Jespersen, Otto. Progress in Language – T01

[Progress in language]

Chapter I.
Introduction.

1. (1) No language is better suited than English
to the purposes of the student who wishes, by means
of historical investigation, to form an independent
opinion on the life and development of language in
general. In English we have an almost uninterrupted
series of written and printed works, extending over
a period of more than a thousand years; and, if we
are not contented with the results to be obtained
from these sources, comparative philology comes in,
drawing its conclusions from all the cognate tongues,
and showing us, with no little degree of certainty, the
nature of the language spoken by the old Germans
at the time when the differentiation of the several
tribes had as yet scarcely begun. The scientific investigations
of our century go still further back: they
have brought together Greek and Latin, German,
Slavonic, Lithuanian, Celtic, Indian and Persian, as
one indissoluble unity;. through a long succession of
parallelisms they have pointed out what is common
to all these languages, and have made it possible
1to some extent to reconstruct the unwritten language
used in intercourse by the ancestral people
several centuries before the era of any languages
historically accessible to us. If we do not
know where the original Arian (or, as it is often
termed, Indo-European or Indo-Germanic) people
lived, we know much about the structure of their
speech.

2. (1) During the course of the ages the language
of the Arians has changed in a multiplicity of ways in
the mouths of different nations; but nowhere has the
original type been more radically modified than in
England. The amount and thoroughness of these
modifications will perhaps be perceived most clearly if
we take some recognised definition of the most essential
features characterising Arian speech, in opposition
to the motley crowd of other tongues. We shall
find that scarcely one of those features is characteristic
of present-day English. Friederich Müller
thus describes the distinguishing traits of the languages
of the Arian type: 11“In the Indo-Germanic languages
root, stem and word are rigorously discriminated”.
In English words such as man or wish no one is able
to make any such separation. “The two categories
of noun and verb are kept clearly from each other”.
Not so in English: e.g., man is generally a noun, but
it is used as a verb when we say, “Man the ship”;
compare also I wish and my wish. “Nouns belong
2to one of three genders, masculine, feminine, or neuter.”
From English grammatical gender has disappeared.
“The distinction between the several grammaticological
categories is here carried out strictly.” This
is not the case in English., where, to mention only
one point, nouns and adverbs may be used as
adjectives.:

3. (2) But if the old order has thus changed,
yielding place to new, the question naturally arises:
Which of these two is the better order? Is the sum
of those infinitesimal modifications which have led
our language so far away from the original state to
be termed evolution or dissolution, growth or decay?
Are languages as a rule progressive or regressive?
And, specially, is modern English superior or inferior
to primitive Arian?

If I am right in my interpretation of the tendencies
of recent philology, the answer cannot be doubtful;
but there is as little doubt that this answer will be
the exact opposite of what an older generation of
linguists would have given as their verdict. It may
therefore be of some interest to examine more closely
the linguistic philosophy of the age that is now going
out. How did the leading men of some thirty years
ago classify and estimate different types of speech,
and what place did they assign to such languages
as modern English?

It would scarcely be possible to find any one man
better suited to represent typically the views here
referred to than August Schleicher. In a series
3of highly important works 12 he has dealt with problems
of classification of languages and linguistic development
in general: by his exceptional knowledge of
a number of languages and as being the first
master builder of that lofty structure, the Arian
“ursprache,” he stands out pre-eminently among his
contemporaries, and exercises a vast influence down
to our own day: in spite of all apparent difference,
it is his ideas that form the basis alike of Max
Müller's brilliant paradoxes and of Whitney's
sober reasonings: he is rightly to be considered the
spiritual father of every comparative philologist of our
own times, notwithstanding the gulf separating his
views from those of some of the younger generation.
Let us, therefore, try to give a short account of his
leading ideas and the manner in which he arrived at
them: our investigation will show the curious spectacle
of a classification and a theory completely outliving
the basis of reasoning on which they were founded.

5. (4) According to Schleicher, the three classes
of languages are not only found simultaneously in
6the tongues of our own day, but they represent
three stages of linguistic development; “to the
nebeneinander of the system corresponds the nacheinander
of history”. Beyond the flexional stage no
language can attain; the symbolic denotation of
relation by flexion is the highest accomplishment
of language; speech has here effectually realised its
object, which is to give a faithful phonetic image of
thought. But before a language can become flexional
it must have passed through an isolating and an agglutinating
period. Is this theory borne out by historical
facts? Can we trace back any of the existing
flexional languages to agglutination and isolation?
Schleicher himself answers this question in the negative:
the earliest Chinese with which we are acquainted
is as monosyllabic as the Chinese of to-day, and the
earliest Latin was of as good a flexional type as are
the modern Romance languages. This would seem
a sort of contradiction in terms; but the Hegelian is
ready with an answer to any objection; he has the
word of his master that History cannot begin till the
human spirit becomes “conscious of its own freedom,”
and this consciousness is only possible after the
complete development of Language. The formation
of Language and History are accordingly successive
stages of human activity. 14 Moreover, as history and
historiography, i.e., literature, come into existence
simultaneously, Schleicher is enabled to express the
same idea in a way that is “only seemingly paradoxical,” 15
7namely, that the development of language
is brought to a conclusion as soon as literature
makes its appearance; this is a crisis after which
language remains fixed; language has now become
a means, instead of being the aim, of intellectual
activity (Sprachvergl. Unters., i., 24). We never meet
with any language that is developing or that has
become more perfect (ibid., 13); in historical times
all languages move only downhill; this is not to
be disputed (“das ist ausgemachte wahrheit,” ibid.,
14); linguistic history means decay of languages
as such, subjugated as they are through the gradual
evolution of the mind to greater freedom (ibid.,
17).

6. (4) This doctrine of an antagonism between
language and history is a pet theory which Schleicher
never abandons; in his first book (ii., p. 134) he
speaks of “die geschichte, jene feindin der sprache”;
and in his Darwinian period he puts it in this way:
“The origin and development of language is previous
to history, properly and strictly speaking…History
shows us nothing but the aging of languages according
to fixed laws. The idioms spoken by ourselves,
as well as those of all historically important nations,
8are senile relics” (Die Bedeut. d. Spr., 27 ; cf. Die
Darwinsche Theorie, 27; D. Spr., 37). 16

7. (5) According to Schleicher, then, we witness
nothing but retrogression and decay; but as the
same view is found as early as Bopp, and as it is
the fundamental belief, more or less pronounced, of
many other linguistic speculators, we are justified in
supposing that with Schleicher the theory is not
really due to the Hegelian train of argument, but
that here, as not unfrequently, reasoning is summoned
to arms in defence of results arrived at by instinct.
And the feeling underlying this instinct, what is it
but a grammar-school admiration, a Renaissance
love of the two classical languages and their literatures?
People were taught to look down upon
modern languages as mere dialects, and to worship
Greek and Latin; the richness and fulness of forms
9found in those languages came naturally to be considered
the very beau idéal of linguistic structure.
To men fresh from the ordinary grammar-school
training no language would seem respectable that
had not four or five distinct cases and three genders,
or that had less than five tenses and as many moods
in its verbs. Accordingly, such poor languages as
had either lost much of their original richness in
grammatical forms (e.g., French, English., or Danish),
or had never had any (e.g., Chinese), were naturally
looked upon with something like the pity bestowed
on relatives in reduced circumstances, or the contempt
felt for foreign paupers.

8. (6) Comparative philologists had one more
reason for adopting this manner of estimating languages.
To what had the great victories won by
their science been due? Whence had they got the
material for that magnificent edifice which had proved
spacious enough to hold Hindus and Persians,
Lithuanians and Slavs, Greeks, Romans, Germans,
and Celts? Surely it was neither from Modern
English nor Modern Dutch, but from the oldest
stages of each linguistic group. The older a linguistic
document was, the more valuable it was to the first
generation of comparative philologists. An English
word like had was of no great use, but Gothic habaidêdeima
was easily picked to pieces, and each of its
several elements lent itself capitally to comparison
with Sanskrit, Lithuanian and Greek. The philologist
was chiefly dependent for his material on the old and
10archaic languages; his interest centred round their
fuller forms; what wonder then if in his opinion they
were superior to all others? What wonder if by
comparing had and habaidêdeima he came to regard
the English word as a mutilated and worn-out relic
of a splendid original? or if in noting the change from
the old to the modern form he used strong language
and spoke of degeneration, corruption, depravation,
decline, phonetic decay, etc., or even adopted for
himself Schleicher's noble simile? “Our words, as
contrasted with Gothic words, are like a statue that
has been rolling for a long time in the bed of a river
till its beautiful limbs have been worn off, so that
now scarcely anything remains but a polished stone
cylinder with faint indications of what once it was”
(Deutsche Spr., 34).

9. (6) Suppose, however, that it would be quite
out of the question to place the statue on a pedestal
to be admired; what if, on the one hand, it was not
ornamental enough as a work of art, and if, on the
other, human well-being was at stake if it was not
serviceable in a rolling-mill: which would then be
the better, — a rugged and unwieldy statue, making
difficulties at every rotation, or an even, smooth,
easy-going and well-oiled roller?

10. (7) Schleicher does not explain by what test
he estimates the comparative merits of languages;
the whole tenor of his linguistic philosophy hinders
him from getting at the only one that really is of
any value: the practical interests of the speaking (or
11talking) community. Schleicher emphatically repeats
on every occasion that linguistics is a natural science;
he never wearies of insisting upon the distinction
between linguistics (or glottics, as he himself terms
it) and the purely historical science of the scholar
(“philology” in the broad or German sense of the
word). Language is to Schleicher a natural object,
just as much as a plant is. And if you object that
language is nothing but human action and has no
material existence, he will answer 17 by defining
language in an entirely materialistic way as the
result, perceptible through the ear, of the action
of a complex of material substances in the structure
of the brain and of the organs of speech with their
nerves, bones, muscles, etc. Anatomists, however,
have not yet been able to demonstrate differences in
the structures of these organs corresponding to differences
of nationality, — to discriminate, that is, the
organs of a Frenchman (quâ Frenchman) from those
of a German (quâ German). Accordingly, as the
chemist can only arrive at the elements which compose
the sun by examining the light which it emits,
while the source of that light remains inaccessible to
him, so we must be content to study the nature of
languages not in their material antecedents but in
their audible manifestations. It makes no great difference,
however; for “the two things stand to each other
as cause and effect, as substance and phenomenon: a
philosopher would say that they are identical”.12

11. (7) I, for one, fail to understand how this can
be, what Schleicher believes it, “a refutation of the
objection that language is nothing but a consequence
of the activity of these organs”. The sun exists
independently of any human observer; but there could
be no such thing as language if there was not besides
the speaker a listener who might become a speaker
in his turn. However this may be, it is certain that
Schleicher never succeeds in establishing a rational
basis for determining the relative value or merit of
different languages. 18

But this is quite easy if we take for our guide an
idea expressed long ago and with considerable emphasis
by Wilhelm Von Humboldt, that language
means speaking, and that speaking means action on
the part of a human being to make himself understood
by somebody else. Then it becomes evident that
that language ranks highest which goes farthest in the
art of accomplishing much with little means, or, in
other words, which is able to express the greatest
amount of meaning with the simplest mechanism.13

12. (7) Rask says 19 that “an elaborate linguistic
structure with a variety of endings in declensions and
conjugations has certain advantages…but it may
be that the advantages of the opposite simplicity are
still greater”. Madvig defends our modern analytical
languages with great vigour. He says that they
are just as good as the old synthetic ones, for thoughts
can be expressed in both with equal clearness;
poverty in grammatical forms is no drawback to a
language. I shall try to show that we are justified in
going still further than these two eminent men, and
saying the fewer and shorter the forms, the better;
the analytic structure of modern European languages
is so far from being a drawback to them that it gives
them an unimpeachable superiority over the earlier
stages of the same languages. The so-called full and
rich forms of the ancient languages are not a beauty
but a deformity.

13. (8) In putting forward these propositions, I am
not treading on entirely new ground. In Jacob
Grimm's singularly clever (though nebulous) essay on
the Origin of Language (1851), I find such passages
as the following: “Language in its earliest form was
melodious, but diffuse and straggling (weitschweifig
und haltlos); in its middle form it was full of intense
poetical vigour; in our own day it seeks to
remedy the diminution of beauty by the harmony of
the whole, and is more effective though it has inferior
means”; he arrives at the result that “human
14language is retrogressive only apparently and in
particular points, but looked upon as a whole it is
progressive, and its intrinsic force is continually
increasing”. The enthusiastic panegyric on the
English language with which he concludes his essay
forms a striking contrast to Schleicher's opinion that
English shows “how rapidly the language of a
nation important both in history and literature can
sink”. 110

14. In recent linguistic literature indications of a
reaction against the prevailing manner of estimating
languages are also found, though the reaction is only
of a sporadic and rather timid character. Thus
Kräuter 211 says: “The dying out of forms and sounds
is looked upon by the etymologists with painful
feelings; but no unprejudiced judge will be able to
see in it anything but a progressive victory over
lifeless material. 312 Among several tools performing
equally good work, that is the best which is simplest
and most handy; this illustration has some significance
for the subject under discussion…That
decay is consistent with clearness and precision, is
shown by French; that it is not fatal to poetry, is
seen in the language of Shakespeare.”

Osthoff says: “We should avoid a one-sided
depreciation of the language of Lessing and Goethe
15in favour of those of Wulfila or Otfried, or vice versâ.
A language possesses an inestimable charm if its
phonetic system remains unimpaired and its etymologies
are transparent; but pliancy of the material
of language and flexibility to express ideas is really
no less an advantage. Everything depends on the
point of view: the student of architecture has one
point of view, the people who are to live in the
house another.” 113

E. Tegnér gives as the conclusion of an interesting
disquisition that “so far from being more perfect
than both the other groups [agglutinating and
isolating] the flexional languages are radically
inferior to them because they impede liberty of
thought”. 214

15. (8) As such utterances are, however, comparatively
isolated, and as the authors quoted, as
well as the great majority of living linguists, are in
many respects still in the toils of Schleicher's system,
I hope that the following attempt to apply consistently
the principle laid down in § II, and to draw
some further conclusions from the results obtained
by comparison of the older and younger stages of
Arian languages, will have some interest for linguistic
students. My design being principally to gain insight
16into historical developments, it will be noticed,
firstly, that I do not attempt to fix the comparative
value of languages that are not closely related to
each other; and, secondly, that the examples I take
are not isolated facts, but typical and characteristic
of the total structures of the languages I am dealing
with.17

Chapter II.
Ancient and modern languages.

16. (9) First, let us look at Schleicher's example:
English had and Gothic habaidêdeima. The English
form is preferable, on the principle that any one who
has to choose between walking one mile or four miles
will, other things being equal, prefer the shorter cut.
It is true that if we take words to be self-existing
natural objects, habaidêdeima has the air of a giant,
and had (like most other words which have been
exposed to phonetic changes carried on through a
long succession of ages) is left a mere pigmy. If,
however, we remember the fact that what we call a
word is really and primarily the combined action of
human muscles to produce an audible effect, we see
that the shortening of a form means a diminution of
effort and a saving of time in the communication of
our thoughts. If had has suffered from wear and
tear in the long course of time, this means that the
wear and tear of people now using this form in their
speech is less than if they were still encumbered with
the old giant habaidêdeima (comp. below, § 92, footnote).

17. (10) But it is not only in regard to economy of
18muscular exertion that the English had carries the
day over the Gothic form. Had corresponds not
only to habaidêdeima, but it unites in one short
form everything expressed by the Gothic habaida,
habaidês, habaidêdu, habaidêduts, habaidêdum, habaidêduÞ,
habaidêdun, habaidêdjau, habaidêdeis, habaidêdi,
habaidêdeina, habaidêdeits, habaidêdeima, habaidêdeiÞ,
habaidêdeina, — separate forms for two or three persons
in three numbers in two distinct moods! It is clear,
therefore, that the English form saves a considerable
amount of brain work to all English-speaking people,
and especially to every child learning the language.
Some one will, perhaps, say that on the other hand
English people are obliged always to join personal
pronouns to their verbal forms, and that this is a
drawback counterbalancing the advantage, so that
the net result is six of one and half a dozen of the
other. This is, however, not entirely the case. In
the first place, the personal pronouns are the same
for all tenses and moods, but the endings are not.
Secondly, the possession of endings does not exempt
the Goths from having separate personal pronouns;
and whenever these are used, the verbal endings which
indicate persons are superfluous. They are no less
superfluous in those extremely numerous cases in
which the subject is either separately expressed by a
noun or is understood from the preceding proposition.
So that, altogether, the numerous endings of the older
languages must be considered uneconomical.

18. (12) If I have shown that the older Arian
19languages burden the memory by the number of their
flexional endings, they do so no less by the many
irregularities in the formation of these endings. Irregularity
may be termed a consequence of flexion —
not, indeed, a logical consequence of any definition of
flexion, for we might very well imagine some language
of the Volapük kind in which all flexions were completely
regular; but, as a matter of fact, such a
language never existed. In Latin, in Greek, in
Sanskrit, in Gothic, in all existing flexional languages
of the same type, anomaly and flexion invariably go
together. If the accidence of Modern English nouns
can be set forth in a few pages, this is not exclusively
due to the fewness of the cases, but also to the fact
that nearly all nouns are declined in pretty much the
same way: but the further back we go in the history
of English or any other cognate language, the greater
is the number of exceptions and anomalies of every
description which we shall encounter. This will
become especially clear when the facts of grammar
are arranged as I have arranged them below (chapter
vi.). And it is not only the forms themselves that are
irregular in the early languages, but also their uses:
logical simplicity prevails much more in Modern
English syntax than in either Old English or Latin
or Greek. But I need hardly point out that growing
regularity in a language means a considerable gain to
all those who learn it or speak it.

19. (12) Let me here quote an interesting remark
made by Friederich Müller in speaking of a
20totally different language: 115 “Even if the Hottentot,”
says he, “distinguishes ‘he,’ ‘she,’ and ‘it,’ and strictly
separates the singular from the plural number, yet
by his expressing ‘he’ and ‘she’ by one sound in
the third person, and by another in the second, and
by his denoting the plural differently according to
person and gender, he manifests that he has no perception
at all of our two grammatical categories of
gender and number, and consequently those elements
of his language that run parallel to our signs of
gender and number must be of an entirely different
nature”. Fr. Müller certainly goes too far in this
glorification of the speech of his own countrymen,
on account of its superiority to that of the poor
Hottentots; for could not the very same thing which
he objects to the Hottentot language be predicated
of his own? “As the Germans express the plural
number in different manners in words like gott–
götter, hand–hände, vater–väter, frau–frauen, etc.,
they must be entirely lacking in the sense of the
category of number!” Or let us take such a language
as Latin; there is nothing to show that dominus
bears the same relation to domini as verbum to verba,
urbs to urbes, mensis to menses, cornu to cornua,
fructus to fructûs, etc.; even in the same word the
idea of plurality is not expressed by the same method
for all the cases, as is shown by a comparison of
dominus–domini, dominum–dominos, domino–dominis,
domini–dominorum. Fr. Müller is no doubt
21wrong in saying that such anomalies preclude the
speakers of the language from conceiving the notion
of plurality; but, on the other hand, it seems evident
that a language in which a difference so simple even
to the understanding of very young children as that
between one and more than one, can only be expressed
by a complicated apparatus, must rank lower
than another language in which this difference has
a single expression for all cases in which it occurs.
In this respect, also, Modern English stands higher
than Latin, Hottentot, or the oldest English.

20. I must pause here a moment to reply to some
objections that have been made to my manner of
viewing these points. It has been said 116 that the
difficulties experienced by a grown-up person in
learning a foreign language are not felt by a child
picking up its mother tongue: children will learn an
inflexional language with the same ease as one which
is analytical; the real difficulties in learning a foreign
language are “those thousands of chicanes caused
by that tyrannical, capricious, utterly incalculable
thing ‘idiomatic usage,’ but this gives little or no
trouble to children learning to talk”. I think, however,
that if any one will listen attentively to children
talking, he will soon perceive that they make a
great number of mistakes, not only in inflecting
strong verbs like regular verbs, etc., etc., but also in
arranging the words of a sentence in a wrong order,
22giving unusual significations to words, using the
wrong prepositions, and, in fact, violating usage in
every possible way. In all this I see evidence of
the labour involved in learning a language, a labour
that is not to be underrated even when the language
is learnt under the most favourable circumstances
possible. And I think there can be no doubt that
the exertion must be greater in the case of highly
complicated linguistic structures with many rules and
still more exceptions from the rules, than in languages
constructed simply and regularly. It is, of course,
impossible actually to prove that it is easier for an
English child to learn to speak English than it was
for a Gothic or Anglo-Saxon child to learn those
languages; but it seems highly probable.

21. Nor is the difficulty of correct speech confined
to the first mastering of the language. Even to the
native who has spoken the same language from a
child, its daily use involves no small amount of
exertion. Under ordinary circumstances he is not
conscious of any exertion in speaking; but such a
want of conscious feeling is no proof that the exertion
is absent. And it is a strong argument to the
contrary that it is next to impossible for you to speak
correctly if you are suffering from excessive mental
work; you will constantly make slips in grammar and
idiom as well as in pronunciation; you have not the
same command of language as under normal conditions.
If you have to speak on a difficult and unfamiliar
subject on which you would not like to say
23anything but what was to the point or strictly justifiable,
you will sometimes find that the thoughts
themselves claim so much mental energy that there is
none left for speaking with elegance or even with
complete regard to grammar: to your own vexation
you will have a feeling that your phrases are confused
and your language incorrect. A pianist may practise
a difficult piece of music so as to have it “at his
fingers’ ends”; under ordinary circumstances he will
be able to play it quite mechanically without ever
becoming conscious of effort; but, nevertheless, the
effort is there. How great the effort is appears when
some day or other the musician is “out of humour,”
that is, when his brain is at work on other subjects or
is not in its usual working order. At once his execution
will be stumbling and faulty.

22. (11) To return to had and habaidêdeima. If we
look at the meaning of these forms we perceive that the
English word has made a great advance on the road
from the concrete to the abstract. It is a well-known
law in psychology that the power of grasping abstract
notions is of comparatively late growth in the individual
as well as in the racc. The development in
language of grammatical forms of a more abstract
character constitutes a great advance upon the earlier
state when there was little beyond concrete terms.
The notion that was formerly expressed by one
inseparable word is now often expressed by means
of a group of pronouns, auxiliary verbs, prepositions,
and other little words, each with a comparatively
24abstract signification. It is one of the consequences
of this change that it has become considerably easier
to express certain minute shades of thought by laying
extra stress on some particular element in the speechgroup.
The Latin cantaveram amalgamates three
ideas into one indissoluble whole; but in the English
I had sung the elements are analysed, so that you
can at will accentuate the personal element, the time
element, or the action. Now, it is possible (who can
affirm and who can deny it?) that the Romans could,
if necessary, make some difference in speech between
cántaveram (non saltaveram), “I had sung” and
cantaverám (non cantabam), “I had sung”; but even
then if it was the personal element which was to
be emphasised, an ego had to be added. Even the
possibility of laying stress on the temporal element
broke down in forms like scripsi, minui, sum, audiam,
and innumerable others. It seems obvious that the
freedom of Latin in this respect must have been far
inferior to that of English. Moreover, in English the
three elements, “I,” “had,” and “sung,” can in certain
cases be arranged in a different order, and other words
can be inserted between them in order to modify
and qualify the meaning of the phrase. Note also
the conciseness of such answers as “Who had sung?”
“I had”; “What have you done?” “Sung”. And
contrast the Latin “cantaveram et saltaveram et
luseram et riseram,” with the English “I had sung
and danced and played and laughed”.

23. (11) In language, analysis means suppleness,
25and synthesis means rigidity; in analytic languages
you have the power of kaleidoscopically arranging
and re-arranging the elements that in synthetic forms,
like cantaveram are in rigid connexion and lead a
Siamese twin sort of existence. The synthetic forms
of Latin verbs remind us of those languages of South
America in which we are told that there is no word
for “head,” or “eye,” but only for “my head,” “your
head,” “his eye,” etc. 117 In one language the verbal
idea (in the finite moods), in the other the nominal idea
is necessarily fused with the personal idea. And if
Latin pater has the advantage over the American words
that it is not always limited to “my father,” or somebody
else's father, it is limited in other ways: it is
one definite number, one definite sex, one definite
case. It is more restricted in its use, more concrete
than necessary; and such a restriction is, or, under
certain circumstances, may be, a hindrance to freedom
or precision of thought In Swedish make, “mate,”
is masculine, and maka feminine; and Tegnér expressly
regrets this distinction, saying: “On account
of the impossibility of separating the stem mak- from
the ‘organically’ coalesced endings -e of the masculine
and -a of the feminine, we cannot give such a
form to the sentence ‘sin make må man ej svika'
as to make it perfectly clear that the admonition is
26applicable to both husband and wife”. 118 In this case
the Danes have advanced beyond their neighbours
by abolishing the distinction and using mage for both
sexes.

24. Most English pronouns make no distinction
of sex: I, you, we, they, who, somebody, etc. And
yet, when we hear that Magyar, and, indeed, the
great majority of languages outside the Arian and
Semitic world, have no separate forms for the masculine
and feminine pronouns of the third person,
that is, make no distinction between he and she, our
first thought is one of astonishment; we fail to see
how it is possible to do without this distinction. But
if we look more closely we shall see it is at times a
great inconvenience to be obliged to specify the sex
of the person spoken about. I remember once reading
in some English paper a proposal to use the word
thon as a personal pronoun of common gender; if it
was substituted for he in such a proposition as this:
“It would be interesting if each of the leading poets
would tell us what he considers his best work,” ladies
would be spared the disparaging implication that the
leading poets were all men.

Now, thon has no great chance of becoming popular,
and the proposal has hardly any significance except
as showing that the want of a genderless pronoun is
sometimes felt. And it is curious to see the different
ways out of the difficulty resorted to in the language
of daily life. First the cumbrous use of “he or she,”
27as in the following sentences: “Everybody to do
just as he or she likes” | Fielding, Tom Jones, i., 174,
“the reader's heart (if he or she have any)” |
Thackeray, Pendennis, iii., 294, “every woman and
man in this kingdom who has sold her or himself” |
G. Eliot, Mill on the Floss, i., 54, “each was satisfied
with him or herself” | Miss Muloch, John Halifax,
Gentleman, ii., 128, “each one made his or her comment”
| C. Doyle, Study in Scarlet, 66, “the murderer
has written it with his or her own blood”. 119 In
many cases he will be used alone in spite of the inaccuracy
which results: compare, for instance: “If
anybody behaves in such and such a manner he will
be punished,” with, “Whoever behaves in such and
such a manner will be punished”.

25. English who is not, like the quis or quæ of the
Romans, limited to one sex and one number, so that
our question “Who did it?” to be rendered exactly
in Latin would require a combination of the four:
Quis hoc fecit? Quæ hoc fecit? Qui hoc fecerunt?
Quæ hoc fecerunt? or rather, the abstract nature of
who (and of did) makes it possible to express such
a question more indefinitely in English than in
any highly flexional language; and indefiniteness in
30many cases means greater precision, or a closer
correspondence between thought and expression.

26. (11) The doing away with the old case distinctions
in English has facilitated many extremely
convenient idioms unknown in the older synthetic
languages, such as: “The girl was given a book” |
“the lad was spoken highly of” | “I love, and am
loved by, my wife” | “these laws my readers, whom
I consider as my subjects, are bound to believe in
and to obey” (Fielding, Tom Jones, i., 60) | “he
was heathenishly inclined to believe in, or to worship,
the goddess Nemesis” (ibid., ii., 165) | “he rather
rejoiced in, than regretted, his bruise” (ibid., iii.,
121) | “many a dun had she talked to, and turned
away from her father's door” (Thackeray, Vanity
Fair, 9) | “their earthly abode, which has seen, and
seemed almost to sympathise in, all their honour”
(Ruskin, Selections, i., 441). 120 Another advantage
is derived from the giving up of the distinctive
forms of the singular and plural in adjectives and
31adjectival pronouns, as is seen from a comparison of
the English “my wife and children” with the French
“ma femme et mes enfants,” or of “the local press and
committees” and “la presse locale et les comités
locaux”. Try to translate exactly into French and
Latin such a sentence as this: “What are the present
state and wants of mankind?” (Ruskin, loc. cit., 405).
In nouns, on the other hand, the two numbers are kept
apart in English., except in a very few words (deer,
sheep, series, cf. § 130). Danish has a somewhat greater
number of words that are alike in singular and plural;
but the advantage of having everywhere the same
indifference to number as is seen in English adjectives
or in Chinese nouns will appear from the words that
a Dane or an Englishman editing a text would use
to express the same idea: “et (singular) eller (or)
flere (plural) ord (indifferently singular and plural)
mangler her” — “some (singular and plural) word (singular)
or words (plural) wanting here”. Cf. also the
expression “a verdict of wilful murder against some
person or persons unknown” where some and unknown
belong to the singular as well as the plural forms;
and Fielding's phrase (Tom Jones, iii., 65): “Some
particular chapter, or perhaps chapters, may be obnoxious”.

27. (13) The languages we have here dealt with
tend evidently in their historical development towards
general instead of special forms; but inseparable
from this tendency is another, to get rid of the
rules of concord. It is a characteristic feature of the
32older Arian languages that the adjective is made to
agree with its substantive in number, gender, and
case, and that the verb of the predicate is governed
in number and person by the subject. The latter
form of concord has disappeared from spoken Danish.,
where, for instance, the present tense of the verb
meaning “to travel” is uniformly rejser in all persons
of both numbers; while the written language till
quite recent times kept up artificially the plural
rejse, although it had been dead in the spoken
language for some three hundred years. The old
inflexion is, to use Madvig's words, “an article of
luxury, as a modification of the idea belonging
properly to the subject is here transferred to the
predicate, where it has no business; for when we say
‘maendene rejse’ (die manner reisen), we do not
mean to imply that they undertake several journeys”. 121

In all of these cases the construction in Danish is
as easy and natural as it generally is in the English
past tense: “It was not her glass, but you that
flattered her”.

30. (14) The “luxury” which Madvig spoke of is
still more striking in the inflexion of nouns and
adjectives. If we compare a group of Latin words
such as opera virorum omnium bonorum veterum with
a corresponding group in a few other languages of a
35less inflexional type: Old English., ealra godra ealdra
manna weorc; Danish., alle gode gamle mænds værker;
Modern English., all good old men's works, we perceive
by analysing the ideas expressed by the several
words that the Romans said really: “work,” plural,
nominative or accusative + “man,” plural, masculine,
genitive + “all,” plural, genitive + “good,” plural,
masculine, genitive + “old,” plural, masculine, genitive.
Leaving opera out of consideration, we find that
“plural” number is expressed four times, “genitive”
case also four times, and “masculine gender”
twice; 122 in Old English the signs of number and
case are found four times each, while there is no
indication of gender; in Danish the plural number
is marked four times, and the case once. And finally,
in Modern English., we find each idea expressed only
once; and as nothing is lost in clearness, this method,
as being the easiest and shortest, must be considered
the best. Mathematically the different manners of
rendering the same thing might be represented by
the formulæ: anx + bnx + cnx = (an + bn + cn)x =
(a + b + c)nx.36

31. (15) This unusual faculty of “parenthesising”
causes Danish., and to a still greater degree English.,
to stand outside of Schleicher's definition of that
family of languages to which they historically belong;
for according to him “the Arian noun (and adjective)
as a living word can never be without a sign indicating
case”. 123 I shall here quote an interesting passage
from one of his books: “The radical difference between
Magyar and Indo-Germanic [Arian] words is
brought out distinctly by the fact that the postpositions
belonging to co-ordinated nouns can be dispensed
with in all the nouns except the last of the
series, e.g., a jó embernek ‘dem guten menschen’ (a
for az, demonstrative pronoun, article; jó, good;
ember, man; -nek -nak, postposition with pretty much
the same meaning as the dative case), for az-nak
(annak) Jó-nak ember-nek, as if in Greek you should
say το ἀγαθο ἀνθρώπῳ. An attributive adjective
preceding its noun always has the form of the pure
stem, the sign of plurality and the postposition indicating
case not being added to it. Magyars say,
for instance, Hunyady Mátyás magyar király-nak (to
the Hungarian king Mathew Hunyady), -nak belonging
here to all the preceding words. Nearly the
same thing takes place where several words are joined
together by means of and.” 22437

32. (15) Now, this is an exact parallel to the
English group genitive in cases like “all good old
men's works,” “the Queen of England's power,” “Beaumont
and Fletcher's plays,” “somebody else's turn,”
etc.; and as this peculiarity of English has developed
in comparatively recent times from a grammatical
construction analogous to the Latin concord (as will
be shown at some length in a subsequent chapter),
we may perhaps be entitled to ask, may not the
absence of concord in Magyar be a comparatively
modern simplification?. In other words, may not
the phenomena of concord be survivals from a primitive
stage of linguistic development? In undeveloped
minds we often find a tendency to be more explicit
than seems strictly necessary, as in the frequent
emphasising of a negation by seemingly redundant
repetitions. In Old English it was the regular idiom
to say: nan man nyste nan Þing, “no man not-knew
nothing”; so it was in Middle English., witness
Chaucer's (C. T. A., 70) “He neuere yet no vileynye
ne sayde In al his lyf unto no maner wight,” and so it
is in the vulgar speech of our own day: says Rob
Jakin (in The Mill on the Floss, i., 327), “There was
niver nobody else gen (gave) me nothin'”; whereas
standard Modern English is contented with one
negation: no man knew anything, etc. Concord
38seems to be a case in point, and this manner of
viewing it will gain in plausibility by the phenomena
of South African grammar treated in the opening of
the next chapter.

Chapter III.
Primitive grammar.

34. (16) Nowhere do the phenomena of concord
seem to grow more luxuriantly than in the languages
of those primitive South African tribes known under
the name of Bantu. I shall give some examples,
chiefly taken from the late W. H. I. Bleek's excellent
grammar; 125 when these interesting facts are
explained, we shall be able to draw some inferences
from them with regard to our own group of languages.

38. (17) By way of contrast I translate a passage
from an article by the German missionary, H.
Brincker: “Another characteristic feature is that
with these people eloquence generally consists, as
it seems, in the employment of a great number of
particles of one or more syllables, most of them untranslatable
and meaningless. What a torrent of
such waste-words (flickwörter) issues from the mouth
of a native orator! Any one who is not familiar
with the language is astonished to think how many
thoughts must have been developed, and yet, at least,
one-third of all the words pronounced were nothing
but those obscure particles, repeated over and over
again, while most of them might very well have been
left out without any loss to the purport of the speech.
Nevertheless, the natives attach a great importance to
the use of these particles”. 12944

This last remark of Brincker's shows that Livingstone
is right in saying that the prefixes are necessary
to the Bantu languages, and that the structure of
these languages is such that the omission of the prefixes
would involve obscurity and ambiguity. But
still Bleek is right in speaking of the repetitions as
cumbersome, just as the endings in the Latin multorum
virorum antiquorum are cumbersome, however
necessary and seemingly indispensable they were to
Cicero and his contemporaries.

39. (18) But what is the origin of this South
African system? The problem has not yet been
completely solved, though Bleek is very much inclined
to consider all Bantu nouns as originally
compound words. As long as each component part
is felt as relatively independent, it is natural, he
argues, that the first part of the compound, which
according to the structure of Bantu languages corresponds
to the last element of our compound words,
should be used as a representative of the whole word.
Bleek illustrates this by means of English examples:
the last syllable of the compound word steamship
might be used to represent the whole word; and thus,
after once mentioning “the steamship,” we might
continue “our ship, which ship is a great ship, the
ship appears, we love the ship”. But in words where
the syllable ship is a derivative suffix, it is incapacitated
45from being used by itself for the purpose of representing
the whole compound noun. Thus in reference
to the word “friendship,” it would be absurd to
continue “our ship, which ship is a great ship, the
ship appears, we love the ship”; but that is just what
the Zulus do, even extending the use to cases in
which the Zulu “derivative prefix” is as little felt to
be an independent element as, say, the -er of steamer.
This is as if in reference to “the steamer,” we should
continue “our er, which er is a great er, the er
appears, we love the er” (p. 107). Bleek very carefully
investigates the several classes of nouns in all
the cognate languages, in order to determine from
the meanings of the words belonging to each class
the original signification of the corresponding prefix,
but he himself acknowledges that great difficulties
attend this task; the want of old literary documents
makes the whole investigation uncertain, as “will be
easily understood by any one who may have tried
to ascertain the original meaning of such English
suffixes as -dom, -ness, -ship, etc., from an analysis of
the nouns formed with them. A comparison of such
nouns as ‘kingdom, martyrdom, freedom,’ etc., may
give us an idea of the present value of the suffix -dom,
and of the meaning which it would give to such
nouns as we can now form with it. But this is a
very different thing from knowing what was the
meaning of the syllable -dom when used independently;
and we imagine that any guess at that
meaning, without tracing it back historically, might
46be far from the truth” (p. 125). I shall mention a
few points of interest in his disquisition.

40. (19) The fifteenth class is characterised by
ku; we saw above ukutanda., “love”. This is identified
by Bleek (126) with the preposition ku, which
corresponds to English to, both in the local meaning,
as in ngi-ya-ku-laba-bantu, “I go to these people,”
and before infinitives, as in ngi-ya-kutanda, which is
literally “I go to love,” and is used as a kind of
future (comp. I am going to love, or je vais aimer); in
u-ku-tanda ku-mnandi, “to love is sweet,” the first ku
is used as a derivative prefix, to which the second ku
refers as a pronoun. Here I may be allowed to
insert an interesting parallel; if such a word as
ukutanda has been named in a previous proposition,
and you want to introduce it later on, say as the
object of some verb, this is achieved by repeating ku
instead of ukutanda (cf. the last sentence in § 35),
exactly as in modern colloquial English., instead of
repeating an infinitive, you may content yourself with
using to as a substitute for it. 13047

41. (20) To the fifteenth Bantu class belong first
the unmistakable infinitives and some words in which
the verbal idea is still more or less easily discernible,
such as ukuchla., “food” (really “eating”), and ukusa,
“morning” (“dawning”), and, secondly, a number of
words which cannot have been originally infinitives;
in many of these, meaning “desert,” “field,” “open
place,” “winter,” “rainy season,” or some other
particularisation of place or time, Bleek says that
“the common origin of the prefix ku- and the preposition
ku- (to) is almost evident”. But whether
we take this “common origin” to mean a development
of the prefix from an original preposition, as
Bleek seems to think, or the development of the
preposition and the noun prefix from some common
source, in any case a good many nouns remain in the
class in the case of which no connexion can be traced
between the meaning of the noun and any of the
different meanings of the preposition. And this
difficulty in seeing reasons for a noun belonging to a
particular class and to no other is still greater in all
the other classes, where it is often nearly impossible
to perceive anything common to all or to most of the
nouns in the class. Nay, where we are able to find a
connexion, it seems in many cases to be a derived
and not an original one; thus a great many names
for living beings are comprised under the first class;
48but it is probable that they were originally adjectives
referred to umuntu and therefore taking the umu-prefix,
which they subsequently kept even in cases
where they were not joined to any umuntu (Bleek,
p. 123).

42. (21) In several of the classes the words have a
definite numerical value, so that they go together in
pairs as corresponding singular and plural nouns (see
the examples above): but though in the more advanced
languages this is carried out pretty regularly,
the existence of a certain number of exceptions shows
that these numerical values cannot originally have
been associated with the class prefixes, but must be due
to an extension by analogy (Bleek, p. 140 sqq.). The
starting-point may have been substantives standing
to each other in the relation of “person” to “people,”
“soldier” to “army,” “tree” to “forest,” “ship” to
“fleet” (ibid., 144); the prefixes of such words as the
latter of each of these pairs will easily acquire a certain
sense of plurality, no matter what they may have
meant originally, and then they will lend themselves
to forming a kind of plural in other nouns, being
either put instead of the prefix belonging properly to
the noun (amazwe, “countries,” 6; ilizwe, “country,”
5), or placed before it (ma-luto, “spoons,” 6 ; luto,
“spoon,” 11). Sometimes we find that instead of
being regulated by the class to which the subject
belongs grammatically, the verb, etc., takes the reminders
of some other class by some constructio ad
sensum, just as in German the “reminder” sie may be
49used in referring to such neuter nouns as weib or
mädchen; instead of erumbi rándye ekúru ráya, “my
eldest brother” (5th class) “has gone away,” you may
hear erumbi uándye omukúru uáya., where the reminders
are of the first class (Bleek, p. 156, note).
As has been mentioned above, the first class comprises
a great many words signifying living beings.

43. (22) Thus an impulse is given to further
deviations and changes; and we are told (Bleek, p. 234)
that in the north-western branch of the Bantu
languages “the forms of some of the prefixes have
been so strongly contracted as almost to defy
identification. Thus prefixes may have been confounded
with each other, and correspondences differing
from the original ones may have arisen through
the force of analogy. At the same time, the concord
appears to be frequently employed in the north-western
languages rather as an alliterative process,
than in its original grammatical sense, or as a division
of nouns into classes”. In one of the languages we
have a two-hundred-year-old grammar by Brusciotto
à Vetralla (see Bleek, i., 9). A comparison of the
language described there with that spoken now-a-days
in the same district (Mpongwe) shows that the class
signs have dwindled down considerably; instead of a
whole syllable we have as a rule only a vowel left;
the phonetic shrinkage has been stronger in these
grammatical elements than elsewhere; 131 the number
50of the prefixes and consequently of the classes has
been reduced from 16 to 10: for instance, classes
11, 14 and 15 have been phonetically amalgamated
(Bleek, 223; cf. 132).

44. (23) Here I shall say good-bye to Bleek and
shall try to obtain from these South African phenomena
some results bearing on the development of
languages in general, and in particular of languages
nearer home than those of South Africa. The reader
will then, I hope, understand that it was not out of
mere caprice that I undertook my rambling excursion
to those far-off regions.

From the historical fact pointed out in the last
section we may safely infer that if we were able to
make acquaintance with the South African languages
at a still earlier epoch, we should meet with a still
greater number of classes than sixteen; and moreover,
that the reminders we should see prefixed to adjectives
and verbs would be still fuller in form.and more
51like whole words. And, maybe, we should then be
still more inclined to doubt the correctness of Bleek's
view, according to which every Zulu noun was originally
a compound word, whose first element was repeated
with the following words of the sentence. He
seems not to have proved or even rendered it probable
that there either is or has been so great a partiality
to composition that all non-compound words should
have disappeared from the language. It would be
very strange indeed if it were so.

46. (24) So far, then, we seem to be on sure
ground. Neither does there seem to be anything
53rash in assuming that accentual law to hold good in
man's first language which we find everywhere in our
own times and which is formulated by Sweet as
follows: 132 “All words that express new ideas are
more or less emphatic; while words that express
ideas already familiar or that can be taken for
granted are unemphatic”; if we begin a story with
the words “A German came to London,” we give
stronger stress to German and London, than when we
go on “. . . the German left London, and went to
Liverpool”. And this feebler stress has as a consequence
a less distinct pronunciation of each of the
sounds making up the words.

47. (24) Add to this another tendency found in
all languages, as far as I am aware, that of shortening
frequently repeated words, especially compound
words when they are no longer felt as compound
words, the meaning being associated with the word
as a whole rather than with the several parts: when
this is the case, it is of no consequence to the speaker
that etymologically the word is, i.e., once was, a compound.
This shortening takes place extremely often
in proper names; 233 in Greek, we see a great many
abbreviations used as pet-names, e.g., Zeuxis for
Zeuxippos, Zeuxidamos, Zeuxitheos, etc., so in Old
High German Wolfo stands for Wolfbrand, Wolfgang,
etc. Icelanders say Sigga for Sigríđr, Siggi for
54Sigurđr, and so on in most languages. Abbreviations
of this character do not belong to any particular time
or to any particular country; they grow luxuriantly
everywhere, and are not at all confined to children's
language or to those cases which are sanctioned by
tradition, like Rob, Jim, Dick, etc. Thus, in the
beginning of this century Napoleon Bonaparte was
generally called Nap or Boney; and Thackeray
constantly says Pen for Arthur Pendennis, Cos for
Costigan, Fo for Foker, Pop for Popjoy, old Col for
Colchicum, etc., etc. This is quite natural; wherever
a person is often spoken of, the speaker is understood
by everybody before he is half through the name,
if it is a rather long one, and therefore he often does
not take the trouble to pronounce the latter part of it.
He thus exemplifies the principle we meet with everywhere:
people do not pronounce distinctly unless they
feel that distinctness is necessary if they are to be
understood; whatever is easily understood from the
context or from the situation is either slurred over or left
out completely. 134 This principle will account alike for
55most of the gradual sound changes in languages, and
for such violent curtailings as cab for cabriolet, caps
for capital letters, the Cri for the Criterion, phiz for
physiognomy, sov for sovereign, or French aristo for
aristocrate, Boule-Miche for Boulevard St. Michel, and
so on. 135

48. (24) Now I fancy it must have been by the
same process that the Bantus have arrived at the
use of umu as a representative of umuntu; the tendency
to use a half-word in this manner may have
been strengthened by the fact that in some cases a
word was felt as a compound, so that the first part
of it could be used independently.

However this may be, so much is certain, that in
these languages we see the origination of pronouns
by natural means; whether Bleek is right
in regarding the beginnings of words as first parts
56of compounds, or whether they stand for complete
words, they are originally nouns, “full words” (not
“demonstrative roots”); and in their function as what
I have called reminders they correspond to pronouns
in our languages; for what else are many pronouns
(especially the personal pronoun of the third person,
the relative, and some of the demonstrative pronouns)
but signs to remind us of what has been mentioned
before?

49. (25) Further, we witness the origin of other
grammatical forms, that are to be classed partly
with the flexional forms of nouns and adjectives
(“our” = wetu when referred to umuntu, but letu
when referred to ilizwe, which is much like Arian
gender; in § 36 we saw something corresponding
to our genitive), partly with verbal endings. And
it should be remembered that we see these forms
come into existence quite naturally from a more
primitive and thoroughly concrete state of language,
without any intention on the part of speakers to
create anything new. They only indulge in the
universal inclination to save oneself trouble, that is,
in this case, to pronounce as few sounds as is compatible
with making oneself understood.

50. (25) Finally we see the development of something
that may be compared to our article; for as
umu was used with other words as a reminder of
umuntu, people seem to have come to look upon it
as a reminder in the word umuntu itself, which was
accordingly understood as umu (a sort of class-sign
57to indicate the grammatical construction of the word,
like the German der, die, das) + ntu, which thus
appropriated to itself the meaning “man”. This
shifting of the popular linguistic conception of the
constituent elements of a word is analogous to the
popular misdividing of anatomy in English into an-atomy,
an being taken for the article, as in an atom,
an attic, etc., and being subsequently subtracted
(“the atomy”); or that of acute into a + cute, the word
cute then being deduced from it. In the ending of
words we see very frequently the same process; a
few centuries ago pease was both singular and plural,
corresponding to Old English singular, pise, plural,
pisan or piosan; then the s was regarded as the
common plural ending and subtracted so as to form
the new singular a pea, which is not found in Shakespeare,
and which is mentioned by Butler (a.d. 1633)
as a cockneyism; in the same manner cherry is for
cherris (cf. French cerise), riddle for riddles (Old
English rædels), and there are many other cases. 136
Now, the same process of subtraction seems to have
obtained, or at any rate to be now in operation, in
Bantu languages, as lexicographers enter the word
which I have mentioned so often, not in the form
umuntu, but as ntu; it is true that Bleek protests
against this division of the word; 237 but if he is
58right from an etymological point of view, he is
perhaps wrong from the point of view of the actual
linguistic instinct of the natives.

51. (25) If I sum up by emphasising the fact
that in the Bantu languages this development of
grammatical signs and categories has gone on indirectly
and through a shortening of longer word-forms,
and not through an extension of shorter words
by means of formal elements, the reader will see how
this long — perhaps too long — disquisition has some
bearing on the comparative grammar of Arian; for
the results arrived at go dead against a great many
of those explanations of the origin of Arian forms
which have hitherto been given by philologists.

52. (26) Madvig's philosophy of language was
on the whole rationalistic; but he certainly in many
respects exaggerated the intellectual faculties of “the
creators of language,” 138 as will be seen very strikingly
in the following passage: “Gender in languages was
created by those who first hit upon (and adopted the
habit of) keeping some particular phonetic modification
of the demonstrative pronoun to indicate the
special shade of signification of the noun; in our
family of languages by those who added to the
pronominal stem the open and soft vowel sound
[a]… the quality of feminine being expressed by
the more soft, open and lingering close of the uttered
59sounds”. Madvig himself had an impression that he
had here resorted to the method of explanation by
means of sound symbolism, of which, he is usually a
fierce (in my opinion, too fierce an) antagonist, for he
says by way of apology: “Such an origin, in which
the character of the sound had a meaning and imparted
it, we must specially imagine for ourselves in
this case rather than in dealing with the formation of
other primitive utterances, because we have not here
to do with the name of some definite conception, but
with a general modification, with the influence of an
incidental condition”. Now, I must confess that I
can more easily imagine to myself primitive man
hitting on a new sound to picture to his ear an
entire perception which impressed him, but which his
language was too poor to express, than fancy him
adding an a to an existing genderless pronoun, in
order thereby to denote the delicacy of das ewigweibliche.
And even if such a conceit might once
come into his head, it is somewhat doubtful if his
contemporaries would be able to see the drift of his
long a and make an appropriate use of it with their
own pronouns.

53. (27) Equally unsatisfactory are many other
explanations that have been put forward by comparative
philologists in their fondness for constructing
hypotheses concerning primitive ages. Indeed, the
history of comparative philology shows how very
short-lived many of these explanations are: here today
and in the waste-paper basket to-morrow! To
60show what sort of hypotheses I am alluding to, I
shall have to quote somebody, for fear people should
say that I am tilting at windmills; and I take a
paper by the clever Norwegian philologist A. Torp, 139
for no other reason than because it is the last paper
of this description that has come into my hands; I
shall add a few criticisms within brackets. He says:
“The common Indo-Germanic [Arian] language
possessed several declensions; but it is a priori
improbable that this should be the original state of
things. The plurality seems necessarily to have
developed out of an earlier unity. [Experience in
historical times, in our family of languages as well as
in that of South Africa, speaks rather in favour of a
development in the opposite direction, from multiplicity
towards comparative uniformity.]… Among the
most primitive elements of language I reckon
particularly those stems that are seemingly formed
from the verbal root by means of the suffix -o, both
on account of the simplicity of their formation [simple
things are pretty often of quite recent growth], their
indefinite signification, as nomina agentis, as denoting
products, as abstract terms, etc. [this is no decisive
proof, the word “abstract” must create suspicion, if
nothing else], and their number. [If the old languages
of our family were dead, it would be possible by
means of the same arguments to prove that the weak
61verbs in English were the most primitive.]…
Later on, certain endings were joined to these stem
forms, the language intending thereby a more
definite denoting of the case relations (case endings).
[Language neither can nor does intend anything;
those who speak it intend nothing but to be understood
at the moment; therefore they do not add
anything to denote more definitely something of
which they can have no notion.] These case endings
have long been justly looked upon as consisting of
pronominal stems. [It is possible that this may turn
out in the end to be the correct view; but hitherto
there is not one single ending with regard to which it
has been shown with any degree of probability how a
pronoun could modify the meaning and function of a
noun in that particular way.]… Thus -s in the
nominative singular is certainly the same pronominal
stem as that which is used as a demonstrative pronoun
in the form so, se [but what is the origin of so, se
itself?]; the -m of the accusative is the same element
as that found in me, the pronoun of the first person,
which was in all probability originally a demonstrative
pronoun also. [Would it not be safer to confess that
one has not the slightest idea of the derivation of -m
than to bring forward an explanation presupposing
violent changes of sense, without making the least
attempt to commend such an assumption by adducing
hypothetical connecting links?]… the ā- stems are,
I fancy, formed by an element -ā, which was, I
suppose, properly a pronoun used to denote the
62feminine gender, being added to the o- stems before
these had yet adopted case endings”.

54. (28) If theories about the origin of things are
not to be worthless, they must on every point be
substantiated by analogies from processes going on
now-a-days, and capable of direct observation and
control. We must, accordingly, ask ourselves: Do
we ever witness the genesis of any new flexional
endings or similar elements? If we do, we cannot
be far wrong in thinking that those formal elements
of language whose origin lies far back in pre-historic
times, must have arisen in similar ways and through
the same agencies.

Now, there is one method of accounting for the
genesis of the elements we are here speaking of,
which seems so natural and obvious that it is no
wonder that very extensive use has been made of
it ever since the first beginnings of comparative
philology, namely, the agglutination theory. According
to this theory two words, originally independent
of each other, so often stand together that at length
they are combined into one indissoluble unity; one
of the two gradually loses its stress, and finally becomes
nothing more than a suffix of the other.
Thus, without the least doubt, the Scandinavian
passive voice originates in an agglutination of the
active verb and the pronoun sik; Old Norse, Þeir
finna sik, “they find themselves,” or “each other,”
gradually becomes one word, Þeir finnask; SwediSh.,
de finnas; Danish., de findes, “they are found”.
63Similarly the future tense of the Romance languages:
Italian, finirò; French, je finirai, “I shall end,” from
finire habeo (finir ho, finir ai), “I have to end”. The
Scandinavian suffixed article is a third case in point,
if we are allowed to consider it as a kind of flexion:
Old Norse, mannenn (manninn), accusative; Danish.,
manden, “the man”; Old Norse, landet (landit);
Danish., landet, “the land,” for original mann, land
+ the demonstrative pronoun enn, neuter et. 140

55. On the strength of these formations it has
been concluded that all derivative and flexional endings
had a similar history, that is, they were all
independent words before they became agglutinated
to, and fused with, the main word. This is the
theory prevalent among all the leading linguists, not
only of the times of Bopp and those of Schleicher,
but also of quite recent days. Thus Whitney says:
“Suffixes of derivation and inflexion are made out
of independent words, which, first entering into union
with other words by the ordinary process of composition,
then gradually lose their independent character,
and finally come to be, in a more or less mutilated
and disguised form, mere subordinate elements, or
indicators of relation”. And again: “The grand
conclusion, however, at which historical study has
surely and incontrovertibly arrived, is that all the
grammatical apparatus of languages is of secondary
growth; the endings of declension and conjugation,
64the prefixes and suffixes of derivation, were originally
independent elements, words, which were first
collocated with other words, and then entered into
combination and were more or less thoroughly fused
with the latter, losing their primitive form and meaning,
and becoming mere signs of modification and
relation; hence, that the historically traceable beginnings
of speech were simple roots; not parts of
speech even, and still less forms”. 141

H. Paul says: “The strictly normal origin of
all formal elements in language is always composition”;
and in criticising the particular manner in
which this process has been supposed to work, he
still assumes the truth of the general theory: “the
first foundation of derivation and of flexion was
created by the coalescence of originally independent
elements; but then, as soon as these foundations
had come into existence, they had to serve as patterns
for formations by analogy”. 242

Brugmann says: “What is included under the
names of stem-formation and flexion depends on a
uniting and more or less close fusion of originally
independent elements”. 343

G. v. d. Gabelentz expresses himself to the same
effect: “As far as authenticated facts of linguistic
history go, all external expedients of derivation and
65accidence originate in agglutination, that is, in the
adding of originally independent words”. 144

Similar expressions might be adduced from other
eminent philologists, such as Tegnér (who holds
that the transition from agglutination to flexion
constitutes a retrogression), Sweet, and Herman
Möller. 245

56. Now, of course it cannot be denied that similar
processes may have been going on at any time, and
that some flexional forms of old Arian may have
arisen in this way. But when the inference is that
they are all to be explained in this manner, and that
here we have the key to flexion in general, great exception
may be taken. First, the number of actual
forms proved beyond a doubt to have originated
through agglutination is very small; the three or
four instances named above are everywhere appealed
to, but are there so many more than these? And are
they numerous enough to justify so general an assertion?
Secondly, these three or four instances can,
at any rate, prove nothing as to the genesis of flexion
in general from agglutination preceded by isolation;
for in all of them the elements were fully flexional
before the fusion (cf. Ital., amerò, amerai, amerà,
etc.; Old Norse, finnask, fannsk; mađrenn, mannenn,
mansens). What they show, then, is really nothing
66but the growth of new flexional formations on an old
flexional soil. Thirdly, it may be objected to the
theory that, assuming it to be true, we should expect
much more regular forms than we actually find in the
old Arian languages; for if one definite element was
added to signify one definite modification of the idea,
we see no reason why it should not have been added
to all words in the same way; as a matter of fact,
the Romance future, the Scandinavian passive voice
and definite article present much greater regularity
than is found in the inflexion of nouns and verbs in
old Arian.

57. (28) And finally, the agglutination theory must
cease to be thought the only possible way of accounting
for the origin of flexional endings, as soon as
we are able to point out certain endings which
undoubtedly have originated in quite a different
manner. Such endings, however, are -en in English
oxen, German ochsen, and -er in German rinder,
lämmer. Here originally -en and -er belonged to the
word through all cases and all numbers; ox was an
n-stem in the same way as, for instance, Latin
homo(n), hominem, hominis, etc., or Greek kuõn,
kuna, kunos, etc., are n-stems; cf. Sanskr. ukšan-; and
the other words were originally es- and os-stems,
comparable to Latin genus, generis from older genesis,
Greek genos, gene(s)os, for original s develops
regularly through z to r in the Germanic languages,
whenever it is preceded by a weak vowel. No one,
in considering the Latin forms homines or genera,
67would dream of the possibility of the syllables in (en)
and er becoming the sign of the plural, when the
same syllables appeared in the singular as well. Yet,
in Germanic, where the declension was originally
strictly analogous to that of Latin, this has actually
come to pass: the final syllables of the nominative
and accusative singular were dropped by a regular
phonetic change, while in the plural n and r were
kept because they were protected by a following
syllable, which had first to be worn away. The
result is that now plurality is indicated by an ending
which had formerly no such function (which indeed
had no function at all); for if we look upon the actual
language, oxen is = ox (singular) + the plural ending
-en, and similarly rinder = rind (singular) +
the plural ending -er; only we must not on any
account imagine that the forms were originally
thus welded together (agglutinated). 146 Compare also
the history of the English possessive pronouns; Old
English min and Þin keep the n throughout as
forming part and parcel of the words themselves;
but in Middle English the n is dropped first before
nouns beginning with a consonant (my father–mine
uncle; it is mine), and then before a vowel as well,
68but only when the pronouns are used attributively
(my father, my uncle — it is mine). The distinction
between my and mine, thy and thine, which was
originally a purely phonetic one, like that between a
and an, gradually acquires a functional value, and
serves to distinguish a conjoint from an absolute
form; and as the former was the more commonly
used, it came to be looked upon as the proper
form, while the n of mine was felt as an ending
serving to indicate the absolute function. That
this is really the instinctive feeling of the people
is shown by the fact that in dialectal and vulgar
speech the n is added to his, her, your, and their, to
form the absolute pronouns hisn, hern, yourn, and
theirn.

58. (29) If we apply such considerations to the
forms of primitive Arian speech, we shall be led to
a change of front similar to that made in historic
phonology when, instead of the i of the Greek elipon
being considered as the root vowel and the ei of leipo
as a strengthening of i, ei began to be taken as the
original and fuller form, of which i was a weakening.
And where the old school could only imagine language
taking the most direct course possible we must realise
the fact that it often takes the most unexpected
round-about ways to reach its goal. It cannot but
be beneficial always to remember that the signification
borne at one time by a word or a word-element
is very often widely different from the
original one, and that sometimes an element which
69had primarily no signification at all may gradually
acquire a signification of great importance. Many
endings may have acquired their special modifying
force in a way analogous to that seen in the French
pas. In the oldest French, ne alone is sufficient to
express the negation; then it became habitual to
strengthen the negation by the addition of such
superfluous words as pas, “a step,” goutte, “a drop,”
mie, “a crumb,” or the like, just as we say in English.,
“not a bit, not a scrap”. Pas became the most
common of these expletives; and little by little it
grew to be as indispensable in most sentences as the
ne itself. Nay, now it is even more so, for pas has
so completely appropriated to itself the negative
meaning as to be used for “not” wherever there is
no verb in the sentence (Pas de ça!) and in the
colloquial style even with a verb, the word which
originally carried the negative meaning being entirely
ousted (C'est pas vrai!). A similar indirect
course has been taken by the French jamais,
“never, which now means the exact opposite of
its etymological value (Latin jam + magis, “now +
more”).

59. (29) Many signs of the times seem to presage
a change of front in the modern science of language.
Numerous cases of agglutination formerly accepted
have been proved by modern criticism to be untenable;
nobody now thinks that the Germanic weak preterite
is a compound of did (loved = love did), or that the r
of the Latin passive is a disguised se; and after
70Prof. Sayce's attack 147 it seems no longer possible to
derive the person-endings of the verbs from personal
pronouns. There is decidedly a growing disinclination
to bring forward the kind of explanations by agglutination
which were formerly so rife: not a few philologists
carry positivism to the length of rejecting as mere
metaphysical speculation any attempt at explaining
the old forms; and the fresh explanations which are
now given by the masters of the science of language
are most of them indirect ones. I shall illustrate
this by referring briefly to a few important investigations
of recent date.

60. (30) The first of these is by the chief of the
Leipzig school of philology, Karl Brugmann. In
his paper Das Nominalgeschlecht in den indogermanischen
Sprachen, 248 he puts the question: How
did it come about that the old Arians attached
a definite gender (or sex, geschlecht) to words like
foot, head, house, town, the Greek pous, for instance,
being masculine, kephalē feminine, oikos masculine,
and polis feminine? The generally accepted explanation,
according to which the imagination of mankind
looked upon lifeless things as living beings, is, Brugmann
says, unsatisfactory; the masculine and feminine
as grammatical genders are merely unmeaning forms
and have nothing to do with the ideas of masculinity
71and femininity; for even where there exists a natural
difference of sex, language often employs only one
gender. So in German we have der hase, die maus,
and “der weibliche hase” is not felt to be self-contradictory.
Again, in the history of languages we often
find words which change their gender exclusively on
account of their form. 149 Nothing accordingly hinders,
us from supposing that grammatical gender originally
meant something quite different from natural
sex. The question, therefore, according to Brugmann,
is essentially reduced to this: How did it
come to pass that the suffix -a was used to designate
female beings? At first it had nothing to do with
femininity, witness the Latin aqua, “water, etc.; but
among the old words with that ending there happened
to be some words denoting females: mama, “mother”
(also with the meaning of “mother's breast,” Latin
mamma, French mamelle, and “aunt,” German
muhme), and gena, “woman” (compare English
quean, queen). Now, in the history of some suffixes,
we see that, without any regard to their original
etymological signification, they may adopt something
of the radical meaning of the words to which they
are added, and transfer that meaning to new formations.
In this way mama and gena became the
starting-point for analogical formations, as if the
72idea of female was denoted by the ending, and new
words were formed, e.g., Latin dea, “goddess,” from
deus, “god”; equa, “mare,” from equus, “horse,” etc.
Other suffixes probably came to denote “feminine
sex” by a similar process.

61. (30) On account of the nature of the subject,
Brugmann's investigation is more convincing in its
negative criticism than in its positive conclusions.
It must decidedly be greeted as a wholesome change
that he does away with such explanations as those
above mentioned, according to which the a in equa,
etc., was a pronoun or a phonetic modification signifying
the feminine quality, and having signified this
from the very beginning. The division of Arian
nouns into three genders, and the concord which is
a consequence of that division (adjectives, etc., being
made to agree with their nouns in gender), is, in
fact, nothing but a class division analogous to that of
the Zulu language described above. The analogy
will be still more striking if we compare Arian,
not with Zulu, but with the neighbouring, but
totally unconnected, Hottentot language, for there
a class division has been employed to distinguish
natural sex which had nothing to do with sex
originally. 150

62. (31) The second instance of the beginning
tendency towards indirect explanations of grammatical
phenomena which I shall quote, is the important
and learned book by the Berlin Professor Johannes
73Schmidt: Die Pluralbildungen der indogermanischen
Neutra (Weimar, 1889). In this work Schmidt
conclusively proves what before him some scholars
had suspected, 151 namely, that the common Arian plural
in -a was originally neither neuter nor plural, but, on
the contrary, feminine and singular. The forms in -a
are properly collective formations like those found,
for instance, in Latin, opera, -æ “work,” comp. opus,
“(a piece of) work”; Latin, terra, “earth,” comp. Oscan,
terum, “plot of ground”; pugna, “boxing, fight,”
comp. pugnus, “fist”. This explains among other
things the peculiar syntactic phenomenon, which is
found regularly in Greek and sporadically in the
Asiatic branch of the Arian family, that a neuter
plural subject takes the verb in the singular. The
Greek toxa is often used in speaking of a single bow;
and the Latin poetic use of guttura, colla, ora, where
only one person's throat, neck, or face is meant,
points similarly to a period of the past when these
words did not denote the plural. We can now also
see the reason of this -a being in some cases the
plural sign of masculine nouns: Lat., loca from
locus, joca from jocus, etc.; Gr., sita from sitos. Joh.
74Schmidt refers to similar plural formations in Arabic;
and we may call to mind our friends the Bantus,
whose plural prefixes were, as we have seen, originally
no more signs of plurality than the Arian
-a. And thus we are constantly reminded that language
must often make the most curious détours to
arrive at a grammatical expression for things which
appear to us so self-evident as the difference between
he and she, or that between one and more.
Simplicity in linguistic structure — that is, expressive
simplicity — is not a primitive, but a derived
quality.

63. (32) Comparative philology did not attain a
scientific character till Rask and Bopp established
the principle that the relationship of two languages
had to be determined by a thorough-going conformity
in the most necessary parts of language,
namely, besides suffixes and similar elements incapable
of independent existence, pronouns and
numerals, and the most indispensable of nouns and
verbs. But if this domain of speech, by preserving
religiously, as it were, the old tradition, affords infallible
criteria of the near or remote relationship of
different languages, may we not reasonably expect
to find in the same domain some clue to the oldest
grammatical system used by our ancestors? And
what sort of system do we then find there? We see
such a declension as I, me, we, us: the several forms of
the “paradigm” do not at all resemble each other,
as they do in more recently developed declensions;
75we find masculines and feminines such as father,
mother; man, wife; bull, cow; while such methods of
derivation as are seen in count, countess; he-bear, she-bear,
belong to a later time; we meet with verbal
flexion such as appears in am, is, was, been, which
forms a striking contrast to the more modern method
of adding a mere ending while leaving the body of
the word unchanged.

64. (33) The general impression left by these and
many similar instances is, that the grammatical
system of our remote ancestors was, to say the least
of it, very unsystematic and far from simple. Things
which belong, or to us would seem to belong, closely
together, were widely sundered as regards their
linguistic expression. And it is only by a slow and
gradual development that conformity and regularity
are brought about, especially in those words which
are in most constant use. The rarer a word is, the
more difficult it is to remember its several forms
unless they resemble one another; accordingly, rare
words are more exposed to being accommodated on
the spur of the moment to the most regular patterns
of inflexion. These regular patterns being more
present to the speaker's mind, he pays no regard to
the fact that the word in question “ought properly to be
irregular”. Nor is it the rarer words alone which
are reduced to rule: even in the case of the more
frequently recurring words the levelling influences
are at work; a greater and greater number of cases
will run together, and irregularities will gradually
76disappear. Those little words which are used every
minute, pronouns and so on, are uttered and heard so
very often that their forms acquire an extreme power
of resistance. And yet, even in these words we
observe the great work of simplification going on.
Let us take one of the clearest instances of all. The
flexion of the second personal pronoun, which was
universal in English some four hundred years ago,
namely, nominative singular thou, accusative singular
thee, nominative plural ye, accusative plural you,
has now in ordinary conversational and prose
language given place to perfect simplicity and uniformity:
nominative singular you, accusative singular
you, nominative plural you, accusative plural you.
But if we look closer into the history of this important
change, which will form the subject of a subsequent
chapter, we shall see that a great many most
widely different circumstances (phonetic, syntactic,
and social) have concurred to produce so complete a
revolution.

65. (33) To turn to the case of nouns, we cannot
imagine even in the most primitive grammar such
violent flexional changes as that seen in I, me, where
a totally different root is needed. Nevertheless, we
find in the oldest Arian languages plenty of comparative
violent changes taking place in the declensions,
as when different cases of the same noun have
different accentuation and different gradation (ablaut);
or as when in some of the most frequently occurring
words some cases are formed from one “stem” and
77others from another. Thus in the common Arian
word for “water,” Greek has preserved both stems:
nominative hudōr, genitive hudat-os, where a stands
for an original n or (ə)n which appears in some of
the other related languages. Whatever the origin of
this change of stems, 152 it is a phenomenon belonging
only to the earlier stages of our languages; 253 in the
later stages we always find a simplification, one
single form running through all cases; this is either
the nominative stem, as in English water, German
wasser (corresponding to Greek hudōr), or the oblique
case-stem, as in the Scandinavian forms, Old Norse
vatn, Swedish vatten, Danish vand (corresponding to
Greek hudat-), or finally a contaminated form, as in
the name of the Swedish lake Vättern (Noreen's
explanation) or in Old Norse and Danish skarn,
“dirt,” which has its r from a form like the Greek
skōr, and its n from a form like the Greek genitive
skatos (older skəntos). The simplification is carried
furthest in English., where the identical form water is
not only used unchanged where in the older languages
different case-forms of the noun would have been
used (the water is cold; he drinks water; the surface
78of the water; he fell into the water; he swims in
the water), but also where it serves as a verb (did
you water the flowers?) or as a quasi-adjective (a
water melon, water plants). We see here an approach
to the Chinese type of speech which we shall glance at
in the next chapter.79

Chapter IV.
The history of chinese and of word-order.

66. (34) In Chinese each word consists of one
syllable, neither more nor less. The parts of speech
are not distinguished: tá means, according to circumstances,
great, much, magnitude, enlarge. Grammatical
relations such as number, person, tense,
case, etc., are not expressed by endings and similar
expedients; the word in itself is invariable. If a
noun is to be taken as plural, this as a rule must be
gathered from the context; and it is only when there
is any danger of misunderstanding, or when the
notion of plurality is to be emphasised, that separate
words are added, e.g., kì, “some,” šú, “number”. 154
The most important part of Chinese grammar is that
dealing with word-order: tá kuok = “great state,” or
“great states”; but kuok tá means “the state is great,”
or, if placed before some other word which can serve
as a verb, “the greatness (size) of the state”; tsi niu
“boys and girls,” but niu tsi “girl” (female child),
etc. 255 Besides words properly so called, or, as the
Chinese grammarians term them, “full words,” there
80are several “empty words” serving for grammatical
purposes, often in a wonderfully clever and ingenious
way. Thus čī 156 has besides other functions that of
indicating a genitive relation more distinctly than it
would be indicated by the mere position of the words;
mîn (people) lik (power) is of itself sufficient to signify
“the power of the people,” but the same notion is
expressed more explicitly by mîn čī lik. The same
expedient is used to indicate different sorts of connexion;
if čī is placed after the subject of a sentence
it makes it a genitive, thereby changing the sentence
into a sort of subordinate clause: wâng paò mîn =
“the king protects the people”; but if you say wâng
čī paò mîn yéû (is like) fú (father) čī paò tsi, the whole
may be rendered, by means of the English verbal
noun, “the king's protecting the people is like the
father's protecting his child”. Further, it is possible
to change a whole sentence into a genitive; for
instance, wâng paò mîn čī taò (manner) k'ò (can) kién
(see, be seen), “the manner in which the king protects
(the manner of the king's protecting) his people is to
be seen”; and in yet other positions čī can be used
to join a word-group consisting of subject and verb,
or of verb and object, as an attribute to a noun; we
have participles to express the same modification of
the idea: wâng paò čī mîn, “the people protected
by the king”; paò mîn čī wâng, “a king protecting
the people”. Observe here the ingenious method
of distinguishing the active and passive voices by
81strictly adhering to the natural order and placing the
subject before and the object after the verb. If we
put ì before, and kú after, a single word, it means “on
account of, because of” (cf. English for…'s sake); if
we place a whole sentence between these “brackets,”
as we might term them, they are a sort of conjunction,
and must be translated “because”.

67. (35) These few examples will give the reader
some faint idea of the language of the Celestial
Empire; and, if the older generations of scholars are
to be trusted, we have to picture to ourselves the
primeval structure of our own language (in the root-period)
as something analogous. Thus Schleicher 157
says: “The structure of all tongues indicates that
their oldest form was essentially identical, with that
which has been kept unchanged in some languages
of the very simplest structure (e.g., Chinese)”. Similar
utterances might be adduced from the writings of
Max Müller, Whitney, etc; and the same view. is
also held by the renowned Chinese scholar J. Edkins,
in his book on The Evolution of the Chinese Language
(1888), from the preface of which I quote the following:
“Chinese remains possessed of a primitive
order of words, and a monosyllabic structure. These
peculiarities give it a claim to be a direct descendant
of the mother-tongue of humanity, but it is not itself
that mother-tongue…there is no other language,
or family of languages, which can be more reasonably
assumed to be the speech first used in the world's
grey morning than can the Chinese.”82

68. (36) However, different considerations have
tended to shake this faith in the primitiveness of the
Chinese language. As early as 1861, R. Lepsius,
from a comparison of Chinese and Tibetan, had
derived the conviction that “the monosyllabic character
of Chinese is not original, but is a lapse from
an earlier polysyllabic structure”. And Mr. Edkins,
whose identification of Chinese with “the speech
first used in the world's grey morning” I just now
quoted, has been among the foremost to examine
the evidence offered by the language itself for the
determination of its earlier pronunciation. This, of
course, is a much more complicated problem in
Chinese than in our alphabetically-written languages;
for a Chinese character, standing for a complete
word, may remain unchanged while the pronunciation
is changed indefinitely. But by means of dialectal
pronunciations in our own day, of remarks on
pronunciation in old Chinese dictionaries, of transcriptions
of Sanskrit words made by Chinese Buddhists,
of rhymes in ancient poetry, of phonetic or partly
phonetic elements in the word-characters, etc., etc., 158
it has been possible to demonstrate — with comparative
certainty on the whole, though undoubtedly
with no small uncertainty in many particulars — that
Chinese pronunciation has changed considerably, and
that the direction of change has been, here as elsewhere,
towards shorter and easier word-forms. Above
all, consonant groups have been simplified.83

69. (36) It is not impossible, as I think, that
certain peculiarities in the living pronunciation of
Chinese might, if correctly interpreted, lead us to
the same conclusion. I refer to the change sometimes
wrought in the meaning of a word by the adoption
of a different musical tone. 159 Thus wang with
the “lower even” tone means “king,” in the “departing”
tone it means “to become king”; lao, according
to the tone in which it is spoken, is either “work”
or “pay the work”; tsung with the “lower even” tone
means “follow,” with the “departing” tone it means
“follower,” and with the “upper even” tone “footsteps”;
haò is “good,” and haó is “love”. Nay,
meanings so different as “acquire” and “give” (sheu)
or “buy” and “sell” (mai) are only distinguished by
the musical accents. 260

Edkins makes an attempt to account for such
changes, which I must confess I am not entirely able
to follow; pointing with the hand seems to play
some part in it, and then, “the wordmaker wanted
the words ‘to love’ and ‘to sell,’ and he formed them
out of ‘good’ and ‘buy’ by adding an intonation
existing in his environment”. Similarly, though
with much greater clearness, V. Henry says: “Is
not the process transparent? In primitive language,
84in order to say ‘I love,’ one would say hao, ‘this is
good (for me),’ accompanying this syllable with a
gesture to take the place of the words that were
understood; this gesture influenced the accent or
tone of the syllable… Further, we have in Chinese
mái (to sell) and mài (to buy)… This double
form seems to relate the history of exchange between
men, as we find it in all works treating of political
economy: unaccented mai probably denoted the rudimentary
bargain or truck; but as this term had in
each particular case to be exactly defined, the party
saying mai (I acquire) would accompany the syllable
with a centripetal gesture to indicate that the object
came to him, and the other party who said mai (I
cede) would naturally make the opposite gesture.
The effect of this mimicry has been a divergent modification
of the sound of the root”. 161 Even granting
the possibility of the gesticulation affecting the tone of
the voice in the two cases, — which does not, however,
seem quite beyond question, — this explanation presents
some serious difficulties; for what if the speaker
wanted to say “you buy”? Then the theory would
make us expect the same gesture (and therefore the
same tone) as in “I sell,” contrary to the actual fact.
And one does not see which gesture and which tone
would have to be chosen to express the notion, “he
sells to her,” if both the persons spoken of were
absent at the time. Or are we to suppose that men
85at some remote period spoke only in the first person
singular? Besides, the theory only assists us in the
case of a very few pairs of words, and leaves us
entirely in the dark about some of the above examples
and numerous others. We must, therefore,
be excused for looking about for another explanation;
and I think I am able to suggest one. In the
Danish dialect spoken in Sundeved (in Prussian
South Jutland) two purely musical tones are distinguished,
one high and the other low. Now these
tones often serve to keep words or forms of words
apart that would be perfect homonyms but for the
accent, exactly as in Chinese. Thus na with the low
tone is “fool,” but with the high tone it is either the
plural “fools” or else a verb “to cheat, hoax”; ri,
“ride,” is imperative or infinitive according to the
tone in which it is spoken; jem in the low tone is
“home,” and in the high “at home”; and so on in a
great many words. 162 Now, in this language we need
not go to gestures to explain the origin of these
tonic differences, for the explanation is obvious to
anybody familiar with the history of Scandinavian
languages. The low tone is found in words originally
86monosyllabic (compare standard Danish nar,
rid, hjem), and the high one in words originally dissyllabic
(compare Danish narre, ride, hjemme). The
tones belonging formerly to two syllables are now
condensed on one syllable (originally, I suppose, in
the form of a circumflex or compound tone). 163

Although, of course, the Chinese tonic differences
cannot in every respect be paralleled with those
found in the Scandinavian languages, I see no reason
why we should not set forth the provisional hypothesis
that the above-mentioned pairs of Chinese words
were formerly distinguished by derivative syllables or
flexional endings and the like, which have now disappeared,
without leaving any traces behind them
except in the tones. This hypothesis is perhaps
rendered more probable by what seems to be an
established fact — that one of the five tones, at least in
87the Nan-king pronunciation, has arisen through the
dropping of final consonants (p, t, k).

70. (37) However this may be, the death-blow
was given to the dogma of the primitiveness of
Chinese speech by Ernst Kuhn's lecture Ueber
Herkunft und Sprache der Transgangetischen Völker
(Munich, 1883). He compares Chinese with the
surrounding languages of Tibet, Burmah, and Siam,
which are certainly related to Chinese, and have
essentially the same structure; they are isolating,
have no flexion, and word-order is their chief grammatical
instrument. But the laws of word-position
prove to be different in these several languages, and
Kuhn draws the incontrovertible conclusion that it is
impossible that any one of these laws of word-position
should have been the original one; for that would
imply that the other races have changed it without
the least reason and at a risk of terrible confusion.
The only likely explanation is that these differences are
the outcome of a former state of greater freedom.
But if the ancestral speech had a free word-order, to
be at all intelligible it must have been possessed of
other grammatical appliances than are now found in
the derived tongues; in other words, it must have
indicated the relations of words to each other by
something like or corresponding to our derivatives
or inflexions.

71. (38) To the result thus established by Kuhn,
that Chinese cannot have had a fixed word-order
from the beginning, we seem also to be led if we
88fully and thoroughly consider the question, what is a
fixed word-order? And is primitive man likely to
have arranged his words according to such fixed
order? A Chinese sentence is arranged with the
same logical precision as the direction of an English
letter, where the most specific word is placed first,
and each subsequent word is like a box comprising
all that precede: Miss-Emily-Brown-23-High-Street-Brighton-Sussex-England.
The only difference is
that a Chinaman would reverse the order, beginning
with the most general word and then in due order
specialising. 164 The logical consistency in both cases
is the same.

Now, is it probable that primitive man, that unkempt,
savage being, still half brute, who did not
yet deserve the proud generic name of homo sapiens,
but would be better termed, if not homo insipiens, at
best homo incipiens, is it probable that this urmensch,
who was little better than an unmensch, should have
been able at once to arrange his words, or, what
amounts here to the same thing, his thoughts, in such
a perfect order? I should prefer to suppose that
logical, methodical, orderly thinking and speaking
have only been attained by mankind after a long and
troublesome struggle. And above all an exact order
of words as a grammatically significant element of
speech is what we should, least of all, look for in the
case of primitive man, whose thoughts and words
are most likely to have come to him rushing helter-skelter
89in wild confusion. Nay, “a fixed word-order”
is without doubt to be considered the highest, finest,
and accordingly latest developed expedient of speech
to which man has attained. The rules of word-position
have too long been the Cinderella of linguistic
science — how many even of the best grammars are
wholly or almost wholly silent about them! Thus,
with regard to the Bantu languages it was only from
a short remark made incidentally by Bleek (ii., 108)
that I got a little bit of information, which was, however,
of the greatest importance to me, namely, that
these languages do not make use of a fixed word-order
to indicate changes of meaning. Not a word was to be
found on this point in the rest of my authorities.
And although in English a change of word-order
will in many cases completely alter the meaning of
the proposition, this subject is, in many grammars,
treated very inadequately, if at all. Yet there is no
denying that the theory of word-order, of its importance,
and of the mutual relation between this and
other grammatical expedients, offers a great many
problems to the thoughtful student of language.

72. (39) To take one of these problems: What is
the reason of the prevalence of the word-order–subject-verb-object — in
English., Danish., French, Chinese,
to mention only a very small number of languages?
The fact of “that heathen Chinee” using the same
order as ourselves precludes the supposition, often
resorted to in such cases, that one of the European
nations has borrowed the usage from one of the others,
90and shows the phenomenon to be founded in the very
nature of human thought, though its non-prevalence
in most of the older Arian languages goes far to
show that this particular order is only natural to
developed human thought.

Again, a question is commonly indicated by an
inversion of the usual order of words, as when
we say: “Has John got his hat?” — did ever any
philological writer examine the rise of this interrogatory
form in those languages where it is found,
and the extension of its use? Originally, in all the
old Germanic tongues as well as in Latin, etc., inversion
was very often used without any interrogatory
sense being denoted by it; and traces of this state
of things are still to be found, especially when the
verb is not the first word of the sentence (“About
this time died the gentle Queen Elizabeth”), and in
parenthetic sentences (“‘Oh, yes,’ said he”). In
German, especially in the ballad style, it is still possible
to begin a sentence with the verb: “Kam ein
schlanker bursch gegangen”. But it is well worth
noticing that in German, as well as in the other
modern languages of Western Europe, such an
arrangement is generally avoided, so that in those
cases where the speaker wants to give a prominent
place to the verbal idea by putting it before the
subject proper, he will, so to speak, satisfy his
grammatical instincts by putting a kind of sham
subject before the verb; English speakers will use
there; Danes, der; in German es is used; in French,
91il; in SwediSh., det: “there comes a time when…,”
“der kommer en tid, da…,” “es kommt eine zeit
wo…,” “il arrive un temps où…,” “det kommer
en tid daa”. 165

The inverted word-order, then, was not originally peculiar
to interrogatory sentences; a question was
expressed, no matter how the words were arranged,
by pronouncing the whole sentence, or the most
important part of it, in the peculiar interrogatory
rising tone. This manner of indicating questions is,
of course, still kept up in our modern speech, and
it is often the only thing to show that the words
are to be taken as a question (“John?” “John is
here?”). But although there was thus a natural
manner of expressing questions, and although the
inverted word-order was used in other sorts of
sentences as well, yet in course of time there came
to be a connexion between these two things, so that
putting the verb before the subject came to mean a
question and to be felt as implying a question. The
proof of this is that that rising of the tone which is
natural to questions is far less marked in an inverted
sentence like “Is John here?” than if you ask the
question by means of the sentence “John is here?”
the word-order leading you to expect a question in
92the former case and a statement in the latter. This
obliteration of the interrogatory tone is especially
easy to see in the pronunciation of indirect questions
in the form of direct ones, such as (Dickens, David
Copperfield, ii., 384) “Dora asked me would I let her
give me all her money to keep” | (George Eliot, Mill
on the Floss, i., 252) “he had meant to imply would
she love as well in spite of his deformity”.

73. (39) Now, after this method of indicating
questions had become comparatively fixed, and after
the habit of thinking first of the subject had become
all but universal, these two principles entered into a
conflict, the result of which, in three of the languages
here specially dealt with, has been in many cases a
compromise, the interrogatory word-order carrying
the day formally, while really the verb, that is to say,
the verb which means something, is placed after its
subject. In English., this is attained by means of the
auxiliary do: instead of Shakespeare's “Came he not
home to-night?” (Romeo, 1045). we now say, “Did
he not, or Didn't he, come home to-night?” and so in
all cases where a similar arrangement is not already
brought about by the presence of some other
auxiliary, “Will he come?” “May he come?”
etc. 166 In Danish., the verb mon, used in the old
93language to indicate a vague futurity, fulfils to a
certain extent the office of the English do; up to
some two hundred years ago mon was really an
auxiliary verb, followed by the infinitive: “Mon nogle
miles færd Vel være saadan larm og saadan fare værd?”
(Holberg); but now the construction has changed, the
indicative is used with mon, as in “Mon den gudinde er,
der plages saa af galde?” (ibid.), and mon must be considered
no longer a verb but an interrogatory adverb;
“Mon han kommer?” differs from “Kommer han?”
in being more indefinite and vague: “Will he come,
do you think?” 167

These remarks will, I hope, show the interest of
many problems connected with the history of word-position,
and also throw a little light on some of
those strange ways by which languages must often
travel to arrive at new grammatical categories and
new forms of expression.

75. (40) I now pass to two questions of the
greatest importance to the main subject of this book.
First, What is the relation between freedom in word-position
and a complicated system of inflexions?
How is it that in historical times simplification of
grammar always goes hand in hand with the development
of a fixed word-order? Is this accidental,
and is there no connexion between the two phenomena?
Or, is there a relation of cause and effect
between them?

I dare say most readers, after bestowing some
little thought on the question, will agree in answering
the last question in the affirmative, and in seeing
that a fixed word-order is the prius, or cause, and
96grammatical simplification, the posterius, or effect.
It is. however, by no means rare to find underlying,
in a more or less latent way, people's notions of these
things, the theory that the inflexional endings were
first lost “by phonetic decay,” or “by the blind operation
of sound laws,” and then a fixed word-order
had to step in to make up for the loss of the previous
forms of expression. But if this were true, we should
have to imagine an intervening period in which the
mutual relations of words were indicated in neither
way; a period, in fact, in which speech would be
unintelligible and consequently practically useless.
The theory in question is therefore untenable. It
follows that the fixed word-order must have come in
first: it would come quite gradually as a natural
consequence of greater mental development, and
general maturity: when the speaker's ideas no longer
came into his mind helter-skelter but in orderly
sequence. If before the establishment of some sort
of fixed word-order any tendency to slur certain
final consonants or vowels had manifested itself, it
could not then have become universal, as it would
have been constantly checked by the necessity that
speech should be intelligible, and therefore those
marks which showed where the several words belonged
to, should not be obliterated. But when
once each word was placed at the exact spot where
it properly belonged, then there was no longer anything
to forbid the endings being weakened by assimilation,
etc., or their being finally dropped altogether.97

76. (40) To bring out my view, I have been obliged
in the preceding paragraph to use expressions that
must not be taken too literally; I have spoken as
if the changes referred to were made “in the lump,”
that is, as if the word-order was first settled in every
respect, and after that the endings began to be
dropped. The real facts are, of course, much more
complicated, changes of the one kind being interwoven
with changes of the other in such a way as to
render it difficult, if not impossible, in any particular
case to discover which was prius and which posterius.
We are not able to lay a finger on one spot and say:
Here final m or n was dropped, because it was now
rendered superfluous as a case-sign on account of the
accusative being invariably placed after the verb, or
for some other such reason. But, nevertheless, the
essential truth of my hypothesis seems to me unimpeachable.
Look at Latin final s. Cicero (Orat.
48, 161) expressly tells us — and a good many inscriptions
corroborate his words 169 — that there existed
a strong tendency to drop final s; but the tendency
did not prevail. The reason seems obvious; try the
effect in a page of Latin prose of striking out all
final s'es, and you will find that it will be extremely
difficult to determine the meaning of many passages;
a consonant playing so important a part in the endings
of nouns and verbs could not be left out without
loss in a language possessing so much freedom in
98regard to word-position as Latin. Consequently, it
was kept; but in course of time word-position became
more and more subject to laws; and when,
centuries later, after the splitting up of Latin into the
Romance languages, the tendency to slur over final
s knocked once more at the door, it met no longer
with the same resistance as before; final s disappeared,
first in Italian and Roumanian, then in
French. In French the disappearance took place
towards the end of the Middle Ages, and some cases
of survival are still found in actual pronunciation;
in SpaniSh., final s is just now, at the end of the
nineteenth century, beginning to sound a retreat.

77. (42) The answer to the second question hinted
at in § 75 cannot now be doubtful. The question is
this: Is it beneficial to a language to have a free word-position?
Or, on the other hand, is the transition
from freedom to greater strictness in this respect to
be termed not loss but progress?

78. (43) It will be objected that freedom to arrange
your words as you please is a great advantage. To
this I answer: We must beware of letting our judgment
be run away with by a word. Because freedom
is desirable elsewhere it does not follow that it should
be the best thing in this domain; just as above we
did not allow the phrase “wealth of forms” to impose
upon us, we must here be on our guard against
the word “free”. It will be an easy matter to turn
the tables, if instead of inquiring into the advantages
of freedom we put the question in this way: Which
is preferable, order or disorder? It is true that
viewed exclusively from the standpoint of the speaker,
100freedom would seem to be a great advantage, as it is
a restraint to him to be obliged to follow strict rules;
but an orderly arrangement is decidedly in the interest
of the hearer, as it facilitates very considerably
his understanding of what is said to him; and therefore,
though indirectly, it is in the interest of the
speaker also, because he speaks for the purpose of
being understood, for we may leave out of account
those persons who speak solely for their own pleasure.
Add to this that the want of a fixed order of words
necessitates for the speaker the use of a more circumstantial
and clumsy wording, including a great many
reminders and so on, and you will see that even from
the speaker's point of view a fixed word-order has
not a few advantages.

79. (43) If it be urged in favour of a free word-order
that we owe a certain regard to the interests of
poets, it must be taken into consideration, first, that
we cannot all of us be poets, and that a regard to all
those of us who resemble Moliere's M. Jourdain in
speaking prose without being aware of it, is perhaps
after all more important than a regard to those very
few who are in the enviable position of writing readable
verse; secondly, that a statistical investigation
would, no doubt, give as its result that those poets
who make the most extensive use of inversions and
other antiquated word-positions are not among the
greatest of their craft; and, finally, that in those
languages which have turned word-order to profit as
a grammatical expedient, — at least, in those that I
101am acquainted with — so many methods are found of
neutralising this restraint, in the shape of particles,
passive voice, constructions of sentences, etc., that no artist
in language (and that is what every poet should
be) need despair.

80. Observe, however, the nature of my arguments,
in favour of a strict word-order, and you will notice
that they imply a reservation of no small significance.
Most languages have some rules of word-position
which are like certain rules of etiquette, in so far that
you can see no reason for their existence, and yet
you are obliged to bow to them. Historians may, in
some cases, be able to account for their origin and
show that they had a raison d'être at some remote
period; but the circumstances that called them into
existence then are now no more, and now the rules
are felt to be restraints with no concurrent advantage
to reconcile us to their observance. No praise is due
to rules of position of this sort, and in estimating
languages we should, as far as possible,
take this point too into consideration: What is the
proportion between useful and useless rules of word-position?

According to Mr. Johannson, these passages show the
disadvantages of doing away with formal distinctions,
for the sentences would have been clear if each
separate case had had its distinctive sign; “the
greater the wealth of forms, the more intelligible the
speech”. And they show, moreover, that such
ambiguities will occur, even where the strictest rules
of word-order are observed (bei der festgeregelsten
103stellung…[beispiele] die eine ganz regelmässige
wortfolge aufweisen). I shall not urge that this is
not exactly the case in the last sentence, if die
schwester and dein weib are to be taken as accusatives,
for then an should have been placed at the very end
of the sentence; nor that, in the last sentence but one,
the mention of George Eliot as the “konkurrent” of
Fernan Caballero seems to show a partiality to the
Spanish authoress on the part of the writer of the
sentence, so that the reader is prepared to take
welche as the nominative case; freilich would seem to
point in the same direction. But these, of course,
are only trifling objections; the essential point is that
we must grant the truth of Mr. Johannson's contention
that we have here a flaw in the German language;
the defects of its grammatical system may and do
cause a certain number of ambiguities. Neither is it
difficult to find out the reasons of these defects, by
considering the structure of the language in its entirety,
and by translating the sentences in question
into a few other languages and comparing the results.

It is the fact that countless sentences of this sort
are perfectly clear, which leads to the employment
of similar constructions even where the resulting
sentence is by no means clear; but if all, or most,
words were identical in the nominative and the dative,
like gott, or in the dative and genitive, like der indierin,
constructions like those used would be impossible
to imagine in a language meant to be an
intelligible vehicle of thought. And the ultimate
reason of the ambiguities will thus be seen to be the
105inconsistency in the formation of the several cases.
But this inconsistency is found in all the old languages,
of the Arian family: cases which in one gender or in
one declension are kept perfectly distinct, are in
others identical. 172 While in Latin patres filios amant
or patres filii amant are perfectly clear, patres consules amant
allows of two interpretations; and in how
many ways cannot such a proposition as Horatius et
Virgilius poetæ Varii amici erant be construed?
Such drawbacks seem to be inseparable from the
structure of the highly flexional Arian languages;
although they are not logical consequences of a
wealth of forms, yet historically they cling to those
languages which have the greatest number of grammatical
endings. And as we are not here concerned
with artificial Volapüks, but with natural languages,
we cannot accept the above quoted verdict of Johannson's:
“The greater the wealth of forms, the more
intelligible the speech”. In fact, the author himself
seems to have a scruple about it, for he adds in a
footnote: “I do not, of course, mean to lend my
sanction to a luxuriant and clumsy wealth of forms
106such as that found for instance in the Bantu languages,
but I always have in my mind the wealth of forms
(formenschatz) found in Arian languages;” unfortunately,
he does not tell us which of the several Arian
languages he will regard as the beau ideal in which
he finds the golden mean; are eight, or seven, or six,
or perhaps five, distinct cases the ne plus ultra?

82. Secondly, we consider the position of words
in Mr. Johannson's sentences, and we discover that
Modern High German still enforces some old rules of
word-order.which have been given up in the other
cognate languages, where they were formerly in
common use. The most important of these is that of
placing the verb last in subordinate sentences; in
two of the examples it is this rule which causes
the ambiguity, which would accordingly have been
avoided in a principal sentence: Die deutsche zunge
klingt und singt gott ivi himmel lieder; or, die d. z.
klingt und gott im himmel singt lieder | sie erreicht
freilich nicht die spanische dichterin; or, die sp. d.
erreicht sie freilich nicht. In one of the remaining
sentences the ambiguity is caused by the rule that the
verb must be placed immediately after an introductory
adverb: if we omit the doch the sentence becomes
clear: Die gesellschaft der indierin würde; or, die
gesellschaft würde der indierin lästig gewesen sein.
All of which exemplifies the distinction between useless
and useful rules of word-position. Word-position in
German is comparatively strictly regulated, but generally
by arbitrary rules; if, therefore, you change the
107order of words in a German sentence, you will often
find that the meaning is not changed in the least, but
the result will be an unidiomatic construction (bad
grammar); while in English a transposition will often
result in perfectly good grammar, only the meaning
will be an entirely different one from that of the
original sentence. I do not mean to say that the
German rules of position are all useless, and the English
all useful; but only that in English word-order is
utilised to express difference of meaning to a far
greater extent than in German, which stands in this,
as in many other respects, on a lower plane of
development than English.

83. Before leaving Mr. Johannson, I must remark
that as word-order in those languages which make the
proper use of it is used much more consistently than
any endings ever are in actually existing languages,
it is not only more convenient, but also clearer than
flexions. The alternatives, accordingly, are not, as
he puts it, the avoidance of misunderstandings on the
one hand, and the sparing of flexional endings on
the other; for in the evolution of languages the
discarding of old flexions is perfectly consistent with
the development of simpler and more regular
expedients that are rather less liable to produce
misunderstandings than were the old endings. When
Mr. Johannson writes, “In contrast to Jespersen I
do not consider that the masterly expression is the
one which ‘is wisely silent,’ and consequently leaves
the meaning to be partly guessed at, but the one
108which is able to impart the meaning of the speaker or
writer clearly and perfectly” — he seems to me rather
wide of the mark. For, just as in reading the
arithmetical symbol 234 we are perfectly sure that
two hundred and thirty-four is meant, and not three
hundred and forty-two, so in reading or hearing
“The boy hates the girl” we cannot have the least
doubt who hates whom. If in any way the understanding
of English (or Chinese, etc.) sentences
depended on guesswork like a missing word competition
(or a missing flexion competition), well, then the
language could not be said to be “wisely silent”.

84. I must here turn to another critic, Prof. D. K.
Dodge, who, in reviewing my Fremskridt i Sproget, 173
says: “To cite one example, which figures in almost
every English Rhetoric as a violation of clearness:
‘And thus the son the fervid sire address'd’. The
use of a separate form for nominative and accusative
would clear up the ambiguity immediately.” No
doubt it would; but so would the use of a natural
word-order. If the example is found in almost
every English Rhetoric, I am happy to say that such
ambiguous sentences are scarcely, if ever, found in
other English books. No person in his sober senses
would ever speak so; no prose writer would ever
indulge in such a style; and in the whole range of
my reading in English poetic literature I do not
remember a single instance of so bold an inversion,
except where the context would unmistakably show
109which word was to be taken as the nominative. 174
And even if such examples are here and there to be
found, the only thing they can prove is this, that
a violation of the rules of grammar entails want of
clearness, and in present-day English such an arrangement
of words is to be considered as a fault to be
classed almost with the use of dominum as a nominative
in Latin.

Those who regret the want of separate forms for
nominative and accusative, etc., seem generally to be
considering how a language might be constructed
which would combine the greatest clearness with
simplicity and freedom; they see some drawbacks in
the language that is most familiar to them, and they
cannot help exclaiming: Oh, how easy it would be
to remedy the defects, if only we had separate forms,
etc. This manner of regarding linguistic problems
presents no very great difficulties, especially as
nobody will take you to task and call upon you
actually to construct a language such as you dream
of, one that would be perfect in every detail. People
are apt to forget that these are really nothing but
barren speculations with not the slightest scientific
significance, and that the really important questions
are, firstly, What is the direction of change in languages,
as they actually exist? And secondly, Is this
or is it not a direction towards progress?

85. My answer is: Languages tend on the whole
110more and more to utilise word-position for grammatical
purposes; and this is really a progressive
tendency, directly progressive, because it is in itself
the easiest and nicest linguistic expedient, rendering
the task of speaking easier, and involving less effort
on the part of the listener; indirectly, because it facilitates
the great work of simplification in language by
making the unwieldy forms used of old to indicate
concord, etc., more and more superfluous. The substitution
of word-order for flexions means a victory of
spiritual over material agencies.

Word-position has acquired grammatical significance;
and if we ask how this has come to pass, we
get the same answer as before, when we were considering
other grammatical instruments: it has come
by a slow growth, without any intention on the part
of the speakers. By little and little, people accustomed
themselves to arrange their words after the
same pattern, until those case-endings which had
hitherto been the primary grammatical sign to indicate
subject and object, or to show what noun an
adjective belonged to, and those tones which had
been the chief means of indicating a question, became
gradually more subordinate and were finally
made wholly or nearly superfluous. Grammatical
meaning was first expressed by certain more material
instruments, independent of word-position, then by
the same instruments with the words arranged in a
fixed order, and finally by order, independent of
those original instruments.111

Chapter V.
The development of language.

86. (44) We have seen in the preceding investigations
that the downhill theory does not hold good
for languages in historic times; on the contrary,
languages seem to be on the whole constantly
progressive, not only with regard to the development
of their vocabulary, where nobody ever denied it, but
also in grammar, where philologists of the old school
were able to see only decay and retrogression. And
besides establishing this progressive tendency, we have
also incidentally seen some at least of the often unexpected
ways which lead languages to develop new
grammatical forms and expressions. We are thus
prepared to enter into a criticism of that theory concerning
the prehistoric development of Arian speech
which has met with greatest favour among philologists,
and which has been expounded with greatest precision
and consistency by Schleicher. The theory, as will
be remembered, was this: an originally isolating
language, consisting of formless roots, passed through
an agglutinating stage, in which formal elements had
been developed, although these and the roots were
112mutually independent, to the third and highest stage
found in flexional languages, in which formal elements
penetrated the roots and made inseparable unities
with them.

87. (45) First, as regards the postulated root
stage, we have seen how the support which Chinese
was supposed to lend to the theory has broken down.
But also from other quarters the belief in such a
starting-point has been shaken. An investigation of
Old Arian phonetic laws has led some philologists to
doubt the supposition, which is essentially due to
the old Indian grammarians, that roots were always
monosyllabic; and now many prefer fancying the
roots as dissyllabics. A more important reason for
objecting to the theory seems to be this, that we
cannot imagine people expressing themselves by
means of a language consisting exclusively of roots
such as those given by Sanskrit scholars; the highly
abstract significations assigned to them (“breathe,
move, be sharp or quick, blow, go,” etc.) would in
themselves be sufficient to preclude the idea of such
a language existing as a practically useful means of
communication — especially between savage or worse
than savage beings. No; of a certainty, roots never
were spoken words; and there is no doubt a great
deal of truth in such expressions as these: “a root is
only something imaginary, an abstraction” (Pott);
“the root is an ideal object” (Lyngby 175); “roots are
not natural entities, but investigators' hypotheses.
113Speakers seem to me to have spoken, from the first,
and to speak now, without any general consciousness
of their existence” (Ellis 176). It seems, then, that the
correct view of the nature of roots is that they are
abstractions of that which is common to a group of
words which are felt as etymologically related. But,
according to this, the root is not older than the
words that are “derived” from it; and consequently,
in spite of the opposition of most living comparative
philologists, we can speak with perfect justice of
Greek, French, or English roots. Why not speak of
a French root roul, found in rouler, roulement, roulage,
roulier, rouleau, roulette, roulis? This only becomes
unjustifiable if, in putting down roul as the root, we
fancy we have historically explained the origin of the
words in question, or if we suppose that roul is a root
which at some time existed independently of the
derived words; for then the linguistic historian steps
in and objects that the words have been formed, not
from a root, but from a real word, and one which is
not even itself a primary word, but a derivative, Latin,
rotula, a diminutive of rota, “wheel”. 277 To the
popular instinct sorrow and sorry are undoubtedly
related to one another; and a student of Living
English should respect this feeling, and say that the
words now belong to the same root; but a thousand
114years ago they had nothing to do with each other,
and belonged to different roots (Old English., sorg,
“care,” and sarig, “wounded,” “afflicted”). If all
traces of Greek and Latin, etc., were lost, a linguist
would have no more scruples about connecting scene
with see than most illiterate Englishmen have now.
But who will vouch that the Arian roots found in our
dictionaries have not originated in similar ways to
the roots roul-, sorr-, and see-?

88. (46) According to Schleicher and his disciples,
the root stage was succeeded by the agglutinating
stage, in which the main part of the word was unchanged,
while formal elements might be added as
prefixes or suffixes. Now, as only very few languages
present the same kind of structure as Chinese (which
represented the first class), and as, on the other hand,
only two families of languages (the Arian and
Semitic) are allowed a place in the third or flexional
class, this intermediate class is made to include the
great majority of human tongues. Consequently,
languages of the most widely different types are
brought together under the heading of “agglutination,”
and it becomes next to impossible to form any
idea of what is properly the connecting link between
these languages. The definitions generally given
seem to have been taken from Ural-Altaic languages,
and to have been thence transferred with more or
less of constraint to all the rest. The consequence
is that in reading, for instance, in Schleicher or Fr.
Müller, their descriptions of languages which are
115termed agglutinating, one is perpetually startled and
driven inwardly to confess that one is unable to see
any difference between the grammatical forms of
these languages and those which in Latin and Greek
we call flexions. It is especially so in dealing with
so complicated a language as Basque; here the verbal
forms indicate not only the person of the subject, but
also that of the object proper and the object of reference
(dative); and, further, “the Basque language distinguishes
in the verbal flexion when a man, a
woman, or a person who commands respect is spoken
to: the two first forms are familiar; the third is
generally used. Thus, dut, ‘I have’ (generally speaking);
diat,‘I have’ (to a man); dinat, ‘I have’ (to a
woman).” 178 On the whole, the forms are so manifold
that we understand how Larramendi, in his legitimate
pride at having been the first to reduce them to a
system, called his grammar El Imposible Vencido,
“The Impossible Overcome”. To give some notion
of this jumble of forms I copy a few from Prof.
Sayce's Introduction to the Science of Language (ii.,
212): det, I have it; aut, I have thee; ditut, I have
them; dizut, I have it for thee; dizutet, I have it for
you; dizkizut, I have them for you; diot, I have it
for him. Can this be called an invariable root with
endings added loosely, and easily separated?

89. Some philologists have maintained that in
French in the coalescence of the pronoun with the verb
we have really something corresponding to the Basque
116verbal forms (or to the American incorporations).
They will say that the spelling of je and the other
pronouns as separate words inj'aime, il te le disait
goes for nothing, and that the pronouns are really
part and parcel of the verb as much as the corresponding
elements in the exotic languages mentioned;
if French had had no literary tradition, jaime, jelaime,
etc., would probably have been written in one word.
There seems, however, to be a difference; the French
elements are much more felt as independent of one
another than can be the case in Basque, etc. This
is shown first by the possibility of varying the pronunciation:
il te le disait may be pronounced either
[itləizɛ] or [itəldizɛ] (or, even in more elevated style,
iltələdizɛ); secondly, by the regularity of these joined
pronominal forms, for they are always the same,
whatever the verb may be; and, lastly, by their
changing places in certain cases: te le disait-il?
dis-le-lui, etc. And, at any rate, the verbal form is
totally independent of the pronoun, as seen in “Jean
disait a sa mere; disons ça à sa mère,” etc. All of
which is impossible in the Basque forms, in which
you can no more separate the pronominal and verbal
elements or make them change places than you could
the am and o in the Latin amo.

90. (44) The term agglutinative is still less applicable
to such languages as some of those spoken
by North American Indians and Eskimos, where the
incorporation of expressions for various subordinate
ideas into the verb is carried to such an extent that
117the whole utterance forms one inextricable web,
which can hardly be termed either a word or a sentence,
and into which the several elements enter, often in
hardly recognisable shape. We are here nearly as far
removed as possible from the simplicity and
lucidity which distinguish those languages to which
the term agglutination seems first and with greatest
propriety to have been applied, namely, Finnic,
and Magyar and their cognate tongues. And yet,
even with regard to these latter, an eminent authority
on all of these languages writes: “The difference
between these and the so-called flexional languages,.
to which our tongue belongs, is in many points comparatively
vague, and there are here found not a
few formations which can with perfect justice be said
to rest on flexion”. 179

91. (47) The third stage, according to Schleicher,
was flexion, characterised by the highest union of
content and form, the root itself being subject to
change to express modifications of the meaning,,
especially for grammatical purposes. Here we must
first notice what Schleicher himself admits — that in
flexional languages we find a great many things,
which cannot be called flexion as he defined it. In
his view they are survivals of the previous stages of
isolation and agglutination through which these
languages passed in prehistoric times. And next we
must remember that originally no modification of
118meaning was associated with those inner changes in
the root. If in Greek we have the three forms of
gradation (ablaut) lip, leip, and loip, they owe their
origin to differences of accent; and if they are used
in three different tenses of the verb (e-lipon I left,
leipō I leave, leloipa I have left) the tense relation
itself is expressed by means of endings (and beginnings),
but not at all, or at least neither originally nor
exclusively, by vowel gradation. So too in the more
recent phenomenon, mutation (umlaut): where it is
used as a means of indicating a plural, as in goose –
geese, Danish gaas — gæs, this is a comparatively
modern development: originally the plural was
expressed by an ending (-es, -iz), which in course of
time modified the root vowel, and then, some time
afterwards, was itself dropped. Here, then, we have
again an originally accidental change of the word,
which has eventually been made to do duty as an
expedient for signifying a change of sense. And,
curiously enough, in no other language has this been
done in a greater degree than in that language which,
according to Schleicher, shows the deepest decline
from the flexional golden age, viz., English. In
English more than anywhere else change of vowel
alone, without any concurrent change of endings or
the like, is used to distinguish different shades of
meaning: as, for instance, sing, sang, sung, song. But
if we should ask whether Schleicher is right in looking
upon this as the highest and most perfect of formal
means in language, we must, I am afraid, express
119ourselves a little more reservedly, as this inner change
cannot be used with complete equality and regularity
in all roots.

92. (48) What is, after all, the essential characteristic
of flexion? I can find no better answer than
this: Flexion means inseparableness of the word
itself (the “full word” of Chinese grammar) and the
formal elements (the “empty words” of the Chinese).
The Latin amo shows flexion, the English I love does
not, because the idea is here dissolved into two independent
parts. 180 But if flexion is thus interpreted
120as inseparableness, as opposed to analysis, it will be
easily seen from all the above investigations that
there exist many different gradations of both; in no
single language do we find either synthesis or analysis
carried out with absolute purity and consistency.
Everywhere we find a more or less. Latin is
synthetic in comparison with French, French analytic
in comparison with Latin; but if we were able to see
the direct ancestor of Latin, say two thousand years
before the earliest inscriptions, we should no doubt
find a language so synthetic that in comparison with
it Cicero's would have to be termed highly analytic.

93. (48) Our principal conclusion, then, is this:
the old theory which imagined the prehistoric development
of Arian speech from roots through agglutination
to flexion is untenable. The only way
of arriving at sound hypotheses with regard to prehistoric
times is by examining the development which
takes place in epochs historically accessible to us.
If, in historic times, we find definite and comprehensive
laws of evolution, we cannot help assuming the
same laws as valid for prehistoric times as well; if
history shows us certain lines of direction, followed
by all languages which are in process of change, we
cannot avoid the conclusion that languages have
changed along the same lines as long as human
beings have spoken; so that to imagine the state of
121primeval language we have only to follow these lines,
backwards beyond the earliest period of which we
have any tradition.

Now, Modern English as compared with Old English;
Modern Danish., Swedish and Norwegian as compared
with Old Norse; Modern Low German as
compared with Old Saxon; Modern High German
as compared with Old High German (all modern
Germanic tongues as compared with the Gothic of
Wulfila); Modern French, Italian, etc., as compared
with Latin; Modern Greek as compared with Old
Greek; Modern Persian as compared with the language
of the Avesta (“Zend”) and the cuneiform
inscriptions; Modern Indian dialects as compared
with Prakrit and Sanskrit — all of these show, though
in different degrees, the same direction of change;
the grammatical forms of the modern languages are
all shorter, fewer, simpler, more abstract and more
regular; those of the older languages in general
longer, more complicated, more concrete and more
irregular. Semitic languages present, as I understand,
similar phenomena. And we find traces of an
evolution in the same direction in those languages where
the want of early documents, or the peculiar
character of the early documents, hinders us from
following the historic development with the same
exactitude as in the languages just mentioned (see
the sections above on Bantu and on Chinese). We
seem therefore justified in believing that the pre-Arian
languages spoken in a remote past by our ancestors
122were still more complicated than the oldest languages
we are now acquainted with; they must certainly in
many points have presented similar features to those
found in Basque or in those entangled, polysynthetic
Indian languages, where the sentences consist in
intricate words or word-conglomerations, embodying
in one inseparable whole such distinctions as subject,
verb, direct and indirect objects, number, tense, mood,
etc., and being therefore very clumsy and imperfect
instruments for the expression of thought.

94. But here it will be — as in fact it has been —
objected that this polysynthesis and incorporation
cannot be primitive, as we see similar phenomena
which have developed in quite recent times. The
French incorporation of pronominal forms has been
mentioned above; it cannot be called a case in point.
Prof. Möller says: “In English ‘entangling’ (or
amalgamation, sammenfiltringen) is growing luxuriantly:
's [-z] = is, has; 'd = had, would; ‘ll = will;
don't, won't, can't, etc.” But these developments
cannot be paralleled with flexion or polysynthesis;
for, however closely together he's or John's ( = John is)
is generally pronounced, it is, and is felt to be, two
words, as is shown by the possibility of transposition
(Is he ill?) and of intercalation of other words (John
never is ill). As for don't, won't, shan't, and can't,
they are more like amalgamations of the verbal with
the negative idea. Still, it is important to notice that
the amalgamation only takes place with a few verbs
all of them belonging to the auxiliary or “empty-word”
123class. Therefore, in saying “I don't write,”
etc., the full word is not touched by the fusion, and is
even allowed to be unchanged in cases where it would
have been inflected had the auxiliary not been used;
compare Iwrite, he writes, I wrote, with the negative
expressions I don't write, he doesn't write, I didn't
write. It will be seen, especially if we take into
account the colloquial or vulgar form for the third
person he don't write, that the general movement here
as elsewhere is really rather in the direction of
“isolation” than fusion; for the verbal form write is
cleared of all signs of person and tense, the person
being indicated separately, and the tense sign being
joined to the negation. So also in interrogative
sentences; and if that tendency which can be observed
in Elizabethan English had prevailed — as
some day it will perhaps — of using the “emphatic”
form, I do write, in positive statements even where no
special emphasis is intended, English verbs (except a
few auxiliaries) would have been entirely stripped of
all those elements which to most grammarians constitute
the very essence of a verb, namely, the marks
of person, number, tense, and mood, write being the
universal form, beside the quasi-nominal or adjectival
forms writing and written.

95. Prof. Herm. Möller holds, in opposition to my
views, that the history of language does not show a
continual progressive tendency, but rather a sort of
gyration. He admits that many regular forms have
been substituted for irregular ones; but, on the other
124hand, he finds that some formations which were regular
in earlier stages become irregular in modern languages,
and he infers that regularity and irregularity
and regularity once more go on continually alternating.
Similarly he notices that the Latin flexional
future amabo has been succeeded by the analytical
expression amare habeo, which in its turn is fused
into a new flexional form amerò, aimerai; and from
this he evolves a similar law of rotation from flexion
through analysis to flexion once more, or, as he puts
it in another place: first flexionless analysis, then
agglutination, then flexion, and then again absence
of flexion. But these results are only arrived at by
considering a comparatively small number of phenomena,
and not by viewing the successive stages
of the same language as wholes and deriving general
inferences as to their typically distinctive characters. 181
For if we find that two regular forms have become
irregular, but that in the same period ten irregular
forms have been succeeded by regular ones; or if for
every two instances of new flexions springing up we
see ten older ones discarded in favour of analysis or
isolation, are we not entitled to the generalisation
that anomaly and flexion tend to give way to regularity
and analysis? Prof. Möller seems to be under
the same delusion as a man who in walking over a
mountainous country thinks that he goes down just
as many and just as long hills as he goes up, while
125on the contrary each ascent is higher than the preceding
descent, so that finally he finds himself unexpectedly
many thousand feet above the level he
started from.

96. (49) On every point our investigation has led
us to scepticism with regard to the system of the old
school of philology. But while we perceive that their
inferences were drawn too hastily and from insufficient
materials, and while we feel tempted totally
to reverse their system, we must be on our guard and
not establish too rigid and too absolute a system ourselves.
It would not do simply to reverse the order
of the three stages of evolution, and say that flexion
is the oldest stage, from which language tends through
an agglutinative stage towards complete isolation;
for flexion, agglutination, and isolation do not include
all possible structural types of speech, nor do these
words with sufficient definiteness characterise the
successive stages of those languages whose history is
comparatively best known. The possibilities of development
are so manifold, and there are such innumerable
ways of arriving at more or less adequate
expressions for human thought, that it is next to
impossible to compare languages of different families.
Even if it is, therefore, probable that English., FinniSh.,
and Chinese are all simplifications of primitive
flexional or even incorporating languages, we cannot
say that Chinese, for instance, was at one time in
structure like English., and at some other time like
Finnish. English was once a flexional language, and
126is now in some respects agglutinative, in others isolating
or nearly so. With the reservation made in this
paragraph, we may say that on the whole languages
tend always in the exactly opposite direction to that
indicated by Schleicher, namely, from polysynthetic
flexion through agglutination to flexionless isolation.
But it will, perhaps, be preferable to state the same
idea thus: the evolution of language shows
a progressive tendency from inseparable
irregular conglomerations to freely and
regularly combinable short elements.

Schleicher's system is to be likened to an enormous
pyramid; only it is a pity that he should make its
base the small, square, strong Chinese root-word, and
suspend above it the inconvenient flexion-encumbered
Indo-Germanic sentence-word. Structures of this sort
may with some adroitness be made to stand; but
their equilibrium is unstable, and sooner or later they
will inevitably tumble over.

97. (50) Although it will be seen that in a great
many particulars the views advanced in these
chapters have been previously enunciated with more
or less of clearness by other philologists, I do not
think that my theory of the progressive tendency and
direction of language has been expounded before by
any one. It is true that I find the following passage
in Prof. Sayce's Introduction to the Science of Language
127(i., pp. 85-87): “In pursuance of Bopp's method, but
independently of the distinctive theories of his school,
Waitz, the anthropologist, has propounded a new
theory of language…the incorporating languages
of America, in which an individual action is represented
by a single sentence pronounced as one word,
are a survival of the primitive condition of language
everywhere. It is only gradually that the different
parts of speech are distinguished in the sentence, and
words formed by breaking up its co-ordinated elements
into separate and independent wholes…The
agglutinative tongues in which the subordinate parts
of a sentence are brought into duly dependent relation
to the principal concept are more highly advanced
than the inflexional…An isolating language like
the Chinese stands on the highest level of development,
since here the sentence has been thoroughly
analysed and each member of it rendered clear and
distinct, their relations to one another being determined
by position alone. Chinese, therefore, has
given concrete expression in language to the philosophic
analysis of ideas…Waitz's theory of
speech is the theory of an anthropologist who, as the
student of the master-science, is better able to decide
upon the origin of language than the comparative
philologist with whom the existence of language has
to be assumed. No science can of itself discover the
genesis of its subject-matter.”

98. (50) It will be understood that after reading
this exposition of a theory which harmonised so completely
128with the whole tenor of my own thoughts, I
eagerly seized an opportunity of consulting the first
volume of Theodor Waitz's Anthropologie der Naturvölker
(1859). It would have been so very pleasant
to refer to the authority of the eminent anthropologist
and to cry out to comparative philologists: Here you
see, that more than thirty years ago, an outsider propounded
a clear, consistent, and undoubtedly correct
theory, which you have kept disregarding for all these
years! But, oh! how great was my disappointment
when on reading, and reading repeatedly, the section
in question, I was utterly unable to find this fundamental
theory. Waitz as an anthropologist cherishes
a profound respect for philologists, and speaks of the
reliable results of their method in determining the
races of mankind as opposed to those which can be
gained by measurings of skulls and the like; certainly,
it never entered into his mind to overthrow the
edifice of linguistic science, or to start new theories on
the development of the different types of speech.

On the contrary, Waitz makes a cardinal point of
the fixed character of linguistic structure, and consequently
keeps at a respectful distance from Max
Müller's (i.e., Schleicher's) view, according to which
the three types of speech have developed out of one
another with the isolating languages as the starting-point.
The reverse evolution with isolation as the
topmost stage is evidently very far from his thoughts,
for he does not set so very great store by Chinese.
“The wholly asyndetic isolating languages (so we
129read on p. 276) leave our thoughts almost entirely to
themselves; they give hardly any hints as to their
organisation, and leave our single ideas, which correspond
to separate words, to stand by the side of each
other in unrelated independence (in beziehungsloser
selbständigkeit); the speaker is not led to analyse
them, and must rest contented with marking a few
rough distinctions between principal and subordinate
conceptions.” It will be seen that this is not exactly
the same thing as the view attributed to Waitz by
Sayce. “In opposition to this (we further read) the
polysynthetic languages force the speaker as much
as possible to grasp each conceptual whole as a unity,
to join subordinate ideas as closely as possible to the
principal idea, to view as it were at a glance the whole
situation, that has to be rendered in speech, and not
to make the modifications be added piecemeal and
little by little to what is the chief element of thought;
these languages hinder the decomposition of ideas to
a far greater degree than the first (or isolating) class.”
As for flexional languages, Waitz seems to look upon
them as standing higher than the others, but his expressions
are somewhat vague and partly contradictory;
on pp. 275 and 277 he says it is characteristic
of flexion that secondary subordinate elements of
thought are expressed by sounds which have no
meaning of their own, but are inserted as integral
portions of the main word; while on p. 276 we read:
“The fundamental idea of flexion is that the principal
and the subordinate elements of thought remain independent
130and separate, and never coalesce into a
single word”. Sayce quotes this passage, but I fail
to understand how Waitz's expressions can be interpreted
as implying the inferiority of flexional
languages.

99. (51) But, if the theory I looked for was not to
be found in Waitz's work, it is in Sayce's; although
he does not give it as his own, and although
he can hardly be said to accept it. I seize the
opportunity of acknowledging the great influence
Prof. Sayce's works on linguistics have had on me;
his suggestive remarks have often made me take up
lines of thought which perhaps I should not have
been led to, if it were not for him. So much the
more must I from my point of view regret that this
bold opposer of the idola of the ordinary linguistic
school is in some very important points as much
warped by prejudices as most other philologists.
Though he repeatedly hints at the difficulties of
drawing a sharply-defined line of division between
agglutinating and flexional languages, yet he holds
that there is a great gulf fixed between them, and he
says: “The Finnic idioms have become so nearly
inflexional as to have led a recent scholar to suggest
their relationship to our Arian group; nevertheless,
they have never cleared the magical [!] frontier
between flexion and agglutination, hard as it may be
to define, since to pass from agglutination to inflexion
is to revolutionise the whole system of thought and
language and the basis on which it rests, and to
131break with the past psychological history and
tendencies of a speech”. 182 Revolutions do, however,
take place in the world of languages, even if they take
more time than it takes the French to change their
constitutions: if a thousand years suffices to change
a type of speech like that of King Alfred into the
totally different one of Queen Victoria, then the
much longer period which palæontologists and
zoologists accord to mankind on this earth could
work still greater wonders. In spite of such
expressions as this, “Species passes gradually into
species, class into class,” Sayce stands, with regard
to those three or four types of speech which are
distinguished by linguists, in much the same attitude
which naturalists kept with regard to the notion of
“species” before Darwin came; he uses the same sort
of expressions, e.g.:“With all this gradual approximation
the several types of language still remain fixed
and distinct”.

100. (51) Neither is he right in his manner of
viewing the value of phonetic attrition (see above, §
16); he speaks, for instance, of Chinese as a “decrepit”
language (i., 372), that “has been affected by phonetic
decay to an enormous extent” (ibid.), and “the whole
speech has grown old and weather-beaten. It is the
Mandarin dialect which chiefly shows these marks of
ruin” (ii., 221). We are here reminded of Schieicher's
words (above, § 6) that the languages which we speak
now-a-days in North-western Europe are “senile
132specimens of speech,” and we search in vain for any
real thought in connexion with such expressions. A
language which is old, weather-beaten, and decrepit,
which is perhaps — that is only one step further — falling
into dotage and second childhood! what can
this mean? Are the English no longer able to
express their thoughts by means of their language?
Is the speech of the Chinese like that of an old toothless
crone, whose ideas and sounds are equally
incoherent?

Similarly, Sayce does not see the value and
significance of the simplification of the Arian noun-declension
which has taken place in historic times;
he says: “The history of the noun is one of continuous
decay…Long before the age of Arian
separation,… the creative epoch had passed, and
the cases and numbers of the noun had entered on
their period of decay” (ii., 149-150). And although
the pages he devotes to the relative estimation of
languages (i, 374 sqq) contain many excellent and
suggestive remarks, and begin by stating the true
principle, “what we really mean when we say that
one language is more advanced than another, is that
it is better adapted to express thought”; yet the
writer does not go the full length of his own opinion,
for on the very next page he tells us that it is all a
matter of taste: “Preferences of this kind can as
little be referred to an absolute standard as preferences
in the matter of personal beauty. The European, for
instance, has a wholly different ideal of beauty from
133the Negro, and the Negro from the Mongol.” On
some of the most vital points, Sayce has not attained
to a settled and consistent belief.

101. (52) In favour of the theory here expounded
it may be said that it leads on every point to a monistic
view; while Schleicher, though clearly perceiving
that all science and philosophy tends in our days
towards monism, 183 is yet by the very nature of his
standpoint obliged to set up a dualism in some decisive
points. Thus, he establishes an opposition
between phonetic decay and simultaneous development
of richer resources in syntax and style; while
according to our view the evolution in both departments
goes hand in hand, if we consider phonetic
evolution rightly as an evolution towards shorter and
easier forms.

Inseparable from this is another dualism of Schleicher's,
according to which grammar falls into two sharply
divided parts: on the one hand, phonology
and morphology, “the nature side of language,”
which is to be treated as a natural science by the
“glotticist”; and, on the other, syntax and style, “the
more spiritual side of language, which is to a greater extent
134subject to the free determination of the individual,”
and which is therefore to be treated by the literary
student (der philologe) on the historical method. 184
In contrast to this view it must be asserted that there
is only one method for the whole of the science of
language, and that a separation of grammar into two
divisions, treated independently of each other, has
only been, and can only be, injurious to the right
understanding of linguistic phenomena; for form and
meaning always influence each other to a degree
unsuspected by readers of philological periodicals.
Fancy just for one moment a division of a dictionary
into two parts, one of them containing the forms of
words without the least regard to their significations,
and the other marshalling up nothing but the meanings.
But as syntax is nothing but the theory of
the functions, i.e., meanings, of the grammatical
forms — this expression taken in its widest signification,
including also word-position and tones — it will
be seen that many recent “grammars on a comparative
basis” correspond only too closely to the first
part of the supposed dictionary. And this one-sidedness
cannot possibly be conducive to scientific progress.

102. (53) The most important of Schleicher's dualisms,
however, is that of two periods of directly opposite
tendencies, a prehistoric period of progress,
evolution, or construction, and an historic period
of retrogression, decay, or destruction. In opposition
135to this view we must assert that the moment of a
nation's entering into history is of no consequence
at all for the direction of linguistic change, which
goes on in an essentially identical manner now and
in the days of old. If history has any influence at
all on linguistic evolution, it seems to be only that
of accelerating the movement along the same lines
as before; the languages of those nations whose lives
have been most agitated by historical events have
gone farthest in evolution. Besides the more lively
exchange of thoughts, mixture of races may count
here for much (see below, §§ 140, 143); an interesting
contrast is that between the slow development of
Lithuanian, which is rendered so precious to the
antiquarian philologist by the great number of old
forms which it has kept, and the rapid evolution of
English., which on account of its great number of
directly observable changes is an inestimable mine
to the philosopher of linguistic history.

On the other hand, literature, which Schleicher
places side by side with history, certainly, though
perhaps not so powerfully as generally supposed, has
the effect of retarding the tendencies of change in
language by keeping older forms alive for a longer
time than if language was only transmitted orally.
But these accelerating and retarding agencies have
no influence on the direction of change.

If the theory arrived at in the preceding chapters
is really and completely monistic, and requires us at
no point to assume any breach of continuity, it must
136also throw some light on that vexed question, the
origin of speech. I think it does, and it will be my
task in the last section of this book to show that; but
before venturing out into that chaos of grey theories,
it will be well for some time still to continue studying
the “golden tree of life” in the development of some
special points of the English language.137

Chapter VI.
English case-systems, old and modern.

[In Old English]

103. (54) The arrangement of inflexions current
in grammars, according to which all cases of the same
noun, all tenses, persons, etc., of the same verb, are
grouped together as a paradigm, is not a truly
grammatical one: what is common to Old English
dæg — dæge –dæges — dagas — dagumi — daga, — for instance,
is not the flexional element, but the word, or
stem of the word; the tie between all these forms,
accordingly, is not of a grammatical, but of a lexical
character. That such an arrangement may offer some
advantages from a practical point of view cannot,
indeed, be denied; but, on the other hand, it causes
many things to be wrested from one another which
belong together grammatically, e.g., the termination
-um, which is common to the dative plural of all
the flexional classes. Besides, it forces us to separate
from one another the two parts of grammar which
treat respectively of the forms of words and of their
uses. In the latter, we must of needs deal with (say)
all datives under one head, all genitives under
another, and so forth, while in accidence these forms
138are distributed according to declension classes. Such
a disjunction, however, of accidence and syntax,
beyond what is strictly necessary, is doubtless injurious
in every respect (cf. above, § 101). At any
rate, this paradigmatic arrangement of grammatical
phenomena will not answer the purposes of this
chapter, where we seek to get as perspicuous a survey
as possible of the grammatical forms of two distinct
stages of one and the same language.

104. (55) Many works of comparative philology,
however, employ another arrangement. In this each
case is dealt with more by itself, so that either
(as in Schleicher's Compendium) the accusative
singular, for example, is treated separately in each
language, or (as in Brugmann's Grundriss) the mode
of formation of one definite case in one definite class
of nouns (i-stems, etc.) is followed out through all the
allied tongues. According to this arrangement all
those facts are brought into a single class which are
related to one another from the point of view of a
student of comparative philology; but, as an inevitable
consequence, the survey of the forms of any one
language (or stage of language) is obscured; the
unity of time and place is effaced; and, moreover, we
get only a formal conception of the phenomena. The
morphological element has been brought to the
front at the expense of the syntactical, which has to
be treated in another section, so that the constant
reciprocal action of form and function is generally
lost sight of.139

105. (56) Lastly, we come to that I will term the
purely grammatical arrangement. The grammar of a
language is, as it were, an answer to the question,
What general means of expression does such and
such a language possess? 185 Now, by the purely
grammatical arrangement the methods of expression
existing in a particular language at a particular time
are tabulated in such a manner that those forms come
together which are grammatically analogous. By this
arrangement, forms which belong together from a
dictionary point of view, e.g., dæg, dæge, are wrested
from one another, and the same may be the case
with forms which belong together historically, e.g.,
Old English nominative plural neuter hof-u and
word; it is true that they were once formed with the
same ending, but an Englishman of King Alfred's
time could not possibly be aware of this point of
agreement. Clearly by this mode of treatment the
individual element, by which I mean that which is
peculiar to each language or to each successive stage
of language, is brought more distinctly into view;
we are, moreover, enabled to survey the potentialities
of development of each particular language: we see
plainly where the differences between the various cases
are so well marked that they can easily be kept
distinct, and where they bear such a close resemblance
to each other in form or function, or in both alike, as
to run the risk of being levelled and blended.

In an ideal language it would be an easy matter to
140carry out such an arrangement: since each modification
of meaning would have its own expression, which
would be constant for all cases and quite unambiguous,
a separation of accidence and syntax would be
precluded, ipso facto; whether we should say, the
genitival relation is expressed by -a, or -a denotes
the genitive, would be quite immaterial.

The two arrangements stand to one another as the
two parts of a dictionary, in one of which the form
(say, some German or French vocable) is given, and
the signification sought (in other words, the English
equivalent is appended): in the other, the meaning is
the known quantity, and the appended part is the German
or French term which was required to be known.141

107. (58) Before attempting to give a synopsis,
arranged upon these principles, of English case-systems
at different epochs of the growth of the
language, I have to premise with regard to Old
English that, as a matter of course, I shall have to
give, in the main, West-Saxon forms, though for a
thorough understanding of the historical process of
development of Standard English it would have been
better if I had been in a position to avail myself of a
Mercian, or, still better, a London grammar representing
the language as spoken about the year 800.
Again, in stating the function, I shall have to be very
brief, and content myself with merely giving names,
leaving it to the reader to understand by “dative”
(for example) — not the notion of dative in itself, for
such a notion has no existence, but — “Old English
dative”. For the particular use which English people
of a thousand years ago made of their dative case, I
must refer to the Old English syntax, which is,
unfortunately, still to be written. In the present
chapter I can give nothing but a skeleton-like scheme,
which does not aim at completeness.

108. (59) It will not fail to meet with general
approval that, in drawing up this scheme, I have
followed Sievers's excellent Angelsächsische Grammatik
(2 Aufl., 1886). In accordance with my
general views, however, as stated above, I shall differ
from Sievers in paying much more regard than he
does to what would naturally appear to King Alfred
and his contemporaries as the significant element in
142language: I shall have to separate word and case-ending,
as far as this is feasible, in the same manner
as the instinctive linguistic sense of that time would
have done, regardless of the prehistoric-condition of
things. Old English eage, for instance, is historically,
it is true, an n-stem; but for my present purposes
I shall have to look upon it as consisting of eag + the
nominative ending -e, the genitive being eag + an, and
so on. We want a special term for this distinction;
and I propose to call the substantial part of the word,
felt as such by the instinct of each generation as
something apart from the ending (eag in the example
chosen), the kernel of the word, while eagan is the
historic “stem”. No doubt, in some cases it will
depend on a more or less arbitrary choice, how much
of the traditional form is to be treated as kernel and
how much as ending. For instance, eage itself might
be said to be the kernel, the genitive ending being -n,
before which the e of the kernel is changed into a.
This division would, however, seem to be unnatural
for Old English; although so much must be granted,
that in Middle English we must look upon eie (not
ei) as the kernel, to which the ending -n is affixed in
the nominative plural. 186

The fact is, that along with the perpetual wearing
away of words there is often an alteration in the
feeling as to the relations of kernel and ending.
143Now a little more, now a little less may be included
in one or the other, exactly as when one generation
considers the sound-combination anaddere as consisting
of a + naddere, whilst the next looks upon it as
an + addere (Modern English., an adder), or when mine
uncle is transmuted into my nuncle.

109. (60) It will be seen that if Old English eage
is said to be an n-stem, what is meant is this, that
at some former period the kernel of the word ended
in -n, while, as far as the Old English language
proper is I concerned, all that is implied is that the
word is inflected in a certain manner. If, therefore,
in the following pages, I shall speak of n-stems, i-stems,
etc., it is only as designations for classes of
declension. It follows, however, from my view that
we are not properly entitled to put down, e.g., wyrm
as an i-stem, for by doing so we should fail to give
a true picture of the real condition of things in the
Old English period. If a modern linguist is able to
see by the vowel-mutation (umlaut) that wyrm was
an i-stem, an Englishman of that time could not
have suspected any such thing, as the endings of the
several cases of wyrm are identical with those of
(the o-stems, e.g.) dom. When Sievers reckons wyrm
among i-stems, or gives sige as an es- os-stem, he
is writing for the benefit of those who take only
a secondary interest in Old English grammar, and
care chiefly for the way in which it reflects prehistoric
phenomena. He is thinking little of those other
students who make the first object of their investigation
144the mutual relations of the facts of a language
at a definite historical epoch, and who go to the
study of Old English partly for the sake of seeing
the mechanism of this particular idiom as an organically
connected whole, partly with a view to seeking
in it the explanation of later developments of the
English language.

Italicised letters indicate the stem (class of declension):
–o (words like dom, hof, word; by others
termed a-stems), i, etc.; c = those consonantal stems
which do not form part of some larger group, such as
n, r. What is said about the â-class applies likewise
to the wâ-stems with a long vowel or a diphthong
preceding the w (S, § 259), so that, in mentioning wâ,
I only mean those in which the w is preceded by a
consonant (S, § 260); the jâ-stems are only referred
to when they present deviations from the other â-stems
(g p); abstr.= words like strengu (S, § 279).
n a p n ob must be read: nominative and accusative
plural of neutral o-stems consisting of an originally
short syllable.

I. Morphological classification.

111. (62) The Old English language used the
following formal means to denote case-relations: —

A. The kernel of the word unchanged.

(1) n a s. o, jo (except lm), wo, i (l)f, u lmf, r, nd, c
mn, c lf [dom hof word, here secg cyn(n) rice westen,
bearu searu (beadu), ben, feld hond, fæder modor,
freond, fot scrud, boc]. — Also N i b [wlit, S, § 263,
anm. 5].146

(2) n s f. (not a s.) â l, jâ (wâ) [ar, sib(b) gierd
(beadu)]; L also i (l)f [ben], -e being used in a s.

(3) d s. some o [(æt) ham, (to) dæg and a few
more, S, § 237, anm. 2], of r only fæder sweostor; r. u
lf [hond] and s [dogor S, § 289]; L c lf [ac, etc., S, §
284, anm. 2].

(4) g s. r 188 [fæder brođor, etc.]; r. L u lf [hond].

(5) n a p. o ln, jo bn, wo, c n [word, cyn(n), searu,
scrud]; also, though not exclusively, some r [brođor
dohtor 289 sweostor], nd [freond hettend], c m [hæleđ
monađ], s n [lamb for lambru by a complete transition
to the o-class].

B. Vocalic endings.

112. (63) –a.

(1) n s m. n [guma; N also f]; L u b [suna].

(2) a s m. L u b [suna].

(3) d s. u [suna 390 felda 491 dura 392 honda 593], also often
words in -ung [leornunga, S, § 255]; also mæda, S, § 260.

(4) g s. u bm, f [suna, 694 dura honda; r. lm felda 695].
(5) n a p. u bm [suna], 796 f [dura 397 honda]; r. u lm
[only hearga 798 appla 799]. — i lm r. [leoda]. — â [giefa 4100
147ara 1101], also instead of -e in i lf and abstr. [bena,
strenga]. — And finally L o bn [hofa, S, § 237, anm. 5].

(6) g p. wherever the ending is not -ana, -ena, -ra,
see below [doma 2102 hofa 2103 worda, 2104 her(i)g(e)a secg(e)a
enda cynna ric(e)a westenna, bearwa searwa, giefa 2105
ara, 2106 sibba gierda, beadwa mædwa, win(ige)a spera,
bena, suna felda dura honda (strenga?), fota scruda
hnuta boca fæd(e)ra freonda]; r. n [bæcistra, S, § 276,
anm. 1]. -a is also found in g p. in neutral adjectives
when used as substantives [goda], Cosijn Altws. Gr.,
ii., § 49.

113. (64) –e.

(On ifor classical O. E. e, see S, §§ 132 f, 237 anm. 2,
246 anm. 1, 252 anm. 1, 263 anm. 1, 269 anm. 2.)

(1) n a s. jo lm [ende], i bmn [wine spere] bf
[only dene 3107], jâ r. [-nisse -nysse, generally -nes], n
nt [eage].

(2) n s. n f [tunge 3108]; N also r. m.

(3) a s. â [giefe are]; abstr. [strenge]; L also i lf
[bene 4109].

(4) d (i) s. (on the difference between the older
instr. in -i (-y) and the dative in -ae, see Sievers, P. B.
148Beitr., viii., 324 f.; in classical O. E., this distinction is
no more found) — everywhere except u and n and the
rest of consonantic stems, where, however, -e begins
to crop up (S, §§ 273 anm. 2, 274 anm. 1, 280 anm. 2,
281, 286). Accordingly -e is found, e.g., in [dome
hofe worde, her(i)ge secge ende cynne rice westenne,
bearwe searwe, giefe are, sibbe gierde, headwe
mæd(w)e, wine spere, bene, strenge; felde for older
felda, r. dure nose flore eage fote freonde]. — Also
neutr. adj. used as substantives [gode], Cosijn, ii., §
49.

(5) i s. distinct from d s. only in some pronouns
and st adj. [micle]; it occurs comparatively seldom,
see Cosijn, ii., §§ 38-48.

(6) g s. â [giefe are], i lf [bene], abstr. [strenge],
c bf [hnute] lf [burge boce, etc., used concurrently
with mutated forms; ace muse and others without
mutation, S, § 284, anm. 1]; r. u f [dure S, § 274,
anm. 1].

(7) n a p. i bm [wine 1110 -ware], lm a few words
[Engle], lf [bene 2111], thence also â [giefe are]; st
m(f) [gode], also nd polysyllabics [hettende, besides
-nd, -ndas].

(8) Mutated d s. og n a p. c bf [hnyte].

114. (65) – u.

(On -o see S, §§ 134 f, 237 anm. 4 and 5, 249, 252,
269 anm. 2 and 5, 279.)

(1) n s. u b [sunu duru]; â b [giefu], abstr.
[strengu], c bf [hnutu].149

(2) a s. u b [sunu duru]; L â b and abstr., S, §§
253 anm. 2, 279.

(3) d s. u b [sunu duru; generally –a], â b and
abstr. as in (2).

(4) g s. L â b and abstr. as in (2).

(5) n a p n. o b [hofu; L also l: wordu, see on
polysyllabics, S, § 243],jo l [ric(i)u] and polysyllab.
[westen(n)u], (wo: u for -wu, searu), i b [speru],
similarly st b which have however often -e from m
[hwatu].

n a p m f u b L [sunu duru]; r [brođru dohtru,
which form also other plurals].

(6) (i s. horu Elene 297 from horh.)

C. Nasal endings.

115. (66) –um.

(1) d s. st. [Þiosum, godum]. — ? miolcum, heafdum,
see Kluge, Pauls Grundr., i., 386.

(2) d p. everywhere [domum hofum wordum,
her(i)gum secg(i)um endum cynnum ric(i)um westennum
bearwum searwum, giefum arum, sibbum gierdum,
nearwum, winum sperum Englum, benum, sunum
feldum durum hondum, gumum 1112 tungum eagum,
strengum, fotum hnutum bocum, fæd(e)rum, freondum,
lombrum L lambum].

On -an, -on for -um see § 116.

–m.

(1) d s. pron. [him đæm hwæm].

(2) d p. in some words after a vowel, for -um
150[cneom beside cneowum, S, § 250, nr. 2; fream, etc.,
S, § 277, anm. 2], numerals [twæm Þrim].

116. (67) –an (–on).

(1) d g s. and n a p. n [guman tungan eagan].

(2) a s. n m. and f. [guman tungan].

(3) n s. L weak adj.

(4) for -um L.

(5) g p r L [eastran, S, § 276, anm. 1; weak adj.
§ 304, anm. 2].

–n
for -an in some words after a vowel [frean, etc., S,
§ 277, anm. 2; beon tan, S, § 278, anm. 2].

117. (68) –ena [N ana].
g p n [gumena tungena eagena]; L also in o and
â, especially b [carena, S, § 252, anm. 4], not jâ.

–na.
g p in a few words [sceona, etc., S, § 242, anm. 2,
N treona, § 250, nr. 2; Seaxna, etc., § 264; n l after
r and g: larna eagna, § 276, anm. 1, oxna, § 277, anm.
l, gefana Sweona, § 277, anm. 2].

118. (69) –ne.
a s m. pron. [hi(e)ne Þone Þi(o)sne hwone] and st.
[godne].

D. Endings containing s.

119. (70) –as.

n p m. o [domas],jo [her(i)g(e)as endas], wo [bearwas],
u l [feldas], r only fæderas; becomes moreover
frequent in i [winas], u b [sunas], nd [also -ras:
wealdendras, S, § 286, anm. 2].151

(G s. in -as r.; perhaps Beowulf, 63, 2453, 2921.)

120. (71) –es.

(1) g s m n. o [domes hofes wordes], jo [her(i)ges
secges endes rices westennes], wo [bearwes searwes],
u l [feldes], nd [freondes hettendes], c m [fotes]; -es
becomes frequent in u b [sunes], n [eages eares], r
[fæderes]; N also in most other stems.

(2) n a p. for -as L, S, § 237, anm. 3.

–s.
g s. very rare: eas (Oros., 17, 23; Chron., 896, 918,
919, 922) cus, S, 284, anm. 4, sæs, S, § 266, anm. 3
(also n a p).

E. Other endings.

121. (72) –ra.

g p. p r o n. [hiera (heora) Þara], st. [godra], nd
polysyll. [hettendra]; 1113 = r + a: s n [lombra cealfra,
etc.; cildra also in texts which in n p have cild].

–re
g d s f p r o n. [Þaere Þisre], st. [godre].

–or, –ru.
n a p n s [lomber, see Schmidt, Pluralb., 149,
lombru 2114].

–rum.
d p n s in the same words as -ru.

–đ
might be considered a case-ending in hæleđ, monađ,
ealođ, d g s, n a p; but the words are generally inflected
regularly.152

F. Changes in the kernel.

G. A totally different kernel.

Frequent in pron. [ic — me — wit — unc — us, etc.;
se–Þone, etc.].

123. (74) Those were the means used in Old
English to denote case-relations; but we have not in
our lists mentioned all the changes undergone by
Old English words, for alongside of these significative
changes we find a great many others which do
not play any part in distinguishing cases. I shall
briefly indicate the most important of these incidental
changes.

(1) I-mutation, in isolated cases of i s. o [hwene,
æne, S, § 237, anm. 2], in d s. c bf [hnyte] and r. u
[dyre]. Where the i-mutation is found through
all cases as in cynn, it does not concern us here.153

(2) U-mutation, o n a d p n [gebeodu from gebed;
it disappears at an early period, leaving perhaps but
one tracc. in the differentiation of cliff and cleeve, see
Murray's Dict. and my Studier over Engelske Kasus,
§ I98]; other instances of u-mutation, see S, §§
241, 253, anm. 1, Cosijn, ii., p. 3 (cneoht); comp. also
cucu, cwices, Sievers, P. B. Beitr., ix., 259.

(3) Interchange of æ and a, found with greatest
phonetic regularity in st. adj. [hwæt, hwates hwate],
while in the nouns (of the o-class) æ is carried through
in the singular and a in the plural [dæg, dæges –
dagas]. After a palatal consonant we have the
peculiar change seen in geat, gatu, which is by-and-by
levelled out in different ways. Note also gærs,
grasu. For the still more complicated change in
magu mæcge(s), plural mæcga(s) magum, see Kluge,
Literaturblatt f. germ. u. rom. Philol., 1889, 134, and
Paul's Grundriss, i., 368.

(4) Interchange of long æ and long a: mæg, magas;
in ān ænne, long a and short æ interchange.

(5) Interchange of single and double consonant:
cyn, cynnes, S, § 231; in the nominative cynn is also
found, and it is not easy to see if the difference is
only a graphical one or indicates a real difference in
pronunciation. There is a tendency to utilise the
difference for sense-distinguishing purposes in mann,
“man,” and man, corresponding to French homme –
on, or still more closely to Danish mand, man, see
Cosijn, ii., p. 47.

(6) Interchange between final voiceless and medial-voiced
154consonants: wulf, wulves (written wulfes), hus,
huze (written huse), bæÞ bađas; see my Studier over
Engelske Kasus, § 193 ff.

(7) The related interchange between h and g: beah,
beages; h also interchanges with w: horh, horwes,
the adj. ruh, ruwes (old “grammatical change,”
determined by Verner's law), and finally there is
often an interchange between h-forms and forms with
no consonant, but with contractions and perhaps
lengthening of the vowel: furh, furum (? fūrum),
sc(e)oh, sc(e)os, feoh, dative, feo. Here we very
often see levellings, the h-less form being as a rule
generalised.

(8) Interchange between forms with and forms
without w: treo, treowes, later on levelled both
ways: treo, treos; treow, treowes; compare also
sna(w), S, §§ 174, nr. 3, 250, anm. 1. The forms are
differentiated in æ “law” and æw “marriage,” S, §
269, anm. 3.

(9) Interchange between e or i, u or o and the
corresponding vowel-like consonants j and w: here,
herias, herigas, hergeas, herigeas; bearu, bearwas (L
bearuw, bearuwas).

(10) Interchange between the advanced and
palatalised open g in dæg and the back open g in
dagas; 1122 so also byrig, burgum. In the latter word
155we have four sound changes: (a) the vowel of the
principal syllable; (b) the vowel of the svarabhaktisyllable,
which is also often left out; (c) the voiceless
and voiced consonants, see above sub 6 and 7; (d) the
palatalised and unpalatalised consonants.

(11) Vowel change in unstressed syllables, due to
an old gradation (ablaut): -ung, ingum (S, § 255,
anm. 1; see however Cosijn, ii., pp. 21, 22); brođor,
bređer; morgen, mergen; see, for instance, Oros., L.
194, 12, on mergen = C. 92, 40, on morgen. 1123

(12) Interchange between a full vowel in final
syllables and a weakened one in the middle of the
word: rodor, roderas, S, § 129.

(13) Interchange between preserved and omitted
weak vowel: engel, engles; deofel, deofles; see
especially S, § 144. At a later period this leads sometimes
to a differentiation of consonants, pointed out for
engel by Napier, see the Academy, March 15, 1890, p.
188.

(14) Interchanging vowel quantity is probable
before many consonant groups; an indubitable case
in point is cīld, cildru.

124. (75) A comparison of Old English with Proto-Arian
will show that a good many case-endings have
been given up, and that similarly the change of accent
and that of vowels (by gradation) have disappeared
from the declension; nor does the Germanic interchange
of consonants according to Verner's law play
156any part in the declension (compare, however, §
123, 7 and 11). 1124 Wherever the Old English language
shows traces of these phonetic changes, it is always
so that one form has been carried through in all
cases, so that the other is only shown by the corresponding
word in other connected languages, or by
other derivatives from the same root. See on these
traces especially Joh. Schmidt in Kuhn's Zeitschrift,
XXVI., p. 8 ff., and Pluralbildungen der idg. Neutra,
passim; Kluge in Kuhn's Zeitschr., XXVI., p. 92 ff.;
and in Paul's Grundriss, i., p. 387 f.

125. (76) It is of greater importance to our subject
to examine the extent in which cases which were distinguished
either at an earlier stage of the language
or in other Old English words, have coalesced in one
and the same word. Such coalescence of cases is found
very frequently, though sometimes the form which
is identical with that used in another case is not the
only one in use for that particular case.

(1) a s. = n s. in all words except (a) â [giefu ar,
accusative giefe are]; from this class the distinction is
transferred to i l [ben, bene, instead of the older ben,
ben], while on the other hand the late O. E. levelling,
by which for instance lufu comes to be used through
the whole of the singular, obliterates the distinction.
(b) n mf [guma tunge, accusative guman tungan].
(c) pron. and st. mf.

(2) d s. = n a s.: (a) in some o-stems in certain
connexions [ham, etc., see § 111, 3], also treo and
157similar words, (b) jo l [ende rice], (c) i mnb [wine
spere]. (d) u b [suna and sunu; duru]. (e) fæder
sweostor; also L r ac boc, etc.

(3) d s. = a s. besides the words mentioned under
(2): n mf [guman tungan].

(4) instr. = dative everywhere except in some pron.
and st. mn., even there not strictly distinguished.

(5) g s. = n s.: r [fæder brođor, etc.], r u bm [suna].

(6) g s. = a s.: â, jâ, wâ [giefe are sibbe gierde
beadwe mæd(w)e], n mf [guman tungan], r [fæder,
etc.]; L i lf [bene], u bm [suna].

(7) g s. = d s.: â, jâ, wâ; lf [bene], u [suna dura
honda, r. felda], n mfn [guman tungan eagan], c lf
[bec, etc.], r [only fæder sweostor], pron. f [hiere
Þaere Þisse Þisre], st. [godre].

(8) n p. = n s.: o ln [word], jo bn [cynn], wo n
[searu], i bm [wine], u bm [suna and sunu], u bf
[duru], r: brođor dohtor sweostor, nd [freond hettend].

(9) n p. = a s. besides those under (8): n mf
[guman tungan], L also â [giefe, are], jâ [sibbe
gierde], wâ [beadwe mæd(w)e], i lf [bene].

(10) n p. = d s.: i bm [wine], i lf [bene], u [suna
and sunu, felda dura honda], n [guman tungan eagan],
c [fet hnyte bec], r: sweostor, nd [friend hettende];
also L the f mentioned in the end of (9).

(11) n p. = instr. s.: st. m [gode].

(12) n p. = g s: u [suna felda dura honda], n
[guman tungan eagan], c lf [bec], r: brođor dohtor
sweostor; L the same words as in (9) and (10);
finally L m when -es came to be used for -as.158

(13) a p. = n p., so that the numbers (8-12) apply
also to a p; the only exceptions are: we — us(ic),
ge — eow(ic).

(14) d p. = d s.: pron. [Þæ m Þi(o)sum], st. [godum],
also weak adj. fgodan], S, § 304, anm. 3.

(15) g p. = n a s.: u bm L [suna].

(16) g p. = d s.: u [suna felda dura honda].

(17) g p. = g s.: u [suna felda dura honda].

(18) g p. = n a p. â [giefa ara], jâ [sibba gierda],
wâ, [beadwa mæd(w)a], i lf [bena], u [suna felda
dura honda], r: dohtra.

126. (77) This list, which does not include indeclinabilia
like strengu, shows that the chances of
mistakes were pretty numerous in Old English
declensions. Take the form suna; it may be any
case, except only dative plural; sunu is everything
except genitive (singular and plural) and dative
plural; dura is everything except nominative, accusative
singular and dative plural; fæder may be
any case in the singular; so also sweostor, which may
moreover be nominative or accusative plural; the
only thing we can affirm on such forms as guman or
tungan is that they are neither nominative singular,
dative plural, nor genitive plural, and in a late text
we cannot even be sure of that, and so on.

II. Syntactical classification.

128. (70) The Old English language has no expressions
for the following syntactic categories, which
were found in the Arian parent speech: (1) the dual
number; the only exceptions are wit, unc(it) uncer
and git, inc(it), incer; the nouns duru, nosu, and
breost, in which traces of the old dual have been
found by comparative philologists, were no doubt
during the whole of the Old English period, and perhaps
even much earlier, felt as singulars, and sculdru
as a plural; (2) the vocative case, unless one feels
inclined to consider the use of the definite form of the
adjective in leofa freond, etc., as a sort of vocative. 1125160

Finally, three or four cases have coalesced to form
the Old English dative, the old instrumental being,
however, in some words distinct from the dative.

[In Modern English]

The three forms sound alike, and the systematic
difference now made between them is quite recent.
Before the middle of the eighteenth century they were
all of them written alike; thus we find for instance
in the original editions of Shakespeare, Kings, ladies,
for the three cases. The apostrophe was at that time
used (without any regard to case-function) where a
syllable was added in pronunciation (Thomas's), or
where the spelling -es was still commonly used, the
apostrophe being then used to indicate that no new
syllable was to be pronounced (compare the modern
spelling stabb'd); in Shakespeare you will find, e.g.,
earth's as a genitive singular and prey's as a nominative
plural. Sometimes the apostrophe is even in our
days used before the plural ending; thus in Shakespeare's
161Twelfth Night (ii., 5, 96) the spelling “her
very C's, her U's, and her T's” is kept unchanged in.
modern editions; and the same manner of spelling
may be found also in proper names, especially when
they are not familiar to English readers (Hrolf's, in
Carlyle, Heroes, 29); similarly in fly's (carriages) as
opposed to the more familiar flies; compare also the
Spectator, No. 80, where Steele speaks of the manner
in which people use “their who's and their whiches”. 1126
Conversely the apostrophe is not written before every
s denoting the genitive: whose, its, hers, yours being
the received spelling, while it is true that some people
write her's and your's.

In dealing with the forms of the spoken language
I shall, however, for convenience' sake give them in
their usual spelling, though it would, of course, have
been more consistent had I written all my examples
phonetically. The abbreviations will be the same as
in the Old English section, as far as they are needed;
“a.” means the modern accusative, dative, or common
oblique case (him, etc.); “abs.” stands for the
absolute form of the possessive pronouns (mine, etc.).

I. Morphological classification.

A. The kernel of the word unchanged.

130. (81) A. The kernel of the word unchanged.

(1) n a s. in all words; as exceptions might be
mentioned those few pronouns which have separate
forms for the accusative (me, us, him, her, them).162

(2) n a p. (a) you. (b) sheep and deer 1127 (c) the
ordinary compounds of -man, gentleman and gentlemen
being pronounced alike; so postmen, policemen, etc.
(d) some words ending in -s [z]: e.g., means, species.
(e) many words are unchanged in the plural in
special connexions, especially after numerals and
collectively: six pair of gloves; twenty-three snipe;
people, fowl, fish, cattle, etc.

B. The ordinary s ending.

131. (82) B. The ordinary s ending.
(that is: the sounds -iz added to a sibilant [s, z, Sh.,
zh];
the sound -s after a voiceless non-sibilant;
the sound -z after a voiced non-sibilant)

(a) g s. in all nouns and some pronouns: prince's,
duke's, king's, whose, somebody's.

(b) n a p. in the majority of nouns and some
pronouns: princes, dukes, kings, somebodies.

(c) g p. in the same words as under (b), if the g p.
can at all be used: princes', dukes', kings' (somebodies').

(d) The same ending denotes the idea of genitive
in all those plurals which are not formed by the
addition of -s: men's, gentlemen's, children's.

(e) absolute: ours, yours, hers, theirs.

C. Other endings.

132. (83) C. Other endings.

-s.

n a p. in dice; comp. also pence, halfpence.163

-n.

(a) n a p. in oxen.

(b) abs. in mine.

D. Change in the kernel.

133. (84) D. Change in the kernel.

(1) without any ending.

n a p.: men, women, geese, teeth, feet, mice, lice.
The plural forms these and those might be mentioned
here or perhaps better under (3), as -se [z] is felt as
a sort of plural ending.

(2) with the ending -ren (or -n).
n a p. children (brethren).

(3) with the -s ending.

n a (g) p. wives (and wives') and others in f; paths
and others in th, houses, the change in the kernel
consisting here in the substitution of the voiced for
the voiceless sound. 1128

As an ulterior case in point might be mentioned
the frequent omission of the Þ- sound in such plurals
as months, sixths, elevenths, etc. In words ending
in -nd the plural is frequently pronounced without
the d: soun(d)s, etc. We are perhaps allowed to
consider Shakespeare's rhyming downs and hounds
together (Venus and Ad., 677) as an early instance
of this pronunciation.

(4) an entirely new kernel
is finally used to distinguish cases in some pronouns:
I, me, we, us, etc.164

134. (85) Coalescence of formerly distinct cases is
found very extensively.

n a p. = n a s. in the words mentioned above, A 2.

g p. = g s. consequently in nearly all the same
words.

The three cases: gen. sg., nom. (and acc.) pl., and
gen. pl., have become identical in nearly all words, so
that you can very soon enumerate the very few words
in which they differ from each other, namely: —

All the three cases are different: child's, children,
children's; similarly with man, woman, and finally
with a few words where the gen. pl. is, however,
scarcely used at all: tooth, goose, mouse, louse; dice,
pence, oxen; compounds on the model of son-in-law
would belong here if genitive plurals, like sons-in-law's,
were not universally avoided.

g s. different from n a p., which is identical with g
p.: wife's, wives, wives' and the other word mentioned
under D 3.

The two genitives are different from the two nominatives
in the nouns mentioned under A 2.

135. (86) A comparison with Old English will
show that all the vocalic and most of the nasal case-endings
have been abandoned; the changes of the
kernel have been considerably limited so that more
particularly those which were not in themselves
sufficient to distinguish cases have been given up;
further we see that one difference, which was unknown
to Old English., has been made subservient to case-distinguishing
purposes (O. E. genitive wulfes, nominative
165plural wulfas, both of them pronounced with v;
modern, wolf's, wolves), and finally the provinces of
the unchanged kernel and of the s form have been
very considerably extended.

II. Syntactical classification.

136. (87)

Here, as in a few places above, I have silently
omitted the exceptional forms of the personal pronouns.

137. (88) A comparison with Old English will here
show that — apart from a few pronouns, which distinguish
a nominative and an objective case — the old
nominative, accusative, dative and instrumental cases
have coalesced to form a common case, which shows
moreover a few traces of the fact that the old genitive
plural grew to be formally identical with the common
case of the singular number (e.g., a twopenny stamp, a
five pound note).

138. (89) The question naturally arises, How has
it come about that the Old English system of declensions
has been so completely metamorphosed? Is it
possible to point out any single cause as the effectual
agent in bringing about this revolution?166

139. (90) Such an influence from Norman-French,
however, is contradicted by various considerations,
partly of a general, partly of a special nature. It
would, indeed, have been at least imaginable, supposing
167that the two constituent elements of the population,
the French-speaking and the English-speaking, had
been co-equal in numbers. But this was not the case.
Moreover, it is admitted that the vast majority of the
conquered people spoke English and never learned
to speak French; they were not, therefore, exposed
to having their sense of the grammatical structure of
their native dialects impaired by commixture with
foreign modes of speech. And, where influence from
the foreign idiom could not be avoided, it must have
taken place essentially in the same manner as French
and English influence each other at the present day,
by the adoption, that is, of single words, which are
then incorporated, substantially, into the native system
of grammar. 1130 Just as a modern Frenchman inflects
the loan-words leader, sport, in accordance with the
laws of his own language, and turns the English verb
stop into stopper (stoppant, etc.), — just as, when some
composite expression passes into his language, he does
not shrink from forming such a derivative as strugg(le)-for-lifeur
(Daudet), — precisely in the same manner did
the English peasant act when he caught up a word
from the courtly speech of the Normans. Quite
instinctively he affixed to it his own terminations
without troubling himself for a moment whether they
would or would not “agree with the new roots”.

140. (91) But, whilst the Norman Conquest exerted
no direct influence on English grammatical
168structure, there can be no doubt that it went far
to accelerate the development of change indirectly.
This was principally due to the fact that England
was for some centuries without that retarding and
conservative influence which will always make itself
felt wherever cultivated classes speaking a “refined”
speech exist side by side with a proletariat whose
linguistic peculiarities are branded as vulgarisms, or
as downright solecisms. Any such control as comes
from an upper class whose more old-fashioned language
is looked upon as a model, and, partly at least,
imitated by the lower classes, was precluded at the
period we are speaking of, inasmuch as the upper
classes did not speak English., or, at best, spoke only
bad English. In consequence of this, not only was
the literary tradition of the English language lost or reduced
to a minimum, but even in its oral transmission,
which is always the more important matter, and was
especially so then, one element was wanting which
generally assists in stemming the tide of revolutionary
tendencies.

141. (92) If now we look at the only detail in
English accidence for which a Norman descent is
claimed (namely, the plural -s 1131), some remarks will
169have to be made which perhaps have not been all
propounded before.

(1) The growth of the plural -s cannot be separated
from that of -s in the genitive case. Now the latter
gained ground even more rapidly and extensively
than the plural -s, and French influence is here utterly
unimaginable. Why, then, resort to it with regard
to the other ending?

(2) The plural in -s was long before the Conquest
extended to many nouns which had formerly had
other endings, belonging to the i- and u- classes, as
also to some of the consonant stems (wyrmas, winas,
sunas, hæleđas, etc., see § 119). This shows that the
tendency of the language would have been the same
even if William the Conqueror had never crossed the
Channel.

(3) -S became universal in the North at an earlier
date than in the South, where we should expect to
find French influence strongest, but where -en seems
for a long time to have had better chances of prevailing
in all nouns than -s.

(4) In Old French -s was not used to the same extent
as now as a plural ending; indeed, it can hardly
be called a plural sign proper, as it was in the most
numerous and important class of nouns the sign of
the nom. sg. and of the acc. pl., but not of the
nom. pl. If, therefore, an Englishman of (say) the
thirteenth century used the -s in the nom. pl., he was
in accord with the rules of his native tongue, but not
with those of French.170

(5) If -s was due to the Normans, we should expect
it in the plural of the adjectives as well as of nouns;
but, as a matter of fact, adjectives take it extremely
rarely, 1132 and hardly except in those cases where a
Romance adjective is placed after its noun. Everywhere
else, Middle and Modern English adjectives
have no -s in plural, agreeing therein with the old
native tradition, but not with French grammar.

(6) And, finally, it is worth noting that the two
endings, Norman -s without any vowel, and English
-es (originally -as) with the vowel pronounced, were
kept distinct for about four hundred years in English;
they are not confounded till, in the fifteenth century,
the weak e disappears in pronunciation.

142. (93) Thus, at the one definite point where
the theory of French influence has been advanced
with regard to accidence, it is utterly unable to stand
the test of historical investigation. And it is the
same case, I believe, with many of the assertions put
forward of late years by E. Elnenkel with regard
to a French influence exerted wholesale on English
syntax. 2133 Einenkel's method is simplicity itself. In
171dealing with any syntactical phenomenon of Middle
English., he searches through Tobler's Verm. Beiträge
zur Frz. Grammatik and the ever-increasing literature
of German dissertations on Old French syntax, in
quest of some other phenomenon of a similar kind.
As soon as this is discovered, it is straightway made
the prototype of its Middle English analogue, sometimes
in spite of the French parallel being perhaps
so rare a use that even Tobler himself can only point
out a very few instances of it, whilst its English
counterpart is of everyday occurrence. In several
cases French influence is assumed, although Einenkel
himself mentions that the phenomenon in question
existed even in Old English., or, not unfrequently,
though it must be considered so simple and natural
a development as to be quite likely to spring up
spontaneously in a variety of different languages. A
little knowledge of Scandinavian languages would,
for example, with regard to many points have convinced
Einenkel that these present the very same
phenomena which when occurring in English he
explains from Old French.

143. (94) A far greater influence than that exercised
upon English by the Gallicised Normans must
be ascribed to the Danish. Wikings, who for such a
long space of time were acting a prominent part in
Britain, and whose significance for the life of the
172English people cannot easily be over-estimated. As
for the language, it should be borne in mind that the
tongue spoken by the Danes was so nearly akin with
the native dialects that the two peoples could understand
one another without much difficulty. But it
was just such circumstances which made it natural
that many nuances of grammar should be sacrificed,
the intelligibility of either tongue coming to depend
mainly on its mere vocabulary. It is in harmony
with this view that the wearing away and levelling of
grammatical forms in the regions in which the Danes
chiefly settled was a couple of centuries in advance of
the same process in the more southern parts of the
country.

A fully satisfactory solution of the question of the
mutual relations of North English and Scandinavian
at that time must be regarded as hopeless on account
of the small number, and generally inadequate
character, of linguistic records; and, unless some
fresh sources become accessible to us, we shall
probably never learn clearly and unequivocally which
points of correspondence in the two languages are
attributable to primitive affinities, which others to
loans from one language to the other, or, finally, how
much may be due to independent parallel development
in two areas which offered such striking
analogies in so many essential particulars. But, as
I hold, any linguistic change should primarily be
explained on the basis of the language itself, while
analogues from other languages may serve as illustrations
173and help to show what in the development
of a language is due to psychological causes of a
universal character, and what is, on the other hand,
to be considered the effect of the idiosyncrasies of the
particular idiom.

In other words: a phonetic law which operates
“blindly,” i.e., without regard to the signification,
causes the Old English unstressed vowels -a, -e, -u, to
become merged in an obscure -e in Middle English;
as these endings were very often distinctive of cases,
the Old English cases were consequently lost.
Another phonetic law was operating in a similar
manner by causing the loss of the final -n, which was
equally utilised, though in a different way, in the
Old English declension. Upon this I have to remark,
first, that beside the phonetic laws must at all events
be mentioned analogy. It is this which, for example,
has led to the levelling of the nominative
plural and dative plural: if phonetic decay had been
174the only factor, Old English stanas and stanum would
still have been distinguished from one another, namely
as stones and stone; whereas, in fact, the former form
has been extended to the dative. This, however,
must by no means be interpreted as an objection to
Dr. Murray and the scholars who hold his view, and
who are as fully alive to this principle of explanation
as anybody else.

145. (97) I have stated elsewhere my reasons for
disbelieving in the axiom of the so-called young
grammarian school of the blind working of sound
laws, and in the theory of sound laws and analogy
sufficing between them to explain everything in
linguistic development. 1135 Here I shall add, with regard
to the special question concerning us in this
chapter, that the young grammarians' view does not
look deep enough in its search for explanations. If
simplification of forms is to be attributed in the main
to the phonetic law of unstressed terminations, what,
then, is the cause of the phonetic law? And if, on the
other hand, analogy has played an important part
in this development, the question arises, if it is not
possible to suggest causes why the principle of
analogy should have thus asserted itself.

Let us for a moment suppose that each of the
terminations -a, -e, -u, bore in Old English its own
distinctive and sharply defined meaning, which was
necessary to the right understanding of the sentences
in which the terminations occurred. Would there in
175that case be any probability that a phonetic law
tending to their levelling could ever succeed in
establishing itself? Most certainly not; the all-important
regard for intelligibility would have been
sure to counteract any inclination towards a slurred
pronunciation of the terminations. Nor would there
have been any occasion for new formations by
analogy, as the endings were already sufficiently
alike.

146. (98) The above comparative survey of the
declensions of Old and Modern English furnishes an
answer to the questions proposed, and makes the
whole causality appear in a much clearer light than
would be possible by any other arrangement of the
grammatical facts: the cause of the decay of the Old English
apparatus of declensions lay in its manifold
incongruities. The same termination did not always
denote the same thing; the same case was signified
now by this, now by that means; many relations
plainly distinguished from each other in one class of
words were but imperfectly, if at all, distinguishable
in another class. And yet there is a still further
cause of mixture and confusion which our arrangement
does not bring out — the one, namely, which is latent
in terms like dative, accusative, etc. In fact, these
terms have no clear and definite meaning in the case
of Old English., any more than in the case of kindred
tongues; in many cases it did not even matter which
of two or more cases the speaker chose to employ.
Thus, not a few verbs existed which were employed
176now with one, now with another case; and it was often
impossible to perceive any accompanying difference
of meaning. 1136 And so also with other parts of speech:
the preposition on, as applied to time, sometimes
governed the dative (instrumental), sometimes the accusative:
thus we find in close succession (Chron., 979,
C), on Þys geare…on Þone sunnandæig; (ibid.,992,
E.) on Þere nihte đe hi on đone dæi togædere cumon
sceoldon; 2137 similarly (Oros., 136, 23 and foll.) on
westeweardum Þisses middangeardes,… on easteweardum
Þeosan middangearde (comp. same page, l. 7),
and so on.

147. (99) This condition of things naturally gave
rise to a good deal of uncertainty, which manifested
itself partly in a rather inaccurate pronunciation of the
endings, partly in the use of them in places where
they did not belong.

This now and then happened in such a manner as
to bring about coincidences of sound without assisting
clearness, nay, even at its expense, as, for instance,
is the case when we find in the Cura Past., 166, 2
and 20: to anra đara đreora burga, instead of anre
(see Sweet's note in his A. S. Reader, p. 191).
Generally, however, such uses of endings on analogy
177are apt to crop up in such places particularly where
the traditional terminations are not sufficiently distinct,
or where cases have been levelled which it is
important should be kept apart. For example, giefa
stands alike for the nominative plural and the genitive
plural, and misapprehensions are the consequence.
These are obviated by the extension to the nominative
and genitive respectively of the termination -e from
the i-class and -ena from the n-class (nominative
giefe, genitive giefena).

But if the transmutations, phonetic as well as non-phonetic,
of the old declensions took their rise from
the numerous inconsistencies of the system and its
want of fixed boundaries, formal or functional, then
what is described above as the true grammatical
arrangement exhibits the prospects of the various
cases and endings in their struggle for existence.
By its aid we are, in some measure, in a position to
cast the horoscope of the whole system and predict
the main features of its destinies.

148. (100) The vocalic terminations (B) were
evidently the least distinct and least sharply defined;
each of these had many values, nor were they
uniformly distributed in the different classes of
inflexion. Here accordingly every succeeding generation
when it came to learning the language was
offered only scanty points of support and a great
many chances of going wrong. It is therefore not
surprising that these endings were confounded and
effaced and in a later period entirely dropped, as
178there was no well-defined barrier between the use of
the bare kernel of the word, and the kernel plus the
vocalic termination -e, in which the endings -e, -a, -u,
had at that time been merged.

The nasal endings were possessed of greater power
of resistance. But they, too, were doomed, chiefly
owing to the exceedingly common use of the ending
-an in the weak forms of adjectives, where it was of
no consequence whatever for the signification, and
could therefore be neglected without any loss. In
the case of verbal forms, too, where endings in -n
occurred also, they did not perform any function of
sufficient importance to check the tendency to drop
the sound in pronunciation; in fact, at an early
period we meet with collocations like binde we, binde
ge, mote we, etc., in which the -n had fallen away
(Siev., § 360).

149. (101) Where, on the other hand, the -n was
protected by a following vowel, it could withstand the
levelling tendencies better. This would be especially
the case in the genitive plural, because of the distinctive
meaning of this genitive. The same thing is
also particularly true of the two -s endings, each of
which was confined to a sharply limited sphere of use.
The -s is too important to be left out; if, on the other
hand, the two endings -as and -es are levelled in the
Middle English -es, this is mainly due to the influence
exercised by the other endings. As -a and -e were
not distinctive enough in point of meaning to oppose
a strong resistance to the tendency prevailing in all
179languages to obscure vowels in weak syllables, nay,
even invited this tendency, -as and -es had to submit
to the resulting “phonetic law”. This they did
without any very great detriment to intelligibility,
the connexion in which they occur being nearly
everywhere sufficient to show whether the genitive
singular or nominative plural was meant, especially
after the rule had been established by which the
genitive is always placed before its governing word
(see chapter viii.).

As regards the prospects which changes of kernel
have of maintaining themselves, we can only be
certain of this much, that those which have become
attended with inherent change of signification are, by
a natural consequence, more likely to be permanent
than the others, which are more liable to be affected
by levelling tendencies, inasmuch as a new regular
form which agrees with the shape of the word in other
cases is sure to be understood as well as, or even
better than, the traditional one. But, on the other
hand, forces tending to change pronunciation are
continually at work, and these give rise to fresh
changes of kernel; we may mention, for instance, the
laws of quantity which have split up the Old English
sceadu into the two Modern English words shade and
shadow. To foretell the durability of such modifications
is, of course, a matter of impossibility.

150. (102) To sum up, setting aside changes of
kernel, the other modifications of the nouns in Old
English declensions are of a character to enable us
180to form an opinion on the main features of their
destinies by considering the reciprocal relations of
phonetic expression and inward signification, the more
so as it was just the least ambiguous endings (-as,
-es) that were used to denote the syntactical relations
which are the most distinctive and appear to be
the most indispensable in language, viz., plurality
and connexion (genitive). Logically to define
the other case-relations is a matter of much more
difficulty: the dative and accusative cases often come
in contact with each other, and both have also some
points of agreement with the nominative. Hence
arises the chance of endless confusions, even where the
forms are sharply distinguished (see the next chapter). 1138
In fact, there is every occasion, be it said incidentally,
alike from a formal and syntactical point of view, to
prefer the arrangement of the cases prevalent in Denmark
since Rask — nominative, accusative, dative,
genitive — to any other, and more especially to that
still current in Germany, where the genitive is placed
between the nominative and the accusative.181

Chapter VII.
Case-shiftings in the pronouns.

151. (103) In the Oldest English pronouns we
find the nominative, accusative, and dative cases
distinct both in point of accidence and syntax, although
in a few pronouns there is no formal difference
between the nominative and accusative (in the
plurals of the third person (hie); in the neuter (hit,
hwæt, etc.), in the feminine form heo or hie).

The first step in the simplification of this system
is the abandonment of the separate forms mec, Þec,
usic, eowic, uncit, incit, which are used only in the very
oldest texts as accusatives distinct from the datives
me, Þe, us, eow, unc, inc, and which are soon ousted by
the latter forms. By parallel developments occurring
somewhat later, the old dative forms hire (hir, her),
him and hwam (whom) are made to fill the offices
held hitherto by the old accusatives heo, hine and
hwone. In some of the southern counties hine is,
however, preserved up till our times in the form of
[ɐn], see Ellis, Early Engl. Pronunciation, v., p. 43;
in the literary transcription of these dialects this is
written 'un, e.g., in Fielding's Tom Jones (Squire
182Western, etc.), and in Thackeray's Pendennis (i., 62,
“Show Mr. Pendennis up to 'un”). 1139 In the plural, also,
the dative form has expelled the old acc.; hem (O. F.
him, heom; preserved in familiar and vulgar speech:
“I know 'em”) and the later them are originally
datives; 2140 the neuter singular, on the other hand, has
preserved the old accusative forms hit (it), Þæt (that),
hwæt (what), at the expense of the old datives.

The reason of this constant preferring of the dative
forms in the person-indicating pronouns is no doubt
the fact that these pronouns are used as indirect objects
more often than either nouns or adjectives; 3141 at any
rate, it is a phenomenon very frequently found in
various languages; compare Danish ham, hende, dem,
hvem, originally datives, now also accusatives and
partly even nominatives (while it is true that in mig
and dig the acc. has outlived the dative); North
183German wem for wen 1142 French lui as an absolute
pronoun (while the acc. has carried the day in elle,
eux, elles; moi and toi may be either); Italian lui,
lei, loro, 2143 etc.

Simplification has gone further in the case of the
pronouns of the second person than in that of the
others; in fact, if we were to believe the ordinary
grammars, the substitution of you for ye is the only
point in which a deviation from the old system has
taken place. But ordinary grammars are not always
trustworthy; in laying down their rules they are too
184apt to forget that the English language is one thing,
common-sense or logic another thing, and Latin
grammar a third, and that these three things have
really in many cases very little to do with one another.
Schoolmasters generally have an astonishing
talent for not observing real linguistic facts, and an
equally astonishing inclination to stamp everything
as faulty that does not agree with their narrow rules;
and the precepts inculcated in the school-room have
no doubt had some influence in checking natural
tendencies, though the following pages will suffice to
show that the best authors have in many points deviated
more from the rules laid down in grammars
than is generally supposed.

153. (105) Many of the phenomena I shall treat of
have, as a matter of course, been noticed and partly
explained by modern grammarians of the historical
school; I shall specially mention Koch, Hist. Gramm.,
ii. (especially p. 244 1144); Mätzner, Engl. Gramm., ii.
passim; Abbott, A Shakespearian Grammar, § 205
ff.; A. Schmidt, Shakespeare-Lexikon; Storm,
Englische Philologie, 1881, p. 207 ff.; Gummere,
The English Dative-Nom. of the Person. Pron., in
American Journ. of Philol., iv.; W. Franz, Die
185dialektspr. bei Dickens, Engl. St., xii., 223 f., and Zur
syntax des älteren Neuenglisch, ibid., xvii., 212 ff.;
Kellner, in the Introduction to Caxton's Blanchardyn
(EETS. Extra Series 58).

On the whole these authors content themselves
with a purely lexical treatment of the matter, giving
for instance all the examples of I for me and vice
versâ under one head, and only occasionally offering
an explanation of some phenomena; the fullest and
most satisfactory explanations are found in Storm's
excellent work. In the following sections I shall
attempt a systematic arrangement according to the
psychological or phonetic principles underlying the
phenomena and causing speakers or writers to use
another case than that exacted by the rules of ordinary
grammar. I shall first take those classes of case-shiftings
which are of a more general character and
may occur more or less frequently in all languages of
our type, giving last those which belong more specially
to English or to one particular period of English.

It must be specially mentioned that in many of the
sentences quoted two or even more causes of shifting
have operated concurrently.

I. Relative attraction.

II. Blendings.

Note also Sh., Cor., i, 1, 236, “And were I anything
but what I am, I would wish me only he,” where he is
the only natural form, as him would only obscure the
meaning of the phrase. 2147 In R. Haggard, Cleopatra, ii.,
189121, “rather than I would see her thy wedded wife and
thou her loving lord,” we have an approach to the
phenomenon mentioned below, § 164.

Storm (E. Philol., 211) has some modern quotations
(from Dickens, who writes also: “Leave Nell
and I to toil and work”), and quotes the Norwegian
[and Danish] colloquial lad vi det for lad os det. In
the corresponding Dutch construction both the nom.
and acc. are allowed: “laat mij nu toonen” as well
as “laat ik nu toonen” (let me now show); similarly
“laat hem [hij] nu toonen, laat ons [laten wij] nu
toonen, lat hem [laten ze] nu toonen”. 1150 In a passage
from Guy of Warwick, 3531, “Let hym fynde a sarasyn
And y to fynde a knyght of myn,” we have a transition
case between this phenomenon and that dealt
with in § 164.

A similar confusion after the verb make is found
in Sh., Temp., iv., 1, 217, “mischeefe which may make
this island Thine owne for ever, and I thy Caliban
for aye thy foote-licker”; here Caliban forgets the
first part of his sentences and goes on as if the beginning
had been “this island shall become”. So also
in Rich. II., iv., 1, 216, “[God] make me, that nothing
191haue, with nothing grieu'd, And thou with all pleas'd,
that hast all atchieu'd”.

In these cases the nominative is used in spite of
grammatical rules requiring the acc. because the
word is thought of as the subject; this is even,
though rarely, the case after a preposition; in Roister
Doister, p, 72, I find: “Nay as for they, shall euery
mother's childe die;” and a phrase in a letter that
is read aloud twice in the same play runs the first
time “as for all them that woulde do you wrong” (p.
51), but the second time “as for all they” (p. 57).
In § 170 ff. we shall see some more instances of the
nominative, as the case proper to the subject, getting
the better of the acc. required by earlier grammatical
rules. 1151

157. (107) Other contaminations leading to confusions
of two cases are found here and there. In
Sh., Temp., ii., 1, 28, we read: “Which, of he, or
Adrian…First begins to crow?” This is a
blending of “Which, he or A.,” and “Which of [the
two] him and A.,” or else of may be a printer's error
for or, as conjectured by Collier. In Sir Andrew's
interruption, Tw. N., ii., 5, 87, “[you waste the treasure
of your time with a foolish knight. — ] That's mee I
warrant you,” me is due to the use of the accus.
in the preceding sentence (= with me); immediately
afterwards he says: “I knew ‘twas I”; in Malvolio's
speech, “If this should be thee,” thee is similarly the
192object of the preceding I loue. Comp, Thack,, Pend.,
iii., 87, “If ever I saw a man in love, that man is
him”. The opposite result of the contamination is
found in Sh., Troil., ii., 3, 102, “Achillis hath inveigled
his foole from him. — Who, Thersites? — He” (= who is
it? it is he); parallel cases occur at every moment in
colloquial language.

159. (113) Save (sauf) presents similar phenomena
of confusion, although it is comparatively seldom
found as a preposition, as in Matth. Arnold, Poems, i.,
159, “For of the race of Gods is no one there, save
me alone”; and in Tennyson, p. 319, “Who should
be king save him who makes us free?” 2155 In Chaucer
sauf (save) is very common with nom. (B., 474, 627;
G., 1355; I., 25; L. G. W., 1633; Morris, ii., 221, 493;
342, 801),so also in Shakespeare (Tw. N., iii., 1, 172;
Cæs., iii., 2, 66, etc.), and in modern poets (e.g., Byron,
iv., 332, “Who shall weep above your universal grave,
save I?”). Where the word is not meant as the
subject, the accusative is used (e.g., Chaucer, B., 4491,
195“Save yow I herde neuere man so singe”; where,
however, one MS. (H) has ye. An example of an abnormal
use of the nom. is Shak., Sonn. 109, 14, “For
nothing in this wide universe I call, save thou, my
rose”.

The accusative is always used in than whom (found
in Shakespeare, Love's L., iii,, 180, in Milton, etc.);
Alford is right in observing that than who is here
excluded because the expression does not admit of
an elliptical construction. I only once remember
200having found than who, namely in the sentence, “Mr.
Geo. Withers, than who no one has written more
sensibly on this subject,” and then it occurs in the
book on The King's English (p. 338) by Mr. Washington
Moon, who is constantly regulating his own and
others' language by what in his view ought to be,
rather than what really is the usage of the English
nation.

III. Anacoluthia.

What is the reason of the accusative in Sh., Cymb.,
v., 4, 70, “we came, our parents and vs twaine”?

In some of these sentences the construction might
be called a kind of apposition; in others we have
208something closely resembling the absolute participle,
of which more will be said below, § 183; the use of
an “unconnected subject” may have favoured the
substitution of the modern “absolute nominative” for
the old “absolute dative”.

But this use of a nominative with the infinitive does
not occur often enough to be a permanent feature of
the English language.209

IV. Influence from the Noons.

167. (116) The absolute absence of any formal
distinction between the nominative and the objective
cases in the nouns and adjectives, as well as in the
neuter pronouns it, that, and what, must of course
do a great deal towards weakening the sense of case
distinctions in general.

An English friend of mine once told me about a
clergyman who in one of his sermons spoke constantly
of your immortal I, but was sadly misunderstood
by the congregation, who did not see why the
eye should be more immortal than any other part
of the body. It is perhaps to avoid such misinterpretations
that the Latin form is sometimes used,
as in Thack., Pend., iii., 363, “every man here has his
secret ego likely”.

Compare the use of me in other exclamations: O(h)
me! Woe me! Ah me! Ay me! (Milt., P. L., iv.,
21286, etc.), Aye me detested! (Sh., Tw. N., v., 142), Alas
me! (Keats, Eve of St. Agnes, xii.), Me miserable!
(Milt., P. L., iv., 73), etc. The use of me in dear me!
gracious me! and other apologies for oaths is probably
due to the analogy of the corresponding use of the
pronoun as an object after a verb, as in bless me! etc.
So perhaps also in Shak., 1 H. IV., ii., 3, 97,“Gods
me, my horse”.

V. Position.

170. (119) Word-order is to no small extent
instrumental in bringing about shiftings of the
original relation between two cases. In Old English
prose the subject is already placed before the verb
in nearly every sentence; the exceptions are almost
the same as in Modern German or Danish; thus
inversion is the rule after adverbs such as Þa (while,
curiously enough, the subject precedes the verb
where the clause is introduced by hwæt Þa or efne
Þa). By-and-by these exceptions disappear or are
reduced to a minimum, so that in Modern Eiiglish
the order, subject, verb, object, is practically invariable. 1167
Cooper defines the difference between the
nom. and the acc. in the pronouns in the following
manner: 2168 “I, thou, he, she, we, ye, they, verbis anteponuntur,
me, thee, him, her, us, you, them, postponuntur
verbis & præpositionibus”. However naïve the
grammarian may find this definition, it contains a
213good deal of truth; this is the perception of the distinction
between the two forms which in the popular
instinct often overrides the older perception according
to which the use of I and me was independent of
position.

172. (120) As regards Shakespeare's use of who
in the objective case, it must suffice to refer to Al.
Schmidt's Lexicon; under the interrogative pronoun
he gives fifteen quotations for the use in question,
and then adds an etc., which, to any one familiar
with the incomparable accuracy and completeness
of Schmidt's work, is certain proof that examples
abound; finally he names nineteen places where
the old editions do not agree. Under the relative
pronoun he adduces twelve quotations for who as an
acc., followed again by an etc., and by eleven references
to passages in which the oldest editions give
different readings. It is well worth noting that
where such variations of reading are found it is
nearly always the earliest edition that has who and
the later editions that find fault with this and replace
it by whom; most modern editors and reprinters
add the -m everywhere in accordance with the rules
of grammars, showing thereby that they hold in
215greater awe the schoolmasters of their own childhood
than the poet of all the ages. 1170

Shakespeare also uses whoever as an accusative;
whomever does not occur in his works; he also
sometimes uses who after a preposition (see Abbott, §
274, and add to his examples, R. III., i., 3, 54), but
this seems now obsolete, because the natural word-order
is to place the preposition at the end of the
sentence, as Shakespeare does himself in numerous
passages; for instance, As, iii., 2, 327, “He tell you
who Time ambles withall, who Time trots withall,
who Time gallops withal, and who he stands stil
withal”. It seems, then, as if the last refuge of the
form whom is the combination than whom, where it
had originally nothing to do; but as this combination
belongs more to literary than to everyday
language, who is now to be considered almost as a
common case; compare what Sweet writes to Storm:
“I think many educated people never use whom at
all; always who”.

173. (121) A great many verbs which in Old
English were impersonal have become personal in
Modern English., and one of the causes which most
contributed to this change was certainly word-order.
The dative, indicating the person concerned, was
216generally placed immediately before the impersonal
verb; the reason of this position was undoubtedly
the greater interest felt for the person, which caused
the word indicating him to take a prominent place
in the sentence as well as in the consciousness of the
speaker. And so this “psychological subject,” as it
has been termed, eventually became the grammatical
subject as well. But other circumstances favoured
the same tendency. Some verbs in O. E. admitted
of both a personal and impersonal construction, e.g.,
recan, “to care”; compare from the thirteenth century
the Ancr. Riwle, p. 104, where one MS. has “ʒ if heo
beođ feor, me ne recched,” and another “Þach ha
beoh feor, naut I ne recche”. In one case, two originally
distinct verbs grew to be identical in pronunciation
by a purely phonetic development, namely O. E.
Þiyncan, “seem” (German dünken), impersonal, and
Þencan, “think” (Germ, denken), personal. In the
former the vowel y by the usual process lost its liprounding
and so became i; in the latter e was raised
to i before the back nasal consonant, as in O. E. streng,
Mod. string, O. E. hlence, mod. link, O. E, Englaland,
Mod. England, pronounced with [i]; compare also the
history of the words mingle, wing, cringe, singe, etc.

The number of verbs that have passed from the impersonal
to the personal construction is too great for
me here to name them all; I shall refer to the lists
given by Koch, Gram., ii., § 109; Mätzner, ii., p.
198 ff.; Einenkel, Streifzüge, p. 114 ff.; and Kellner,
Blanchardyn, p. xlvii. ff. But I shall supplement
217the remarks of these scholars by attempting to
analyse the psychological agencies at work in the transition;
I shall for this purpose print those examples
from my own collection which seem to be the most
illustrative, confining myself generally to only a few
of the most usual verbs coming under this head.

In Ch., M. P., 5, 114, “[thou] dauntest whom thee
lest,” some of the manuscripts read thou, probably in
order to avoid the two accusatives after each other.

The last quotation is of especial interest as showing
a sort of blending of no less than three constructions:
the impersonal construction with us lyketh as a third
personal sg. with no object, the old personal construction,
where like means “to please,” us lyken ye, 1174
and finally the modern personal use, we lyken yow;
the continuation “and ever han doon” (= “and we
have always liked you”) shows that the last construction
was at least half present to Chaucer's mind.221

Other blendings of a similar nature are found with
think; me thinks and I think are confused in me
thinke, found, for instance, in a sermon of Latimer's
(Skeat's Specimens, xxi., 176); 1175 thinks thee? and
thinkst thou? give thinkst thee? in Shakespeare's
Hamlet, v., 2, 63 (folio; the old quartos have thinke
thee; some modern editors write thinks't thee, as if contracted
for thinks it thee; but this is hardly correct,
as this verb is very seldom used with it, at least when
a personal pronoun is added).

These we may consider either the oldest examples
of who as an accusative (centuries before any hitherto
pointed out), or else the oldest examples of O. E.
Þyncan and lystan used personally. 2176 I suppose, however,
222that the correct way of viewing these sentences
is to say that the two tendencies, neither of which
was strong enough to operate by itself, here combined
to bring about a visible result.

In some cases the personal construction has not
become universal, as in the case of ail (O. E. eglan).
Though Dr. Murray is able to show the personal use
of the word in a quotation as early as 1425, and
though Shakespeare never uses it impersonally (comp.
also Marlowe, Jew, 1193, “What ayl'st thou”), the old
construction still survives. The reason is undoubtedly
the fact that the verb is so very often used in the
223common formula: What ails him? (her, etc.), where
the personal pronoun is placed after the verb; see, e.g.,
Sirith, 337; Chauc., B., 1170, 1975, 4080; H., 16; M.
P., 3, 449, etc., etc,; Tennyson, p. 132: “What ail'd
her then?” G. Eliot, Mill, i., 80, “there's nothing ails
her”.

181. (135) I must here also mention the peculiarity
of the English language by which not only what
would be the direct object of the active verb but
other parts of the sentence may be made the subject of
a passive verb. As I have not collected sufficient
materials to give an exhaustive treatment of this interesting
subject, I shall confine myself to a few remarks.
There can be little doubt that nouns were employed
227in this way as “free subjects” of passive verbs at an
earlier time than pronouns in which the nom. and the
acc. had distinct forms. I shall arrange my examples,
under four heads. 1183

Compare the somewhat analogous phenomenon in
Ancr. R., 6, “sum is old & atelich & is đe leasse dred
of” (is dred of is a sort of passive of habben dred of);
here, however, we have rather a continuation of the
old use of of as an adverb = “thereof”.

182. This extension of the passive construction is
no doubt in the first place due to the effacement of
the formal distinction between the dative and the
accusative; but a second reason seems to be the same
fact which we met with before in the case of verbs
originally impersonal: the greater interest felt for the
231person makes the speaker place the noun or pronoun
by which the person is indicated before the direct
object, as in the sentence: “He gave the girl a gold
watch”. This makes it natural that in the passive voice
the dative should be placed at the very beginning of
the sentence: “The girl was given a gold watch”.
But this position immediately before the verb is
generally reserved for the subject; so the girl, though
originally a dative, comes to be looked upon as a
nominative, and instead of “her was given a gold
watch,” we say, “she was given a gold watch”. On
the other hand, the nature of these constructions
reacts on the feeling for case-distinctions in general;
for when “I was taught grammar at school” comes
to mean the same thing as “me was taught grammar,”
or “she was told” as “her was told,” etc., there is one
inducement the more to use the two cases indiscriminately
in other sentences as well, or at least to
distinguish them in a different way from that which
prevailed in the old language.

183. No doubt the position before the verb has
also been instrumental in changing the old absolute
dative (as seen, for instance, in Chron., 797, “Gode
fultomiendum, God helping”) into the modern nominative.

It is true that these sentences are modern and
penned long after the absolute nom. had been settled;
but although I have no old quotations ready to hand,
similar expressions may and must have occurred at
any time.

184. (129) Having dealt (in §§ 170-183) with the
substitution of the nominative for an original accusative
234or dative before the verb, we shall now proceed
to the corresponding tendency to use an objective case
after the verb where a nominative would be used in
the old language. This is, of course, due to the
preponderance of the instances in which the word
immediately following the verb is its object. 1189 The
most important outcome of this tendency is the use of
me after it is. I have already had occasion to mention
a few connexions in which the accusative will
naturally come to be used after it is (see §§ 154 and
157); to these might be added accusatives with the
infinitive, as in Greene, Friar. Bacon, 10, 57, “Let it be
me”. But even where there is no inducement of that
kind to use me, this form will occur after it is by the
same linguistic process that has led in Danish to the
exclusive use of det er mig, where some centuries ago
the regular expression would have been det er jeg, and
which is seen also in the French c'est, used in Old
French with the oblique form of nouns and then also
of pronouns, c'est moi, etc. 2190

185. (129) The nom. accordingly seems to have
been the natural idiom, just as det er jeg was in
236Danish a few centuries ago, and as det är jag is still
in Sweden; but now it is otherwise, and it is me must
be reckoned good English., just as det er mig is good
Danish. In Shakespeare (besides the passages accounted
for above) we find the accusative used in
three passages, and it is well worth noting that two
of them are pronounced by vulgar people, viz., Two.
Gent., ii., 3, 25, “the dogge is me” (the clown
Launce), and Lear, i., 4, 204, “I would not be thee”
(the fool; comp. Pericl., ii., 1, 68, “here's them in our
country of Greece gets more,” spoken by the fisherman);
the third time it is the angry Timon who
says: “[I am proud] that I am not thee” (iv., 3, 277).
The stamp of vulgarity would have disappeared completely
by now from the expression had it not been
for grammar schools and school grammars; even to
the most refined speakers it's me is certainly more
natural than it's I. 1192 And Shelley has consecrated the
construction as serviceable in the highest poetic style
by writing in his Ode to the West Wind: “Be thou,
spirit fierce, my spirit! Be thou me, impetuous one!”

Latham, Ellis, Sweet and Alford defend it is me as
the only natural expression; the reason of their not
extending this recognition of the objective case
equally to the other persons will be found below
237(§ 194); yet in Thackeray's Vanity Fair, p. 163, a
young lady says It's her; and in Cambridge Trifles,
p. 96, an undergraduate says It couldn't be them — to
mention only two examples.

186. (130) Not only the predicate but also the
subject itself is liable to be put in the accusative after
the verb. Shall's (= shall us) for shall we is found
six times in Shakespeare. As four times it means
exactly or nearly the same thing as let us (Cor., iv.,
6,148, “Shal's to the Capitoll”; Wint., i., 2,178; Cymb.,
v., 5, 228; Pericl., iv., 5, 7), it is probable that this
idiom is originally due to a blending of let us and
shall we (compare the corresponding use of a nom.
after let, § 156). But it has been extended to other
cases as well: Tim., iv., 3, 408, “How shal's get it?” |
Cymb., iv., 2, 233, “Where shall's lay him?” Towards
the end of the last century shall us was common in
vulgar speech according to Sam. Pegge, 1193 who adds:238

“The Londoner also will say — “Can us,” “May us,”
and “Have us”. Storm quotes (p. 209) from
Dickens some instances of vulgar shall us, can't us, do
us, hadn't us; is this phenomenon still living in the
mouth of uneducated people? I do not call to mind
a single instance from the Cockney literature of the
last ten years or so.

Here the phrase be damned, or its substitute be
hanged, has become an exclamation, and to you is
added as if “I say” was understood; compare also
Hail to thee (Middle Engl. heil be Þow); farewell
to you; welcome to you; good-bye to you. 1196

188. (132) The position after the verb has probably
had no small share in rendering the use of thee (and
you) so frequent after an imperative, especially in the
240first Modern English period; the usage is still seen in
the poetical phrase “Fare thee well”. Here we have,
however, a concurrent influence in the use of a reflexive
pronoun (without the addition of self) which
was extremely common in all the early periods of the
language, and which did not perceptibly alter the
meaning of the verb to which it was added. 1197 This
reflexive pronoun was sometimes originally added in the
accusative case,e.g., afterrestan (see Voges,p. 333),but
generally in the dative; this distinction, however, had
obviously no significance for any but the very earliest
stages of the language. As now it made no difference
whatever whether the speaker said I fear or I fear me
(compare, for instance, Marlowe, Jew, 876, with 1110),
the imperative would be indifferently fear or fear
thee (fear yow); 2198 but it was equally possible with the
same meaning to say fear thou (fear ye), with the
usual addition of the nominative of the pronoun to
indicate the subject. Examples from Malory of the
latter combination: 73, “go ye” | 74, “telle thow” |
75, “doubte ye not,” etc. etc. 3199 In other words: after
an imperative a nominative and an accusative would
241very often be used indiscriminately. Thus, Care ye not
(Malory, 72) means exactly the same thing as care
not yow (ibid., 135); stay thou (Sh., Cæs., v., 5,
44) = stay thee (3 H. VI, III., 2, 58); get ye
gon (Marlowe, Jew, 1226) = get you gone (common,
Sh.); stand thou forth (Sh., All, v., 3, 35) = stand thee
by (Ado, iv., 1, 24); turn ye unto him (Isaiah, xxxvi.,
6; Ezek., xxxiii., 11) = turn you, at my reproof (Prov.,
i., 23); turn you to the stronghold, ye prisoners of
hope (Zech., ix., 12); turn thee unto me (Ps., xxv.,
16) = turn thou unto me (ibid., lxix., 16 1200); fare ye
well (Sh., Merch., i., 1, 58 and 103) = fare you well
(ibid., ii., 7, 73); seldom as in Tim., i, 1, 164, Well
fare you, fare thou well (Temp., v., 318) = fartheewell
(Tw. N., iii., 4, 183); far-thee-well (ibid., iii., 4, 236);
far thee well (ibid., iv., 2, 61); sit thou by my bedde
(Sh., 2 H. IV., 5, 182) = sit thee downe vpon this
flowry bed (Mids. N., iv., 1,1; also with the transitive
verb set thee down, Love's L., iv., 3, 4, in some editions
emended into sit!).

189. (132) It will now be easily understood that
thee (or you) would be frequently added to imperatives
where the thought of a reflexive pronoun would
not be very appropriate; in hear thee, hark thee, look
thee and similar cases, Voges finds a reflexive dative,
242whereas Al. Schmidt quotes them under the heading
“thee for thou”; it is rather difficult to draw a line
here. When Troilus says (act iv., 5, 115): “Hector,
thou sleep'st, awake thee” no less than three grammatical
explanations are applicable: awake may be
intransitive, and thee the subject (Al. Schmidt), awake
is intransitive, but thee is a reflexive dative (Voges,
l. c, p. 372), and finally, awake may be a transitive
verb having thee as its object (comp. Murray's Dict.);
but whichever way the grammatical construction is
explained, the meaning remains the same. 1201

It is evident that all this must have contributed
very much to impair the feeling of the case-distinction,
and it should be remarked that we have here a
cause of confusion that is peculiar to the pronouns of
the second person . 2202243

It would be a waste of paper and ink to give
examples from more recent times, as they abound
everywhere; I shall therefore only state the fact that
in the modern use of himself and themselves (and
244herself?) we have a dative used as a nominative (or
rather as a common case), and that this was formerly
the case with me self and us self (or us selue, seluen)
as well, which have now been ousted by myself and
ourselves. 1204

It will be seen that, if my explanation is the correct
one, we have here an influence of word-position of
quite a different order from that pointed out in the
rest of this section. Dr. Sweet, 1206 while accepting this
explanation as far as the Elizabethan idiom is concerned,
thinks that when between you and I or he saw
John and I is said now-a-days, it is due to the
grammatical reaction against the vulgar use of me for I.

VI. Phonetic influences.

193. (137) I now come to the last but by no
means the least important of the agencies that have
brought about changes in the original relations between
the cases of the pronouns. I mean the influence
of sound upon sense.

If you glance at the list of pronominal forms
printed in § 152 you will see that seven of them
rhyme together, the nominatives we, ye, he, she, and
the accusatives me, thee. After the old case-rules
had been shaken in different ways, instinctive feeling
247seized upon this similarity, and likeness in form has
partly led to likeness in function.

A century later the same holds good with the verb
lust in Roister Doister: ye (pp. 12 and 51), me (12), he
(42), she (87); there are two exceptions: hym (43), I (44).

The phonetic similarity is used to mark the contrast
in Sh., Macb., iii., 4, 14 (1035), “'Tis better thee
without then he within”; see W. A. Wright's note:
“It [Banquo's blood] is better outside thee than
inside him. In spite of the defective grammar, this
must be the meaning.”

The common answer which was formerly always
Not I! (thus in Shakespeare, see Al. Schmidt, Sh. Lex.,
p. 565 a, bottom of the page) is now often heard as
Not me! while the corresponding form in the third
person does not seem to be Not him! even in vulgar
speech, but always Not he! At least, I find in the
Cockney Stories, Thenks awf'lly, London, 1890, p. 82,
“Not'e!” 2209251

P. Greenwood, Grammatica Anglicana, 1210 mentions
among errors committed by plerosque haud mediocri
eruditione praeditos: “He spake it to shee whose
fountaines is dried up,” and he adds: “Non mirum si
vulgus barbare omnino loquatur, cum qui docti, et
sunt, et habentur, tam inscite, et impure scribunt”.

197. (140) Phonetic influences may have been at
252work in various other ways. If the vowel of the
nominative Þu was weakened when the word was
unstressed the result would be Þe [đə], exactly like a
weakened form of the accusative Þe. This is, I take
it, the explanation of the nominative Þe found so
often in the Ayenbite of Inwit (A.D. 1340) in such
combinations as Þe wylt, Þe miʒt, Þe ssoldest. As u
is undoubtedly weakened into e in Huannes comste,
“whence comest thou” (Ayenb., 268), as te stands
certainly for Þu in Robert of Gloucester, 10792 seiste,
3150 woste, 4917 ʒifst' us, 1211 and as similarly to is
weakened into te in the Ayenbite as well as in (parts at
least of) the Ancren Riwle, this phonetic explanation
seems to me, as it did to Mätzner, 2212 more probable than
the two other explanations given by Gummere 3213 and
Morris. 4214

As, however, this use of Þe for Þu is only found in
a few texts (also in Sir Beues of Hamtoun, see Engl.
Studien, xix., 264), we cannot ascribe to it any great
influence on the later development.

198. (170) Similarly a you pronounced with weak
253sentence-stress will be reduced to ye or even to the
short vowel i, written y. This is especially the case
in stock phrases like thank you (thanky), God be with
you (Good-bye, 1215 the oo-vowel is probably introduced
from the other forms of salutation: good-morrow,
good-night, etc., the naming of God being thus
avoided; in Shakespeare it is also written God buy
you), God give you good even (in Shakespeare Godgigoden,
Godigoden, God dig you den). Harky (hark'ee)
and look'ee may contain ye, weakened for you (§ 188),
or the nominative ye. I am inclined to think that
this phonetic weakening of you is the cause of the
unstressed ye after verbs, which is found so very
frequently from the beginning of the sixteenth
century, although it is impossible in each single
instance to distinguish the ye which originates in this
way from ye's called forth by the other circumstances
dealt with in this chapter.

200. (142) It will be noticed that these phonetic
tendencies cannot possibly have had any influence
on the case-relations of most pronouns; weaken
the vowel of me as you like or drop it altogether,
the remaining m' is not brought one bit the nearer
to the nom. I. But in the pronouns of the second
person there is this peculiarity, that the cases are
distinguished by the vowel only; if the vowel is
left out it becomes impossible to tell whether the
nominative or the accusative is meant — one more
reason for the old distinction to become forgotten.

In Chaucer thee is elided, see Cant. T., B., 1660, in
thalighte. In Greene's Friar Bacon, 12, 78, “For
ere thou hast fitted all things for her state,” we must
certainly read th'hast (see also the same play, 13, 37).
In countless passages, where modern editions of
Shakespeare read you're the old folio has y'are, which
must no doubt be interpreted ye are. But when we
find th'art (for instance, Cor., iv., 5, 17 and 100, mod.
256edd. thou'rt), is this to be explained as thou art (thu
art) or as thee art? Similarly th'hast (mod. edd.
thou'st), th'hadst (mod. edd. thou hadst); in Macb.,
iv., 1, 62 (1312), “Say if th'hadst rather heare it from
our mouthes,” it is specially difficult to decide in
favour of one or the other form on account of the
peculiar constructions of had rather (see above, §
180).

201. (143) There is one more thing to be noticed.
Where the pronouns are combined with the verbal
forms commencing with w, those forms are preferred
that contain rounded vowels. The past subjunctive
of y'are is in Shakespeare you're (Cymb., iii., 2, 76,
“Madam, you're best consider”); the second person,
corresponding to I'le for I will, is not ye'le, 1219 but you'le
(Marlowe, Jew, 708), or more frequently you'll. Now
I take it to be highly probable that these forms were
heard in the spoken language at a much earlier period
than they are recorded in literature, that is, at a time
when you was not yet used as a nom., and that they
are contracted not from you were, you will, but from
ye were, ye will (? ye wol), the vowel u being thus a
representative of the w of the verb. 2220 If this is so,
257we have here yet another reason for the confusion of
ye and you, as the contracted forms you'll and you're
would be felt instinctively as compounds of you and
will or were. For thou wert we find thou'rt; 1221 for
thou wilt similarly thou'lt (e.g.,Marl., Jew, 1144 ; often
in Shakespeare, who also, though rarely, writes thou't).

It is in accord with this that in Roister Doister
(printed 1566) you rhymes with thou (pp. 31 and 32),
with now (pp. 15, 43, 48, 53, 60, 63 and 70), and with
inowe (p. 18).

Now the [au] form of you is extinct; the current
pronunciation [ju·] or [juw] must be due to a natural
lengthening of the originally unstressed form [ju],
when it was used with stress. 2224 The existence of
the form [ju] at the time of Shakespeare may be
concluded from the pun in Love's Labour, v., 1, 60.

203. (145) In thou, on the other hand, it is the
fuller form with [au] that is now heard solely: this
260is quite natural because the word is now never found
in colloquial language, so that only the emphatic
pronunciation of solemn or ceremonial speech has
survived. But when the two pronouns thou and you
were used pari passu in ordinary conversation, their
sounds were alike; you and thou formed correct
rhymes, exactly as thee and ye did. 1225 But to the
formal likeness corresponded a functional unlikeness:
you is not the same case as thou, but as thee, and ye
has the same case-function as thou. Are not these
cross-associations between sound and sense likely to
have exerted some influence on the mutual relations
of the forms?

When matters stand thus, and when the feeling for
case-distinctions is shaken in a multiplicity of ways,
must not countless confusions and blendings take
place in ordinary careless conversation? The speaker
begins to pronounce a ye, but, half-way through, he
falls into the more familiar manner of address, and
thus he brings about the compromise you, which is
accordingly in many instances to be considered a
sort of cross between ye and thou; you = y(e) + (th)ou.
Such blendings of two synonyms, where the resulting
word consists of the beginning of one and the end of
262the other word, are by no means rare in language;
Shakespeare has rebuse = rebu(ke) + (a)buse (Shrew,
i., 2, 7), and Tennyson: be dang'd = da(mned) +
(h)anged (Works, p. 618); but the nearest parallel to
our case, that I know of, is the Scottish pronoun thon
= th(at) + (y)on (see Murray, Dial. South. Counties,
p. 186), where in two synonymous pronouns the very
same two sounds are interchanged as in the case
before us. 1226 In you there are, as we have seen, many
more inducements at work, 2227 which all of them concur
in causing the cross to be rapidly recognised and
accepted by everybody.

You is practically non-existent in that work; I
find it only on p. 252 (Works, ed. Tauchnitz, vol. iv.),
“And you, ye new And scarce-born mortals,” and p.
224, where it is used in the indefinite signification of
the French on.

The old ye has yet another refuge, namely, in
grammars, where it renders the separate plural forms
of other languages, Latin vos, German Ihr, etc. If
this small domain is excepted, the English seem
never to feel any inconvenience from their language
having the same form for the singular and the plural
in this pronoun; if a separate form is now and then
required for distinction's sake the want is easily
remedied — after the Chinese fashion, see § 66 — by the
addition of some noun: you people, you gentlemen, you
girls, you chaps, you fellows, etc.

209. (150) Here I end my survey of the various
case-shifting agencies and of their operations. As
already mentioned, it extremely often happens that
in the same sentence two or more causes co-operate
to make the speaker use a different case from what
we should expect, or rather from what the grammar
of an earlier stage of the language would require.
The more frequently such concurrences occur, the
greater the vitality of the new manner of using the
270case in question. We saw in § 178 that two separate
tendencies, whose effects do not appear properly till
some two hundred years later, were powerful enough
when co-operating to bring about a visible (that is, an
audible) result. And on reading again the quotations
used to illustrate the first sections of this chapter you
will find that the forms in e supply a comparatively
greater contingent than the other forms, showing thus
the concurrence of the associations treated in § 193.
The facts which have been brought to light will, moreover,
have made it clear that with the pronouns of the
second person more shifting agencies were at work
than with the rest (§§ 188,189,193-204), the result being
that the original case-relations have been completely
revolutionised in these pronouns. In the case of
I and me, too, some special causes of changes in
the case-relations have been pointed out (§§ 192,
193); but they proved to be much less powerful
than those seen in the second person, and operated
besides in opposite directions, so that the same simplicity
as that found in you was here impossible.
Finally, we have seen that the invariable position of
who before the verb has caused it to become a common
case, whom being relegated to a very limited
province which it did not properly belong to.

210. (151) There is one factor I have not taken
into account, though it is nearly everywhere given as
explaining the majority of case-shiftings in a great
many languages, — I mean the tendency to let the objective
case prevail over the subjective case. My reason
271is simply that this tendency cannot be considered
as a cause of case-shiftings; it does not show us how
these are called forth in the mind of the speaker; it
indicates the direction of change and the final result,
but not its why and wherefore. Nay, in English., at
least, it does not even exhaustively indicate the
direction of change, as will be gathered from some
points in the above exposition: the nominative
carries the day in the absolute construction, in who
and in the (vulgar) combination between you and I;
note also the change of the case used with the old
impersonal verbs. Still, it must be granted that the
nominative generally has the worst of it; this is a
consequence of the majority of the case-shifting
agencies operating in favour of the accusative; thus,
while it is only the position immediately before the
verb that supports the nominative, the accusative is
always the most natural case in any other position;
see, for instance, the treatment of than as a preposition.

211. (152) This will afford an explanation of the
fact that wherever we see the development of special
emphatic or “absolute” pronouns as opposed to conjoint
pronouns (used in direct conjunction with the
verb), the former will as a rule be taken from the
originally oblique cases, while the nominative is restricted
to some sort of unstressed affix to the verb.

Such a development is not carried through in
Standard English., which has formed the principal
subject of our investigations. But if we turn to the
272dialects now existing in England, we shall find this
distinction of absolute and conjoint pronouns made
very frequently. A thorough examination of the
case-relations of living dialects would present very
great interest, although it would rather show the
results of similar developments to those found in the
literary language — with many deviations, it is true
— than throw any fresh light on the agencies at work
or the causes of the changes effected. These are
best investigated in the literary language, because
we there have materials from so many succeeding
centuries that we are often enabled to discover the
first germs of what living dialects would only present
to us as a development brought to a definite (or preliminary)
conclusion. For this reason, as well as for
the obvious one that the dialects of our own days
have not been so fully and reliably treated, especially
with regard to syntax, as to render a satisfactory
exposition possible, I shall content myself with a
few remarks only on the pronouns in the dialects.

212. (153) In the dialect of the southern counties
of Scotland, so admirably treated by Dr. Murray
an emphatic form, originating in the old accusative,
is used very much as the corresponding forms in
French, e.g., Thaim ‘at hæs, aye geates mair; mey, aa
canna gang (moi je ne peux pas aller); yuw an' mey ‘ll
gang ower the feild. “He gave it to you” = hey
gæ ye'd; “he gave it to YOU” = hey gæ yuw'd;“he
gave IT to you” = hey gæ ye hyt; “he gave IT to
YOU” = hey gæ yuw hyt.273

For the dialect of West Somerset, Elworthy gives
no less than six series of forms, viz., for the nominative:
(1) “full” forms, used when the nominative
stands before its verb with emphasis; among these
forms we notice the old objective forms dhee and
yùe; perhaps also uur, “her,” if Dr. Murray is not
right in considering it as the old nom. heo; (2) unemphatic
forms used before the verb, generally the
same forms as in the first series, only weakened [ee
– ye?]; (3) interrogative enclitic forms, among
which [ees] us is noticeable as being used exactly
as the Shakespearian us in shall's, see above, § 186;
in the third person pl. um = O. E. heom is used in
the same manner; and (4) unconnected forms, all of
them old accusatives, except he (ee), compare § 196,
and dhai. Then for the objective case we have two
series of forms: (1) the unemphatic, of which we note
the second person pl. ee = ye and the third person
sg. masc., un, n = O. E. hine, see § 151; and (2) emphatic
or prepositional, among these aa˙y concurrently
with mee, and wee with uus (§ 196), and on the
same principle also ee˙ (he) and shee˙; finally dhai.
Whom has here as well as in Scotch been completely
superseded by who.

Compare also you were better do it, where you was a
dative and is now the subject in the nominative, and
where simultaneously were has changed imperceptibly
from the third person singular (it being understood)
to the second person pl. or sg. In handing something
to some one you will often say, “Here you are!”
meaning, “Here is something for you, here is what you
want”. I think that this phrase too contains an old
dative; and perhaps, some centuries ago, in handing
only one thing, people would say, “Here you is!” 1237

215. If now finally we ask: Are the changes described
in this chapter on the whole progressive?
276the answer must be an affirmative one. Although
for obvious reasons (see § 64) pronouns are more apt
to preserve old irregularities than other classes of
words, we find instead of the old four irregular forms,
thou, thee, ye and you, one form carried through uniformly;
the same uniformity is, as far as case is concerned,
observable in the self-forms as compared with
the old he self, hine self, etc., and who shows almost
the same indifference to cases. Then there is some
progress in syntax which does not appear from the
scheme just given. Many of the uncertainties in the
choice of case exemplified in the early sections of the
chapter are owing to a want of correspondence between
the logical and grammatical categories; for instance,
when a word might be logically, but not
grammatically, the subject. Sometimes, also, one
grammatical rule would require one case, and another
equally applicable rule a different one. The inconsistency
was particularly glaring where the logical
(and psychological) subject was to be put in quite
another case than that generally used to denote the
subject; and here, with the old impersonal verbs and
in the absolute construction, logic has completely conquered
the old grammar. The rule which is entirely
incompatible with the old state of things, that the
word immediately preceding the verb is logically and
grammatically the subject of the sentence, has been
carried through on the whole with great consistency.
And in the great facility which the English have now
acquired of making the real psychological subject
277grammatically the subject of a passive sentence, the
language has gained a decided advantage over the
kindred languages, an advantage which Danish is even
now struggling to acquire, in spite of the protests of
the schoolmaster grammarians. Thus we see that
many phenomena, which by most grammarians would
be considered as more or less gross blunders or “bad
grammar,” but which are rather to be taken as
natural reactions against the imperfections of traditional
language, are really, when viewed in their
historical connexion, conducive to progress in language.278

Chapter VIII.
The english group genitive.

216. To a mind trained exclusively in Latin (or
German) grammar such English constructions as
“the Queen of England's power,” or “he took
somebody else's hat,” must seem very preposterous;
the word that ought to be in the genitive case (Queen,
somebody) is put in the nominative or accusative,
while in the one instance England, whose power is
not meant, and in the other even an adverb, is put in
the genitive case. Similarly, in the case of “words in
apposition,” where it might be expected that each
would be put in the genitive, as in “King Henry the
Eighth's reign,” only one of them takes the genitive
ending.

217. In an interesting and suggestive article, “Die
genetische erklärung der sprachlichen ausdrucksformen”
(Englische Studien, xiv., 99), H. Klinghardt
makes an attempt to explain this as well as
other peculiarities of English grammar (the passive, in
“the request was complied with,” “he was taken no
notice of,” “with one another,” etc.), by the power
of the accent. “In English,” he says, “unstressed
279vowels are weaker than in German; and the distinction
between stressed and unstressed syllables
greater. So it is with the stressed words of a
sentence in relation to the unstressed words surrounding
them; the action of stress therefore reaches
farther than in German; emphatic words are capable
of gathering around them a greater number of weak
words than in German… The [German] pupil
will now understand how easily and conveniently in
English small groups of words, such as King Henry the
Eighth, are joined together under one accent, and are
inflected, put in the Saxon genitive, etc., exactly in
the same manner as single words.”

218. I do not think that this theory is the correct
one, and I shall state my objections. In the first
place, we are not told which word in the group is
invested with that powerful accent that is said to
keep the group together. Nothing hinders us from
pronouncing a group like “King Richard the Second's
reign” at one moment with strong stress on Richard
(as opposed to, say, Edward II.) and at the next with
great emphasis on the numeral (as opposed to Richard
the Third); we may also pronounce the two words
with even stress; yet in all of these cases the grammatical
construction is the same. Next, if we adopt
Dr. Klinghardt's theory, we must assume an historical
change in English accent which seems to be supported
by no other fact. And thirdly, the theory fails completely
to account for the difference between the
final s in genitives like Queen of England s or sister-in-law's,
280and the internal s in plurals like the queens of
England or sisters-in-law.

Before venturing to propose a new explanation it
will be well to look somewhat closely at the historical
development of the several phenomena with which we
are here concerned. I shall group my examples
under six heads.

I. Attributive words.

220. In Chaucer we find no single trace of an
inflected genitive of any attributive adjective; the
rapid disappearance of the s in the gen. may to
a great extent be due to the analogical influence of
the weak forms of the adjective, in which after the
loss of the final n the endings were the same for the
genitive as for all the other cases.

In present-day English most adjectives are placed
before their nouns, and then are never inflected; an
adjective put after its noun is only capable of assuming
the genitive s in cases like Henry the Eighth's;
it is impossible to say, for example, the women
present's opinions. Comp. Marlowe, Jew, 242, “That
you will needs haue ten years [genitive!] tribute
past” (= the tr. of ten years past).

II. Words in apposition.

222. But in a great many cases, where we have
this word-order — and it is, indeed, the order most
frequently used throughout the M. E. period 1239 — there
can be no doubt that the last word is put in the
nominative (or common) case. The leaving out of
the case-sign is rare in Old English., but extremely
283common in Middle English; in Modern English it
is getting rarer again. The phenomenon is to be
classed with those mentioned above, § 163.

A.S. Chron., E., 855, “To Karles dohtor Francna
cining” | A. R., 148, “Moiseses hond, Godes
prophete” | ibid., 244, “Þuruh Iulianes heste
Þe amperur” | 352, “Ine Jesu Cristes rode,
mi louerd” | Ch., Hous of F, 142, “Seys
body the king” | 282, “The kinges meting
Pharao” | Ch., B., 431, “Kenulphus sone,
the noble king of Mercenrike” | F., 672, “The
god Mercurius hous the slye” | L. G. W.,
1468, “Isiphilee the shene, That whylom
Thoas doghter was, the king” | Malory, 70,
“By my faders soule Vtherpendragon” |
91, “Gaweyn shalle reuenge his faders deth
kynge Loth” | 126, .“In his wyues armes
Morgan le fay” | Marl., Tamburl., 193, “In
the circle of your fathers armes, The
mightie Souldan of Egyptia” | Greene,
Friar B., 2, 10, “To Bacon's secret cell, A
friar newly stall'd in Brazennose” | Sh., 1
H. IV., ii, 4, 114, “I am not yet of Percies
mind, the Hotspurre of the North, he that
killes me some sixe or seauen dozen of
Scots” | Matt., xiv., 3 (Auth. V.), “For
Herodias' sake, his brother Philip's wife” |
Wycherley (Mermaid Ser.), 24, “He has
now pitched his nets for Gripe's daughter,
the rich scrivener” | Tennyson, 322, “Merlin's
284hand, the Mage at Arthur's court” | Mth.
Arnold, Poems, i., 191, “Doubtless thou
fearest to meet Balder's voice, Thy brother,
whom through folly thou didst slay”. 1240

224. When the governing word is not expressed,
the s-ending is — or was — often added to the first
noun exclusively; Lindley Murray says (Grammar,
8th edit, p. 262) that of the three forms, “I left the
parcel at Smith's, the bookseller”: or “at Smith, the
bookseller's”: or “at Smith's, the bookseller's,” —
the first is most agreeable to the English idiom; and
if the addition consists of two or more nouns, the case
seems to be less dubious; as, “I left the parcel at
Smith's, the bookseller and stationer”. This does not
now apply to a group consisting of a title and a proper
name, as it did formerly, witness the first two of the
following quotations, which would in modern speech
be King Alexander's and Admiral Presane's. Even
the last example does not seem to be now very
natural; and custom is perhaps more and more in
favour of saying “at Smith, the bookseller's,” or “at
Smith's, the bookseller's,” unless “the bookseller” is
only part of a phrase, e.g., “at Smith's, the bookseller
in Trinity Street”. At least, this is the opinion
of Mr. G. C. Moore Smith.

Guy of Warw., 7921, “Hyt [the helme] was Alysawndurs
the kynge” | ibid., 8714, “Hyt [the
cuntre] ys admyrals Presane” | Sh., H. V., i.,
2, 105, “Inuoke his warlike spirit, and your
great vnckles, Edward the Black Prince”
| Thack., P., i., 259, “He managed to
run up a fine bill at Nine's, the livery stablekeeper”
| ibid., ii., 199, “I remember at poor
Rawdon Crawley's, Sir Pitt Crawley's
286brother” | Beaconsf., Loth., 16, “Villas like
my cousin's, the Duke of Luton”.

225. When one of the words in apposition is a personal
pronoun a special difficulty arises from the
genitive proper being here replaced by a possessive
pronoun. What is the genitive of “we, the tribunes”?
It would be a little awkward to say “our, the tribunes'
power,” and so most people would probably say with
Shakespeare (Cor., iii., 3, 100), “the power of vs the
tribunes”.

226. It will be noticed that in most cases it is
perfectly immaterial to the meaning of the passage
whether we take both as qualifying the pronoun or
the following substantive, as each of us has only one
head, one throat, one life, etc. But in other instances
the same consideration does not hold good; when
we read, for instance, in John Halifax, Gent., ii., 76,
“the name set both our thoughts anxiously wandering,”
the meaning cannot be that each of them had only
got one wandering thought, so that both must certainly
here be taken as a genitive case. But the tendency
goes undoubtedly in the direction of taking both as a
nominative, the construction being avoided whenever
that would be obviously impossible: I suppose it
would be fruitless to search through the whole of
the English literature for a connexion like “both our
four eyes,” although, indeed, Fielding writes (Tom
Jones, iii., 45): “Both their several talents were excessive”
(each had several talents); compare ibid.,
iii., 66, “The two ladies who were riding side by
side, looking steadfastly at each other; at the same
moment both their eyes became fixed; both their
horses stopt,” etc.

III. Connected by a preposition.

230. Two nouns are connected by a preposition,
e.g., father-in-law, the Queen of England. In old times
such word-groups were not felt as inseparable units,
as they are now; witness Chaucer, B., 3870, “Ageyn
Pompeius, fader thyn in lawe”. Consequently, when
they were to be used in the genitive, they were
separated by the governing word; this was the
universal practice up to the end of the fifteenth
century.

Ch., B., 3442, “of kinges blood Of Perse is she
descended” | B., 3846, “Philippes sone of
Macedoyne” | E., 1170, “for the wyues loue
of Bathe” | M., iv., 108, “That was the
kynge Priamus sone of Troye” | Malory,
45, “The dukes wyf of Tyntagail” | 127, “I
am the lordes doughter of this castel” | 141,
“The kynges sone of Ireland” etc.

The same construction is resorted to even in more
recent times whenever the ordinary construction
would present special difficulties. It is possible to
denote a lady as “she in the cap,” but how about the
genitive case of such a group? Shakespeare says:
“What's her name in the cap?” (L. L. L., ii., 209) — “For
honour of former deeds' sake” would be rather heavy;
so Milton puts it (Sams. Ag., 372), “For honour's sake
of former deeds”. Compare also Sh., 1 H. IV.,iii., 2,
293119, “The Archbishops grace of York” = the Archbishop
of York's grace = his Grace the Archbishop
of York.

Sometimes, but very rarely indeed, an ambiguity
may arise from this sort of construction, as in the
well-known puzzle: “The son of Pharaoh's daughter
was the daughter of Pharaoh's son”.

In ordinary language the construction is found
only with the preposition of and in the words son-in-law, 1244
295etc., so also the Commander-in-Chief's levees
(Thack., Esmond, i., 345) and perhaps: “for God in
Heaven's sake”. But in dialects it is used with other
prepositions as well; Murray gives as Scotch (Dial
of the Southern Counties, p. 166): “the man-wui-the-quheyte-cuot's
horse”; and Elworthy quotes from
Somersethire (Gramm. of the Dial. of W. Soms.,
p. 157): Jan Snéok uwt tu Langvurdz duung kee,
“John Snook out of Langford's donkey”; Mr. Buurj
tu Shoaldur u Muutuns paig, “Mr. Bridge of the
Shoulder of Mutton's pig”.

232. What is the genitive of some of them, any of
you, one of us? There is some difficulty here, and the
reason of it is the same as we met with before, viz.,
the difference between a genitive proper and a possessive
pronoun, cf. § 225. In olden days, when a
partitive relation could be expressed by the gen. pl.,
we occasionally find formations like these: A. R.,
204, “hore summes nome” (the name of some of them),
where the genitive ending is tacked on to the nom.,
or Orrm, l. 2506, “& all onn ane wise fell till eʒʒÞer
Þeʒʒress herrte” (to the heart of either of them),
where it is added to the old gen. pl.

Dr. Murray once told me that it would be possible
for a Scotchman to add the s to the whole of such
a combination (“Is this ony of you's?”),and that you
might even, though rarely, in colloquial English hear
“This must be some of you's”. I have some suspicion
that this construction is a little less rare in
colloquial language when there is a word added in
apposition to you: “Is this any of you children's?”

IV. Defined by a following adverb.

But some English friends have corroborated my
conjecture that it would be more natural to say, e.g.,
“It doesn't matter whose ever it is,” than “whoever's,”
which would indeed, according to some, be impossible
in this connexion; and if the elements of the word
are separated, who of course is inflected, as in Sh.,
R. III., iv., 4, 224, “whose hand soeuer”.

V. Conjunction.

234 When one word should properly govern two
or more genitives, connected by and or some other
conjunction, it makes some difference whether the
governing word is placed after the first or after the
last of the genitives.

In view of all these examples, it will not be easy to
lay down fully definite and comprehensive rules for
determining in which cases the group genitive is
allowable and in which the s has to be affixed to each
member; the group construction is, of course, easiest
when one and the same name is common to two
persons mentioned (Mr. and Mrs. Brown's compliments),
or when the names form an inseparable group
(Beaumont and Fletcher's plays; Macmillan & Co.'s
publications). On the whole, the tendency is towards
using the group genitive, wherever no ambiguity is
caused by it.

VI. A relative clause.

In English dialects the phenomenon seems to be
very widely spread; thus in Scotland (Murray, p.
308166), “The-màn-ăt-ye-mæt-yesterday's dowchter”; in
Cheshire (Darlington, E. D. S., xxi., p. 55), “I've
just seen Jim Dutton, him as went to 'Meriky's weife”
= the wife of J. D., the man who went to America;
in Somersetshire (Elworthy, Gr., 15), “That's the
woman what was left behind's child,” i.e., that is the
child belonging to the woman who was left behind.

240. After thus passing in review all the different
kinds of group genitives, 1249 it remains for us to find an
explanation that will account for all the facts mentioned.
It is obvious that the reason of our phenomenon might
309be sought either in the nature of the compound group,
or in that of the ending and its function.

It might perhaps be urged that the phenomenon
was due to the natural instinct taking the Queen of
England or King Henry the Eighth as one inseparable
whole, that would allow of no case-ending separating
its several elements. The case would then be a
parallel to the German treatment of those word-groups
which, like sack und pack, grund und boden,
have been fused together to the extent of making it
impossible to inflect the former word and say, e.g.,
mit sacke und packe or grundes und bodens; indeed,
we here, though very rarely, may find something
corresponding to the English group genitive; thus,
Wieland has “des zu Abdera gehörigen grund und
bodens”. 1250 But an inspection of the above collected
examples will show that the explanation does not
hold good; for in the majority of cases we have not
only group-compounds, but also free groups 2251 inflected
like single words. This feeling of connectedness may
310have gone for something in the development of
the modern word-order where the genitive of the
Queen of England is placed before the governing
noun, instead of the old “the Queen's crown of
England”; and it undoubtedly plays some part in
the cases mentioned in § 237 (A and B's); but it
gives no satisfactory explanation of the difference
between the plural the Queens of England and the
genitive the Queen of England s.

241. As the nature of the group fails to give an
answer to our question we turn our attention to the
ending, and the first thing that strikes us is that we
find no trace of the group genitive with any of the
O. E. genitive endings -a, -ra, -an, -e, -re, etc. (cf. § 127),
but only with -(e)s. It is not till this ending has
practically superseded all the other ways of forming
the genitive that our phenomenon begins to make
its appearance. In other words, the first condition
of forming genitives of whole groups as if they were
single words is that the manner of formation of
genitives should be on the whole uniform. Where
the genitive is formed irregularly, as is now only the
case with the personal pronouns, we have had until
the present day only rudimentary and feeble attempts
at group genitives.

242. Now, if we were to ask: What is the reason
of this regularity in the formation of English noun
genitives? then any student that is at all acquainted
with modern linguistic theories and methods would
be out with the answer: “Why, it is due to analogy;
311the s-ending has gradually been extended to the
whole of the vocabulary, the analogy of those nouns
which had an s-genitive in O. E. prevailing over
the others”.

Very good; the answer is obviously correct. And
yet it is not entirely satisfactory, for it does not
account for the difference observable in many words
between the formation of the genitive and that of the
plural. In the latter, too, the s-ending has been
analogically extended in pretty much the same way
as in the former; but how is it that we so often see
the irregular plural preserved, whereas the genitive
is always regular? We have the irregular plurals
men, children, oxen, geese, etc., as against the regular
genitives man's, child's, ox's, goose's, etc. In the days
of Chaucer and Shakespeare the plural and the
genitive of most words ending in f, e.g., wife and life,
were identical, wives and lives being said in both
cases; why has the analogy of the nom. sg. been
more powerful in the genitive (modern wife's, life's)
than in the plural?

The only explanation, as far as I can see, lies in
the different function of the two endings; if we put
a singular word into the plural, the change affects
this word only; its relation to the rest of the proposition
remains the same. But if, on the other
hand, we put a word in the genitive case which was
in the nominative, we change its syntactical relation
completely; for the function of a genitive is that of
closely connecting two words.312

243. There is yet another thing to be noted.
The O. E. genitive had many different functions;
we may broadly compare its syntax to that of the
Latin genitive. We find in Old English possessive,
partitive, objective, and descriptive genitives; genitives
governed by various adjectives and verbs, etc. And
the position of the genitive is nearly as free as it is
in Latin. But if you will take the trouble to read a
few pages of any Old English prose book, of the
Anglo-Saxon chronicle, of King Alfred, or of Ælfric,
you will soon observe that where the Old English
genitive might be rendered by a genitive in Modern
English, it nearly always precedes its noun; where
the word-order is different, the old genitive construction
has, in the majority of cases, been abandoned.
It is a significant fact that the only surviving use of
the English genitive is a prepositive one; the word-order
“the books my friend's” for “my friend's
books” is, and has been for many centuries, as
impossible in English.,as it is frequent in German:
“die bücher meines freundes”.

244. We are now in a position to draw our conclusions.
The s is always wedged in between the two
words it serves to connect; it is, accordingly, felt as
belonging nearly as much to the word following it
as to the preceding one. Nay, it is now more
important that the s should come immediately before
the governing word than that it should come immediately
after the noun which it turns into a genitive
case. It is now partly a suffix as of old, partly a
313prefix; if we were allowed to coin a new word we
should term it an interposition.

This peculiar development gives us the clue to the
problems mentioned above. If the s of the genitive
is more loosely connected with the word it belongs
to than is the s (or other suffix) of the plural, that
is the reason why it tolerates no change in the body
of the word: the old plural wives may remain; but
the genitive (originally wives also) must be made
to agree with the nominative — and so it becomes
wife's. 1252

And we now see clearly why such groups as the
Queen of England, when put in the genitive, affix the
s to the last word of the group, but when put in the
plural, to the first.

246. Now, let us sum up the history of the genitive
ending s.

In the oldest English it is a case-ending like any
other found in flexional languages; it forms together
with the body of the noun one indivisible whole, in
which it is often impossible to tell where the kernel
of the word ends and the ending begins (compare
315endes from ende and heriges from here); the ending is
only found in part of the vocabulary, many other
genitive endings being found elsewhere.

As to syntax, the meaning and the function of these
genitive endings are complicated and rather vague;
and there are no fixed rules for the position of the
genitive in the proposition.

In course of time we witness a gradual development
towards greater regularity and precision. The partitive,
objective, descriptive and some other functions
of the genitive become obsolete; the genitive is
invariably put immediately before the word or words
it governs: irregular forms disappear, the s-ending
only surviving as the fittest, so that at last we have
one definite ending with one definite function and one
definite position. If the syntactical province of the
genitive has been narrowed in course of time, the
loss — if such it be — has been compensated, and
more than compensated, as far as the s-ending is
concerned, by its being now the sole and absolute
sovereign of that province; its power is no longer
limited to some masculine and neuter nouns nor to
one number only; it rules irrespective of gender and
number.

247. In an Old English genitive the main (“full”)
word and the case-forming element are mutually dependent
on each other, not only in such genitives as lufe or
suna or bec or dohtor, but also in the more regular formations
in -es; one part cannot be separated from
the other, and in the case of several words belonging
316together, each of them has to be put in the genitive
case: anes reades mannes | Þære godlican lufe | ealra
godra ealdra manna weorc, etc.

It is important to notice that here historically
attested facts show us in the most unequivocal way a
development — not, indeed, from an originally self-existent
word to an agglutinated suffix and finally
to a mere flexional ending, but the exactly opposite
development of what was an inseparable part of a
complicated flexional system to greater and greater
emancipation and independence.

Appendix to chapter VIII.
“Bill stumps his mark,” etc.

248. The tendency to turn the genitive ending
into an independent word meets with, and is to a
certain degree strengthened by, a phenomenon that
has originally nothing to do with it; I mean, the
expression of a genitive relation by a common case
plus a possessive pronoun. The best known instance
of this is “for Jesus Christ his sake” in the Common
Prayer Book.318

This peculiar idiom is not confined to English:
it is extremely common in Danish., Norwegian and
Swedish dialects, in Middle and Modern Low German,
in High German (Goethe: “Ist doch keine menagerie
So bunt wie meiner Lili ihre!”), in Magyar, etc.
In English the phenomenon has been noticed by
many grammarians; 1256 and if any one wishes to see
other or more instances than those from which I have
tried to form an idea of the origin and character of
the idiom, it is to their works that I must refer him.

249. In most cases the phenomenon is a form of
that anacoluthia which I have already had occasion
to mention (see § 162), and which consists in the
speaker or writer beginning his sentence without
thinking exactly of the proper grammatical construction
of the word that first occurs to him, so that he
is subsequently obliged to use a correcting pronoun.
As this want of forethought is common everywhere
and at all times, we find the grammatical irregularity
in many languages, 2257 and it is naturally very frequent
when a lengthy clause is introduced: it is also often
319resorted to where a foreign name is introduced that
does not conform to the native declensions.

The similarity between this sentence from the
Revised Version and “the man I saw yesterday's
father” is conspicuous.

252. Sprung as it is, then, from various sources,
this makeshift genitive now converges with and meets
323the originally totally different interpositional descendant
from the old flexional s-genitive, so that the
two formations become often practically indistinguishable. 1259
The similarity is of a purely phonetic nature;
his would, of course, be pronounced with weak stress,
and in unstressed words in the middle of a sentence
h is scarcely if at all audible (as in the rapid pronunciation
of “he took his hat,” etc.; compare also
it for older hit, and 's for has). Thus, Þe bissop his
broÞer, etc., in the B-text of Layamon, may be only
another way of writing bissopis or bissopes. 2260

Compare the treatment of the verbal form is: that's,
there's, this is. In Elizabethan English., it was treated
similarly. I saw't, for't, do't, upon't, done't, etc. So
also us (comp. mod. let's): upon's, among's, upbraid's,
behold's, etc.

In its struggle for an independent existence, the
s-interposition seemed likely to derive great assistance
from the concurrence of the his-construction.
But the coincidence was not to last long. On the
one hand, the contraction of the weak his seems to
have been soon given up, the vowel being reintroduced
from the fully stressed form, even where the h
was dropped (he took 'is hat); on the other hand, the
limited signification of the possessive pronoun counteracted
the complete fusion which would undoubtedly
have taken place, if his had been common to all
genders and to both numbers, instead of being confined
to the masc. (and in former centuries the neuter)
sg. A formation like “Pallas her glass” (quoted by
Abbott from Bacon) does not fit in with the rest of
the system of the language, and “Pallas his glass”
would jar upon English ears because his is too much
felt as a pronoun denoting sex.327

Chapter IX.
Origin of language.

I. Method.

256. Goethe, in his Dichtung und Wahrheit, relates
how in Strasburg he was in constant intercourse
with Herder at the time when the latter was engaged
in writing his prize essay for the Berlin Academy on
the origin of language; and how he read the manuscript,
although, as he confesses himself, he was very
little prepared to deal with that subject; “I had,” he
says, “never bestowed much thought on that kind of
thing; I was still too much engrossed by present
things (zu sehr in der mitte der dinge befangen) to
think about their beginning or end”.

If it is not presuming too much to compare oneself
with Goethe, even in so small a matter, and one,
moreover, of so negative a character, I must confess
that I too, like Goethe, have given most study to
languages as they are now-a-days, to the “middle” of
languages; the earlier stages I have studied mainly,
if not exclusively, in so far as they are capable of
throwing light upon the languages which are still
living: I have therefore only an imperfect and sporadic
328knowledge of the vast literature which deals with
the origin of speech; and the impressions left by
occasionally reading some book or short paper on the
subject have not encouraged me to master that literature
more systematically. Under these circumstances I
felt greatly relieved to come across the following verdict
of Whitney's: “No theme in linguistic science is
more often and more voluminously treated than this,
and by scholars of every grade and tendency; nor
any, it may be added, with less profitable result in
proportion to the labour expended; the greater part
of what is said and written upon it is mere windy
talk, the assertion of subjective views which commend
themselves to no mind save the one that produces
them, and which are apt to be offered with a
confidence, and defended with a tenacity, that are in
inverse ratio to their acceptableness. This has given
the whole question a bad repute among sober-minded
philologists.” 1262

257. Although I look upon all previous attempts
to penetrate the secret with very much the same
feelings as those of the fox in the fable, when he
noticed that all the traces led into the den, and not a
single one came out, I shall ask my readers to join
me in casting a rapid glance at those theories which
have hitherto been most generally accepted as containing
the clue of our problem. In mentioning them
I shall make use of those nicknames by which they
329are familiar to readers of the discussion between Max
Müller and Whitney.

First comes the old bow-wow theory: Primitive
words were imitative of sounds; man copied the
barking of dogs and thereby obtained a natural word
with the meaning of “dog” or “bark”.

The next theory is the ding-dong or “nativistic”
theory: according to this there is a somewhat mystic
harmony between sound and sense; “there is a law
which runs through nearly the whole of nature, that
everything which is struck rings. Each substance
has its peculiar ring. Gold rings differently from tin,
wood rings differently from stone; and different
sounds are produced according to the nature of each
percussion. It was the same with man.” Language
is the result of an instinct, a “faculty peculiar to man
in his primitive state, by which every impression from
without received its vocal expression from within”.
But this “creative faculty which gave to each conception
as it thrilled for the first time through the brain
a phonetic expression, became extinct when its object
was fulfilled” (Max Müller, who has, however, abandoned
this theory).

The pooh-pooh theory derives language from interjections,
instinctive ejaculations called forth by pain
or other intense sensations or feelings.

The fourth and last of these theories is the jo-he-ho,
first propounded by Noiré, 1263 and subsequently adopted
330by Max Müller: under any strong muscular effort it
is a relief to the system to let breath come out
strongly and repeatedly, and by that process to let
the vocal chords vibrate in different ways; when
primitive acts were performed in common, they would,
therefore, it is said, naturally be accompanied with
some sounds which would come to be associated with
the idea of the act performed, and stand as a name
for it; the first words would accordingly mean something
like “heave” or “haul”.

258. Now, these theories — which, by the way, it is
rather difficult to represent with perfect impartiality
in a few lines — denounce and combat each other;
thus Noiré and Geiger, in their explanation of the
origin of speech, think it perfectly possible to do
entirely without sound imitation, or onomatopoeia.
And yet what would prevent our uniting these
several theories and using them concurrently? It
would seem to matter not so very much whether the
first word uttered by man was bow-wow or pooh-pooh,
for the fact remains that he has said both one and
the other. Each one of the theories — save, perhaps,
the ding-dong, which is hardly anything but a rather
misty variation of the interjectional theory — is able to
explain parts of language, but still only parts, and not
even the most important parts — the main body of
language they hardly seem even to touch.331

To all the theories, that of Noiré only excepted, it
may further be objected that they are too individual;
they do not touch language as a means of human
intercourse; as Ellis puts it, “The Pooh-pooh! the
Bow-wow! and the Ding-dong! theories might serve
for Robinson Crusoe. With Man Friday would begin
real language — attempted and partially effected interchange
of thought by mouth and ear”. 1264 Moreover,
they all tacitly assume that up to the creation of
language man had remained mute or silent; but this
is most improbable from a physiological point of
view. As a rule we do not find an organ already
perfected on the first occasion of its use; an organ is
only developed by use.

259. As to the bow-wow theory in particular, it is
in the first place rather an unlucky hit that the dog's
cry should have been chosen of all others; for naturalists
maintain that dogs did not learn to bark till
after their domestication (one might perhaps wish that
they had not learned then!). But apart from this —
and we might of course just as well use some other
animal's cry to name the theory after; there is abundance
of choice — it still seems rather absurd, as remarked
by Renan, to set up this chronological
sequence: first the lower animals are original enough
to cry and roar; and then comes man, making a
language for himself by imitating his inferiors.

To the advocates of the pooh-pooh theory it must be
objected that they do not go deep enough when they
332take interjections for granted, without asking where
they originate. This is a question which philologists
have entirely disregarded; but natural science has
offered an explanation of at least some of our interjections.
In Darwin's interesting work on The Expression
of the Emotions, which it is not to the credit
of the science of language to have overlooked, purely
physiological reasons are given for the feeling of contempt
or disgust being accompanied by a tendency
“to blow out of the mouth or nostrils, and this produces
sounds like pooh or pish”. And Darwin goes
on to say: “When any one is startled or suddenly
astonished, there is an instantaneous tendency, like-wise
from an intelligible cause, namely, to be ready
for prolonged exertion, to open the mouth widely, so
as to draw a deep and rapid inspiration. When the
next full expiration follows, the mouth is slightly
closed, and the lips, from causes hereafter to be discussed,
are somewhat protruded; and this form of
the mouth, if the voice be at all exerted, produces,
according to Helmholtz, the sound of the vowel O.
Certainly a deep sound of a prolonged Oh! may be
heard from a whole crowd of people immediately after
witnessing any astonishing spectacle. If, together
with surprise, pain be felt, there is a tendency to contract
all the muscles of the body, including those of
the face, and the lips will then be drawn back; and
this will perhaps account for the sound becoming
higher and assuming the character of Ah! or Ach!”

260. It is a common feature of all previous attempts
333at solving the question that the investigator has conjured
up in his imagination a primitive era, and then
asked himself: How would it be possible for men or
manlike beings, who have hitherto been unable to
speak, to acquire speech as a means of communication
of thought? Not only is this method followed, so to
speak, instinctively by everybody, but we are even
positively told (by Marty) that it is the only method
possible. In direct opposition to this assertion I
should like to advance the view that it is chiefly and
principally due to this method and to this manner of
putting the question that the result of all attempts to
solve the problem has been so very small. Linguistic
philosophers have acted very much as the German
did in the well-known story, who set about constructing
the camel out of the depths of his inner consciousness.
Hegel began his philosophy with pure non-existence,
and thence took a clean jump to pure existence; and
our philosophers make the same jump with regard to
language. But jumps are dangerous if you have no
firm ground to take off from!

If we are to have any hope of success in our investigation,
we must therefore look out for new methods
and new ways; and there are, as far as I can see, only
two ways which lead us to where we may expect to
see new views opened before us over the world of
primitive language.

261. One of these has its starting-point in the
language of children. On a great many points biologists
have utilised the discovery that the development
334of the individual follows on the whole the same course
as that of the race; the embryo, before it arrives at
full maturity, will have passed through essentially the
same stages of development which in countless generations
have led the whole species to its present level.
Would it then be surprising if the course of development
by mankind at large of the faculty of speech
and the mental conceptions therein implied should be
humbly mirrored to us in the process by which any
child learns to use its vocal organs to communicate
its thoughts?

This idea has obviously been present, more or less
consciously, to many; and children's language has
often been invoked to furnish illustrations and parallels
of the process gone through in the formation of
primitive language. But I cannot help thinking that
philologists have generally been guilty of an erroneous
inference in applying this principle; inasmuch as
they have taken all their examples from a child's
acquisition of an already existing language. The
fallacy will be evident if we suppose for a moment
some one endeavouring to imagine the evolution of
music from the manner in which a child is now-a-days
taught to play on the piano. Manifestly the modern
learner is in quite a different position to primitive
man, and has quite a different task set to him: he has
an instrument ready to hand, and melodies already
composed for him, and finally a teacher who understands
how to draw these tunes forth from the instrument.
It is just the same thing with language: the
335task of the child is to learn an existing language, that
is, to connect certain sounds heard from the mouths
of its fellow-creatures with the same ideas which the
speakers associate with them, but not in the least to
frame anything new. No; if we are seeking some
parallel to the primitive acquisition of language, we
must look elsewhere and go to baby language as it is
spoken in the first year of life, before the child has as
yet begun to “notice” and to make out what use
is made of language by grown-up people. Here, in
the child's first purposeless murmuring, crowing, and
babbling, we have real nature sounds; here we may
expect to find some clue to the infancy of the
language of the race.

262. The second way hinted at above is likely to
yield more important results; it is exactly the opposite
of that followed by the propounders of the usual
theories. They make straight for the front of the
lion's den; we have seen that this is fruitless (vestigia
terrent!) and we will therefore try and steal into the
den from behind. They think it logically correct,
nay necessary, to begin from the beginning; let us,
for variety's sake, begin from the “middle” of things,
from languages as accessible at the present day, and
let us attempt from that starting-point step by step
to trace the backward path. Perhaps in this way we
may reach the very first beginnings of speech.

The method I recommend is, in other words, to
trace our modern nineteenth-century languages as far
back in time as history and our materials will allow
336us; and then, from this comparison of present English
with Old English or Anglo-Saxon, of Danish with
Old Norse, and of both with “Common Germanic,”
of French and Italian with Latin, of Modern Persian
with Zend, of modern Indian dialects with Sanskrit,
etc., to deduce definite laws for the development of
languages in general, and to try and find a system
of lines which can be lengthened backwards beyond
the reach of history. If we should succeed in discovering
certain qualities to be generally typical of
the earlier as opposed to the later stages of languages, 1265
we shall be justified in concluding that the same
qualities obtained in a still higher degree in the beginning
of all; if we are able within the historical
era to demonstrate a definite direction of linguistic
evolution, we must be allowed to infer that the direction
was the same even in those primeval periods for
which we have no documents to guide us. But if
the change witnessed in the evolution of modern
speech out of older forms of speech is thus on a
larger scale projected back into the childhood of
mankind, and if by this process we arrive finally at
uttered sounds of such a description that we cannot
help thinking that this is no longer a real language,
but something antecedent to language — why, then
the problem will have been solved; for transformation
is something we can understand, while a creation out
337of nothing can never be comprehended by human
understanding; it can at best be left to stand as a
religious postulate, as a miracle or a crux.

This, then, will be the object of the following rapid
sketch: to search the several departments of the
science of language for general laws of evolution —
most of them have already been indicated and
discussed at some length in the opening chapters of
this volume — then to magnify the changes observed,
and thus to form a picture of the outer and inner
structure of some sort of speech more primitive than
the most primitive language accessible to direct
observation.

II. Sounds.

263. First, as regards the purely phonetic side of
language, we observe everywhere the tendency to
make pronunciation more easy, so as to lessen the
muscular effort; difficult combinations of sounds are
discarded, those only being retained which are pronounced
with ease. In most languages therefore
only such sounds are used as are produced by expiration,
while inbreathed sounds and “clicks” or suction-stops
are not found in connected speech. In civilised
languages we meet with such sounds only in interjections,
as when an inbreathed voiceless l (generally
with rhythmic variations of the strength of breathing
and corresponding small movements of the tongue) is
used to express enjoyment, especially the enjoyment
caused by eating and drinking, or when a click formed
338with the tip of the tongue (generally, but rather
inadequately spelled tut in our alphabet, which is not
at all adapted to the writing of such sounds) is used
to express impatience; in drivers' shouts to their
horses some other clicks occur. In some very
primitive South-African languages, on the other hand,
these and similar sounds are found as integral portions
of words; and Bleek's researches render it probable
that in former stages of these languages they were in
more extensive use than now. We may perhaps
draw the conclusion that primitive languages in
general were extremely rich in all such difficult
sounds.

264. Of much more far-reaching consequence is
the following point. In some languages we find a
gradual disappearance of differences of musical accent
(or pitch); this has been the case in Danish., whereas
Norwegian and Swedish have kept the old tones;
so also in Russian as compared with Servo-Croatian.
With regard to the tones in use in most early languages
it is extremely difficult to state anything with certainty,
as written documents scarcely ever indicate such
things; still, we are fortunate enough in the works
of old Indian, Greek and Latin grammarians to have
express statements to the effect that pitch accents
played a prominent part in those languages, and that
the intervals used must have been comparatively
greater than is usual in our modern languages. No
doubt the same thing may be asserted with regard
to languages spoken now-a-days by savage tribes,
339though here too our materials are very scanty, as
most of the writers who have made a first-hand study
of such languages have not had the necessary qualifications
for undertaking this kind of investigation;
nor can this astonish us, seeing how imperfectly tonic
accents have been hitherto studied even in the best-known
European languages. Here and there, however,
we come across some information about peculiar
tonic accents, as, for instance, in the case of some
African languages. 1266

265. So much for word-tones; now for the sentence
melody. It is a well-known fact that the
modulation of sentences is strongly influenced by the
effect of intense emotions in causing stronger and
more rapid raisings and sinkings of the voice. I
may here refer to the excellent introduction to Herbert
Spencer's essay on The Origin and Function
of Music, where the illustrious author examines the
340influence of the feelings on the loudness, quality or
timbre, pitch, intervals, and rate of variation of the
sounds uttered. “The utterances grow louder as the
sensations or emotions, whether pleasurable or painful,
grow stronger… The sounds of common conversation
have but little resonance; those of strong
feeling have much more. Under rising ill-temper the
voice acquires a metallic ring… Grief, unburdening
itself, uses tones approaching in timbre to those of
chanting; and in his most pathetic passages an
eloquent speaker similarly falls into tones more
vibratory than those common to him… While
indifference or calmness will use the medium tones,
the tones used during excitement will be either above
or below them; and will rise higher and higher, or
fall lower and lower, as the feelings grow stronger…
Extreme joy and fear are alike accompanied by shrill
outcries… While calm speech is comparatively
monotonous, emotion makes use of fifths, octaves, and
even wider intervals… The remaining characteristic
of emotional speech which we have to notice is
that of variability of pitch… On a meeting of
friends, for instance — as when there arrives a party of
much-wished-for visitors — the voices of all will be
heard to undergo changes of pitch not only greater
but much more numerous than usual.” 1267341

These facts and considerations all seem to point
to the conclusion that there once was a time when
all speech was song, or rather when these two actions
were not yet differentiated; but I do not think that
this inference can be established inductively at the
present stage of linguistic science with the same
amount of certainty as the statements I am now
going to put forth as to the nature of primitive
speech.

267. Linguistic evolution seems constantly to
display a tendency to shorten words. Besides the
shortening processes shown in such instances as cab
for cabriolet and bus for omnibus (above, § 47), and
haplologies, by which one of two succeeding similar
sounds or sound-groups is discarded as in the
pronunciation [wustə] for Worcester, in England for
Englaland, in simply for simplely, in the familiar or
vulgar pronunciations of library, February, probably,
literary, mama as [laibri, Febri, probli, litri, ma˙], etc.,
in Latin nutrix for nutritrix, stipendium for stipipendium,
tuli for tetuli, etc., in French controle for contrerôle,
idolâtre for idololâtre, Neuville for Neuveville,
colloquial [talœ˙r] for tout a l'heure, in Italian cosa
for che cosa, qualcosa for qualchecosa, etc., etc., 1269
and finally shortenings by subtraction, such as pea
for pease, adder for nadder (§ 50); besides these more
sporadically-occurring processes we find that a great
many of the constant phonetic changes of every
343language result in the shortening of words: vowels
in weak syllables are pronounced more and more
indistinctly and finally disappear altogether; final
consonants are dropped (as is perhaps best seen by a
comparison of the pronunciation and the spelling of
Modern French: the spelling will be found to retain a
great many sounds which were formerly pronounced);
initial consonants are often as unstable (see, for
instance, Engl. kn, gn and wr, where the k, g and w,
were formerly sounded), and in the middle of words
assimilation and other causes lead to similar results.
Every student of historical linguistics is familiar with
numerous examples of seemingly violent contractions,
which have really been wrought by regular and
gradual changes continued through centuries: lord,
with its three or four sounds, was formerly laverd, and
in Old English hlaford; nay, the Old Germanic form
of the same word contained indubitably as many as
twelve sounds; Latin augustus has in French through
aoust become août, which now consists only of two
sounds [au], or, in a very widely spread pronunciation, of
only one sound [u]; Latin oculus (oculum) has shrunk
into four sounds in Italian occhio, three in Spanish ojo,
and two in French œil. These are everyday occurrences,
while lengthenings of words (as in English
sound from Fr. son, M. E. son, soun) are extremely rare.
The ancient languages of our family, Sanskrit, Zend,
etc., abound in very long words; the further back we
go, the greater the number of sesquipedalia. This fact
inspires us with distrust of the current theory, according
344to which every language started with monosyllabic
roots; even the rare agreement on this point of two
otherwise such fierce opponents as Professors Max
Müller and Whitney cannot make us accept the
theory; and the bull of excommunication issued by
the latter 1270 must not deter us from the heresy of
saying: If the development of language took the
same course in pre-historic as in historic times — and
there is no reason to doubt it — then we must imagine
primitive language as consisting (chiefly at least) of
very long words, containing many difficult sounds,
and sung rather than spoken.

III. Grammar.

268. Can anything be stated about the grammar
of primitive language? Yes, I think so, if we continue
backwards into the past the lines of evolution resulting
from the investigations contained in the preceding
chapters. Ancient languages have several forms
where modern languages content themselves with
fewer; forms originally kept distinct are in course of
time confused, either through a phonetic obliteration
of differences in the endings, or through analogical
extension of the functional sphere of one form; the
single form good is now used where O. E. used the
345forms god, godne, gode, godum, godes, godre, godra, goda,
godan, godena; Ital. uomo or French homme corresponds
to Lat. homo, hominem, homini, homine. Where
the modern language has one or two cases, in an earlier
stage it had three or four, and still earlier even seven
or eight. The same thing is seen in the flexion of
the verb: an extreme, but by no means unique
example, is the English cut, which can serve both as
present and past tense, both as singular and plural,
both in the first, second and third persons, both in
the infinitive, in the imperative, in the indicative, in
the subjunctive, and as a past participle; compare
with this the old languages with their separate forms
for different tenses and moods, for two or three
numbers, and in each for three persons; and remember,
moreover, that the identical form, without
any inconvenience being occasioned, is also used as a
noun (a cut), and you will admire the economy of the
living tongue. A characteristic feature of the structure
of languages in their early stages is that each
form of a word (whether verb or noun) contains in
itself several minor modifications which, in the later
stages of the language, are expressed separately, e.g.,
by auxiliary verbs or prepositions. Such a word as
Latin cantavisset, unites into one inseparable whole
the equivalents of six ideas: (1) “sing,” (2) pluperfect,
(3) that indefinite modification of the verbal idea
which we term subjunctive, (4) active, (5) third person,
and (6) singular.

269. These general tendencies of the later stages
346of language may be properly denoted by the term
“analysis”. If, however, we accepted “synthesis”
as the designation of the earliest stage we should
be guilty of inconsistency: for as synthesis means
composition, putting together, it presupposes that
the elements “put together” had at first an independent
existence; and this we deny. Therefore,
whoever does not share the usual opinion that all
flexional forms have originated through independent
words gradually coalescing, but sees that we have
sometimes to deal with the reverse process of inseparable
parts of words gradually gaining independence
(§§ 50, 57, 246 ff.), will have to look out for a better
or less ambiguous word than synthesis for the condition
of primitive speech. What in the later stages
of language is analysed or dissolved, in the earlier
stages was unanalysable or indissoluble; “entangled”
or “complicated” would therefore be better renderings
of our impression of the first state of things.
In Latin homini nobody is able to see where the
designation of “man” ceases, or which element
signifies the dative case and which the singular
number.

270. The direction of movement is towards
flexionless languages (such as Chinese, or to a certain
extent Modern English) with freely combinable elements;
the starting-point was flexional languages
(such as Latin or Greek); at a still earlier stage we
must suppose a language in which a verbal form
might indicate not only six things like cantavisset,
347but a still larger number, in which verbs were perhaps
modified according to the gender (or sex) of the
subject, as they are in Semitic languages, or according
to the object, as they are in some American Indian
languages. But that amounts to the same thing as
saying that the borderline between word and sentence
was not so clearly defined as in more recent times;
cantavisset is really nothing but a sentence-word, and
the same holds true to a still greater extent of the
sound-conglomerations of Indian languages. It is,
indeed, highly characteristic of the primitive mind,
and a subject of constant astonishment to those who
study the languages of savage races, that a thing by
itself cannot be conceived or spoken of: it is an
utter impossibility for a savage to think of “knife,”
for instance, by itself; his power of abstraction
is not sufficiently developed; but he can perfectly
well say, “give me that knife,” or “he plunged the
knife into the hart”. It will be noticed that in
speaking of “sentence-words” as the original units
of language I do not use that expression in exactly
the same sense as certain linguistic writers, who
exemplify their notion of primitive sentence-words
by such modern instances as “Fire!” or “Thief!”
In my opinion primitive linguistic units must have
been much more complicated in point of meaning, as
well as much longer in point of sound.

271. Another point of great importance is this:
in early languages we find a far greater number of
irregularities, exceptions, anomalies, than in modern
348ones. It is true that we not unfrequently see new irregularities
spring up, where the formations were
formerly regular; but these instances are very far
from counterbalancing the opposite class in which
words once irregularly inflected become regular, or
anomalies in syntax, etc., are levelled. The tendency
is more and more to denote the same thing by the
same means in every case, to extend the ending, or
whatever it is, that is used in a large class of words
to express a certain modification of the central idea,
until it is used in all other words as well.

Primitive language no doubt had a superabundance
of irregularities and anomalies, in syntax and word-formation
not less than in accidence. It was capricious
and fanciful, and displayed a luxuriant growth
of forms, entangled one with another like the trees
in a primeval forest. Human minds of those times
disported themselves in these long and intricate words
as in the wildest and most wanton play. Primitive
speech was certainly not, as is often supposed, 1271
distinguished for logical consistency; nor, so far as
we can judge, was it simple and facile: it is much
more likely to have been extremely clumsy and unwieldy.
Renan rightly reminds us of Turgot's wise
saying: “Des hommes grossiers ne font rien de
simple. Il faut des hommes perfectionnés pour y
arriver.”349

IV. Vocabulary.

272. If we turn to the inner side of language,
that is, to the meaning connected with the words, we
shall find a development parallel to that noticed in
grammar; and indeed, if we go deep enough into the
question, we shall see that it is really the very same
movement that has taken place here. The more
advanced a language is, the more developed is its
power of expressing abstract things. I use this term
“abstract” not in the narrow sense of some logicians,
who make it cover only such words as “whiteness”
or “love”; but in a wider sense, so as to denote also
the so-called general terms. Everywhere language
has first attained to expressions for the concrete and
special. In accounts of barbaric people's languages
we incessantly meet with such phrases as these:
“The aborigines of Tasmania had no words representing
abstract ideas; for each variety of gum-tree and
wattle-tree, etc., etc., they had a name; but they had
no equivalent for the expression ‘a tree’; neither
could they express abstract qualities, such as ‘hard,
soft, warm, cold, long, short, round’”; or, “The
Mohicans have words for cutting various objects, but
none to convey cutting simply; and the Society
Islanders can talk of a dog's tail, a sheep's tail, or a
man's tail [?], but not of tail itself. The dialect of
the Zulus is rich in nouns denoting different objects
of the same genus, according to some variety of
colour, redundancy, or deficiency of members, or
some other peculiarity, such as ‘red cow,’ ‘white
350cow,’ ‘brown cow,’ etc.” 1272 Some languages have no
word for brother, but only for “elder brother” and
“younger brother”; others can only express “hand”
as being either “my hand” or “your hand” or “his
hand,” and so on. In Cherokee, instead of one word
for “washing” we find different words, according to
what is washed: ku-tuwo, “I wash myself”; ku-lestula,
“I wash my head”; tsestula, “I wash the head of
somebody else”; kukuswo, “I wash my face”;
tsekuswo, “I wash the face of somebody else”; takasula,
“I wash my hands or feet”; takunkela, “I wash
my clothes”; takutega, “I wash dishes”; tsejuwu,
“I wash a child”; kowela, “I wash meat”. Many
savage tribes possess specific appellations for a
number of shades of relationship which we can only
express by a combination of two or three words, etc.,
etc.

In old Germanic poetry we find an astonishing
abundance of words translated in our dictionaries by
“sea,” “battle,” “sword,” “hero,” and the like:
these may certainly be considered as remains of an
earlier condition of things, in which each of these
words at present only differing in form had its separate
shade of meaning, which was subsequently lost. 2273
The nomenclature of a remote past was undoubtedly
351constructed upon similar principles to those which we
still come across in a word-group like horse, mare,
stallion, foal, colt, instead of he-horse, she-horse,
young horse, etc. So far, then, primitive speech had
a larger vocabulary than later languages.

273. While our words are better adapted to express
abstract things and to render concrete things with
definite precision, they are comparatively colourless.
The old words, on the contrary, spoke more immediately
to the senses, they were manifestly more suggestive,
more graphic and pictorial; while to express
one single thing we are not unfrequently obliged to
piece the image together bit by bit, the old concrete
words would at once present it to the hearer's mind as
an indissoluble whole; they were, accordingly, better
adapted to poetic purposes. Nor is this the only way
in which we see a close relationship between primitive
words and poetry.

274. If we try mentally to transport ourselves to a
period in which language consisted of nothing but
such graphic concrete words, we shall discover that,
in spite of their number, even if taken all together, they
would not suffice to cover everything which needed
expression; a wealth in such words is not incompatible
with a certain poverty. Words will accordingly
often be required to do service outside of their proper
sphere of application. That a figurative or metaphorical
use of words is a factor of the utmost importance
in the life of all languages, is a well-known fact;
but I am probably right in thinking it played a more
352prominent part in old times than now. In course of
time a great many metaphors have become stiffened
and worn out, so that nobody feels them to be metaphors
any longer. Examine closely such a sentence
as this: “He came to look upon the low ebb of morals
as an outcome of bad taste,” and you will find that
nearly every word is a dead metaphor. 1274 But the
better stocked a language is with those ex-metaphors
which have now become regular expressions for definite
ideas, the less need is there for going out of
your way to find new metaphors. The expression
of thought tends therefore to become more and more
mechanical or prosaic.

V. Conclusion.

275. If now we try to sum up what has been inferred
about primitive speech, we see that by our
backward march we arrived at a language whose
units had a very meagre substance of thought, and
this as specialised and concrete as possible; but at
the same time the phonetic body was ample; and
the bigger and longer the words, the thinner the
thoughts! Much cry and little sense! No period
has seen less taciturn people than the first framers of
speech; primitive speakers were not reticent and
reserved beings, but youthful men babbling merrily
on, without being so very particular about the meaning
of each of their words. They did not narrowly
weigh every syllable, — what were a couple of syllables
more or less to them? They chattered away for the
mere pleasure of chattering, resembling therein many
a mother of our own times who will chatter away to
baby without measuring her words or looking too
closely into the meaning of each; nay, who does not
care a bit for the consideration that the little deary
does not understand a single word of her affectionate
eloquence, and perhaps is not even able to hear it.
But primitive speech — and we return here to an idea
354thrown out above — still more resembles the speech
of little baby himself, before he begins to listen
properly to the words of grown-up people and to
frame his own language after the pattern of theirs;
the language of our remote forefathers was like that
ceaseless humming and crooning with which no
thoughts are as yet connected, which merely amuses
and delights the little one. As Preyer has it, there
is a period in the life of a child when his tongue is
his dearest toy and best plaything. 1276 Language
originated as play, and the organs of speech were
first trained in this singing sport of idle hours.

Which of the two, hunger or love, was the more
powerful in producing germs of speech? To be sure,
it was not hunger or that which is connected with
hunger: mere individual self-assertion and the struggle
for material existence. This prosaic side of life has
355only been capable of calling forth short monosyllabic
interjections, howls of pain and grunts of satisfaction;
but these are isolated and not capable of much further
development; they are the most immutable portions
of language, and remain now on essentially the same
stand-point as thousands of years ago.

277. It is quite otherwise with love; as far as I
see, linguistic considerations and generalisations point
towards essentially the same source of language as
that which Darwin arrived at by other paths: the
effort to charm the other sex. To the feeling of love,
which has left traces of its vast influence on countless
points of the evolution of organic nature, are due not
only the magnificent colours of birds and flowers:
it inspired the first songs, and through them gave
birth to human language as well.

278. If after spending some time over the deep
metaphysical speculations of German linguistic philosophers
you turn to men like Madvig or Whitney, you
are at once agreeably impressed by the sobriety
of their reasoning and their superior clearness of
thought; but if you look more closely, you cannot
help thinking that they imagine our primitive ancestors
after their own image as serious and well-meaning
men endowed with a large share of common-sense.
By their laying such great stress on the communication
of thought as the end of language and on the
usefulness to primitive man of being able to speak to
his fellow-creatures about matters of vital importance,
they leave you with the impression that these “first
356framers of speech” were sedate, alderman-like citizens,
with a prominent sense for the purely business and
matter-of-fact side of life; indeed, according to Madvig,
women had no share in the creating of language.
Speech seems chiefly to have been instituted as a
vehicle of important communications and judicious
reasonings.

279. In opposition to this rationalistic view I
should like, for once in a way, to bring into the field
the opposite view: the genesis of language is not to be
sought in the prosaic, but in the poetic side of life;
the source of speech is not gloomy seriousness, but
merry play and youthful hilarity: in primitive speech
I hear the laughing cries of exultation when lads and
lasses vied with one another to attract the attention
of the other sex, when everybody sang his merriest
and danced his bravest to lure a pair of eyes to throw
admiring glances in his direction. Language was
born in the courting days of mankind: the first
utterance of speech I fancy to myself like something
between the nightly love lyrics of puss upon the tiles
and the melodious love songs of the nightingale.

280. Strong, however, as must have been the
influence of love, it was not the only feeling which
tended to call forth primitive songs. 1277 Any strong
357emotion, and, more particularly, any pleasurable
excitement, will result in song. Singing, like any
other sort of play, is due to an overflow of energy,
which is discharged in “unusual vivacity of every
kind, including vocal vivacity”. Out of the full
heart the mouth sings! Mr. Spencer has a good
many quotations to the effect that savages will sing
whenever they are excited: exploits of war or of
the chase, the deeds of their ancestors, the coming of
a fat dog, any incident, “from the arrival of a stranger
to an earthquake,” is turned into a song; and most
of these songs are composed extempore. “When
rowing, the Coast-negroes sing either a description
of some love intrigue or the praise of some woman
celebrated for her beauty. In Loango the women
as they till the field make it echo with their rustic
songs.” Park says of the Bambarran: “They lightened
their labours by songs, one of which was composed
extempore, for I was myself the subject of it”. In
some parts of Africa nothing is done except to the
sound of music. They are very expert in adapting
the subjects of these songs to current events. The
358Malays amuse all their leisure hours with the repetition
of songs, etc. One of Mr. Spencer's quotations
aptly illustrates the way in which primitive
men, as I fancy, struck up their songs long before
language was developed for the communication of
ideas: “In singing, the East African contents himself
with improvising a few words without sense or rhyme
and repeats them till they nauseate”.

Nor is this sort of singing on every and any occasion
confined to savages; it is found wherever the
in-door life of civilisation has not killed out open-air
hilarity; formerly in our Western Europe people
sang much more than they do now. The Swedish
peasant Jonas Stolt, writing about 1820, says: “I have
lived in a time when young people were singing from
morning till eve. Then they were carolling both out- and
in-doors, behind the plough as well as at the
threshing-floor and at the spinning-wheel. This is all
over long ago: now-a-days there is silence everywhere;
if some one were to try and sing in our days as we
did of old, people would term it bawling.” 1278

281. The first things that were thus expressed in
song were, to be sure, neither deep nor wise; how
could you expect it? Even now the thoughts
associated with singing are generally neither very
clear nor very clever; like humming or whistling,
singing is often nothing more than an almost
automatic expression of a mood; “and what is
359not worth saying can be sung”. Besides, it has been
the case at all times that things transient and trivial
have been readier to find expression than Socratic
wisdom. But the frivolous use ground the instrument,
and rendered it little by little more serviceable
to a multiplicity of purposes, so that it became more
and more fitted to express everything that touched
human souls.

282. Men sang out their feelings long before they
were able to speak their thoughts. But they did
not originally sing in order to communicate their
ideas or feelings; in fact, they had not the slightest
notion that such a thing was possible. They “sang
but as the linnet sings” — this word is truer of primitive
men and women than ever it was of the late poet
laureate. They little suspected that in singing as
nature prompted them, they were paving the way for
a language capable of rendering minute shades of
thought; just as they could not suspect that out of
their coarse pictures of men and animals there should
one day grow an art enabling men of distant countries
to speak to each other. As is the art of writing to
primitive painting, so is the art of speaking to primitive
singing. And the development of the two vehicles of
communication of thought present other curious and
instructive parallels. In primitive picture-writing,
each sign meant a whole sentence or even more —
the image of a situation or of an incident being given
as a whole — ; this developed into an ideographic
writing of each word by itself; this system was
360succeeded by syllabic methods, which had in their
turn to give place to alphabetic writing, in which each
letter stands for, or is meant to stand for, one sound.
Just as here the advance is due to a further analysis
of language, smaller and smaller units of speech
being progressively represented by single signs, in an
exactly similar way, though not quite so unmistakably,
the history of language shows us a progressive
tendency towards analysing into smaller and smaller
units that which in the earlier stages was taken as
an inseparable whole.

283. While an onomatopoetic or echo-word like
bow-wow and an interjection like pooh-pooh were at once
employed and understood as signs for the corresponding
idea, this was not the case with the great bulk
of language. Just as we have seen above with regard
to many details of grammatical structure 1279 that by
indirect and round-about ways they acquired other
meanings than they had had originally, or acquired
meanings where they had originally had none, so it
was also with language at large. Originally a jingle
of empty sounds without meaning, it came to be
an instrument of thought. If man is, as Humboldt
has somewhere defined him, “a singing creature, only
associating thoughts with the tones,” we must answer
the question: How did this association of sense and
sound come about? I think we can arrive at forming
361some idea of that process by remembering what has
been said above on the signification of primitive words.
This we must imagine to have been concrete and
special in the highest degree. There are, however,
no words whose signification is so concrete and
special as proper names, — not such proper names as
our modern John or Jones or Smith, which have become
so common as to be scarcely proper names any
longer; but proper names of the good old kind, borne
by and denoting only one single individual. How
easily might not such names spring up in a primitive
state such as that described above! In the songs of
a particular individual there would be a constant recurrence
of a particular series of sounds sung with
a particular cadence; no one can doubt the possibility
of such individual habits being contracted in olden as
well as in present times. Suppose, then, that “In the
spring time, the only pretty ring time,” a lover was
in the habit of addressing his lass “With a hey and
a ho, and a hey nonino!” his comrades and rivals
would not fail to remark this, and would occasionally
banter him by imitating and repeating his “hey-and-a-ho-and-a-hey-nonino”.
But when once this had
been recognised as what Wagner would term a person's
“leitmotiv,” it would be no far cry from mimicking
it to using the “hey-and-a-ho-and-a-hey-nonino”
as a sort of nick-name for the man concerned; it
might be employed, for instance, to signal his arrival.
But when once proper names were given, common
names (or nouns) would not be slow in following;
362we see the transition from one to the other class constantly
going on, names originally used exclusively
to denote an individual being used metaphorically to
connote that individual's most characteristic peculiarities,
as when we say of one man that he is “a Croesus”
or “a Vanderbilt,” and of another that he is “no
Bismarck”. We may also remind the reader of the
German schoolboy who stated in his history lesson
that Hannibal swore he would always be a Frenchman
to the Romans. 1280 This is, at least, one of the ways
by which language arrives at designations for such
ideas as “rich,” “statesman,” and “enemy”. Names
of tools are in some cases proper names, used originally
as some term of endearment, as when in
thieves’ slang a crowbar or lever is called a betty or
jemmy; English derrick, as well as the German and
Scandinavian word for a picklock (German, dietrich;
Dan., dirk; Swed., dyrk), is nothing but the proper
name Dietrich (Derrick, Theodoricus); compare also
the history of the words bluchers, jack (boot-jack, jack
for turning a spit, a pike, etc., also jacket), pantaloon,
hansom, to burke, to name only a few examples.

284. Again, we saw above that the further back
we went, the more the sentence was one indissoluble
whole, in which those elements which we are accustomed
to think of as single words were not yet
separated. But it is just sentences of this sort whose
genesis we can imagine with greatest ease on the
supposition of a primitive period of meaningless
363singing. If a certain number of people have together
witnessed some incident and have accompanied
it with some sort of impromptu song or
refrain, the two ideas are associated, and later on the
song will tend to call forth in the memory of those
who were present the idea of the whole situation.
Suppose some dreaded enemy has been defeated and
slain; the troop will dance round the dead body and
strike up a chant of triumph, say something like
“Tarara-boom-de-ay!” This combination of sounds,
sung to a certain melody, will now easily become
what might be called a proper name for that particular
event; it might be roughly translated, “The
terrible foe from beyond the river is slain,” or “We
have killed the dreadful man from beyond the river,”
or “Do you remember when we killed him?” or
something of the same sort. Under slightly altered
circumstances it may become the proper name of the
man who slew the enemy. The development can
now proceed further by a metaphorical transference
of the expression to similar situations (“There is
another man of the same tribe: let us slay him as we
did the other!”); or by a blending of two or more
of these proper-name melodies. I can give nothing
but hints; but does not the reader begin now dimly
to see ways by which primitive “lieder ohne worte”
may have become, first, indissoluble sentences, and
then gradually combinations of words more and more
capable of being analysed? And does not this
theory explain better than most others the great part
364which chance and fortuitous coincidence always seem
to play in languages?

285. Language, then, began with half-musical
unanalysed expressions for individual beings and
events. Languages composed of such words and
sentences are clumsy and insufficient instruments of
thought, being intricate, capricious and difficult.
But from the beginning the tendency has been one of
progress, slow and fitful progress, but. still progress
towards greater and greater clearness, regularity,
ease and pliancy. No one language has arrived at
perfection; an ideal language would always express
the same thing by the same, and similar things by
similar means; any irregularity and ambiguity would
be banished; sound and sense would be in perfect
harmony ; any number of delicate shades of meaning
could be expressed with equal ease: poetry and
prose, beauty and truth, thinking and feeling would
be equally provided for: the human spirit would
have found a garment combining freedom and gracefulness,
fitting it closely and yet allowing full play to
any movement.

The end.365