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

Sayce, Archibald-Henry. Principles of Comparative Philology – T02


The substance of the first eight of the following
chapters was originally delivered in the form of
lectures at Oxford in the early part of 1873. The
last chapter is a subsequent addition, which should
strictly be regarded as an appendix of the first.
The detailed treatment of a single philological
principle in it is so disproportionate to the general
plan of the book, that its introduction can only be
defended on the double ground of the great and
far-reaching influence of analogy, and the scant
attention it has hitherto received. In tracing its
action, I have had to review all the various parts
of the science of language; and practical illustrations
of this kind may be more efficacious than
pages of abstract argument, in showing clearly
what I believe to be comprehended within the
limits of Glottology, and in summing up the results
I have tried to make good in the preceding

For the sphere and subject-matter of Comparative
Philology is very liable to be unduly narrowed.
viiThe danger lies not so much with the followers of
Steinthal, who consider language too exclusively
from the psychological and à priori point of view,
as with those who, treading in the footsteps of
Schleicher, would assimilate its study to the
fashionable physical sciences, and demand a place
for it by the side of chemistry or physiology.
The belief that philology is etymology has passed
away, only to be succeeded by the belief that it is
phonology. But this belief, whether consciously
or unconsciously held, cannot be too strenuously
resisted. Etymology and phonology are parts of
Comparative Philology, but they do not constitute
Comparative Philology. This is a science of far
more comprehensive reach; a science, too, which
takes its start not from matter, but from mind.
Glottology is an historic, as opposed to a physical,
science; and its object is to trace the development
of the human intelligence as expressed in
the outward and enduring monuments of speech.
Language is at once the creator and the mirror of
society; and it is in and through society that the
human mind has attained its present level of
civilisation. Our science, therefore, by comparing
the linguistic relics of social change and thought,
by classifying sounds, and words, and sentences,
by tracing out the history of forms and syntax,
and determining the laws which govern speech,
will work back to the progressive intelligence that
produced them, and will tell us with the certainty
of scientific knowledge, better than all the flints of
viiiAbbeville or the skulls of Bruniquel, how “man,
the speaker,” first raised himself above the level
of the brute, how society progressed from an hive-like
communism to the republics of Greece and
the kingdoms of modern Europe, and how the
fairy world of mythology, the instincts of an unrevealed
religion, the philosophic systems of East
and West, have grown out of the manifold imaginings
of the mind as it struggled to express
itself in language. To understand the present,
and to provide for the future, we must know the
past; and the key to this is given us by scientific
philology. The very problems which face the
logicians of our day as they once faced Aristotle
and Bacon, will never be solved until it is recognised
that, instead of building up a so-called
science on a narrow basis of empiric observation
like the great Stagirite, or determining like Hegel
the laws of being from the standpoint of modern
European speech, we must begin at the opposite
end, and learn from Glottology how the framework
and character of our thought originated, and
wherein it differs from that of other races in the
past and present.

Phonology and letter — change, comparative
grammar and comparative mythology, the history
of words and their meanings, the origin of flection
and the nature of roots, — such are the subjects
with which scientific philology has to deal;
and the construction of an universal language
is the practical object towards which it aims.
ixUnder the head of Comparative Grammar is included
comparative syntax, a most important
branch of study, but one which is only beginning
to be worked. A thorough-going investigation of
it may throw light on the difficult question as to
the possibility of a mixed grammar; and Mr Edkins
already believes that he can detect the influence
of Semitic idiom upon the doctrine of the relative
and the definite article in Greek. The origin
of language itself must be left to other sciences
to reveal, but there is no reason to despair of our
eventually determining this problem of problems.
Glottology, however, has to postulate the existence
of conscious and articulate speech; all that
it can do is to point the way to the true solution
of the riddle, to show what is the conclusion towards
which its body of facts and evidence is
tending. But this does not prevent the solution
of the riddle being of the utmost importance to
it; on the contrary, like the law of gravitation in
astronomy, a knowledge of the genesis of speech
will bind together the empirical generalisations of
language, and give the reason for their special
character. We cannot properly be said to know
a subject, or to trace the course of its development,
until we are able to resolve it into its
original elements, and to discover how and out of
what it arose.

The following pages, it will be seen, are rather
critical than constructive. New theories have indeed
been put forward in regard to mythology,
xand such points as gender or number; but the
chief feature of the first seven chapters of the
book is a criticism of certain generally-received
hypotheses which underlie a good deal of current
philological reasoning, but which do not stand, as
it seems to me, the test of facts. These hypotheses
may be reduced to three axiomatic assumptions,
against which the present rough-hewn
work, however devoid of the graces of style, and
bristling with uncouth words, is intended to be a
protest. The belief that the Aryan languages are
the standard of all others, and that the generalisations
gathered from their exceptional phenomena
are laws of universal validity; the substitution of
the mechanical and the outward for the intellectual
and the inward; the confusion between the convenient
classifications of science and actual divisions
into natural “families,” — these are the
three assumptions which, though maintained unconsciously,
and rejected by most students when
presented in their crude form, are yet the real
causes of certain fashionable theories which have
even been elevated into “the most unquestionable
results of modern philology.” First and
foremost among these is the doctrine of a graduated
evolution of speech through an isolating and
agglutinative into an inflectional stage — a doctrine
which rests upon the second assumption, and
explains the forms of grammar by the accidents
of phonetic decay. When will it be recognised
that the growth of most of our present flections
xiout of independent words indicates not a primitive
agglutination, but a pre-existing inflectional instinct
or analogy, which they could but follow,
and that the near approach of certain members of
the agglutinative group — the Finnic idioms, for
example — to some of the phenomena of inflection,
only proves the fixed character of their mental
point of view, which remained true to its agglutinative
type although the outward crust of language,
the phonetic expression of the inward
thought, had done its utmost to bring about a

Had it been remembered in what language
really consists, we should have heard less of letters
and more of sounds, less of outward form and
more of inward meaning, less of phonetic decay
and more of analogy; the philologist would have
betaken himself to the study of living speech
rather than of dead literature, and have learned
that, instead of starting with the written crystallised
word, he should have begun with the only
actual whole of which language knows — the sentence.
Had the sentence been made the basis of
research, little would have been said of an agglutinative
background to Aryan speech, or of a time
when men talked with one another in roots. But,
in fact, the larger part of the strange hypotheses
which the discovery of roots has called forth, are
mostly dependent on the first assumption. I feel
confident that the world would never have heard
of “pronominal roots” had the Turanian tongues
xiibeen the primary subject of inquiry, nor would the
supposed necessity of finding biliteral radicals
have made such wild havoc in the Semitic family.
Even the term “family” itself calls up erroneous
ideas. The days are passed indeed when philological
and ethnological unity were imagined to be
identical, but we still picture to ourselves a
“family of languages” like a family in social
life, except that it springs not from two ancestors,
but from one. Such pictures, however, are but
the convenient symbols of working science, and if
pressed too literally, lead to conclusions the reverse
of the truth. Simplicity and unification are the
latest result of time, and instead of forcing all the
known dialects of the world under a few neatly-labelled
classes or “families,” we should rather
wonder that more waifs and strays have not come
down to us out of the infinite essays of early

The arguments with which I have endeavoured
to combat these and similar views are founded
upon three or four postulates. Language is social,
not individual, interpreting the society of the past,
and interpreted by the society of the present; it
starts with the sentence, not with the word; it is
the expression of thought, so that all explanations
of its phenomena which rest contented with its outward
form alone must be inadequate or erroneous;
and its study, if carried on by the light of the
comparative method, ought to embrace all the
manifold operations and products of thought which
xiiiare embodied in spoken utterance. These are the
principles which underlie the following pages, and
will furnish the key to what I have written.
Throughout I have presupposed an acquaintance
with Professor Max Müller's “Lectures on the
Science of Language,” to whose world-wide popularity
Comparative Philology owes its present
position and its present charm. My indebtedness
to their wealth of illustration will be apparent to
every reader, and the familiar character of the
work has relieved me of the necessity of encumbering
my book with frequent references to it.

A. H. Sayce.
Queen's College, Oxford,
May 1874.xiv