should be taken as reflecting my opinion about the other aspects
of Moore's philosophy.

2. Let me state at the outset that the word 'common sense' is


used by Moore in a somewhat unusual sense. Ryle has rightly
drawn attention to the fact that 'common sense' is ordinarily
used to stand for "a particular kind and degree of untutored
judiciousness in coping with slightly out of the way, practical
contingencies".2 To make common sense therefore a partisan in
philosophical dispute would appear monstrous to common sense
itself. For, 'common sense', in its ordinary unphilosophical use,
does not stand for a set of beliefs or a set of propositions like
those listed by Moore. Men who possess common sense of course
do believe in many or even all of these propositions; but so also
do men who, we say, lack common sense. Moore, therefore, when
he takes upon himself, as one of his philosophical jobs, the task
of defending common sense in the sense of defending a set of

* First published in the Indian Journal of Philosophy, II, 1960, NO.4, 1-10.
1 J.
H. Muirhead (ed.), Contemporary British Philosophy (Second Series), London,
1925, pp. 193-223.
2 G. Ryle, Dilemmas, Cambridge, 1954, p. 3.
ON G. E. MOORE'S DEFENCE OF COMMON SENSE I7I

propositions in which he along with most of us believes, must be


using the word 'common sense' in a very uncommon way.

2.I. In reply to the above, one may quite well agree to dis-
pense with the word 'common sense' while appealing instead to
the beliefs that are common to the plain men of Europe and
North America! This is in fact what Thomas Reid means when
in his Reflections on the Common Theory of Ideas, he says, re-
ferring to Berkeley's philosophy:
If a plain man, uninstructed in Philosophy, has faith to
receive these mysteries, how great must be his astonish-
ment! ... After his mind is somewhat composed it will
be natural for him to ask his philosophical instructor:
Pray, Sir, are there then no substantial and permanent
beings called the sun and moon, which continue to exist
whether we think of them or not? ... 3
It is the beliefs of 'the plain man, uninstructed in Philosophy'
which, it might be suggested, Moore was defending. This how-
ever cannot be Moore's intention. For, firstly, Moore certainly
does not believe in all that the plain man, uninstructed in phi-
losophy, believes to be true. And he seeks to defend the truth
only of some of the beliefs of the plain man. Secondly, the beliefs
of the plain man, depending largely upon his religious and cul-
tural background, may - and, in fact, do - include a large num-
ber of beliefs which Moore, I presume, would not undertake to
defend. 4

3. Let me therefore pursue this point a little further with a


view to bringing out the nature of the beliefs Moore sought to
defend. It seems to me that Moore's interest consists as much in
defending his belief in certain propositions as in proving certain
beliefs of philosophers to be false. Many philosophers have be-
lieved in such propositions as 'Time is unreal', 'There are no
other selves', 'Matter is unreal', etc .. Moore's purpose is to show
that these propositions contradict the beliefs of common sense.
N ow the truth of the proposition 'Time is unreal' contradicts the

3 I am indebted to my friend Eberhard Bubser for pointing out this passage to me.
4 When the words 'common sense', 'commonsense beliefs' are used in the rest of
this paper, they are to be understood in the light of the remarks in para. 2.
I72 ON G. E. MOORE'S DEFENCE OF COMMON SENSE

truth of the proposition 'Time is real'. The belief that Matter is


unreal contradicts the belief that Matter is real. Since it often
happens that what a person is refuting throws light upon what
he is at the same time defending, we may presume that Moore
seeks to defend the beliefs 'Matter is real', 'Time is real', etc.
But of what kind are these latter beliefs? Can we attribute these
beliefs to the 'plain man, unschooled in Philosophy'? The plain
man, unless he is also an unschooled philosopher, does not
bother about such propositions as 'Time is real' or 'Matter is real'.
Defending the truth of these propositions may therefore be taken
as amounting to defending a certain philosophical theory and
not what a plain man believes in.

3.1. In reply to the above criticism, it may be suggested that


although the plain man does not say, or explicitly formulate his
belief by saying 'Matter is real' or 'Time is real', yet the other
propositions5 which he believes to be true certainly imply the
truth of the propositions 'Matter is real' and 'Time is real'. Some
examples of these other propositions which the plain man belie-
ves to be true are: 'Here is my right hand which I am raising up',
'There exists at present a human body which is my body', 'I was
born in the year I928'. Belief in the truth of these propositions
implies belief in the reality of matter, time and space and there-
fore contradicts the philosophers' beliefs in the unreality of mat-
ter, time and space. The contradiction, therefore, which Moore
detects subsists, not between the said philosophical beliefs and
certain beliefs of common sense, but between the said philosophi-
cal beliefs and certain other beliefs implied by the above-men-
tioned beliefs of common sense. That the said philosophical
beliefs are not necessarily incompatible with the truth of the
propositions in which common sense believes is admitted by
Moore; but he nevertheless reminds us that the philosophical
propositions may be understood in such a way that they con-
tradict the common sense beliefs. In other words, although
the proposition 'Matter is unreal' is not incompatible with the
proposition 'There is a human hand here', the former proposition
may be so understood (or formulated, analysed, or interpreted)
that belief in it amounts to believing that the latter proposition
5 Contemporary British Philosophy (Second Series), p. 200.
ON G. E. MOORE'S DEFENCE OF COMMON SENSE I73

is false. It seems to me that in stressing this latter possibility,


that is to say, the possibility that the said philosophical beliefs
may contradict the said beliefs of common sense, Moore is on the
wrong side.
For, first, as I have already emphasized, the said philosophical
beliefs do not directly contradict the commonsense beliefs; they
contradict only certain other beliefs which are implied by these
commonsense beliefs. On no interpretation of them, that is to
say, on no interpretation either of the philosophical beliefs or of
the commonsense beliefs would they come to a direct conflict.
But even as to this indirect conflict, two questions should be
raised: first, what is the nature of these other beliefs which are
implied by the commonsense beliefs? Secondly, can we at all
say that these other beliefs, whatever may be their nature, are
implied by the commonsense beliefs? I have already suggested
my answer to the first of these questions: these other beliefs are
themselves not beliefs of the plain man but are as much philo-
sophical beliefs as the philosophical beliefs which they contradict.
So the conflict is between two sets of philosophical beliefs.
To the second question, my answer would be in the negative.
The said commonsense beliefs do not imply the philosophical
beliefs 'Matter is real', 'Time is real', 'Space is real', etc.

3.2. It may be suggested in reply to the above that the proposi-


tion 'Matter is real' could be understood in such a manner that
it would thereby become an implicate of the commonsense be-
liefs. With this I agree, but so far as I can see, the proposition
'Matter is real' if suitably interpreted so as to become an impli-
cate of the commonsense beliefs would be, in effect, nothing
other than a restatement of these latter beliefs. In that case,
to believe in the proposition 'Matter is real' would be the same
thing as to believe in all propositions like This is a human hand',
etc., etc. If the proposition 'Matter is real' is thus nothing but
shorthand for a number of propositions in which commonsense
believes, then only it is not a philosophical proposition and is also
an implicate of commonsense beliefs; but in that case it would
not be the contradictory of the philosophical proposition
'Matter is not real'. To sum up : the philosophical proposition
'Matter is not real' contradicts the proposition 'Matter is real'
I74 ON G. E. MOORE'S DEFENCE OF COMMON SENSE

only when the latter proposition is a philosophical proposition,


but in that case the proposition 'Matter is real' is not an impli-
cate of the commonsense beliefs and therefore no contradiction
could be shown to subsist between these beliefs and the philo-
sophical proposition 'Matter is real'. On the other hand, if the
proposition 'Matter is real' is taken to be an implicate of com-
monsense beliefs, then it would not be a philosophical proposi-
tion but would be reduced to a mere shorthand for the innumer-
able propositions about physical objects in which we believe
ordinarily; but in that case it would cease to be the contradic-
tory of the philosophical belief that matter is unreal. Again, no
contradiction could be shown to subsist between the philosophi-
cal belief that matter is unreal and the commonsense beliefs.
It follows that in no case does the philosophical proposition
'Matter is unreal' contradict the commonsense proposition like
'Here is a human hand'. Moore's defence is therefore not called for.

3.3. Nor would it help to insist that what Moore is doing is to


defend, not beliefs, but ordinary use of words. For neither of the
two propositions 'Matter is real' and 'Matter is unreal' makes an
ordinary use of the word 'real'. Both make philosophical uses.
Malcolm has drawn attention to the fact the doubt which Moore
aims at dispelling by asserting 'I know for certain this is a human
hand' is a philosophical doubt and that his use of 'know' in this
context is not an ordinary use. 6 What I want to insist on is that
in his zeal to defend common sense Moore has ended up by dis-
torting it.
From what has been said before, it would follow that there is no
question of the same proposition being true from the common-
sense point of view and false from the philosophical point of view.

3+ I cannot imagine common sense saying 'Time is real', for


the assertion 'Time is real' is uttered only when the doubt 'Is
time real?' is dispelled. And I wonder if common sense is ever
haunted with this last doubt. Common sense, on the other hand,
may be haunted, given suitable circumstances, by the doubt
'Is this a real tree?' and this doubt is dispelled by the assertion
'This is a real tree' or by the assertion 'This is only a painted one.'
6 In Philosophical Review, 1949.
ON G. E. MOORE'S DEFENCE OF COMMON SENSE I75
4. This brings us to a certain paradox which belongs to what I
should like to call the existential situation of the philosopher - a
paradox out of which Moore develops a set of arguments against
the philosopher who denies the reality of matter and of time. No
philosopher, Moore seems to be insisting, has ever been able to
hold such views consistently. "One way in which they have be-
trayed this inconsistency is by alluding to the existence of other
philosophers. Another way is by alluding to the existence of the
human race, and in particular by using "we" in the sense in which
any philosopher who asserts "we sometimes believe propositions
that are not true" is asserting ... that very many other human
beings. .. have had bodies and lived upon the earth ... "7
The philosopher in the course of his philosophical activity assu-
mes the reality of those very objects that his philosophy regards as
unreal. This is indeed a paradox. But what does it point to? Does
it show that a philosophical transcendence is not possible? I
would rather say that the paradox would not exist if philosophi-
cal transcendence of common sense were not possible. Just be-
cause there is this paradox, philosophical transcendence is a fact.
I am aware that there are philosophers who would deny the para-
dox and a paradox is denied the moment you resolve it, - either
following Moore or following the Absolutist! I for one do not
believe that resolution of such paradoxes is either necessary or
possible. They are there; they have to be recognized as such. In
fact, they provide the tragic ethos that characterizes the exis-
tence of the philosopher. They neither call for a rejection of the
philosophic pursuit in favour of the certainty of common sense
nor do they call for a denial of the common sense beliefs in
favour of the philosophic truths.

4.I. Imaybetold that though common sense has its limitations


yet the limitations themselves belong to common sense so that
the philosopher could love common sense as Cowper loved Eng-
land in spite of all her faults. 8 The point that concerns us here
is whether it is possible to transcend common sense. Making use
of the analogy of the poet's England, let me suggest that al-

7 Contemporary British Philosophy (Second Series), p. 203.


8 I am indebted to my friend and colleague Prof. K. K. Banerjee of Jadavpur
University for suggesting this metaphor.
176 ON G. E. MOORE'S DEFENCE OF COMMON SENSE

though one loves one's homeland in spite of all her faults, one
can transcend that love to reach a wider love of humanity. What
however is more important is that one understands one's love
only when one can contemplate it from a distance. What I
wish to suggest ;s this: the true character of common sense belief
as a belief cannot be revealed to me unless I can look at it from
outside, as a neutral spectator - that is to say, unless in so far as
I philosophize, I suspend my beliefs, neutralize them as it were,
do not live in them, do not let myself to be merged in them, and
so on. It is true that I have thereby to experience an existential
paradox to which I have just now referred.
There are certain limitations that fall within that whose limi-
tations they are: they fall within it in the sense that you can
grasp them while confining yourself to the same level of exper-
ience. But there are certain other limitations - which are really
fundamental - which you can grasp only when there is a radical
transcendence of the level of experience concerned. The inac-
curacies, inadequacies, hesitations, ambiguities and the vague-
ness of common sense belong to the first group of limitations. I
would even say that when science corrects common sense, it
improves upon limitations of the first kind. Science does not
therefore bring about a radical reformulation of the notions of
common sense. A radical transcendence, and therefore a funda-
mental understanding, of common sense requires what has been
characterized as a neutralization of common sense beliefs or
what Husserl would have called a 'phenomenological bracketing'
of them. Moore - should I say even at the risk of appearing
audacious? - has not given us a genuine philosophy of common
sense, for he has not gone into the roots of common sense be-
liefs. He has not exhibited these beliefs as beliefs. He has not
been able to do this, for he wanted to defend common sense.
Thereby he played the role of a partisan and not of an enquirer.

5. In the light of the above remarks on Moore's defence of


common sense, it will be now of interest to pay some attention
to the very puzzling proof of an external world which Moore has
advanced in his British Academy Lecture. After going through
Moore's proof, one is left wondering what precisely could have
led Moore to advance such a proof. Which philosophers he could
ON G. E. MOORE'S DEFENCE OF COMMON SENSE I77
have had in mind, that is to say, to which philosophers was he
attributing the view that there is no external world? 'Berkeley!'
of course, is the first choice. But as we know, Berkeley certainly
did not mean to deny the existence of the external world in the
sense in which Moore proves it. Nor was Moore, in trying to give
a proof, refuting what Kant called 'problematic idealism', that
is to say, the position that we never know for certain that there
is an external world. Moore, of course, has something to say
against 'problematic idealism'; it is in this context perhaps that
he draws the distinction between knowing something and pro-
ving something. His main proof however is concerned with
showing not that we know for certain but that there is an exter-
nal world. Presumably, he thought that Berkeley had denied the
external world. Whatever that may be, let us go into his proof.
By 'external things', he means 'things outside of our minds'.
Things to be met with in space are of course things outside of our
minds, though not all things outside of our minds (e.g., pains or
visual images of animals) are to be met with in space. Now if Moore
can prove that there are two things to be met with in space, it
would follow that there are two things outside of our minds.
"By holding up my two hands, and saying, as I make a certain
gesture with the right hand, "Here is one hand", and adding, as
I make a certain gesture with the left, "and here is another",
Moore claims to have given a most rigorous demonstrative
proof of the existence of the external world.
Moore himself has given expression to his apprehension that
what he proves may be accepted as true but may at the same
time be declared as unimportant. But quite apart from that,
does the proof succeed?
I am aware that in questioning Moore's premiss I am in com-
pany with many of Moore's critics. Malcolm 9 has, for example,
questioned if Moore is justified in saying 'I know here is a human
hand'. I, however, wish to urge a quite different point. I would
say that Moore cannot, on the basis of his theory of perception,
say with certainty 'this is a human hand'.
Just consider some features of his own analysis. The proposi-
tion 'I am now perceiving a human hand' is analysed into (Moore

9 In Philosophical Review, 1949.


178 ON G. E. MOORE'S DEFENCE OF COMMON SENSE

says: is a deduction from 10) two further propositions: 'I am


perceiving this' and 'This is a human hand'. He is sure about 'I
am perceiving this', but what exactly is known thereby he is
not sure of. The analysis which is accepted is that the principal
subject of the proposition 'I am now perceiving this' is a sense-
datum. And he is besides sure that this sense-datum is not a
hand. He finds reasons to doubt - although he himself does not
doubt - that this (i.e., to say, the sense-datum) is a part of the
surface of the hand,11 How can he under such circumstances be
sure of the proposition 'This is a human hand?' The distinction
between 'knowing a proposition to be true' and 'not knowing the
correct analysis of the proposition' does not help us here.

5.1. In his essay on Hume's philosophy, written much earlier,


Moore admits that it is quite impossible for anyone to prove, as
against the sceptic, that one knows any external fact. 'I can
only prove that 1 do by assuming that in some particular instance,
I actually do one',12 On this, Stebbing remarks: 'The notion that
we may have a reason, though not a logically conclusive reason
for certain statements concerning direct observation, is, I be-
believe, one of Moore's important contributions to philosophy.'13
Hume showed that demonstrative knowledge of matters of fact
is not possible; Moore, I would suppose Stebbing to mean, shows
that even our non-demonstrative knowledge of the external
world has its own certainty which should not be underestimated
just because it is other than demonstrative certainty. This, if it
were Moore's contention, would have been ranked as one of his
valuable insights. When we turn however to Moore's reply, we
are disappointed: Moore rejects this suggestion,14 1 can appre-
ciate why Moore should object to the use of the word 'probable'
in connexion with our knowledge of the external world. Steb-
bing in fact is aware of the misleading associations of this word.
Let me call the type of knowledge Stebbing had in mind 'non-
demonstrable certainty.'
Instead of emphasizing this 'non-demonstrable certainty' of
10 Contemporary British Philosophy (second series), p. 217.
11 Ibid., p. 218.
12 Philosophical Studies, p. 160.
13 In P. A. Schilpp (ed.), The Philosophy of G. E. Moore, p. 524.
14 Ibid., p. 677.
ON G. E. MOORE'S DEFENCE OF COMMON SENSE I79
the external world and instead of exhibiting the phenomenolo-
gical nature and roots of that certainty, Moore proceeds to offer
a rigorous demonstrative proof. And no wonder that he should
fail. The external world is neither in need of nor is capable of a
logical proof. That such a proof is necessary is what the sceptics
persuade us to believe though knowing fully well that we would
not succeed. Moore has succumbed to their persuasion and has
offered a proof that hopelessly fails.

6. To sum Up: Moore is wrong in presenting common sense as a


party in philosophical disputes. He was misled into thinking that
philosophical statements could come in conflict with common
sense beliefs. A philosophical understanding of common sense
requires a measure of transcendence of the level of common
sense; it must be added that even Moore in his distinction be-
tween the common sense beliefs and their correct analysis makes
room for transcendence. As in his defence of common sense, so
also in his proof of an external world, Moore's task is ill-conceived.
What is important for us is to realize that there is a common
source of the two errors: in his eagerness to combat the specu-
lative philosophers, he misses the proper task of a truly pheno-
menological philosophy both of common sense beliefs and of our
belief in the external world.
PART THREE
CHAPTER XVII

REFLECTIONS ON THE NY A Y A
THEORY OF A VA Y A VIPRA TY AK$A *

It is well known that the Nyaya advocates an extreme form of


direct realism and maintains that what we directly perceive are
physical objects and not some intermediate entities called var-
iously by philosophers 'ideas', 'contents' or even 'sense-data'l.
Gotama's sutras 2.1.31 - 2.1.36 and Vatsayana's commentaries
on them contain arguments which may be regarded as consti-
tuting a very effective defence of what has come to be called the
physical object language as against the sense-datum language.
Gotama's, as well as his commentator's, direct interest however
is twofold. In the first place, they are out to refute the sugges-
tion that perception is not an independent source of knowledge
but a variety of inference. In the course of this refutation, they are
led to their second point: they try to show that the object of
perception, that is to say, the physical object, is not a mere as-
semblage of parts but a true unity of some unanalysable kind.
Our task in this paper will be to bring out the relevance of these
arguments in the light of contemporary discussions of the
problem of perception.
It would at once be appreciated that the view that perception
is a kind of inference is logically connected with the view that
what we directly perceive are sense-data and not physical objects.
For on this latter view the transition from the sense-data which
alone are directly given to the physical object which we say we
* First published in the Journal of the Indian Academy of PhilosoPhy, Vol. I,
No. I,pp. 30-41.
['Avayavipratyak~a' = perception (pratyak~al of a whole (avayavil]
1 In its theory of savikalpa perception, however, Nyaya is led to grant a pecu-
liarly intermediate status to certain epistemic entities. This does not affect the basic
direct realism of the system which is maintained with the help of the theory of nir-
vikalpa perception in which the object is directly given free from all epistemic ad-
juncts.
I84 THE NYAYA THEORY OF AVAYAVIPRATYAK$A

perceive can be effected only through some kind of inference -


or if you like by some process of logical construction. In any case,
the physical object which we say we perceive is not really per-
ceived but either inferred or 'constructed'. If the modern logical
constructionist claims that the Nyaya refutation of the inferen-
tial theory does not affect him for he too abandons inference in
favour of logical construction, we may in that case implore him to
have patience; for the Nyaya has another point directly effective
against him: this is the Nyaya contention that the physical
object is a not further analysable unity. After showing how
these Nyaya arguments constitute an effective plea in favour of
the physical object language, we shall enquire into the precise
nature of the unity characterising the obj ect of perception.
The view which Gotama seeks to refute is thus stated by him
in sutra 2.I.3I: "perception should really be regarded as infe-
rence, for we apprehend directly only a part of the object. Our
knowledge of the object is based only on such partial apprehen-
sion."2 I say I perceive the yonder tree. But do I really perceive
the tree? An obj ect is properly said to be perceived by me only when
our knowledge of it is caused by the contact of some sense organ
of mine with that very object. But are my eyes in contact (with-
out taking into consideration that peculiar sense in which the
Nyaya speaks of such contacts) with the entire tree? Certainly
not with the hind part hidden away from my sight, certainly
not with the interior of the trunk, and so on. Only a part of the
tree do I see. Why then do I say that I perceive the tree? What
I actually do is that from this part which I really see I infer that
there is a tree over there, so that my perception of the tree is
really inference. The inference may be implicit or explicit or
associational or a self-conscious process of reasoning; in any case,
the same argument holds good. Further, there may be difference
of opinion as to what exactly the part is. It is quite possible
that Gotama has in his mind the Buddhist theory that the tree,
in fact any object, is nothing but a mere aggregate of atoms, and
this is how Vatsayana interprets the intention of the sutra. But it
is also possible that the parts are nothing but the various quali-
ties, the colour, the shape, the size, etc. which make up the ob-
2 This is a rather free rendering of the sidra which runs thus: "Pratyak$amanu-
mdnamekadesagrahanddupalabdhe!r," •
THE NYAYA THEORY OF AVAYAVIPRATYAK~A I8S

ject, for as is well known the dravya, according to the cepamkhya is


a collection of such qualities (gu1Jasamghata). Or in conformity
with common sense, the parts of the tree may be regarded simply
as its trunk, branches, leaves etc. In any case, the point under
consideration is, if it is only a part or some parts that I see, what
justifies me in saying that I perceive a tree over there.
Gotama and his commentator have advanced the following
objections against such a view: -

1. What I call the tree is either the mere assemblage of its


parts or it is more than such assemblage. On either alternative the
inferential theory would not be tenable. Consider the first alter-
native that the tree is nothing but a mere assemblage. Only one
part of this assemblage is directly seen. What then is inferred?
The theory can only hold that from seeing this part, say the front
part, we infer the other parts that are not seen. But neither the
front part nor the hind part taken by itself is the tree. Hence
what is inferred is not the tree but only the unseen part of the
tree. My knowledge of the tree thereby has not been proved to be
inferential. Similarly on the theory that the tree is more than the
assemblage of parts, the inferential theory has no better chance of
success. For in order to be able to infer B from A it is necessary
either that B must be analytically contained within A or that B
must always have been observed together with A. In the present
case, only the second possibility is open. And yet since on this
theory the whole is never perceived along with the part, the
whole could not possibly be inferred from the part that is given.
Hence the tree which is a whole that is more than the mere sum
of its parts could not possibly be inferred on this theory.

2. We do not directly apprehend the mere part (Naca eka-


deSopalabdhi!t), for the whole, the tree itself, is also given through
that part. This in fact is the central argument of the Naiyayika.
This thesis has two parts each of which requires separate treat-
ment. In the first place, it has to be shown that the tree - in fact
any physical object - is more than a mere assemblage of its parts,
that it is something new over and above the parts. But next it
has also to be shown that perception of this whole does not re-
quire perception of all its parts, so that the whole may be per-
I86 THE NYAYA THEORY OF AVAYAVIPRATYAK!?A

fectly legitimately said to be perceived even if one is not directly


perceiving all its parts. The view criticised here (that the tree is
not being directly perceived for only a few of its parts are visible)
is based on the wrong assumption that we must perceive all the
parts of a whole in order to perceive the whole itself. The Nyaya
rejects this assumption. In fact the perception of any of its parts
may suffice for the perception of the whole.

3. Nyaya therefore takes great pains to prove that there is a


whole that is other than the mere sum of the parts. Gotama de-
votes the sutras z.I.33-36 to this point but again returns to it
in a later context in the sutras 4.z.4-IZ. The commentator
Vatsayana and the later Naiyayikas develop this theme in great
detail, their principal target for attack being the Buddhist. Here
we shall only sum up the main Nyaya arguments relevant from
our present point of view.
(a) If the tree be regarded as a mere aggregate of atoms, then
the atoms being themselves supersensible, the tree itself would
remain unsensed. If the tree remains unsensed, so also shall its
colour, shape, size, position, etc. We could never also know it
as a tree, that is to say, we could never perceive the universal
treeness in it. Thus all the padarthas such as dravya, gU1Ja, karma,
samanya, etc. would remain unperceived. But actually we do
apprehend dravya; we also apprehend its quality, its activity,
its class, etc. 3
(b) I not only apprehend the tree over there but also apprehend
it as one object. What could be the proper referent of such a
statement as "This is one object and that is another", - asks
Vatsayana. 4 Is the referent of such a statement itself one or
many? If it be one, then that would amount to recognising that
the whole is one object, and not a mere sum of its parts. If, how-
ever, there be only parts, how can a mere plurality be referred to
as one object. The sense of unity ('This is one') and the sense of
plurality ('These are many') cannot refer to the same object. 5
The former refers to one object, not to a mere aggregate. The
latter to an aggregate, but not to a unity. "But" it may be asked,

3 Nyayasutra 2.1.34.
4 Commentary on NyayasUtra 2.1.35.
5 Uddyotakara, Vartika on Nyayasurta 2.I.35.
THE NYAYA THEORY OF AVAYAVIPRATYAK!?A 187
"do we not refer to a wood or to an army as one object, although
the wood or the army really is a mere aggregate of many dif-
ferent things? Why then would it not be possible in a similar
manner to refer to a tree as one object though in reality it is a
mere aggregate ?"6 To this, the Nyaya replies in the following
manner. It is true that we mistake from a distance, or on account
of other dO$as, a mere aggregate or a plurality for a unity, so that
instead of saying 'These are many', we say 'This is one'. Such a
sense of unity is no doubt erroneous. But such an erroneous
sense of unity is possible, only if there are other cases, where our
sense of unity is right.7 But if, as the Buddhist contends, all sense
of unity is erroneous, then even those cases where, all are agreed,
a plurality is mistaken as a unity would remain unexplained. 8
Vatsayana anticipates the modern phenomenalists' view that we
do have a right sense of unity, not of course in the case of our
apprehension of a tree, but certainly in the case of our apprehen-
sion of a sense-datum, and has two replies to offer. First, since we
have one instance of a sense of unity being right (e.g. 'This is one
sound') and another instance of unity being wrong, i.e. mis-
placed(e.g. 'This wood is one object'), some satisfactory reason
needs to be advanced, before assimilating the case under con-
sideration ('This is one tree') to the one rather than to the other:
the phenomenalist has given no satisfactory reason for assim-
ilating it to the latter type. Secondly, the phenomenalist has
no satisfactory reason for regarding what we call sound as one
entity. Leaving apart Vacaspati's reminder that some of the
Vaibha~ikas in fact did regard even a sound as an aggregate of
atoms, we have to remember the extreme difficulty - perhaps
impossibility - of identifying a sense-datum as one sense-datum.
Even if we are ever rightly able so to identify, the proper pro-
cedure would be to assimilate 'This is one sound' type of state-
ment to 'This is one tree' - type rather than go the other way
about. Vatsayana however has his own special reason for regard-
ing 'This is one tree' as a right application of the concept of
unity and not as a case of error. For we also say, 'This is a single,
6 Nyiiyasutra 2.I.36.
7 Vatsayana, "Atasminstaditi pratyayasya pradhiiniipek§itatviit pradhaniisidhi[r,"
commentary on Nyayasutra 2.I.36.
8 Uddyotakara, "Mithyiipratyaya apyete na bhavanti pradhiiniibhiiviit" Viirtika
on Nyiiyasutra 2.I.36.
ISS THE NYAYA THEORY OF AVAYAVIPRATYAK;>A

big, and banyan tree'; in such a statement as this, bigness, one-


ness, and the specific and the generic characters of the tree are
apprehended as belonging to the same locus (samiiniidhikara1Ja)
- which proves that that which is one is also a tree and is big etc.
Further, when we say that 'the two (A and B) are in contact',
the contact is apprehended as belonging to the same locus as
two-ness: in such a case it would amount to distorting the im-
plication of such statement if we say, instead, that it is really the
many parts which are in contact.
(c) Now if there is such a whole that is other than the mere
assemblage of parts, how is this whole related to the parts? The
Buddhist might argue that since no satisfactory relation is con-
ceivable why not abandon the hypothesis of such a distinct
whole?9 Either the parts are in the whole or the whole is in the
parts, no third alternative being conceivable. The part cannot be
in the whole for anyone part cannot reside in the entire whole,
the part and the whole having different extensions. Nor could it be
said that a part resides in a certain region (ekadda) of the whole,
for that region would itself be a part of the whole and there are
no regions other than the parts. Exactly similar arguments can
be used to prove that the whole cannot reside in the parts: the
entire whole cannot reside in anyone part, the latter being smal-
ler in extension than the former and since the whole does not
have regions other than the parts, it cannot also be said that it is
one part in one region and another part in another region. Thus
no relation between the whole and the parts being conceivable, it
would be safer to conclude that the so-called whole is a mere
assemblage of parts.
To this Buddhist argument, Nyaya replies as follows. Nyaya,
of course, does not accept the position that the parts reside in the
whole. This position, seemingly acceptable to common sense,
owes its obviousness to the unreflective identification of the
whole with the sum of the parts. Once, however, the distinction
and the peculiar unity of a whole are admitted, the parts cannot
be accommodated within it; for the spaces - as the critic has
pointed out - to which the parts may be allotted are themselves
parts. The same difficulty, however, does not really vitiate the
position that the whole resides in the parts. The critic's question,
9 Gautama 4.2.6.
THE NYAYA THEORY OF AVAYAVIPRATYAK;>A r89

"Does the whole in its entirety reside in a part or does it do so


only partially?" is ruled out ab initio, for the words 'entire' and
'partial' - as Vatsayana points out -- 10 have no application to a
thing that is one. The word 'entire' means 'all of many things'
and the word 'partial' means 'some amongst the many'.ll Hence
both apply only to a plurality. The whole, according to Nyaya,
however, is one and not a mere aggregate; hence there is no
question of treating it either in its entirety, or partially.l2 The
whole, therefore, is present in each of the parts as well as in their
totality. 13

4. It has now to be shown that perception of all the parts is not


a necessary condition of the perception of the whole residing in
those parts. If this could be shown, it would follow that the tree
may be legitimately said to be perceived even if some of the parts
remain unperceived. What is necessary is that some of the parts
should be perceived. The argument is simple. Since, as has just
been shown, the whole itself, as a distinct entity, is present in
each of its parts, the apprehension of a part involves the appre-
hension of the whole. It cannot be said that I am perceiving only
a part or only some of the parts of the tree. True, there are some
parts that remain hidden from me. But nevertheless I do perceive
the tree through those parts that are exposed to me. Of the parts
that are many, some are perceived, and some not. But the tree is
one and not many. Being one, if it is perceived, it cannot also be
unperceived, nor can it be perceived partially. Hence, I do per-
ceive the tree though I do not perceive all its parts.
We sometimes speak of seeing more or less of a thing. From
my window I can see only one side of the school building. I go
round the building, see its other sides and say I have now seen
more of it. For Nyaya, all such statements are in a sense legiti-
10 Commentary on Nydyasutra 4.2.II: Tavimau krtsnaikadesasabdau bhedavi$ayau
naikasminnupapadyate, bheddbhdvdditi".
11 Ibid.: Krtsnamityanekasyase$abhidhdnam, ekadesa iti ndndtve kasyacidabhid-
hanam".
12 When we speak, however, of the tree in its entirety or in its parts, these words
are applied not directly to the tree itself but to the parts, only secondarily to the tree.
13 There is thus a parallelism between the way the whole resides in each part and
the way a universal resides in each instance of it. This parallel justifies the inclusion
of both relations under the same type: samavdya. (The parallel, however, fails, in
that the whole also resides in the totality of its parts; while the universal does not
reside in the totality of its instances).
I90 THE NYAYA THEORY OF AVAYAVIPRATYAK!?A

mate, but in another important sense misleading. If from the be-


ginning I know it as a house, I have perceived the whole, and
although I might go on increasing my knowledge of its parts, it is
the same house which I continue perceiving; I do not, in that
sense, come to perceive more of that entity.
What happens when I see the mere edge of a building so that I
am not in a position to say what I see, a house or a pillar, a monu-
ment or a bandstand ?14 I do see a part of whatever the whole may
be. But do I also see that whole? The Nyaya account of such a
situation is based on the following important considerations: -
Though the perception of a part involves the perception of any
one of the wholes residing in that part, it cannot be laid down as a
general rule that the perception of a part necessarily involves the
perception of a definite whole, resident in it. Let P be a part of W
and Q a part of P. Both the wholes P and Ware resident in Q.
There are also many other wholes P', P", .... resident in Q.
It is quite possible that on perceiving Q I do not perceive W but
perceive P or any other of the wholes residing in it. But I must,
whenever I perceive Q, perceive it as part of a whole, in case it is
so, that is to say, in case a whole is resident in Q.
5. Is it possible to maintain - as the modern sense-datum
theorists do - that our perception begins (no matter whether the
beginning is understood in the logical or in the psychological
sense) with noticing the bare sense-data? Do we not, to start
with, perceive a bare patch of colour, and only afterwards come
to know of the physical object that is so coloured? Nyaya, it
is now apparent, rejects this view. According to Nyaya, we never
sense a mere colour, but always perceive a coloured object. The
colour is always perceived as characterising the physical object,
and the fact of illusory appearance need not lead us to revise
this account.
The point, however, to which we want to draw attention es-
pecially is this: within the categorial structure of the Nyaya on-
tology, the perception of a mere property is impossible. The per-

14 This case is different from the case of doubt or error where I perceive a part, no
doubt, but there are also certain vitiating factors (do$as) that render the perception of
the whole impossible, or there are conditions that render certain knowledge impossible
and give rise, instead, to doubt.
THE NYAYA THEORY OF AVAYAVIPRATYAK~A I9I

ception of a quality (gu1;ta) is possible only through the via media


of a substance (dravya). A quality is contacted through a relation
of samyukta samavaya, which is a complex relation entailing a
samyoga relation with the substance in which the quality is per-
ceived as inhering. To this, one might reply that such a via media
is necessitated only by the categorial structure of the Nyaya
ontology. True; but we may remind the sense-datum theorist that
his view that what we sense is a mere colour is no less necessitated
by the way he defines both 'sensing' and 'sense-datum'.
Thus according to Nyaya, we do not pass from the part to the
whole nor do we pass from the sense-datum to the physical object.
The part and the whole, the substance and the quality, are given
together though not in the same manner.

II
Thus far it was our task to elaborate the arguments of Gauta-
ma and his commentator, with a view to exhibiting their re-
levance to the contemporary discussions of the problem of per-
ception. What has been said would suffice, it is hoped, to show
that the Nyaya distinction between avayavin and avayava is a
most valuable means of rehabilitating the physical object lan-
guage. It is now left for us to ask what precisely is the nature of
the unity of the avayavin? The Nyaya, of course, tells us two
things about it. It is in the first place something other than the
mere assemblage of parts. And secondly the unity of the whole
resides in each of the parts in the relation of samavaya, while it is
wrong to say that the parts reside in the whole. In the following, we
shall attempt a phenomenological interpretation of these two
points, and in doing so, we shall, of course, depart from the naive-
ontological attitude of the Nyaya-Vaise;;ika system. Before,
however, we undertake this, it is necessary to draw attention to
certain unsatisfactory features of the Nyaya account.
In the first place, it should be borne in mind - and this is not
exactly pointing out a drawback of the theory - that Nyaya
does not bring out the exact difference between the mode of
perceiving the part and the mode of perceiving the whole. One of
the ways of doing this would have been to say that whereas the
part is perceived through the relation of samyoga, the whole
I92 THE NYAYA THEORY OF AVAYAVIPRATYAK$A

which resides in the part by the relation of samavaya is perceived


through the relation ot samyukta-samavaya. Nyaya does not
say this tor two obvious reasons. Two substances (dravyas) can-
not, in accordance with the categorial structure of Nyaya on-
tology, enter into a samavaya relation (which by definition is
reserved only for certain cases that do not include the case where
the relata are two dravyas not related as whole and part), and the
whole being a dravya can only enter into samyoga relation with the
sense organ. Further the supposition that the whole is perceived
through the indirect relation of samyukta-samavaya would lead to
the following grave difficulty. The part with which the sense organs
are said to be in contact is itself a whole some of whose parts
are unperceived; this whole, then, it must be said, is perceived
exactly like the bigger one of which it is a part, indirectly through
the relation of samyukta samavaya. In this manner, since every
part is a whole, we would be led to the unacceptable position that
there is direct contact through samyoga only with the last con-
stituents or atoms which are not any more wholes but which,
according to Nyaya, are imperceptible.
Nyaya, therefore, holds that both the part and the whole are
apprehended through samyoga. A whole is contacted through
samyoga, although not all its parts are so contacted. Inevitably,
however, different orders of samyoga would have to be admitted
and the privilege of having first-order samyoga would be ac-
credited only to the last imperceptible atoms. Though, however,
such different orders of samyoga would have to be postUlated,
nothing phenomenologically corresponds to this in the subjecti-
ve mode of awareness; for in each case, I have the same kind of
anuvyavasaya, 'I perceive ... ' The reason for this is that every
perceptible part is itself a whole. Since no perceptible substance
could have the primary order of samyoga, the idea of its absolute
givenness, in one of its senses, is here ruled out. This, in effect,
means that every whole is perceived through some of its parts;
and so on. Though in this manner every whole is perceived
through some of its parts, this however is not the same as saying
that it is inferred and not perceived. This, in fact, is the only
manner in which a substance could possibly bedirectlyperceived.l 5
15 The case, however, with either gU7J-as or universals is different, for these are not
wholes made of parts, not avayavins. Though these are perceived through some com-
THE NYAYA THEORY OF AVAYAVIPRATYAK$A I93

The real difficulty with the Nyaya account does not lie here.
It has to be sought in the arbitrary way it seeks to limit the con-
ception of avayavin. Firstly, not all conjunctions of parts, it is
said, give rise to true wholes. It is only a special kind of con-
junction that is regarded as giving rise to a true whole. Now it
seems to me that the N yaya- Vaise~ika theory is not quite clear
about the precise nature of this special kind of conjunction. It
also further seems to me that no strict line of demarcation can be
drawn between that conjunction of substances that gives rise to
a true whole and that conjunction which does not, for one reason,
amongst others, that the Nyaya-Vaise~ika philosophy believes in
the separability of the parts of even an avayavin (For the last
mentioned reason, the unity of the avayavin cannot be regarded
as what has come to be called an organic unity). No strict line of
demarcation, in that case, could be drawn amongst physical
objects, between genuine wholes and pseudo-wholes. The Nyaya-
Vaise~ika philosopher supposedly has, in mind, the idea that
genuine wholes are produced in a manner or in a sense in which
the pseudo-wholes are not. What precisely this manner is, in
what sense the idea of production applies to one and not to the
other case, and how precisely the conjunction of parts in the one
case differs from that in the other, - these are questions on which
no further light could be thrown except by referring to the way a
potter makes a pot or a carpenter a table. Perhaps the maker
alone knows the secret, but who then can be sure that some
whole has a maker and some others have not? Certainly the
separate physical objects are conceived by Nyaya as true ava-
yavins, in so far as the system believes in a maker for them all
on the analogy of a potter. But why, then, are we debarred from
treating the world as a whole as the ultimate avayavin? Nyaya
a voids this consequence by taking God's authorship of the world to
mean not that God has produced the world as a whole, regarded
as one single entity, as the potter produces a pot, but only that
His authorship pertains to each physical object taken sepa-
rately.
We thus notice the extreme difficulty of limiting the concept

posite relation like samyukta-samaviiya, yet in that relation it is given all by itself
and is not contacted through any of its parts (though again the gu'{/-a is contacted
through a substance, and a universal through its instances).
I94 THE NYAYA THEORY OF AVAYAVIPRATYAK~A

of avayavin only to some wholes. It cannot be said that to regard


the world as a whole as an avayavin would conflict with the
idea of external relation and the pluralism that are basic to the
Nyayaontology. For if it did then the very admission of the limi-
ted avayavin would have given rise to the same difficulty though
on a lesser scale.
It cannot also be argued that just as we directly perceive a
physical object as avayavin - for avayavitva is rightly held in the
theory to be perceptually known - so also we apprehend some of
the physical objects as wholes that are not again parts of bigger
wholes. In the words of Nyaya, we could say that carama-
avayavitva is not perceptually determinable. That a given whole
is not the part of a bigger whole is not given to perception.
The problem, therefore, with which we are faced is how best
to assimilate the central point of the Nyaya conception of
avayavi-pratyak$a into a satisfactory theory of perception with-
out limiting it in the way Nyaya does. To this task, then, we now
turn.

III

A satisfactory theory of perception must have to avoid the


atomistic conception of the given of the Humean-phenomenalis-
tic sort. It has also to avoid the conception of the given which
entails that the physical object is inferred but not perceived:
this is the view which Gotama and Vatsayana combat. The
physical object is directly perceived, though never all its parts
are: this is a point which a sound theory of perception must ad-
mit if it is to be honest to phenomena. But there are two direc-
tions in which it is possible to extend and amplify the thesis, not
in the interest of the theory, but at the dictates of phenomena.
In the first place, it is necessary to ask: what exactly is the
mode in which the whole is present in the part? The Nyaya-
V aise~ika answer is, of course, "through the relation of samavaya" .
But samavaya is also the relation by which qualities are present
in substances, and universals in particulars. By including all
these cases under one common type, we, of course, economise,
but the idea of economy, however indispensable it may be in
THE NYAYA THEORY OF AVAYAVIPRATYAK!?A I95
constructing a theory, may sometimes lead to a distortion of the
phenomena. What is common to them all is a certain insepara-
bility of the relata (ayutasiddhatva) , but beyond this they are
most unlike each other. Hence to say that the whole is present in
the parts by the relation of samaviiya, is not only not very illu-
minating but may prove misleading. The whole is indeed present
in the parts but not as one thing in another, not as a quality in a
substance, nor as a universal in the particular. The whole is pre-
sent in the part as the intended. The part, qua part, is, as it were,
saturated with an intention that refers beyond it to the whole.
The whole is, ontologically speaking, made up of parts. But phe-
nomenologically speaking, it is the fulfilment of the intention
awakened by the part. The perception of a part of the wall
arouses the intention that demands fulfilment and would not
be completely fulfilled short of the perception of all the parts.
In this way, we can give sense both to the fact that we know more
or less about the whole and to the fact that the whole was given
right from the beginning. The whole, as intended, was given right at
the outset, but a progressive fulfilment of this intention is possi-
ble. Since, however all the parts can never come to be apprehen-
ded, a complete fulfilment of this intention is an unattainable
ideal, an endless process of approximation. This, however, does
not negate the other fact that it is always the same whole that is
given as the all-pervading intention.
Secondly, it is necessary to make room for the phenomenon
that our perception of a physical object is never that of an iso-
lated self-complete atom but is always out of a situation, and that
likewise it tends to pass beyond itself, or within itself for further
determination. Now it seems possible to account for this charac-
ter of our perceptual experience if we lift the arbitrary limitation
imposed upon the concept of avayavin, so that every physical
object awakens an intention towards self-transcendence inas-
much as it is constituent of a bigger whole. The Nyaya idea of a
samuhiilambana knowledge, though based on an undeniable
phenomenon, is inadequate to account for that aspect of per-
ceptual experience to which I am drawing attention; for in a
piece of samuhiilambana knowledge, all objects are determi-
nately apprehended and possess equal importance, whereas in all
perceptual experience there is always, besides the focus (which
196 THE NYAYA THEORY OF AVAYAVIPRATYAK~A

may contain one or more substantive) a fringe and a tendency of


the focus either to pass beyond itself or to determine itself more
precisely (corresponding to what Husserl calls the outer and the
inner horizons). It cannot be said that the world is atomic and
that therefore the supposed continuity and dynamism within
perceptual experience is inadmissible, for we cannot argue from
what the world is like to what our experience should be like.
The correct procedure should be to start from the nature of
experience from the phenomenologically descriptive datum and
then - if one is still interested in ontology - to base one's con-
ception of the world upon it. It cannot be argued that the sup-
posed continuity - the fringe and the tendency - are not discer-
nible in the linguistic expression of a perceptual knowledge
which takes the simple form 'This is a jar'; for linguistic ex-
pression is not always a sure index to the nature of experience.
It simplifies the datum by limiting itself always to the fOCUS 16 and
this simplification, one might suppose, is in part due to the prac-
tical interest that hovers around.
A satisfactory theory of perception has to make room for these
two phenomenological data: the discreteness and the continuity
of our perceptual experience. It is not a phenomenalistic ato-
mism of the Humean-Buddhist type but only a physical object
atomism of the Nyaya type that does justice to the former aspect.
But a physical object atomism should have to accommodate the
other aspect of our perceptual experience; otherwise, it, too,
would be faulty by inadequacy. The continuity is no doubt
interrupted, as the flow of mental states, in general, is interrup-
ted in deep sleep or in swoon. But in every perception, there is a
tendency towards self-transcendence or towards further self-
determination, and this intention is imbedded not in our awa-
reness, but in the very object of perceptual awareness, in so far as
the objects of perceptual awareness constitute, by virtue of such
intentions, one single world.
Before closing this discussion I shall consider only two amongst
the many objections that might possibly be levelled against the

16 Nyaya also recognises a similar limitation to the claim of linguistic expression


to be index of experience: there is always, as a Nyaya rule runs, an unexpressed
qualifier except in the case of universals. See Mohanty, Gangesa's Theory of Truth,
Santiniketan, 1966.
THE NYAYA THEORY OF AVAYAVIPRATYAK~A I97

theory of perception briefly sketched above. It may be urged that


if the theory is true, then there would be no limit to what is
given, so that it should, by implication, follow that in any per-
ceptual situation the world as a whole is given, which is absurd.
In the second place - and this shall follow from the above - there
would remain no room for any other indirect, say, e.g. inferential,
mode of knowledge. Both the points together lead one to press
the question: what in that case is not given?
As to the first question, my answer would be twofold. It has
been said above that there is no general rule that the perception
of a part necessarily involves the perception of any definite whole
resident in it, though I must perceive some whole resident in it.
Now in that case even if we admit that the world as the largest
whole is resident in every part of it, it would not necessarily
follow that the world as a whole is given in every perceptual
situation. Since the world as a whole is not in fact given, we could
only postulate some form of inadequacy inherent in every per-
ceptual situation to present the world as a whole.
Secondly, saying that every perceptual situation gives rise
to an intention that takes us beyond it does not imply that all
such intentions are fulfilled. In fact, such intentions are more
often than not frustrated owing to the intervention of a stronger
cognitive or other interest.
Finally, it should be recognised that the world as a whole is
not one object amongst others. It is rather the horizon or the
dimension within which perceptions and their intentions are
possible.
It should be clear from the above remarks that the theory of
perception suggested in this paper does not necessarily lead to the
absurdity that everything is all at once given. The distinction
between the given and the not-given is not meant to be oblite-
rated.
CHAPTER XVIII

NYAYA THEORY OF DOUBT*

The Nyaya logic contains a theory of doubt. A preoccupation


with the nature, origin and structure of doubt seems out of place
in a logical system inasmuch as logic has been taken to be con-
cerned, speaking rather broadly, with formally valid thought
abstracted from its psychological context. Now, Nyaya logic -
in fact all Indian logic - does not conform to this conception.
It is in a broad sense coextensive with, and indeed indistinguish-
able from, a theory of knowledge, and concerns itself with all
kinds of knowledge, the non-propositional and the invalid ones
not excluding. In a narrower sense it is of course a theory of
inference. 1 But even as a theory of inference, (i) it does not
concern itself with the bare form, though some amount of for-
malism has been developed, and (ii) it does not separate logic
from psychology in a way in which western formal logic has done.
Consequently, it is as much interested in the psychological con-
ditions of the origin of a certain type of knowledge, say e.g.
of inference, as in the conditions of its logical validity. 2
It is in the light of these remarks about the general nature
of the Nyaya logic that we are to understand the reasons for its
preoccupation with doubt. For, inquiry (or, as the Nyiiyabhii$ya
says, pramii1Jairarthaparik$a1Jam i.e. the attempt to determine

* First published in the Visva Bharati Journal of Philosophy, III, 1966, 15-35.
1 Viitsiiyana for example defines Nyaya as Pratyak$dgamdsritamanumdnam
(Bha~ya on NyayaSl1tra 1.1.1.).
2 What is more, the Nyiiya goes to the extent of holding that a formally invalid
inference is even psychologically impossible, the socalled hetvdbhdsas being, not
errors in inference, but conditions which render an inference psychologically as well
as logically impossible.
NYAYA THEORY OF DOUBT I99
the nature of the object with the help of the various sources of
true knowledge) presupposes a prior state of doubt; though the
Nyaya allows for the case where we make an inference even when
there is prior certainty, there being however a special desire to
infer. The fact remains however that apart from such cases of
intellectual curiosity to provide reasons for what one already
knows for certain the most important stimulation for making an
inference is provided by a doubt about the presence of the
siidhya in the pak$a (e.g. of the fire in the hill).
It is further important to bear in mind the fact that for the
N yaya, as for most systems of Indian philosophy, doubt is a species
of knowledge, so that if I have a doubt of the form 'Is S p or not?",
most Indian logicians would say that I am having a knowledge-
though not a valid one about S. This rather strange contention,
so much at variance with both the philosophical and the ordi-
nary usages of the English word 'knowledge' may be accounted
for in either of two ways. It may be either that the Indian phi-
losophers, supported by the conventions of the Sanskrit language,
are using the word in such a wide sense as to include even doubt
and error. Or, it may be - and this seems to me to be the more
reasonable account - that the Sanskrit word 'Jiiiina' should not
be rendered into the English word 'knowledge', so that doubt and
error are species of]iiiina but not of knowledge. 'Jiiiina' means any
conscious state which is characterised by a reference to an object
beyond it, and surely doubt and error are states in which we are
conscious of something. To be conscious of something amounts,
according to the Nyaya, to having a jiiiina about that object.
There are various classifications of jiiiina, the most usual one
being into anubhuti and smrti (memory). The former may conve-
niently be defined as all jiiiina other than memory. Anubhuti
again is usually subdivided into pramii (or true) and apramii (or
false). A true jiiiina is one in which the object is known exactly as
it is, and a false one is one in which the object is known as what it
is not. 3 False jiiiina is either doubt or error. It may be noted that
the exact equivalent of the English word 'knowledge', in this
scheme, is 'pramiijiiiina'. Doubt is a kind of false jiiiina.
Since it has now been pointed out that 'jiiiina' is not strictly
synonymous with 'knowledge', we shall henceforth in this paper
8 Memory also is apramii, but not in the sense in which doubt or error is so.
200 NYAYA THEORY OF DOUBT

use the word 'knowledge' as if it were so synonymous, and leave


the matter at that with the hope that there would be no further
scope for misunderstanding.

II

The Siddhiintamuktiivali defines doubt as a knowledge which


is ekadharmikaviruddhabhiiviibhiivaprakiirakam, i.e. a knowledge
which has (two) contradictory prakiiras - one positive and the
other negative - but referring to the same substantive. From
amongst the host of definitions to be found in the Nyaya liter-
ature, this one may be singled out for its precision and simpli-
city, and it may be worthwhile to fix upon it. For an explanation
of the definition it is of course necessary to prefix a few words
about the concept of prakiira. 4
It is well known that according to the Nyaya, knowledge is
ontically formless (niriikiira) and owes its determinations to its
object. It is however capable of being logically analysed. Possi-
bility of such analysis presupposes that knowledge has forms of
its own in a quite different sense. But what precisely is this sense?
The Nyaya no doubt advocates a direct realism, and holds that
knowledge in an important sense has no forms of its own, that
it is niriikiira, its specific forms being derived entirely from its
object. However, the Nyaya also believes in the possibility of
analysis of knowledge, which presupposes that knowledge has
its constituent logical elements and relations.
In primary unreflective attitude, knowledge is directed to-
wards its object but not towards itself. The content of knowledge
is brought to light only in the reflective subsequent attitude. In this
reflective awareness, it is the contents of the primary knowledge
that are directly intended, whereas the object of the primary
knowledge is intended only as ancillary (pucchalagna). All
such contents of knowledge which reflection discovers are
brought under one category, technically called 'vi$ayatii',
which again is further subdivided into three sub-categories:

4 For a more detailed account of this and the allied concepts see Mohanty, Gan·
geSa's Theory 0/ Truth, Santiniketan, 1966, Introduction. See also Ingalls, Materials
/01' the Study 0/ Navya Nyaya Logic, Harvard, 1951.
NYAYA THEORY OF DOUBT 20I

'vise$yatii', 'prakiiratii' and 'samsargatii'. 'Vise$yatii is the general


title for all knowledge contents referring to substantives, 'pra-
kiiratii' for all contents referring to adjectives, and 'samsargata'
for those that refer to relations. 5
For illustration, consider the knowledge expressed in the judg-
ment 'Thisisapot' (Ayamghato). This knowledge maybe analysed,
at the first instance, into the following contents:
(i) a prakiirata referring to (the Nyaya would elliptically
say, 'attached to') pot-ness;
(ii) a vise$yata referring to the pot (in so far as potness quali-
fies the pot);
(iii) a prakiirata referring to the pot (in so far as the pot is a
determination of the mere this);
(iv) a vise$yata referring to the this (in so far as it is deter-
mined by the pot);
(v) a prakiiratii referring to this-ness (in so far as it qualifies
the this);
and (vi) a second vise$yata referring to the this (in so far as it is
qualified by this-ness).
Let us now introduce a few symbolical devices with a view to
facilitate a schematic representation of these contents in their
mutual interrelations.
We symbolise a vise$yata by enclosing the name forthe corres-
ponding element within the braces { } , and a prakiirata by
enclosing the name for its corresponding element within the bra-
ces ( ).
Thus '{this}' and '(pot)' would read as 'the vise$yatii referring
to this' and 'the prakiiratii referring to pot' respectively.
If and when a certain prakarata determines or limits a certain
vise$yata we shall simply write the symbol for the vise$yata first
and write that for the prakiirata after it. Thus '{this} (pot)'
would read as 'the vise$yatii referring to this is determined by the
prakiirata referring to pot'.
Enclosing the whole analysiens by the brackets [ ] and

5 Further elaboration of these concepts, howsoever necessary, would take us


beyond the scope of the present paper. See reference given under footnote 4. Be it
noted here, to avoid any further misunderstanding, that the epistemic distinction
between 'viSe$yata' and 'pl'akal'ata' does not correspond to the ontological distinction
between substance and attribute. Nor is it the same as the logical distinction between
'subject' ('uddesya') and 'predicate' ('vidheya').
202 NYAYA THEORY OF DOUBT

writing 'K' before it, we shall symbolise the knowledge whose


logical structure is exhibited within the outermost brackets. Thus
'K [{this} (potness)] is to be read as 'the knowledge whose vise-
$yatii referring to this is limited by the prakiiratii referring to
potness'.
, would symbolise 'and' and '",-,' would symbolise
'not'. Thus 'K [{s} (p). {t} (q)]' would read as 'the knowledge
whose vise$yatii referring to s is limited by the prakiiratii referring
to p, and whose vise$yatii referring to t is limited by the prakiiratii
referring to q'. But 'K [{s} ( (p).(q) )]' would read as 'the know-
ledge whose vise$yatii referring to s is limited by two prakiiratiis,
one referring to p and the other referring to q.'
'K [{s} ("'-'p)]' would read as 'the knowledge whose vise$yatii
referring to s is limited by the prakiiratii referring to the negation
or abhiiva of p'. But 'K [{s "'-' (p)]' would read as 'the knowledge
whose vise$yatii referring to s is not limited by the prakiiratii
referring to p. 'K [{s} "'-' ("'-' p)]' would on the other hand read as
the knowledge whose vise$yatii referring to s is not limited by the
prakiiratii referring to the negation of p'.
The Nyaya defines a niscaya or a certain knowledge as one
which, not having "'-' p as a prakiira has p as a prakiira (where, as
here, p is a term-variable).6 Following the symbolic conventions
stated above, we may then define a niscaya as
K [{s} (p). {s} "'-' ("'-' p)] ... (r)
As contrasted with niscaya, a doubt may then be defined as a
knowledge which has two mutually incompatible predicates
(ekadharmikaviruddhabhiiviibhiivaprakiirakam) one of which is the
negation of the other. In the case of the doubt 'Is this a man or a
lamp-post?', two mutually incompatible predicates are being
employed. It is however not sufficient for a knowledge to be called
a doubt that it should have two incompatible predicates, for it
may be - as in the case of the so-called samuccayajiiiina - that the
two incompatible predicates are referred to two different sub-
jects (e.g. 'This is a man and that is a lamp-post'). If samucca-
yajiiiina is to be represented symbolically as
K [{s} (p). {t} (q)] ... (2), a doubt has to be represented as

6 "tadabhavaprakarakam tatprakarakamjfianam (Muktavali on Karika 129).


NYAYA THEORY OF DOUBT 203

K [{s} (p). {s} (q)], where p and q are mutually incompatible


predicates ... (3)
Thus there are two essential components of doubt. In the first
place, the predicates must be mutually incompatible. Secondly,
they must be referred to the same subject. We shall enquire a
little more into each of these.
That the predicates of a doubt should be incompatibles is
suggested by the connective 'or.'7 The Riimarudri defines incom-
patibility thus: "virodhasca tadadhikara1Jiivrttitvam". 8 On this
definition, to say that p and q are incompatibles would mean that
one of them is never present in the locus of the other. The author
of the Nyiiyatiliivati is not satisfied with this definition of in-
compatibility, for in that case the definition of doubt would
illegitimately apply to such a case of error as 'This conchshell is
yellow' whose analysis may be stated as K [{this} (conchshell-
ness). {this} (yellowness)], where, granted that conch-shells are
always white and never yellow, the two predicates are incom-
patibles according to the definition of incompatibility given
above. It is necessary therefore that the two predicates should be
logical contradictories, 'p' and '~ p', one of which is bhiiva and
the other abhiiva. Doubt in that case would be defined as 'bhii-
viibhiivaprakiirakajiiiinam' and symbolically represented, in-
stead of (3), as:
K [{s} (p). {s} (~p)] ... (4)

The adjective 'viruddha' (incompatible, or opposed) may still


be regarded as not redundant inasmuch as there may be cases
where 'p' and '~ p' may be predicated of the same subject with-
out it being a case of doubt, as e.g. in the judgment 'The yonder
tree both has and has not contact with a monkey.' In this case,
contact with the monkey (kapisamyoga) and absence of such a
contact may both be rightly predicated of the same thing at the
same time, for it may have the contact in one part of it, and ab-
sence of it in another part. Dinakari therefore suggests that the
adjective 'viruddha' has significance. Even the contradictory

7 "Vtiktirathasca virodha", Kirantivalipraktisa/t (Gu~a), The Princess of Wales


Saraswati Bhavana Texts Series, p. 135.
8 SiddhtintamukttivaH with DinakarI and Ramarudn.
20 4 NYAYA THEORY OF DOUBT

predicates must be really incompatibles. Doubt then is viruddha-


bhaviibhavaprakarakajiianam. 9
I think, however, that the adjective 'viruddha' is superfluous.
The two predicates 'contact' and 'absence of contact' become in-
compatibles as soon as the subject is further determined. That
the subject of both is to be the same implies, strictly understood,
that the two subjects must have the same limitor (avacchedaka).
Referred to the same tree in the same part of it at the same time,
i.e. in the same spatio-temporallimitations, the two predicates
would certainly constitute a case of doubt.
This shows the importance of the second component of the
definition of doubt, namely, the requirement that the two pre-
dicates must be referred to the same subject. The idea of sameness
is deceptive. Consider a deliberate contradictory judgment
(aharyajiiana) of the form This hill without fire has a fire in it'
('nirbahni parvato bahnimaniti') whose form is'S which is ,-...., P is
p.' Here two mutually opposed and contradictory predicates are
being referred to the same subject, and yet we do not have a
case of doubt. Following Ramarudri, we may say that the ana-
lysis of such a case shows its form to be
K [{{s} (,-...., p)} (p)] ... (5)
which is different from K [{s} (p). {s} (,-...., p)]. The difference,
likely to be obvious from a mere inspection of the two schemata,
lies in the fact that in the case of (5) ,-...., p is predicated of mere s
whereas p is predicated of s as qualified by ,-...., p, while in (4)
both p and,-...., p are predicated of one and the same s.
The definition may nevertheless seem to apply to cases of
deliberate contradictory judgments of the form "S is both p and
,-...., p" which clearly are not cases of doubt. However, such judg-
ments may be regarded as carrying only sabdajiiiina, so that the
prakiiras are not 'p' and' ,"",p' themselves but the property that 'P'
and ',-....,p' are prakiiras (virodhinanaprakiirakatvaprakaraka).
This serves to distinguish a mere sabda awareness of a contra-
diction from a doubt where p and '""'p themselves are the prakaras
(virodhinanaprakaraka). This ingenuous distinction drawn by
Vardhamana both in Lilavatiprakasa 10 and in Kira1Javali-
9 Ibid., p. 480.
10 Chowkhamba ed., p. 414.
NYAY A THEORY OF DOUBT 205

prakasa,u may have its source in the consideration that a sabda


knowledge is mediated through a sentence, so that in this case
of merely verbal knowledge, 'p' and '-..p' directly qualify the
sentence and the fact of their so qualifying the sentence may then
be regarded as qualifying the resulting knowledge. In our sym-
bolism, this case may be represented as
K [{s} . ( (p). (-.. p) )] ... (6)
Gadiidhara holds that the two contents or vi$ayatas belonging
to a doubt have the following three properties:
(i) One of them is incompatible with the other in the sense
that one acts as a hindrance (pratibandhaka) to the other:
(ii) nevertheless, the two are co-present;
and (iii) the one content belongs to the knowledge only as qua-
lified by the other, and therefore not as an independent
content. 12
The Naiyayikas have further discussed the question - not of
great importance though - if a doubt has two predicates or four.
Added to this is the controversy, touched upon earlier, as to
whether in case there are two alternatives they are both positive
incompatibles or logical contradictories. We get accordingly three
schemata which may be exhibited as follows:
(a) Two positive alternatives theory:
K [{s} (p). {s} (q)], p and q being incompatibles.
(b) One positive and the other negative alternative theory:
K [{s} (p). {s} (-..p)J
(c) Four alternatives theory:
K [{s} (p). {s} (-..p). {s} (q). {s} (-..q)]
There seems nevertheless to be something about a doubt which
escapes the attempt to analyse it logically. We may grant that
the Nyaya is not committing the obvious error of mistaking a
doubt-sentence for a propositional one. What the Naiyayika
seeks to analyse is not the sentence, not the proposition certainly
in this case - for there is no proposition here - but the inana as
apprehended in reflective awareness. In spite of all this we may

11 Loc. cit., p. I38.


12 Hence Gadadhara's rather forbidding definition: "svavacchinnaprativadhy-
atanirupitapl'ativandhakatva-svasamanadhikara'("yobhayasambandhena vi~ayataviSi$ta­
vi$ayatti.salijiianam samSayapadarthalt"_ (Pramti.nyavada Gadadharl).
206 NYAYA THEORY OF DOUBT

nevertheless point out that the above analysis still misses some-
thing essential to doubt qua doubt.
There were amongst the Naiyayikas some who sought to reduce
a doubt-sentence to the compresence of two contradictory asser-
tions 'S is p' and'S is not p'. This view traditionally ascribed to
the author of Ratnako$a is voiced by GangeSa, when he in course
of an argumentation with the Mimamsakas, contends that doubt
is nothing but such joint predication.13 Happily, this view is not
shared by GangeSa himself, for he tells us soon after that doubts
are characterised by Kotyutkatatva, i.e. difference in the relative
strength of the alternative predicates. In a mere compresence of
two predications, the question of relative strength of the alter-
natives would not arise. Vacaspati refers to three possibilities
from this point of view: either the affirmative predicate (p) is
relatively stronger, or the negative predicate ('"" p) is the stronger
one, or it may be that both the alternatives are equally strong. 14
In any case, doubt would involve an oscillation of the mind be-
tween the two alternatjves: it is this which he has in mind when
Vardhamana so aptly characterises doubt as doliiyitiinekakotika,15
i.e. as a knowledge where there is, as it were, an oscillation be-
tween the alternatives. I think, it is this state of the mind, this
doliiyitatva that is an essential character of doubt and should be
added to the structural analysis explained above, - unless of
course it could be shown that such a character follows from the
structure revealed in (4). I do not know however how this could
be shown.
Distinction should nevertheless be drawn between doubt and
question. Doubt is no doubt one of the sources of enquiry, though
not all doubt is so. There are doubts that are not important
enough and are just set aside and do not initiate any enquiry
whatsoever.16

13 "Viruddhobhayaropasamagridvayasamajadubhyaropa eka eva bhavati sa eva


samsayalt" (Pramav-yavada, Darbhanga edition, p. 91).
14 Ibid., p. 92.
15 Vardhamana, Kirav-avaliguv-aprakasa, p. 130.
16 Vardhamana, uNa ca jijiiasajankam iiitinam samsayalt upek$aniyasamsayav-
yaptelt" (ibid., p. 183).
NY AY A THEORY OF DOUBT 207

III

Having given an outline of the structural analysis of doubt


given by the Naiyayikas, we may now turn to certain ancillary
issues concerning it. There is in the first place the question of
classifying doubt into various types, and there is secondly the
question regarding the causes of doubt. The Naiyayikas have
generally taken up these two issues together and have classified
doubt according to its origination. The Naiyayikas are not all
agreed about any of these issues, and Gotama's sutra on this has
been subjected to conflicting interpretations. Vatsayana, Uddy-
otakara and Vacaspati differ amongst themselves, not to speak
of their differences from the Navya Naiyayikas. It is difficult to
evolve an agreed formula. I give below what seems to me to be
an account which cuts across the divergences of opinion about
the causes of doubt.
These causes may be divided into two groups: the general
causes and the specific causes. By the general causes of doubt we
mean those factors which must be present so that any doubt at
all may occur. They are in other words causes of doubt qua doubt.
The specific causes are the causes only of specific kinds of doubt,
and are not therefore to be regarded as causes of doubt qua
doubt. If doubts are to be grouped in accordance with their
origination, it is only these latter, namely the specific causes that
are to be taken into consideration.
A. The general causes may be brought under two sub-heads, the
positive and the negative.
(a) The positive general causes of doubt are two: (i) dharmi-
jiiiina and (ii) vise$asmrti.
(i) In the first place, a doubt qua doubt presupposes a knowl-
edge of some sort of the dharmi, i.e. the substantive of which the
two mutually contradictory predicates are predicates. It should
be obvious that this knowledge of the dharmi should be a ni-
scaya i.e. a certainty, and cannot itself be a doubt, for otherwise
the latter doubt would presuppose a further dharmijiiiina, thus
leading to an infinite regress. Consider the doubt 'Is this a man
or not ?'. Here though the doubter is not sure whether this is a
man or not, he has a certain apprehension of the object here
208 NYAYA THEORY OF DOUBT

before him as a this, and may be along with some other generic
characters. Gangesa suggests two reasons why this factor should
be regarded as an essential precondition of all doubt qua doubt.
If dharmijfiiina were not required for all doubts, there ought not to
have been the rule that all doubts must have some substantive.l7
In fact, however, doubts are of the form K [{s} (p). {s} (,....., p)],
and not of the form K [(p). ('"" p)]. Further, the property that in
doubts one of the alternatives may be stronger than the other
(kotyutkatatva) cannot be explained otherwise, for in the know-
ledge K [(p). (,....., p)], 'p' and ''"'' p' should have no difference in
status; any difference which they may have must be in their
relation to the's' which is being apprehended as's'.
(ii) Mere dharmijfiiina is not enough to produce a doubt. More-
over mere knowledge of's' as's' does not explain why the doubt
should have the predicates 'p' and' '"" p' and not the predicates,
let us say, 'q' and ',....., q'. We need therefore another positive,
general condition, namely a remembrance of the two alternatives
'p' and ',....., p'. This is what is called vise~asmrti, 'p' and ',....., p'
being the vise~as or specific characters. It may also be called
kotismrti for they are also called the kotis or alternatives.
(b) The negative general condition necessary for all doubts
qua doubt is non-perception of the specific characters as belong-
ing to the substantive (vise~a-adarsana). Definite knowledge of
the presence of any of the specific characters in the substantive is
a hindrance to doubt. If the supposed doubter knew for certain
that s has p, or if he knows for certain that s has,....., p, then the
doubt 'Is s p or ,....., p?' would not obviously arise. Hence the ab-
sence of such specific knowledge is a necessary condition of all
doubt qua doubt.1 8
B. While the conditions listed under A are necessary for there
being any doubt at all, there are however other special causes of

17 "Dharmijiianam ca samsayahetu!t, anyatha samsaye dharminiyama!t kotyut·


katatvam ca na syat" (loc. cit., p. 92).
18 It may be mentioned that both the factors (a) (ii) and (b) are implied in Gotama's
sutra where doubt is characterised as being vise$apek$a, which may mean either
viSe$adarsana (apek$a=adarsana) or vise$asmrtyapek$a corresponding to the above
two factors respectively. There is a third possible interpretation according to which
this expression means "the desire to apprehend the specific characters" (apek$a=
akank$a or desire); but we cannot include such desire amongst the necessary pre-
conditions of doubt qua doubt. There may be doubts even when such desire is absent.
Moreover, often the desire to ascertain is a consequence rather than a cause of doubt.
NYAYA THEORY OF DOUBT

specific types of doubts. Thus doubts may be caused by either (a)


perception of the common character (samanadharmopapatti) , or
(b) perception of an uncommon character (anekadharmopapatti),
or (c) hearing contradictory views expressed by parties opposed
over an issue (vipratipatti), or (d) reflection on the absence of any
concommitance between being experienced and being real and
between not being experienced and being unreal (Upalabdhi-
anupalabdhi-avyavastha). Let us explain each one of these with
suitable examples.
(a) A doubt of the form 'Is s p or not?' may arise from per-
ception of some character common to both 'p' and 'not-p' (pro-
vided of course, it is accompanied by perception of s as s, non-
perception of the specific characters 'p' and 'not-p', and re-
membrance of those specific alternatives). By a common cha-
racter is here meant any character or characters which are pre-
sent where p-ness is present, but is also present where p-ness is
not present, i.e. which may be accompanied either by p-ness or
by not-p-ness. Seeing something at a distance (s as a mere this),
and perceiving its height, size, shape etc. which could very well
belong to a man or to a dead tree trunk, one might doubt 'Is this
a man or a dead tree trunk'? If at this stage he could detect any
of the specific properties which goes only with manhood, then
his doubt would give place to the certainty 'This is a man'. Now
it seems clear that in doubts arising from the perception of a
common character in the above sense, the alternatives tend to
be positive contraries instead of being logical contradictories,
a point recognised by Vacaspati when he says that doubts of
this kind are characterised by vidhipradhanya.
(b) A doubt may arise also from the perception of an uncom-
mon character. On perceiving some uncommon character in
some substantives s, one may be haunted by the doubt what
character it shares or has in common with others. The Nai-
yayika's favourite example is this: if sound is known merely as
possessing soundness (which is its distinguishing and in that sense
uncommon character) one may very well doubt if it over and
above this possesses eternality or non-eternality, for both are
compatible with soundness:
The distinction between cases under (a) and those under (b)
is apt to be overlooked. In case (a), one perceives in s a cha-
210 NYAYA THEORY OF DOUBT

racter X which is common in the sense that it as a matter a fact


accompanies and is consistent with both the alternatives. It
accompanies p-ness as well as not-p-ness. In case (b), a character
x is perceived in a s such that x belongs to s alone, and x is con-
sistent with both, but not known to accompany either of, p-ness
and not-p-ness.
(c) When in course of a disputation, the opposed and contrast-
ing parties put torward their respective theses, a hearer is very
likely to be overcome by a doubt as to which of the theses is the
correct one. The older Naiyayikas take this as a special case of
doubt, where the doubt is sabda, i.e. generated by hearing and
understanding of the words uttered by the disputing parties.
In such a case, of course, doubt arises not in the mind of the
disputationists, for each of them is convinced of the correctness
of his own contention, but in a neutral observer (madhyastha)
who is confused by the mere statements of the contradictory
positions advocated jn the absence of any decisive supporting
arguments. There must be absence, in other words, of anyata-
rasadhakahetu. Such doubt, once it has arisen, cannot be re-
moved by the mere collective judgment (or, sampratipatti) of the
form 'A holds p to be the case, and B holds not-p to be the case'.19
What is necessary is the ascertainment which of the two is really
the case.
N aiyayikas are divided over the issue whether a doubt arising
from this special cause is to be called sabda or manasa. The ques-
tion in other words is: does the doubt arise through hearing, or
does it arise through the operation of mind (manas)? Raghunatha
defends the former alternative and has the older authorities on
his side. 20 Viswanatha argues in favour of the latter alternative,
and makes use of the premise that sabda as a rule is a source of
certainty so that by itself it cannot generate doubt. 21 What
happens according to him is that the statements of the disputing
parties give rise to remembrance of the alternatives. This latter
knowledge, then, provided all the other required conditions are
present, gives rise to doubt which therefore is manasa and not
sabda.

19 See Vatsayana Bha$ya on Nyaya Satra 2.1.6.


20 See Phanibhusan Tarkabagish on sutra 2.1.6. (Nyayadarsana Vol. II).
21 .. Sabdavyaptijiianadinam niScayamatraianakatvasvabhavat".
NYAYA THEORY OF DOUBT 2II

(d) Being an object of experience is not a sure mark of being


real. Epistemological objecthood mayor may not be accompa-
nied by ontological independence. Both the real water and the
water-in-the-mirage are objects of experience. Both the real
snake and the illusory snake-in-the-rope are seen. Therefore from
the mere fact that something is being experienced one cannot
make sure as to whether the experienced something is also real or
not. There may therefore arise in such a case doubt about its
reality or unreality.
Similarly not being an object of experience is not a sure mark
of unreality. The unreal of course may not be experienced. But
so also frequently is the real. Not all that is real is experienced.
Therefore, from the mere fact that something is not being ex-
perienced nothing can be ascertained as regards its reality or
unreality. There may therefore arise in such a case doubt about
the reality or unreality of what is not being experienced.
Attempts have been made to explain case (d) in other ways.
Consider the possibility (d'): Supposing I am having a knowledge
which certainly possesses the generic character of experience-ness
or jiianatva. This generic character however is consistent with,
and is accompanied by, either of the two specific characters, the
property of having a real object (sadvi$ayakatva) and the pro-
perty of having an unreal object (asadvi$ayakatva). If we do not
experience any of these specific characters in the knowledge
under consideration there may be a doubt in accordance with
rule (a).
However, there is a difference between (d) and (d'). In (d),
the doubt concerns the object of the experience under consider-
ation. The object being an epistemological object mayor may
not be ontologically real. In (d'), on the other hand, the doubt is
about the knowledge or experience itself. Its being an experience
does not entail either that it is sadvi$yayka (true) or that it is
asadvi$ayaka (false).
It has also been contended by others22 that (d) is a special
case of another rule (d"): doubt about the truth of a knowledge
gives rise to doubt about the reality of the object of that know-
ledge (prama1Jyasamsayat vi$ayasamsayal:t). Let K be a know-

22 See for example Dinakari on Siddhlintamuktlivali on klirikli 180.


212 NYAYA THEORY OF DOUBT

ledge having 0 for its object. If for any reason I have a doubt of
the form 'Is K true or not?' this would generate a further doubt
of the form 'Is 0 real or not?'
There is again a nice point of difference between (d) and (d").
In case of (d) what causes doubt about the reality or unreality of
o is not a prior doubt in the truth of K but the perception of the
generic character of 0 as an object of knowledge, this character
being consistent with both the reality and unreality of O.
The importance of the rule (d") - and one reason why it cannot
be reduced to any other - is that though a prior certainty about
an object (arthaniscaya) rules out the possibility of doubt about
the same object, nevertheless such doubt may be caused by an
intervening doubt in the truth of that initial certainty. The
sequence in such cases may be set down thus:
1. Certainty, 'K', about O. 2. Doubt: 'Is K true'?
3. Doubt: 'Is 0 real?'
In the absence of (2), (3) cannot take place when (r) has al-
ready been there, the general rule being that though doubt does
not obstruct certainty (for otherwise doubt would never be
resolved), yet certainty does exclude doubt except in the case
coming under (d' ') .
(e) Another rule, which according to many comes under the
'ea' of Gotama's sutra 1.1.23, is to the effect that a doubt about
the pervaded gives rise to a doubt about the pervader (vyiiPY-
asandehiit vyiipakasandeha).23 Smoke, for example, is pervaded
(vyiipya) by fire which is the pervader (vyiipaka) in relation to it.
Wherever there is smoke, there is fire. Smoke is never present
in any locus of the absence of fire. If a person who knows this
relationship between smoke and fire perceives smoke in a distant
hill and recognises the smoke as the vyiipya of fire, he would
naturally infer, and so arrive at a certainty that the hill also
possesses fire. If however such a person, for whatever reason,
comes to have the doubt whether what looks like smoke is really
smoke or not, he would be led to the further doubt whether the
hill possesses fire or not. Of course, here as before in the case of
inference, it is necessary that smoke should have been earlier
known and in the present recognised to be a vyiipya of fire. It

23 See Ingalls, loco cit. for the relation of pervasion or vyapti.


NYAYA THEORY OF DOUBT 21 3

also holds good that certainty about the vyiipya of anyone of


the alternatives of a doubt would necessarily put an end to the
doubt. Consider the doubt 'Is this a man or not?' As soon as the
doubter comes to perceive clearly such features as hands, feet
etc. in the object before him which is being referred to as this,
his doubt would give place to the certainty 'This is a man', for
the property of possessing limbs is a sure mark of manhood.
Hence, the vise$iidarsana or non-perception of specific characters
- which is one of the general conditions of all doubt qua doubt -
must be taken to include non-perception of the marks (or vyii-
pyas) of the specific characters.

IV

In this section we propose to examine Descartes's doubt with


the help of the Nyaya theory outlined above. Such a confron-
tation, it is hoped, will help us to throw light on both the sides,
and thereby on the nature of doubt qua doubt.
Descartes's doubt applies in the first instance to anything and
everything in the world and also to any and every knowledge and
experience. In his first Meditation and also in The Principles
of Philosophy, Part I, he gives us the grounds of his universal
doubt. These grounds are the following:

1. The senses are often found to mislead us. We cannot there-


fore place absolute confidence in them, for "it would be impru-
dent to trust too much to what has even once deceived us". 24

2. Secondly, "in dreams we perpetually seem to perceive or


imagine innumerable objects which have no existence".25

2a. There are, Descartes argues, "no certain marks by which


the state of waking can ever be distinguished from sleep",26 i.e.
from the state of dreaming.

24 Descartes, The Principles of Philosophy (Everyman ed.), p. 166.


25 Ibid., p. 166.
26 Descartes, Meditations (Everyman ed.), p. 81.
NYAYA THEORY OF DOUBT

3. With regard to the supposedly self-evident truths of mathe-


matics Descartes employs the following two arguments:
a. It is often found that men fall into error even in such mat-
ters, and regard as self-evident what is really false. 27
b. More important for Descartes is this one: We believe that
God who created us is all-powerful. We do not however know for
certain whether this all-powerful God is not a deceiver. It may
therefore be that he created us with the will to deceive us. If on
the other hand the creator is not all-powerful then we shall be
more imperfect and more likely to be under continual deception.

4- There is a final argument which, as would be clear from the


remarks to follow is of the highest importance. We possess a free
will and we are therefore free to withhold our assent from what-
ever is doubtful. In other words, we may suspend our belief in
whatever is not "manifestly certain and undoubted". 28
It seems clear that Descartes's arguments (r) and (z) come
under the Nyaya rule (d), i.e. they are really based upon what the
Nyaya calls Upalabdhi-avyavasthii. In this respect these two
really constitute one argument. They appeal to the fact that
there is no fixed correlation between being an object of exper-
rience and being real. The unreal is as much an object of ex-
perience as the real. What is presented through the senses may
then be unreal, just as what is presented in a dream may seem
to be real. The doubt therefore may be accounted for by (i)
perception (mentally) of the generic character of objectivity
(fniinavi$ayatva) and (ii) uncertainty as to reality or unreality,
arising out of the absence of any settled order in such matters.
The argument (za) however presents great difficulty. What
is necessary for the possibility of a doubt of the form' Am I awake
or am I dreaming?' (= 'Is this a dream or is it a waking exper-
ience ?') is that I should perceive the generic character of ex-
perience-ness (iniinatva) , and yet fail to perceive either of the
two specific characters (which in the present case are the property
of being a dream and the property of being a waking experience)
or their respective marks. The possibility is a priori implied
therein that there are such marks, and that it is possible to dis-
27 The Principles of Philosophy, p. 166.
28 Ibid., p. 166.
NYAYA THEORY OF DOUBT 2I 5

tinguish between the two specific characters though in any given


case one may fail to do so. Descartes however contends that
there are "no certain marks" by which one may be distinguished
from the other. If two properties 'p' and 'q' cannot at all be dis-
tinguished, i.e. no sure mark exists which could serve the purpose,
then there is no question of the non-perception of such marks and
hence no possibility of doubt with regard to them. If on the other
hand there are such marks though Descartes fails to adduce any
then his doubt cannot claim universality. He could then only say
that he could not then and there distinguish the one from the
other. Moreoever, if no sure mark of dream experience were
known to him on what ground could he almost persuade himself
to think that he was then dreaming?29 The point is that a doubt
of the form 'Is S p or q?' requires both that 'p' and 'q' are dis-
tinct with their respective distinguishing marks and that in a
given case there is a non-perception of them. These two conditi-
ons defeat the possibility of a universal scepticism.
The argument (3a) is formally of the same type as the first
argument and is to the effect that where there is the least chance of
error, where in other words there is no upalabdhivyavasthii i.e. no
rule that only the real is experienced, one may reasonably doubt.
This applies as much to sense-perception as to mathematics. And
incidentally it may be pointed out that the argument applied to
mathematical truths is close to the point of view of the Nyaya
logic which does not admit the distinctions between analytic
and synthetic, a priori and a posteriori, self-evident and not-self-
evident truths. The subjective possibility of error being always
there, there may be doubt regarding the truth of any knowledge
whatsoever as also a resulting doubt about the reality of the
object of such knowledge.
However, none of the arguments I-3, though sanctioned by the
Nyaya rules, can be used for the purpose of justifying a universal
scepticism. For among the necessary conditions of doubt qua
doubt there is at least one which constitutes a certainty: this is
the dharmijiiiina or knowledge of the substantive. In any parti-
cular doubt there must be certainty about the dharmi. Basing on
the facts of error, we may have two kinds of doubt: the one of the
same sort as 'Is this a rope or a snake?', another of the philo so-
29 Meditations (Everyman's ed.), p. 81.
216 NYAYA THEORY OF DOUBT

phical kind: 'Is sense-perception valid or not?' or 'Is the world


real or imagined?'. It is easy to show that doubts of the second
kind are self-stultifying, for they question the very reality of
their own respective dharmis which they cannot consistently do.
Two other arguments of Descartes remain to be examined. The
doubt involved in 3(b) may be restated thus:
'Is God who is known to be all-powerful also a deceiver or
not? The doubt so formulated seems to be sanctioned by the
Ny1iyaruleB(b). Here we have an uncommon character belong-
ing to the dharmi i.e. being both the creator of the world and
all-powerful, and we are left in doubt as to which of the two pro-
perties 'being a deceiver' and 'being veracious' - both compatible
with the above uncommon character - further belongs to it.
Such a doubt, if it comes to happen, would no doubt have a
limitless scope with regard to the truth of all our experience.
It would not however apply to our belief in the fact that there is
an all-powerful creator. The argument is to that extent effective,
but loses its force because of the fact that its starting point is a
theological belief from which a reflective philosopher may not
start.
Descartes's Cogito which sets a limit to his doubt may be inter-
preted as the ultimate dharmi, certainty about which is presup-
posed in any doubt. But it should be pointed out that the 'I' is
the dharmi only in the reflective judgments of the form 'I know',
'I perceive' etc., but not in the unreflective judgments of the
form 'This wall is white', 'The yonder bird is a crow' etc. Doubts
being of the form 'Is this wall white or not?', 'Is the yonder bird
a crow or not?', they do not presuppose certainty about the 'I'
as their dharmi. In case however these doubts come to occur
through the instrumentality of doubts of the validity of the
respective knowledges (as per rule d"), then of course certainty
about the 'I' would be presupposed, for the 'I' is the dharmi in
the latter doubts that have the form 'Did I know rightly or
not ?'. If Descartes's doubt is to justify certainty about the Cogito,
then he must be interpreted as having taken to this reflective
way of having first doubted the validity (priimii1Jya) of our
beliefs, and then arrived at the doubt about the reality of the
objects of those beliefs. Such an interpretation is amply borne
out by Descartes's writings.
NY AY A THEORY OF DOUBT 21 7
One of the chief grounds sustaining Descartes's universal scep-
ticism lies hidden in the last of his arguments. The human will, he
writes, is free and so is also free to withhold its assent from what-
ever is doubtful. It must readily be seen that this argument
represents a type of thinking foreign to the Nyaya, and in fact to
all Indian philosophy. Withholding assent or doubting as a func-
tion of the limitless freedom of the will is not recognised as a
possibility in the Nyaya, and therefore the argument (4) does
not conform to any of the Nyaya rules. 3o And yet if anywhere it is
here that we shall find the sources of a truly philosophical doubt.
With a view to looking closer into the nature of this argument
let us ask what is meant by the two expressions "withholding
one's assent" and "whatever is doubtful"?
Withholding one's assent means a deliberate, reflective decision
not to believe, to suspend or neutralise one's belief, to 'bracket'
it as Husserl would say. The motive for doing this - with Des-
cartes, and also with Husserl - is the reflective one of finding
a secure basis for human knowledge, a radical foundation, a
first principle for the sciences. The Nyaya is operating with a
strictly causal-deterministic conception, and within such a frame-
work a doubt could occur only when there are necessary and
sufficient conditions for it. The Naiyayika might seek to include
this reflective doubt within his own deterministic framework by
tracing it to the factor of iccha or desire which is recognised by
him to overpower others. But I wonder if this would help us to
overcome the great difference that subsists between the two
conceptions, which may perhaps be brought to light in still
another way.
Withholding one's assent to a belief does not exclude making
practical use of that belief. Descartes and Husserl, just when they
ask us to doubt, or to practise the epoche, do not suggest that
that would mean giving up and reorientation of our practical
behaviour, of our Lebenswelt based precisely on those beliefs.
On the contrary, Descartes writes:
" ... we ought not meanwhile to make use of doubt in the

30 The only rule which bears a certain semblance to the Cartesian case under
consideration is the Nyaya rule that the factor of desire (iccha) overpowers any other
set of factors tending to produce a contrary result.
218 NYAYA THEORY OF DOUBT

conduct of life". (The Principles of Philosophy, part 1. III.)


And Husserl says the same of his phenomenological epoche: the
epoche will not affect the daily course of practical life; it will
only suspend theoretical judgment about the 'being' of the world.
What is excluded is the possibility of making any theoretical
use of the beliefs concerned. I wonder if the N aiya yika would
approve of this attitude. For him, though ascertainment of the
truth of a belief (pramanyagraha) is not necessary for the appro-
priate practical behaviour, yet non-apprehension of its falsity is
certainly a necessary condition. 31 Now, on the Nyaya analysis,
doubt in the truth of a knowledge has the form:
K [{this knowledge} (truth). {This knowledge} (falsity)]. In so
far as this doubt has falsity as one of its prakaras, the doubt is
an apprehension of falsity in the belief and therefore on the Nyaya
rules would serve as a pratibandhaka or hindrance to the appro-
priate practical behaviour. Normal practical behaviour would
therefore be impeded by a universal scepticism. Descartes and
Husserl do not apprehend this possibility but on the other hand
assure us that their doubt would leave the practical Lebenswelt
untouched. There must then subsist a radical difference between
the two doubts. They must then be not merely different kinds of
doubts, but as doubts different.
The same radical difference comes to light if we examine what
Descartes means by 'whatever is doubtful'. In one meaning of it,
a thing is doubtful if it is in fact an object of doubt. But, for
one thing, this is not all that is there in the ordinary meaning of
that expression; and, for another, if to be doubtful meant to be in
fact an object of doubt, then Descartes' decision to withhold as-
sent from what is doubtful would be trivial; he would then be
asking us to doubt what is being doubted, which would be utterly
pointless. To be doubtful then means to be a possible object of a
doubt. Now on the Nyaya theory nothing possesses any property or
properties which make it liable to be doubted. Nothing by itself,
i.e. by virtue of any of its own properties, is doubtfuL Everything
at the same time is a possible object of a valid knowledge, i.e. a
prameya. Suitable epistemic conditions may however produce in a
person doubt about anything. For Descartes, and for the entire
tradition of Western philosophy there is an important sense in
31 Gangesa, Prama1Jyaviida.
NYAYA THEORY OF DOUBT

which a thing may meaningfully be called doubtful. This is the sense


that the thing could have been otherwise, that its contradictory
is possible, or that it is, though a fact, a contingent one. If there
is any p of which it holds good that both p and not-p are pos-
sible then it is doubtful. It is not necessary. It is, as Descartes
says, uncertain. From all such things we are entitled to with-
hold our assent.
It would be obvious that the Nyaya knows no such categori-
sation of things or facts or propositions into contingent and ne-
cessary. It knows, as said before, no distinction between what is
in tact true but might not have been and what is necessarily true,
between the a posteriori and the a priori, or even between ana-
lytic and synthetic truths. Bare logical possibility or counter-
factual conditionals do not interest it. It gives a logic of facts,
and in this sense is extensional, avoiding modal concepts. 32
It would not therefore approve of a universal scepticism, based
on the notion of logical possibility.
Should we then say that here is an overall limitation of the
Nyaya logic, whose symptoms show themselves in all aspects of
it far beyond the narrow subject matter of the present paper?
Perhaps it is so. It may also well be the case that doubt (or
samsaya) in one sense is exactly what the Nyaya means by it,
and for it the Nyaya logic is well adapted. At the same time,
philosophical doubt, doubt in the reflective level, falls beyond
its scope. And the two doubts, it may well be, are not only dif-
ferent kinds of doubts but are as doubts different. One and the
same logic cannot do justice to both. Descartes may be said to
have erred on the opposite side, when he sought to extend the
logic of ordinary doubt to philosophical doubt, which is the same
as using arguments (r) - (3) to justify the latter.

32 It should be remembered however that the Nyaya logic is not extensional in


the sense that it knows no quantification. But that is an altogether different sense of
extensionality.

You might also like