THE LAWS OF PROGRESS IN LANGUAGE.
The history of almost every language points to the action of
certain general law’s of progress, which laws are psychological as
well as linguistic, _i.e._ they correspond and are parallel to the
growth and progress of the human mind. They may be briefly summed
up by saying that languages advance from exuberance to moderation,
from complexity and confusion to grammatical regularity, and from
synthesis to analysis. The explanation and illustration of these
laws will occupy the present chapter.
1st. _Languages advance from exuberance to moderation, by
eliminating superfluities._
The earliest languages are marked by exuberance,[214]
indetermination, extreme variety, uncontrolled liberty. They
are melodious, but prolix and measureless. Words were invented
independently, spontaneously, as they were required by the tribe or
the individual, with little or no reference to already existing
forms. The absence of literature, the want of political unity, the
habits of a nomadic life, tended to create an immense multitude of
terms and idioms. Among semi-barbarous and wandering communities
the peculiarities which we call dialects existed simultaneously and
side by side.
The Caucasus and Abyssinia present us a number of distinct
languages in a narrow district. The number and variety of the
American dialects is almost as great as that of the several tribes;
and in Oceania it has been asserted that nearly every island
or group of islands possesses a speech which barely offers any
affinity with that of the neighbouring groups.
Unity of speech is the result of civilisation, and it is
preceded by a diversity of forms which subsequently become the
characteristics of particular localities. The steps towards unity
are three; first, we have the confused, simultaneous existence of
dialectic varieties; then the isolated and independent existence
of dialects; and, finally, the fusion of these varieties in a
more[215] extended unity. Thus the earliest Hebrew records contain
traces of idioms which were subsequently the peculiar property of
Aramaic, and we find in the Homeric poems a thousand variations
of form and structure which were afterwards exclusively Æolic, or
Attic, or Doric. The explanation of this fact is to be found in
the consideration that these forms were in Homer’s time the common
property of the old Ionic tongue, and it was not till after ages
that they became appropriated and localised. The supposition that
the rhapsodists employed a judicious selection of idioms, and made
a mosaic out of distinct dialects, has long ago been abandoned as
impossible and absurd.
The process of eliminating superfluities is found in every
language. Redundancy seems to have been necessary to an early stage
of thought, for we find it not only in words but in expressions.
The whole of Hebrew poetry depends on a repetition and enforcement
of the same fundamental thought, so as to gain emphasis and
variety. In children we find a tendency to repeat the same thing
twice, once affirmatively and once negatively, as though the double
assertion gave them an additional security. “It is not you, but
I;” “This letter is not A, but B;” are turns of expression well
known to those who have observed the language of the nursery. It
is surprising to find the same unnecessary tautology existing
very widely in the most advanced literatures. “We have seen with
our eyes and heard with our ears,” is a superfluity which has many
types in the sacred writers; “They were in great numbers, not in
small,” is the translation of a line in the Œdipus Tyrannus, and we
find even a poet of our own times writing--
There saw he where some careless hand
O’er a DEAD[216] _corpse_ had heaped the sand.
There is no doubt that such tautologies are often so far from being
barren, that they give force and precision to the conception which
they convey; but the mischief of them is that they give rise to a
thousand errors of reasoning, and to many minds have the effect of
an argument.
The Spanish fleet you cannot see, because
It is not yet in sight,
or,
Et respondeo
Quia sit in eo
Vis quæ faciat homines dormire.
might be used as the satirical motto of many a treatise both in
science and metaphysics.[217]
There are two processes by which nations get rid of words which
are mere synonyms of other words, and are therefore burdensome.
The one is to drop altogether the superfluous word, or only retain
some one form or application of it; the other is to desynonymise
words by using them each with one special shade of signification.
Thus, when the Greek language obtained the word χρύσος to mean
“gold,” it dropped altogether the word αὔρον, which at one time
it must have possessed, as is clear from a comparison of the
word θήσαυρος with the Latin _aurum_. What are called anomalous
declensions and conjugations are explicable in the same manner,
since ancient idioms are always richer than those which have
undergone the revision of grammarians. It is, in fact, one of the
duties of grammarians to make a choice among the riches of popular
language, and to eliminate all words that are unnecessary. Thus
a boy would be naturally puzzled by being told that φέρω, οἴσω,
ἤνεγκα are parts of the same verb, but it will be easy for him to
understand and remember that these words are, in fact, the débris
of three entirely separate conjugations, parts of which only have
been retained, while the remaining forms have been dropped because
they were in no way needed. Merely capricious varieties have all
been solved into a single verb.
2ndly. _Languages advance from confusion to regularity, from
indetermination to grammar._
What is true of the vocabulary of a language is no less true
of its grammar. Here also simplicity is due to reflection, and
is posterior to the rich complexity of a faculty spontaneously
exercised. Scientific grammar is a subsequent invention; at their
birth languages are lawless and irregular. The reason why the
oldest and least grammatical languages appear to have the longest
grammars, is because the anomalies are all catalogued as though
they were so many rules, and what was once permissible because
it then violated no law of language is ranked as the recognised
exception to a definite order. An Isaiah would have been amazed
at reading the innumerable rules of language by which modern
grammarians suppose him to have been governed; and a Thucydides
would have been hardly less astonished to see his “syllogism of
passion” rigidly reduced to a syllogism of grammar.
At first, until usage had arisen, every body seems to have been
at liberty to invent or adopt conjugations and declensions
almost at his own caprice. “The more barbarous a language,” says
Herder, “the greater is the number of its conjugations.” It has
been a fatal mistake of philology to suppose that simplicity is
anterior to complexity: simplicity is the triumph of science,
not the spontaneous result of intelligence. The Basque language,
which has retained much of the primitive spirit, has eleven
moods; the Caffir language has upwards of twenty. Agglutination
or Polysynthetism[218] is the name which has been invented for
the complex condition of early language, when words follow each
other in a sort of idyllic and laissez-aller carelessness, and
the whole sentence, or even the whole discourse, is conjugated or
declined as though it were a single word, every subordinate clause
being inserted in the main one by a species of _incapsulation_.
This is the case with the Astec, the languages of the Pacific,
and many other languages. The Mongol declines an entire firman,
and even in Sanskrit, flexions so far supersede syntax that
the whole thought is in some sort declined. In Mexican, the
word[219] _Notlazomahiuzteopixcatatzin_, with which they salute
the priests, is easily decomposed into “Venerable priest, whom
I honour as my father;” and in Turkish,[220] the single word
_Sev-ish-dir-il-me-mek_ means “not to be brought to love one
another.” Yet even these are entirely surpassed by some of the
dialects of North America. In the[221] Iroquois, for instance, one
word of twenty-one letters expresses this sentence of eighteen
words: “I give some money to those who have arrived, in order to
buy them more clothes with it.” This one word is an agglomeration
of simple words and roots in a violent state of fusion and apocope.
3rdly. Analogous in great measure to the law which we have been
mentioning (or perhaps we may say a further development of the same
law), is the progress of language _from Synthesis to Analysis_.
We have seen that many ancient languages are polysynthetic
or[222] holophrastic, _i.e._, that they produce the entire thought
or sentence under the form of one complex and rich unity, and
subordinate every word and phrase to the domination of the entire
clause. Even in early Greek and Latin we may find traces of this
“_holophrasis_” in the separation of two parts of the same word
which was permissible by what is called _tmesis_, as for instance
in such expressions as κατὰ δάκρυ χέουσα, and even κατὰ πίονα μήρι’
ἔκηα. In Latin the same licence is far more rare, although we find
it in the lines, “_Inque cruentatus_,” &c., and it was retained
in one or two compounds, as “_Quo te cumque_ ferent.” In both
languages these extreme cases early disappeared, and the startling
audacity of Ennius in the famous
_Cere_ comminuit _brum_,
for “comminuit cerebrum,” would probably have made Virgil stare and
gasp, as much as[223] the modern
O _Jo_ qui terras de cœlo despicis _hannes_.
But although nothing is left in the Indo-European languages but
the faint _traces_ of that sylleptical tendency which seems to
have marked the earliest stage of language, they offer the most
splendid examples of a perfect synthesis. By a facile power of
composition, and by attaching to the verb and noun a variety of
terminations capable of distinguishing the nicest modifications
of meaning, they have produced an instrument of thought almost
unrivalled in accuracy and beauty.
In Greek and Latin one word was enough to express alike the
subject, copula, and predicate; in English, two are always
requisite, and generally three. The single word τύπτω requires
the three words--“I am striking”--to render it; to translate
_amabor_ in English or in German we require four words, “I shall be
loved--_Ich werde geliebt werden_;” and the same is true of many
other parts of the verb; as ἐτετιμήμεθα, _periisses_, “we had been
honoured,” “you would have perished.”
At first sight this analysis may seem to be a defect, but, in point
of fact, it is a development. It is a bad thing for the human mind
to be subjected to the despotism of a rigid grammar, the tyranny of
too perfect a form. As it is the danger of advancing civilisation
and of too refined a society to reduce men to the dead level of
uniformity, and subject every caprice of the individual to the
domination of an unwritten code, called the “laws of society,” so
a language which crystallises every relation in a definite form
tends to cramp and restrain the genius of those who use it. In the
tragedies of Æschylus and the odes of Pindar, marvellous as is the
power which crams every rigid phrase with the fire of a hidden
meaning, we yet feel that the form is cracking under the spirit,
or at least there is a tension injurious to the general effect. A
language which gets rid of its earlier inflections--English, for
instance, as compared with Anglo-Saxon--loses far less than might
have been supposed.
The progress of language from synthesis to analysis is that of the
human intelligence. Later generations find the language of their
ancestors too learned for their own use. For the unity, spontaneous
but often obscure, of the primitive tongues, they substitute an
idiom clearer and more explicit by giving a separate existence to
every subject in the sentence. They break up the conglomerated
jewels of old speech to reset them in an order less dazzling but
more distinct. They sacrifice the magnificence of mystery to the
light of distinct comprehension. Instead of one sentence, out of
whose tangled intricacies flashed, all the more brightly from
contrast, the rays of enthusiasm and genius, they attain to a
logical accuracy which gives to each idea and each relation its
isolated expression. What they lose in euphony, force, and poetic
concision, they gain in the power of marking the nicest shades of
thought; what they lose in[224] elasticity they gain in strength.
If synthetic and agglutinative languages are the best instruments
of imagination, analysis better serves the purposes of reflection.
Splendid efflorescence is followed by ripe fruit.
It is thus[225] that Sanskrit, with its eight cases, six moods, and
numerous inflections, capable of expressing a crowd of secondary
ideas, decomposes first into the Pali (?), Prâkrit, and Kawi,
dialects less rich and learned, but more precise, which substitute
auxiliaries and prepositions for case and tense; and even these
latter, too complex for ordinary use, are gradually displaced by
the more vulgar dialects of Hindostan,--the Hindoo, the Mahrattah,
and the Bengali.
In the same way the Zend, Pehlvi, and Pars-i, are replaced by the
modern Persian. The Zend, with its long and complicated words,
its want of prepositions, and its method of supplying the want by
means of cases, represents a language eminently synthetic. Modern
Persian, on the contrary, is poorer in flexions than almost any
language which exists; it may be said, without exaggeration, that
its whole grammar might be compressed into a few pages. Modern
Greek is the analysis or decomposition of ancient Greek during a
long period of barbarism. The Romance languages are Latin submitted
to the same process; Italian, Spanish, French, and Wallachian,
are merely Latin mutilated, deprived of its flexions, reduced
to shortened forms, and supplying by numerous monosyllables the
learned organisation of the ancient idioms. “The fact then that
the people in Italy, in France, in Spain, in Greece, on the
banks of the Danube and of the Ganges, have been reduced to the
necessity of treating their ancient languages in precisely the same
manner to accommodate them to their wants; and the fact that two
languages, so distant in time and space as the Pali and the Italian
for instance, occupy positions exactly identical in relation to
their mother-tongues, affords the best proof that there is in
the progress of languages a necessary law, and that there is an
irresistible tendency which leads idioms to despoil themselves of
an apparel too learned to clothe a form more simple, more popular,
and more convenient.”[226]
In the Semitic languages we find the progress towards analysis
from various[227] causes less decided, but no less ascertainable.
Ancient Hebrew is remarkable for its agglutination. “Like a child,”
says Herder, “it seeks to say all at once.” It uses one word where
we require five or six. But as we approach the period of the
captivity we find a propensity to replace grammatical mechanisms by
periphrasis, a propensity still more marked in modern or Rabbinical
Hebrew. The later dialects--Chaldean, Samaritan, Syriac--are
longer, clearer, more analytic. These, in their turn, are absorbed
into Arabic, which pushes still farther the analysis of grammatical
relations. But the delicate and varied flexions of Arabic are still
too difficult for the rude soldiers of the early Khâlifs; solecisms
multiply, grammatical forms are abandoned, and for the Arabic of
the schools we get the vulgar Arabic, which is simpler and less
elegant, but in some respects more accurate and distinct.
Even the languages of central and eastern Asia are not entirely
wanting in analogous phenomena. But the facts already adduced
are amply sufficient to prove that, in the history of languages,
Synthesis is primitive, and Analysis, far from being the natural
process of the intelligence, is only the slow result of its
development. And if it be a natural development it must, on the
whole, be considered an advance.
“An instance,”[228] observes Grimm, “unique but decisive, is
alone sufficient to replace all the proofs and arguments which I
have accumulated in my reasoning on this subject. Among modern
languages there is not one which has gained more force and solidity
than the English by neglecting or breaking the ancient rules of
sound, and suffering almost all flexions to drop. The abundance
of medial sounds, the pronunciation of which may be learnt but
cannot be taught, gives to this language a power of expression,
such as perhaps no human language has ever attained. Its highly
spiritual genius and marvellously happy development are due to
the astonishing union of the two most noble languages in modern
Europe, German and Romance. We know the part which each of these
elements plays in the English language; one of them is almost
entirely devoted to the representation of sensible ideas, the other
to the expression of intellectual relations. Yes, the English
language, which has produced and nourished with its milk the
greatest of modern poets, the only one who can be compared to the
classical poets of antiquity (who does not see that I am speaking
of Shakspeare?), may of good right be called an universal language,
and seems destined, like the English people itself, to extend its
empire farther and farther in all quarters of the globe.”
To the laws which we have been considering, many philologists would
be inclined to add a fourth--viz., the _progress to polysyllabism
from a state originally monosyllabic_. Many arguments may
undoubtedly be adduced, which give a _primâ facie_ probability to
this supposition.[229] We will proceed briefly to state them.
It is argued, firstly, that we should have expected _à priori_ a
predominance of monosyllabic roots, because it is unlikely that a
single powerful impression would have expressed itself by more
than one sound. Since one sound would have been sufficient, we
should not be inclined to look for any superfluity. Impression
would provoke expression with the same rapidity that the flash
of lightning is kindled by the shock of two electric clouds. It
must be remembered that the young senses of the human race were
unaccustomed to compound articulations, and neither their ears nor
their tongues would have led them to signify by two sounds or two
syllables an impression essentially single.
_Secondly_, it is said that existing facts prove the likelihood
of this conclusion. Thus, to this day, some nations are unable
to pronounce compound consonants by one emission of the voice.
Such is the case with the Mantschou, and the Chinese can only
utter the word _Christus_ by changing it according to the custom
of his language into _ki-li-su-tu-su_.[230] The Chinese then
may be considered as a language petrified in its first stage of
flexionless and ungrammatical monosyllabism. Thus, in order to
express the plural, they are obliged to add the words, “another”
and “much,” or to repeat the noun twice, expressing “us” by “me
another,” and trees by “tree,[231] tree.” The prayer, “Our Father
which art in heaven,” assumes in Chinese[232] the form “Being
heaven me another (= our) Father who,” a style not unlike the
natural language of very young children.
Thirdly, it is asserted that all existing languages are capable
of being deduced from monosyllabic roots; that even the
triliteral[233] Semitic languages afford abundant evidence of the
fact that the three consonants are only the result of a growth,
since one of the consonants is often weak and unnecessary, and many
of the words expressing simple[234] ideas have only one syllable.
Whatever weight may attach to these considerations, they do not
appear to be convincing. The attempt of Fürst and Delitzch to
get over the fact of Semitic triliteralism is not completely
successful, and no evidence has ever been adduced to show the
causes which could have influenced a language to abandon an
essentially monosyllabic character, or the time when so immense a
change could have taken place.
Chinese, as we know, has been monosyllabic from the earliest
period, and continues so to this day; and even Thibetian and
Burmese,[235] though they have, under the influence of other
languages, made great efforts to attain a grammar, have yet
retained the ineffaceable impress of their original condition. We
therefore reject this fourth law, as one which, even if possible,
is by no means proven. Further discussion of it will, indirectly,
be involved in the following chapter. At best, it can only be
regarded as an artificial hypothesis, occasionally convenient for
the purposes of the grammarian, but not corresponding to any real
condition of the languages as once spoken.