CTLF Corpus de textes linguistiques fondamentaux • IMPRIMER • RETOUR ÉCRAN
CTLF - Menu général - Textes

Malinowski, Bronislaw. The Meaning of Meaning – T02

Supplement I
The Problem of Meaning in primitive

By Bronislaw Malinowski, Ph.D., D.Sc.
Late Professor of Anthropology, University of London.

I. The need of a Science of Symbolism and Meaning, such
as is presented in this volume by Ogden and Richards.
This need exemplified by the Ethnographer's difficulties in
dealing with primitive languages.

II. Analysis of a savage utterance, showing the complex problems
of Meaning which lead from mere linguistics into the
study of culture and social psychology. Such a combined
linguistic and ethnological study needs guidance from a
theory of symbols developed on the lines of the present work.

III. The conception of ‘Context of Situation.’ Difference in
the linguistic perspectives which open up before the Philologist
who studies dead, inscribed languages, and before
the Ethnographer who has to deal with a primitive living
tongue, existing only in actual utterance. The study of an
object alive more enlightening than that of its dead remains.
The ‘Sign-situation’ of the Authors corresponds to the
‘Context of Situation’ here introduced.

IV. Language, in its primitive function, to be regarded as a
mode of action, rather than as a countersign of thought.
Analysis of a complex speech-situation among savages.
The essential primitive uses of speech: speech-in-action,
ritual handling of words, the narrative, ‘ phatic communion ’
(speech in social intercourse).

V. The problem of Meaning in primitive languages. Intellectual
formation of Meaning by apperception not
primitive. Biological view of meaning in early non-articulate
296sound-reactions, which are expressive, significant
and correlated to situation. Meaning in early phases of
articulate speech. Meaning of words rooted in their pragmatic
efficiency. The origins of the magical attitude towards
words. Ethnographic and genetic substantiation of Ogden's
and Richards' views of Meaning and Definition.

VI. The problem of grammatical structure. Where to look
for the prototype of grammatical categories. ‘Logical’
and ‘purely grammatical’ explanations rejected. Existence
of Real Categories in the primitive man's pragmatic outlook,
which correspond to the structural categories of language.
Exemplified on the nature of the noun and of other Parts of


Language, in its developed literary and scientific functions,
is an instrument of thought and of the communication of thought.
The art of properly using this instrument is the most obvious
aim of the study of language. Rhetoric, Grammar and Logic
have been in the past and still are taught under the name of Arts
and studied predominantly from the practical normative point of
view. The laying down of rules, the testing of their validity,
and the attainment of perfection in style are undoubtedly important
and comprehensive objects of study, especially as Language
grows and develops with the advancement of thought and
culture, and in a certain sense even leads this advancement.

All Art, however, which lives by knowledge and not by inspiration,
must finally resolve itself into scientific study, and there is
no doubt that from all points of approach we are driven towards a
scientific theory of language. Indeed, for some time already, we
have had, side by side with the Arts of Language, attempts at
posing and solving various purely theoretical problems of linguistic
form and meaning, approached mainly from the psychological
point of view. It is enough to mention the names of
W. von Humboldt, Lazarus and Steinthal, Whitney, Max Müller,
Misteli, Sweet, Wundt, Paul, Finck, Rozwadowski, Wegener,
Oertel, Marty, Jespersen and others, to show that the Science of
Language is neither new nor unimportant. In all their works,
besides problems of formal grammar, we find attempts at an
analysis of the mental processes which are concerned in Meaning.
But our knowledge of Psychology and of psychological methods
advances, and within the last years has made very rapid progress
297indeed. The other modern Humanistic Sciences, in the first
place Sociology and Anthropology, by giving us a deeper understanding
of human nature and culture, bring their share to the
common problem. For the questions of language are indeed the
most important and central subject of all humanistic studies.
Thus, the Science of Language constantly receives contributions
of new material and stimulation from new methods. A most
important impetus which it has thus lately received has come from
the philosophical study of symbols and mathematical data, so
brilliantly carried on in Cambridge by Mr Bertrand Russell and
Dr Whitehead.

In the present book Mr Ogden and Mr Richards carry over
the study of signs into the field of linguistics, where it assumes a
fundamental importance. Indeed, they work out a new Science of
Symbolism which is sure to yield most valuable criteria for the
criticism of certain errors of Metaphysics and of purely Formal
Logic (cf. Chaps. II, VII, VIII and IX). On the other hand,
the theory has not merely a philosophical bearing, but possesses
practical importance in dealing with the special, purely scientific
problems of Meaning, Grammar, Psychology and Pathology of
Speech. More especially, important researches on Aphasia by
Dr Henry Head, which promise to throw entirely new light on
our conceptions of Meaning, seem to work towards the same Semantic
theories as those contained in the present book. 11 Dr
A. H. Gardiner, one of the greatest experts in hieroglyphic script
and Egyptian grammar — of which he is preparing a new analysis
— has published some remarkable articles on Meaning, where he
approaches the same problems as those discussed by Mr Ogden
and Mr Richards, and solved by them in such an interesting
manner, and their respective results do not seem to me to be
incompatible. 22 Finally, I myself, at grips with the problem of
primitive languages from Papuo-Melanesia, had been driven
into the field of general Semantics. 33 When, however, I had
the privilege of looking through the proofs of the present
book, I was astonished to find how exceedingly well the theories
there presented answered all my problems and solved my difficulties;
and I was gratified to find that the position to which I
298had been led by the study of primitive languages, was not essentially
a different one. I was therefore extremely glad when the
Authors offered me an opportunity to state my problems, and to
outline my tentative solutions, side by side with their remarkable
theories. I accepted it the more gladly because I hope to show
how important a light the theories of this book throw on the
problems of primitive languages.

It is remarkable that a number of independent inquirers, Messrs
Ogden and Richards, Dr Head, Dr Gardiner and myself, starting
from definite and concrete, yet quite different problems, should
arrive, if not exactly at the same results stated in the same terminology,
at least at the construction of similar Semantic theories
based on psychological considerations.

I have therefore to show how, in my own case, that of an
Ethnographer studying primitive mentality, culture, and language,
I was driven into a linguistic theory very much on lines parallel
to those of the present work. In the course of my Ethnographic
researches among some Melanesian tribes of Eastern New Guinea,
which I conducted exclusively by means of the local language,
I collected a considerable number of texts: magical formulæ,
items of folk-lore, narratives, fragments of conversation, and
statements of my informants. When, in working out this linguistic
material, I tried to translate my texts into English, and
incidentally to write out the vocabulary and grammar of the
language, I was faced by fundamental difficulties. These difficulties
were not removed, but rather increased, when I consulted
the extant grammars and vocabularies of Oceanic languages.
The authors of these, mainly missionaries who wrote for the
practical purpose of facilitating the task of their successors,
proceeded by rule of thumb. For instance, in writing a vocabulary
they would give the next best approximation in English
to a native word.

But the object of a scientific translation of a word is not to
give its rough equivalent, sufficient for practical purposes, but
to state exactly whether a native word corresponds to an idea at
least partially existing for English speakers, or whether it covers
an entirely foreign conception. That such foreign conceptions do
exist for native languages and in great number, is clear. All
words which describe the native social order, all expressions
referring to native beliefs, to specific customs, ceremonies, magical
rites — all such words are obviously absent from English as from
any European language. Such words can only be translated into
English, not by giving their imaginary equivalent — a real one
299obviously cannot be found — but by explaining the meaning of
each of them through an exact Ethnographic account of the
sociology, culture and tradition of that native community.

But there is an even more deeply reaching though subtler
difficulty: the whole manner in which a native language is used
is different from our own. In a primitive tongue, the whole
grammatical structure lacks the precision and definiteness of our
own, though it is extremely telling in certain specific ways. Again
some particles, quite untranslatable into English, give a special
flavour to native phraseology. In the structure of sentences,
an extreme simplicity hides a good deal of expressiveness, often
achieved by means of position and context. Returning to the
meaning of isolated words, the use of metaphor, the beginnings of
abstraction, of generalization and a vagueness associated with
extreme concreteness of expression — all these features baffle any
attempt at a simple and direct translation. The ethnographer
has to convey this deep yet subtle difference of language and of
the mental attitude which lies behind it, and is expressed through
it. But this leads more and more into the general psychological
problem of Meaning.


This general statement of the linguistic difficulties which beset
an Ethnographer in his field-work, must be illustrated by a
concrete example. Imagine yourself suddenly transported on
to a coral atoll in the Pacific, sitting in a circle of natives and
listening to their conversation. Let us assume further that there
is an ideal interpreter at hand, who, as far as possible, can convey
the meaning of each utterance, word for word, so that the listener
is in possession of all the linguistic data available. Would that
make you understand the conversation or even a single utterance?
Certainly not.

Let us have a look at such a text, an actual utterance taken down
from a conversation of natives in the Trobriand Islands, N.E. New
Guinea. In analysing it, we shall see quite plainly how helpless
one is in attempting to open up the meaning of a statement by
mere linguistic means; and we shall also be able to realize what
sort of additional knowledge, besides verbal equivalence, is
necessary in order to make the utterance significant.

I adduce a statement in native, giving under each word its
nearest English equivalent:

Tasakaulo kaymatana yakida;
We run front-wood ourselves;
300tawoulo ovanu; tasivila tagine
soda; isakaulo ka'u'uya
oluvieki similaveta Pilolu
we paddle in place; we turn we see
companion ours; he runs rear-wood
behind their sea-arm Pilolu

The verbatim English translation of this utterance sounds at
first like a riddle or a meaningless jumble of words; certainly not
like a significant, unambiguous statement. Now if the listener,
whom we suppose acquainted with the language, but unacquainted
with the culture of the natives, were to understand even the
general trend of this statement, he would have first to be informed
about the situation in which these words were spoken. He would
need to have them placed in their proper setting of native culture.
In this case, the utterance refers to an episode in an overseas
trading expedition of these natives, in which several canoes take
part in a competitive spirit. This last-mentioned feature explains
also the emotional nature of the utterance: it is not a mere statement
of fact, but a boast, a piece of self-glorification, extremely
characteristic of the Trobrianders' culture in general and of
their ceremonial barter in particular.

Only after a preliminary instruction is it possible to gain some
idea of such technical terms of boasting and emulation as kaymatana
(frontwood) and ka'u'uya (rear-wood). The metaphorical use
of wood for canoe would lead us into another field of language
psychology, but for the present it is enough to emphasize that
‘front’ or ‘leading canoe’ and ‘rear canoe’ are important
terms for a people whose attention is so highly occupied with
competitive activities for their own sake. To the meaning of such
words is added a specific emotional tinge, comprehensible only
against the background of their tribal psychology in ceremonial
life, commerce and enterprise.

Again, the sentence where the leading sailors are described as
looking back and perceiving their companions lagging behind on
the sea-arm of Pilolu, would require a special discussion of the
geographical feeling of the natives, of their use of imagery as a
linguistic instrument and of a special use of the possessive pronoun
(their sea-arm Pilolu).

All this shows the wide and complex considerations into which
we are led by an attempt to give an adequate analysis of meaning.
Instead of translating, of inserting simply an English word for a
native one, we are faced by a long and not altogether simple process
301of describing wide fields of custom, of social psychology
and of tribal organization which correspond to one term
or another. We see that linguistic analysis inevitably leads
us into the study of all the subjects covered by Ethnographic

Of course the above given comments on the specific terms
(front-wood, rear-wood, their sea-arm Pilolu) are necessarily
short and sketchy. But I have on purpose chosen an utterance
which corresponds to a set of customs, already described quite
fully. 14 The reader of that description will be able to understand
thoroughly the adduced text, as well as appreciate the present

Besides the difficulties encountered in the translation of single
words, difficulties which lead directly into descriptive Ethnography,
there are others, associated with more exclusively linguistic
problems, which however can be solved only on the basis
of psychological analysis. Thus it has been suggested that the
characteristically Oceanic distinction of inclusive and exclusive
pronouns requires a deeper explanation than any which would
confine itself to merely grammatical relations. 25 Again, the puzzling
manner in which some of the obviously correlated sentences
are joined in our text by mere juxtaposition would require much
more than a simple reference, if all its importance and significance
had to be brought out. Those two features are well known and
have been often discussed, though according to my ideas not
quite exhaustively.

There are, however, certain peculiarities of primitive languages,
almost entirely neglected by grammarians, yet opening up very
interesting questions of savage psychology. I shall illustrate this
by a point, lying on the borderland between grammar and lexicography
and well exemplified in the utterance quoted.

In the highly developed Indo-European languages, a sharp
distinction can be drawn between the grammatical and lexical
function of words. The meaning of a root of a word can be
isolated from the modification of meaning due to accidence or
some other grammatical means of determination. Thus in the
word run we distinguish between the meaning of the root — rapid
302personal displacement — and the modification as to time, tense,
definiteness, etc., expressed by the grammatical form, in which
the word is found in the given context. But in native languages
the distinction is by no means so clear and the functions of
grammar and radical meaning respectively are often confused in
a remarkable manner.

In the Melanesian languages there exist certain grammatical
instruments, used in the flection of verbs, which express somewhat
vaguely relations of time, definiteness and sequence. The most
obvious and easy thing to do for a European who wishes to use
roughly such a language for practical purposes, is to find out what
is the nearest approach to those Melanesian forms in our languages
and then to use the savage form in the European manner. In
the Trobriand language, for instance, from which we have taken
our above example, there is an adverbial particle boge, which,
put before a modified verb, gives it, in a somewhat vague manner,
the meaning either of a past or of a definite happening. The
verb is moreover modified by a change in the prefixed personal
pronoun. Thus the root ma (come, move hither) if used with
the prefixed pronoun of the third singular i — has the form ima
and means (roughly), he comes. With the modified pronoun ay
— or, more emphatical, lay — it means (roughly) he came or he
has come
. The expression boge ayna or boge layma can be approximately
translated by he has already come, the participle boge
making it more definite.

But this equivalence is only approximate, suitable for some
practical purposes, such as trading with the natives, missionary
preaching and translation of Christian literature into native
languages. This last cannot, in my opinion, be carried out with
any degree of accuracy. In the grammars and interpretations of
Melanesian languages, almost all of which have been written by
missionaries for practical purposes, the grammatical modifications
of verbs have been simply set down as equivalent to Indo-European
tenses. When I first began to use the Trobriand language in
my field-work, I was quite unaware that there might be some
snares in taking savage grammar at its face value and followed
the missionary way of using native inflection.

I had soon to learn, however, that this was not correct and I
learnt it by means of a practical mistake, which interfered slightly
with my field-work and forced me to grasp native flection at the
cost of my personal comfort. At one time I was engaged in
making observations on a very interesting transaction which took
place in a lagoon village of the Trobriands between the coastal
303fishermen and the inland gardeners. 16 I had to follow some important
preparations in the village and yet I did not want to miss
the arrival of the canoes on the beach. I was busy registering and
photographing the proceedings among the huts, when word went
round, ‘they have come already’ — boge laymayse. I left my
work in the village unfinished to rush some quarter of a mile to
the shore, in order to find, to my disappointment and mortification,
the canoes far away, punting slowly along towards the beach!
Thus I came some ten minutes too soon, just enough to make me
lose my opportunites in the village!

It required some time and a much better general grasp of the
language before I came to understand the nature of my mistake
and the proper use of words and forms to express the subtleties
of temporal sequence. Thus the root ma which means come,
move hither, does not contain the meaning, covered by our word
arrive. Nor does any grammatical determination give it the
special and temporal definition, which we express by, ‘they have
come, they have arrived.’ The form boge laymayse, which I heard
on that memorable morning in the lagoon village, means to a
native ‘they have already been moving hither’ and not ‘they have
already come here.’

In order to achieve the spatial and temporal definition which
we obtain by using the past definite tense, the natives have recourse
to certain concrete and specific expressions. Thus in the case
quoted, the villagers, in order to convey the fact that the canoes
had arrived, would have used the word to anchor, to moor.
‘They have already moored their canoes,’ boge aykotasi, would
have meant, what I assumed they had expressed by boge laymayse.
That is, in this case the natives use a different root instead of a
mere grammatical modification.

Returning to our text, we have another telling example of the
characteristic under discussion. The quaint expression ‘we
paddle in place’ can only be properly understood by realizing
that the word paddle has here the function, not of describing
what the crew are doing, but of indicating their immediate
proximity to the village of their destination. Exactly as in the
previous example the past tense of the word to come (‘they have
come’) which we would have used in our language to convey the
fact of arrival, has another meaning in native and has to be
replaced by another root which expresses the idea; so here the
native root wa, to move thither, could not have been used in
304(approximately) past definite tense to convey the meaning of
‘arrive there,’ but a special root expressing the concrete act of
paddling is used to mark the spatial and temporal relations of
the leading canoe to the others. The origin of this imagery is
obvious. Whenever the natives arrive near the shore of one of
the overseas villages, they have to fold the sail and to use the
paddles, since there the water is deep, even quite close to the
shore, and punting impossible. So ‘to paddle’ means ‘to
arrive at the overseas village.’ It may be added that in this
expression ‘we paddle in place,’ the two remaining words in
and place would have to be retranslated in a free English interpretation
by near the village.

With the help of such an analysis as the one just given, this or
any other savage utterance can be made comprehensible. In this
case we may sum up our results and embody them in a free
commentary or paraphrase of the statement:

A number of natives sit together. One of them, who has just
come back from an overseas expedition, gives an account of the
sailing and boasts about the superiority of his canoe. He tells
his audience how, in crossing the sea-arm of Pilolu (between the
Trobriands and the Amphletts), his canoe sailed ahead of all
others. When nearing their destination, the leading sailors
looked back and saw their comrades far behind, still on the sea-arm
of Pilolu.

Put in these terms, the utterance can at least be understood
broadly, though for an exact appreciation of the shades and details
of meaning a full knowledge of the native customs and psychology,
as well as of the general structure of their language, is indispensable.

It is hardly necessary perhaps to point out that all I have said
in this section is only an illustration on a concrete example of the
general principles so brilliantly set forth by Ogden and Richards
in Chapters I, III and IV of their work. What I have tried to
make clear by analysis of a primitive linguistic text is that language
is essentially rooted in the reality of the culture, the tribal life and
customs of a people, and that it cannot be explained without
constant reference to these broader contexts of verbal utterance.
The theories embodied in Ogden's and Richards' diagram of
Chapter I, in their treatment of the ‘sign-situation’ (Chapter
III) and in their analysis of perception (Chapter IV) cover and
generalize all the details of my example.305


Returning once more to our native utterance, it needs no
special stressing that in a primitive language the meaning of any
single word is to a very high degree dependent on its context.
The words ‘wood’, ‘paddle’, ‘place’ had to be retranslated in
the free interpretation in order to show what is their real meaning,
conveyed to a native by the context in which they appear. Again,
it is equally clear that the meaning of the expression ‘we arrive
near the village (of our destination)’ literally: ‘we paddle in
place’, is determined only by taking it in the context of the whole
utterance. This latter again, becomes only intelligible when it is
placed within its context of situation, if I may be allowed to coin
an expression which indicates on the one hand that the conception
of context has to be broadened and on the other that the situation
in which words are uttered can never be passed over as irrelevant
to the linguistic expression. We see how the conception of context
must be substantially widened, if it is to furnish us with its
full utility. In fact it must burst the bonds of mere linguistics
and be carried over into the analysis of the general conditions
under which a language is spoken. Thus, starting from the wider
idea of context, we arrive once more at the results of the foregoing
section, namely that the study of any language, spoken by a
people who live under conditions different from our own and
possess a different culture, must be carried out in conjunction
with the study of their culture and of their environment.

But the widened conception of context of situation yields more
than that. It makes clear the difference in scope and method
between the linguistics of dead and of living languages. The
material on which almost all our linguistic study has been done
so far belongs to dead languages. It is present in the form of
written documents, naturally isolated, torn out of any context of
. In fact, written statements are set down with the purpose
of being self-contained and self-explanatory. A mortuary
inscription, a fragment of primeval laws or precepts, a chapter or
statement in a sacred book, or to take a more modern example,
a passage from a Greek or Latin philosopher, historian or poet —
one and all of these were composed with the purpose of bringing
their message to posterity unaided, and they had to contain this
message within their own bounds.

To take the clearest case, that of a modern scientific book, the
writer of it sets out to address every individual reader who will
peruse the book and has the necessary scientific training. He
306tries to influence his reader's mind in certain directions. With
the printed text of the book before him, the reader, at the writer's
bidding, undergoes a series of processes — he reasons, reflects,
remembers, imagines. The book by itself is sufficient to direct
the reader's mind to its meaning; and we might be tempted to
say metaphorically that the meaning is wholly contained in or
carried by the book.

But when we pass from a modern civilized language, of which
we think mostly in terms of written records, or from a dead one
which survives only in inscription, to a primitive tongue, never
used in writing, where all the material lives only in winged words,
passing from man to man — there it should be clear at once that
the conception of meaning as contained in an utterance is false and
futile. A statement, spoken in real life, is never detached from
the situation in which it has been uttered. For each verbal
statement by a human being has the aim and function of expressing
some thought or feeling actual at that moment and in that
situation, and necessary for some reason or other to be made
known to another person or persons — in order either to serve
purposes of common action, or to establish ties of purely social
communion, or else to deliver the speaker of violent feelings or
passions. Without some imperative stimulus of the moment,
there can be no spoken statement. In each case, therefore,
utterance and situation are bound up inextricably with each
other and the context of situation is indispensable for the understanding
of the words. Exactly as in the reality of spoken or
written languages, a word without linguistic context is a mere
figment and stands for nothing by itself, so in the reality of a
spoken living tongue, the utterance has no meaning except in the
context of situation.

It will be quite clear now that the point of view of the Philologist,
who deals only with remnants of dead languages, must
differ from that of the Ethnographer, who, deprived of the ossified,
fixed data of inscriptions, has to rely on the living reality
of spoken language in fluxu. The former has to reconstruct the
general situation — i.e., the culture of a past people — from the
extant statements, the latter can study directly the conditions
and situations characteristic of a culture and interpret the
statements through them. Now I claim that the Ethnographer's
perspective is the one relevant and real for the formation of
fundamental linguistic conceptions and for the study of the life
of languages, whereas the Philologist's point of view is fictitious
and irrelevant. For language in its origins has been merely the
307free, spoken sum total of utterances such as we find now in a savage
tongue. All the foundations and fundamental characteristics of
human speech have received their shape and character in the
stage of development proper to Ethnographic study and not in
the Philologist's domain. To define Meaning, to explain the
essential grammatical and lexical characters of language on the
material furnished by the study of dead languages, is nothing
short of preposterous in the light of our argument. Yet it would
be hardly an exaggeration to say that 99 per cent of all linguistic
work has been inspired by the study of dead languages or at best
of written records torn completely out of any context of situation.
That the Ethnographer's perspective can yield not only generalities
but positive, concrete conclusions I shall indicate at least in
the following sections.

Here I wish again to compare the standpoint just reached with
the results of Messrs Ogden and Richards. I have written the
above in my own terminology, in order to retrace the steps of my
argument, such as it was before I became acquainted with the
present book. But it is obvious that the context of situation, on
which such a stress is laid here, is nothing else but the sign-situation
of the Authors. Their contention, which is fundamental to all the
arguments of their book, that no theory of meaning can be given
without the study of the mechanism of reference, is also the main
gist of my reasoning in the foregoing paragraphs. The opening
chapters of their work show how erroneous it is to consider Meaning
as a real entity, contained in a word or utterance. The ethnographically
and historically interesting data and comments of
Chapter II show up the manifold illusions and errors due to a
false attitude towards words. This attitude in which the word
is regarded as a real entity, containing its meaning as a Soul-box
contains the spiritual part of a person or thing, is shown to be
derived from the primitive, magical uses of language and to
reach right into the most important and influential systems of
metaphysics. Meaning, the real ‘essence’ of a word, achieves
thus Real Existence in Plato's realm of Ideas; and it becomes the
Universal, actually existing, of mediæval Realists. The misuse
of words, based always on a false analysis of their Semantic
function, leads to all the ontological morass in philosophy, where
truth is found by spinning out meaning from the word, its assumed

The analysis of meaning in primitive languages affords a
striking confirmation of Messrs Ogden and Richards' theories.
For the clear realization of the intimate connection between linguistic
308interpretation and the analysis of the culture to which the
language belongs, shows convincingly that neither a Word nor its
Meaning has an independent and self-sufficient existence. The
Ethnographic view of language proves the principle of Symbolic
Relativity as it might be called, that is that words must be treated
only as symbols and that a psychology of symbolic reference must
serve as the basis for all science of language. Since the whole
world of ‘things-to-be-expressed’ changes with the level of
culture, with geographical, social and economic conditions, the
consequence is that the meaning of a word must be always
gathered, not from a passive contemplation of this word, but from
an analysis of its functions, with reference to the given culture.
Each primitive or barbarous tribe, as well as each type of civilization,
has its world of meanings and the whole linguistic apparatus
of this people — their store of words and their type of grammar — can
only be explained in connection with their mental requirements.

In Chapter III of this book the Authors give an analysis of
the psychology of symbolic reference, which together with the
material collected in Chapter II is the most satisfactory treatment
of the subject which I have ever seen. I wish to remark that the
use of the word ‘context’ by the Authors is compatible, but not
identical, with my use of this word in the expression ‘context of
situation.’ I cannot enter here into an attempt to bring our
respective nomenclature into line and must allow the reader to
test the Relativity of Symbolism on this little example.


So far, I have dealt mainly with the simplest problems of
meaning, those associated with the definition of single words
and with the lexicographical task of bringing home to a European
reader the vocabulary of a strange tongue. And the main result
of our analysis was that it is impossible to translate words of a
primitive language or of one widely different from our own,
without giving a detailed account of the culture of its users and
thus providing the common measure necessary for a translation.
But though an Ethnographic background is indispensable for
a scientific treatment of a language, it is by no means sufficient,
and the problem of Meaning needs a special theory of its own.
I shall try to show that, looking at language from the Ethnographic
perspective and using our conception of context of situation,
we shall be able to give an outline of a Semantic theory,
309useful in the work on Primitive Linguistics, and throwing some
light on human language in general.

First of all, let us try, from our standpoint, to form a view of
the Nature of language. The lack of a clear and precise view of
Linguistic function and of the nature of Meaning, has been, I believe,
the cause of the relative sterility of much otherwise excellent
linguistic theorizing. The direct manner in which the Authors
face this fundamental problem and the excellent argument by
which they solve it, constitute the permanent value of their work.

The study of the above-quoted native text has demonstrated
that an utterance becomes comprehensive only when we interpret
it by its context of situation. The analysis of this context should
give us a glimpse of a group of savages bound by reciprocal ties
of interests and ambitions, of emotional appeal and response.
There was boastful reference to competitive trading activities,
to ceremonial overseas expeditions, to a complex of sentiments,
ambitions and ideas known to the group of speakers and hearers
through their being steeped in tribal tradition and having been
themselves actors in such events as those described in the narrative.
Instead of giving a narrative I could have adduced linguistic
samples still more deeply and directly embedded in the
context of situation.

Take for instance language spoken by a group of natives
engaged in one of their fundamental pursuits in search of subsistence — hunting,
fishing, tilling the soil; or else in one of those
activities, in which a savage tribe express some essentially human
forms of energy — war, play or sport, ceremonial performance or
artistic display such as dancing or singing. The actors in any
such scene are all following a purposeful activity, are all set
on a definite aim; they all have to act in a concerted manner
according to certain rules established by custom and tradition.
In this, Speech is the necessary means of communion; it is the
one indispensable instrument for creating the ties of the moment
without which unified social action is impossible.

Let us now consider what would be the type of talk passing
between people thus acting, what would be the manner of its
use. To make it quite concrete at first, let us follow up a party of
fishermen on a coral lagoon, spying for a shoal of fish, trying to
imprison them in an enclosure of large nets, and to drive them
into small net-bags — an example which I am choosing also
because of my personal familiarity with the procedure. 17310

The canoes glide slowly and noiselessly, punted by men especially
good at this task and always used for it. Other experts who
know the bottom of the lagoon, with its plant and animal life,
are on the look-out for fish. One of them sights the quarry.
Customary signs, or sounds or words are uttered. Sometimes a
sentence full of technical references to the channels or patches on
the lagoon has to be spoken; sometimes when the shoal is near
and the task of trapping is simple, a conventional cry is uttered
not too loudly. Then, the whole fleet stops and ranges itself —
every canoe and every man in it performing his appointed task —
according to a customary routine. But, of course, the men, as
they act, utter now and then a sound expressing keenness in the
pursuit or impatience at some technical difficulty, joy of achievement
or disappointment at failure. Again, a word of command
is passed here and there, a technical expression or explanation
which serves to harmonise their behaviour towards other men.
The whole group act in a concerted manner, determined by old
tribal tradition and perfectly familiar to the actors through lifelong
experience. Some men in the canoes cast the wide encircling
nets into the water, others plunge, and wading through the shallow
lagoon, drive the fish into the nets. Others again stand by with
the small nets, ready to catch the fish. An animated scene, full
of movement follows, and now that the fish are in their power the
fishermen speak loudly, and give vent to their feelings. Short,
telling exclamations fly about, which might be rendered by such
words as: ‘Pull in,’ ‘Let go,’ ‘Shift further,’ ‘Lift the net’; or again
technical expressions completely untranslatable except by minute
description of the instruments used, and of the mode of action.

All the language used during such a pursuit is full of technical
terms, short references to surroundings, rapid indications of
change — all based on customary types of behaviour, well-known
to the participants from personal experience. Each utterance is
essentially bound up with the context of situation and with the
aim of the pursuit, whether it be the short indications about the
movements of the quarry, or references to statements about the
surroundings, or the expression of feeling and passion inexorably
bound up with behaviour, or words of command, or correlation
of action. The structure of all this linguistic material is inextricably
mixed up with, and dependent upon, the course of the
activity in which the utterances are embedded. The vocabulary,
the meaning of the particular words used in their characteristic
technicality is not less subordinate to action. For technical
language, in matters of practical pursuit, acquires its meaning
311only through personal participation in this type of pursuit. It
has to be learned, not through reflection but through action.

Had we taken any other example than fishing, we would have
reached similar results. The study of any form of speech used
in connection with vital work would reveal the same grammatical
and lexical peculiarities: the dependence of the meaning of each
word upon practical experience, and of the structure of each
utterance upon the momentary situation in which it is spoken.
Thus the consideration of linguistic uses associated with any
practical pursuit, leads us to the conclusion that language in its
primitive forms ought to be regarded and studied against the
background of human activities and as a mode of human behaviour
in practical matters. We have to realize that language originally,
among primitive, non-civilized peoples was never used as a mere
mirror of reflected thought. The manner in which I am using it
now, in writing these words, the manner in which the author of
a book, or a papyrus or a hewn inscription has to use it, is a very
far-fetched and derivative function of language. In this, language
becomes a condensed piece of reflection, a record of fact or thought.
In its primitive uses, language functions as a link in concerted
human activity, as a piece of human behaviour. It is a mode of
action and not an instrument of reflection.

These conclusions have been reached on an example in which
language is used by people engaged in practical work, in which
utterances are embedded in action. This conclusion might be
questioned by an objection that there are also other linguistic
uses even among primitive peoples who are debarred from writing
or any means of external fixation of linguistic texts. Yet even they,
it might be urged, have fixed texts in their songs, sayings, myths
and legends, and most important, in their ritual and magical
formulæ. Are our conclusions about the nature of language
correct, when faced with this use of speech; can our views
remain unaltered when, from speech in action, we turn our
attention to free narrative or to the use of language in pure social
intercourse; when the object of talk is not to achieve some aim
but the exchange of words almost as an end in itself?

Anyone who has followed our analysis of speech in action and
compares it with the discussion of the narrative texts in Section
II, will be convinced that the present conclusions apply to narrative
speech as well. When incidents are told or discussed among
a group of listeners, there is, first, the situation of that moment
made up of the respective social, intellectual and emotional
attitudes of those present. Within this situation, the narrative
312creates new bonds and sentiments by the emotional appeal of the
words. In the narrative quoted, the boasting of a man to a mixed
audience of several visitors and strangers produces feelings of
pride or mortification, of triumph or envy. In every case,
narrative speech as found in primitive communities is primarily
a mode of social action rather than a mere reflection of thought.

A narrative is associated also indirectly with one situation to
which it refers — in our text with a performance of competitive
sailing. In this relation, the words of a tale are significant because
of previous experiences of the listeners; and their meaning depends
on the context of the situation referred to, not to the same degree
but in the same manner as in the speech of action. The difference
in degree is important; narrative speech is derived in its function,
and it refers to action only indirectly, but the way in which it
acquires its meaning can only be understood from the direct
function of speech in action. To use the terminology of this
work: the referential function of a narrative is subordinate to
its social and emotive function, as classified by the Authors in
Chapter X.

The case of language used in free, aimless, social intercourse
requires special consideration. When a number of people sit
together at a village fire, after all the daily tasks are over, or when
they chat, resting from work, or when they accompany some
mere manual work by gossip quite unconnected with what they
are doing — it is clear that here we have to do with another mode of
using language, with another type of speech function. Language
here is not dependent upon what happens at that moment, it
seems to be even deprived of any context of situation. The
meaning of any utterance cannot be connected with the speaker's
or hearer's behaviour, with the purpose of what they are doing.

A mere phrase of politeness, in use as much among savage
tribes as in a European drawing-room, fulfils a function to
which the meaning of its words is almost completely irrelevant.
Inquiries about health, comments on weather, affirmations of
some supremely obvious state of things — all such are exchanged,
not in order to inform, not in this case to connect people in action,
certainly not in order to express any thought. It would be even
incorrect, I think, to say that such words serve the purpose of
establishing a common sentiment, for this is usually absent from
such current phrases of intercourse; and where it purports to
exist, as in expressions of sympathy, it is avowedly spurious on
one side. What is the raison d'être, therefore, of such phrases as
‘How do you do?’ ‘Ah, here you are,’ ‘Where do you come
313from?’ ‘Nice day to-day’ — all of which serve in one society or
another as formulæ of greeting or approach?

I think that, in discussing the function of Speech in mere
sociabilities, we come to one of the bedrock aspects of man's
nature in society. There is in all human beings the well-known
tendency to congregate, to be together, to enjoy each other's
company. Many instincts and innate trends, such as fear or
pugnacity, all the types of social sentiments such as ambition,
vanity, passion for power and wealth, are dependent upon and
associated with the fundamental tendency which makes the mere
presence of others a necessity for man. 18

Now speech is the intimate correlate of this tendency, for, to
a natural man, another man's silence is not a reassuring factor,
but, on the contrary, something alarming and dangerous. The
stranger who cannot speak the language is to all savage tribesmen
a natural enemy. To the primitive mind, whether among savages
or our own uneducated classes, taciturnity means not only unfriendliness
but directly a bad character. This no doubt varies
greatly with the national character but remains true as a general
rule. The breaking of silence, the communion of words is the
first act to establish links of fellowship, which is consummated
only by the breaking of bread and the communion of food.
The modern English expression, ‘Nice day to-day’ or the Melanesian
phrase, ‘Whence comest thou?’ are needed to get over
the strange and unpleasant tension which men feel when facing
each other in silence.

After the first formula, there comes a flow of language, purposeless
expressions of preference or aversion, accounts of irrelevant
happenings, comments on what is perfectly obvious. Such
gossip, as found in Primitive Societies, differs only a little from
our own. Always the same emphasis of affirmation and consent,
mixed perhaps with an incidental disagreement which creates
the bonds of antipathy. Or personal accounts of the speaker's
views and life history, to which the hearer listens under some
restraint and with slightly veiled impatience, waiting till his own
turn arrives to speak. For in this use of speech the bonds created
between hearer and speaker are not quite symmetrical, the man
linguistically active receiving the greater share of social pleasure
and self-enhancement. But though the hearing given to such
314utterances is as a rule not as intense as the speaker's own share,
it is quite essential for his pleasure, and the reciprocity is established
by the change of rôles.

There can be no doubt that we have here a new type of linguistic
use — photic communion I am tempted to call it, actuated by the
demon of terminological invention — a type of speech in which
ties of union are created by a mere exchange of words. Let us
look at it from the special point of view with which we are here
concerned; let us ask what light it throws on the function or
nature of language. Are words in Phatic Communion used
primarily to convey meaning, the meaning which is symbolically
theirs? Certainly not! They fulfil a social function and that
is their principal aim, but they are neither the result of intellectual
reflection, nor do they necessarily arouse reflection in the listener.
Once again we may say that language does not function here as a
means of transmission of thought.

But can we regard it as a mode of action? And in what relation
does it stand to our crucial conception of context of situation?
It is obvious that the outer situation does not enter directly into
the technique of speaking. But what can be considered as situation
when a number of people aimlessly gossip together? It consists
in just this atmosphere of sociability and in the fact of the personal
communion of these people. But this is in fact achieved by
speech, and the situation in all such cases is created by the
exchange of words, by the specific feelings which form convivial
gregariousness, by the give and take of utterances which make up
ordinary gossip. The whole situation consists in what happens
linguistically. Each utterance is an act serving the direct aim of
binding hearer to speaker by a tie of some social sentiment or
other. Once more language appears to us in this function not as
an instrument of reflection but as a mode of action.

I should like to add at once that though the examples discussed
were taken from savage life, we could find among ourselves exact
parallels to every type of linguistic use so far discussed. The
binding tissue of words which unites the crew of a ship in bad
weather, the verbal concomitants of a company of soldiers in
action, the technical language running parallel to some practical
work or sporting pursuit — all these resemble essentially the primitive
uses of speech by man in action and our discussion could have
been equally well conducted on a modern example. I have chosen
the above from a Savage Community, because I wanted to emphasize
that such and no other is the nature of primitive speech.

Again in pure sociabilities and gossip we use language exactly
315as savages do and our talk becomes the ‘phatic communion’
analysed above, which serves to establish bonds of personal
union between people brought together by the mere need of
companionship and does not serve any purpose of communicating
ideas. “Throughout the Western world it is agreed that people
must meet frequently, and that it is not only agreeable to talk,
but that it is a matter of common courtesy to say something even
when there is hardly anything to say” 19 — as the Authors remark.
Indeed there need not or perhaps even there must not be anything
to communicate. As long as there are words to exchange,
phatic communion brings savage and civilized alike into the
pleasant atmosphere of polite, social intercourse.

It is only in certain very special uses among a civilized community
and only in its highest uses that language is employed
to frame and express thoughts. In poetic and literary production,
language is made to embody human feelings and passions, to
render in a subtle and convincing manner certain inner states
and processes of mind. In works of science and philosophy,
highly developed types of speech are used to control ideas and to
make them common property of civilized mankind.

Even in this function, however, it is not correct to regard
language as a mere residuum of reflective thought. And the
conception of speech as serving to translate the inner processes of
the speaker to the hearer is one-sided and gives us, even with
regard to the most highly developed and specialized uses of speech,
only a partial and certainly not the most relevant view.

To restate the main position arrived at in this section we can
say that language in its primitive function and original form has
an essentially pragmatic character; that it is a mode of behaviour,
an indispensable element of concerted human action. And
negatively: that to regard it as a means for the embodiment or
expression of thought is to take a one-sided view of one of its
most derivate and specialized functions.


This view of the nature of language I have tried to establish
by a detailed analysis of examples, by reference to concrete and
actual facts. I trust therefore that the distinction which I have
explained, between ‘mode of action’ and ‘means of thinking,’
will not remain an empty phrase, but that it has received its content
from the adduced facts. Nothing, however, establishes the
316positive value and empirical nature of a general principle so
completely as when it is shown to work in the solution of definite
problems of a somewhat difficult and puzzling description.

In linguistics we have an intractable subject of this kind in the
Problem of Meaning. It would perhaps be presumptuous for
me to tackle this subject in an abstract and general manner and
with any philosophical ambition, after it has been shown by
Ogden and Richards (Chapters VIII and IX) to be of so highly
dangerous a nature. But I simply want to approach it through
the narrow avenue of Ethnographic empiricism and show how
it looks viewed from the perspective of the pragmatic uses of
primitive speech.

This perspective has allowed us to class human speech with
the active modes of human behaviour, rather than with the
reflective and cognitive ones. But this outside view and wholesale
conception must be still supplemented by some more detailed,
analytic considerations, if we want to arrive at a clearer idea of

In Chapter III of the present work the Authors discuss the
psychology of Sign-situations and the acquisition of significance
by symbols. I need not repeat or summarize their penetrating
analysis, which to me is extremely convincing and satisfactory
and forms the corner-stone of their linguistic theory. I wish
however to follow up one point in their argument, a point closely
related to our pragmatic conception of language.

The Authors reject, and rightly so, the explanations of meaning
by suggestion, association or apperception, urging that such
explanations are not sufficiently dynamic. Of course new ideas
are formed by apperception and since a new idea constitutes a
new meaning and receives in due course a new name, apperception
is a process by which significance is created. But that
happens only in the most highly developed and refined uses of
language for scientific purposes. From our previous discussion
it should be well established that such a type of formulation of
meaning is highly derivative and cannot be taken as the pattern
on which to study and explain significance. And this not only
with reference to savages, but also in our own linguistic life.
For a man who uses his language scientifically has his attitude
towards language already developed by and rooted in the more
elementary forms of word-function. Before he has ever begun
to acquire his scientific vocabulary in a highly artificial manner
by apperception — which, moreover, takes place only to a very
limited degree — he has learnt to use, used and grown up using
317words and constructions, the meaning of which has been formed
in his mind in quite a different manner. And this manner is
primary as regards time, for it is derived from earlier uses; it is
more general, because the vast majority of words thus receive
their meaning; and it is more fundamental, since it refers to
the most important and prevalent uses of speech — those which we
have indicated above as common to primitive and civilized

This manner of formation of meaning we must now proceed
to analyse more in detail, with reference to our pragmatic view of
language. And it will be best done by genetic considerations,
by an analysis of infantile uses of words, of primitive forms, of
significance and of pre-scientific language among ourselves.
Some glimpses of formation of meaning in infancy and childhood
will appear the more important, as modern psychology seems to
be more and more inclined to assign a permanent influence to
early mental habits in the outlook of the adult.

The emission of inarticulate emotional sound and of articulate
speech is a biological arrangement of enormous importance to the
young and adult of the human species, and is rooted deeply in
the instinctive and physiological arrangement of the human
organism. Children, savages and civilized adults alike react with
vocal expression to certain situations — whether these arouse
bodily pain or mental anguish, fear or passion, intense curiosity
or powerful joy. These sound reactions are part of the human
expression of emotions and as such possess, as has been established
by Darwin and others, a survival value or are at least themselves
relics of such values. Anyone in contact with infants and small
children knows that they express without the slightest ambiguity
their mood, their emotion, their need and desire. Concentrating
our attention for the moment on infantile utterances of this
type, it can be said that each sound is the expression of
some emotional state; that for surrounding people it has a
certain significance; and that it is correlated with the outer
situation surrounding and comprising the child's organism — a
situation which makes the child hungry or afraid or pleased or

All this is true of the non-articulate sounds emitted by an
infant, such as gurgling, wailing, squealing, crowing and weeping.
Later on, certain slightly articulated utterances follow, first
syllables — gu, ma, ba, etc. — repeated indefinitely, mixed up and
blurred by other sounds. These sounds serve in a parallel
manner to express certain psycho-physiological states and to
318expend some of the child's energy. They are a sign of health
and they are a form of indispensable exercise. Emission of sounds
is at the earliest and at the later stage of verbal development,
one of the child's main activities, persistent and passionate, as
every parent knows from pleasant and unpleasant experiences

How shall we conceive the formation of meaning at these
earliest stages? Here, in this somewhat different approach, the
pragmatic view of language obtrudes itself again. The child acts
by sound at this stage, and acts in a manner which is both adapted
to the outer situation, to the child's mental state and which is
also intelligible to the surrounding adults. Thus the significance
of sound, the meaning of an utterance is here identical with the
active response to surroundings and with the natural expression
of emotions. The meaning of such a sound is derived from one
of the earliest and most important forms of human activity.

When sound begins to articulate, the child's mind develops
in a parallel manner and becomes interested in isolating objects
from its surroundings, though the most relevant elements,
associated with the food and comfort of the infant, have been
already singled out previously. At the same time, the child
becomes aware of the sounds produced by the adults and the
other children of its surroundings, and it develops a tendency to
imitate them. The existence of a social milieu surrounding the
child is a factor of fundamental biological importance in the
upbringing of the human young and it is also, an indispensable
element in speech formation. Thus the child who begins to
articulate certain syllables soon finds these syllables repeated by
the adults and this paves the way to a clearer, more articulate

It would be extremely interesting to find out, whether and how
far some of the earliest articulated sounds have a ‘natural’
meaning, that is a meaning based on some natural connection
between sound and object. The only fact here relevant I can
quote from personal observation. I have noticed in two children
that at the stage where distinct syllables begin to be formed the
repeated sound, ma, ma, ma… appears when the child is dissatisfied
generally, when some essential want is not fulfilled or
some general discomfort is oppressing it. The sound attracts the
most important object in its surroundings, the mother, and with
her appearance the painful state of mind is remedied. Can it be
that the entry of the sound mama… just at the stage when
articulate speech begins — with its emotional significance and its
319power of bringing the mother to the rescue — has produced in a
great number of human languages the root ma for mother? 110

However this might be, and whether the child acquires some
of its early vocabulary by a spontaneous process or whether all
its words come to it from the outside, the manner in which the
first items of articulate speech are used is. the point which is
really interesting and relevant for us in this connection.

The earliest words — mama, dada, or papa, expressions for
food, water, certain toys or animals — are not simply imitated
and used to describe, name, or identify. Like the previous non-articulate
expressions of emotion, these early words also come to
be used under the stress of painful situations or strong emotions,
when the child cries for its parent or rejoices in her sight, when
it clamours for food or repeats with pleasure or excitement the
name of some favourite plaything of its surroundings. Here
the word becomes the significant reaction, adjusted to situation,
expressive of inner state and intelligible to the human milieu.

This latter fact has another very important set of consequences.
The human infant, helpless in itself and unable to cope with the
difficulties and dangers of its early life, is endowed with very
complete arrangements for care and assistance, resulting from the
instinctive attachment of the mother and, to a smaller extent, of
the father. The child's action on the surrounding world is done
through the parents, on whom the child acts again by its appeal,
mainly its verbal appeal. When the child clamours for a person,
it calls and he appears before it. When it wants food or an object
or when it wishes some uncomfortable thing or arrangement to
be removed, its only means of action is to clamour, and a very
efficient means of action this proves to the child.

To the child, words are therefore not only means of expression
but efficient modes of action. The name of a person uttered aloud
in a piteous voice possesses the power of materializing this person.
Food has to be called for and it appears — in the majority of cases.
Thus infantile experience must leave on the child's mind the
deep impression that a name has the power over the person or
thing which it signifies.320

We find thus that an arrangement biologically essential to the
human race makes the early articulated words sent forth by
children produce the very effect which these words mean. Words
are to a child active forces, they give him an essential hold on
reality, they provide him with the only effective means of moving,
attracting and repulsing outer things and of producing changes
in all that is relevant. This of course is not the statement of a
child's conscious views about language, but it is the attitude
implied in the child's behaviour.

Following the manner in which speech is used into the
later stage of childhood, we find again that everything reinforces
this pragmatic relation to meaning. In all the child's experience,
words mean, in so far as they act and not in so far as they make the
child understand or apperceive. His joy in using words and in
expressing itself in frequent repetition, or in playing about with
a word, is relevant in so far as it reveals the active nature of early
linguistic use. And it would be incorrect to say that such a
playful use of words is ‘meaningless.’ It is certainly deprived of
any intellectual purpose, but possesses always an emotional value,
and it is one of the child's favourite actions, in which he approaches
this or that person or object of his surroundings. When a child
greets the approaching person or animal, item of food or toy,
with a volley of the repeated name, he establishes a link of liking
or disliking between himself and that object. And all the time,
up to a fairly advanced age, the name of an object is the first means
recurred to, in order to attract, to materialize this thing.

If we transfer now this analysis to conditions of primitive
mankind, it will be better not to indulge in essentially imaginary
and therefore futile speculations about the beginnings of speech,
but simply to cast a glance at the normal uses of language as we
see them in empirical observations of savages. Returning to the
above examples of a group of natives engaged in a practical
pursuit, we see them using technical words, names of implements,
specific activities. A word, signifying an important utensil, is
used in action, not to comment on its nature or reflect on its
properties, but to make it appear, be handed over to the speaker,
or to direct another man to its proper use. The meaning of
the thing is made up of experiences of its active uses and not of
intellectual contemplation. Thus, when a savage learns to understand
the meaning of a word, this process is not accomplished by
explanations, by a series of acts of apperception, but by learning
to handle it. A word means to a native the proper use of the thing
for which it stands, exactly as an implement means something
321when it can be handled and means nothing when no active
experience of it is at hand. Similarly a verb, a word for an action,
receives its meaning through an active participation in this action.
A word is used when it can produce an action and not to describe
one, still less to translate thoughts. The word therefore has a
power of its own, it is a means of bringing things about, it is a
handle to acts and objects and not a definition of them.

Again, the same view of meaning results from the active uses
of speech among ourselves, even among those of us, who, on
comparatively rare occasions, can use language in a scientific or
literary manner. The innumerable superstitions — the agnostic's
fear of blasphemy or at least reluctance to use it, the active dislike
of obscene language, the power of swearing — all this shows that
in the normal use of words the bond between symbol and referent
is more than a mere convention.

The illiterate members of civilized communities treat and
regard words very much as savages do, that is as being strongly
bound up with the reality of action. And the way in which they
value verbal knowledge — proverbs, sayings, and, nowadays, news
— as the only form of wisdom, gives a definite character to this
implied attitude. But here I encroach on a field amply illustrated
and analysed in this book.

Indeed, on anyone who has read the brilliant chapters of Ogden
and Richards and grasped the main trend of their argument, it
will have dawned before now that all the argument of this Section
is a sort of foot-note to their fundamental contention that the
primitive, magical attitude towards words is responsible for a good
deal in the general use and abuse of language, more especially in
philosophical speculation. By the rich material cited in Chapter
II, and in Word Magic, by the examples of Chapters VII, VIII,
and IX, and by much of what is incidentally said, we are made to
realize how deeply rooted is the belief that a word has some power
over a thing, that it participates of the nature of the thing, that
it is akin or even identical in its contained ‘meaning’ with the
thing or with its prototype.

But whence is this magical attitude derived? Here the study
of the early stages of speech steps in helpfully and the Ethnographer
can make himself useful to the Philosopher of Language.
In studying the infantile formation of meaning and the savage
or illiterate meaning, we found this very magical attitude towards
words. The word gives power, allows one to exercise an influence
over an object or an action. The meaning of a word arises out
of familiarity, out of ability to use, out of the faculty of direct
322clamouring as with the infant, or practically directing as with
primitive man. A word is used always in direct active conjunction
with the reality it means. The word acts on the thing and the
thing releases the word in the human mind. This indeed is
nothing more or less than the essence of the theory which underlies
the use of verbal magic. And this theory we find based on
real psychological experiences in primitive forms of speech.

Before the earliest philosophical speculation sets in, there
emerges the practice and theory of magic, and in this, man's
natural attitude towards words becomes fixed and formulated by
a special lore and tradition. It is through the study of actual
spells and verbal magic as well as by the analysis of savage ideas
on magic that we can best understand this developed traditional
view of the secret power of appropriate words on certain things.
Briefly it may be said that such study simply confirms our theoretical
analysis of this section. In magical formulæ we find a
preponderance of words with high emotional tension, of technical
terms, of strong imperatives, of verbs expressing hope, success,
achievement. So much must suffice here and the reader is referred
for more data to Chapter II of this book, and to the chapters on
‘Magic’ and ‘The Power of Words in Magic’ in the above
quoted work of mine. 111

It may be of interest to interpret the results of our analysis of
the earliest stages of meaning on the diagram in which the
relations between Symbol, Act of Thought, and Referent are
represented by a triangle at the beginning of Chapter I of this
book. This diagram represents very adequately the said relations
in the developed uses of speech. It is characteristic in this triangle
that the base, indicated by a dotted line, represents the
imputed relation which obtains between a Symbol and the thing
it refers to, its Referent as the Authors name it. In developed
functions of speech, such as are, or at least should be, used in
philosophical speculation or scientific language (and it is chiefly
with these functions that the Authors are concerned in this
book) the gulf of Meaning, as it could be called, is bridged over
only by the Act of Thought — the bent line of the two shoulders
of the triangle.

Let us try to represent by analogous diagrams the earlier stages
of Meaning. At the first stage, when the utterance is a mere
sound-reaction, expressive, significant and correlated with the
situation, but not involving any act of thought, the triangle is
reduced to its base, which stands for a real connection — that
323between sound-reaction and situation. The
first cannot yet be termed a Symbol nor the latter a Referent.

The beginnings of articulate
speech, when, parallel with its
appearance Referents begin to
emerge out of the Situation, are
still to be represented by a
single solid line of actual correlation
(second stage). The
sound is not a real symbol yet,
for it is not used detached from
its Referent.

First stage

image sound-reaction | connected directly with | situation

Second stage

image active sound (semi-articulated or articulated) | correlated with | referent

Third stage

image speech in action | active symbol | used to handle | referent | narrative speech | act of imagery | symbol | indirect relation | language of ritual magic | ritual act | based on traditional belief | mystically assumed relation324

In the third stage we have to distinguish between the three
fundamental uses of language, active, narrative and ritual. Each
of them is made sufficiently clear by the diagram here given,
which must be taken in conjunction with our previous analysis.
The final stage of developed language is represented by the triangle
of Ogden and Richards, and its genetic relation to its
humble predecessors may explain some of its anatomy. First
of all: the possibility of extending the Authors' diagram or pushing
it backwards into primitive speech-uses affords an additional
proof of its validity and adequacy. Further the solid nature of
almost all the bases of our triangles explains why the dotted line
in the final figure shows such tenacity and why it is capable of so
much mischief. The extreme vitality of the magical attitude to
words is explained in our foot-note to this, the theory of the book,
not only by a reference to the primitive uses of language by
savage and no doubt by prehistoric man, but also by its
perpetual confirmation in infantile uses of language and in the
very mechanism by which meaning is acquired in every individual

Some other corollaries might be drawn from our theory of
primitive meaning. Thus we might find in it an additional
confirmation of the Authors' analysis of definition. It is clear
that they are right when they maintain that ‘verbal’ and ‘real’
definition must in the end come to the same thing, and that the
making of this artificial distinction into a fundamental one has
created a false problem. Meaning, as we have seen, does not
come to Primitive Man from contemplation of things, or analysis
of occurrences, but in practical and active acquaintance with
relevant situations. The real knowledge of a word comes through
the practice of appropriately using it within a certain situation.
The word, like any man-made implement, becomes significant
only after it has been used and properly used under all sorts of
conditions. Thus, there can be no definition of a word without
the reality which it means being present. And again, since a
significant symbol is necessary for man to isolate and grasp an
item of reality, there is no defining of a thing without defining a
word at the same time. Definition in its most primitive and
fundamental form is nothing but a sound-reaction, or an articulate
word joined to some relevant aspect of a situation by means of an
appropriate human action. This definition of definition does not,
of course, refer to the same type of linguistic use as the one discussed
by the Authors of this book. It is interesting to see,
however, that their conclusions, which are arrived at by the study
325of higher types of speech, hold good in the domain of primitive
uses of words.


In the course of this essay I have tried to narrow down the
scope of each linguistic problem discussed. At first it was the
principle that the study of language needs an ethnographic background
of general culture, that linguistics must be a section,
indeed the most important one, of a general science of culture.
Then an attempt was made to show that this general conclusion
leads us to certain more definite views about the nature of language,
in which we conceived human speech as a mode of action,
rather than as a countersign of thought. We proceeded then to a
discussion of the origins and early forms of Meaning, as it must
have been experienced by Primitive Man. This gave us the
explanation and showed us the roots of the magical attitude of
man to words. Thus we moved by a series of conclusions, each
more concrete and definite than the previous one.

I wish now to touch upon one more problem, still more definite
and concrete than the others, that namely of the structure of

Every human tongue has a definite structure of its own. We
have types of isolating, agglutinative, polysynthetic, incorporating
and inflectional languages. In every one of them, the means of
linguistic action and expression can be brought under certain
rules, classified according to certain categories. This body of
structural rules with their exceptions and irregularities, the various
classes into which the elements of the language can be ranged, is
what we call ‘the grammatical structure’ of a language.

Language is usually, though, as we have seen, incorrectly,
regarded as ‘the expression of thought by means of Speech
Sounds.’ The obvious idea, therefore, is that linguistic structure
is the result of the rules of human thought, that ‘every grammatical
category is — or ought to be — the expression of some
logical category.’ But it does not require much mental effort
to realize that to hope for such perfect conjugal harmony between
Language and Logic, is far too optimistic: that in actuality
‘they often diverge from one another,’ in fact that they are constantly
at loggerheads and that Language often ill-treats Logic,
till it is deserted by her. 112326

Thus we are faced by a dilemma: either the grammatical
categories are derived from the laws of thought, and we are at a
loss to explain why the two are so ill adapted to each other. Why,
if Language has grown up in the services of Thought, has it been
so little influenced or impressed by its pattern? Or we can, to
escape these difficulties, run on to the other horn of the dilemma
as most grammarians do. They haughtily turn away from the sour
grapes of any deeper probing or philosophy of Language, and
simply affirm that Grammar rules in its own right, by a sort
of divine grace, no doubt; that the empire of Grammar must
continue in its splendid isolation, as a power hostile to Thought,
order, system and common sense.

Both views — the one appealing to Logic for help and the
other indicating an autonomous rule for Grammar — are equally
in disagreement with facts and to be rejected. It is nothing
short of absurd to assume, with the rigid grammarian, that
grammar has grown up as a sort of wild weed of human faculties
for no purpose whatever except its own existence. The spontaneous
generation of meaningless monstrosities in the brain of
Man will not be easily admitted by psychology — unless of course
the brain is that of a rigid scientific specialist. And, general principles
or predilections apart, all human languages show, in spite
of great divergences, a certain fundamental agreement in structure
and means of grammatical expression. It would be both
preposterous and intellectually pusillanimous to give up at the
outset any search for deeper forces which must have produced
these common, universally human features of Language. In
our Theory of Meaning, we have seen that Language serves for
definite purposes, that it functions as an instrument used for
and adapted to a definite aim. This adaptation, this correlation
between language and the uses to which it is put, has left its
traces in linguistic structure. But of course it is clear that we
must not look in the domain of logical thinking and philosophical
speculation for light on the aim and purposes of early human
speech, and so this purely logical view of language is as useless
as the purely grammatical one.

Real categories there are, on which the grammatical divisions
are based and moulded. But these real categories are not derived
from any primitive philosophic system, built up by contemplation
of the surrounding world and by crude speculations, such as
have been imputed to primitive man by certain anthropologists.
Language in its structure mirrors the real categories derived from
practical attitudes of the child and of primitive or natural man
327to the surrounding world. The grammatical categories with all
their peculiarities, exceptions, and refractory insubordination
to rule, are the reflection of the makeshift, unsystematic, practical
outlook imposed by man's struggle for existence in the widest
sense of this word. It would be futile to hope that we might
be able to reconstruct exactly this pragmatic world vision of the
primitive, the savage or the child, or to trace in detail its correlation
to grammar. But a broad outline and a general correspondence
can be found; and the realization of this frees us anyhow
from logical shackles and grammatical barrenness.

Of course the more highly developed a language is and the
longer its evolutional history, the more structural strata it will
embody. The several stages of culture — savage, barbarous,
semi-civilized, and civilized; the various types of use — pragmatic,
narrative, ritual, scholastic, theological — will each have
left its mark. And even the final, powerful, but by no means
omnipotent purification by scientific use, will in no way be able
to obliterate the previous imprints. The various structural
peculiarities of a modern, civilized language carry, as shown by
Ogden and Richards, an enormous dead weight of archaic use,
of magical superstition and of mystical vagueness.

If our theory is right, the fundamental outlines of grammar are
due mainly to the most primitive uses of language. For these
preside over the birth and over the most plastic stages of linguistic
development, and leave the strongest mark. The categories
derived from the primitive use will also be identical for
all human languages, in spite of the many superficial diversities.
For man's essential nature is identical and the primitive uses of
language are the same. Not only that, but we have seen that the
pragmatic function of language is carried on into its highest
stages, especially through infantile use and through a backsliding
of adults into unsophisticated modes of thinking and speaking.
Language is little influenced by thought, but Thought, on the
contrary, having to borrow from action its tool — that is, language
— is largely influenced thereby. To sum up, we can say that the
fundamental grammatical categories, universal to all human
languages, can be understood only with reference to the pragmatic
Weltanschauung of primitive man, and that, through the
use of Language, the barbarous primitive categories must have
deeply influenced the later philosophies of mankind.

This must be exemplified by a detailed analysis of one at
least of the concrete problems of grammar; and I shall choose
for a brief discussion the problem of the Parts of Speech. We
328must turn, therefore, to a stage in the development of the individual
or of mankind when the human being is not interested in
reflection or speculation, when he does not classify phenomena
for purposes of knowledge but only in so far as they enter into
his direct dealings with his conditions of existence. The child,
the primitive man, or the unsophisticated individual has to use
Language as an indispensable means of influencing his social
surroundings. In all this, a very definite attitude develops, a
manner of taking notice of certain items of reality, of singling
them out and connecting them — an attitude not framed in any
system of thought, but expressed in behaviour and, in the case
of primitive communities, embodied in the ensemble of cultural
achievements among which Language looms first and foremost.

Let us begin with the relation of a child to its surroundings.
At the earliest stage, its actions and behaviour are governed by the
wants of the organism. It is moved by hunger and thirst, desire
for warmth and a certain cleanliness, proper conditions for rest
and sleep, a due amount of freedom for movement, and last, not
least, the need of human companionship, and of handling by
adults. At a very early stage the child reacts to general situations
only, and hardly even singles out the nearest persons who minister
to its comfort and supply it with food. But this does not last
long. Even within the first couple of weeks, some phenomena,
some units begin to stand out from the general surroundings.
Human faces are of special interest — the child smiles back and
utters sounds of pleasure. The mother or the nurse is gradually
recognized, as even before that, are objects or vehicles of food.

Undoubtedly the strongest emotional appeal is exercised over
the child by the personality of its mother, and these articles or
vehicles of food. Anyone imbued with Freudian principles
might feel inclined to look here for a direct connection. In the
young of man, as in those of any Mammalian species, the infant
associates with its mother all its emotions about food. Primarily
she is for him a vessel of nourishment. If therefore nutrition is
given by any other means — and it must be remembered that
savage infants are fed with chewed vegetable food almost from
birth, as well as by the breast — the tender feelings by which an
infant responds to maternal cares are probably extended to other
ministrations of food. When one sees the loving attitude of a
modern bottle-fed baby to its bottle, the tender caresses and fond
smiles which it bestows on it, the identity of response to artificial
and natural food-conveyers seems to imply an identical mental
attitude of the infant. If this be so, we gain an insight into a
329very early process of personification of objects, by which relevant
and important things of the surroundings release the same
emotional response as do the revelant persons. However true
may be this suggestion of a direct identification, there is no doubt
that a great similarity exists between the early attitude towards the
nearest persons and objects which satisfy the needs of nutrition.

When the child begins to handle things, play with objects
of its surroundings, an interesting feature can be observed in its
behaviour, also associated with the fundamental nutritive tendency
of an infant. It tries to put everything into its mouth.
Hence the child pulls, tries to bend and ply soft or plastic objects,
or it tries to detach parts of rigid ones. Very soon isolated,
detachable things become of much greater interest and value than
such as cannot be handled in their entirety. As the child grows
up and can move things more freely, this tendency to isolate,
to single out physically, develops further. It lies at the bottom of
the well-known destructive tendency of children. This is interesting,
in this connection, for it shows how one mental faculty of
singling out relevant factors of the surroundings — persons, nutritive
objects, things — has its parallel in the bodily behaviour of the
child. Here again, in studying this detail of behaviour, we find a
confirmation of our pragmatic view of early mental development.

There can also be found a tendency to personify objects of
special interest. By the term ‘personification’ I do not mean here
any theory or view of the child's own. I mean, as in the case of
food items, that we can observe in him a type of behaviour which
does not discriminate essentially between persons and objects.
The child likes and dislikes some of his playthings, gets angry
with them should they become unwieldy; he hugs, kisses and
shows signs of attachment towards them. Persons, no doubt,
stand out first in time and foremost in importance. But even
from this it results that the relation to them is a sort of pattern
for the child's attitude towards things.

Another important point is the great interest in animals. From
my own observation, I can affirm that children a few months
old, who did not take any prolonged interest in inanimate things,
would follow a bird in its movements for some time. It was also
one of the first words which a child would understand; that is,
it would look for the bird when it was named. The interest
shown in animals at later stages of childhood is well known.
In this connection, it is of importance to us, because an animal
and especially a bird with its spontaneous movements, with its
ease of detachment from surroundings, with its unquestionable
330reminiscence of persons, is just such an object as would arouse
the child's interest, according to our theory.

Analysing the present-day savage in his relation to the surroundings,
we find a clear parallel to the attitude just described.
The outer world interests him in so far as it yields things useful.
Utility here of course must be understood in its broadest sense,
including not only what man can consume as food, use for
shelter and implement, but all that stimulates his activities in
play, ritual, war, or artistic production.

All such significant things stand out for the savage as isolated,
detached units against an undifferentiated background. When
moving with savages through any natural milieu — sailing on
the sea, walking on a beach or through the jungle, or glancing
across the starlit sky — I was often impressed by their tendency
to isolate the few objects important to them, and to treat the rest
as mere background. In a forest, a plant or tree would strike me,
but on inquiry I would be informed — ‘Oh, that is just “bush.”’
An insect or bird which plays no part in the tradition or the larder
would be dismissed ‘Mauna wala’ — ‘merely a flying animal.’
But if,on the contrary, the object happened to be useful in one way
or another, it would be named; detailed reference to its uses and
properties would be given, and the thing thus would be distinctly
individualized. The same would happen with regard to
stars, landscape features, minerals, fishes and shells. Everywhere
there is the tendency to isolate that which stands in some connection,
traditional, ritual, useful to man, and to bundle all the
rest into one indiscriminate heap. But even within this tendency
there is visibly a preference for isolated small, easily handled
objects. Their interest in animals is relatively greater than in
plants; greater in shells than in minerals, in flying insects than
in crawling ones. That which is easily detached is preferred.
In the landscape, the small details are often named and treated
in tradition, and they arouse interest, while big stretches of land
remain without name and individuality.

The great interest taken by primitive man in animals forms a
curious parallel to the child's attitude; and the psychological
reasons of both are, I think, similar. In all manifestations of
Totemism, Zoolatry, and of the various animal influences in
primitive folk-lore, belief and ritual, the interest of the savage in
animals finds its expression.

Now let us restate the nature of this general category in which
primitive mind places persons, animals and things. This rough,
331uncouth category is not defined, but strongly felt and well
expressed in human behaviour. It is constructed on selective
criteria of biological utility as well as further psychological and
social uses and values. The prominent position taken up in it by
persons colours it in such a way that things and animals enter
into it with a personified character. All items of this category are
also individualized, isolated, and treated as units. Out of an
undifferentiated background, the practical Weltanschauung of
primitive man isolates a category of persons and personified
things. It is clear at once that this category roughly corresponds
to that of substance — especially to the Aristotelian ousia. But,
of course, it owes nothing whatever to any philosophical speculation,
early or late. It is the rough, uncouth matrix out of which
the various conceptions of substance could be evolved. It might
be called crude substance, or protousia for those who prefer learned
sounds to simple ones.

As we have seen, parallel with the child's early mental attitudes,
and presumably also with those of man in the first stages of his
development, there comes the evolution of significant, articulate
sound. The category of crude substance so prominent in the
early mental outlook requires and receives articulate sounds to
signify its various items. The class of words used for naming
persons and personified things forms a primitive grammatical
category of noun-substantives. Thus, this part of speech is seen
to be rooted in active modes of behaviour and in active uses of
speech, observable in child and in savage, and assumable in primitive

Let us next treat briefly the second important class of words —
the action-words or verbs. The underlying real category appears
later in the child's mental outlook, and it is less preponderant in
that of the savage. To this corresponds the fact that the grammatical
structure of verbs is less developed in savage tongues.
Indeed, human action centres round objects. The child is and has
to be aware of the food or of the ministering person before it can
or need disentangle the act from the agent or become aware of its
own acts. The bodily states of a child also stand out much less
from the situation than the things which enter into the latter.
Thus only at a subsequent stage of the child's development can
we see that it disentangles the changes in its surroundings from
the objects which change. This happens at a stage when articulate
sounds have begun to be used by the infant. Actions such
as eating, drinking, resting, walking; states of the body, such as
sleep, hunger, rest; moods, such as like and dislike begin to be
332expressed. Of this real category of action, state and mood, we
can say that it lends itself to command as well as to indication or
description, that it is associated with the element of change, that
is, time, and that it stands in a specially close connection with the
persons of the speaker and hearer. In the outlook of savages, the
same characters could be noticed in this category; great interest
in all changes referring to the human being, in phases and types
of human action, in states of human body and moods. This
brief indication allows us to state that at the primitive stages of
human speech there must have existed a real category into which
entered all items of change capable of temporal modification,
bearing the character of human mood and of human will, and
bound up with the personal action of man.

When we look at the class of words used to denote items of
this real category, we find a close correspondence between category
and part of speech. The action-word, or verb, is capable
in all languages of grammatical modifications expressing temporal
relation, moods or modes of utterance, and the verb is also closely
associated with pronouns, a class of words which corresponds
to another real category.

A few words must be said about the pronouns. What is the
real category of primitive human behaviour and primitive speech
habits corresponding to that small but extremely vital class of
words? Speech, as we saw, is one of the principal modes of
human action, hence the actor in speech, the speaker, stands to
the foreground of the pragmatic vision of the world. Again, as
Speech is associated with concerted behaviour, the speaker has
constantly to refer to hearer or hearers. Thus, the speaker and
hearer occupy, so to speak, the two principal corner-sites in the
perspective of linguistic approach. There comes then a very
limited, special class of word corresponding to a real category,
constantly in use, easily associable with action-words, but similar
in its grammatical nature to nouns — the part of speech called
pronoun, including a few words only, but constantly in use; as
a rule short, easily manageable words, appearing in intimate
association with the verb, but functioning almost as nouns.
This part of speech, it is obvious, corresponds closely to its real
category. The correspondence could be followed into many
more interesting details — the special asymmetric position of the
third pronominal person, the problem of genders and classificatory
particles, shown especially in the third person. 113333

One point, however, referring to a common characteristic of
nouns and pronouns and dealing with the declension of the various
cases of the noun, must still be touched upon. The real category
of this latter is derived from personified units of the surroundings.
In the child, the first attitude towards items of this category is
discrimination, based on biological utility and on pleasure in
perceiving them. The infant hails them in significant sounds,
or names them with articulate words on their appearance, and
calls for them in need. Thus these words, the nouns, are submitted
to a definite use, that of naming and appeal. To this
there corresponds a subclass of noun-substantives which could
be called the appellative case, and which is similar to some uses
of the vocative and nominative in the Indo-European declension.

In the more developed uses of Language, this becomes a more
efficient adjunct of action. The thing-word comes into a nearer
association with the action-word. Persons are named, by their
names or by pronominal designations in association with what they
do: ‘I go,’ ‘thou comest,’ ‘so-and-so drinks,’ ‘animal runs,’
etc. The name of a person or personified thing is thus used in
a different manner, with a different mode of meaning as an actor,
or technically as the subject of action. This is the use corresponding
to the subjective case in which a noun is always put as the
subject of a predication. It may be said that to this case in nouns
corresponds a class of pronouns, the personal pronouns, I, thou, he.

Action is carried out with relation to certain objects. Things
and persons are handled. Their names, when associated with an
action-word in that manner, stand in the objective case, and
pronouns are used in a special form, viz., that called objective
or reflexive.

Since language is rooted in man's practical interest in things
and persons there is another relationship of fundamental importance,
that namely in which a person can lay a definite claim to
relation with or possession of, another person or thing. With
regard to the surroundings nearest people, there are the ties of
kinship and friendship. With regard to things, there comes the
economic sentiment of possession. The relation of two nouns,
standing to each other as a thing or person related to or possessed
by another thing or person, can be called the genitival or possessive
relation; and it is found as a distinct mode of connecting
two nouns in all human languages. To this corresponds also the
genitive case of European languages in its most characteristic
uses. In pronouns again, there is a special class of possessive
pronouns which expresses relationship.334

Finally, one mode of action towards outer things or people
stands out from the others, namely that determined by
spatial conditions. Without going more into detail on this
subject, I suggest that a definite subclass of substantival uses
can be assumed in all languages — that corresponding to a prepositional

There are still obviously further categories resulting from
man's utilitarian attitude, those of the attributes or qualities of a
thing, characteristics of an action, relations between things,
relations between situations, and it would be possible to show
that adjective, adverb, preposition, conjunction are based on
these real categories. One could proceed also, still dealing on
the one hand with the Semantic Matter-to-be-expressed and on
the other with structural features of Language, to explain these
latter by a reference to real facts of primitive human nature.

This short sketch, however, is sufficient to indicate the method
and the argument, by which such a genetic, primitive Semantics
could be established — a science which, referring to the primitive
attitude of Man towards Reality, would show what is the real
nature of grammatical categories. The results of such primitive
Semantics even in so far as we have indicated them, stand, I
think, in close connection with the results of Ogden and Richards.
Their contention is that a false attitude towards Language and its
functions is one of the main obstacles in the advance of philosophical
thought and scientific investigation, and in the ever-growing
practical uses of language in the press, pamphlet and novel.
Now in this and the previous section, I have tried to show that
such a crude and unsound attitude towards Language and Meaning
must exist. I have tried to demonstrate how it has arisen
and why it had to persist; and I try to trace it even into details of
grammatical structure.

There is one more thing to add. Through later processes of
linguistic use and of thinking, there took place an indiscriminate
and wholesale shifting of roots and meanings from one grammatical
category to another. For according to our view of primitive
Semantics, each significant root originally must have had its place,
and one place only, in its proper verbal category. Thus, the roots
meaning ‘man,’ ‘animal,’ ‘tree,’ ‘stone,’ ‘water,’ are essentially
nominal roots. The meanings ‘sleep,’ ‘eat,’ ‘go,’ ‘come,’ ‘fall,’
are verbal. But as language and thought develop, the constant
action of metaphor, of generalization, analogy and abstraction,
and of similar linguistic uses build up links between the categories
and obliterate the boundary lines, thus allowing words and
roots to move freely over the whole field of Language. In analytic
languages, like Chinese and English, this ubiquitous nature of
roots is most conspicuous, but it can be found even in very primitive

Now Mr Ogden and Mr Richards have brought out in a
most convincing manner the extreme persistence of the old
realist fallacy that a word vouches for, or contains, the reality
of its own meaning. A peep behind the scenes of primitive
root-formation, of the reality of primitive categories and of their
subsequent, insidious collapse, adds an important document to
the Authors' views. The migration of roots into improper places
has given to the imaginary reality of hypostatized meaning a
special solidity of its own. For, since early experience warrants
the substantival existence of anything found within the category
of Crude Substance or Protousia, and subsequent linguistic
shifts introduce there such roots as ‘going,’ ‘rest,’ ‘motion,’
etc., the obvious inference is that such abstract entities or ideas
live in a real world of their own. Such harmless adjectives as
‘good’ or ‘bad,’ expressing the savage's half-animal satisfaction
or dissatisfaction in a situation, subsequently intrude into the
enclosure reserved for the clumsy, rough-hewn blocks of primitive
substance, are sublimated into ‘Goodness’ and ‘Badness’ and
create whole theological worlds, and systems of Thought and
Religion. It must, of course, be remembered that the theory of
Ogden and Richards, and the view here expressed, maintain
most emphatically that Language, and all Linguistic processes
derive their power only from real processes taking place in man's
relation to his surroundings. I have merely touched upon the
question of linguistic shiftings, and it would be necessary to
account for them by the psychological and sociological processes
of barbarous and semi-civilized communities; exactly as we
accounted for Primitive Linguistics by analysing the mind of
Primitive Man — and as the Authors of this book account for the
virtues and imperfections of the present-day language by their
masterly analysis of the human mind in general. 336

11 See the preliminary articles in Brain, to which the Authors also
refer in Chapter X.

22 See Dr Gardiner's articles in Man, January 1919, and in The
British Journal of Psychology
, April 1922.

33 Cf. my article on “Classificatory Particles in the Language of
Kiriwina,” Bulletin of School of Oriental Studies, Vol. II., and Argonauts
of the Western Pacific
, chapter on “Words in Magic — Some Linguistic

41 See op. cit., Argonauts of the Western Pacific — An account of Native
Enterprise and Adventure in the Archipelagoes of Melanesian New
Guinea, 1922.

52 See the important Presidential Address by the late Dr W. H. R.
Rivers in the Journal of the Royal Anthropological Institute, Vol. LII.,
January-June, 1922, p. 21, and his History of Melanesian Society,
Vol. II., p. 486.

61 It was a ceremony of the Wasi, a form of exchange of vegetable
food for fish. See op. cit., Argonauts of the Western Pacific, pp. 187-189
and plate xxxvi.

71 Cf. the writer's article on “Fishing and Fishing Magic in the
Trobriand Islands,” Man, 1918.

81 I avoid on purpose the use of the expression Herd-instinct, for
I believe that the tendency in question cannot strictly be called an
instinct. Moreover the term Herd-instinct has been misused in a
recent sociological work which has, however, become sufficiently
popular to establish its views on this subject with the general reader.

91 Cited from Chapter I of the present work.

101 The correspondence between early natural sounds and the nearest
kinship terms is well known (cf. Westermarck, History of Human
, Vol. I., pp. 242-245). Here I suggest something more:
namely that the natural emotional tone of one of these sounds, ma,
and its significance for the mother, cause her appearance and thus
by a natural process form the meaning of the mama type of words.
The usual opinion is that meaning is given to them, artificially, by
adults. “The terms which have been derived from the babble of
infants have, of course, been selected, and the use of them has been
fixed, by grown-up persons.” (Westermarck, loc. cit.., p. 245)

111 Argonauts of the Western Pacific.

121 I quote from H. Sweet (Introduction to the History of Language),
because this author is one of the cleverest thinkers on language. Yet
even he sees no alternative but Rule of Logic or Anarchy in language.

131 Cf. the writer's article on ‘Classificatory Particles’ in the Bulletin
of Oriental Studies
, Vol. II.