Whitney, William. The Life and Growth of Language – T01

[Life and Growth of Language]

Chapter I.
Introductory : the problems of the science of language.

Definition of language. Man its universal and sole possessor. Variety
of languages. The study of language ; aim of this volume.

Language may be briefly and comprehensively defined
as the means of expression of human thought.

In a wider and freer sense, everything that bodies
forth thought and makes it apprehensible, in whatever
way, is called language ; and we say, properly enough,
that the men of the Middle Ages, for example, speak
to us by the great architectural works which they have
left behind them, and which tell us very plainly of
their genius, their piety, and their valor. But for
scientific purposes the term needs restriction, since it
would apply else to nearly all human action and product,
which, discloses the thought that gives it birth.
Language, then, signifies rather certain instrumentalities
whereby men consciously and with intention represent
their thought, to the end, chiefly, of making it
known to other men : it is expression for the sake of
communication.

The instrumentalities capable of being used for this
purpose, and actually more or less used, are various :
gesture and grimace, pictorial or written signs, and
1uttered or spoken signs : the first two addressed to the
eye, the last to the ear. The first is chiefly employed
by mutes — though not in its purity, inasmuch as these
unfortunates are wont to be trained and taught by
those who speak, and their visible signs are more or
less governed by habits born of utterance ; going even
so far as slavishly to represent the sounds of speech.
The second, though in its inception a free and independent
means of expression, yet in its historical development
becomes linked as a subordinate to speech, and
even finds in that subordination its highest perfection
and greatest usefulness. 11 The third is, as things actually
are in the world, infinitely the most important ; insomuch
that, in ordinary use, “language” means utterance,
and utterance only. And so we shall understand
it here : language, for the purposes of this discussion,
is the body of uttered and audible signs by which in
human society thought is principally expressed, gesture
and writing being its subordinates and auxiliaries. 22

Of such spoken and audible means of expression
no human community is found destitute. From the
highest races to the lowest, all men speak ; all are able
to interchange such thoughts as they have. Language,
then, appears clearly “natural” to man ; such are his
endowments, such his circumstances, such his history —
one or all of these — that it is his invariable possession.

Moreover, man is the sole possessor of language.
It is true that a certain degree of power of communication,
sufficient for the infinitely restricted needs of
their gregarious intercourse, is exhibited also by some
2of the lower animals. Thus, the dog's bark and howl
signify by their difference, and each by its various style
and tone, very different things ; the domestic fowl has
a song of quiet enjoyment of life, a clutter of excitement
and alarm, a cluck of maternal anticipation or
care, a cry of warning — and so on. But these are not ;
only greatly inferior in their degree to human language
they are also so radically diverse in kind from
it, that the same name cannot justly be applied to both.
Language is one of the most marked and conspicuous,
as well as fundamentally characteristic, of the faculties
of man.

Nevertheless, while human language is thus one as
contrasted with brute expression, it is in itself of a
variety which is fairly to be termed discordance. It is
a congeries of individual languages, separate bodies of
audible signs for thought, which, reckoning even those
alone of which the speakers are absolutely unintelligible
to one another, are very numerous. These languages
differ among themselves in every degree. Some
are so much alike that their users can with sufficient
trouble and care come to understand one another ; of
others, even a superficial examination shows abundant
correspondences ; of yet others, similar points of accordance
are rarer, and only discoverable by practised
study and research ; and a great number are to all appearance
wholly diverse — and often, not only diverse
in respect to the actual signs which they use for their
various conceptions, but also as to their whole structure,
the relations which they signify, the parts of
speech they recognize. And this diversity does not
accord with differences of intellectual capacity among
the speakers : individuals of every degree of gift are
found using, each according to his power, the same
3identical dialect ; and souls of kindred calibre in different
societies can hold no communion together. Nor
does it accord with geographical divisions ; nor yet, in
its limits and degrees, with the apparent limits of
races. Not seldom, far greater race-differences are met
with among the speakers of one language, or of one
body of resembling languages, than between those who
use dialects wholly unlike one another.

These, and their like, are the problems which occupy
the attention of those who pursue the science of
language, or linguistic science. That science strives to
comprehend language, both in its unity, as a means of
human expression and as distinguished from brute
communication, and in its internal variety, of material
and structure. It seeks to discover the cause of the
resemblances and differences of languages, and to effect
a classification of them, by tracing out the lines of resemblance,
and drawing the limits of difference. It
seeks to determine what language is in relation to
thought, and how it came to sustain this relation ;
what keeps up its life and what has kept it in existence
in past time, and even, if possible, how it came into
existence at all. It seeks to know what language is
worth to the mind, and what has been its part in the
development of our race. And, less directly, it seeks
to learn and set forth what it may of the history of human
development, and of the history of races, their
movements and connections, so far as these are to be
read in- the facts of language.

No reflecting and philosophizing people has ever
been blind to the exceeding interest of problems like
these, or has failed to offer some contribution toward
their solution. Yet the body of truth discovered in
earlier times has been so small, that the science of language
4is to be regarded as a modern one, as much so
as geology and chemistry ; it belongs, like them, to the
nineteenth century. To review its history is no part of
our present task ; no justice could be done the subject
within the space that could be spared it in this volume ;
and the few words that we can bestow upon it will be
better said in the last chapter than here. Although of
so recent growth, the science of language is already
one of the leading branches of modern inquiry. It is
not less comprehensive in its material, definite in its
aims, strict in its methods, and rich and fruitful in its
results, than its sister sciences. Its foundations have
been laid deep and strong in the thorough analysis of
many of the most important human tongues, and the
careful examination and classification of nearly all the
rest. It has yielded to the history of mankind as a
whole, and to that of the different races of men, definite
truths and far-reaching glimpses of truth which
could be won in no other way. It is bringing about a
re-cast of the old methods of teaching even familiar
and long-studied languages, like the Latin and Greek ;
it is drawing forward to conspicuous notice others of
which, only a few years ago, hardly the names were
known. It has, in short, leavened all the connected
branches of knowledge, and worked itself into the very
structure of modern thought, so that no one who hears
or reads can help taking some cognizance of it. No
educated person can afford to lack a clear conception
of at least a brief connected outline of a science possessing
such claims to attention.

The design of this volume, accordingly, is to draw
out and illustrate the principles of linguistic science,
and to set forth its results, with as much fullness as the
limited space at command shall allow. The study is
5not yet so developed and established as not to include
subjects respecting which opinions still differ widely
and deeply. But direct controversy will be avoided ;
and the attempt will be made to construct an argument
which shall commend itself to acceptance by the
coherence of its parts and the reasonableness of its
conclusions. In accordance with the plan of the series
of treatises into which this enters as a member, simplicity
and popular apprehensibility will be everywhere
aimed at. To start from obvious or familiar truths,
to exemplify by well-known facts, will be found, it is
believed, the best way to arrive with assurance at the
ultimate results sought after. The prime facts of language
lie, as it were, within the easy grasp of every
man who speaks — yet more, of every man who has
studied other languages than his own — and to direct
intelligent attention toward that which is essential, to
point out the general in the midst of the particular
and the fundamental underneath the superficial, in
matters of common knowledge, is a method of instruction
which cannot but bear good fruit.6

Chapter II.
Each individual acquires his language : life
of language.

Language learned, not inherited or made, by the individual ; process of
children's learning to speak ; what this involves, outside the province
of the linguistic student. Origin of particular words. Character
of a word as sign for a conception. Mental training in learning
language ; determination of the inner form of language from without ;
constraint and advantage in the process. Acquisition of a
second language, or of more than one ; learning even of native
speech a never-ending process. Imperfection of the word as sign ;
language only the apparatus of thought.

There can be asked respecting language no other
question of a more elementary and at the same time
of a more fundamentally important character than this :
how is language obtained by us ? how does each speaking
individual become possessed of his speech ? Its
true answer involves and determines well-nigh the
whole of linguistic philosophy.

There are probably few who would not at once reply
that we learn our language ; it is taught us by
those among whom our lot is cast in childhood. And
this obvious and common-sense answer is also, as we
shall find on a more careful and considerate inquiry,
the correct one. We have to look to see what is implied
in it.7

In the first place, it sets aside and denies two other
conceivable answers : that language is a race-characteristic,
and, as such, inherited from one's ancestry, along
with color, physical constitution, traits of character,
and the like ; and that it is independently produced by
each individual, in the natural course of his bodily and
mental growth.

Against both these excluded views of the acquisition
of language may be brought such an array of facts
so familiar and undeniable that they cannot be seriously
upheld. Against the theory of a language as a race characteristic
may be simply set, as sufficient rebutting
evidence, the existence of a community like the American,
where there are in abundance descendants of African,
of Irish, of German, of southern European, of
Asiatic, as well as of English ancestors, all using the
same dialect, without other variety than comes of differences
of locality and education, none showing a
trace of any other “mother-tongue” or “native
speech.” But the world is full of such cases, on the
small scale and on the large. Any child of parents living
in a foreign country grows up to speak the foreign
speech, unless carefully guarded from doing so ; or, it
speaks both this and the tongue of its parents, with
equal readiness. The children of missionary families
furnish the most striking examples of this class : no
matter where they may be in the world, among what
remotely kindred or wholly unrelated dialects, they acquire
the local speech as “naturally” as do the children
of the natives. And it is only necessary that the
child of English or German or Russian parents, born
in their native country, should (as is often done) be put
with a French nurse, and hear French alone spoken
about it, and it will grow up to speak French first and
8French only, just as if it were a French child. And
what is French, and who are its speakers ? The mass
of the people of France are Celts by descent, with,
characteristic Celtic traits which, no mixture or education
has been able to obliterate ; but there is hardly an
appreciable element of Celtic in the French language ;
this is almost purely a Romanic dialect, a modern representative
of the ancient Latin. There are few unmixed
languages in the world, as there are few unmixed
races ; but the one mixture does not at all determine
the other, or measure it. The English is a very striking
proof of this ; the preponderating French-Latin
element in our vocabulary gets its most familiar and
indispensable part from the Normans, a Germanic race,
who got it from the French, a Celtic race, who got it
from the Italians, among whom the Latin-speaking
community were at first a very insignificant element,
numerically. It is useless to bring up further examples ;
the force of those here given will be sufficiently
supported by our later inquiry into the actual processes
of acquisition of language.

So far as the other theory, that of independent production
by each person of his own speech, implies that
each inherits from his ancestors a physical constitution
which makes him develop unconsciously the same
speech as theirs, it is virtually coincident with the first
theory, and the same facts tell with crushing weight
against it ; so far as it is meant to imply that there is
a general likeness in intellectual constitution between
members of the same community which leads them to
frame accordant systems of expression, it is equally
without support from facts ; for the distribution of
human dialects is as irreconcilable with that of natural
capacity and bent as with that of physical form among
9human beings. Every variety of gift is found among
those who employ, each with his own degree of skill
and capacity, the same speech ; and souls of commensurate
calibre in different communities are unable to
have intercourse together.

We come, then, to consider directly the process by
which the child becomes able to speak a certain language — a
process sufficiently under every one's observation
to allow of general and competent criticism of
any attempted description of it. We cannot, it is true,
follow with entire comprehension all the steps of evolution
of the infantile and childish powers ; but we. can
understand them well enough for our purpose.

The first thing which the child has to learn, before
speech is possible, is to observe and distinguish ; to
recognize the persons and things about him, in their
concrete individuality, and to notice as belonging to
them some of their characteristic qualities and acts.
This is a very brief description of a very intricate psychological
process — which, however, it does not belong
to the student of language to draw out in greater detail.
There is involved in it, we may further remark
in passing, nothing which some of the lower animals
may not achieve. At the same time, the child is exercising
his organs of utterance, and gaining conscious
command of them, partly by a mere native impulse to
the exertion of all his native powers, partly by imitation
of the sound-making persons about him : the child
brought up in solitude would be comparatively silent.
This physical process is quite analogous with the training
of the hands : for some six months the child tosses
them about, he knows not how or why ; then he begins
to notice them and work them under command, till at
length he can do by conscious volition whatever is
10within their power. Control and management of the
organs of utterance comes much more slowly ; but the
time arrives when the child can imitate at least some
of the audible as well as the visible acts of others ; can
reproduce a given sound, as a given gesture. But before
this, he has learned to associate with some of the
objects familiar to him the names by which they are
called ; a result of much putting of the two together
on the part of his instructors. Here is seen more
markedly, at least in degree, the superiority of human
endowment. The association in question is doubtless
at the outset no easy thing, even for the child ; he does
not readily catch the idea that a set of sounds belongs
to and represents a thing — any more than, when older,
the idea that a series of written characters represents a
word ; but their connection is set so often and so distinctly
before him as to be learned at last, just as the
connection is learned between sugar and pleasure to
the taste, between a rod and retribution for misbehavior.
And every child begins to know things by
their names long before he begins to call them. The
next step is to imitate and reproduce the familiar name,
usually at first in the most imperfect way, by a mere
hint of the true sound, intelligible only to the child's
constant attendants ; and when that step is taken, then
for the first time is made a real beginning of the acquisition
of language.

Though not all children start with the acquisition
of precisely the same words, yet their limit of variety
is but a narrow one. We may take as fair examples
of at least the very early ones the childish names for
‘father’ and ‘mother,’ namely papa and mamma, and
the words water, milk, good. And we have to notice
especially both how wholly external is the process
11which makes the child connect these particular names
with their respective ideas, and how empirical and imperfect
are the ideas themselves. What is really implied
in papa and mamma, the child does not in the
least know ; to him they are only signs for certain loving
and caring individuals, distinguished most conspicuously
by differences of dress ; and the chance is
(and it not seldom chances) that he will give the same
names to other individuals showing like differences ;
the real relation of male and female parent to child he
comes to comprehend only much later — not to speak
of the physiological mysteries involved in it, which no
man yet comprehends. As little does he understand
the real nature of water and milk ; he knows no more
than that, among the liquids (that name, to be sure,
comes much later, but not till long after the child has
realized the distinction of liquid and solid) constantly
brought before him there are two which he readily distinguishes,
by look and by taste, and to which other
people give these names ; and he follows their example.
The names are provisional, convenient nuclei for the
gathering of more knowledge about ; where the liquids
come from will be learned by and by, and their chemical
constitution, perhaps, in due time. As for good,
the first association of the term is probably with what
has a pleasant taste ; then what is otherwise agreeable
comes to be comprehended under the same name ; it
gets applied to behavior which is agreeable to the parents,
as judged by a standard which the child himself
is far from understanding — and this transfer to a moral
sphere is by no means an easy one ; as he grows up,
the child is (perhaps) all the time learning to distinguish
more accurately between good and bad ; but he
is likely to be at the last baffled by finding that the
12wisest heads in the world have been and are irreconcilably
at variance as to what good really means —
whether it implies only utility, or an independent and
absolute principle.

These are only typical examples, fairly illustrating
the whole process of speech-getting. The child begins
as a learner, and he continues such. There is continually
in presence of his intellect more and better than
he can grasp. By words lie is made to form dim conceptions,
and draw rude distinctions, which after experience
shall make truer and more distinct, shall
deepen, explain, correct. He has no time to be original ;
far more rapidly than his crude and confused impressions
can crystallize independently into shape, they
are, under the example and instruction of others, centred
and shaped about certain definite points. So it
goes on indefinitely. The young mind is always learning
words, and things through words ; in all other cases
as really, if not so obviously, as when, by description
and picture or by map and plan, it is led to form some
inaccurate half-conception of the animal lion or the
city Peking. The formal distinctions made by the inflectional
system of even so simple a language as English,
and by words of relation, are at first out of the
child's reach. He can grasp and wield only the grosser
elements of speech. He does not apprehend the relation
of one and more than one clearly enough to use
the two numbers of nouns ; the singular has to do duty
for both ; and so also the root-form of the verb, to the
neglect of persons, tenses, and moods. It is an era in
his education when he first begins to employ preterits
and plurals and their like. So with the pronouns. He
is slow to catch the trick of those shifting names, applied
to persons according as they are speaking, spoken
13to, or spoken of ; he does not see why each should not
have an own name, given alike in all situations : and he
speaks of himself and others by such a name and such
only, or blunders sorely in trying to do otherwise —
till time and practice set him right. 13 Thus, in every
respect, language is the expression of matured and
practised thought, and the young learner enters into
the use of it as fast as natural capacity and favoring
circumstances enable him to do so. Others have observed,
and classified, and abstracted ; he only reaps
the fruit of their labors. It is precisely as when the
child studies mathematics ; he goes over and appropriates,
step by step, what others have wrought out, by
means of word and sign and symbol ; and he thus
masters in a few years what it has taken generations
and ages to produce, what his unaided intellect could
never have produced ; what, perhaps, he could never
independently have produced a single item of, having
just mental force enough to follow and acquire it :
though also, perhaps, he has capacity to increase it by
and by, adding something new for those to learn who
come after him — even as the once educated speaker
may come to add, in one way and another (as will be
pointed out later), new stores of expression to language.

In all this, now, is involved infinitely more than
linguistic science has any call to deal with and explain.
Let us consider, for example, the word green. Its presence
in our vocabulary implies first the physical cause
of the color, wherein is involved the whole theory of
optics : and this concerns the physicist ; it is for him
to talk of the ether and its vibrations, and of the frequency
14and length of the waves which produce the
sensation of greenness. Then there is the structure
of the eye : its wondrous and mysterious sensitiveness
to just this kind of vibration, the apparatus of nerves
which conveys the impression to the brain, the cerebral
structure which receives the impression : to treat
of all this is the duty of the physiologist. His domain
borders and overlaps that of the psychologist, who has
to tell us what he can of the intuition and resulting
conception, considered as mode and product of mental
action, of the power of apprehension and distinction
and abstraction, and of the sway of consciousness over
the whole. Then, in the hearing of the word green is
involved the wonderful power of audition, closely akin
with that of vision : another sensitive apparatus, which
notes and reports another set of vibratory waves, in
another vibrating medium : it falls, like vision, into the
hands of the physicist and physiologist. They, too,
have to do with the organs of utterance, which produce
the audible vibration ; with their obedience to the directions
of the will : directions given but not executed
under the review of consciousness, and implying that
control of the mind over the muscular apparatus of the
body which is by no means the least of mysteries. We
might go on indefinitely thus, noticing what is included
in the simplest linguistic act ; and behind all would lie
as a background the great mystery of existence and its
cause, which no philosophy has yet been able to do
more than recognize. Every part of this is of interest
and importance to the linguistic scholar, but each in
its own way and degree ; and his specific and central
business is with none of it, but rather with something
else. This, namely : there exists an uttered and audible
sign, green, by which, in a certain community, are
15designated a certain class of kindred shades among the
infinitely varied hues of nature and of art ; and every
person who, by birth or by immigration or as a visitor
(a bodily visitor, or only a mental one, as student of its
literature), comes into the community in question,
learns to associate that sign with the given group of
shades, and to understand and employ it as designating
them ; and he learns to classify the infinity of hues
under that and certain other signs, of like nature and
use. About this pivotal fact all the other matters involved
fall into position as more or less nearly auxiliary ;
from it as point of view they are judged and have
their value estimated. Language, both in its single
items and as a whole, is primarily the sign of the idea,
the sign with its accompanying idea ; and to take any
other department of the questions involved as the central
one is to throw the whole into a false position, distorting
the proportions and relations of every part.
And, as the science of language seeks after causes, endeavors
to explain the facts of language, the primary
inquiry respecting this fact is : how came this sign to
be thus used ? what is the history of its production and
application ? and even, what is its ultimate origin and
the reason of it ? provided we can reach so far.

For there is, recognizably and traceably, a time
when and a reason why many of our words came into
use as signs for the ideas they represent. For example,
a certain other shade of color, a peculiar red, was
produced (with more, of its kind) not many years ago,
as result of the chemical manipulation of coal tar, and
was, reflectively and artificially, called by its inventor
magenta, after the name of a place which a great battle
had recently made famous. The word magenta is just
as real and legitimate a part of the English language
16as green, though vastly younger and less important ;
and those who acquire and use the latter do so in precisely
the same manner as the former, and generally
with equal ignorance and unconcern as to its origin.
The word gas is of much longer standing and wider use
with us, and has its respectable family of derivatives
and compounds — as gaseous, gasify, gas-pipe — and even
its colloquial figurative uses — as when we call an empty
and sophistical but ready talker gassy ; but it was the
wholly arbitrary invention of a Dutch chemist (Van
Helmont), about A. D. 1600. Science was at that time
getting so far along as to begin to form the distinct
conception of an aëriform or gaseous condition of existence
of matter, and this name chanced to be introduced
and supported in a way that commended it to
general acceptance ; and so it became the name, and
for all Europe. The young now for the most part
know it first as the title of a certain kind of gas, made
practically useful in giving light ; but by and by, if
fairly educated, they are led in connection with the
word to form for themselves the scientific idea of which
this is the sign. To trace the history of these two vocables
is to inform ourselves as to the time and the circumstances
of production of the aniline colors, and as
to the taking of a certain important step forward in
scientific thinking. We cannot follow so clearly toward
or to its source the word green, because it is vastly
older, reaching back far beyond the period of literary
record ; but we do seem to arrive by inference at a
connection of it with our word grow, and at seeing
that a green thing was named from its being a growing
thing ; and this is a matter of no small interest as bearing
on the history of the word.

It is not the place here to follow up this line of inquiries,
17and see what is meant by etymologizing, or
tracing the history of words toward their origin ; the
subject is one which will occupy us more properly later.
We touch it in passing merely in order to note that the
reason of first attribution of a sign to its specific use
is one thing, and that the reason of its after employment
in that use is another and a very different thing.
To the child learning to speak, all signs are in themselves
equally good for all things ; he could acquire
and reproduce one as well as another for a given purpose.
In fact, children in different communities do
learn every possible variety of names for the same
thing : instead of green, the German child learns grün,
the Dutch groen, the Swedish grön — all related to our
green, yet not identical with it ; and the French child
learns vert, the Spanish verde, the Italian viride — a similar
group of related yet diverse names ; while the Russian
says zelenüi, the Hungarian zöld, the Turk ishil,
the Arab akhsar, and so on. Each of these, and of hundreds
of others, is obtained in the same way : the child
hears it uttered by those about him under such circumstances
as make plain to him what it signifies ; by its
aid he in part learns to abstract the quality of color
from the colored object and conceive it separately ; and
he learns to combine in one comprehensive conception
the different shades of green, distinguishing them together
from the other colors, as blue and yellow, into
which they pass by insensible gradations. The learner
grasps the conception, at least in a measure, and then
associates his own word with it by a purely external tie,
having been able, if so guided, to form the same association
with any other existing or possible word, and
not less easily and surely. An internal and necessary
tie between word and idea is absolutely non-existent for
18him ; and whatever historical reason there may be is
also non-existent to his sense. He may sometimes ask
“what for ?” about a word, as he does, in his childish
curiosity, about everything else ; but it makes no difference
with the young etymologist (any more than with
the older one) what answer he gets, or whether he gets
an answer ; to him, the sole and sufficient reason why he
should use this particular sign is that it is used by those
about him. In the true and proper meaning of the terms,
then, every word handed down in every human language
is an arbitrary and conventional sign : arbitrary, because
any one of the thousand other words current among men,
or of the tens of thousands which might be fabricated,
could have been equally well learned and applied to
this particular purpose ; conventional, because the reason
for the use of this rather than another lies solely in
the fact that it is already used in the community to
which the speaker belongs. The word exists θέσει, ‘by
attribution,’ and not φύσει, ‘by nature,’ in the sense
that there is, either in the nature of things in general,
or in the nature of the individual speaker who uses it,
any reason that prescribes and determines it.

There is obviously mental training and shaping, as
well as mental equipment, in the process of learning to
speak. The mental action of the individual is schooled
into certain habits, consonant with those of his community ;
he acquires the current classifications and abstractions
and ways of looking at things. To take an
example : the quality of color is so conspicuous, and
our apprehension of it so urged by the infinity of its
manifested differences which are ever before our eyes,
that the conception of color is only quickened and rendered
more distinct by acquisition of the words which
denote it. But in the classification of the shades of hue
19the phraseology of the language acquired bears a determining
part ; they fall into order under and about the
leading names, as white, black, red, blue, green ; and
each hue is tested in the mind by aid of these, and referred
to the one or the other class. And different
languages make different classifications : some of them
so unlike ours, so much less elaborate and complete,
that their acquisition gives the eye and mind a very
inferior training in distinguishing colors. This is still
more strikingly the case as regards number. There
are dialects which are in a state of infantile bewilderment
before the problem of numeration ; they have
words for ‘one,’ ‘two,’ and ‘three ;’ but all beyond is
an undivided ‘many.’ None of us, it is tolerably certain,
would ever have gone farther than that by his
own absolutely unassisted efforts ; but by words — and
only by words ; for such is the abstractness of the relations
of number that they, more than any others, are
dependent for their realization and manageableness on
expression — more and more intricate numerical relations
have been mastered by us, until finally we are
provided with a system which is extensible to every
thing short of infinity — the decimal system, namely, or
that which proceeds by constant additions of ten individuals
of any given denomination to form the next
higher. And what is the foundation of this system ?
Why, as every one knows, the simple fact that we have
ten fingers (“digits”) on our two hands ; and that fingers
are the handiest substitutes for figures, the most
ready and natural of aids to an unready reckoner. A
fact as external and physical as this, and seemingly so
trivial, has shaped the whole science of mathematics,
and, altogether without his being aware of it, gives
form to all the numerical conceptions of each new
20learner. It is a suggestion of general human experience
in the past, transmitted through language into a
law for the government of thought in the future.

The same, in varying way and measure, is true of
every part of language. All through the world of
matter and of mind, our predecessors, with such wisdom
as they had at command, have gone observing, deducing,
and classifying ; and we inherit in and through
language the results of their wisdom. So with the distinctions
of living and lifeless ; of animal and vegetable
and mineral of fish and reptile and bird and insect ;
of tree and bush and herb ; of rock and pebble
and sand and dust. So with those of body, life, mind,
spirit, soul, and their kindred. So with the qualities
of objects, both physical and moral, and with their relations,
through the whole round of the categories :
position and succession, form and size, manner and degree :
all, in their indefinite multitude, are divided and
grouped, like the shades of color, and each group has
its own sign, to guide the apprehension and help the
discrimination of him who uses it. So, once more,
with the apparatus of logical statement : the ability to
put a subject and predicate closely together, and to test
their correspondence by repeated comparison, comes
only by language ; and it is the fruitful means whereby
old cognitions are corrected and new ones attained.
So, in fine, with the auxiliary apparatus of inflections
and form-words, wherein various tongues are most of
all discordant, each making its own selection of what it
will express and what it will leave for the mind to understand
without expression.

Every single language has thus its own peculiar
framework of established distinctions, its shapes and
forms of thought, into which, for the human being who
21learns that language as his “mother-tongue,” is cast
the content and product of his mind, his store of impressions,
however acquired, his experience and knowledge
of the world. This is what is sometimes called
the “inner form” of language — the shape and cast of
thought, as fitted to a certain body of expression.
But it comes as the result of external influence ; it is
an accompaniment of the process by which the individual
acquires the body of expression itself ; it is not
a product of his internal forces, in their free and undirected
workings ; it is something imposed from without.
It amounts simply to this : that the mind which
was capable of doing otherwise has been led to view
things in this particular way, to group them in a certain
manner, to contemplate them consciously in these
and those relations.

There is thus an element of constraint in language-learning.
But it is an element of which the learner is
wholly unconscious. Whatever language he first acquires,
this is to him the natural and necessary way of
thinking and speaking ; he conceives of no other as
even possible. The case could not be otherwise. For
even the poorest language in existence is so much better
than any one's powers could have produced unaided,
that its acquisition would imply a greatly accelerated
drawing out and training of the powers of even the
most gifted being ; the advantage is so great that the
disadvantage entirely disappears before it. We, to be
sure, looking on from without, can sometimes find reason
for regret, saying : “Here is a man of capacities
far beyond the average of the degraded community of
which he is a member ; in justice to those capacities,
he should have had his birth where a higher language
would have developed them into what they were able
22to become ; only,” we should have to add, “this barbarian
tongue raises him far above what he could have
become had he never learned to speak at all.” Moreover,
it is far oftener the case that the individual's linguistic
lot is beyond his deserts ; that he acquires a
language above his level, and would have been better
fitted by a lower dialect.

It is not easy to over-estimate the advantage won by
the mind in the obtaining of a language. Its confused
impressions are thus reduced to order, brought under
the distinct review of consciousness and within reach,
of reflection ; an apparatus is provided with which it
can work, like the artisan with his tools. There is no
other parallel so close, as regards both the kind and the
degree of assistance afforded, as this between words,
the instruments of thought, and those other instruments,
the creation and the aids of man's manual dexterity.
By as much as, supplied with these, man can
traverse space, handle and shape materials, frame textures,
penetrate distance, observe the minute, beyond
what he could compass with his unequipped physical
powers, by so much is the reach and grasp, the penetration
and accuracy, of his thought increased by speech.
This part of the value of speech is by no means easy
to bring to full realization, because our minds are so
used to working by and through words that they cannot
even conceive of the plight they would be in if deprived
of such helps. But we may think, for example,
of what the mathematician would be without figures
and symbols.

In respect to this general training and equipment
of the mind for work, the first acquisition of a language
does for the individual what can never be repeated
later. When we first take hold of an additional
23language, we cannot help translating its signs into
those we already know ; the peculiarities of its “inner
form,” the non-identity and incommensurability of its
shaped and grouped ideas with those of our native
speech, escape our notice. As we gain familiarity with
it, as our conceptions adapt themselves to its framework
and operate directly through it, we come to see
that our thoughts are cast by it into new shapes, that
its phraseology is its own and inconvertible. Perhaps
it is here that we get our most distinct hint of the element
of constraint in language-learning. Certainly,
the exceptionally-gifted Polynesian or African who
should learn a European language — as English, French,
German — would find himself prepared for labor in departments
of mental action which had before been inaccessible
to him, and would realize how his powers
had been balked of their best action by the possession
of only the inferior instrument. The scholars of the
Middle Ages, who employed the Latin for the expression
of their higher thought, did so partly because the
popular dialects had not yet become enriched to a capacity
for aiding the production of such thought and
for expressing it.

But in all other respects, the learning of a second
language is precisely the same process as the learning
of a first, of one's own “mother-tongue.” It is the
memorizing of a certain body of signs for conceptions
and their relations, used in a certain community, existing
or extinct — signs which have no more natural and
necessary connection with the conceptions they indicate
than our own have, but are equally arbitrary and conventional
with the latter ; and of which we may make
ourselves masters to a degree dependent only on our
opportunities, our capacity, our industry, and the length
24of time devoted to the work ; even coming to substitute,
if circumstances favor, the second language in our
constant and ready use, and to become unfamiliar with
and forget its predecessor.

We realize better in the case of a second or “foreign,”
than in that of a first or “native” language,
that the process of acquisition is a never-ending one ;
but it is not more true of the one than of the other.
We say, to be sure, of a child who has reached a certain
grade that he “has learned to speak ;” but we
mean by this only that he has acquired a limited number
of signs, sufficient for the ordinary purposes of the
childish life, together with the power, by much practice,
of wielding them with adroitness and general correctness.
There are, probably, only a few hundred
such signs, all told ; and outside their circle, the English
is as much an unknown language to the child as is
German, or Chinese, or Choctaw. Even ideas which
he is fully able to grasp when put into his acquired
phraseology are unintelligible if expressed as grown-up
men would naturally write them ; they must be translated
into childish phrase. What he has is especially
the central core of language, as we may call it : signs
for the most commonly recurring conceptions, words
which every speaker uses every day. As he grows
older, as his powers develop and his knowledge increases,
he acquires more and more ; and in different
departments, according to circumstances. He who has
to turn at once to the hard work of life may add to the
first childish store little besides the technical expressions
belonging to his own narrow vocation ; he, on
the other hand, who devotes years to the sole work of
getting himself educated, and continues to draw in
knowledge through the rest of his life, appropriates
25constantly larger stores, and rises to higher styles of expression.
The ordinary vocabulary of the educated, including
a great variety of the technical terms of special
branches of knowledge with which the educated man
must have at least a degree of acquaintance, he may
come to understand and to use with intelligence ; but
there will be whole bodies of English expression which
he cannot wield, as well as styles to which he does
not attain. The vocabulary of a rich and long-cultivated
language like the English may be roughly estimated
at about 100,000 words (although this excludes a
great deal which, if “English” were understood in its
widest sense, would have to be counted in) ; but thirty
thousand is a very large estimate for the number ever
used, in writing or speaking, by a well-educated man ;
three to five thousand, it has been carefully estimated,
cover the ordinary needs of cultivated intercourse :
and the number acquired by persons of lowest training
and narrowest information is considerably less than
this. Nowhere more clearly than here does it appear
that one gets his language by a process of learning, and
only thus ; for all this gradual increase of one's linguistic
resources goes on in the most openly external
fashion, by dint of hearing and reading and study ; and
it is obviously only a continuation, under somewhat
changed circumstances, of the process of acquisition of
the first nucleus ; while the whole is parallel to the beginning
and growth of one's command of a “foreign”
tongue.

The same thing, however, appears clearly enough,
if we consider more narrowly the somewhat shifting
relations between our linguistic signs and the conceptions
for which they stand. The relation is established
at first by a tentative process, liable to error and subject
26to amendment. The child finds out very soon that
names do not in general belong to single objects alone,
but rather to classes of related objects ; and his power
of noting resemblances and differences, the most fundamental
activity of intellect, is from the first called
into lively action and trained by the constant necessity
of applying names rightly. But the classes are of every
variety of extent, and in part determined by obscure
and perplexing criteria. We have noted already the
natural and frequent childish error of using papa and
mamma in the sense of ‘man’ and ‘woman ;’ the child
is puzzled, by and by, by finding that there are other
papas and mammas, though he must not call them so.
An older child he learns to call, for example, George ;
but he finds that he must not say George of other kindred
beings ; there is another word, boy, for that use.
But then, again, he makes acquaintance with still other
Georges ; and to find the tie that binds them into a
class together is a problem quite beyond his powers.
A variety of creatures of very diverse appearance he
learns to call dog ; but he may not take the same liberty
with horse ; though mules and donkeys are much
more like horses than greyhounds and lapdogs are like
terriers, they must be carefully distinguished in appellation.
A sun in a picture is still a sun ; and in a cultivated
community the child soon gets his imagination
trained to recognize the pictured representations of
things, and to call them by the same names, while
still distinctly aware of the relation between thing and
picture ; while a grown-up untutored savage is completely
baffled by such a counterfeit, seeing in it only a
confusion of lines and scratches. A toy house or tree
is to have the title house or tree ; but a kind of toy human
being has the specific name of doll. The words
27of degree have their peculiarities of application : near
is sometimes at an inch of distance, sometimes at a
rod ; a big apple is not nearly so big as a little house ;
a long time means a few minutes or a few years.
The inconsistencies of expression are numberless ; and
till added experience explains them, there is room for
misapprehensions and blunders. Moreover, there are
cases in which the difficulty is much more persistent,
or is never wholly removed. Fish even adult apprehension
makes to include whales and dolphins, till scientific
knowledge points out a fundamental difference
as underlying the superficial resemblance.

But it is especially in regard to matters of which
the knowledge is won in a more artificial way, that the
beginner's ideas are vague and insufficient. For example,
children are apt to be taught the names and
definitions of geographical objects and relations without
gaining any real comprehension of what it all
means ; a map, a more unintelligible kind of picture, is
little better than a puzzle ; and even older children, or
grown men, have defective conceptions which are only
rectified by exceptional experience in after-life. Localities,
of course, are most incorrectly imagined by those
who have not seen them. Of Sedan, Peking, Hawaii,
Chimborazo, every well-instructed person knows enough
to be able to talk about them ; but how imperfectly do
we conceive them, as compared with one who has lived
at or near them ! We have to be extremely careful, in
teaching the young, not to push them on too rapidly,
lest we prove to have been building up a mere artificial
and empty structure of names, without real enlightenment.
And yet, something of this is unavoidable, a
necessary incident of instruction. A host of grand
conceptions are put before the youthful mind, and kept
28there by a paltry association or two, while it is left for
after-development to fill them out to more nearly their
true value. The child is ludicrously unable at first to
know what is meant by God, or good, or duty, or conscience,
or the world, even as sun and moon, weight
and color, involve infinitely more than he has an inkling
of ; but the word, in each case, gives him a definite
nucleus, about which more and ever more knowledge
may be grouped ; he makes a constant approach toward
the right conception, even if it be one to which no human
wisdom has yet attained. For the condition of
the child, after all, differs only in degree from that of
the man, and in no very great degree. Our words are
too often signs for crude and hasty, for indefinite and
indefinable, generalizations. We use them accurately
enough for the ordinary practical purposes of life ; and
most of mankind go through life content with that, letting
instruction and experience bring what improvement
they may ; few have the independence, even if
they had the time and ability, to test every name to
the bottom, drawing precise limits about each. For
the most part, we are loose thinkers and loose talkers,
misled into error in an infinity of cases by our ignorance
of the terms we glibly use. But even the wisest
and most thorough of us is met by the impossibility of
giving to speech a preciseness of definition which should
exclude misunderstanding and unsound reasoning — especially
as to matters of subjective import, where it is
hard to bring conceptions to a sharp test. And so the
differences of view, even of philosophers, take on the
form of verbal questions, controversies hinge on the
interpretation of a term, and every writer who aims at
exactness has to begin with definitions — to which, then,
he finds it impossible to be faithful ; some antagonist
29or successor, perhaps, shows him to have failed of exactness
at a critical point, and tumbles into ruins the
whole magnificent structure of fancied truth which he
had erected.

We see from all this, it may be observed, how far
language is from being identical with thought. It is
so just as much as the mathematician's figures and
symbols are identical with his conceptions of mathematical
quantities and relations ; and not one whit
more. It is, as we noticed at the outset, the means
of expression of thought, an instrumentality auxiliary
to the processes of thought. An acquired language
is something imposed from without upon the
methods and results of mental action. It does, indeed,
as a frame-work imposed upon a growing and developing
body, give shape to that which underlies it, determining
the “inner form ;” and yet it is everywhere
loose and adjustable. While working by it, the mind
also works under it, shifting and adapting, changing
and improving its classifications, working in new knowledge
and better insight. Thus far we have emphasized
the passive receptive work of the mind in dealing with
language, because that is, especially at the outset, the
bulk of its work ; in the following chapters we have
to take account of its more independent and creative
activity.

But nothing that has been said is to be misconstrued
into meaning that the mind is not, in all its work, essentially
an active and creative force, or that it gets by
instruction a faculty which, it did not before possess.
All that is implied in the power to speak belongs indefeasibly
to man, as a part of his natural endowment ;
but this power is guided in its development, and determined
in the result it attains, by the example and instruction
30of other minds, already developed. It does
nothing which it might not have done alone, under
favoring circumstances, and with sufficient time — the
life-time, namely, of a few score or hundred generations ;
but for what it actually does, both as regards
the how much and the how, it has to thank those about
it. Its acquisition of language is a part of its education,
in just the same manner and degree as the other
parts of education.31

Chapter III.
The conservative and alterative forces in language.

Other side of life of language ; growth and change ; question of its mode
and cause. Illustrative passage from oldest English, or Anglo-Saxon ;
exposition of its differences from modern English : differences
of pronunciation ; abbreviations and extensions ; changes of
meaning ; of phraseology and construction. Classification of linguistic
changes.

We have seen in the foregoing chapter that the individual
learns his language, obtaining the spoken signs
of which it is made up by imitation from the lips of
others, and shaping his conceptions in accordance with
them. It is thus that every existing language is maintained
in life ; if this process of tradition, by teaching
and learning, were to cease in any tongue upon earth,
that tongue would at once become extinct.

But this is only one side of the life of language. If
it were all, then each spoken dialect would remain the
same from age to age. In virtue of it, each does, in
fact, remain nearly the same ; this is what maintains
the prevailing identity of speech so long as the identity
of the speaking community is maintained — aside
from those great revolutions in their circumstances
which now and then lead whole communities to adopt
the speech of another people. This, then, is the grand
32conservative force in the history of language ; if there
were no disturbing and counteracting forces to interfere
with its workings, every generation to the end of time
would speak as its predecessors had done.

Such, however, as every one knows, is very far from
being the case. All living language is in a condition
of constant growth and change. It matters not to
what part of the world we may go : if we can find for
any existing speech a record of its predecessor at some
time distant from it in the past, we shall perceive that
the two are different — and more or less different, mainly
in proportion to the distance of time that separates
them. It is so with the Romanic tongues of southern
Europe, as compared with their common progenitor the
Latin ; so with the modern dialects of India, as compared
with the recorded forms of speech intermediate
between them and the Sanskrit, or with the Sanskrit
itself ; and not less with the English of oar day, as
compared with that of other days. An English speaker
even of only a century ago would find not a little in
our every-day speech which he would understand with
difficulty, or not at all ; if we were to hear Shakespeare
read aloud a scene from one of his own works, it would
be in no small part unintelligible (by reason, especially,
of the great difference between his pronunciation and
ours) ; Chaucer's English (500 years ago) we master by
dint of good solid application, and with considerable
help from a glossary ; and King Alfred's English (1000
years ago), which we call Anglo-Saxon, is not easier to
us than German. All this, in spite of the fact that no
one has gone about of set purpose to alter English
speech, in any generation among the thirty or forty
that have lived between us and Alfred, any more than
in our own. Here, then, is another side of the life of
33language for us to deal with, and to explain, if we can.
Life, here as elsewhere, appears to involve growth and
change as an essential element ; and the remarkable
analogies which exist between the birth and growth
and decay and extinction of a language and those of an
organized being, or of a species, have been often enough
noticed and dwelt upon : some have even inferred from
them that language is an organism, and leads an organic
life, governed by laws with which men cannot
interfere.

Plainly, however, we should be overhasty in resorting
to such an explanation until after mature inquiry
and deliberation. There is no primâ facie impossibility
that language, if an institution of human device,
and propagated by tradition, should change. Human
institutions in general go down from generation to
generation by a process of transmission like that of
language, and they are all modified as they go. On
the one hand, tradition is by its very nature imperfect
and inaccurate. No one has ever yet been able to prevent
what passes from mouth to ear from getting altered
on the way. The child always commits blunders,
of every kind, in his earlier attempts at speaking : if
careful and well trained, he learns later to correct
them ; but he is often careless and untrained. And
all through the life-long process of learning one's
“mother-tongue,” one is liable to apprehend wrongly
and to reproduce inexactly. On the other hand, although
the child in his first stage of learning is more
than satisfied to take what is set before him and use it
as he best can, because his mental development is far
short of that which it represents, and its acquisition is
urging him on at his best rate of progress, the case
does not always continue thus with him : by and by
34his mind has grown up, perhaps, to the full measure
of that which his speech represents, and begins to exhibit
its native and surplus force ; it chafes against the
imposed framework of current expression ; it modifies
a little its inherited instrument, in order to adapt this
better to its own purposes. So, to have recourse to an
obvious analogy, one may, by diligent study under instructors,
have reached in some single department — as
of natural science, mathematics, philosophy — the furthest
limits of his predecessors' knowledge, and found
them too strait for him ; he adds new facts, draws new
distinctions, establishes new relations, which the subsisting
technical language of the department is incompetent
to express ; and there arises thus an absolute
need of new expression, which must in some way or
other be met ; and it is met. Every language must
prove itself able to signify what is in the minds of its
speakers to express ; if unequal to that, it would have
to abdicate its office ; it would no longer answer the
purposes of a language. The sum of what all the individual
speakers contribute to the common store of
thought and knowledge by original work has to be
worked into the “inner form” of their language along
with and by means of some alteration in its outer form.

Here, then, at any rate, are two obvious forces, having
their roots in human action, and constantly operating
toward the change of language ; and it remains to
be seen whether there are any others, of a different
character. Let us, then, proceed to examine the changes
which actually go on in language, and which by their
sum and combined effect constitute its growth, and see
what they will say as to the force that brings them
about.

Thus although, from one point of view, and and
his are the only words in the Anglo-Saxon passage
which are the same also in the English — and not even
those really, since their former pronunciation was somewhat
36different from their present — from another point
of view everything is English, excepting se, ‘the,’ and
hî, ‘they’ — and even those, virtually ; since they are
cases of inflection of the definite article and third personal
pronoun, of which other cases (as the, that, they,
and he, his, him) are still in good use with us. Both
the discordance and the accordance are complete, according
to the way in which we look at them. We
will proceed to examine the passage a little in detail,
in order to understand better the relations between the
older and the newer form.

In the first place, their pronunciation is even more
different than is indicated by the written text. There
are at least two sounds in the Anglo-Saxon which are
unknown in our present speech : namely, the h of
cnihtas, which was nearly or quite the same with the
ch of the corresponding German word knecht, and the y
of hyngrede, which was the German ü and French u, an
u(oo)-sound with an i(ee)-sound intimately combined
with it. On the other hand, there are sounds in the
English which were unknown to the Anglo-Saxon.
Our so-called “short o,” of on, was no ancient sound ;
nor was the “short u” of hegun, pluck, which had
then the vowel-sound of book and full ; nor was the
“short i” of his, which was more like the French
and German short i, not markedly different in quality
from the true long i, our so-called “long e,” oree-sound.
All these are examples of the manifold changes of English
pronunciation during the thousand years since Alfred — changes
which have altered the whole aspect of
our orthoëpy and orthography. And others of them
are illustrated in the passage : for instance, our knight
and eat show protractions of the short vowels of cniht
and etan, each typical of a whole class of cases ; and
37the lengthened i has been changed into a diphthong,
which we call “long i” simply because it has taken the
place of our former long i(ee) ; while we call the real
long i of eat by the false name of “long e” for the
same reason.

Again, we may observe in the forms of many words
the effects of a tendency toward abbreviation. Reste
and hyngrede have lost with us their final e, which in
Anglo-Saxon, as now in German and Italian, made an
additional syllable. Ongunnon, pluccian, and etan
have lost both vowel and consonant of a final syllable ;
and these syllables were the distinctive endings, in the
first word of the plural verbal inflection (ongan, ‘I or
he began,’ but ongunnon, ‘we or they began’), in the
other two of the infinitive. In œceras, ‘acres,’ and
cnihtas, ‘knights,’ though we have saved the final s of
the plural ending, it no longer makes an additional
syllable. And in sôthlîce, ‘soothly’ (i. e. ‘truly, verily’),
there is a yet more marked abbreviation, to which
we shall presently return.

On the other hand, ear, ‘ears,’ and fôr, ‘fared,’
have been extended in modern time by the addition of
other pronounced elements. It was the rule in Anglo-Saxon
that a neuter noun of one syllable, if of long
quantity, had no (nom. or accus.) plural ending. With
us, every noun, of whatever gender or quantity (save
a few exceptions, of which we need take no account
here), takes s as its plural sign. As for fôr, the Anglo-Saxons
conjugated faran, ‘fare,’ as they did dragan,
‘draw,’ and said fôr, ‘fared,’ like drôh, ‘drew’ (compare
the corresponding German fahren fuhr and tragen
trug) — that is to say, faran was to them a verb of
the “irregular,” or “old,” or “strong” conjugation.
But for a long time there has existed in English speech
38a tendency to work over such verbs, abandoning their
irregularly varying inflection, and reducing them to
accordance with the more numerous class of the “regularly”
inflected, like love, loved ; and fare is one of the
many that have undergone this change. The process
is quite analogous with that which has turned ear into
ears : that is to say, a prevailing analogy has been
extended to include cases formerly treated as exceptional.

In connection with ear comes to light another very
striking difference between the ancient and modern
English : the Anglo-Saxon had grammatical gender,
like the Greek and Latin and German ; it regarded ear
as neuter, but œcer and dœg as masculine, and, for instance,
tunge, ‘tongue,’ and dœd, ‘deed,’ as feminine ;
to us, who have abolished grammatical gender in favor
of natural sex, all are alike neuter.

We turn now to consider a few points relative to
the meaning of the words used. In fôr we find a
marked difference of sense as well as of form. It is
part of an old Germanic verb meaning ‘go,’ and is
traceable even back into the earliest Indo-European, as
the root par, ‘pass’ (Skt. pârayâmi, Gk. περάω, Lat.
ex-per-ior) ; now it is quite obsolete in any such sense
as this, and rather unusual even in that of ‘getting on,’
‘making progress :’ “it fared ill with him.” Again,
œcer meant in Anglo-Saxon a ‘cultivated field,’ as does
the German acker to the present day ; and here, again,
we have its very ancient correlatives in Sanskrit agra,
Greek ἀγρός, Latin ager ; the restriction of the word
to signify a field of certain fixed dimensions, taken as
a unit of measure for fields in general, is something
quite peculiar and recent. It is analogous with the like
treatment of rod and foot and grain, and so on, except
39that in these cases we have saved the old meaning
while adding the new.

Among the striking peculiarities of the Anglo-Saxon
passage is its use of the words Hœlend, ‘healing one’,
reste-dœg, ‘rest-day,’ and leorning-cnihtas, ‘learning-knights’
(i. e. ‘youths under instruction’), in the sense
respectively of ‘Savior,’ ‘sabbath,’ and ‘disciples.’
Though all composed of genuine old Germanic materials,
they were nevertheless recent additions to the
language. The introduction of Christianity had created
a necessity for them. For the new idea of the
Christian Creator and Father, the old word god, ennobled
and inspired with a new meaning, answered
English purposes well enough. But there was no current
name applicable to the conception of one who
saved men from their sins, making them whole or hale ;
and so the present participle of the verb hœlan, ‘make
hale, heal,’ was chosen to represent σωτήρ, and specialized
into a proper name, a title for the one Savior. It
is the same word which, in German, is still current as
Heiland. Reste-dœg, as name for the sabbath, needs
no word of explanation or comment. As for leorning-cnihtas,
rendering discipuli and μαθηταί, its most
striking characteristic, apart from its rather lumbering
awkwardness, is the peculiar meaning which it implies
in cniht, ‘knight.’ Between our knight, a word of
high chivalric significance, and the German knecht,
‘servant, menial,’ is a long distance : both show a deviation,
the one in an upward and the other in a downward
direction, from the indifferent ‘youth, fellow,’
which lies at the bottom of the use of the word in our
Anglo-Saxon compound.

But a not less noteworthy point in the history of
these words is that in our later usage they have all become
40superseded by other terms, of foreign origin.
The Anglo-Saxon did not, like our English, resort freely
to foreign stores of expression for the supply of new
needs. It was easier then to accept the new institutions
of Christianity than new names for them. We
have wonderfully changed all that, under the operation
of causes which will come up for notice hereafter
(chapter vii.) ; and in place of the three new Saxon
names we have put other yet newer ones : two Latin-French,
disciple and savior, and one Hebrew, sabbath.
The substitution exemplifies a capital trait in English
language-history.

Our attention being thus directed to the introduction
of new elements into Anglo-Saxon, we will note
another case or two of the same kind of linguistic
change in another department. Sôthlîce is an adverb,
answering to our ‘truly.’ We recognize in the first
part of it our sooth, a word now almost obsolete — quite
so, as far as ordinary use is concerned. Its second part,
lîce, is our ly. But it is also a case-form (instrumental)
of an adjective lîc, our like, which was appended to the
noun sôth, ‘truth,’ forming a compound adjective (or
adjectival derivative) equivalent to truth-like, and completely
analogous to truthful, from truth and full.
Our adverbial ending ly, then, by which most of our
adverbs are made, and which to us is only a suffix, is
really the product of alteration of a case-form of a
compounded adjective, a word originally independent.
Instead of using, like the modern German, the base or
crude-form of an adjective as adverb — that is to say, in
the formal grammatical character of adaptedness to
qualify a verb or adjective rather than a substantive
— we have wrought out for that purpose a special form,
of which the history of development may be followed
41step by step to its origin, and which is exclusively the
property of our language among its kindred Germanic
dialects.

A second case is brought before us in hyngrede.
Its preterit ending de is not, like the adverbial ly, exclusively
English ; it is rather, like the adjective lîc, a
common Germanic possession. Without dwelling here
at length upon its history, we will only observe that it
is, like lîce, traced back to an independent word, the
preterit did, which was in remote Germanic time added
to some verbal derivative, or other part of speech, to
form a new style of past tense, when the yet older processes
of preterit formation had become no longer manageable.

There are also changes of construction in our passage
which ought not to pass without a moment's notice.
The word leorning-cnihtas is object, not subject,
of hyngrede ; and the construction is that peculiar one
in which the impersonal verb, without expressed subject,
takes before it as object the person affected by the
action or feeling it signifies. This is still a familiar
mode of expression in German, where one freely says
mich hungerte, ‘me hungered,’ for ‘I hungered ;’ and
even we have a trace of it, in the obsolescent methinks,
German mich dünkt — that is, ‘it seems to me.’ Again,
the infinitives pluccian and etan, being by origin verbal
nouns and having properly the construction of
nouns, are directly dependent, as objects, on the transitive
verb ongunnon. We make the same construction
with some verbs : so, he will pluck, he must eat,
see him pluck, let him eat ;and even after began shortened
to 'gan it is allowed ; 14 but in the vast majority of
eases we require the preposition to as “infinitive sign,”
42saying “began to pluck and to eat.” This preposition
was not unknown in Anglo-Saxon ; but it was used
only where the connection pretty manifestly favored
the insertion of such a connective ; and the infinitive
after it had a peculiar form : thus, gôd to etanne, ‘good
unto eating,’ and so ‘good to eat.’ The to which at
the period of our specimen-passage was a real word of
relation has now become the stereotyped sign of a certain
verbal form ; it has no more independent value
than the ending an of pluccian and etan — which, indeed,
it in a manner replaces ; though not, like -ly and
-d, combined with the word to which it belongs, its office
is analogous with theirs.

We will notice but one thing more in the passage :
the almost oblivion into which sôth, our sooth, has fallen.
Only a small part of the great body of English-speakers
know that there is such a word ; and no one
but a poet, or an imitator of archaic style, ever uses it.
We have put in place of it true and truth, which of old
were more restricted to the expression of faithfulness,
trustworthiness.

The brief sentence selected, we see, illustrates a
very considerable variety of linguistic changes ; in fact,
there is hardly a possible mode of change which, is not
more or less distinctly brought to light by it. Such
are, in general, the ways in which a language comes to
be at a later period different from what it has been at
an earlier. They are matters of individual detail ; each
item, or each class of accordant items, has its own time
and occasion, and analogies, and secondary causes, and
consequences ; it is their sum and collective effect which
make up the growth of language. If we are to understand
how language grows, we must take them up and
examine them in their individuality. This, then, is the
43subject which is now for some time to occupy us : an
inquiry into the modes of linguistic change, and their
causes, nearer and remoter.

We have already rudely made one classification of
these linguistic changes, founded on the various purpose
which they subserve : namely, into such as make
new expression, being produced for the designation of
conceptions before undesignated ; and such as merely
alter the form of old expression ; or, into additions and
alterations. It will, however, suit our purpose better
to make a more external division, one depending upon
the kind of change rather than upon its object. In
carrying this out, it will be practicable to take everywhere
sufficient notice of the object also.

This classification is obviously exhaustive ; there
can be no change in any language which will not fall
under one or other of the three classes here laid down.44

Chapter IV.
Growth of language : change in the outer form of
words.

Relation of the word to the conception it designates, as conditioning the
possibility, and the mutual independence, of its changes of form and
meaning. Tendency to ease or economy in changes of form. Abbreviation
of words ; examples ; its agency in form-making ; loss
of endings. Substitution of one sound for another ; examples of
vowel and consonant change ; Grimm's law ; underlying causes of
phonetic change ; processes of utterance ; physical or natural scheme
of spoken alphabet ; its series and classes ; distinction of vowel and
consonant ; syllabic or articulate character of human speech. General
tendencies in phonetic change. Limits to phonetic explanation.
Change of form by extension of a prevailing analogy.

In this chapter we have to take up and illustrate
the first division of the first class of linguistic changes,
that which includes alterations of the uttered and audible
forms of words. But first it will be well to call
attention anew to certain general principles (already
hinted at in the second chapter), which are of fundamental
importance as underlying the whole subject of
verbal alteration, whether in respect to shape or sense.
And we shall best attain our object by discussing a
selected example.

Let us take a familiar word, found in most of the
languages of modern Europe, and having a well-known
history — the word bishop. It comes, as almost every
45one is aware, from the Greek ἐπίσκοπος (episkopos).
This, again, is a derivative from the root sleep, ‘see,
look,’ with the prefix epi, ‘at ;’ and so it means by origin
simply ‘inspector, overseer ;’ in the early formative
period of the Christian church, it was selected as official
designation of the person to whom was committed
the oversight of the affairs of a little Christian community :
and both word and office are still readily recognizable
in our bishop and its use. But we have cut
down the long title into a briefer one, by dropping its
first and last syllables : and we have worked over into
new shape most of its constituent sounds : we have
changed the first p into a different but closely kindred
sound, its corresponding sonant, b ; the sk, a sibilant
with following palatal mute, has been as it were fused
together into the more palatal sibilant, sh, a simple
sound, though it is written with two letters, just because
of its usual derivation by fusion of two simple
sounds into one ; and the o-sound of the second syllable
has been neutralized into what we usually call the
“short u” sound — and the result is our word, with two
syllables instead of four, and with five sounds instead
of nine, and among those five only two, the consonant
p and the vowel i, which were of the nine. The German,
in its bischof, has altered even the final p. The
French, again, has made out of the same original a
very different looking product, évêque, which does not
contain a single sound that is found either in the English
word or in the German ; it comes, by another set
of changes, from evesc, for episk. In Spanish, the word
is made into obispo, by yet another process, and this is
further shortened in the Portuguese bispo. The Danish,
finally, shows the extreme of abbreviation, in the
monosyllable bisp. While these changes have been
46going on, the meaning of the word has been not less
altered. The official who was, when first named, merely
overseer of the interests of a little band of timid
proselytes to a new and proscribed faith, half-expectant
martyrs, has risen immensely in dignity and power,
along with the rise of the religion to importance, and
to preeminence in the state ; he has become a consecrated
prelate, charged with spiritual and temporal
authority through an entire province — a kind of ecclesiastical
prince, yet still wearing his old simple title.

From this word, taken as a type, we may learn
many things, which a wider induction, from innumerable
examples, would only confirm.

First, the name had its origin in a need which arose
at a particular time and place in the progress of human
history. A new religion came into being, and required
organization of its votaries ; and this made a call for
technical designations of its officials — which, as in all
similar cases, were then without difficulty found : not
bishop only, but priest and deacon, and so on. The
words were, in fact, already in existence, as general
terms, ready, like the people who should wear them, to
be selected and set apart to this specific office. What
should come of it further, whether the new titles should
rise to importance and attain wide currency, depended
on the after-fate of the system to which they belonged.

Again, the word bishop did not describe, either fully
or accurately, the office which it was used to designate.
Mere ‘looking on’ or ‘looking over’ was not what
men expected of the person elected ; the barest hint of
his official duty is contained in the term. But, imperfect
as it may have been as a description, it was sufficient
as a designation. The description would have
47needed to be a long one, and varied to suit the circumstances
of each new place and time ; the title
answered its desired purpose equally well in all circumstances.

Hence also, as little did the retention of the title
depend upon the maintenance of just that kind and degree
of relation between its etymological meaning and
the office it denominated which had existed at the outset.
Even what etymological appropriateness it once
possessed was no longer of any account, when once it
had become established in use as name of the office.
It passed, with the institution to which it belonged,
into the keeping and use of great communities which
did not speak Greek and had no knowledge of what it
originally signified, and it served its purpose with them
just as well as if they had understood its whole history.
From the moment when it became an accepted sign for
a certain thing, its whole career was cut loose from its
primitive root ; it became, what it has ever since continued
to be, a conventional sign, and hence an alterable
sign, for a certain conception, but a variable and
developing conception.

In this fundamental fact, that the uttered sign was
a conventional one, bound to the conception signified
by it only by a tie of mental association, lay the possibility
both of its change of meaning and of its change
of form. If the tie were a natural, an internal and
necessary one, it would seem to follow that any change
in either would have to be accompanied by a change in
the other. But in the case taken, while the idea has
expanded into greatness, the word has been shrinking
in its proportions, and is nowhere more than a fragment
of its former self. The only tendency which we
can discover in its treatment is a tendency toward
48economy of effort in its utterance ; it has been reshaped
to suit better the convenience of those who used it.
In the forms which it has assumed, we can plainly trace
the influence of national habits. The Germanic races
accent prevailingly the first syllable of their words ;
they have, then, while retaining the old accented syllable
with its accent, cast off the one that preceded it.
The French, on the other hand, accents its final syllable
(which is regularly the Latin accented syllable) ; it,
accordingly, drops all that followed the accented -písk-,
but retains the initial syllable which the others rejected.
And the other various alterations of form
which the word has undergone may be paralleled with
classes of similar alterations in other words of the same
language ; all apparently made to humor the ease of
the speakers.

In treating separately, therefore, the subjects of
change of form and change of meaning in words, we
are not parting two necessarily connected and mutually
dependent processes, but only recognizing a natural
independence. A word may change its form, to any
extent, without change of meaning ; it may take on an
entirely new meaning without change of form. As a
matter of fact, the words are few or none which have
not done both ; and, in taking up either, we shall have
to use examples which illustrate the other as well. All
the material of language exhibits more or less the
working of all the processes of growth ; but it will not
be hard to direct our attention, exclusively or especially,
now to the one and now to the other of them.

And, as regards change of form, we have to recognize,
as the grand tendency underlying all the innumerable
and apparently heterogeneous facts which it
embraces, the disposition, or at least the readiness, to
49give up such parts of words as can be spared without
detriment to the sense, and so to work over what is
left that it shall be more manageable by its users, more
agreeable to their habits and preferences. The science
of language has not succeeded in bringing to light any
more fundamental law than this, even any other to put
alongside of it ; it is the grand current setting through
universal language, and moving all its materials in a
given direction — although, like other such currents, it
has its eddies, where a counter-movement on a small
scale may seem to prevail. It is another manifestation
of the same tendency which leads men to use abbreviations
in writing, to take a short cut instead of going
around by the usual road, and other like things — in
which there is no harm, unless more is lost than gained
by the would-be economy : then, indeed, it becomes
rather laziness than economy. Its operation, as manifested
in language, is of both kinds, true economy and
lazy wastefulness ; for it works on with blind absence
of forethought, heedless, in part, of the results to
which it leads.

The character of the tendency is seen most clearly
in the abbreviation of words ; obviously, nothing else
is needed to explain the gradual reduction of form
which has ever been going on in the constituents of
every language. We noticed above (p. 38) sundry examples
of innocent abbreviation made by us in the
words of our specimen-passage : the most striking was
our knights (i. e. naits) for cnihtas, a loss of two pronounced
elements besides the shortening by a syllable.
It is easy to perceive in all these cases the tendency to
ease at work ; and we appreciate in the last the comparative
difficulty of uttering a k-sound before an n :
the class of words in which we have dropped it off is
50not a small one (e. g. knife and knit, gnaw and gnarl).
And the German ch-sound (of ich, etc.) belonging to the
h of cniht, itself coming by phonetic change from an
earlier k, is one which English organs have taken a distaste
to, and have refused longer to produce. Sometimes
they have left it out altogether (with compensatory
prolongation of the preceding vowel), as in the
word before us ; sometimes they have changed it into
f, as in draught and laugh. In ongunnon, ‘begun,’
however, and in pluccian and etan, ‘pluck’ and ‘eat,’
we have instances of that kind of loss which is akin to
wastefulness ; for the lost final syllables are those which
showed the grammatical form of the words, being plural
ending and infinitive ending. Regrettable as they
may be, the history of our language, and of the others
related with it, has been from the beginning marked
with such losses, whereby grammatical distinctions have
been let go, along with the forms on which the speakers'
consciousness of them depended. To show this
more fully, we will for a moment follow the history of
the on, the now lost ending of ongunnon. In the oldest
form to which it can be traced, it was anti, probably
the relic of an independent pronoun or pronouns, distinguishing
the third person plural in all verbal inflection.
In the Latin it is shortened to unt, but still perfectly
distinctive. In the oldest Germanic (Mœso-Gothic),
it is and in the present tense, but in the preterit
already contracted to un. The corresponding ending
in the first person plural was masi, also of pronominal
derivation ; this, after passing through such intermediate
forms as Sanskrit mas, (Doric) Greek μες, Latin
mus, and Slavonic mŭ, had become in Gothic am in the
present, um in the perfect. In German, we find only
en in both first and third person, the slight difference of
51um and un having been obliterated ; but the second
person has et, different from the other two ; in the Anglo-Saxon,
this distinction has gone the way of the rest,
and we have left only a general ending on, separating
all the plural persons alike from the singular ; and
finally, the English has swept away even this remnant
of a former elaborate system.

Another example of the earlier effects of the same
tendency in our passage is fôr, ‘fared,’ the brevity of
which, like that of English monosyllables generally, is
the result of a long succession of abbreviating processes.
Its earliest traceable form is papâra ; but even that
shows the loss of a personal ending ti, which it must
have had at the outset, and which is still represented to
us in the present tense by the t of German fährt, and
the th or s of our fareth or fares.

It was pointed out above (p. 41) that in the lîce of
sôthlîce we have the full case-form of a compounded
adjective, out of which has been made later the adjective
and adverbial suffix ly. Here is illustrated another
department of the action of the abbreviating tendency ;
its aid is essential to the conversion of what was once
an independent word into an affix, an appended element
denoting relation. So long as the word which enters
into combination with another retains its own shape
unaltered, the product is a compound only ; but when,
by phonetic change, its origin and identity with the still
subsisting independent word are hidden, the compound
becomes rather a derivative. Phonetic abbreviation has
made the difference between godly, for example — a
formed word, containing a radical and a formative element — and
godlike, a mere compound. Just so, in German,
the adjective suffix lich has become distinct from
gleich (which has, besides, a prefix) ; and in that language
52göttlich and göttergleich stand in the same manner
side by side, the one a derivative and the other a
compound. At an earlier period of Germanic language-history,
the same influence helped to convert the compound
hyngre-dide, ‘hunger-did,’ into the grammatical
form hyngre-de, ‘hunger-ed ;’ and, in vastly more ancient
time, to shape over certain pronominal elements
into the personal endings anti, masi, and ti, spoken of
above.

Thus the tendency to economy, in the very midst of
its destructive action, is at the same time constructive.
It begins with producing those very forms which it is
afterward to mutilate and wear out. Without it, compound
words and aggregated phrases would remain ever
such. Its influence is always cast in favor of subordinating
in substance what is subordinate in meaning, of
integrating and unifying what would otherwise be of
loose structure — in short, of disguising the derivation
of linguistic signs, making them signs merely, and signs
easy to manage. The point is one to which we shall
have to return in discussing (in the seventh chapter) the
third great class of linguistic changes, the production
of new words and forms.

But while the tendency is everywhere one, the ways
in which it manifests itself by abbreviation are very
various, each needing for its explanation a full understanding
of the habits of the language in which it appears.
The Germanic languages are all characterized
by a pretty strong accentual stress, laid in general on
the first or radical syllable of their words, derivative or
inflectional, and on the first members of compounds.
This mode of accentuation is itself an example of phonetic
change ; for it belongs to none of the related languages,
not even to the Slavonic, generally regarded as
53nearest of kin with the Germanic. A result of it has
been that at a later time, and quite independently in
the different Germanic languages, the endings or suffixes,
of inflection or derivation, have generally lost
their distinctive vowels, and come to be spoken with
the more neutral e : this change belongs, for example,
to the transition from Old to Middle High-German, and
from Anglo-Saxon to Old English. To it is also in
part due (though also to a more mental willingness to
abandon distinctions formerly established and maintained)
the extensive loss of endings to which these
languages have been subjected, and which appears most
of all in our English. In French, the history of change
has been somewhat different : there has been no general
shift of the place of the accent as compared with
Latin ; but there has been a wholesale abbreviation and
loss of whatever in Latin followed the accented syllable,
which has accordingly become (leaving out of account
the mute e) the final one of every regular French
word : so peuple from pópulum, faire from fácere,
prendre from prehéndere, été from both œstatem and
státum. This last example — été from statum — draws
aside our attention for a moment to a class of alterations
which, by a curious turn, end in the extension of
a word's syllabic form. To the Gallic peoples who
adopted Latin speech, the utterance of an s before a
mute — k, t, or p — seemed a difficulty which should be
avoided : just as to us, later, the utterance of a g or k
before n (in gnaw, knife, etc.). But, instead of dropping
the trying letter, they at first prefixed a vowel to
it, to make it more manageable, producing such words
as escape (Lat. scapus), esprit (spiritus), estomac (stomachus).
And then, by an actual abbreviation, and a
common one, the sibilant has in later times been usually
54dropped out, and a large class of words like école
(schola), époux (sponsus), and étude (studium), is left in
the French vocabulary. Another consequence of the
same difference of accent is the greater mutilation of
the radical part of the word in the Romanic languages
(especially French) than the Germanic ; and many of
its results have passed into English : thus, preach (Fr.
prêcher) from prœdicare, cost (Fr. coûter) from constare,
count (Fr. compter) from computare, blame (Fr. blâmer)
from blasfemare (Gr. βλασφημεῖν). Words, however,
like such and which (A.-S. swylc and hwylc, Scotch
whilk, Germ, solch and welch), from so-like and who-like,
show plainly that this disguising fusion of two
parts of a word is by no means limited to the French
part of English.

One conspicuous result of these processes is the
presence of numberless “silent letters” in the written
forms of languages like French and English, in which
the omission of sounds formerly uttered has been going
on during the period of record by writting. Such
letters are relics of modes of utterance formerly prevalent.

This must suffice by way of illustration of the tendency
to ease as manifested in abbreviation. But the
other mode of its action, consisting in the alteration of
the retained elements of words, the substitution of one
sound for another, is quite as extensive, and much more
intricate and difficult. We have already noted examples
of it : the abbreviate piskop, we saw, has been
mouthed over into bishop ; and we reviewed above (p.
37) some of the principal differences which separate
our vowel-utterance of our vowel-system, especially, has
been completely broken up by these changes, the pervading
55nature of which is attested by the strange names
we give to our vowel-sounds. The original and proper
sound of a is that in far, father : what we call “long
a” (fate) is really long ē, the nearest correspondent in
quality to the “short e ” of met, which we continue to
call by its right name because we have not generally altered
its ancient sound ; our “short a” (fat) is a new
tone, intermediate between a (far) and e (fate), and
none of our letters was devised for its representation.
In like manner, our “long e” (mete) is really a long ī
and what we call “long i” (pine) is a diphthong, ai.
And, on the other side, our “long u” (pure) is not
even a diphthong, but a syllable, yu, composed of semivowel
and vowel, and our “short ο” (not) and “short
u” (but) are new sounds, having nothing to do with
“long o” and “long u,” and, of course, possessing no
hereditary and rightful representatives in our alphabet.
It is somewhat as if we were to call our elms “tall lilacs,”
and our rose-bushes “short maples.” That our
written vowels have from three to nine values each, is
owing to the fact that we have altered their original
unitary sounds in so many different ways during the
historic period ; and there lies yet further back another
like history of change. This kind of change has been
carried on upon a larger scale in English than in almost
any other known language ; but its effects are found
abundantly in every other : the French, for example,
has given to the old Latin u a mixed i and u sound (the
German ü), and has converted the old diphthong ou
into an u(oo)-sound (being curiously paralleled in both
respects by the ancient Greek) ; it has taken a strange
fancy for the diphthongal oi (nearly equal to our wa of
was), and substitutes it for all manner of ancient sounds :
as in moi for mē, crois for credo, mois for mensis, quoi
56for quid, foi for fides, loi for legem, noir for nigrum,
noix for nucem ;and so on.

The vowels are much more liable to wholesale alteration
than are the consonants, and in our specimen-passage
the indications of consonantal change are rather
scanty. Ofer, however, has become over with us, by
the conversion of a surd into its corresponding sonant
sound, a phenomenon of very wide range and great
frequency in language ; and the same change has passed
upon the final s of his and œceras, making of it a z,
though without change of spelling. But if we look
further away, among the tongues kindred with ours, we
shall discover signs in plenty of consonantal mutation.
Dœg is in German tag, with t for d, and hyngrede is
hungerte ; and if we were to go through the whole vocabulary
of the two languages, we should find this the
prevailing relation, and be led to set up the “law” that
English d and German t correspond to one another.
Again, etan is essen in German, with an s-sound for t :
and this, too, is a constant relation ; nor is it otherwise
with thâ, which is German die, with d for th. But etan
and essen answer to Latin edere, Greek ἔδω, Sanskrit
ad ; and thâ and die are the two regular Germanic
forms of the old pronominal root ta (Gr. το, etc., Skt.
tad, etc.) : and these, too, are general facts ; insomuch
that comparative grammarians are led to set up the
“law” that a t-sound, as found in most of the languages
of our family, is regularly a th in part of the Germanic
dialects and a d in others ; that a d-sound, in like manner,
is a t or an s ; and that to English d and German t
an aspirate, th or dh, corresponds in Greek and Sanskrit.
This is, indeed, the famous “Grimm's Law,” of
the permutation or rotation of mutes in Germanic
speech. It is only an example — to be sure, an unusually
57curious and striking example — of what is universally
true between related languages : their sounds, in
corresponding words, are by no means always the same ;
they are diverse, rather, but diverse by a constant difference ;
there exists between them a fixed relation,
though it is not one of identity. Hence, in the comparison
of two languages, a first point to which attention
has to be directed is this : what sounds in the one,
vowel or consonantal, correspond to what sounds in the
other. This condition of things is only a necessary result
of the fact, already noted, that the mode of pronunciation
of every language is all the time undergoing
a change : a change now more and now less important
and pervading, but never entirely intermitted ; and that
no two languages change after precisely the same fashion.
In presence of such a phenomenon as that last instanced,
the student of language has to inquire which
(if any) of the sounds, t, d, th, dh, s is in any given
case the original, through, what steps of successive
change each varying result has been reached, and, if
within his reach, what cause has governed the course of
mutation.

And, heterogeneous as the facts may at first sight
appear, the student soon finds that they are very far
from being a mere confusion of lawless changes ; they
have their own methods and rules. One sound passes
into another that is physically akin with it : that is to
say, that is produced by the same organs, or otherwise
in a somewhat similar manner ; and the movement of
transition follows a general direction, or else is governed
by specific causes. This has caused the processes of
articulation to be profoundly studied, as part of the
science of language. And such is the interest and importance
of the study that we cannot avoid dwelling
58upon it here a little : not long enough, indeed, to penetrate
to its depths, but at least until we are able to gain
some idea of our spoken alphabet as of an orderly system
of sounds, and of the lines and degrees of relationship
which bind its members together, and help to determine
their transitions.

The organs by which alphabetic sounds are produced
are the lungs, the larynx, and the parts of the mouth
above the larynx. The lungs are, as it were, the bellows
of the organ ; they simply produce a current of
air, passing out through the throat, and varying in rapidity
or force according to the requirements of the
speaker. The larynx is a kind of box at the upper end
of the windpipe, and contains what is equivalent to the
reed of the organ-pipe, with the muscular apparatus for
its adjustment. From the sides of the box, namely,
spring forth a pair of half-valves, of which the membranous
edges, the “vocal chords,” are capable of being
brought close together in the middle of the passage,
and made tense, so that the passing current of air sets
them in vibration ; and this vibration, communicated to
the air, is reported to our ears as sound. In ordinary
breathing, the valves are relaxed and retracted, leaving
a wide and rudely triangular opening for the passage
of the air. Thus the larynx gives the element of tone,
accompanied with variety of pitch : and how important
a part of speaking this latter is, only they can fully realize
who have heard the performance of an automatic
speaking-machine, with its dreadful monotone. Above
the vibrating reed-apparatus is set, after the fashion of
a sounding-box, the cavity of the pharynx, with that of
the mouth, and the nasal passage ; and movements of
the throat and mouth-organs under voluntary control so
alter the shape and size of this box as to give to the
59tone produced a variety of characters, or to modify it
into a variety of tones — which are the sounds of our
spoken alphabet. A concise description of voice, then,
is this : it is the audible result of a column of air emitted
by the lungs, impressed with sonancy and variety
of pitch by the larynx, and individualized by the mouth-organs.

To describe in detail the construction of the vocal
apparatus, and the movements of the muscles and cartilages
and membranes which cause and modify the
vibrations, belongs to physiology ; to determine the
form and composition of the vibrations which produce
the audible variety of effects upon the ear, belongs to
acoustics : the part of phonetics, as a branch of linguistic
science, is to follow and describe, as closely as may
be, the voluntary changes of position of the mouth-organs,
etc., which determine the various sounds. These
are in part easy of observation, in part much, more difficult ;
but the main points, nearly all that we need to
take account of here, are within the reach of careful
and continued self-observation. And no one can claim
to have any proper understanding of phonetic questions,
unless he has so studied that he fairly follows and
understands the movements that go on in his own
mouth in speaking, and can arrange his spoken alphabet
into a systematic and consistent scheme. Such a
scheme, for the ordinary sounds composing the English
alphabet, we will attempt here to set up.

Every alphabetic system must start from the sound
a (of far, father) ; for this is the fundamental tone of
the human voice, the purest intonated product of lungs
and throat ; if we open the mouth and fauces to their
widest, getting out of the way everything that should
modify the issuing current, this is the sound that is
60heard. Upon this openest tone various modifications
are produced by narrowing the oral cavity, at different
points and to different degrees. The less marked modifications,
which, though they alter decidedly the quality
of the tone, yet leave predominant the element of tone,
of material, give rise to the sounds which we call vowels.
But the cavity may be so narrowed, at one and
another point, that the friction of the breath, as driven
out through the aperture, forms the conspicuous element
in the audible product ; this, then, is a sound of
very different character, a fricative consonant. And
the narrowing of the organs may be pushed even to
the point of complete closure, the element of form, of
oral modification, coming thus to prevail completely
over that of material, of tone : the product, in that case,
is made distinctly audible only as the contact is broken ;
and we call it a mute.

This brief statement suggests the plan on which the
systematic arrangement of every human alphabet is to
be made. It must lie between the completely open a
(far) and the completely close mutes ; these are its
natural and necessary limits ; and it may be expected
to fall into classes according to the intermediate degrees
of closure. But there are also other lines of
relationship in it. Theoretically, an indefinite number
of mute-closures are possible, all along the mouth,
from the lips to as far back in the throat as the organs
can be brought together ; in practice, however, they
are found to be prevailingly three : one in the front,
made by lip against lip, the labial closure, giving p ;
one in the back of the mouth, made against the soft
palate by the rear upper surface of the tongue, the
palatal (or guttural) closure, giving k ; and one intermediate
between the other two, made by the point or
61front of the tongue against the roof of the mouth near
the front teeth, the lingual (or dental) closure, giving t.
These are the only mute-closures found in English, or
French, or German ; or even in the majority of tongues
in the world. And the same tendency toward a triple
classification, of front, back, and intermediate, appears
also in the other classes of sounds, so that these arrange
themselves, in the main, nearly upon the lines of gradual
closure proceeding from the neutrally open a (far) to
the shut p, t, k. This adds, then, the other element
which is needed in order to convert the mass of articulate
utterances into an orderly system. We have below
the English alphabet arranged upon the plan described,
and will go on to consider it in more detail.

image vowels. | sonant. | semivowels. | nasals. | surd. | aspiration. | sibilants. | spirants. | mutes. | palatal series | lingual series | labial series

Along with k, t, p, in the first place, go their nearest
kindred, g, d, h. These are their sonant (or vocal,
phthongal, intonated) counterparts. In the former,
namely, there is no audible utterance, but complete
62silence, during the continuance of the closure ; the antithesis
to a is absolute ; the explosion is their whole sensible
substance. In the latter there is, even while the
closure lasts, a tone produced by the vibration of the
vocal chords, a stream of air sufficient to support vibration
for a very brief time being forced up from the
lungs into the closed cavity or receiving-box of the
pharynx and mouth. This is the fundamental distinction
of “surd” and “sonant” sounds ; anything else is
merely a consequence of this and subordinate to it ; the
names strong and weak, hard and soft, sharp and flat,
and so on, founded (with more or less of misapprehension
added) upon these subordinate characteristics, are
to be rejected. The difference between pa and ba,
then, is that the sonant utterance begins in the former
just when the contact is broken, and in the latter just
before ; in ab, it continues a moment after the contact
is made ; in aba, it is uninterrupted and continuous :
and so also with d and g.

But there is a third product of the same three positions
of mute-closure. By dropping, namely, the veil
of the palate, which in ordinary utterance closes the
passage from the pharynx into the nose, the intonated
current of b, d, g is allowed entrance to the nose and
exit there : and the result is the class of nasals (or “resonants”),
m, n, and ng (as in singing). Here, though
there is closure of the mouth-organs, the tone is so
sonorous and continuable that the breach of contact, or
explosion, is reduced to a very subordinate value, and
the class belongs high up in the alphabet, toward the
vowels.

As a general rule (exceptions to it are not common),
any language that has either of these three products
of a given mute-closure will have also the other
63two : thus, the presence of a p in the alphabet implies
also that of a b and an m ; and so on.

In the older tongues of our family, and even in
some modern ones, both of our own and of other families,
there are fourth and fifth products of the same
articulating positions, made by letting slip a bit of
breath or flatus, a brief h, after the simple mute ; turning
a p or b into a ph or bh (pronounced as written),
and so on. These are called aspirate mutes, or, briefly,
aspirates.

Next to the mutes in regard to degree of closure
are the class of so-called “fricatives,” defined above as
containing a rustling or friction of the breath through
a narrowed aperture as their main element. If the
lips are brought together in loose instead of close contact,
and the breath forced out between them, there is
heard an f-sound ; or, if the breath be intonated, a v-sound.
These, however, are not precisely our English
or French (nor the general German) f and ν ; for, in
the latter, the tips of the teeth are brought forward
and laid upon the lower lip, and the expulsion is made
between them ; giving a product somewhat differently
shaded, a dentilabial instead of a purely labial sound.
A relaxation of the lingual contact, in like manner,
gives the s and z sounds ; and that of the palatal gives
the German ch (its sonant counterpart is very rare).
Practically, however, it is found convenient to divide
the fricatives into two sub-classes : s and z have a peculiar
quality which we call sibilant or hissing ; and the
same is shared by the sh and the zh (in azure, vision)
sounds, which are produced farther back upon the
roof of the mouth, or in a more palatal position.
These two pairs, accordingly, we set by themselves,
as lingual and palatal “sibilants.” Then, along with
64the f and v, as akin with them, especially in their
dentilabial variety, we have the two English th-sounds,
surd in thin and sonant in then (written dh in the
scheme), real dentilinguals, produced between the
tongue and teeth. These four, with the (German)
ch-sound, we class as “spirants.” Historically, they
have a special kinship in that they are all alike frequent
products of the alteration of an aspirate mute ;
hence it is that they are so often, in various languages,
written withph, th, ch (=kh).

A like tendency to the points of oral action already
defined appears in the vowels, the opener tone-sounds.
An i (in pīque, pĭck) is a palatal vowel, made by an
approach of the flat of the tongue toward the palate
where its contact produces a k ; an u (rūle, pŭll) involves
a rounding approach of the lips, the organs
whose contact makes a p (although not without accompanying
action at the base of the tongue also). And
between a (far) and i stands e (thēy, thĕn), made by a
less degree of palatal approach, as ο (nōte, ŏbey) between
a and u. And again, the sound of fat, man
(œ in the scheme) stands between a and e, as that of
āll, whăt (ă in the scheme) between a and o. Representing
for the moment the pure fricatives by kh and
ph, we have the palatal series α œ e i kh k, and the
labial series a A o u ph p, which are true series all the
way through, made by gradually increasing degrees of
approximation of the same parts of the mouth until
complete closure is reached.

There is still one class to be noticed : that of the
semivowels, or sounds which stand nearly on the division-line
between vowel and consonant. I (pique) and
u (rule) are the closest sounds we can make with retention
of the predominant tone-quality which constitutes
65a vowel. But so close are they, that it is only necessary
to abbreviate them sufficiently, making them merely
starting-points from which to reach another vowel-sound,
in order to convert them into consonants, y and
w ; these differ, at the utmost, only infinitesimally in
articulating position from i and u. And with them belong
the r and l, lingual semivowels, used in many languages
also as vowels ; the l, even in English, in able,
eagle, etc. The r is produced between the tongue-tip
and the roof of the mouth, and is so generally trilled or
vibrated that trilling is apt to be given as its distinctive
characteristic ; the l sets the tip of the tongue
against the roof of the mouth, but leaves the sides
open for the free escape of the intonated breath.

We have one more pair of simple vowels, that in
hūrt and hŭt (ə in the scheme), the specific quality of
which is due to a dimming action along the whole
mouth rather than an approach at a definite point or
points, and which are thus a duller kind of a ;they
are put in the centre of the vowel-triangle rather because
they belong nowhere else than because they
belong precisely there.

The distinctions of long and short vowel, although
in English they always involve differences of quality
as well as of quantity, and the three compound vowel-sounds
or diphthongs, ai (“long i” of aisle, isle), au
(out, how), and Ai (oil, boy), are for simplicity's sake
left unnoticed in the scheme. And it remains only
to find a place in it, and a definition, for the somewhat
anomalous h. We have seen that in the classes of
mutes and fricatives the sounds go in pairs, one produced
by mere breath, the other by intonated breath,
forced through the same position of the organs ; while
this is not the case with the remaining and opener
66classes of sounds. We may define the difference in a
general way thus : after a certain degree of closeness is
reached, simple breath is sufficiently characterized to
give a constituent to the alphabet for every articulating
position ; short of that degree, only tone is fully
distinctive ; surd breath, though somewhat differentiated
in the several positions, is not enough so to
furnish a separate alphabetic element in each ; the
various breaths count only as one letter — namely, the
h. The h, the pure aspiration, is an expulsion of flatus
through the position of the adjacent letter, whether
vowel, semivowel, or nasal ; in English, it occurs only
before a vowel, or before w and y, in such words as
when and hue. It is, then, the common surd to the
three classes of sonant sounds just mentioned.

The scheme thus drawn up and described may be
taken as a general model, on the plan of which the
spoken alphabet of any language may best be arranged,
in order to the determination of its internal relations
and to its comparison with other alphabets. Though
not accurate to the very last detail, it exhibits more of
the relations of alphabetic sounds, and exhibits them
more truly, than any other plan that can be adopted.
And, restricted as it is in number of sounds, as compared
with the immense variety — not less than three or
four hundred — which enter into human speech, it yet
includes those sounds which make up the bulk of all
human speech, and of which many of the others are
slightly differentiated variations. The possible number
of human articulations is theoretically infinite ; but
practically it is rather narrowly limited ; and a system
like our own, which contains about forty-four distinctly
characterized sounds, is hardly excelled in richness,
among tongues ancient or modern.67

Our scheme is to be valued, especially, as putting
in a true light the relations of vowel and consonant :
which, though their distinction is of the highest importance
in phonetics, are by no means separate and
independent systems, but only poles, as it were, in one
continuous unitary series, and with, a doubtful or neutral
territory between them : they are simply the opener
and closer sounds of the alphabetic system. Upon their
alternation and antithesis depends the syllabic or “articulate”
character of human speech : the stream of utterance
is broken into articuli, ‘joints,’ by the intervention
of the closer sounds between the opener, connecting
the latter at the same time that they separate them,
giving distinctness and flexibility, and the power of
endlessly variable combination. A mere succession of
vowels passing into one another would be wanting in
definite character ; it would be rather sing-song than
speech ; and, on the other hand, a mere succession of
consonants, though pronounceable by sufficient effort,
would be an indistinct and disagreeable sputter.

Another advantage of the same arrangement consists
in its illustration of the general historical development
of the alphabet. The primitive language of our family
had not half the sounds given in the scheme ; and those
which it had were the extreme members of the system :
among the vowels, only a, i, and u, the corners
of the vowel triangle ; among the consonants, mainly
the mutes, along with the nasals m and n, which are
also mutes as concerns their mouth-position ; of the
whole double class of fricatives, only the s. The l was
not yet distinctly separated from the r, nor the w and
y from u and i. There has been a filling-up of the
scheme by the production of such new sounds as are
intermediate in character, made by less strongly differentiated
68positions of the organs. We may fairly
say that, in the process of time, with greater acquired
skill in the art of utterance, men's organs have come
to be able to make and use more nicely distinguished,
more slightly shaded tones than at first. This is no
mere loose poetic expression ; nor, on the other hand,
does it imply any organic change in the organs of
utterance. The case is only as in any other department
of effort : the higher skill is won by the advanced or
adult speakers, and the shape which they give to their
inherited speech becomes the norm toward which new
learners have to strive, attaining it when they can.

In the process, too, is involved an evident manifestation
of the tendency to ease. Not, indeed, that the
new sounds are in themselves any easier than the old ;
on the contrary, judged by some tests, they are harder :
they are not so readily learned and reproduced by children ;
they are not so frequently met with in the general
body of human languages. But they are easier to
the practised speaker, in the rapid movements of continuous
utterance, when the organs are making constant
quick transitions between vowel and consonant, between
opener and closer positions. To reduce the length of
swing of these transitions, by reducing the openness of
the open sounds and the closeness of the close ones, is an
economy which the articulating organs — of course, unconsciously — find
out for themselves by experience and
learn to practise. It is the most general kind of assimilating
influence exerted by consonant and vowel upon one
another : each class draws the other toward itself ; the
vowels become more consonantal ; the consonants become
more vocalic. Hence the prevailing direction of
phonetic change is from the extremities toward the middle
of the alphabetic scheme : the mutes become fricatives ;
69the a (far) is changed to e (they) and i (pique), or
to o (note) and u (rule). Movement in the contrary
direction is by no means unknown ; but it is exceptional
or under special causes : it is, as we have called it above,
the eddy in the current. The central classes, of nasals
and semivowels, which are least exposed to this general
movement, are also, on the whole, the least convertible
of the alphabetic sounds. To illustrate the effects of
the tendency : in Sanskrit (the least altered, phonetically,
of the tongues of our family), the a (far) is full thirty
per cent. of the whole utterance ; and we can easily
reason back to a time when a and the mutes were three
quarters of the sounds heard in continuous speech ; in
English, the most altered, a is only about half of one
per cent. of our utterance, while i (pique, pick) and ə
(hurt, hut), the closest and thinnest of the vowels, are
over sixteen per cent. ; and the fricatives have become
rather more common than the mutes (each class, about
eighteen per cent.). 15

We have called this a process of assimilation ; and
under the same comprehensive head may be grouped
the greater part of the other phonetic changes that
occur in language. The combinations of elements to
form words, their contraction by the omission of light
vowels, often bring into contact or into proximity
sounds which cannot be so uttered without too much
muscular exertion : it is eased by adapting the one to
the other. For example, many combinations of surd
consonant with sonant have that degree of difficulty
which we call impossibility (this is only a matter of
degree) ; and nothing is more frequent in all language
70than the interchange of surd and sonant utterance.
There is also a more general movement here : since the
sonant elements in connected speech are (including the
vowels) much more numerous than the surd, the general
weight of the assimilative force is in the direction
of sonancy, and surds are converted into sonants more
often than the reverse.

There is a degree of assimilation effected in vowels
by the consonants with, which they come into immediate
connection ; yet the cases are rather sporadic and
often doubtful. The influence of vowels on other
vowels, even when separated from them by consonants,
is more marked, and leads to some important
classes of phenomena. The difference between man
and men is ultimately due only to the former presence
of an i-vowel in the plural ending, which colored by
anticipation the preceding vowel : in Icelandic, the
effect is still plainly illustrated in the forms degi and
dögum from dagr. In the Scythian languages, on the
other hand, it is the final vowel of the base which
assimilates that of the following suffixes, as will be
noted hereafter (chapter xii).

Though assimilation is the leading principle in the
mutual adjustment of sounds, its opposite, dissimilation,
is not altogether unknown, as the close recurrence of
two acts of the same organs is felt as burdensome, and
avoided by the alteration of one of them.

Not only the parts of the same words, in their combination,
but also separate words, in their collocation,
affect one another ; and the influence expresses itself
particularly in their final elements. There are various
circumstances which help to condition this. In our
own and the majority of other families of speech, the
formative or less indispensable element comes last, and
71is the one least efficiently conserved by the sense of its
importance. Moreover, all experience shows that an
“open syllable,” one ending with, an open or vowel
sound, is easier, more “natural” to the organs, than a
closed one, ending with a consonant. A mute, indeed,
is hardly audible as final, unless the contact is broken
again with a puff of flatus ; and something of the same
disability clings also to the other consonants. The difficulty
is one which English-speakers can hardly realize,
since they allow freely every consonant in their alphabet
(with the accidental exception of the zh-sound) at
the end of a word, or of a syllable, before another consonant ;
but the Polynesian dialects, for example, admit
no groups of consonants anywhere, and end every
word with a vowel ; the literary Chinese has no final
consonant except a nasal ; the Greek, none save ν, σ, ρ
(n, s, r) ; the Sanskrit allows only about half a dozen,
and almost never a group of more than one ; the Italian
rarely has any final consonant ; the French silences, as
a rule, all save c, f, l, r ; the German tolerates no final
sonant mutes : and so on.

But the principle of ease does not find its sole exercise
in the work of assimilation. Nothing is more frequent
than for a language to take a dislike, as it were,
to some particular sound or class of sounds, and to get
rid of it by conversion into something else. We found
an example of this above in the old English h-sound of
cniht, etc. Most of the tongues of our family have cast
out the ancient aspirate mutes, changing them to simple
mutes or to spirants. The Greek early rejected the
y-sound, and then the w : the latter, as the “digamma,”
just prolonging its existence into the historical period.
Curious caprices, discordances between different languages
as to their predilections and aversions, come
72abundantly to light in this department of phonetic
change. Yet more exceptional and puzzling are the
cases of interchange between two sounds : for example,
the Armenian mutual exchange of surd and sonant
(Dikran for Tigranes, and so on) : to which the cockney
confusion of w and v, and of the presence and
absence of an initial h, furnishes a familiar, if undignified,
parallel. And of a comparative difficulty which
is at least as the square of the number of elements involved
is “Grimm's Law” of permutation of mutes,
illustrated above (p. 57). Phonetic science is not yet
far enough advanced to deal successfully with facts like
this ; no attempted explanation of the particular phenomenon
in question does much more than ignore its
real difficulties.

It must be carefully noted, indeed, that the reach of
phonetics, its power to penetrate to the heart of its
facts and account for them, is only limited. There is
always one element in linguistic change which refuses
scientific treatment : namely, the action of the human
will. The work is all done by human beings, adapting
means to ends, under the impulse of motives and the
guidance of habits which are the resultant of causes so
multifarious and obscure that they elude recognition
and defy estimate. The phonetist is never able to put
himself in an à priori position ; his business is only to
note the facts, to determine the relation between the
later and the earlier, and to account for the change as
well as he can, showing of what tendencies, in which of
their forms, it may be accounted the result. The real
effective reason of a given phonetic change is that a
community, which might have chosen otherwise, willed
it to be thus ; showing thereby the predominance of
this or that one among the motives which a careful
73induction from the facts of universal language proves
to govern men in this department of their action.

The tendency of phonetic change is so decidedly
toward the abbreviation and mutilation of words and
forms that it has been, suitably enough, termed “phonetic
decay.” Under the impulse to ease, the component
elements of speech are first unified, then unbuilt
and destroyed. It is the processes of combination (to
be treated of in the seventh chapter) that open a wide
field for the action of the tendency ; if language had
always remained in its original simple state, the sphere
of change would have been a greatly restricted one, and
the effects far less comparable to decay.

Before quitting the subject of changes of external
form, we must give a moment's attention to a class of
changes which bear a very different character, although
their cause has its points of analogy with those which
we have been considering : the class, namely, of which
we found instances in our modern ears and fared (p.
38), as compared with the earlier ear and fôr. When
phonetic corruption has disguised too much, or has
swept away, the characteristics of a form, so that it
becomes an exceptional or anomalous case, there is
an inclination to remodel it on a prevailing norm. The
greater mass of cases exerts an assimilative influence
upon the smaller. Or, we may say, it is a case of mental
economy : an avoidance of the effort of memory involved
in remembering exceptions and observing them
accurately in practice. The formal distinction of plural
from singular was one which our language was
never minded to give up. Of all the plural signs, the
one which had the most distinctive character was s.
The attention of the language-users became centred
upon this as an affix by which the plural modification
74of sense was made, and they proceeded to apply it in
words where it had not before been used ; and the
movement, once started, gathered force in its progress,
until it swept in nearly all the nouns of the language.
So with the verb. By the numerical predominance of
forms like loved from love, the addition of a d got itself
more conspicuously associated with the designation of
past time ; and men began to overlook the cases which
by right of former usage ought to be made exceptions.
Considerable numbers of verbs, in the middle age of
our tongue, thus changed, like fare, their old mode of
conjugation for a new. But the tendency is ever at
work, and on a small scale as well as a large ; and, of
course, especially among those whose acquisition of their
language has not been made complete and accurate.
Children, above all others, are all the time blundering
in this direction — saying gooder and badder, mans and
foots, goed and comed, even brang and thunk — and
items of such products creep not seldom into cultivated
speech. Its was made in this way, in the sixteenth
and seventeenth centuries ; we have gained thus
the double comparatives lesser and worser ; many are
led to say plĕad (like rĕad) instead of pleaded, and
even to fabricate such unsupported anomalies as proven
for proved. And the principle is often appealed to in
explaining the processes of earlier language-making.
The force of analogy is, in fact, one of the most potent
in all language-history ; as it makes whole classes of
forms, so it has power to change their limits.75

Chapter V.
Growth of language : change in the inner content
of words.

Wide reach and variety of this change ; underlying principles : looseness
of tie between word and meaning ; principle of economy ; class-names
and proper names. Illustrations : the planets and their kin.
Restriction of general terms to specific use ; extension of specific
terms to wider use. Figurative extension ; illustrations, head, etc. ;
forgetfulness of derivation. Growth of intellectual vocabulary from
physical terms ; of means of formal expression from material terms ;
auxiliaries, formal parts of speech ; phrases.

We come next to consider the other grand department
of change in the existing material of language —
namely, that of the inner content or meaning of words.
This is just as vast a subject as the preceding ; and, if
possible, even more irreducible in its immensity and in
finite variety to the dimensions of a chapter. The processes
of phonetic change have been worked out with
great industry by numerous students of language and
brought into order and system, and the comparatively
restricted and sensible movements of the organs of
speech investigated in order to form a concrete basis for
their explanation ; but no one has ever attempted to
classify the processes of significant change, and the
movements of the human mind under the variety of
circumstances defy cataloguing. Yet we may hope
76within reasonable space to lay out at least the foundations
of the subject, and to trace some of the chief
directions of movement.

It has been already pointed out that the separate
possibility of external and internal change rests upon
the nature of the tie, as a merely extraneous and unessential
one, which connects the meaning of a word with
its form. Were the case otherwise, the two kinds of
change would be mutually dependent and inseparable ;
as it is, each runs its own course and is determined by
its own causes ; even though the history of the two
may often touch, or go on for a time in close connection.
We also saw that words were assigned to their
specific uses (so far as it is possible to trace their history)
each at some definite time in the past, and for
reasons which were satisfactory to the nomenclature,
though they did not make the name either a definition
or a description of the conception ; and that the name,
once given, formed a new and closer tie with the thing
named than with its own etymological ancestor. We
took as illustration of this the word bishop, originally
simply ‘overseer ;’ claiming that it was only a specimen
of the way things regularly go on in language. It
is just so, for example, with priest, formerly πρεσβύτερος,
presbyter, elder, literally ‘older person ;’ so
with volume, though no longer ‘rolled,’ as when the
name was given ; with book, though not now a block of
‘beech’-wood ; with paper, now made of other material
than papyrus ; with gazette, which has ceased to be
sold for a Venetian ‘penny ;’ with hank, which has infinitely
outgrown the simple ‘bench’ of the money-changer
in the market-place, while the bankrupt has
vastly worse trials to endure than having his ‘bench
broken ;’ with candidate, though one in such a position
77is no longer expected to be ‘dressed in white ;’
with copper and muslin, which come now from other
quarters than Cyprus and Mosul ; with lunatic, even if
we discredit the moon's influence on the disorder ; with
Indian, though the error of the Spanish navigators,
who thought they had discovered ‘the Indies’ in
America, was detected a good while ago — and so on indefinitely.

We may see in all this something of the same principle
of ease or economy which we found to underlie
the changes of form. Were it altogether as easy, when
the shape of one's conception alters a little, or more
than a little, to fling away its old name and make a new
one ; were it as easy, when a new conception presents
itself, to give it an appellation before unheard-of, as to
stretch a familiar term a little to cover it, then might
there perhaps be no such thing as significant change in
human speech ; as it is, the old material of language is
constantly suffering extension and transferral to new
uses, obstructed by no too intrusive sense of original
meaning. Again, in virtue of the same principle, our
words are, almost universally, class-names. There is,
if narrowly enough regarded, a degree of individuality
about every being, thing, act, quality, which would justify
it in laying claim to a separate appellation ; but
language would be utterly unmanageable if it were
made up of such appellations ; and, in practice, having
named an individual thing, we apply the same name to
whatever other things are enough like it to form a class
with it. And thus, as we noted in the second chapter,
the acquisition of language is the adoption of certain
classifications ; herein consists a large share of its value
as a means of training. The classes, to be sure, are of
very different extent : there are even some — as sun,
78moon, God, world — which, have a natural restriction to a
single member. Then, again, there are classes of which
the individuals in their separateness rise to such importance
for us that we give each in addition a name belonging
to it as an individual only, or a “proper name,”
as we call it : such are the persons of our community,
our pet animals, streets, towns, and other localities, the
planets, months, week-days, and the like. In this class-use
is an additional facilitation of significant change ;
for every class is liable to revision, in consequence of
increased knowledge, keener insight, and consequent
change of criteria.

We shall best establish these fundamental principles,
and win suggestion of a classification for the
modes of change, by glancing over a series of illustrations.

In the olden time, certain heavenly bodies which, as
they circled daily about the earth from east to west,
had also a slower and more irregular movement in the
opposite direction among their fellows, were by a little
community in the eastern Mediterranean called planêtês,
because the word in their language meant ‘wanderer.’
From their use, we imported it into our own
tongue in the form planet, mutilated in shape and having
no etymological connection with any other of our
words. The class included the sun and moon not one
whit less than Jupiter and Mars ; it did not include the
earth. But within two or three centuries past, we have
acquired new knowledge which has led us to alter this
classification, and give a new value to its nomenclature.
We see now that, in a truer sense, the sun is not a
planet, but that the earth is one ; and planet has been
changed to mean, not a ‘wandering star’ as viewed
from earth, but a body that moves about a central sun.
79The moon is no longer precisely a planet, but a secondary
planet, a satellite. Having thus altered the conception
designated by moon, we are ready, when the telescope
discloses to us like satellites of other planets, to
convert this unique appellation into a class-name, calling
them all alike moons. So also with sun ; having
found that the sun is essentially akin with the fixed
stars rather than with the planets, we put him into the
linguistic class of fixed stars, or we call the fixed stars
suns.

The class of planets is one of those already referred
to, of which each separate member calls for an individual
designation, or “proper name.” Apart from
the sun and moon, however, they did not so impress
the popular mind as to receive popular titles, and it fell
to the learned, the astronomo-astrologers, to christen
them. These, though they did their work reflectively,
were not altogether arbitrary in their selection ; they
took the names of gods, since Sun and Moon were already
names of gods as well as of luminaries ; and they
distributed the names — Jupiter, Saturn, Mercury, Mars,
Venus — under the guidance of motives which we can
at least in part recognize : Mercury, for example, the
swift messenger of the divinities, had the most rapidly
moving and changeful of the class called after him.
Then, by a like transfer, the alchemists gave the god-and
planet-name to the most mobile of the metals.
And now, though the god Mercury is only a memory
of a state of things long gone by, Mercury and mercury
are still words of familiar use in our vocabulary ;
we even shut up mercury in a tube and bid him, as
Jupiter used to do, go up and down, to tell us what the
weather is. Again, the Frenchman calls the middle day
of his week ‘Mercury's day’ (Mercredi), though without
80being well aware of it, and yet less comprehending
why : it is because, in the distribution by the astrologers
of the hours through the whole week to the planets in
their order, the first hour of that day fell to the regency
of Mercury. Then, once upon a time, these
Latin day-names were mechanically turned into German
shape for the use of Germanic peoples, and Mercurii
dies became Woden's day, our Wednesday : and so with
the rest. Certainly a most curious history of transfer,
which brings out of a series of reflective acts of nomenclature
made by learned heathen — and not without
Christian aid, since the planetary day-names would have
remained to Europe, as to India, a mere astrologers'
fancy, but for Christianity and its inheritance of the
Jewish seven-day period as a leading measure of time —
a little group of some of the commonest and most truly
popular terms in our language ! The same words, moreover,
have been made to answer other purposes : the
astrologers held that a person born under the special influence
of a certain planet was characterized by a corresponding
disposition ; and those dispositions we still
call mercurial, jovial, saturnine ; martial and venereal,
on the other hand, come from the office of the divinities
themselves.

Again, we use sun and moon to designate ‘day’ and
‘month,’ saying “so many suns,” “so many moons.”
Here is simply a striking ellipsis : we mean really “so
many [revolutions of] sun or moon” — counting, however,
the revolutions on different principles ; else a sun
would be a ‘year.’ Then month, which is only a derivative
form of moon, has been transferred to designate
an arbitrary period of twenty-eight to thirty-one
days, having nothing whatever to do with the moon's
movement. Further, a moon (or lune) is in fortification
81a crescent-shaped outwork : an analogy, this time, of
shape merely. Nor is it meant to imply that the moon
is always, or usually, of this shape ; but only that she
is the most conspicuous object in nature that ever assumes
the shape. If we want to be more precise, we
say “crescent-shaped.” But here also is an ellipsis, and
of the most striking kind ; for crescent literally means
simply ‘growing,’ and does not contain even a hint of
the moon. Moreover, the moon does not have this
shape all the time she is “growing,” but only at a particular
period, and she has it just as much when decreasing
as when increasing ; so that crescent really means
‘[resembling the moon at a certain stage of her] growing
[as also of her waning].’ It is good English, too,
to talk of a moon-struck idler as mooning around, although,
we should indignantly deny the belief in lunar
influences which suggested the expressions.

This may seem like an aimless roaming through one
department of our vocabulary ; but its heterogeneousness
is due to the character of the facts with which we
have to deal, and is an important part of the value of
the illustration. It is simply impossible to exhaust the
variety of significant change in linguistic growth : there
is no conceivable direction in which a transfer may not
be made ; there is no assignable distance to which a
word may not wander from its primitive meaning.
There is no such, thing as a concise and exhaustive classification
of such variety ; all we can do is to point out
some of the main divisions, the leading directions in
which the movement goes on, neglecting the unclassified
and perhaps in part unclassifiable residue.

One of the largest classes (already more than once
hinted at) has a striking example in crescent. Crescent,
‘growing,’ is a word of the widest application ; a young
82child or tree, an aggregating crystal, a new-built fire, a
beginning reputation, an evolving cosmos, are really as
much crescent as a young (so, by a figure, we call it)
moon. To seize upon the word as specific title of the
growing moon, then, is to commit a very bold and arbitrary
act of restriction. But the act is also open to
objection on another side. It takes account of only a
single, and that a very trivial, characteristic of an object
which has many others. All we can say in reply is that
nomenclature is a free and easy process, and that such
objections count for nothing as against the demands of
convenient expression. The case was the same with
bishop, ‘overseer,’ as we saw above ; it was the same
with green, ‘growing ;’ it was the same with planet,
‘wanderer.’ It is believed by the etymologists that
moon itself comes in a similar way from a root meaning
‘measure ;’ our satellite having been thus designated,
in remote ages, because of her office in measuring
the longer intervals of time : “so many moons.”
Certainly, her Latin name luna is for lucna, and related
with lux, and so describes her simply as a ‘shiner.’
And sun goes back, it is believed, to an equivalent
source. Comparative philology claims to have shown
(as will be noticed hereafter) that the earliest appellations
of specific things were in general won in precisely
this way, the germs of speech being expressions for
acts and qualities. However that may be, it is certain
that, through the whole history of language since, the
method has been in constant use : epithets of things,
representing some one of their various attributes, become
the names of things, through every department
of nomenclature. Our etymologies are apt to bring us
back finally to some so general, comprehensive, colorless
idea, that we almost wonder how it can have given
83birth to such strongly-marked progeny. All the varied
and definite meanings of post (to take a further example
or two) go back to the sense of ‘put, placed.’ The
idea of rolling is specialized into the muster roll and
the breakfast roll, the roll of the drum and rolls of
fat ; by a longer route, it comes to us in the form of
the actor's rôle ; and a slight addition makes of it control,
of which the connection with its original escapes
all but skilled and curious eyes.

Another leading principle, of the first order of importance,
is somewhat contrary in its effects to that
which we have been considering : it is the principle of
extension, as opposed to restriction, of the sphere of
meaning of a term. A name won by specialization begins
an independent career, which ends in its gaining
the position of head of a tribe. Mr. Miller, named by
the specializing process from his vocation, becomes the
father of a multitude of Millers, so named from their
relation to him, without the least regard to their vocations.
And he may turn out the founder of a sect, who
shall call themselves Millerites after him, and make his
name as conspicuous an element in the nomenclature of
theology as is already that of Arius or Nestorius. The
butterflies were first named in the species which showed
itself butter-colored as it flew : the title is extended,
heedless of the differences of color, to every other kindred
species. Our recent examples showed us sun and
moon made class-names. Crescent develops a group of
new uses out of the fortuitous presence of the figure
on the Mohammedan standard. No one knows precisely
why the rose was so entitled : the botanist has made
it the type of a whole order of quite diverse plants,
which he terms rosaceœ, ‘rose-like.’ A great part of
our acquisitions of new knowledge go to swell old established
84classes, expending themselves, so far as language
is concerned, in the extension of existing class-names.
To take an example of the most obvious kind : the discovery
of every new animal or plant or mineral stretches
a little not only the scope of those widest terms, but
also of a whole series of subordinate ones. And sometimes
the change rises to conspicuous value. The zoologist's
conception of horse, for example, has undergone
no slight modification by the recent discovery in
the American West of numerous fossil species, of
greatly varying size and structure. Every exploring
naturalist, in fact, is all the time illustrating, in an
openly reflective way, in his naming of species, the two
principles which direct a great part of the world's less
conscious nomenclature. Having in his hands a new
plant, he at once proceeds to classify it : that is to say,
to determine of what current class-names it must swell
the content : he finds it, we will suppose, a plant, and a
phenogamous, a dicotyledonous, a rose-like plant, and
finally a rubus or ‘blackberry.’ But it has peculiarities
which entitle it to a specific designation ; and this must
be gained by the other method : the nomenclator selects
the quality which he will describe, and christens it megalocarpus,
‘big-fruited,’ gracilis, ‘elegant,’ or the like ;
or he gets a suggestion from the locality, the situation,
the circumstances of discovery ; or he connects it with,
some still more extraneous matter : so, for instance, he
compliments his friend Smith by naming it Smithii.

The extension of a name's application, however, involves
a great deal that is far less plain and legitimate
than all this. Not only a true accordance in generic
character, but relations of an infinitely looser kind, are
used to tie together the classes that go under one name.
We saw lately a heathen god, a planet, a metal, a temperament
85and a day of the week, all forced into unnatural
union under the title mercury. Since fruit is
apt to be green when not fully ripe, green becomes a
synonym for ‘unripe’ (and so we can commit the familiar
linguistic paradox that blackberries are red when
they are green) ; and then, in less elegant diction, it is
again shifted to signify ‘immature, not versed in the
ways of the world.’ Such transfers we are wont to call
figurative ; they rest upon an apprehended analogy,
but one generally so distant, subjective, fanciful, that
we can hardly regard it as sufficient to make a connected
class. Instances of this kind lie all about us, in our
most familiar words ; and this department of change is
of so conspicuous importance in language-history that
we must dwell upon it a little longer. Our minds delight
in the discovery of resemblances, near and remote,
obvious and obscure, and are always ready to make
them the foundation of an association that involves the
addition of a new use to an old name. Thus, not only
an animal has a head, but also a pin, a cabbage. A bed
has one, where the head of its occupant usually lies —
and it has a foot for the same reason, besides the four
feet it stands on by another figure, and the six feet it
measures by yet another. More remarkable still, a river
has a head : its highest point, namely, where it heads
among the highlands — and so it has arms ; or, by another
figure, branches ; or, by another, feeders ; or, by
another, tributaries ; and it has a right and left side ;
and it has a bed, in which, by an unfortunate mixture
of metaphors, it runs instead of lying still ; and then,
at the farthest extremity from the head, we find, not its
foot, but its mouth. Further, an army, a school, a sect,
has its head. A class has its head and its tail ; and so
has a coin, though in quite a different way. A sermon
86has its heads, as divided by their different headings ;
and we can beg to be spared anything more “on that
head.” A sore comes to a head ; and so, by one step
further away from literalness, a conspiracy or other disorder
in the state, the body politic, does the same. We
give a horse his head, which he had before our donation ;
and then we treat in the same way our passions —
that is to say, if by their overmastering violence we lose
our heads. And so on, ad infinitum.

These side or figurative uses of a word do not perplex
us ; they do not even strike us as anything out of
the way ; they are part and parcel of the sphere of application
of the word. For it is an important item in
this process of transfer that we gradually lose our sense
of the figure implied, and come to employ each sign as
if it had always been the simple and downright representative
of its idea. Here we see again the willingness,
which has been already pointed out, and which is
essential to the prosperous development of a language,
to forget the origin of a name when once it is won, to
let drop the old associations and suggestions which belonged
to it in virtue of its etymology, and invest it
with a new set appertaining to its present use. Perhaps
there is in English hardly a more striking example
of this than our word butterfly, a name of utterly prosaic
and trivial origin, but which has become truly poetic
and elegant, as we think in connection with it of
the beautiful creatures it designates, and not one in a
thousand has ever had come into his head the idea that
it literally means ‘a fly of butter-color.’ The relics of
forgotten derivations, of faded metaphors, are scattered
thickly through every part of our vocabulary. It is, to
our apprehension, in the nature of a word to have its
figurative as well as its literal uses and applications, we
87inherited our vocabulary in that condition ; and, by new
discoveries of analogies and new transfers of meaning,
we are all the time adding to the confusion — if it were
a confusion. Sometimes the connection between the
different senses is obvious on the least reflection ; sometimes,
again, it is so obscure that we cannot find it, or
that we conceive it wrongly ; ordinarily, we do not concern
ourselves about the matter ; we use each word as
we have learned it, leaving to the lexicographer to follow
up the ramifications to their source in its primitive
or etymological meaning.

A conspicuous branch of the department of figurative
transfer, and one of indispensable importance in
the history of language, is the application of terms having
a physical, sensible meaning, to the designation of
intellectual and moral conceptions and their relations.
It is almost useless to attempt to illustrate this ; the examples
would come crowding in too numerously to be
dealt with : we will merely notice a few of those which
happen to be offered in the preceding paragraph. Perplex
means ‘braid together, intertwine.’ Simple is
‘without fold,’ as distinguished from what is double, or
‘two-fold ;’ in simplicity and duplicity we have a
moral contrast more distinctly brought to view ; application
contains the same root, and denotes an actual
physical ‘folding or bending to’ anything, so as to fit
it closely ; while imply intimates a ‘folding in.’ Important
means ‘carrying within ;’ that is, ‘having a
content, not empty.’ Apprehension signifies literally
the ‘taking hold’ of a thing. Relation is a ‘carrying
back,’ as transfer is a ‘carrying across’ in Latin, and
metaphor nearly the same thing in Greek. To invest is
to ‘put into clothes ;’ to develop is to ‘unwrap.’ Trivial
is what is found ‘at the street-crossings ;’ anything
88is obvious which meets us ‘in the way,’ which occurs to,
or ‘runs against’ us. Derivation involves the curiously
special idea of drawing off water ‘from the bank’ of
a river, for irrigation or the like. To suggest is to
‘carry under,’ or supply, as it were, from beneath, not
conspicuously — and so on. All these are from the
Latin part of our language, which furnishes examples
in the greatest abundance, because our philosophical and
scientific vocabulary comes mainly from that source ;
but there is plenty like it in the Saxon part also. Wrong
is ‘wrung’ or ‘twisted,’ as its opposite right is ‘straight ;’
and downright involves the same figure as upright, as
having nothing oblique or indirect about it. A striking
example needs no comment. To forget is the opposite
of to get, but signifies only a mental loss. We see
things that never come before our bodily eyes. And
point out, let drop, follow up, lay down, come into the
head, out of the way, are instances of phrases that show
plainly a similar shift of application. In fact, our
whole mental and moral vocabulary has been gained
precisely in this way ; the etymologist feels that he has
not finished tracing out the history of any one of its
terms until he has hunted it back to the physical conception
in which, by the general analogies of language,
it must have had its origin.

Thus, as the general movement of human knowledge
is from the recognition of sensible objects to an
ever finer analysis of their qualities and determination
of their relations, and to the apprehension of more recondite
existences, objects of thought, so, as the accompaniment
and necessary consequence, there is a movement
in the whole vocabulary of language from the
designation of what is coarser, grosser, more material,
to the designation of what is finer, more abstract and
89conceptional, more formal. Considered with reference
to the ends rather than the methods of expression, there
is no grander phenomenon than this in all language-history.
But the evolution of the intellectual vocabulary
is only one division of the movement ; there is another
to which a few moments' attention must be given.

We have a verb, be, bearing the purely formal grammatical
office of connecting a subject with its predicate.
Such a connective is wanting in many languages, which
are obliged simply to set the two elements side by side,
leaving their relation to be supplied by the mind. Its
conjugation is made up of various discordant parts ;
which, however, agree in the quality of derivation from
roots having a distinct physical meaning : am, is, are,
come from as, which signified either ‘breathe’ or ‘sit ;’
was, were, from vas, ‘abide ;’ be, been, from bhû,
‘grow.’ The French has filled up its scheme of the
same verb from the Latin stare, ‘stand.’ The development
of meaning here is analogous with what we have
been considering, a case of transfer and extension — extension
so wide that it has effaced all that was distinctive
in the words ; we may call it an attenuation, a fading-out,
a complete formalizing, of what was before
solid, positive, substantial.

The same general connective be, when used with the
past participle of a transitive verb, becomes an “auxiliary,”
making a whole conjugation of what we call
“passive” forms — “I am loved,” etc. ; with a present
participle, it makes a like scheme of “continuous” or
“imperfect” tenses — “I am loving,” etc. It thus enters
just as fully into the service of formal grammatical
expression as the formative endings of languages of
other habit than ours. We have many other words of
which the history and present application are nearly
90the same. There is do, which, from the original physical
notion of ‘set, place,’ has been extended and formalized
into expressing efficient action of every kind —
do good, do one's best, do to death, and so on ; and
which also does service as verbal auxiliary — I do love,
did I love ? etc. Again, the Latin root cap (carpere)
means ‘seize, grasp.’ Its Germanic correspondent is
hab, in Gothic haban, German haben, our have. But
here the more physical sense of ‘grasp’ has almost disappeared
(we have it in Germ. handhabe, our haft, the
part of an instrument that is ‘grasped’ by the hand) ;
in its place has come the more conceptional one of ‘possess.’
So also with the Latin habere, the relation of
which to capere on the one hand and haben on the other
is a puzzle to the etymologists. Finally, this too has
been turned to use in verbal expression, and by a transfer
which, though illustrated in the history of many
languages, must be called a very remarkable one. Present
possession often implies past action : habeo cultellum
inventum, habeo virgulam fissam, habeo digitum vulneratum,
‘I possess my knife found (recovered after
loss), I possess a twig that is split, I have a wounded
finger :’ here the several conditions have been preceded
by the several acts, of finding, splitting, wounding. On
this absurdly narrow basis is built up the whole immense
structure of the “perfect”-tense expression :
the phrase shifts its centre of gravity from the expressed
condition to the implied antecedent act ; and I
have found the knife, ich habe das Messer gefunden,
j'ai trouvé le couteau, become indicators of a peculiar
variety of past action contemplated as completed : further
examples are the Sanskrit kritavân, ‘[I am] possessing
[something] done,’ i. e. ‘I have done ;’ and
Turkish dogd-um, ‘striking mine,’ i. e. ‘I have struck.’
91The next step is to forget how have came by its “perfect”
meaning, and to use it with all sorts of verbs,
where an etymological analysis would make nonsense :
as in I have lost the knife, I have lived (German and
French the same) ; and, in English, even I have come,
where the other languages still say, more properly, ‘I
am come.’

But the same verb has other auxiliary work to do.
The phrases habeo virgulam ad findendum, j'ai une
verge à fendre, ich habe ein Aestchen zu spalten, I have
a twig to split (for splitting), as plainly imply a contemplated
future action. They become formal verbal
expressions when, by a like shift of emphasis and apprehended
connection with that noted above, the construction
is changed to I have to cut a twig, and the
noun is viewed no longer as object of the have, but
rather of the other verb, the infinitive ; and yet more
completely when (again as above) the construction is so
extended that we say I have to strike, I have to go, I
have to be careful. We thus have a phrase denoting
obligation to future action, developed out of the same
expression for ‘seizing’ which is also used to denote
past action. The French has gone still further. Not
emphasizing, as we do, the idea of obligation, it uses
the same phrase as simple expression of futurity ; and
more, it combines the auxiliary into one word with the
other verb — je fendrai (for je fendre ai, i. e. j'ai à
fendre) ; in which no French speaker, unless philologically
educated, ever recognizes the elements of the combination.

Once more, the English is peculiar in expressing a
causative sense by the same agency : I had my horse
shod, I will have the book bound, point to a different
aspect of the action, setting it forth as something
92brought about, though not executed, by the actor. It is
merely a turning-up to view of another of the many
implications involved in the state of possession.

All our verbal auxiliaries come after a like fashion.
Behind our shall and will, as signs of future action, lies
a history of transfers and extensions. One step back,
I shall means ‘I owe, am under obligation ;’ I will ‘I
intend, purpose.’ Both are examples of that important
little class of Germanic verbs called “preterito-presential,”
because (by a change just the opposite of that
which we noticed above) they have won their present
meaning through a “perfect” one. And shall, it is
claimed, goes back finally to ‘I have offended,’ and
hence ‘am under penalty ;’ will, to ‘I have selected’
(yet more primitively, ‘have enclosed or surrounded’).
The Greek κέκτημαι ‘I have acquired’ (colloquial English,
I have got), for ‘I possess,’ is a parallel here : indeed,
both Greek and Sanskrit have one of the very
verbs that compose the Germanic class : Skt. véda, Gr.
οἶδα, Goth. wait, Germ. weiss, ‘I wot or know :’ literally,
‘I have seen.’ And the Latin furnishes a very
notable parallel to the shifts of construction we have
been instancing, in its use of the accusative as “subject”
of an infinitive : it all grew out of an inorganic
extension of such constructions as dicit te errare, ‘he
declares you to err.’ Toward this we have in English
at least a near approach in phrases like “for him to err
is a rare thing,” where we have almost forgotten that
for logically connects him with rare : “to err is a thing
rare for him.” Another kindred case is the infinitive
in passive sense in German causative phrases : er liess
sich nicht halten, ‘he did not let himself be held ;’ literally,
‘did not let [any one] hold him.’

This kind of change is by no means limited to verbal
93constructions, as a few examples from other parts of
the grammar will show. In Anglo-Saxon there was no
such word as of, as distinguished from off ; their separation,
in form and meaning, is a piece of very recent
word-history. Off is the earlier sense, as the more material :
though itself, as preposition, a sign of relation,
and therefore formal as compared with our general vocabulary.
But in of we have all limited and definable
relation extinguished ; the word is a token of the most
indefinite appurtenance, the absolute equivalent of a
genitive case-ending, a link between a noun and its
modifying noun, sign of the adjective relation of one
noun to another. The French de has a history not unlike
this. Almost as striking an example is our for,
originally the same word with fore, ‘before, in front
of ;’ in German the word has taken on a threefold form
for its various offices, in vor, für, and the inseparable
prefix ver — each of more attenuated quality than its
predecessor. To retains in general its ancient office as
designating approach ; but as “sign of the infinitive”
it is as purely formal as of itself ; in to have, for example,
it is nothing more than a kind of modern substitute
for the old ending an of haban : we have absolutely
lost from memory its real value, as that of a preposition
governing a verbal noun.

But there is another shift of construction lying back
of the whole class of prepositions. The oldest of them
were originally — as many of them still continue also
to be — adverbs, modifiers of verbal action, only aiding
to determine the noun-case which that action should
take as its further adjunct. Here is a whole part of
speech, of an especially formal character, developed from
those of more material aspect and office. The conjunctions
are another case of the same kind, though into
94the details of their history we have no time here to enter.
And the articles, sometimes ranked as a separate
part of speech, are likewise altered and faded words :
their originals, to be sure, were formal enough ; but
they are etherealized formals : the definite article is
a demonstrative, from which the full demonstrative
force has been withdrawn ; the indefinite article comes
by a similar process of attenuation from the numeral
‘one.’

The great variety and prominent importance of this
department of change of meaning tempt to protracted
illustration ; and no brief array of examples can do it
justice : but we must content ourselves with only one
more. Alongside the conjunctions, the relative pronouns
are by far the most important of the connectives
by which we bind together separate assertions, making
a period out of what would otherwise be a loose aggregation
of phrases. They are pronouns with conjunctive
force ; they fasten distinctly to their antecedent an assertion
which would otherwise be connected with it
only by implication. There are plenty of languages in
the world which have no such syntactical apparatus ;
and we, too, could make shift to get on well enough
without it. To say “my friend had had a fever ; he
was not quite recovered ; he was looking pale and ill,”
is fully sufficient to enable the hearer to combine the
circumstances in their proper relations. We only put
into expression the necessarily implied mental act when
we say “my friend, who had had a fever from which
he was not quite recovered, was looking ill ;” and we
have no small variety of other ways of putting the
same thing : “he was looking ill because (or, for) he
had had” etc. ; or, “my friend, being not yet recovered
from a recent fever, was looking ill,” and so on.
95The various modes of statement are devices for presenting
to more special attention one and another aspect of
a fact and its causes ; their possibility is an added decoration
rather than a substantial resource of speech ; they
serve a rhetorical purpose. But the relatives, which,
though not indispensable, are an agency we could hardly
afford to miss, are only a comparatively recent acquisition.
They are demonstratives and interrogatives put
to a new use ; employed first with pregnant allusion to
an antecedent, then gaining such allusion as an essential
element. The construction was in a forming and doubtful
state in our earliest English, and who and which
won their relative force only considerably later.

It is by no means only in verbal phrases and other
examples of the reduction of terms of independent
meaning to formal value that language exhibits its characteristic
tendency toward oblivion of original meaning
and disregard of etymological concinnity. Most tongues
are full of idiomatic phrases, which, when we attempt
to analyze them, are often obscure or meaningless or
absurd, and which nevertheless constitute no small part
of the strength and charm of expression. Take place
is a fair English example ; the same expression in German,
Platz nehmen, means ‘sit down,’ while to represent
our meaning the German says rather Statt finden,
‘find stead.’ In French we may instance avoir beau,
literally ‘to have beautiful,’ used to intimate the uselessness
of an action : il a beau s'excuser, ‘he tries in
vain to excuse himself ;’ or en vouloir, literally ‘wish
about it,’ but meaning ‘bear a grudge.’ And between
the three equivalent expressions there is, il y a, literally
‘it has there,’ and es gibt ‘it gives,’ it is hard to
choose the one which implies the most curious twist
of meaning. The very abundance and heterogeneousness
96of the material here discourage more extended
illustration.

It is, as has been already said, impossible to exhaust
the variety of significant change in linguistic growth.
Whole volumes, full of interest and instruction, have
been produced upon this subject alone ; and if our object
were general interest and instruction, we should
not quit the theme here. We should dwell, for instance,
upon the curious fate which, while some words
fade to the thinnest skeleton, almost shadow, of substantial
value, crowds others with, pregnancy and force
— like home, comfort, tact (literally ‘touch’), taste, humor
(‘moisture’) ; upon the contrast between words
which from a low or an indifferent origin rise to dignity,
and those which from a respectable origin sink
into contempt (we had above, p. 40, an example of both
these changes in the same word, our knight and the
German knecht) ; between words which become so conventionally
inexpressive that we seek for newer and
more positive phraseology, and those which, dealing
with delicate subjects, become too directly suggestive,
and are replaced in refined usage by others which hint
more remotely at the intended sense ; between words
which for no assignable reason become the fashion, and
others which as causelessly come to be looked askance
at and avoided. Some of these cases will call for remark
farther on, in other connections : for the present
we must be satisfied with having noticed at least the
principal tendencies, those which have most influence
on the growth of language.

Chapter VI.
Growth of language : loss of words and forms.

Loss of words ; its causes ; obsolescent and obsolete words. Loss of
meanings. Loss of grammatical forms and the distinctions conveyed
by them ; examples ; excess of this loss in English.

We saw above (in the third chapter) that loss of
what had constituted the material of a language was an
appreciable element in that constant change and development
which we called its growth. Even such a process
of subtraction is fairly enough to be reckoned as a
part of growth ; just as the growth of organic beings
consists in removal as well as in resupply. And our
preliminary illustrations showed us that the loss might
consist either in the disappearance of complete words
from a vocabulary, or in the disappearance of the signs
of grammatical distinction.

The reduction of a vocabulary by loss of its words
is a matter so simple that we shall not need to spend
much time upon it.

As all the items of a given language are kept in existence
only by being taught and learned, it is evident
enough that the cessation of this process of tradition with
regard to any item will bring about its annihilation.
Existence, in speech, is use ; and disuse is destruction.
Whatever leads to disuse leads to loss ; and there is
98nothing else that can have that effect. And there are,
accordingly, two principal ways in which loss can occur.

In the first place, the disappearance from before the
attention of a community of the conceptions designated
by certain words occasions the disappearance of those
words. If anything that people once thought and
talked about comes to concern them no longer, its
phraseology goes into oblivion — unless, of course, it be
preserved, as a memory of the past, by some of those
means which culture supplies. It has been so, for example,
with the old heathen religion of our Germanic
ancestors. Once, the names of Thor and Woden, of
Tuis and Freya, and the rest of them, were as common
on English lips as those of Christ and the Virgin Mary,
of St. Peter and St. Paul, are nowadays ; but, save for
their fortuitous and generally unrecognized retention in
the names of the days of the week, they have become
extinct in the speech of common life, and are known
only to curious students of antiquity. The same thing
is true of a host of words belonging to the vocabulary
of the ancient arts and sciences, the ancient institutions
and customs. The technical terms of chivalry mostly
fell out as those of modern warfare came in ; those of
astrology, as this was crowded from existence by astronomical
science. Only, we have here and there, not
always consciously, in our present speech, reminiscences
of the old order of things, in the shape of words transferred
to new uses. Even so common and indispensable
a term as influence is said to be of astrological origin,
denoting in its early use only the bearing of the
heavenly bodies on human affairs ; disaster is etymologically
a mishap due to a baleful stellar aspect ; and
we have already noted jovial, saturnine, mercurial, as
names for dispositions that were regarded as produced
99by the influence of planets. In like manner, part of
the vocabulary of hawking, when that mode of securing
game went out of use, was transferred to the new
apparatus : as an especially noticeable instance, musket
was the name of a certain small hawk.

But, in the second place, words are crowded out of
use, and so out of life, by the coming into use of other
words which mean the same thing, and which for some
cause, definable or not, win the popular favor, and supplant
their predecessors. Of this process we found
examples in our specimen-passage : the honest Saxon
derivatives or compounds Hœlend, reste-dœg, learning-cnihtas,
are replaced in our usage by the outlandish
terms Savior, sabbath, disciple, and have themselves
disappeared. And this is but a specimen of a process
of wide reach and abundant results in English. In consequence
of the Norman conquest, a considerable body
of French words was poured in upon our language, and
gradually accepted and put to service as an integral part
of it. To no small degree, indeed, as a direct enrichment
of English, speech, by furnishing expression for
new ideas, or French synonyms for Saxon words, each
useful in its own style and connection : like brotherly
and fraternal, outlandish and foreign, forgive and pardon,
rot and decay, hue and color, stench and odor, foresight
and providence. But to a considerable extent also
there was an over-enrichment, which, the requirements
of practical use did not justify ; and the intrusion of
the new caused an extrusion of the old. Thus a host
of Saxon words gave place to substitutes of foreign
origin : nothing would be easier than to add to the
examples given above numberless others, like wanhope
displaced by despair, ayenbite by remorse, inwit by
conscience, and so on.100

Nor is it by foreign importation alone that words of
native growth become superfluous, and are dropped out
of a language. There are cases in abundance of a
word's simply going out of fashion, becoming obsolescent
and then obsolete, by an act of supersession attributable
only to what we call chance or caprice. We
have one or two fair examples of it in our specimen-passage,
as already pointed out (pp. 39, 43) : namely
fôr and sôth. In Anglo-Saxon, the verb faran ‘fare,’
was in frequent and familiar use in the simple sense of
‘go’ or ‘pass.’ Gân, ‘go,’ was also good English, with
its irregular preterit eode, ‘went ;’ likewise gangan,
‘gang,’ with gêng, ‘ganged ;’ and wendan, ‘turn, wend,’
with wende, ‘turned, went.’ Out of this, as it was
found, somewhat wasteful provision of words for
‘going,’ our later English has made arbitrary selection
of go and went, dropping the rest — or else, as in
the case of fare, restricting them to special uses. In
a similar way, equus has gone out of use as name for
‘horse’ in all the descendants of the Latin, and has
been replaced by caballus, which was originally a word
of inferior dignity, like our nag ; although, in chivalry,
etc., it has since come to honor enough : so magnus has
been superseded by grandis, and pulcher by bellus ; and
so, in French, vulpes has been given up for renard,
which is the German Reinhart, a proper name, by
which a fox was at one time popularly called, much as
we call a dog “Tray.” It may even happen that an
important word dies out, without provision of any full
substitute : so the Anglo-Saxon weorthan, corresponding
to the German werden, ‘become.’ Doubtless the transfer
to its present meaning of become (literally ‘come by,
get at, get’) caused the oblivion of the older and more
legitimate synonym ; and with this went the possibility
101of such distinctions as the German makes abundantly
by means of werden : especially, that of the true passive
es wird gebrochen, ‘it is getting broken,’ i. e. ‘is undergoing
fracture,’ as against es ist gebrochen, ‘it is broken,’
i. e. ‘has undergone fracture ;’ whence, further, the
necessity for such awkward, but naturally formed and
really unavoidable phrases as it is being broken.

By these means, there is in every language a certain
amount of obsolescent material, in various stages : some
words that are only unusual, or restricted to particular
phrases (like stead, in in stead alone) ; some that belong
to a particular style, archaic or poetical ; some that have
become strange and unintelligible to ordinary speakers,
though formerly in every-day use ; some that survive
only in local dialects. And the older records of any
tongue, if preserved, show words in greater or less number
that are gone past recovery.

It is hardly necessary even to spend, in passing, a single
word upon the somewhat analogous loss, by words
and phrases, of their old meanings, although this may
also involve, in its manner and degree, a reduction of
the resources of expression. The examples of transfer
of meaning given in the last chapter have shown also
sufficiently that the process is not always, though it may
be usually, an addition of new meanings without an
abandonment of the old. It may be, too, that the substantial
sense of a word remains to it, while its accessory
suggestiveness is altered ; so when Milton speaks
of ladies who “from their eyes rain influence,” we miss
the whole poetic significance of the line if we do not
know the astrological allusion it involves. In reading
older authors, we are constantly liable to this loss or
misunderstanding, often skimming a mere surface comprehension
off that which has a profound meaning, or
102deluding ourselves with a belief that we understand
where the real sense escapes us.

A subject of greater consequence and deeper reach
in language-history is the loss of old distinctions of
grammatical form. Of this, our illustrative sentence
brought to light several striking examples, already
briefly noticed by us. By the wearing off, under the
prevalent phonetic tendency, of the old infinitive ending
an (Middle English and German en), our infinitive
as a verbal form is no longer different from the root of
verbal inflection. And yet we do not fail to appreciate
distinctly enough the idea of the form, and have even
(as we saw) fabricated a new sign to as a kind of substitute
for the obliterated suffix. Again, having lost all
such signs of plurality as the final on of ongunnon, we
no longer distinguish the plural of a verbal tense formally
from the singular except in am and are, was and
were : yet here, also, the difference made by us between
singular and plural nouns and pronouns, scantily supplemented
by the absence of a personal ending in they love
as compared with he loves, seems still to keep up in full
life the old distinction. The se and thâ, however, as
singular and plural respectively, and the former of them
as specifically masculine (the feminine was seo, and the
neuter thœt), are examples of a class of grammatical distinctions
which have gone by the board, swept clean
away, so that we have forgotten that they ever existed :
namely, the variation of an adjective word for gender
and number and case. The Anglo-Saxon adjective had
a fuller inflection than the German, almost as full a one
as the Greek or Latin ; it even had a double one, definite
and indefinite, like the German ; and the language
still retained the old system of concord, of formal correspondence
between a substantive and its qualifier or
103representative, which, founded on the original identity
of substantive and adjective, is one of the glories of a
completely inflective language ; but since we have lost
it, we have never thought of missing or regretting it ;
and no one of us would be easy to convince that, when
we say good men, there would be anything gained by
giving the word good a different form from that which
it has in good man,. And yet less, from that which it
has in good women. For the distinctions of gender
have been extirpated even in our nouns. To us, the
name or appellation of a person is masculine or feminine
only according as the person is male or female ; and of
sex in the lower animals we make very small account ;
while our Anglo-Saxon ancestors were as much under
the dominion of that old artificial grammatical distinction
of all the objects of thought as masculine, feminine,
and neuter, on a basis only in small part coinciding with
actual sex, as are the Germans now, or as were the
Greeks and Latins of old : it was one of the original
and characteristic features of that language from which
all these, and most of the other tongues of Europe, are
descended. The French has suffered the same loss only
partially, having saved the distinction of masculine from
feminine, but confounded neuter and masculine together
by the obliteration of their respective marks of difference.
But also the old scheme of cases in our nouns
has become a wreck and a remnant, although the distinctions
on which it is founded are just as necessary
a part of language as ever. The English has no dative,
and no accusative except in a few pronouns (him, them,
whom, etc.) ; the French is still poorer, having not even
a possessive ; although it makes in a few pronominal
words a somewhat evanescent distinction of subject and
object. We have also nearly parted with our subjunctive,
104which in German is as rich in forms as is the
indicative.

The English is, in truth, of all the languages of its
kindred, the one which most remarkably illustrates that
mode of linguistic change consisting in the loss of formal
grammatical distinction by synthetic means ; there
is no other known tongue which, from having been so
rich in them, has become so poor ; none which has so
nearly stripped its root-syllables of the apparatus of suffixes
with which they were formerly clothed, and left
them monosyllabic. All this has come about mainly
through the instrumentality of the tendency to ease and
abbreviation, a tendency which in this department of its
working, especially, makes truly for decay ; the conservative
force, the strictness of traditional transmission, has
not been sufficient to resist its inroads. Much of the
loss has been the work of the last few centuries ; and
there is no difficulty in pointing out causes which have
at least quickened it. When men learn a strange language,
by a practical process, they are apt especially to
make bad work with its endings ; if they get the body
of the word, its main significant part, intelligibly correct,
they will be content to leave the relations to be
understood from the connection. This was what helped
the decay of the Latin tongue, and its reduction, in the
mouths of Italians, Celts, Iberians, and others, into the
corrupted and abbreviated shape of the modern Romanic
dialects ; and the irruption into England of the French-speaking
Normans, and their fusion with the Saxon-speaking
English, added an appreciable element of
force to a tendency which was perhaps already sufficiently
marked in the later Anglo-Saxon.

But it is only in degree that the English differs
herein from the other languages of its family, and from
105those of other families. The tendency to abbreviation
for ease, for economy of effort in expression, is a universal
and a blind one ; destruction lies everywhere in
its path. The same process which, by a disguising
fusion and integration of elements once independent,
makes a word or form, goes straight on to its contraction
and mutilation — and in early language as certainly,
though not necessarily so rapidly, as in later. There is
believed to be hardly anything, if anything at all, earlier
in the structure of our language than the first-personal
endings, mi in the singular, masi in the plural.
Yet these are already economized alterations of something
still more primitive ; the masi, especially, so
changed that the comparative philologists dispute as
to its derivation. All that we have left of either of
them in English is the solitary m of am (for as-mi).
And every language related with ours has something of
the same loss to show ; and like losses in every other
department of inflection and of derivation.

The forms, even of the richest known languages,
embody and bring to distinct consciousness only a small
part of the infinity of relations which subsist among
the objects of thought, and which the mind impliedly
recognizes, even when it does not direct attention to
them by expression. Hot one of those which are expressed,
any more than those which have not found embodiment,
is absolutely essential to successful speech.
When it has attained expression, the mind which contemplates
it is not dependent upon its audible sign, but
may even be made carelessly secure by this, and, while
realizing the idea, permit itself to drop the sign as not
indispensable. But we may note for our consolation
that, unless a people is undergoing actual degradation
in quantity and quality of mental work, it does not
106lose what it once possessed in the way of inflectional
apparatus without providing some other and on the
whole equivalent means of expression. The style of
expression may become very much changed, without
any real loss of expressiveness. The downfall of the
case-system was accompanied by the uprise of the class
of prepositions ; the loss of pronominal elements in the
form of personal endings led only to their more extended
use as independent words ; the impoverishment of
the scheme of moods and tenses was compensated by
the introduction of a rich apparatus of auxiliaries, capable
of expressing nearly all the old distinctions, along
with a host of new ones.

This brings us, however, as we have already been
repeatedly brought, to face the remaining department
of change of language — namely, the addition of new
resources of expression ; and to that we now turn.107

Chapter VII.
Growth of language. Production of new words and
forms.

Special importance of this mode of linguistic change ; objects attained by
it. These objects partly gained without external additions ; enrichment,
definition, multiplication of meaning in old words. Provision
of new styles of expression. External additions ; borrowing from
other languages ; its kind and degree ; excess of it in English. Invention
of new words ; onomatopœia. New words made by combination
of old ones ; production of forms by this method ; its wide reach
and importance ; internal formative changes really the result of external
additions. Differentiation of the form of a word in different
uses. Multiplication of the uses of a word by derivative apparatus ;
conversion of one part of speech into another.

In our examination of the methods of change or
growth in language, we have finally to consider the subject
of acquisition of new material, of the means whereby
the waste incident to phonetic decay is made up, and
expression for new thought and knowledge provided.
These means have been already in part set forth or
alluded to ; for all the modes of linguistic growth so
intertwine and interact that it is impossible to discuss
any one of them, however succinctly, without taking
more or less account of the others.

This last mode of change, it may be remarked in
introduction, constitutes in a higher and more essential
108sense than any of the others the growth of language,
and ought to bring most distinctly to light the forces
actually concerned in that growth.

The general object attained by additions to language
is obviously the extension and the improvement of expression,
supply of representative signs for new knowledge,
amendment in the representation of old knowledge.
But, as we must first observe, these ends are to
no small extent gained without any apparent change in
language. In part, by new syntactical combinations of
the old materials of speech, by putting together old
words into new sentences : and this is plainly a department
of the use of language by which great results are
won ; hosts of new cognitions and deductions are thus
provided for. And yet, this work cannot go on without
more or less affecting the inner content of the terms we
use, changing the limits and even the whole character
of the conceptions which they represent. If, for example,
we say “the sun rises, shedding light and heat on
the earth,” the sentence is one which (or its equivalent
in other languages) might have been uttered, so far as
concerns the items of which it is made up, at any time
since the infancy of speech and knowledge : but how
different the real meaning which it stands for as employed
by us, and by a modern boor or an ancient sage !
Rise to us, as applied to the sun, is only a concession to
appearances ; we do not care to take the trouble to say
that the earth has been rolling over till now our spot
of it comes within reach of the sun's rays ; and as to
rising and falling, it is only since Newton discovered
the great cosmic law of gravitation that we really know
what the words denote. It is a much shorter time since
we learned that light and heat are modes of motion of
matter, apprehended by certain effects which they produce
109on our sensitive organization. And the transformation
which sun and earth have undergone in our
minds needs no more than an allusion. The example
is, no doubt, an extreme one ; yet it is a perfectly fair,
even normal, illustration of what becomes in speech of
one most important part of the new knowledge we
acquire. This kind of change is ever operating like a
ferment through the whole material of language, incorporating
without outward show the changed apprehensions,
the clearer cognitions, the sharpened distinctions,
which are the result of gradual intellectual growth. It
is, as we have called it before, the mind of the community
all the time at work beneath the framework of
its old language, improving its instruments of expression
by adapting them to new uses.

In fact, all the ground over which we went in the
fifth chapter, treating the alterations of meaning as individual
changes, of various kind and direction, we
might properly enough here go over again, having in
view the purposes which the changes are made to subserve.
That, however, would take too much time ; and
we must content ourselves with briefly pointing out certain
aspects of the subject.

How great, in the first place, is the sum of enrichment
of language by this means, may be seen by observing
the variety of meanings belonging to our words.
If each of them were like a scientific term, limited to a
definite class of strictly similar things, the number
which the cultivated speaker now uses would be very
far from answering his purposes. But it is the customary
office of a word to cover, not a point, but a
territory, and a territory that is irregular, heterogeneous,
and variable. A certain noted English lexicographer
thought he had performed a great feat when he
110had reduced the uses of good to forty varieties, besides
an insoluble residue of a dozen or two of phrases ; and,
though we need not accept all his distinctions as valuable,
their number at any rate indicates a real condition
of things. No student who remembers his occasional
despair as (in early stages of his studies) he has glanced
over the lists of meanings of Greek and Latin words in
his dictionaries, trying to find the one that fitted the case
in hand, will question that foreign words, at any rate,
have a perplexing variety of signification ; but the case
is precisely the same with the foreigner who uses an
English dictionary. It is the duty of the competent
lexicographer, in any language, to reduce the apparent
confusion to order by discovering the nucleus, the natural
etymological meaning from which all the rest have
come by change and transfer, and by drawing out the
others in proper relation to their original and to one another,
so as to suggest the tie of association by which each
was added to the rest — if he do not find (as is not very
rarely the fact) that the tie is doubtful or undiscoverable.
If we were to count in our words only those degrees of
difference of meaning for which in other cases separate
provision of expression is made, the 100,000 English
words would doubtless be found equivalent to a million
or two. As an extreme example of what this mode of
enrichment can do, there is in existence one highly cultivated
tongue, the Chinese, all the growth of which
has had to be by differentiation of meaning, since it rejects
all external additions ; and it has only about 1,500
words : what a host of discordant and hardly connectable
meanings each word is compelled to bear may be
easily imagined.

The particular mode of transfer by which new expression
is most abundantly won is the figurative (as set
111forth and illustrated in the last chapter but one). But,
rich as are its contributions to the absolute needs of expression,
especially in the department of intellectual
and relational language, they are by no means limited
to that. The mind not only has a wonderful facility in
catching resemblances and turning them to account, but
it takes a real creative pleasure in the exercise, and derives
from it desirable variety and liveliness of style.
The power is strikingly illustrated in the case of men
whose life-occupations run in restricted lines, and who
have little general culture ; when they come to talk
upon matters less familiar, they see constant analogies
between these and their staple subjects of thought, and
their discourse is redolent of the “shop.” So especially
the sailor talks as if all the world were a ship, and with
a piquancy and raciness which, as illustrated in the nautical
stories, is full of charm to us land-lubbers ; and
many a term or phrase of this origin has passed into
our general English tongue. And if we would see how
far the phraseology of the mine and the card-table can
be made to go in figurative substitution for ordinary
speech, we may read, in Mark Twain's “Roughing It”
(chap, xlvii.), that amusing (and, in this aspect, instructive)
account of the interview between the preacher and
the gambler who wants to get his late exemplary partner
decently buried. For a more dignified example,
take the constant recurrence of the Vedic poets to the
cattle-yard and the pasture for the staple of their comparisons,
and for the suggestion of many a term used
later, without any sense of a figure involved in it, to
express human conditions. So far as this is odd or undignified,
it forms the largest element of what we call
“slang,” and we frown upon it ; and properly enough ;
but yet it is only the excess and abuse of a tendency
112which is wholly legitimate, and of the highest value, in
the history of speech. It seeks relief from the often
oppressive conventionality, even insipidity, of words
worn out by the use of persons who have put neither
knowledge nor feeling into them, and which seem incapable
of expressing anything that is real. In the exuberance
of mental activity, and the natural delight of
language-making, slang is a necessary evil ; and there
are grades and uses of slang whose charm no one need
be ashamed to feel and confess ; it is like reading a narrative
in a series of rude but telling pictures, instead of
in words.

A meaningless conventionality, to be sure, has also
its special uses, as in the forms of social intercourse,
where we are sometimes called upon to disguise instead
of disclosing our thoughts by speech. To take an example
or two of the simplest kind — we say “how do you
do ?” to an acquaintance, but should feel imposed upon
if he answered by detailing all the symptoms of his
health ; we begin a letter to one whom we really detest
with “my dear sir,” and at the end declare ourselves his
“obedient servant,” though we should resent a single
word from him which bore the semblance of a command.
And so in many other cases : to devise more sincere
phrases would seem blunt and odd, an unbecoming intrusion
of our personality. Then, again, there are subjects
of decency or delicacy, with, reference to which we
have to pick our expressions very carefully, if we would
not offend or disgust. It is one of the most striking
illustrations possible of the dominion which words have
won over our thoughts, that we will tolerate in indirect,
figurative, merely suggestive expression what would be
repulsive in direct statement. Here, by an effect contrary
to that which we noticed above, a term perhaps
113becomes after a time, by frequent use, too directly significant,
and we have to devise a new one, less lively.

Thus, independently of any marked increase of
knowledge and multiplication of conceptions, as well as
in connection with this, the instrument of expression is
continually undergoing alteration for the better, by being
applied to more varied and defter modes of use.
The same methods of increase serve both the one purpose
and the other. We have perhaps already given
sufficient attention, in the fifth chapter, to that most
general and grandest of movements of signification,
which carries words over from a more material and substantial
value toward one that is more conceptual and
formal, in its two departments of the making of intellectual
expression and the production of form-words —
in the former, turning more to the uses of new thought ;
in the latter, more toward the completion of the expression
of old thought ; and we may proceed to take
up the other and more conspicuous part of growth, consisting
in external additions to language, the accession
of new words to the vocabulary.

And we may best begin with that particular mode
of external increase which is the most extraneous of all
— namely, the bringing into a language of words borrowed
out of other languages. Borrowing, in greater
or less degree, is well-nigh universally resorted to ; there
is hardly a dialect in the world, of which the speakers
ever come in contact with those of another dialect,
which has not taken something out of that other. What
comes most easily after this fashion is names for articles
and institutions of foreign growth, which, on making
their acquaintance, and deeming them worthy of introduction
or adoption, we often find it convenient to call
by the names given them by their former possessors.
114So the banana is a tropical fruit, with its own tropical
title ; and the nations of continental Europe mostly call
anana, for the same reason, the fruit for which we have
chosen to provide the more native appellation of pineapple — i. e.
such, an apple as, judging from its cones, a
pine might bear if it tried to be an apple-tree. So also
with the institution of the tabu, of which the Polynesian
name has fairly won a place in more than one
European tongue. A language like ours — since we
come in contact with nearly all the nations of the world,
and draw in to ourselves whatever we find of theirs that
can be made useful to us, and since even our culture derives
from various sources — comes to contain specimens
from dialects of very diverse origin. Thus, we have
religious words from the Hebrew, as sabbath, seraph,
jubilee ; certain old-style scientific terms from the
Arabic, as algebra, alkali, zenith, cipher, besides a considerable
heterogeneous list, like lemon, sugar (ultimately
Sanskrit), sherbet, magazine ; from the Persian, caravan,
chess, shawl, and even a word which has won so
familiar and varied use as check ; from Hindi, calico
and chintz, punch and toddy ; from Chinese, tea and
nankeen ; from American Indian languages, canoe and
moccasin, guano and potato, sachem and caucus. Some
of these are specimens out of tolerably long lists ; and
there are yet longer from sundry of the modern European
languages, as the Spanish, and Italian. All together,
they do not make up any considerable proportion
of English speech ; but they have for us a high
theoretical importance, as casting light upon the general
process of names-giving, of which we shall treat more
particularly in the next chapter. It is by no process of
organic growth, assuredly, that we put a certain title
upon a certain thing because some far-off community, of
115which we know little and for which we care less, gave
it that title ; yet this makes, when once in use, just as
good English as the words that belong to the very oldest
Saxon families, or that “came in with the Conqueror.”

This last expression, however, reminds us that there
is another kind and rate of borrowing in which our language
indulges, more or less in common with others.
All the leading nations of Europe have received their
culture and their religion, directly or indirectly, from.
Greece and Rome. Some of them, indeed — as the various
tribes of Italy, the Celts of Gaul, the Celtiberians of
the Spanish peninsula — took so much from Rome that,
along with the rest, they accepted also her speech, in
mass, and now talk a nearly pure Latinic dialect. With
the others, there followed only a result akin with that
which we have been noticing above ; in connection with
new ideas and institutions, they took the names by
which these were known to their more original possessors.
Thus there came to be numerous Latin and Greek
words in the Germanic, the Slavonic, and the Celtic
tongues. Not a few of them occur in the oldest Anglo-Saxon ;
and they abound in the German vocabulary,
even in those parts of it which have an original aspect.
The dependence of Europe on the classical sources for
knowledge, arts, and sciences, continued long. Latin
was everywhere read and written by the learned, almost
as the only language worthy of such high uses ; and
even now its study is a pervading element in education.
This kept fully alive the habit of resorting to the stores
of Latin expression to satisfy all those needs of the
learned which the more regular growth of the popular
speech did not supply. In a certain way, it was easier
for those modern tongues which are themselves derived
116from the Latin to do this than for others ; but we must
not estimate their advantage too highly, observing how
little we ourselves borrow from the Anglo-Saxon, or
from any other Germanic language. The Latin and
Greek alone have occupied such a position that all Europe
could resort to them for the enrichment of its multifarious
speech. In other parts of the world, other
languages have stood in a like place. To the scores of
tribes and nations of discordant speech in India, the
Sanskrit has long been the sacred and literary dialect,
and its literature the fountain of higher thought and
knowledge ; and all the cultivated tongues of modern
India have come to be full of Sanskrit words, as the
European tongues are of Latin. The Persians, a thousand
years and more ago, were forced to receive a new
religion and constitution at the hands of their Arab conquerors,
and modern Persian is almost more Arabic
than Persian. The Turks burst into Persia as a wild
uncultivated horde, with nearly everything to learn save
war and plunder ; and their present written style is
more crowded with Persian and Arabic than the most
extreme Johnsonese with Latin. The Japanese made
themselves, fifteen centuries since, the pupils of the
Chinese ; and they have absorbed the Chinese vocabulary
almost bodily into their own language.

The English, then, is not at all peculiar in its borrowing
freely from other tongues to enrich its vocabulary ;
it is merely peculiar among European languages
for the extent of its borrowing from tongues only remotely
akin with itself. A trustworthy estimate of the
derivation of the words found in our great dictionaries
makes nearly five sevenths of them to be of classical
derivation, and only about two sevenths native Germanic :
the sum of derivatives from other quarters —
117only a thousand or two — being of no account in such
an estimate. Of course, the words do not enter into
the ordinary combinations of practical use in any such
proportion as this, because our commonest terms, the
bulk of the material of ordinary speech and nearly all
its machinery, are Germanic. In the list of words used
by Milton, for example, full two thirds are classical ;
but in a page anywhere of Milton's poetry the classical
element is only ten to thirty per cent. ; and even in
Johnson's style its proportion is but little greater.

For this preponderance, in one aspect, of the borrowed
material in English speech, there are easily assignable
reasons. The Norman invasion, leading to a
long antagonism and final fusion of a French-speaking
with a Saxon-speaking race, brought in by violence, as
it were, a great store of French words, of Latin origin,
and thus made it comparatively easy to bring in without
violence a great many more. And the deadening of the
native processes of composition and derivation and inflection,
caused in part by the same great historical
event, made the language more incapable of meeting
out of its own resources any great call for new expression.
So, when the pressing exigencies of the last century
or two, almost unexampled in their urgency, arose,
the resource of borrowing, already much availed of, was
drawn upon almost to excess. When a community is
living quietly on, with no marked accumulation of the
fruits of mental activity, ruminating its old conceptions
and slowly elaborating new, the purely natural increase,
proceeding slowly and unconsciously from the great body
of speakers, will be likely to serve all needs. But when
science and art and philosophy are making rapid advances,
when new branches of knowledge are springing
up, one after another, each calling for a whole vocabulary
118of new terms, when infinite numbers of new facts and
new objects are coming to notice, then the native modes
of growth, of even the most fertile language, will be
taxed beyond their capacity to provide a nomenclature
for all. The call is in very great part for technical vocabularies,
words for learned use ; and the learned find
what they want most conveniently in the learned languages.
They gain in addition the practical advantage
that all the inheritors and continuers of a common civilization
thus possess something like a common dialect,
in which to denominate those conceptions in which they
have a joint interest closer than that which they have
with the mass of their countrymen. Our five sevenths
of classical material are mainly words of learned use
only, which the young child does not acquire in order
to “speak English,” and which the uneducated man
never learns ; a host of them are of rare occurrence
even in books. But any one of them may come, under
the conditions of practical life, to be as familiar as
material of less artificial origin : cases of this kind are
gas, Thursday and its kin, dahlia, petroleum, telegraph,
photograph.

There are degrees of kind as well as of extent in
the process of borrowing. What is most easily taken
out of the stores of one language to be added to those of
another is the names and epithets of things, nouns and
adjectives ; verbs, much less easily ; particles, hardly
at all ; apparatus of derivation, prefixes and suffixes,
very sparingly ; and apparatus of inflection, endings of
declension and conjugation, least of all. Even English
is nearly unmixed in its grammar ; its articulating parts,
the elements that bind ideas together and show their
relations, that make sentences, are almost exclusively of
Anglo-Saxon origin. For this reason, notwithstanding
119the preponderance of classical material in its wider vocabulary,
the English is still rightly reckoned a Germanic
language.

Of the out-and-out invention of new words, language
in the course of its recorded history (for we do
not now speak of its initial stage) presents only rare
examples. Sometimes, however, a case occurs like that
of gas, already noticed as having been devised by an
ancient chemist, as artificial appellation for a condition
of existence of matter which had not before been so
distinctly apprehended as to seem to require a name.
Along with it, he proposed blas for that property of the
heavenly bodies whereby they regulate the changes of
time : this, however, was too purely fanciful to recommend
itself to general use, and it dropped out of sight
and was forgotten, while the other came to honor.

More frequent than such words as this, which only
by a lucky hit gain life and a career, are those in which
the attempt has been made in a rude way to imitate the
sounds of nature : as when the cuckoo and the pewee
and the toucan were named from their notes ; or as in
some of the descriptive words like crack and crash, hiss
and buzz, which are by no means all old, but have been
made, or shaped over into a pictorial form, within no
long time. We call such words onomatopœias, literally
‘name-makings,’ because the Greeks did so : they could
conceive of no way in which absolutely new language-material
should be produced except by such imitation.

We pass now to notice another process, whereby
there comes into being for the uses of expression material
which is only in a certain sense new, but which
nevertheless furnishes notable enrichment to speech,
and in more than one department ; a process which the
general history of language shows to be more important
120than any other. It is the composition of words, the
putting two independent elements together to form a
single designation. Our illustrative passage furnished
us one or two examples of it : namely, reste-dœg, ‘rest-day,’
and leorning-cnihtas, ‘learning-knights,’ i. e. ‘pupils.’
Such a word is logically an abbreviated descriptive
phrase, with the signs of relation, the ordinary
inflections or connectives, omitted ; the two main ideas
are put side by side, and the mind left to infer their relation
to one another from the known circumstances of
the case. It is so far an abnegation, for the sake of
brevity and convenience, of the advantages of a language
which has formative elements and form-words.
The undefined relation may be of every variety : thus,
a headache is a pain in the head ; a head-dress, a dress
for the head ; a headland, a point of land comparable
to a head ; a headsman, a man for cutting off heads ;
headway, motion in the direction of the head (of any
animal but man) ; thus, also, a steamboat is a boat propelled
by the force of steam ; a railroad or railway is
a road laid with rails ; a buttercup or butterfly is a cup
or fly having the color of butter : and so forth. Such
a word, again, is formally characterized by a unity of
accent ; this is the chief outer sign of combination,
binding the word together — although it is not enough
of itself to make a compound ; else the man and have
gone and shall go and their like would be compounds
also. Nothing is simpler or more common than for a
language to form such compounds. Yet their frequency
is very different in different languages : the Sanskrit
abuses the liberty of making them ; the Greek, the
Latin, the German, are examples of tongues which use
them abundantly, yet with wise moderation ; the French
has most nearly lost the power of their production.
121Though in English they are far from being as numerous
as in German, our speech is pretty full of them,
the words quoted above may serve as examples of what
is done in this way to increase the resources of expression.
How ready the language-users are to forget the
source of the compound, to lose the separate impression
of its constituent words, to use it as a unitary sign for
the conception to which it is attached, and then to disguise
and integrate it by phonetic change, has been already
pointed out, and need not be here further dwelt
on or exemplified. But a most important department
of its action is in a direction which calls for a little additional
illustration.

Among the many adjectives which we sometimes
combine with nouns to form compound adjectives, there
are those which, in virtue of their meaning and consequent
wide applicability, we use with special frequency,
forming considerable classes of compounds with a common
final element. A typical instance is full, German
voll, which is added to nouns enough, and in a sufficiently
general sense, to be made a kind of suffix, its
own specific force being lost : dutiful and plentiful
are equivalent to duteous and plenteous. Its opposite
is less, German los ; not our adjective less, but, as the
German indicates and as the older forms of our language
prove, loose ; here the originally independent
word has been so disguised by phonetic change as to
have become absolutely an adjective suffix. Ly (of godly,
homely, etc.) has been already fully enough explained
(p. 41), as coming, by a different sort of phonetic change,
from like. And a certain case-form of this compounded
adjective, we saw, was by a change of office converted
into a nearly universal adverbial suffix : thus, truly,
plentifully. The French adverbial ending ment is in
122like manner from the Latin ablative mente : grandement,
‘grandly,’ is by origin grandi mente,‘with great
mind.’ Our some in wholesome (German sam in heilsam)
is altered from older sam, and identical with same
in the sense of ‘like.’ There are noun-forming suffixes,
also, which own a like origin. The plainest cases among
them, perhaps, are ship, German schaft, in lordship,
herrschaft, and their like ; and dom, German thum, in
kingdom, wisdom, königthum, weisthum : the former
comes from shape, the latter from doom. We have
glanced above at a case or two of verbal tense-making
after the same fashion. The don of hyngredon (plural
of hyngrede, p. 42) was in Gothic dêdum, an evident
auxiliary, our did, which, at a time very early in the
common history of the Germanic dialects (for it is
found in them all, though not in any even of their
nearest relatives), was added to some verbal word to
make a verbal form, with the final result that the two
became fused together into one, even as we now add it
to a verbal word, the infinitive, to make a verbal phrase,
I do love, I did love, only without fusion. Quite parallel
with this is the fusion of the present of the verb ‘to
have’ with the infinitive in the Romanic languages, to
make their modern future, as donner-ai, ‘shall give,’
when compared with our verbal phrase I have to give,
its unfused equivalent. Abundant traces of the same
sort of composition, fusion, and resulting production of
a new verbal form, are to be seen in the Latin, whose
imperfect in bam, future in bo, and perfect in ui or vi,
are generally acknowledged to contain as their endings
certain forms of the verb which in our language is the
substantive verb to be. And even the Greek and Sanskrit
have like compound forms to show, of earlier and
later date : one, the future in Skt. syâmi, Gr. σω.is
123believed to go back to the primitive period of linguistic
growth in our family of languages.

These are some of the plainest among the numerous
examples which might be brought forward, going to
show that suffixes of derivation and inflection 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, in more elaborate structures. The
auxiliary processes of oblivion and attenuation and
transfer of meaning, and of disguise and abbreviation
of form, are simply the same here as in all the other
cases we have treated ; they are essential parts of the
making of forms ; for so long as the independent word,
in its individual shape and meaning, is plainly recognized
in the combination, so long does this remain a compound
rather than a form : our ful, for example (German voll),
is not so truly a suffix as ly (lich), because the independent
adjective is too apparent in it ; a disguising alteration
is needed to help make an affix — a “formative element,”
as it is properly termed, in distinction from the
“radical element,” the root or base, or the crude-form,
to which it is appended.

Now it is by no means all, or even the largest part,
of our existing formative elements, suffixes of derivation
and inflection, of which the origin in this method
can be actually proved ; and if we are to believe nothing
respecting language which does not rest on positive
evidence, we shall never make the principle of combination
go far toward explaining the growth of language.
But it would be highly unreasonable to demand everywhere
such proof. The disguising effect of the two
124principles of change which bear their part in every new
formation is such that after a time we may be able only
to guess, or not even that, at its origin. We could not
explain the ly from modern English alone ; we could
not be certain as to the d of loved without the help of
the Gothic ; nor as to the σω of the Greek future without
the Sanskrit. Every period of linguistic life, with
its constantly progressing changes of form and meaning,
wipes out a part of the intermediates which connect a
derived element with its original. There are a plenty
of items of word-formation in even the modern Romanic
languages which completely elude explanation.
Mere absence of evidence, then, will not in the least
justify us in assuming the genesis of an obscure form
to be of a wholly different character from that which is
obvious or demonstrable in other forms. The presumption
is wholly in favor of the accordance of the one
with the other ; it can only be repelled by direct and
convincing evidence. And, in actual fact, linguistic
study does not bring to light any such evidence ; its
trustworthy results go rather to prove that the combination
of independent element with element has been
from the beginning, in the languages of our family, the
fertile and the sufficient method of new external growth,
has furnished the needed supply of fresh material, which
then, under the action of the other processes, has been
applied to meet the needs of expression. We shall
have, by and by, to review in brief the history of early
development of these languages, as explained by the
comparative philologists upon the principle here stated.

But a part of our forms, derivative and inflectional,
appear to be made by internal modification rather than
external addition. We say boy and boys, indeed, but we
also say man and men ; we say love and loved but also
125rēad and rěad ; and then there is that wide-reaching and
most important phenomenon in Germanic language,
the variation of radical vowel, in large classes of words
like sing, sang, sung, and song ; like break, broke, and
breach ; like bind, bound, bond, and band. The Greek
has a kindred but less conspicuous change in a considerable
body of verbs and verbal derivatives like λείπω,
ἔλιπον, λέλοιπα ; like τρέπω, ἔτραπον, τέτροφα, τρεπτός,
τράπηξ, τρόπος ; etc. These are seeming violations of
the principle of new growth by external addition, by
combination ; if, however, they can be shown to be,
after all, its results, they will rather lend it a strong
support.

Let us begin with rēad rěad, as the most recent and
the plainest case. In the Anglo-Saxon, this verb and
the little class that go like it had no such difference of
vowel between present and preterit ; and they had in
the preterit the same added ending as other “regular”
or new verbs : the forms were rœdan, ‘rēad,’ rœdde,
‘rĕad.’ But here came in the phonetic principle of easy
utterance : the penult of rœdde had a long vowel before
a doubled consonant ; it was lightened by shortening
the vowel — a proceeding so customary in all Germanic
speech that it has led to the frequent orthographic
device of marking a vowel as short by doubling the consonant
after it. When, then, in the further course of
abbreviation, by loss of final vowels, both forms were
reduced to monosyllables, the double pronunciation of
the final consonant was lost, and the difference of vowels
was left alone to mark the difference of tense. The
case is, on the one hand, analogous with lēave lĕft, feel
fĕlt, etc., where there is a shortening of the vowel for
a like cause, the occurrence of two consonants after it,
but where the consonant group has been preserved ;
126and, on the other hand, it is analogous with set, put, and
their like, which have also lost their preterit ending,
but, having a short vowel in the present, never established
a difference between the two tenses, and so have
the same form in both. The distinction of rēad rĕad,
lēad lĕd, etc., is thus a mere phonetic accident ; a final
turning to account, for the purposes of grammatical expression,
of a difference which arose secondarily, as the
unforeseen consequence of an external addition, when
that addition had been lost by phonetic decay. Such
a distinction is wont to be termed “inorganic,” as distinguished
from one like loved from love, which answers
just the purpose for which it was at first intended.

As for man men, that is a case of what in German
is termed umlaut, or “modification of vowel,” a phenomenon
of wide range in Germanic language, but of
which the results are reduced almost to a minimum in
English. It was originally the alteration of an a-sound
to an e-sound by the assimilating influence of a following
i (see above, p. 71) : a change, therefore, which depended
on the character of the case-ending, and had
nothing whatever to do with the distinction of plural
from singular ; it was even the fact in Anglo-Saxon
that one of the singular cases (dative) had e, and two of
the plural cases (genitive and dative) had a. But, after
exercising their assimilative influence, the endings were
lost (like the second d which had shortened the long
vowel of read) ; and the dative and genitive (plural)
were lost as separate forms ; and so man and men were
left to stand over against one another as singular and
plural. And because this difference of vowel was sufficient
to distinguish, the two numbers, linguistic usage
did not go on, as in a multitude of other cases (e. g. in
ears for ear : see p. 38), to add an s for the same purpose.
127Here, again, is an application to the purposes of
grammatical distinction of a difference which was accidental,
inorganic, in its origin.

To enter into a full discussion and explanation of
the remaining case, the ablaut, or variation of radical
vowel, in bind, bound, band, bond, and their like, would
take a great deal more time than we can afford to it,
and would bring up some obscure and difficult points,
as to which the opinions of investigators are still at
variance. But we should find in it nothing different, as
regards the essential principles involved, from what the
other two examples have furnished us. The preterit,
the participle, the derivative noun, had originally their
external formative elements — the first its reduplication,
as in cano cecini, τρέπω τέτροφα, haldan haihald ;
the other two their endings of derivation — there was
no difference of vowel. And when the difference first
appeared, it was not significant, any more than that of
felt from feel, of (German) männer from mann ; it
was developed under purely euphonic influences ; it
involves, in its various manifestations, the weakening
of an original a-sound, the strengthening of an i or
u-sound when accented, and a fusion of the preterit reduplication
with the root. There is nothing here to
call for the admission of an exception to the general
rule that, in our languages, forms are made by an external
accretion of elements which were at first independent
words.

The fact, however, is here brought to light, and constitutes
an addition of some importance to the means
of enrichment of language, that accidental differences
are seized upon and turned to account by being put to
new uses. A word thus, as it were, divides into two or
more, each of which then leads an independent life.
128Some notable examples of this we have seen already :
the Anglo-Saxon ân has become in English the numeral
one and the article an or a ; of has become off and of ;
also and as, like German also and als, are representatives
of one original ; so fore and for, like German
vor, für, ver ; through and thorough are a very peculiar
divorcement, with accompanying conversion of an adverb
into an adjective ; outer and utter are two sides of
one word and one idea ; cónduct and condúct are specimens
of a large class of couplets, distinguished by accent
alone ; minúte and mínute (mínit) are a convenient
distinction, which we might wish we had also for the
two uses of second ; and genteel, gentle, and gentile are
all alike the Latin gentilis, and in their variety of meaning,
as well as in their common derivation from a root
signifying simply ‘to be born,’ are a striking example
of the possibilities of linguistic mutation.

The method of growth out of the native resources
of a language, by putting its materials together into new
combinations, and so making new names for things, and
sometimes new forms, is of course one of much, slower
operation than the importation of learned and technical
terms from abroad, especially when this is pushed to
such an extreme as in our speech. Above all, in the
making of forms, its progress is almost insensibly gradual,
and its results are few. It cannot well take less
than generations to pass an element originally independent
through those changes of shape and meaning
which it must undergo in order to become a suffix. As
a set-off against this, to be sure the results, once attained,
are of very wide application. When, for example,
did is worked down into a preterit ending, we apply
it to make past tenses for all our new verbs, however
many they may be ; and there are few adjectives in the
129language which may not form their corresponding adverb
with ly, little as most of them would endure composition
with like. But if we take into consideration
the whole long course of life of a language, extending
through thousands of years, and also the sum of human
languages in all parts of the world, few of which, comparatively,
are placed in circumstances to derive much
advantage from borrowing, it is of the utmost importance.
It is capable of providing, along with variation
of meaning, and variation of form under phonetic
change, all the new material which is needed for the
ordinary development of expression ; it is also able,
with the same help, to transform by degrees the grammatical
character of a language, adding new distinctions,
and supplying the place of those that are lost by the
wearing-out processes.

In connection with this, we have to note one more
important department of the means of enrichment of a
language : namely, the capacity, belonging to every
tongue that has any share of an inflective character, of
multiplying the applicabilities, and so the usefulness, of
its material, new or old, by adding formative elements
to it, by putting it through the processes of inflection
and derivation. By no means all the formative apparatus
which a language possesses can be turned to use in
this way ; the English distinctions, for example, of he
and him and they and them, of man and men, of give
and gave, of sit and set, of true and truth, of land and
landscape, though inflective, are dead, and we can no
longer make new forms by their help. But to any noun
which we import we may add an s for the possessive
and plural, as telegraphs ; from any verb we can make
a little scheme of inflectional forms, as telegraphest, telegraphs,
telegraphed (pret. and part.), telegraphing (part.
130and infin.). Then we have our suffixes for turning a
noun into an adjective, as telegraphic ; a number of
these, as ful, less, ous, ish, y, are still sufficiently alive
to admit of practical application. Then, besides that
we can turn any adjective, on occasion, into a noun — as
the good, the beautiful, and the true — we have a suffix
ness, of very wide applicability, for abstracts. And the
ly will convert almost any adjective into an adverb, as
telegraphically. The verb, too, has its instruments of
mutation : telegraph, for instance, makes telegrapher
and telegraphist and telegraphy. And, on the other
hand, there are means of turning nouns and adjectives
into verbs : we say harden and roughen, and revolutionize
and demoralize, and so on. This last is in all languages
the principal means whereby the stock of verbal
expression is increased, and new starting-points are obtained
for further development : such “denominative”
verbs, as they are called, abound in every member of
our family, in every period of its history. All depends
upon the power which language has of treating its stock
of formative elements in the same way as its more material
elements. Let a certain modificatory syllable,
however reduced to formative value, once come to occur
in forms enough to get itself distinctly associated in the
minds of speakers with a certain modification of meaning,
and it is further applied when that modification
needs to be expressed, just as naturally as a connective
or an auxiliary is similarly used. A notable example of
how an element of extraneous origin can come into a
language, and by slow extension finally work its way up
to such a use, is afforded by ize and ism and ist, which,
though ultimately of Greek origin, and imported by us
through the French, have made themselves part of our
living apparatus of derivation, and are even abused, in
131a half-artificial and affected way, by low speakers and
writers, to the formation of such monstrosities as walkist,
hair-cuttist.

It is of high importance, if we would understand
the structure of any language, to distinguish its living
apparatus of inflection and derivation from that which
is only recognizable in its older words as having been
formerly alive. And it is in great part by the deadening
of such means of multiplication of expression that
a language like ours gains its peculiar character, as
a prevailingly analytical speech. Each tongue has its
own way in this regard : the French is poorer even
than English in apparatus of derivation ; the Slavonic
tongues, as the Russian, are vastly richer than either
Germanic or Romanic.

The English retains a peculiar relic of its former
capacities as an inflective language, in its power to turn
one part of speech directly into another, without using
any external sign of the transfer. The tongues of our
family had in old time a formal means of making “denominative”
verbs out of nouns and adjectives ; we
have mainly worn out and lost the means, but we make
the verbs almost more freely than ever : thus, to head
an army, to foot a stocking, to hand a plate, to toe a
mark, to mind a command, to eye a foe, to book a passenger,
to chair a candidate, to table a resolution, to
stone a martyr, to scalp an enemy : and so on indefinitely.
The examples show that the relation of the
action to the conception expressed by the noun is of the
greatest possible variety, determined in each case only
by its known conditions, as apprehended by the mind
of speaker and hearer. An equally peculiar capacity is
that of transmuting without ceremony a noun into an
adjective : thus we say a gold watch, while the Frenchman
132must say ‘a watch of gold,’ and the German ‘a
golden watch,’ or else, by actual composition, ‘ a gold-watch ;’
so also, a steam mill, as against the French ‘a
mill by steam’ and the German ‘a steam-mill ;’ so a
China rose ; and so on. This comes from a relaxation
of the bonds of composition ; the division, as it were,
of a loose compound like gold-mine into its parts, and
an attribution to the name itself in separate use of an
office rightfully belonging to it only when it loses its independence
by union with another. This changeableness
of office is something very different from the original
indefiniteness of uninflected languages. Our apprehension
of the different office of verb, noun, and adjective
is kept clear enough by the numerous words which have
only one and not another of these characters ; we preserve
the distinction even after abandoning its sign ; and
thus have by inheritance more of the power of increasing
the resources of expression than makes any outward
show in our language.133

Chapter VIII.
Summary : the name-making process.

Review of the processes of change ; their contribution to name-making.
Degrees of reflectiveness in name-making. Antecedence of the conception
to its sign ; illustrations ; examination of arguments used
against this view. Sources of the material of names ; artificiality
of the tie between name and idea. Etymological inquiries ; character
of the reasons for names ; a science of morphology. Force
concerned in name-making ; the linguistic faculty ; false views and
their grounds examined. Part taken by the community in the process ;
its relation to the action of individuals.

We have now finished our compendious review of
the individual processes — at least, the leading ones — of
which is made up the growth of languages like ours.
In order to understand the historical movement of any
language at a given period, we need to analyze it into
such parts as these, and to see how, separately and together,
they are working ; to note the kind and degree
of activity of each, and trace, if possible, the causes
that determine their difference. In our exposition and
illustration, we have had in view especially their agency
in the recent and present growth of English ; and we
cannot spend the time, nor is it necessary, to take any
more notice of their different operation in other languages
than we have already incidentally done, and
shall have occasion in the same way to do hereafter.
134We go on, rather, to consider certain general principles,
mainly derivable in the way of inference from the details
we have had before us, and bearing upon the general
process of name-giving, or the provision of signs
for conceptions. The other departments of linguistic
change, as we have already seen, are of comparatively
subordinate importance and not difficult of explanation ;
but to understand fully the means whereby language
compasses the expression of whatever calls for expression
is to comprehend the essential nature of linguistic
growth, and even that of language itself.

We will begin by noticing that a part of the name-giving
process, at any rate, is easy enough to understand ;
it goes on in the broadest daylight. When a
human being is born into the world, custom, founded
in convenience, requires that he have a name ; and those
who are responsible for his existence furnish the required
adjunct, according to their individual tastes,
which are virtually a reflection of those of the community
in which they live. English-speaking parents
do not give a Chinese or a Sioux name, nor vice versâ ;
the saint to whom his natal or christening day is sacred,
a conspicuous public character, a relation from whom
expectations are entertained, or something else equally
unessential, directs their choice ; no matter what, so
long as the individual is named, and with such a name
that neither the community who call him by it, nor he
himself later, shall revolt and insist on another appellation.
Such an act as this may seem to have little to do
with general language ; but that depends upon circumstances :
the proper name Julius has ended in our calling
a month July ; the nickname Cœsar has given the
title to the heads of two great nations, Germany and
Russia (kaiser, czar) ; the christening of the baby Vespucci
135as Amerigo has led to America and American.
So also with a planet : Herschel had the naming of
Uranus, and Leverrier of Neptune ; only they too were
guided by the already established usages of language
and the consequent preferences of the community ; the
name of Georgium sidus, with which, in the former
case, it was unworthily sought to flatter a monarch, was
frowned upon, and dropped out of sight. The discoverers
of the asteroids enjoy the same privilege ; and
under the same conditions. So with all scientific discoverers ;
they exercise a prerogative, yet under limitations ;
they must respect the prejudices of their fellows,
and they must prove their right as nomenclature : in
the scientific community, as every one knows, the claims
of rival name-makers are very sharply discussed, under
government of nicely-established rules. So with inventors
likewise : to each is conceded a limited right to
give a name, or to determine the acceptance of a name
given by some one else, to what he has produced. Nor
is the case different anywhere in the technical vocabularies
of art, of science, of philosophy. The metaphysician
who draws a new distinction denominates it ;
he is even allowed — always with restrictions — to recast
the whole vocabulary of his department, for his own
special convenience ; and if the other philosophers
are convinced of the usefulness of the change, they
ratify it.

All this is done under the full review of consciousness.
There is first the apprehension of something as
calling for expression, or for better expression, and then
the reaching out after, and the obtaining in some way,
the means of expression.

But just this, only with variety in the degree of
consciousness involved, is the nature of the process of
136name-making in all its varieties. If it were not so, language
would consist of two discordant parts, one made
in this way, and one in some other. Let us consider it
a little more particularly, with reference to some of the
principles involved.

First, there is always and everywhere an antecedency
of the conception to the expression. In common phrase,
we first have our idea, and then get a name for it. This
is so palpably true of all the more reflective processes
that no one would think of denying it ; to do so would
be to maintain that the planet, or plant, or animal,
could not be found and recognized as something yet
unnamed until a title had been selected and made ready
for clapping upon it ; that the child could not be born
until the christening-bowl was ready. But it is equally
true, only not so palpable, in all the less conscious acts,
all the way down the scale to the most instinctive. The
principle of life, for example, was called animus ‘blowing’
or spiritus, ‘breathing,’ because the nomenclature
had a dim, to us a wholly insufficient, apprehension of
something within the bodily frame, distinct from it,
though, governing and directing it, something which
could come to an end while the body continued in existence ;
and because the breath seemed a peculiar manifestation
of this something, its stoppage being the most
conspicuous sign of the latter's death : they seized the
expression for an already formed conception as undeniably
as did the anatomist who, by an equally bold figure,
first applied inosculation to the observed connection
of the arteries and veins. Every figurative transfer
which ever made a successful designation for some
non-sensible act or relation, before undesignated, rested
upon a previous perception of analogy between the one
thing and the other : no one said apprehend of an idea
137until he had felt the resemblance between the reaching-out
of the bodily organs after a physical object they
want to handle and the striving of the mental powers
toward a like end ; we repeat the act when we say “you
don't get hold of my meaning.” No one said “a thought
strikes me,” or “occurs to me” (i. e. ‘runs against me’),
or “comes into my head” (German, fällt mir ein, ‘falls
in to me’), except as result of an analogy which his
mind had discovered between the intellectual and the
physical action. When a certain new shade of red had
been produced by the creative ingenuity of modern
chemistry, the next thing was to give it a name ; and
magenta was pitched upon, by a perfectly conscious process,
because historical causes had at about that time
given a celebrity to the town Magenta : the name was
not a whit more indispensable to the conception of the
color than, at a period so much more ancient that we
cannot get back to it, the name green had been to the
conception of its color : men said green when they had
observed the distinction of this from other colors, and
its especial appurtenance to ‘growing’ things. And if
we were to trace the etymology of any other similar
word, we should find it of the same character. Nor is
the genesis of form-words and forms unlike this. Off
was changed to a (virtual) sign of the genitive case,
and to to an infinitive sign, by a long succession of
steps, each of which was a putting of the word to
a use slightly different from that which it had served
before, in order to answer a felt need of expression ;
and nothing other than this is implied in the making
of loved, of donnerai, of amabam, of δώσω, of asmi
(am).

We might go over the whole list of illustrations
given in the preceding chapters, and as many more as
138we chose to take, without finding a case different from
these. The doctrine that a conception is impossible
without a word to express it is an indefensible paradox
— indefensible, that is to say, except by misapprehensions
and false arguments. One or two of these it may be
worth while to notice more particularly.

It is wont to be assumed by those who oppose the
antecedence of the idea to the sign, that this opinion
implies the elaboration by thinkers of a store of thoughts
in advance, and then the turning back and naming them
by a conscious after-thought. Here is an inexcusably
gross misrepresentation. There is implied, rather, that
each act of nomenclature is preceded by its own act of
conception ; the naming follows as soon as the call for
it is felt : even, it may be, before the need is realized ;
the forward step in mental action may be so small in
each particular case that only after many have been
taken in the same direction is the removal noticed, when
reflection chances to be applied to it. Every conceptual
act is so immediately followed as to seem accompanied
by a nomenclatory one. Or, an inkling of an idea is
won ; it floats obscurely in the mind of the community
until some one grasps it clearly enough to give it a name ;
and it at once takes shape (perhaps only a delusive
shape), after his example, in the minds of others. The
immense gain in clearness of apprehension, in facility of
handling, conferred upon a conception by its naming, is
not for a moment to be denied : only those are in error
who would transform this advantage into an absolute
necessity. Not less is their error by whom the acknowledged
impossibility that the mind should do without
language the work which it actually does is transferred
to each single minute mental action. It might just as
well be claimed that a man cannot ascend to the summit
139of St. Peter's, or go from Rome to Constantinople, because
in each case the distance is vastly greater than the
length of his legs. In point of fact, he takes one step,
upward or onward, at a time, and makes each newly-won
position a starting-point for further motion ; and in this
way he can go just as far as circumstances and his natural
powers allow. Just so with the mind ; every item
of knowledge and of self-command that it conquers it
fixes in assured possession by means of language ; and it
is always reaching out for more knowledge, and gaining
additional control of its powers, and fixing them in the
same way. It is, as we have repeatedly seen already,
always at work under the surface of speech, recasting
and amending the classifications involved in words, acquiring
new control of conceptions once faintly grasped
and awkwardly wielded, crowding new knowledge into
its old terms — all, on the whole, by and with, the help
of language, and yet in each individual item independently
of language : and there is nothing in the production
of new signs that is different from the rest. The
mind not only remodels and sharpens its old instruments,
but also makes its new ones as it works on.

Again, in making provision of expression for new
conceptions, the names-giving faculty gets its material
simply where it can most conveniently, not inquiring
too curiously whence it comes. Virtually, the object
aimed at is to find a sign which, may henceforth, be
linked by association closely to the conception, and used
to represent it in communication and in the processes
of mental action. To attempt more than this would be
useless indeed, when the tie by which each individual
holds and uses his whole body of expression is only this
same one of association. As we saw abundantly in the
second chapter, the child gets his words by learning
140them from others' lips, and connecting them with the
same conceptions that others do. Questions of etymology
are naught to him, as even the question what language
he shall acquire at all. But those questions are
not really anything more to the adult ; nay, not even to
the learned etymologist, so far as concerns his practical
use of speech. The most learned of the guild can only
follow for a brief distance backward the history of most
words ; and, near or far, he comes to a reason identical
with that of the peasant : “It was the usage :” a certain
community, at a certain time, used such and such a
sign thus and so ; and hence, by this and that succession
of partly traceable historical changes, our own usage
has come to be what it is. We have had to notice over
and over again, above, the readiness on the part of
language-users to forget origins, to cast aside as cumbrous
rubbish the etymological suggestiveness of a
term, and concentrate force upon the new and more
adventitious tie. This is one of the most fundamental
and valuable tendencies in name-making ; it constitutes
an essential part of the practical availability of
language.

Even when there is no conspicuous transfer, when
the changes of use are so slight and gradual that each
new application stands closely connected with its predecessor,
there is no real persistency of original value, and
the point finally reached is often enough so far off from
the place of starting that the one cannot be seen from
the other — as when, in one of our examples above, a
word (have) of which the ultimate radical idea is ‘seize,
grasp,’ has become in one and the same language a sign
of possession in every kind, physical and moral, and
likewise of past action, of future obligation, and of
causation. There is nothing in the least abnormal
141about such a case ; every language has a plenty like it to
show. But every language has also cases in abundance
of a more summary distant transfer, making the reasons
that underlie the current use of words so trivial or so
preposterous that, if use were heedful of incongruities,
the words could not stand a moment. Two forms, for
example, of the great forces that govern matter, electricity
and magnetism, are named, the one from a Greek
word for ‘amber,’ the other from an obscure province of
Thessaly ; merely because the first electric phenomena
observed by the founders of our civilization appeared in
connection with the rubbing of a bit of amber, and because
the stones that exhibited to them the magnetic
force came from Magnesia. Galvanism seems more
worthy, because there is a certain propriety in our honoring
the man who initiated our acquaintance with
this department of phenomena ; yet, after all, it is
rather petty to link such an element to the name of an
Italian doctor. Tragic, tragedy, and all their train,
come, by some tie of connection not yet fully understood,
from the Greek word for a ‘he-goat ;’ comic and
comedy, probably from that for ‘village,’ the same with
our home. Many of the examples already used in other
connections might well be recalled here, as equally suiting
our present purpose ; but it is surely unnecessary to
go further ; our thesis is already sufficiently proved. If
a direct and necessary tie had to be established even at
the outset between idea and sign, new inventions would
be constantly coming into speech, instead of showing
themselves, as at present, the rarest of phenomena.
The reason why we resort instead to the store of old
material is, like all the rest, simply one of convenience.
And perhaps, after all, the most telling fact of wide
range is that the stores of expression of a wholly
142strange language are, when once the way is opened,
drawn upon without stint ; and. we English-speakers
come to call things innumerable by certain names for
the very unphilosophical reason that certain communities
in southeastern Europe, a long time ago, called
things more or less resembling these by names somewhat
similar.

Our doctrine must not at all be understood as implying
that there is no reason why anything is called as it
is : there is in every case a reason ; only the present use
of the name is not dependent on it ; it cannot always be
found out ; and, if found, it is grounded on convenience,
not on necessity of any kind. It amounts to this :
the conception in question is thus designated because
that other was formerly so and so designated ; and the
same is true of the latter also ; another earlier designation
of a more or less kindred conception lay back of it
— and so on, as far back toward the beginning as our
limited vision can reach. Our tracing of the etymology
of a word is the following-up of a series of acts of
name-making, consisting chiefly in the new applications
of old material — with the accompanying, but independent,
changes of form. And every one of those acts
was one of choice, involving the free working of the
human will ; only under the government, as always and
everywhere, of conditions and motives. In order completely
to understand and judge it, we need to put ourselves
precisely in the nomenclator's place, apprehending
just his acquired resources of expression and his habits
of thought and speech as founded on them ; realizing
just his insight of the new conception and his impulse
to express it. But this, of course, is wholly out of our
power ; the à priori position is one we can never assume ;
we can only deal with the case à posteriori,
143reasoning back toward the mental condition from the act
in which it is manifested.

Hence it is evident in what sense alone there can be
a science of morphology, or of the adaptations and readaptations
of articulate signs to the uses and changes
of thought. As implying the existence of necessary
laws of significant development, which are to be traced
out and made to explain the phenomena underlain by
them, no such science is possible ; as classifying and
arranging the infinite variety of actual facts, and pointing
out the directions in which the movement takes place
more than in others, it has a most useful work to do.
What has been done above, in the fifth chapter, is only
a beginning ; the subject is one which would reward a
deep and comprehensive investigation, embracing the
languages of many or all families.

Once more, there is nothing in the whole complicated
process of name-making which, calls for the admission
of any other efficient force than the reasonable
action, the action for a definable purpose, of the speakers
of language : their purpose being, as abundantly shown
above, the adaptation of their means of expression to
their constantly changing needs and shifting preferences.
This great and most important institution, though carried
forward from step to step of its existence in its
condition as heretofore existing, by the incessant process
of teaching and learning, is at the same time in no part
or particle out of reach of the altering action of those
who learn and use it. If convenience require that the
word learned and hitherto only used in a certain sense
or group of senses, and having a certain form, be applied
to an additional sense, or change its application from the
old to a new, and be shaped a little differently, the thing
is done, and no one can hinder it ; if practical use is for
144any reason no longer served by a word, it drops out of
use and is no more ; if practical need, again, call for provision
of new expression, it is in one way or another obtained,
the particular way depending on the conditions
of the particular case. Nor is there any peculiar faculty
of the mind, any linguistic instinct, or language-sense,
or whatever else it may be called, involved in the process ;
this is simply the exercise in a particular direction
of that great and composite faculty, than which no other
is more characteristic of human reason, the faculty of
adapting means to ends, of apprehending a desirable
purpose and attaining it. It is different only in its accidents — namely,
the kind of object aimed at and the
kind of material used — and not in its essential nature,
from that other process, not less characteristic of human
reason, the making and using of instruments. No exercises
of reason, in fact, as we nave already once or
twice remarked, are so closely and instructively parallel
as these two.

This point is obviously one of the most fundamental
and vital importance in the philosophy of language.
There are those still who hold that words get themselves
attributed to things by a kind of mysterious
natural process, in which men have no part ; that there
are organic forces in speech itself which — by fermentation,
or digestion, or crystallization, or something of the
sort — produce new material and alter old. No one,
however, has ever managed, if indeed any one has ever
attempted, to show these forces in actual operation, or
to analyze and set forth their way of working and the
results it produces in detail, exhibiting their product
item by item. Take any individual bit of linguistic
growth, and it is found and acknowledged to be the act
of a human being, working toward definable ends under
145the government of recognizable motives, even though
without any reflective consciousness of what he is accomplishing :
and it is manifestly absurd to recognize one
force in action in the items and another in their sum.
If we refuse to examine the items when forming an
estimate of the force, and only gaze with admiration at
the great whole, there is no theory so false that we may
not for a time rest in it with satisfaction. But we
might with the same reason regard the pyramids, in our
wonder at their immensity and grandeur, as great crystals,
produced by the infinite organizing forces of Nature,
as ascribe language to organic powers contained
within itself ; the moment we come to examine their
component parts, we find everywhere the marks of
human workmanship ; and we ourselves are all the
time building similar structures, even if not upon so
grand a scale as the men of old. The general laws or
general tendencies of language, well enough called by
that name if we do not let ourselves be deceived by the
terms we use, are really only laws of human action, under
the joint guidance of habit and circumstance. As
for setting them up as efficient causes, that is sheer
mythology ; we might as well erect into forces the laws
which govern the development of political institutions,
or the tendencies which in any country, at a given
time, are leading to the victory of one party over another :
it all resolves itself at last into the action of individual
minds, capable of choice, under wide-reaching
motives and inducements, which are recognizable in
their general operations, though not in the detail of
their working upon each mind.

One great reason why men are led to deny the
agency of the human will in the changes of speech is
that they see so clearly that it does not work consciously
146toward that purpose. No one says to himself, or to
others : “Our language is defective in this and that particular ;
go to now, and let us change it ;” any more
than he says : “All things carefully considered, this particular
word in our speech can well enough be spared ;
let us cast it out.” The end aimed at — and not even
that with full consciousness — is the supply of a need
of expression, or the attainment of a more satisfactory
expression. An exigency arises, a conjuncture
in which the existing available resources are not sufficient
for the speaker's ends ; and, in one or other of
the various ways described above, he adds to them, to
answer his present purpose. Or the opportunity offers
itself, and is seized, for a short cut, a new and more
attractive path, to a point accessible enough in old ways.
A person commits thus an addition to language without
ever being aware of it ; any more than the parents who
name their son reflect that they are thus virtually making
an addition to the city directory. If he will well
understand it to be in this sense, every one is welcome
to hold that alterations of speech are not made by the
human will ; there is no will to alter speech ; there is
only will to use speech in a way which is new ; and the
alteration comes of itself as a result. So it was not by
the exertion of his will that the reptile, creeping over
the muddy surface of a Permian or Jurassic shore, made
a record of himself for the human geologist to study, a
few million years later ; and yet, if he had not voluntarily
taken the steps, under sufficient inducement, there
would have been no record.

We must not, indeed, commit the error of ascribing
too much consciousness even to the act of satisfying
the momentary impulse which produces the alteration.
Thus, for example, in phonetic change. A word is produced
147by a highly intricate succession of acts on the
part of the vocal organs ; a careless and unheeded omission
of any one of them results in a mutilation of the
word, or a slight relaxation of the energy of articulation
affects the character of one of the sounds in the compound ;
and as the word answers its purpose just as
well as before, it passes without notice, and the act is
repeated, and becomes first customary, then constant.
This is, in fact, the normal method of phonetic corruption ;
yet no sensible person would ever think of
recognizing any other agency at work than the speaker
himself, acting voluntarily — any more than he would
attribute it to some force operating from outside if a
man, on coming to a ditch which he had been used to
leap every day, should some time put forth an insufficient
exertion of force, and should fall in. If there
were penalties of this sort following slips in utterance,
the subject of phonetic change would make but a small
figure in our comparative grammars. And this is not
the only way in which careless or slovenly handling of
language leads to change. A very large department of
alterations has no other source, but is due to the omission
of distinctions, the blunders of mistaken analogy, on
the part of those who have not carefully studied and do
not bear accurately in mind the proper uses of the words
they employ. And yet, here just as much as in the case
of the naturalist who cons his Greek and Latin dictionaries
in search of a name for a new mineral or plant,
the act of change is the work of the speaker, and of him
alone.

Another reason for holding the false view which we
are now combating is that every person is conscious of
his inability to effect a change in language by his own
authority and arbitrarily ; and what he cannot do, he is.
148sure that nobody can do. And that is true enough ; in
a sense, it is not the individual, but the community, that
makes and changes language. We must be careful,
however, to see clearly in what sense, lest we fail signally
to understand the subject we are examining.
There is implied here a point of high importance in
linguistic philosophy, one which we have already had
more or less in view, but have not taken up for direct
consideration : namely, the part which the community
of speakers, as distinguished from the individual speaker,
have to play in language-making.

The community's share in the work is dependent
on and conditioned by the simple fact that language is
not an individual possession, but a social. It exists (as
we shall notice more particularly in the fourteenth chapter),
not only partly, but primarily, for the purpose of
communication ; its other uses come after and in the
train of this. To the great mass of its speakers, it exists
consciously for communication alone ; this is the use
that exhibits and commends itself to every mind. That
would have no right to be called a language which only
one person understood and could use ; and there is not,
nor has ever been, any such in existence. Acceptance
by some community, though but a limited one, is absolutely
necessary in order to convert any one's utterances
into speech. Hence arise the influences which guide
and restrain individual action on language. In the first
place, an individual's alterations and additions, if not
adopted by others and kept up in their tradition, die
with him, and never come to light at all. But again,
even if he were careless of offending the prejudices or
shocking the taste of his fellows, he would not, at any
rate, pass the limit of being intelligible to them ; and
this would be by itself a powerful brake to check his
149arbitrary action. But such a brake is unnecessary, because,
in the third place, each individual feels, in the
main, the governing force of the same motives which
sway the minds of his fellows. He does not himself
incline, any more than they would incline to allow him,
to abandon the established habits of speech and go off
upon a tangent, toward some new and strange mode of
expression. Everything in language goes by analogy ;
what a language is in the habit of doing, it can do, but
nothing else ; and habits are of very slow growth ; a lost
habit cannot be revived ; a new one cannot be formed
except gradually, and almost or quite unconsciously.
And the reason of this lies in the common preferences
of the speakers. We signify the fact popularly by saying
that such and such a thing is opposed to the “genius
of the language ;” but that is merely a mythological
term ; the German calls the same thing the Sprachgefühl,
‘speech-feeling,’ or ‘linguistic instinct :’ both
are expressions of a convenient dimness, under which,
inexact thinkers often hide an abundance of indefinite
or erroneous conceptions. What is really meant is the
sum, or resultant, of the preferences of the language-users,
as determined by the already existing material
and usages of their speech ; outside of certain narrow
limits of variation, they are not themselves tempted to
suggest, nor will they ratify and accept as suggested by
any one, new meanings, new phrases, new words.

Our recognition of the community as final tribunal
which decides whether anything shall be language or
not, does not, then, in the least contravene what has
been claimed above respecting individual agency. Some
one must lead the way for the rest to follow ; if they do
not follow, he falls back or stands alone. The community
cannot act save by the initiative of its single members ;
150they can accomplish nothing save by its cooperation.
Every new item in speech has its own time and
occasion and place of origination ; it spreads from, one
to another until it wins general currency, or else it is
stifled by general neglect. Only, of course, it is not
necessary that every single change should start from a
single point. There are some toward which the general
mind so distinctly inclines, which lie so close outside of
and within reach from the present boundaries of usage,
that they are made independently by many persons, in
many places, and thus have a variety of starting-points
from which to strive after currency. Probably it was
thus with its, when, two or three centuries ago, it was
crowded into English speech, against the outspoken
opposition of educated and “correct” speakers, by the
force of its apparent analogy with the general store of
English possessives ; probably the same was the case
with is being done, the corresponding passive form to
the continuous active is doing, as is done corresponds to
does — a phrase which, against a like opposition, has not
yet made its place entirely good in the best English
usage. Phonetic changes are especially likely to be thus
general, instead of solitarily individual, in their origin.
A very notable example is seen in the Germanic umlaut,
or modification of vowel (see above, p. 71) ; which,
since it is wanting in the Gothic, cannot have belonged
to the Germanic branches before their separation, but
was later developed independently in the High-German,
the Low-German, and the Scandinavian dialects, doubtless
as the final and accordant working-out of habits of
utterance which were already present in the unitary
Germanic dialect.

Having thus recognized the nature of the force
which, notwithstanding the strictness of linguistic tradition,
151is all the time altering the traditionary material,
and seen in what ways and under what inducements it
acts, we have next to view the same force, in the same
modes of action, as causing not only the variation of a
single language from age to age of its existence, but
also, under the government of external circumstances,
its variation in space, its divarication into dialects.152

Chapter IX.
Local and class variation of language : dialects.

Dialectic differences within the limits of a single language ; individual,
class, and local peculiarities of speech. What makes a language
one. Influences favoring or restraining dialectic differences ; effect
of culture. Illustration : Germanic language-history ; Romanic.
Centripetal and centrifugal forces ; separate growth causes dialectic
division ; examples. Verbal correspondences prove common
descent of words and languages ; cautions as to applying this principle.
Degrees of relationship. Constitution of Indo-European
family and evidences of its unity. Universality of families and
dialectic relations. Relation of terms “language” and “dialect.”

Our inquiries into the phenomena of speech have
thus far shown us that the mass of each one's language
is acquired by him by a process of learning, of direct
acquisition of what is put before his mind by others ;
that, however, each one is at the same time a partner
in the work of changing the language : contributing, indeed,
only an infinitesimal quota toward it, in exact proportion
to his importance in the aggregate of speakers
by whom the language is kept in existence, yet doing
his part in a sum which is all made up of such infinitesimal
parts, and would not exist without them. The tradition
of speech is carried on by him and such as he is ;
its modification is due to no other agency. Every item
of difference between new speech and old, whether in
153the way of alteration or of addition, has its separate origin,
beginning in the usage of individuals, and spreading
and seeking that wider acceptance which alone
makes language of it ; and it has its time of probation,
during which it is trying to establish itself.

But if this is true, then there must be in every existing
language, at any time, processes of differentiation
not yet fully carried out, words and forms of words in
a state of transition, altering but not altered ; words and
phrases under trial, introduced but not general ; words
obsolescent but not yet obsolete ; old modes of pronunciation
beginning to seem strange and affected, new
modes coming into vogue — and so on, through the
whole catalogue of possible linguistic changes.

And this is, in fact, precisely the state of things, in
every language under the sun : a state of things only
explainable by the causes which we have been considering.
It exists even in our own speech ; although here,
for reasons to be presently adverted to, the conditions
are more opposed to it than almost anywhere else in the
world. We must be careful not to overrate the uniformity
of existing languages ; it is far enough from
being absolute. In a true and defensible sense, every
individual speaks a language different from every other.
The capacities and the opportunities of each have been
such that he has acquired command of a part of English
speech not precisely identical with any one else's : the
peculiarity may be slight, but it is certainly there.
Then, what is yet more obvious and yet more important,
the form of each one's conceptions, represented by
his use of words, is different from any other person's ;
all his individuality of character, of knowledge, education,
feeling, enters into this difference. And yet
again, few if any escape the taint of local and personal
154peculiarities of pronunciation and phraseology, peculiarities
which, because more conspicuous than the others,
are more often noticed by us and called dialectic. This
last shades off into the more wide-spread and deeper
differences of district and class ; every separate part of
a great country of one speech has its local form, more
or less strongly marked — even where, as in America,
there are no old inherited dialects, of long standing,
such, as prevail in Britain, in Germany, in France : in
short, almost everywhere. Every class, however constituted,
has its dialectic differences : so, especially, the
classes determined by occupation ; each trade, calling,
profession, department of study, has its technical vocabulary,
its words and phrases unintelligible to outsiders ;
the carpenter, the iron-maker, the machinist, the miner,
not less than the physician, the geologist, or the metaphysician,
has occasion every day to say many things
which would not be understood by a man of any of the
other classes mentioned, if not exceptionally well-informed.
Then there are the differences in grade of
education ; the highly cultivated have a diction which
is not in all its parts at the command of the vulgar ;
they have hosts of names for objects and ideas of educated
knowledge, which (like dahlia, petroleum, telegraph,
instanced above) may perhaps some time work
their way down into the lower rank, becoming universal,
like is and head, and long and short, instead of
class-words only ; and, yet more especially, the uncultivated
have current in their dialect a host of inaccuracies,
offenses against the correctness of speech — as
ungrammatical forms, mispronunciations, blunders of
application, slang words, vulgarities ; all of these, perhaps,
analogous with alterations which the cultivated
speech, as compared with its predecessors, has undergone,
155and some of them destined to become at a future
time the established usage of the whole language ; but
as yet kept down in the category of errors by the resistance
of the higher classes to their acceptance and
use. Finally, there are the differences of age : the nursery,
in particular, has its dialect, offensive to the ears of
old bachelors ; and older children have their language
at least characterized by limited vocabulary.

Every one of all these differences is essentially dialectic :
that is to say, they differ not at all in kind, but
only in degree, from those which hold apart acknowledged
dialects. They all fall, as regards their origin,
under the classes of change already laid down : they are
deviations from a former standard of speech which have
hitherto acquired only a partial currency, within the
limits of a class or district ; or they are retentions of a
former standard, which the generality of good speakers
have now abandoned. In illustration of this latter
class, we may note in passing that no small number of
what the English stigmatize as Americanisms are cases
of survival from former good usage, and that, on the
other hand, much of what we regard as the peculiarities
of Irish pronunciation is also old English, more faithfully
preserved by the Irish than by the more native
speakers. Of course, it is as wrong to be lagging in the
rear of the great moving body of the usages of a language
as to be rushing on in advance, or flying off to
one side. When the speech of the best speakers
changes, those who do not conform have to be ranked
in a lower class.

And yet, despite all these varieties, the language is
one ; and one for the simple reason that, though the
various individuals who speak it may talk so as to be
unintelligible to one another, they may also, on matters
156of the most familiar common interest, understand one
another. As the direct object of language is communication,
the possibility of communication makes the unity
of a language. No one can define, in the proper sense
of that term, a language ; for it is a great concrete institution,
a body of usages prevailing in a certain community,
and it can only be shown and described. You
have it in its dictionary, you have it in its grammar ; as
also, in the material and usages which never get into
either dictionary or grammar ; and you can trace the
geographical limits within which it is used, in all its
varieties.

It is an obvious corollary from the view we have
taken of the forces governing the growth of language,
and of the way in which they act, that the quasi-dialectic
discordances existing within the limits of the same language
in the same community will be greatest where
the separation of classes and sections is greatest. The
necessity of communication is the restraint upon the
alterative processes, and communication is the means
whereby any alteration actually made is adopted by
all : whatever, then, makes communication most lively
and penetrating, through all regions and all ranks, will
tend to preserve the unity of speech most strictly
through the whole community. On the other hand,
all that dulls the forces of communication, and lets a
people break up into tribes, or into widely-sundered
castes or classes, tends to increase the discordance of the
forms comprehended together in the general language.

Different causes exert in this way a different influence.
On the one hand, in a barbarous condition of
society the discordances of class and occupation are at
their lowest. All members of the same community
stand substantially upon the same level ; with but insignificant
157exceptions, they have the same knowledge,
the same skill, the same habits ; the collective wealth of
thought and its expression is not too great for each person
to grasp and wield the whole of it. On the other
hand, local differences are at their highest point, since
it is only civilization and culture that can bind together
into one the parts of a great community. The influences
of barbarism, beyond narrow limits, are prevailingly
segregative ; a wild race that multiplies and
spreads widely breaks up into mutually jealous and hostile
divisions, within each of which linguistic changes
run their own independent course. Every element of
culture that finds its way in exercises a conservative influence,
tending both to preserve the language from
change and to preserve its unity throughout the territory
it occupies. The rise of a national feeling of so
high an. order that it reverences the deeds and the words
of past generations, and leads to the production of a
national literature, is obviously conservative, because it
amounts to setting up a norm of correct speech, by
which men's minds shall be influenced in judging, for
acceptance or rejection, the individual proposals of
change. A written literature, the habit of recording
and reading, the prevalence of actual instruction, work
yet more powerfully in the same direction ; and when
such forces have reached the degree of strength which
they show in our modern enlightened communities,
they fairly dominate the history of speech. The language
is stabilized, especially as regards all those alterations
which proceed from inaccuracy ; local differences
are not only restrained from arising, but are even wiped
out, so far as the effect of education extends. There is
also a state of things intermediate between the two extremes
of barbarism and all-pervading culture : namely,
158where there is culture which reaches only a particular
class, a minority, of the community, its conserving influences
being mainly limited to that class. This alone
possesses the records of the language, and, using them
as models, propagates its speech nearly unaltered, while
the language of the mass goes on changing unchecked.
There comes thus to be a separation of the originally
unitary speech into two parts : a learned dialect, which
is the old common language preserved, and a popular
dialect, which is its altered descendant ; and the latter,
perhaps, finally crowds the former out of existence, and
becomes, in its turn, the cultivated speech of a new
order of things. Such has been, for example, the history
of the Latin, and of the later dialects descended
from it, and now become the vehicles of great and noble
literatures ; such, also, that of the now cultivated languages
of modern Aryan India, in their relations to the
Sanskrit.

Let us suppose, then, that there is a definite community
X, of one speech. It is divided — not, of course,
by definite or fixed lines — into the various local parts
A, B, C, etc., and into the classes, whether social, vocational,
or educational, A, B, C, etc., and a, b, c, etc. ;
the various divisions variously overlapping and overlying
one another. The common speech is, like all living
speech, in a condition of constant growth and change ;
this change being possible, and actually occurring, only
by such acts of alteration as we have considered in detail
above, each arising at a point or points in one or
more divisions, and spreading thence by communication
to the rest. What arises thus in A, or B, or C, becomes
at length the possession of all — if, indeed, it does not
continue within certain limits, as a merely local dialectic
word or mode of expression. So what arises in A or a
159goes through the rest — unless it remain within the boundaries
of a class, as a technical term, a high-caste expression,
a popular blunder or vulgarism, or something
of the sort. And the amount and value of these various
residua, constituting the minor discordances which
may consist with general agreement and unity, is various
according to such determining circumstances as we
reviewed briefly in the paragraph next preceding : no
language is or can be without them, but they are very
different in different languages.

This whole state of things is dependent on historical
conditions, as concerns its continuance and
changes. Let us take our hypothetical case to represent
the German language as it was at and after the beginning
of our era. Here, while the divisions of class
and occupation were comparatively unimportant, those
of locality, A, B, C, etc., were very marked : so much
so, indeed, as to make it improper to speak of the
whole as one language ; besides innumerable minor discordances,
there were sections the speech of each of
which was not intelligible to the rest ; and if no new
force had been introduced, things might have gone on
thus to the end of time, the local discordances constantly
deepening and widening. But a new and controlling
force was introduced : that of Greco-Roman, soon to
become European, civilization : this led the way to institutional
and political unity. But not for a long time
did it win the predominance in the domain of language.
At first, each local division had its own separate culture ;
the beginnings of literature were produced, and are in
part still extant, in one and another local form of speech,
fully intelligible only within limits. But at length,
early in the sixteenth century, the fullness of time was
come ; political and educational conditions had reached
160a point where a movement toward an educated — and so,
in a certain sense, an artificial — unity of speech could
be made with success. A certain local form of speech,
A — which, to be sure, had already gained a degree of
currency as a class-form also — was definitely adopted by
the educated as their dialect, A, the style of German
which should thenceforth alone be written, and looked
up to as a model, and taught in the schools. And its
authority has ever since gone on increasing, with the
extension of the power of civilization and education, till
now an outsider almost looks upon it as the sole German
speech. That, however, it is far enough from being ; it
is still only A, the German of a class, though of a class
which the conditions of modern civilization have made
the dominant and the growing one. B, C, and D, etc.,
still subsist ; there are whole regions of Germany where
the local dialect is unintelligible to him who is versed
only in the literary language ; but they divide among
them, for the most part, only the classes of lower education,
E and F, etc. ; and they, as well as the classes of
vocation, a and b and c, etc., feel profoundly and in
various ways the influence of the learned speech. A
is the predominant speech, modifying and shaping everything
else in German usage, and even promising, if the
forces of education should ever attain that overwhelming
degree of importance, to sweep out of existence all
the other varieties, save those of occupation.

Not, however, as we must next notice, over the
whole territory occupied by High or Low German
tribes. There were at least two local varieties — we may
call them E and F — which did not fall under the unifying
influences that brought all the rest within the dominion
of A. One, E, the English, was cut off by distance
and inaccessibility, and consequent independence.
161The Germanic Angles and Saxons, who carried a German
dialect across the North Sea into Britain, and with
it displaced the old Celtic speech, have passed, in their
separateness, through a series of changes analogous with
those of their former fellow-countrymen. Their own
secondary divisions, of whatever kind — whether local, as
E′, E″, E‴, etc., or of class, as E′, E″, etc. — have been
in a similar manner brought under the controlling influence
of another literary dialect, of like origin with that
of Germany. And in the northeastern district of continental
Germany, the Netherlands, political independence,
with the consequent isolation of general interests,
had a kindred result ; while the rest of Low Germany,
speaking by local division forms of German speech not
less peculiar than those of the Anglo-Saxons and Dutch,
uses the High-German literary dialect as its learned
speech, the corner Holland and the colony England have
given an equivalent literary value to their separate Low-German
dialects. No matter how the local varieties
A and B and C become separated, so that what passes
in each is not participated in by the others, their development
will take a different course, and they will in
time become separate tongues.

The same forces, in like modes of action, but with,
abundant differences of detail, are seen at work in producing
the modern Romanic languages, descendants of
the Latin. When the arms and civilization and polity
of Rome carried her speech all through Italy, and over
great regions outside of Italy, it was already divided by
education into class-varieties. All were transmitted together ;
and the learned dialect — A, as we may call it,
in accordance with our use of this sign above — has been
kept up in its complete purity even to the present day,
by appropriate and adequate means, though in a constantly
162diminishing class. The lower forms of speech,
B, C, etc., had their full influence in laying the foundations
of the new history. The changes of Latin went
on, all the more rapidly for its having passed into the
keeping of races who had learned it at second hand, by
an outside pressure ; and, as the forces of communication
were very far from being sufficient to keep the
immensely extended community one, it broke up, by
differentiation within geographical limits, into a correspondingly
numerous array of local forms, for which it
would take several alphabets to provide sufficient symbols ;
and historical circumstances, which in their main
character and influence admit of being distinctly pointed
out, led to one here and another there — as C, and F,
and I, and P, and S, and W — being adopted as the
learned dialects of great regions, and used for literary
and educational purposes, not only by their own native
speakers, but also by those of the rest — which, like the
German dialects, still subsist as the uneducated patois
each of its own district.

It would be very easy to push, this illustration indefinitely,
but to carry it further is quite needless. The
methods of linguistic change detailed above, and governed
in their historical workings by the antithesis between
the initiatory action of the individual, and the
regulating action of the community in accepting or rejecting
his proposals — this has been all we have needed
to explain the historical phenomena instanced ; and this,
and this only, is sufficient to explain all the rest. It
may be fairly and confidently claimed that there is no
known case which cannot thus be solved. Individuals
are the diversifying or centrifugal force in the growth
of speech ; for, as there are no two persons absolutely
alike in countenance, so there are no two identical in
163character and education, and the shaping influence exerted
by each on the speech he has learned will be
slightly different from that of every one else. But just
so far as communication extends, like the centripetal
force, which dominates the other, and keeps the moving
body upon a certain track never too far remote from the
centre, the individualities are curbed and restrained, and
their jarring action forced into and held in accordance.
Or, in terms of our recent hypothesis, just so long as
every change which arises in the local parts A and B
and C, and so on, works its way through all the rest,
passing the ordeal of their acceptance or rejection, so
long will the language X remain one. It may and will
alter from age to age ; it may even become so changed
in two or three centuries (as English has actually become
in a thousand years) that its speakers at one and the
other end of that period would not, if they could be
brought together, understand one another at all ; yet,
at every period, all the community would understand
each other, because it would have changed alike in the
minds and mouths of all. But separate, in any way you
please, the parts A and B and C from one another, so
that the changes in each are made in that alone, and do
not extend into the rest, and the peculiarities of each
will begin to be confined to itself ; what we call dialectic
growth will set in ; the process of divarication into
diverse languages will have begun. A brick wall, high
enough and long enough, between the sections, would
perfectly accomplish their division, and initiate dialectic
divergence ; only, of course, if the separation takes place
by local removal, so that the sections are brought into
different external circumstances of nature and occupation,
and under different historical influences, the process
of linguistic divergence will be quickened.164

This cutting off, by cessation of communication, of
a common regulative influence over the never-ending
changes of speech, may seem a very slight cause of divergence ;
and so in truth it is ; but it is fully sufficient
to account for all the phenomena of dialectic growth.
No matter how small the angle may be between two
lines starting from the same point ; if they are protracted
far enough, their extremities may be found any given
distance apart. And the angle of dialectic divergence
is practically an increasing one ; the two lines of development
curve asunder. At the outset, namely, the
sum of guiding analogies in each is almost precisely the
same ; identity of material, and of habits of its use, is,
as it were, a continuance of the common momentum, carrying
the two on in almost the same direction ; and independent
accordant results of this community of original
habit may, as we have more than once seen above, continue
to appear for a long time, even indefinitely. But
each bit of difference that creeps in lessens the accordance ;
new habits arise, special disturbing influences set
in, and the distance comes at last, perhaps, to be rapidly
instead of slowly increased. The history of our English,
as compared with the Low-German dialects from which
it sheered off in the fifth and sixth centuries, is as striking
an example of this as could be desired.

Again, as dialectic discordance only arises in consequence
of linguistic growth, and as the maintenance of
an original condition of speech unchanged would do
away with all possibility of difference of speech among
the separated parts of the community which formerly
spoke it as one together, it is evident that the rate of
divergence must depend in great degree upon the general
rate of growth. And, as we have seen, the influences
of barbarism and of civilization are directly opposed
165to one another in this regard, although they are
by no means the only determining influences which
quicken or retard the alterative processes. It is the
predominant forces of civilization which, by a two-fold
action, have kept the language of the two great divisions
of English-speakers nearly accordant, notwithstanding
the broad ocean that rolls between them : first, by making
actual communication between them easier and
closer than between two tribes of rude people separated
only by a few miles of mountain or of plain, by a forest
or a river ; indeed, even by giving them, as it were, in
their common literature, a great body of speakers who
are all the time communicating with both ; and, in the
second place, by so restraining the activity of the alterative
processes that their results have time to reach and
permeate both divisions. Absence of the same conserving
influences causes the French of the habitans of Canada
and the German of the colony in Pennsylvania to
differ far more widely from the dialects of the countries
whence these colonists came.

The most instructive attainable example of dialectic
growth, on the whole, is that presented us in the Romanic
languages, because we have there a most important
and widely-spread body of highly cultivated
languages, each with its legion of subsidiary dialectic
forms ; and also — what is nowhere else to be had in anything
like the same measure — the very mother, the
Latin, from which they have all sprung. The student
of language finds in them a whole world of facts to study
and compare, to trace out in their origin and in the laws
which have produced them. And his task, though in
part simple and easy, is also in no small part difficult
and baffling ; for even here, under the eyes of history,
as it were, though hidden from them, have gone on
166changes which seem to defy investigation, producing
results which cannot be carried back to their sources.
Let us look at a specimen or two of the process of divarication,
as it has passed upon some of the materials
of the Latin original.

The Latin had a word for ‘brother,’ frater. In
French, the word, in the abbreviated form frère, still
bears the old office. But in Italian and Spanish, the
same word, having undergone still greater mutilation —
as Spanish fray, Italian frate and fra — signifies only a
‘brother’ of some ecclesiastical order, a friar, as we call
it, by yet another form of the same name. So, for
‘brother’ in its original and proper sense, each language
has had to provide a new word : the Italian takes the
diminutive fratello ; the Spanish puts to use the Latin
germanus, ‘nearly related,’ and says hermano. Again,
the Latin had the name mulier for a ‘woman,’ distinctively
as woman, besides femina for ‘female,’
woman or other. In Spanish, now, the former is still
retained, altered to muger, in nearly its ancient meaning ;
but in Italian, as moglie, it signifies only ‘wife’ or
‘spouse ;’ and in French it has utterly disappeared. In
French, femme, the representative of the other Latin
word, has become the general name for ‘woman,’ adding
also the meaning of ‘wife ;’ while for ‘female’ has
come to be used femelle (like Italian fratello for Latin
frater). For ‘woman,’ the Italian has shaped a new
word, donna, out of later Latin domina ‘mistress ;’
and the Spanish uses for ‘lady’ the same word donna,
besides señora, a feminine of modern make to senior,
‘older person.’ These are fair specimens of how the
original material of a language gets worked over, in
form and in meaning, in the keeping of the severed descendants
of that language. If we looked into the class
167of verbs, we should find the same condition of things.
The verb ‘be,’ for example, is made up of a remnant
of the forms of the Latin esse, pieced out in all the
dialects with parts of stare, ‘stand :’ so the French
étais, été, are stabam, status, with remarkable alterations
of form, one of which has been commented on above
(p. 54). And French aller, ‘go,’ is put together by
adding parts of Latin ire, ‘go,’ and parts of vadere,
‘walk,’ to a main stock of very obscure origin, representing
Latin adnare, ‘arrive by water,’ or aditare,
‘make one's adit, or arrival,’ or something of the sort.

Turning now to the Germanic dialects, our own nearest
relatives, we find the same kind of resemblance in
difference everywhere prevailing. The Germanic words
for ‘brother’ — as Netherlandish broeder, German bruder,
Icelandic brodhir, Swedish and Danish broder and bror
— are not less obviously the variations of one original
than are the Romanic products of frater. The old
Germanic weib, ‘woman,’ is found in most of the
modern languages, in easily recognizable forms, with
its former value ; but in modern English its representative
wife has become restricted (like Italian moglie) to
a married woman. And there is another ancient word,
Gothic quens and quinon, which in some dialects is the
accepted name for ‘woman,’ instead of the other, but
which in English has undergone the curious fate of being
divided into two terms, of lofty and humble meaning,
queen and quean. Our verbs be and go, too, like
their Romanic equivalents, are made up of fragments
from various roots, pieced together partly in more
ancient, partly in more modern times. Both we have
already noticed elsewhere in passing (pp. 90, 101) ; it is
unnecessary here to enter into any further detail respecting
them.168

From these and all the other innumerable correspondences
of the Germanic dialects we cannot possibly
help drawing the same conclusion which is taught us
by a comparison of the Latin with its descendants. It
is not one whit less certain that wife and weib and vif
and the rest are the variously altered representatives of
a single primitive Germanic vocable, than that moglie
and muger come from the Latin mulier. We may not
always, or often, be able to restore by inference the Germanic
word with a certainty equal to that inspired by
the actually preserved Latin word ; but that makes no
difference. We believe in the former existence of the
grandfather of a group of cousins, whom we have never
seen because he died long ago, just as thoroughly as in
the present existence of one whom we find still living
in the midst of another group. According to our experience
of how things go on in the world of human
beings and in that of words, there is no other possibility.
The processes of linguistic change, working regularly
on in the way in which we see them working in
the present and the recently past historic periods, are
fully sufficient to account for the existence in certain
languages of groups of words more or less resembling
one another yet not identical ; and there is no need that
we resort to adventurous hypotheses for its explanation.

This, legitimately generalized, gives us the great
principle that genuine correspondences, of whatever
degree, between the words of different languages, are
to be interpreted as the result of derivation from one
original : relationship, in words as in men, implies descent
from a common ancestor. And what is true of
the words of two languages is true of the languages
themselves : languages made up of related words must
be descended from a single common language.169

Only, to this principle need to be applied certain
cautions and corrections. Two sources of error require
to be guarded against in its use. First, words are borrowed
out of one language into another, as was fully
explained and illustrated in the seventh chapter. Certain
elements in English are of common descent with
elements in the Romanic and in many other of the
world's languages ; they have been handed over from
the tradition of one people into that of another : and
though there is so far a community of tradition, it does
not imply general relationship of the languages. Secondly,
accidental correspondences occur between words
which have no historical connection : so, for example,
between Greek ὅλος and our whole, between Sanskrit
loka and Latin locus, between Mod. Greek ματι, ‘eye,’
and Polynesian mata ‘see,’ and so on. These two difficulties
impose upon the comparer of languages the
necessity of increased caution in his work, and warn
him against over-hasty conclusions. An instance or two,
or a few instances, of verbal correspondence are not
sufficient to prove anything. But accidental resemblances
have their limit ; and it is in general possible to
distinguish borrowed material, so as not to be misled by
it into false inferences. The linguist looks to see both
how many and how close the asserted correspondences
are, and in what part of the vocabulary they are found.
If we did not know by external information the history
of English, we could still recognize it beyond all question
as essentially a Germanic dialect, by noticing what parts
of its material accord with the Germanic tongues, and
what part with the Romanic.

But relationship in language, as in genealogy, is a
thing of degrees, and for the same reason. The French,
Spanish, and Italian are cousins, on grounds which we
170have already sufficiently noticed ; but each is a group
of yet more closely related dialects. And so also among
the Germanic languages : the English belongs to a Low-German
group, still occupying the northern shores of
Germany, whence the ancestors of the English came ;
there is likewise a High-German group, occupying the
central and southern part of Germany ; and there is a
Scandinavian group, holding in possession Denmark,
Sweden and Norway, and Iceland ; moreover, there is
a single dialect, the Mœso-Gothic, of which limited records
are saved from extinction, and which represents
alone yet another group, of unknown extent. From
these minor groupings precisely the same inference is
to be drawn as from the larger ones : they represent
historical centres of more recent divergence, of the
same kind and by the same means as the others.

Nor does the finding of correspondences and tracing
of relationships end here. Between the Germanic brothar
and the Latin frater there is a pretty evident resemblance,
which becomes still more evident when we
put alongside of them other words of the same class, as
German mothar,fathar, and Latin mater, pater. But
there are yet other groups of languages which show
similar signs of relationship : we find in Greek φρατήρ
(meaning, to be sure, only a member of a confraternity,
like fray and fra, as noticed above) and μήτηρ and
πατήρ ; and, in Sanskrit, bhrâtar and mâtar andpitar ;
and the Persian and Celtic and Slavonic tongues have
in the same words correspondences which are like these,
though not quite so striking. These are telling indications
of an original relationship among all the groups
of languages mentioned : outcroppings, as it were, of
a vein which invites further exploration. For, in the
first place, the correspondences are too numerous and
171wide-spread and close to be explained with the slightest
show of plausibility as the result of chance ; and then,
there appears to be equally small hope of accounting
for them by borrowing. How should all these widely-sundered
tribes of men, found at the dawn of history
in every variety of cultural condition, have obtained
from a common source, or by transmission from one
to another, names for conceptions like these, the formation
of which must have accompanied the first development
of family life ? Plainly, all probabilities are
against it.

No confident conclusion, however, as to so important
a fact should be built on narrow foundations ; and
we look further, into other classes of words. There are
no savages in the world so undeveloped that they cannot
count ‘one, two, three’ — even though there are
those who have gone no further than that by their own
powers, but are either destitute of the higher numbers,
or have borrowed them from races more advanced.
If we find these numerals accordant in the languages
we have named, it will be a very strong piece of evidence
corroborative of that furnished by the names of
relationship. And the accordance exists, and is of the
most striking character, not only in these numerals, but
in all that follow : dwa is the common basis of the
various words for ‘two,’ and tri of those for ‘three,’
through the whole great mass of dialects. The pronouns,
again, are a class of words in which the suspicion
of borrowing is, if possible, even less to be entertained ;
and here also, in such words as those for ‘thou’ (two)
and ‘me’ (ma), in the demonstrative ta and the interrogative
kwa, we find a degree of agreement which is
quite beyond the power of accident to have produced.

Yet once more, we have seen (p. 119) that inflectional
172apparatus, grammatical structure, is most of all out of
the reach of a language that is borrowing from another.
But through all the grammatical apparatus of these
groups of dialects, when we can reach far enough back
in their history to find it preserved in a distinct form,
we discover an accordance not less convincing. Thus,
in the verbal inflection, there are the various alterations
of an original ending mi for the first person singular,
and of masi for the first plural ; of si and tasi for the
second person, and of ti and anti for the third ; of a
reduplication forming a perfect tense, of a sign of the
optative mood, and so on. In noun declension the
traces are more obscure and scanty, but still perceptible
enough. The comparison of adjectives is everywhere
by the same means. Participles and other derivative
words show the same suffixes of derivation.

In short, there is a superabundance of evidence going
to prove that the speech of all the peoples we have
mentioned, filling most of Europe, ancient and modern,
and an important tract of Asia, is related, in the sense
in which we have used that word above. There is no
theoretic reason against such a fact ; rather, every conclusion
drawn from the phenomena of existing speech
makes directly in its favor. We know that the separation
and isolation of the different parts of a once unitary
community must necessarily bring about a separation
of its language into different dialects ; and we
know that this process may go on repeating itself, over
and over again ; and that, at the end, those dialects
which parted latest will (apart from special altering
forces), though unlike, be least unlike and most like one
another, while those which parted earliest will be least
like and most unlike one another : and we know of no
other way in which this likeness in unlikeness can be
173brought about. We infer, then, that all the languages
in question are the divaricated representatives of a single
tongue, spoken somewhere and somewhen in the
past by a single limited community, by the spread and
dispersion of which all its discordances have in the
course of time grown up. Such a grand congeries of
related languages, in different degrees, we are accustomed
to call a “family :” a name taken, by an allowable
figurative transfer, from the vocabulary of genealogy.

This is an example of the way we are to proceed to
examine and classify all the various languages which
the earth contains. The first steps in it are easy enough.
It takes no conjurer to discover that London English and
Yorkshire English and Scotch English and negro English,
even, are all one language ; and no observant person,
probably, who learns German or Dutch or Swedish,
fails to see that he has in hand a tongue akin with his
own. But it takes a more penetrating and enlightened
study to pick out the signs of original unity amid the
greatly more conspicuous differences of English, French,
Welsh, Russian, Romaic, Persian, and Hindi ; and it
requires especially a resort, in the case of each language,
to the older tongues of its own nearer kindred,
which have preserved the ancient common material
with less change. Only the learned and experienced
investigator, therefore, can be trusted to push the work
of classification safely to its extreme limits ; and the
classification of all human tongues is only attainable by
the labors of a great number of investigators, each
learned in his own special department. Nor has it
been even thus by any means finished ; yet much has
been done toward it : the vast majority of languages
have been grouped together by their affinities into families
174and branches of families ; and the results of this
classification have to be briefly reviewed by us in the
following chapters.

For, as might be expected to follow from the principles
laid down above as determining dialectic growth,
there is not a language in the world which does not exist
in the condition of dialectic division, so that the
speech of each community is the member of a more or
less extended family — unless, indeed, there may be here
and there an isolated language so nearly extinct as to
be used only by the narrowest possible community : by
a few families, or a single village. Even languages of
so limited area as the Basque in the Pyrenees, as some
of the tongues in the Caucasus, have their well-marked
dialectic forms ; because an uncivilized people can
hardly break up even into camps, and still maintain
that communication which alone can keep their speech
a unit.

This linguistic condition of the earth runs parallel,
in the closest manner, with its social and political condition.
At the very beginning of history, and even as
far beyond as archaeological science can penetrate, the
earth is all peopled, more or less thickly, with a seemingly
heterogeneous mass of clans and tribes and nations.
But not even the most heterodox naturalist who
holds to a variety of origins for the human race believes
these all to have sprung out of the ground, as it were,
where they stand : they come from the multiplication
and dispersion of a certain limited number of primitive
families, if not, as many think, from that of a single
family. So with language : at the first attainable period
of our knowledge of it, whether by actual record
or by the inferences of the comparative student, it is in
a state of almost endless subdivision ; and yet every
175sound linguist holds, and knows that he has the most
satisfactory reasons for holding, that this apparent confusion
is a result of the extension and divarication of a
certain limited number of primitive dialects — whether
of a single one, is a question which we shall have later
to consider our right to determine. At the earliest
historical period, too, the darkness of barbarism covers
the earth in general ; the centres of culture are but two
or three, and their light spreads but a very little way,
and is even in constant danger of being extinguished
by the greatly superior brute force of the uncultivated
masses around. Hence the divaricating forces in linguistic
growth are also in the ascendant ; dialects go on
multiplying, by the action of the same causes that had
already produced them. But wherever civilization is at
work, an opposite influence, in linguistic as in political
affairs, is powerfully operating. Out of the congeries of
jarring tribes are growing great nations ; out of the Babel
of discordant dialects are growing languages of wider and
constantly extending unity. The two kinds of change
go hand in hand, simply because the one of them is dependent
on the other : nothing can make wide unity of
speech except extended community ; nothing but civilization
can make extended community. As, through the
ages of recorded history, the power as well as the degree
of civilization has been constantly growing, till now it is
the predominant force, and the uncivilized races subsist
only by the toleration of the civilized — if even that ; so,
by external forces, every act and influence of which is
clearly definable, the cultivated languages have been
and are extending their sway, crowding out of existence
the patois which had grown up under the old order of
things, gaining such advantage that men are beginning
to dream of a time when one language may be spoken
176all over the earth. And, though the dream may be
Utopian, there is not an element of the theoretically
impossible in it ; only a certain condition of external
circumstances is needed to render it inevitable.

It is possible so to misunderstand these facts in the
wide history of human speech as to believe that language
actually began in a condition of infinite dialectic
division, and has been from the outset tending toward
concentration and final unity. But that is possible only
by a total failure to comprehend the forces that are at
work in the growth of language, and the modes of
their interaction. Tell the ethnologist that the beginnings
of the human race were an indefinite number of
unconnected individuals, who first coalesced into families,
and these into clans and tribes, and these into confederacies,
whence came nations, and whence may yet
come, by the same natural tendency to unity out of diversity,
a single homogeneous race all over the earth —
and he will hardly pay the theory the compliment even
of laughing at it. And the corresponding linguistic
view is really just as absurd ; only, from the greater
obscurity or unfamiliarity of the conditions involved,
not so palpably absurd, and therefore not so ludicrous.

Before closing this chapter, we must notice for a
moment the meaning of the terms language and dialect,
in their relation to one another. They are only two
names for the same thing, as looked at from different
points of view. Any body of expressions used by
a community, however limited and humble, for the
purposes of communication and as the instrument of
thought, is a language ; no one would think of crediting
its speakers with the gift of dialect but not of language.
On the other hand, there is no tongue in the
world to which we should not with perfect freedom
177and perfect propriety apply the name of dialect, when
considering it as one of a body of related forms of
speech. The science of language has democratized our
views on such points as these ; it has taught us that
one man's speech is just as much a language as another
man's ; that even the most cultivated tongue that exists
is only the dialect of a certain class in a certain locality
— both class and locality limited, though the limits may
be wide ones. The written English is one of the forms
of English, used by the educated class for certain purposes,
having dialectic characters by which it is distinguished
from the colloquial speech of the same class,
and yet more from the speech of other classes or sections
of the English-speaking community — and each one of
these is as valuable to the comparative student of language
as their alleged superior. But English and
Dutch and German and Swedish, and so on, are the
dialects of Germanic speech ; and the same, along with
French and Irish and Bohemian, and the rest, are the
dialects of the wider family whose limits we have
drawn above. This is the scientific use of the terms ;
in the looseness of popular parlance, an attempt is made
at the distinction of degrees of dignity and importance
by means of the same words, as when the literary language
of a community is alone allowed the name of
language, and the rest are styled dialects. For ordinary
purposes the usage is convenient enough ; but it has no
acceptableness on other grounds ; it forms no part of
linguistic science.178

Chapter X.
Indo-european language.

Genetic classification. Indo-European family ; its names ; its branches
and their earliest records : Germanic, Slavo-Lettic, Celtic, Italic,
Greek, Iranian, and Indian ; doubtful members. Importance of thie
family ; value of its study to the science of language. Time and
place of original community impossible to determine. Scientific
method of studying its structural history ; form-making by composition
and integration ; sufficiency of the principle. Resulting doctrine
of original radical monosyllabism ; Indo-European roots. Development
of forms : structure of verb, of noun ; pronouns ; adverbs
and particles ; interjections, their analogy with roots. Question of
order of development, and time occupied. Synthetic and analytic
structure.

Having examined, with all the fullness which the
space at our command allows, the foundation on which
a genetic classification of the languages of the world
reposes, we are ready to undertake a brief view of that
classification, as established by the researches of linguistic
scholars. We have seen that correspondence in the
material of different languages, if existing in measure
and kind beyond what can be accounted for as the result
of accident or of borrowing, is explainable only as
due to the separate tradition of an originally common
tongue, a tradition which preserved a part of the original
usages, while it modified or discarded other parts,
179or introduced what was new, to such an extent as to
obscure, and perhaps even to hide, the evidences of former
connection. As an example, we glanced at an outline
of the great family of related tongues to which
our own belongs, and noticed a limited but sufficient
specimen of the evidence on which is founded the general
belief in its unity as a family. We have now to
go on and lay down more definitely the constitution of
this family, and to sketch its structure and its structural
history.

It is called, in the first place, by a variety of names,
no one of which has fully established itself in general
use. We will employ “Indo-European,” as having on
the whole the best claim ; it was deliberately adopted
by Bopp, the great expounder of the relations of the
family, and is as widely used as any of the others.
Most of Bopp's countrymen now prefer “Indo-Germanic,”
for no other assignable reason than that it contains
the foreign appellation of their own particular
branch, as given by their conquerors and teachers, the
Romans. Others, rejecting both these titles as cumbrously
long, say instead “Aryan,” which also has a
wide and perhaps a growing currency ; the chief objection
is, that it properly belongs only to the Asiatic
division, composed of the Iranian and Indian branches,
and is still needed and widely used to designate that
division. “Sanskritic,” from the oldest and in some
respects the leading language of the family, and “Japhetic,”
from the son of Noah to whom are attributed
as descendants in the Genesis some of the people speaking
its various dialects, are terms of limited and now
obsolescent employment.

The Indo-European family, then, is composed of
seven great branches : the Indian, the Iranian or Persian,
180the Greek, the Italic, the Celtic, the Slavonic or
Slavo-Lettic, and the Germanic or Teutonic.

Taking these up in their inverse order, we have
first the Germanic branch, in the four principal divisions
already noted : 1. The Mœso-Gothic, or dialect
of the Goths of Mœsia, preserved only in parts of a
Bible-version made by their bishop Ulfilas in the fourth
century of our era, being long ago extinct as a spoken
language. 2. The Low-German languages, still spoken
in the north of Germany, from Holstein to Flanders,
and across in the neighboring England, and including
two important cultivated tongues, the Netherlandish
and the English. English literary monuments go back
to the seventh century, Netherlandish to the thirteenth ;
and there is an “Old-Saxon” poem, the Heliand, or
‘Savior,’ from the ninth, and Frisian literature from
the fourteenth. 3. The High-German body of dialects,
represented at the present day by only a single literary
language, the so-called German, of which the literature
begins with the Reformation, in the sixteenth century ;
back of this, the New High-German period, lie a Middle
and an Old High-German period, with their literatures
in various somewhat discordant dialects, reaching
back into the eighth century. 4. The Scandinavian
division, written in the forms of Danish, Swedish,
Norwegian, and Icelandic. The Icelandic monuments
go back to the twelfth and thirteenth centuries, and are
in point of style and content older than anything in
High or Low German : the Edda is the purest and
most abundant source of knowledge for primitive Germanic
conditions. The Icelandic is also, especially in
its phonetic state, the most antique of living Germanic
dialects. Besides these literary remains, there are brief
Runic inscriptions, generally of but a word or two, going
181back, it is believed, even to the third or second century.

The Slavonic branch has always lain in close proximity
to the Germanic, on the east ; it has been the
last of all to gain historical prominence. Its eastern
division includes the Russian, Bulgarian, Servian and
Croatian, and Slovenian. The Bulgarian has the oldest
records ; its version of the Bible, made in the ninth
century, in the same region where the Gothic version
had been made five centuries earlier, has become the
accepted version, and its dialect the church language,
throughout the Slavonic division of the Greek church.
The Russian is by far the most important language of
the whole branch ; it has remains from the eleventh
century ; some of the southern dialects present specimens
from a yet remoter date. To the western division
belong the Polish, the Bohemian, of which the Moravian
and Slovakian are closely kindred dialects, the
Sorbian, and the Polabian. There is nothing in Polish
earlier than the fourteenth century ; Bohemian records
are believed to go back to the tenth.

This branch is often called the Slavo-Lettic, because
it is made to include another sub-branch, the Lettic or
Lithuanic, which, though considerably further removed
from the Slavonic than any of these from the rest, is
yet too nearly related to rank as a separate branch. It
is composed of three main dialects : the Old-Prussian,
extinct during the past two centuries, the Lithuanian,
and the Livonian or Lettish ; all clustered about the
great bend of the Baltic. The Lithuanian is the most
important and the oldest, having records from the middle
of the sixteenth century. It exhibits in some respects
a remarkable conservation of ancient material
and form.182

The Celtic branch is one which from the beginning
of history has been shrinking in extent, till it now occupies
only the remotest western edges of the immense
region of western and central Europe which it formerly
possessed. Not enough is known of the ancient Celtic
dialects of northern Italy, of Gaul, of Spain, to show
what was their place in the sub-classification of the
branch. The preserved dialects compose two groups,
usually called the Cymric and Gadhelic. The Cymric
includes the Welsh, with “glosses” from the ninth century
or thereabouts, and a literature from the twelfth,
but of which part of the substance is probably older,
even up to the sixth ; the Cornish, which became extinct
as a vernacular about the end of the last century,
leaving a considerable literature nearly as old as the
Welsh ; and the Armorican of Brittany, so nearly allied
to the Cornish that it is believed to belong to fugitives
from that part of England ; its earliest records are of
the fourteenth century. The Gadhelic group includes
the Irish, which has monuments going back to the end
of the eighth century, the Scotch Gaelic, of which the
earliest remains are attributed to the sixteenth, and the
insignificant dialect of the Isle of Man.

The Italic branch is represented among living languages
only by the Romanic dialects, so called as being
all descended from the dialect of Rome, the Latin. We
have already noticed some particulars affecting their
history and their importance. They all rose at not far
from the same period — namely, the eleventh to the thirteenth
centuries — out of the condition of local patois,
products of the corruption of the popular speech while
the Latin continued the language of the learned. Fragments
of French are oldest, coming from the tenth
century ; its literature begins one or two centuries later ;
183the earliest Italian, Spanish, Portuguese, are from the
twelfth, or hardly earlier. These four are the conspicuous
modern members of the group. But there was also,
in the eleventh to the fourteenth centuries, a rich literature
of the chief dialect of southern France, the Provençal,
which, except for a recent sporadic effort or two,
has been ever since unused as a cultivated tongue.
There exists, too, in the northern provinces of Turkey,
in Wallachia and Moldavia, a broad region of less cultivated
Romanic speech, witness to the spread of Roman
supremacy eastward : it is destitute of a proper literature.
Moreover, certain dialects of southern Switzerland
are enough unlike Italian to be ordinarily ranked
as an independent tongue, under the name of Rhæto-Romanic,
or Rumansh.

The ancient members of the Italic branch, coordinate
with the Latin, were long ago crowded out of existence ;
but a few remains of them are still left, especially
of the Umbrian, north from Rome beyond the
Apennines, and of the Oscan of southern Italy. The
Latin itself, in its oldest monuments, reaches hardly
three centuries beyond the Christian era, appearing
there in a form which seems very strange, and is hardly
intelligible, to those who have learned only the cultivated
dialect of the last century B. C.

The Greek branch attains a much greater age, those
masterpieces of human genius, the poems of Homer,
being nearly or quite a thousand years older than our
era. From about 300 B. C., all Greek is written in the
Attic or Athenian dialect, as all modern German literature
in the New High-German ; but before that time,
as in the Old High-German period, each author used
more or less distinctly his own local dialect ; and in this
way, as well as, more widely but less abundantly, by
184inscriptions and the like, we have a tolerably full representation
of the local varieties into which the Greek
had divided in prehistoric times. There is, of course,
a similar variety of dialects now ; but only one is written,
and it is called Modern Greek, or Romaic ; it is
less altered from the classic Greek than is the Italian
from the Latin. Notwithstanding the wide sway of
Greek civilization, the spread of Greek empire under
Alexander and his successors, and the unexcelled character
of the language, the latter has had a limited and
inconspicuous career as compared with the Latin : out
of Greece itself, it is spoken only on the islands and
shores of the Adriatic, and along the northern and
southern edges of Asia Minor.

The next branch is the Persian, or properly Iranian,
since Persia is only one among the many provinces constituting
the territory of Iran (Airyana, the home of
the western Aryans). It has two ancient representatives :
the Old Persian, or Achæmenidan Persian, of
Darius and his successors ; and the language of the
Avesta, the so-called Zend, or Avestan, or Old Bactrian.
The former, of determinate date (five centuries B. C.), is
read in the cuneiform inscriptions, recently deciphered ;
of the other, the date is unknown ; it may be older or
younger. The Avesta is the Bible of the Zoroastrian
faith, of which the date and place of origin are obscure ;
it is believed to reach beyond 1000 B. C. ; and if parts of
the record are, as they claim to be, from Zoroaster himself,
they have this antiquity. The modern votaries of
the religion, and the keepers of its sacred books, are the
Parsis of western India, fugitives from Mohammedan
persecution in their native land. With the Avesta,
they have preserved a version of it in the Huzvâresh or
Pehlevî, of the time of the Sassanids, a dialect of peculiar
185and problematical character. The Modern Persian
literature, abundant and rich, begins from about
A. D. 1000, after the country had been ground over in
the Mohammedan mill.

These are the members of the main body of Iranian
speech. The Kurdish is only a strongly-marked dialect
of the same stock. The Ossetic, in a little province of
the Caucasus, is plainly, but more distantly, related.
The Armenian, of which the considerable literature goes
back to the fifth century — but, it is recently claimed,
with cuneiform fragments a thousand years or more
older — is also of Iranian type. Finally, the Afghan,
near the border of Iran and India, is usually reckoned
as Iranian, but by some recent trustworthy authorities
regarded as rather Indian.

The branch of Indo-European language in India does
not cover the whole of that vast territory ; the Dravidian
race, which was doubtless crowded out by the intrusive
Aryans in the north, still occupies the main
central part of the southern peninsula, the Dekhan.
The earliest of Indo-European tongues is the Sanskrit,
especially its earlier or Vedic dialect, the dialect of the
religious hymns, which, with auxiliary literature of
somewhat later date, became the Bible of the Hindus,
the so-called Veda. At the period of the oldest hymns,
the Sanskrit-speaking peoples appear to have been not
yet in possession of the great Ganges basin, but nearly
or quite confined, rather, to the valleys of the Indus and
its branches, in the northwestern corner, the region
bordering nearest on Iran. The date is incapable of
being determined with any exactness ; probably it was
nearly or quite 2000 B. C. The classical Sanskrit is a
dialect which, at a later period, after the full possession
of Hindustan and the development of Brahmanism
186out of the simpler and more primitive religion and
polity of Vedic times, became established as the literary
language of the whole country, and has ever since maintained
that character, being still learned for writing and
speaking in the native schools of the Brahmanic priesthood.
From the fact that inscriptions in a later form
of Indian language are found dating from the third century
B. C., it is inferred that the Sanskrit must at least
as early as that have ceased to be a vernacular tongue.
The next stage of Indian language, to which the inscriptions
just referred to belong, is called the Prakritic.
One Prakrit dialect, the Pali, became in its turn
the sacred language of southeastern Buddhism, and is
still taught and learned as such in Ceylon and Farther
India ; the others are represented partly in the Sanskrit
dramas, as the unlearned speech of the lower orders of
characters, and partly by a limited literature of their
own. Finally, there are the modern dialects of India,
numerous and various, but rudely classifiable under
the three comprehensive names of Hindî, Mahrattî, and
Bengâlî, having literatures of more recent origin. The
so-called Hindustânî, or Urdu, is Hindî with a great infusion
of Arabic and Persian words, introduced by Mohammedan
influence.

The boundaries of this great family are more distinctly
drawn than those of any other. But they are
not absolute or immovable. There are one or two
isolated tongues in Europe which may yet be pronounced
Indo-European. Thus, the Skipetar, or language
of the Albanians, on that part of the west coast
of European Turkey which lies close opposite the heel
of Italy : it is believed to be the representative of the
ancient Illyrian, and more probably Indo-European than
anything else. And the Etruscan, the obscure and
187much-discussed tongue of that peculiar people with
whose relations to the early Romans, until finally conquered
and assimilated by Rome, every school-boy is
familiar, after being assigned to almost every distant
race on earth, is now (1874) declared Indo-European
and Italican by scholars of such rank and authority that
the conclusion must stand as probable until completely
refuted. It is evident enough that in theory such, cases
of doubtful classification are to be expected. There is
no limit to the degree to which a language may, by
special disturbing causes, become altered in its material
and structure, even to the effectual disguise of its original
relationships.

There are many reasons why the Indo-European
family is of predominant importance among the languages
of the world ; why it has thus far received a
very large share of the attention of linguistic scholars,
and must always continue to receive, even if not the
same share as hitherto, yet a larger than any other family.
The least of these reasons is that it is our own
family ; though that is, after all, no illegitimate plea in
enhancement of the interest with which it is invested
for us. Of more importance is the circumstance that
it belongs to the race which has long been the leading
one in the history of the world, and which at the present
day, as for some time past, has not even a rival. The
grand and highly-developed institutions of great nations
are those which most demand and best repay study.
The tongues and the history of the Greeks and Romans
are that part of antiquity which will continue to form,
even as it constitutes at present, a leading subject in all
liberal education. And the whole history of Indo-European
language will have its share by reflection in
this educational value, because it casts light on the study
188of Greek and Latin, of the Romanic languages, of the
Germanic languages, of the Slavonic languages, on all
that is nearest and dearest to those nations which are
pursuing the study.

But there are other and more imperative reasons
why the study of Indo-European language has been the
training-ground of the science of language ; why the two
have almost grown up together, and in the minds of
some have even perhaps been confused and identified
with one another. The student has at best a most imperfect
and fragmentary record before him. If the
whole history of human speech were represented by a
great sheet of paper, the part of it to be marked as
known, or as accessible to direct knowledge, would be
almost ludicrously small. For most human races, only
the present spoken dialects lie within reach ; then a few
lines of light run back into the past to various distances
toward the Christian era ; a much smaller number beyond
that point ; four or five, probably, into the second
thousand years before Christ ; and only one, the Egyptian,
to a time considerably remoter yet. And how
much of language-history, as of human history in every
department, may lie behind even that point, we are
only recently beginning to realize. Such being the
condition of the whole field, how was a fruitful beginning
to be made except just as it has been made — namely,
by taking up that body of historically-related facts
which was widest in its range, deepest and most abundant
in its penetration of the past, and most advanced
in its development out of original conditions ? By
grasping this and reducing it to manageable order, discovering
the general hidden under the particular, tracing
tendencies and laws, the student might hope to acquire
the ability to deal with other like bodies of facts, of
189narrower range and offering less abundant facilities.
The character of preeminence in this line belongs to the
Indo-European, beyond dispute and beyond comparison :
where we have equal or greater penetration of the past,
as in Egyptian, Chinese, and the Semitic tongues, there
is either (as in the two former) a peculiar jejuneness of
development, or at any rate (as in the last) a variety
and wealth which is greatly inferior. To blame philologists,
therefore, for their devotion hitherto to Indo-European
study is in the highest degree unreasonable ;
one might as properly blame historians for their devotion
to the study of European civilization and of its
sources in the past. To cast reproach upon them, moreover,
for their attention to the past, to the partially
understood records of extinct and almost forgotten
tongues, and to declare that the true and fruitful field
for linguistic research is the living and spoken dialects
of the present day, is not less narrow and erroneous. It
overlooks the character of linguistics as a historical science ;
it forgets that the explanation of the present is
by the past, and that the record of by-gone conditions
casts on existing conditions a light that nothing else could
yield. More precisely, it exaggerates and pushes forward
unduly the equally true fact that the comprehension
of the past is complete only by the help of the present.
It would be most unfortunate to check the zeal of those
who are submitting present language to the most rigorous
investigation, especially on its phonetic side, or to
cast the slightest reflection on the deep and far-reaching
value of their work ; there is hardly another more
promising direction of linguistic inquiry : only they, on
their side, should refrain from impliedly contemning
their predecessors, and should realize that they are striking
in now when the way is prepared for making their
190labors fruitful. So the minute study of the customs,
institutions, beliefs, and myths of rude peoples now existing
was, not long ago, comparatively a mere matter
of curiosity ; it gains its most valuable bearing from the
study of civilization in its historical development. It
was of little use to watch, and study nebulœ until geology
and astronomy together had learned so much about
the constitution and history of our solar system as to
have found out how to interpret the facts observed.

So also, in the claims here put forth as to the priority
and preeminence of the Indo-European tongues as
a subject of linguistic study, there is nothing which,
must be in the slightest degree understood as depreciating
the importance of the study of other families, even
its indispensability to the comprehension of Indo-European
language itself. The science of language is what
its name implies, a study of all human speech, of every
existing and recorded dialect, without rejection of any,
for obscurity, for remoteness, for lowness of development.
The time has come when questions are rising
in abundance in the history of Indo-European speech
which cannot possibly be answered until the languages
of lower organization are more thoroughly understood.
And it must be distinctly laid down as a fundamental
principle in linguistics, that no fact in human expression
is fully estimated, until it is seen in the light of related
facts all through the domain of universal expression.
Only, it is not possible, in philology any more than in
other branches of study, to help letting facts arrange
themselves along certain leading lines, and converge
their light where light is most desired.

We have reached, as was seen above, the certain
conclusion that all the known Indo-European languages
are descended from a single dialect, which must
191have been spoken at some time in the past by a single
limited community, by the spread and emigration of
which — not, certainly, without incorporating also bodies
of other races than that to which itself belonged by origin — it
has reached its present wide distribution : even
as a similar process, in historical times, has brought its
two leading modern branches to fill the New World, a
region almost vaster than that which it occupies in the
Old. Of course, it would be a matter of the highest
interest to determine the place and period of this important
community, were there any means of doing so ;
but that is not the case, at least at present. As for the
time, the less said about that the better, in this transitional
period of opinion as to the age of man on the
earth. The question whether the first man was born
only 6,000 years ago, or 12,000, or 100,000, or 1,000,000,
as the new schools of anthropology are beginning to
claim, is one of which the decision must exercise a controlling
influence on that which we have here in view.
As for the testimony of language itself, there is none,
of any authority ; the philologists will doubtless claim
that they do not see how to compress the growth of
Indo-European language into the shortest of the periods
named, but they have not yet found a rule with which
to measure the time they actually need. To give even
a conjecture at present would be foolish.

Nor is the place perceptibly easier to determine.
Man has ever been a migratory animal, and if he has
had a million years, or a tenth part of the number, to
wander in, it must be next to impossible to fix the
starting-point of any division of the race. How little
could be inferred as to the history of movement of the
Celts from their present distribution ! If some barbarous
race had conquered and exterminated or absorbed
192the Germans of the continent, what erroneous conclusions
might not be drawn from their presence only in
Scandinavia and Iceland ! And there are probabilities
of just as baffling occurrences in the history of the Indo-Europeans.
Men have long, and on well-known grounds,
been accustomed to look upon the southwestern part of
Asia as the cradle of the human race ; and, mainly
der the influence of this opinion, so long rooted that it
sways the minds even of those who reject the authority
of the testimony on which it is founded, it is by many
asserted with great confidence that the Hindu-Kush
mountain-region, or that Bactria, is the Indo-European
cradle : the only bit of tangible evidence which they are
able to allege being that that is the region where the
Iranians and Indians separated, and that the Iranian and
Indian dialects are the most primitive of the family. But
to plead this is equivalent to maintaining that slowness
or rapidity of change in language is dependent on stability
or change of place in the speaking community : which
is so grossly wrong that it needs no refutation. In fact,
the condition of these languages is reconcilable with any
possible theory as to the original site of the family. As
to the interconnections of the different branches with one
another, the best scholars have for some years past been
settling down upon the opinion that the separation of
the five European branches from one another must have
been later than their common separation from the two
Asiatic branches, which latter then continued to exist as
one community almost down to the historical period.
Upon this last point, there is unanimity of opinion ; the
oldest forms of Persian and Indian speech are as closely
like one another as, for instance, the more dissimilar of
the Germanic dialects ; the two branches are ranked together
under the common name of “Aryan ;” and the
193Indian branch is thought to have parted from the common
home in northeastern Iran not very much earlier
than 2000 B. C. Within the European grand division,
the Germanic and Slavonic are by nearly all regarded
as specially related ; opinions are more nearly divided
as to whether the Celtic is a wholly independent branch,
or closely akin with the Italican. In all this there is
evidently nothing which should point our eyes definitely
toward an original home. The separation of Aryan
from European may just as well be due to a spread
and migration of the former into Asia as of the latter
into Europe : and localities in Europe as well as in
Asia have actually been pitched upon by eminent
scholars. But it is useless to pretend to come to a
definite conclusion where the data are so indefinite.
Evidences of real weight bearing on the question may
possibly yet be found ; but certainly none such have
been hitherto brought to light.

Owing to the exceptional abundance of the material
for study of the history of Indo-European speech, and
the amount of study which has been devoted to it, it is
far better understood than is the history of any other
division of human language. Partly, therefore, because
of the high intrinsic interest of the subject, and partly
as a standard of reference in the treatment of the structural
growth of other languages, we have to follow out
in a little detail, though still with all possible brevity,
the ascertained history of the common foundation of the
Indo-European languages.

But we have first to consider the question — if, indeed,
it can be called a question — as to how the prehistorical
periods of language are to be investigated. Not
even the Indo-European has more than a small part of
its history illustrated by contemporary documents : how
194are we to learn anything beyond the point where the
records fail us ? The answer, it is believed, is a plain
and a confident one : we have to study the forces at
work under our observation, and the methods of their
working ; and we have to carry them back into the past
by careful analogical reasoning, inferring from similar
effects to similar causes, just as far as the process can be
made to work legitimately, never assuming new forces
and modes of action except where the old ones are absolutely
incapable of furnishing the explanation we are
seeking — and, even then, only under the most careful
restrictions. This is the familiar method of the modern
inductive sciences ; and its applicability to the science of
language also is beyond all reasonable doubt. The parallel
between linguistics and geology, the most historical of
the physical sciences, is here closest and most instructive ;
and it has often been resorted to for illustration. The
geologist infers the mode of formation of ancient sand-stones
and conglomerates from that of modern sand-banks
and gravel and pebble-beds ; and so on, through
the whole series of strata, sedimentary and eruptive ; he
accounts for the occurrence of fossils by the engulfing
or burying of extant species. And the true geologic
method has been so thoroughly worked out, and is so
strictly applied, that the scientific man who abandons it,
and resorts to arbitrary hypotheses, even to account for
facts which for the time seem unexplainable by ordinary
means, is at once put down as “unscientific,” and
bidden to wait until the growth of knowledge shall
bring around the possibility of solving his problem,
if it shall finally be found soluble, in an admissible
way.

Of course, the circumstances and conditions of action
of the same forces may differ greatly. The admission
195of the unity of geologic history by no means implies
that the earth has always worn the same aspect as at
present ; it is even a prevailing opinion among geologists
that the whole solar system was once a nebulous
mass of whirling vapor ; but this result is reached by
the inductive method. The essential unity of linguistic
history, in all its phases and stages, must be made the
cardinal principle of the study of language, if this is to
bear a scientific character. To assume outright, as some
do, either explicitly or impliedly, that ancient modes of
language-making were and must have been different
from modern, and that the former are not to be judged
by the latter, would, if linguistic science were as matured
and well-established a branch of study as geology,
be enough to exclude the assumer from the ranks of
scientific linguists. Here, again, the difference of conditions,
of the grade of historic development, has to be
fully allowed for ; and the student may arrive at the
recognition of a primitive condition of language to
which the present is as unlike as a civilized country,
teeming with the public and private works of its inhabitants,
is unlike the wilderness through which the savage
roams ; or even as the existing cosmos is unlike the
nebulous chaos ; yet the present must be regarded as
the consequence of a gradual accumulation of results in
one unbroken line of action. We must beware, too, of
claiming that we understand the present forces and their
action in all points so thoroughly that we can judge the
past by them completely, or even that processes which
would now strike us as anomalous may not come hereafter
to appear regular ; but we are authorized to refuse
to admit them until a clear case shall be made out in
their favor ; they are never to be granted as postulates.

Now we have seen above, in the chapters devoted to
196detailed examination of the changes of language, that
the general effort of language-making is toward the provision
of expression, for the needs of communication
and the uses of thought, by such means as lie most availably
at hand ; that a prominent part of the movement
is the reduction of coarser and more physical, material,
sensible designations to finer and more formal uses, both
by constant shifts of meaning, by the attenuation of
words once of full material meaning to the value of
form-words, and by the conversion of words formerly
independent into formative elements, suffixes and prefixes,
signs of modified meaning or of relation attached
to and forming part of other words. In the earliest
traceable condition of our language, the use of formative
elements was the prevailing means of denoting
relations, so much so as to constitute the distinctive
characteristic of the common Indo-European language
and to explain this feature is to explain Indo-European
growth.

It was in the simple practice of composition that we
found (p. 120 seq.) the germ of synthetic form-making ;
and we noticed a number of real forms as made by this
means, with the help of only those tendencies which are
universally prevalent in human speech. The adverbial
endings ly and (French) ment, the tense-signs d and
(French) ai, the derivative suffixes less and dom, and so
on, are, in all respects, precisely as true and as good formative
elements as anything in Indo-European speech ;
it is only the historical student, not the speaker, who
knows them as different from the s of loves and the th
of truth, which go back for their origin to a period
greatly remote in comparison. And all form-making
of which we know anything in the historical period is
of this same kind, by external accretion ; all the cases
197of an apparently different character (we exemplified
them by man and men, rēad and rĕad, sing and sang)
being demonstrably inorganic, accidental, results of the
putting to use of a difference of secondary value,
wrought out by phonetic change from forms originally
made by concretion.

This being so, we are required by the principles of
inductive investigation to endeavor to make this sole
recognizable method of formation found active in historical
times explain the growth of Indo-European language
in the ancient times. If it is sufficient, we are
not only not called upon, but actually forbidden, to
bring in any other method to aid ; or, at any rate, nothing
but the most direct and cogent evidence can have
the right to compel our admission of any other. And
such evidence is by no means to be found in our simple
inability to trace any given element or elements, or even
a great many such, to the independent words out of
which they grew, and to describe the series of changes
of form and meaning which converted the one into the
other. The linguistic record is too hopelessly fragmentary
for that. As every period in the changeful
life of the earth, denudes or covers up or dislocates a
part of the record of geological succession, so the
changes of every age contribute to break the continuity
of linguistic succession, in every part — in the transiers
of meaning, in the formation of words, in the
making of means of derivation. While there is so
much in the peculiar and recent formations of even the
Germanic and Romanic languages that baffles the inquirer
and seems to defy explanation, it would be most
unreasonable to expect that words and forms of vastly
more ancient growth will be completely and in all parts
amenable to analysis. If we can find any trustworthy
198evidences of the operation of the method of combination
in the earliest synthetic forms, we have the right
to assume it, in default of proof to the contrary, to
have been the sole operative principle, then as well as
later.

And it is claimed by the leading school of comparative
philology that the principle in question is actually
sufficient to account for the whole structure of Indo-European
language ; that the latter presents no forms
which demand the admission of any other genesis than
by addition of element to element ; that wherever, by
our analytical processes, we succeed in detaching from a
word a subordinate part, indicating some modification
or relation of a radical idea, there we are to recognize
the trace of a formerly independent word, which has
lost its independence and become an affix, by the same
processes which have made love-did into loved, true-like
into truly, habere habeo into aurai, verâ mente into vraiment,
and so on.

But in this doctrine is involved another very important
one : that, namely, of a primitive body of monosyllabic
roots as the historical beginnings of Indo-European
speech-development. Its necessity as a corollary
from the former is clear enough : if all formative
elements come by accretion and integration, then only
that can have been original which is left when these
have been stripped off, to the very last one : and what
is left is the root ; and it is, in our family of language,
a monosyllable. This is the doctrine actually held by
most students of language ; the dissidents are few, and
have nothing to say, in defense of their unbelief, except
what is easily refuted as misapprehension or want
of logical consistency. Though at first sight repellent
to some, it involves nothing that has a right to trouble
199the scientific inquirer, any more than the acceptance of
a primitive state of rudeness with reference to the arts
of life or the condition of knowledge. And as there
are races now living on the earth which have never
gained command of more than the simplest tools, modes
of dress and shelter, and the like, so (as we shall see
more particularly in the twelfth chapter) there are those
which have never developed their language out of this
radical stage. If we see in later times conjugational
and declensional inflections formed and brought into
use, there can be no invincible obstacle in the way of
our reasoning back to a time when such things did not
exist ; if we see parts of speech like prepositions, conjunctions,
and articles coming into being, we may regard
as possible a period when the first distinction of parts of
speech was made. Whether such possibilities were ever
realities, is a matter to be determined by sufficient scientific
evidence.

It is to be noticed that this doctrine does not commit
us to the recognition of any actually traceable list of
roots as being the beginnings of development in our
family. If it shall be shown hereafter — as it is already
shown, or at least made probable, with regard to some —
that any of the elements now generally regarded as
roots are of composite structure, containing a formative
element fused with a root (as in our count, cost, preach,
etc., noticed above, p. 55), this will only push the name
and quality of roots one step further back. The firm
foundation of the theory of roots lies in its logical necessity
as an inference from the doctrine of the historical
growth of grammatical apparatus. It is to be noticed
further that the question of roots as the historical
beginnings of language is quite distinct from that
of the origin of language, which we do not take up until
200later (fourteenth chapter) : the one is exclusively linguistic,
the other partly anthropological.

The Indo-European roots, then, are the elements of
speech which existed prior to the whole development of
the means of grammatical distinction, before the growth
of inflection, before the separation of the parts of
speech. They indicated each some conception in entire
indefiniteness as concerns its relations, neither viewed
as the concrete name of an object, nor as attribute only,
nor as predicate ; but as equally ready to turn to the
purpose of any of the three. This is a state of things
which we, with our habits of speech and thought, find
it very hard to realize, but which is brought comparatively
within reach of our apprehension by making
acquaintance with existing tongues of a low grade of
development. The roots, however, are not all of one
homogeneous class ; there is a little body of so-called
pronominal or demonstrative roots which are distinguished
from the rest as signifying position or direction
with reference to the speaker, rather than any more
concrete quality. They are very few, and of the simplest
phonetic form : a vowel only, or a consonant with
following vowel. That they are ultimately distinct
from the roots of the other class, and were not rather
developed out of these by attenuation of meaning, as
form-words in the later stages of language-history, many
students of language are very loath to believe, and not
without reason ; but the distinction is one which must,
it seems, at any rate be admitted as antecedent to the
whole growth of Indo-European forms ; nor have the
attempts to identify the one class with the other been
as yet at all successful. The point is one of which, the
complete solution will probably be possible only when
the languages of lower order shall have come to be more
201widely and deeply understood ; perhaps the early development
of such a class of form-words was the first sign
of that linguistic aptitude which has always distinguished
this family, and prepared the way for its afterevolution.
The other class, commonly called verbal or
predicative roots, were significant in general of such
acts and qualities as are apprehensible by the senses,
and were much more numerous, counting by hundreds :
examples are stâ (Greek ἵστημι, Lat. stare, our ‘stand’
etc.), dâ. ‘give’ (δίδωμι, dare), par, ‘pass’ (πέραω, ex-per-ior, fahren, fare,
etc.), wid, ‘see’ (οἶδα, video, weiss,
wot, etc.), and so on.

An early (perhaps the first) and most important act
in the history of linguistic development out of these
rather scanty beginnings was that whereby a separation
was made between noun (substantive and adjective) and
verb. The essence of a verb is that it predicates or
asserts ; and the establishment of a distinct form by
which predication shall be signified has by no means
been reached in all languages. There are many tongues
which do not formally distinguish giving (adjective or
substantive) and gift from gives : they put the subject
and predicate side by side, as ‘he giver,’ ‘he good,’ and
leave the mind to supply the lacking copula. The making
of a verb is nothing more than the establishment
of certain combinations of elements in an exclusively
predicative use, the supplying of a copula in connection
with them and not with others. This was accomplished
by adding certain pronominal elements to the verbal
element : dâ-mi, dâ-si, dâ-ti ; the former having already
gained at least a quasi-personal significance, as
designating that which is nearer or remoter. Precisely
how we shall explain dâ-mi, for instance — whether as
meaning more ‘give I,’ or ‘giving (adj.) I,’ or ‘giving
202(subst.) mine,’ or ‘giving here’ — seems a matter not
worth contending about ; since, at the period in question,
noun and adjective and verb were equally present
in the one element, and pronoun and adverb in the
other ; and there was as yet no distinction of ‘I’ and
‘mine.’ The combinations adduced above gave three
verbal persons ; they were made exclusively singular in
number by the addition of a plural and a dual, usually
explained (through many difficulties of detail) as formed
by a composition of pronominal elements in the ending :
masi, for example, being ma-si ‘I [and] you,’ i. e.
‘we.’ The forms thus made contained no implication
of time, were not properly a “tense ;” but a past was
by-and-by made by prefixing an adverbial element, the
“augment” of the Greek, pointing to a ‘then’ as adjunct
of the action : a-dâ-mi, ‘then give I,’ i. e. ‘I
gave ;’ and the form, by reason of the accented addition
at the beginning, was shortened at the end, to ádâm
(Skt. ádâm, Gr. ἔδων) — whence the distinction between
secondary and primary endings, conspicuous in some of
the languages of the family. But yet another tense, of
completed action, was made by reduplication or repetition
of the root : dâ-dâ-mi, ‘give-give I,’ i. e. ‘I have
given’ (the reduplication being then variously abbreviated) ;
and this in Latin and Germanic has become the
general preterit, the augment-tense having been lost ;
our sang, held, etc., are its descendants. As handed
down to us, however, few of the “present” tenses of
Indo-European verbs are of the simple formation above
illustrated ; more usually, the root appears in some way
extended, either by another reduplication (Skt. dâdâmi,
Gr. δίδωμι), or by the addition of sundry formative
elements (Lat. cer-no, cre-sco, Gr. δάμ-νη-μι, δείκ-νυ-μι,
etc., etc.) : all of them supposed to have been at first
203means employed for denoting the continuousness of an
action, like our am giving, though they later lost their
restriction to this sense. In some verbs, along with the
new present and its continuous preterit or proper “imperfect,”
the preterit and moods of the simpler root
were retained in use, with a more undefined past meaning,
becoming the Greek (and Sanskrit) “second aorist”
(as ἔδων, ádâm, beside imperfect ἐδίδων, ádadâm). For
other verbs, an accordant tense was made apparently by
composition with a second root as, ‘be,’ making what is
called in Greek the “first aorist.” Besides these, a
future, also supposed to contain the same auxiliary, was
made before the separation of the branches, and is best
retained in Greek and Sanskrit ; the full form of its
suffix is syâmi : Sanskrit dâ-syâmi, Greek δώσω (older
δωσιο), ‘I will give.’ There were some imperative
persons, with no special mood-sign, but with peculiar
endings. Of other moods, there were a subjunctive
and an optative, marked by insertions between root and
ending, of somewhat doubtful character. Then, finally,
there was a reflexive or “middle” voice for all these
various forms, with its characteristic in the personal
endings themselves : an extension of them, prevailingly
explained as a repetition, once with subjective value,
once with objective.

This appears to have been the entire fabric of the
Indo-European verb prior to the separation of the
branches. It has been variously preserved, contracted,
expanded, in the later history of the branches. The
Sanskrit has preserved most faithfully the outward
forms ; the Greek has best retained the original uses,
and has added most, so that its verb is far the richest in
the family. The Latin lost much, but added a great
variety of modern formations. The Germanic lost all
204save present and perfect, with their optative (called by
us subjunctive), and with the imperative ; apart from
the preterit with did, often already referred to, its new
additions have been made in the way of analytic combination.
To follow out further the details of the verb-history,
interesting as the task would be, would take us
too long.

The genesis of the noun as a part of speech, in its
two forms, substantive and adjective, was implied in
that of the verb : when one set of forms became distinctly
verb, the rest were left as noun. And everything
in Indo-European speech from predicative roots
is by origin either verb or noun, a form, either of conjugation
or of declension. On the other hand, the further
we go back, the less are substantive and adjective
distinguished from one another ; they are made by the
same suffixes, they share the same inflection : things, in
fact, are named from their qualities ; and whether the
quality-denoting word shall be used attributively or
appellatively is at the outset a matter of comparative
indifference ; though the two come finally to be distinct
enough. The characteristic of the noun is the case-ending,
as that of the verb is the personal ending ; case
and number are to the noun what person and number
are to the verb, fitting it to enter into definite relations
in the sentence. The Indo-European cases are seven,
besides the vocative, which is not a case in the same
sense with the rest, since it stands in no syntactical
relation with anything else. The accusative is the to-case,
marking that toward which the action of the verb
is immediately directed, and hence becoming also the
case of the direct object ; the ablative is the from-case ;
the locative, the at- or in-case ; the instrumental, that of
adjacency or accompaniment, then of instrument or
205means — the by-case, in both senses of by. Then the
dative is the for-case, and the genitive the of-case, that
of general relation or concernment. The nominative,
finally, is the case of the subject, and its ending, so far
as at present appears, more formal than that of any
of the others ; the vocative is most often accordant
with it, and has, at any rate, no inflectional sign of its
own.

The subject of the genesis of the case-endings is
much more obscure than the history of the verb. The
genitive suffixes show most signs of kindred with the
ordinary suffixes of derivation. Pronominal elements
seem clearly visible among some of the rest, but every
point is too doubtful to allow of summary presentment ;
and for more than this there would be no room here.
How the distinctions of number are combined with
those of case is by no means plain ; the endings of singular,
dual, and plural have the air of being independent
of one another, nor are there demonstrable indicators
of number, such as in languages of lower type are
often found inserted between theme and ending. Yet
the earliest language is mainly free from that diversity
of modes of inflection according to which, in the middle
period, words are arranged in different “declensions.”
First, uniformity, at least approximate, of declension in
all words ; then correspondence in the declension of
themes having the same final ; then, the characteristic
finals being lost, a confusion of declensions — such has
been the general history of development.

One more matter of distinction, that of gender, is
so mixed up with those of case and number as not to be
completely separable from them. The problem of the
treatment of this element in Indo-European language is
still very far from being completely solved. Its foundation
206must, of course, lie in the distinction of sex in
those creatures which have conspicuous sex ; but such
constitute only an exceedingly small part of the creation ;
and the distinctions of gender involve everything
that exists, and in a manner which is only in the smallest
part accordant with natural sex. The world of untraceably
sexual or of unsexual objects is not, as with
us, relegated to the indifferent “neuter ;” great classes
of names are masculine or feminine partly by poetical
analogy, by an imaginary estimate of their distinctive
qualities as like those of the one or the other sex in the
higher animals, especially man ; partly by grammatical
analogy, by resemblance in formation to words of gender
already established. At any rate, in the common
Indo-European period, all or nearly all attributive words
were inflected in three somewhat varying modes, to indicate
generic distinctions ; and the names of things
followed one or other of these modes, and were masculine
or feminine or neuter. The distinction was
partly in the case-ending, partly in the derivative theme
or base, though there was hardly a suffix, derivative
or inflectional, that was rigidly of one gender only ; it
was most marked as characterizing the feminine ; masculine
and neuter were hardly separated except in the
nominative and accusative cases.

The noun-inflection was shared also by the pronouns,
in all the three varieties of case, number, and gender.
In those demonstrative words, however, which acquired
a specific personal character, as denoting the speaker
and the spoken-to, gender was undistinguished. And
the words of pronominal origin exhibit certain irregularities
of inflection, different from those of the general
mass of nouns.

Although a case-ending of itself makes a noun, and
207there are many primitive Indo-European nouns which,
are made by such alone, the great mass of them have
other elements interposed between root and ending,
which we call suffixes of derivation ; and these even
come, in time, to be divided into two well-marked
classes : primary, or such as are appended directly to
verbal roots ; and secondary, or such as are added only
after other derivative endings. Of these, likewise, too
few among the most ancient ones are recognizable in
their independent character, and traceable through their
changes of application, to allow of our illustrating here
the method of their growth. But though the subject
is full of obscurity in its details, there is no mystery in
the principles involved : the processes which have
formed modern suffixes are fully capable of having produced
also the ancient ones.

As the two sides of meaning and application in the
predicative or verbal roots are verb and noun, so in the
demonstrative (which do not make verbs) the two sides
may be said to be pronoun and adverb. From the latter
class come those earliest words of place and direction,
readily convertible also into words of time, which are
of adverbial quality. Yet even these are claimed by
some to be properly case-forms of pronouns ; and the
rule is laid down that everything in language is by origin
an inflected form either of verb or of noun. At any
rate, the class of adverbs, when once brought into
existence, receives abundant accessions of this kind,
through its whole history, down to the latest, from
which we have already drawn examples (pp. 41, 122).
Prepositions, in our sense of the term, are of yet more
recent origin, created a separate part of speech by the
swinging away of certain adverbs from apprehended
relation to the verb, and their connection in idea with
208the noun-cases which their addition to the verb had
caused to be construed with it. We see them coming
into distinct existence in the oldest languages of the
family, as the Sanskrit ; and their increase of number
and consequence ever since is apparent. Conjunctions,
though we nowhere find them absolutely wanting, are
of secondary origin, being among the most characteristic
products of the historical development of speech. To
be able to put clauses together into periods, with due
determination of their relation to one another, is a step
beyond the power to put words alike determmately together
into clauses.

These are the Indo-European “parts of speech :”
that is to say, the main classes of words, having restricted
application and definite connection, into which the
holophrastic (‘equivalent to a whole phrase’) utterances
of a primitive time have by degrees become divided ;
the separated parts, members, of what was once an undistinguished
whole. But there is one other class, the
interjections, which are not in the same and the proper
sense a “part of speech ;” which, are, rather, analogous
with those all-comprehending signs out of which the
rest have come by evolution. A typical interjection is
the. mere spontaneous utterance of a feeling, capable of
being paraphrased into a good set expression for what
it intimates : thus, an ah ! or an oh ! ; may mean, according
to its tone, ‘I am hurt,’ or ‘am surprised,’ or ‘am
pleased,’ and so on ; only there is no part of it which
means one of the elements of the statement while
another part means another. Yet, such creatures of
conventional habit in regard to expression have we become
by our long use of the wholly conventional apparatus
of language, that even our exclamations have
generally a conventional character, and shade off into
209exclamatory utterance of ordinary terms. A man's feelings
must be very keenly touched in order to draw out
of him a purely natural interjection, in which absolutely
no trace of the acquired habits of his community
shall be perceptible. And the interjectional employment
of common words, or of incomplete phrases, is a
very common thing in the general use of speech ; emotion
or eagerness causing the usual set framework of the
sentence, the combination of subject and predicate, to
be thrown aside, and the conspicuous or emphatic elements
to be presented alone — a real abnegation of the
historical development which, under the growing dominion
of consciousness over instinct and of reason
over passion, has wrought the sentence out of the root.

In this too brief and imperfect sketch of the history
of Indo-European speech, no attempt has been made to
define the order in which, the parts of the inflectional
development followed one another. Success is not to be
hoped for in any such attempt until the history of less
highly developed and of almost undeveloped languages
shall be far better understood than it is at present.
For, to reason these matters out on Indo-European
ground alone is at any rate impossible : the period lies
too far back, its evidences are too fragmentary and
difficult of interpretation ; we are not competent to
judge them. As to the impossibility of determining
the absolute time occupied by the history, enough, perhaps,
has been already said : that it should have taken
less than a very long time, there is no reason whatever
for believing. The whole was a series of successive
steps, of which one led to another and these to yet
others ; a growth of habits which were in themselves
capacities also ; and each step, the formation of each
habit, was a work of time, not less in the olden time
210than it would have to be in the modern period : though
whether a work of not less time, we can hardly venture
to say, since the rate of growth may fall under the government
of conditions which we cannot, as yet, fully
appreciate.

There has also been, so far as synthetic structure is
concerned, an evident climax, followed by an anti-climax,
in this history. During the immense prehistoric
period, and prior to the separation of the branches from
one another, the inflectional system of the noun, and
less distinctly that of the verb, reached a fullness which
has since undergone a gradual reduction. Not that
there has been generally a diminution of ability to express
distinctions ; but means of another kind have been
more and more resorted to : auxiliaries, form-words, instead
of suffixes, formative elements in words ; and
these later means we are accustomed to call analytic, as
distinguished from synthetic. He might have loved
and he will be loved, as contrasted with their Latin
equivalents amavisset and amabitur, may be taken as
typical examples of the two modes of expression. This
fact has been adduced as evidence against an original
radical condition of language, by some scholars, who
prefer to assume a primitive period of excessive polysyllabism.
But with evident injustice ; the argument
would be a good one only if no such thing as the making
of forms were known in language, but only their
wearing-out and loss. If we see how collocation and
combination and integration and mutilation and corruption
all work in succession on the same material in
every part of language, producing forms and destroying
them again, it is plainly within the competency of the
changing circumstances and habits of the language-making
community to give the history of development a
211climactic form. The constructive methods, once inaugurated,
are made effective up to the provision of a
sufficient apparatus for the expression of relations ; and
for a time, until this point is reached, their efficiency is
greater than that of the destructive processes, which
also have been all the time at work — then the relation
is gradually reversed, and there is more wearing-out
than replacement by synthetic means, though this latter
also never entirely ceases ; collocations remain such,
instead of going on to combination and integration ;
there is still abundant new provision, but it is of
another sort. The habit of construction has changed ;
though to a very different degree in the divided parts
of the great community. If there is a law which
governs this climactic phase of development, it has not
yet been worked out and exhibited ; nor is it likely
ever to be so, although we can trace some of the determining
influences which have contributed to bring about
the effect.

It is time now for us to leave the family which has
so long occupied us, and to review, in a much briefer
manner, the structure of the other grand divisions of
human language. But, founding upon the example of
historic growth which we have just been studying, it is
desirable first to turn our attention to some general
features of the doctrine of linguistic structure.212

Chapter XI.
Linguistic structure : material and form in language.

The distinction of material and form ; examples : number, gender, case,
etc., in nouns ; comparison and concord of adjectives ; time, mood,
and other distinctions in verbs. Form by position. Inferences.
National and individual prejudices ; comparative value of different
languages. A language represents the capacity of its makers.
Rude beginnings of all speech.

To understand, in a general way, the structure of
Indo-European speech, in its character and its uses, is
to us no difficult task ; the subject is already more or
less familiar. Though, the parts of this structure
which our own language still possesses are but fragmentary,
they are at least akin with the rest, and lead
the way to the knowledge of the whole. It is comparatively
a question only of less and more ; and many of
us know the more, as exhibited in those tongues of the
family which have retained a larger share of the original
structure, or have supplied its loss more fully. We
cannot, however, go on profitably to examine the character
of other languages without discussing a little,
by way of introduction, the principles of grammatical
structure. It will be possible to do this, sufficiently for
our purpose, in a wholly simple and unpretentious manner,
drawing illustration from phenomena with which
213almost every one is familiar, and especially out of our
own English.

The distinction of the more material and the more
formal, relational parts of expression has been noticed
and illustrated by us often already. The s of brooks,
for example, is formal in relation to brook as material ;
the added letter indicates something subordinate, a
modification of the conception of brook, the existence of
it in more than one individual : it turns a singular into
a plural. Men has the like value as regards man, the
means of making the same formal distinction having
come to be of a different kind from the other, an internal
change instead of an external. Brooks and men
are not mere material ; they are “formed” material,
signs for conceptions with one important characteristic,
number, added. But then, by simple contrast with
them, brook and man are also “formed ;” each implies,
not by a sign, but by the absence of an otherwise
necessary sign to the contrary, restriction to a single
article of the kind named. According to our habits of
speech, no one of these words, no one of our nouns in
general, can be used without a distinct recognition by
the mind of the number of things signified.

But there are many other definable qualities or circumstances
belonging to brooks and men besides number.
They are, for example, of very different sizes.
And we have a similar formal means, though only a
very limited one, of signifying this : a small brook is to
us a brooklet ; a small man, a mannikin. It is perfectly
conceivable that a language should take constant
cognizance of this element of size, distinguishing always
the large, the medium, and the small individuals of a
kind, by diminutives and magnificatives. The Italian
almost does as much as that, by a peculiarity which has
214grown up in it since it became a separate language.
But while we call a small brook a brooklet, we call a
large one a creek, or a river, or something of that sort ;
or we apply small and large to it, in all their varying
degrees : and so with giant and dwarf, and all the
limiting adjectives, as applied to man. All this classification
which is made by independent words is as truly
expression of form as is that which is made by affixes.
Another equally real quality, the differences of which
are apparent in every case that comes before the mind,
is, in many animals, age ; and we can say man, lad,
boy, child, infant, etc., as horse and colt, cow and calf,
and their like ; and the Latin senex and German greis
show the extension of the same system in the other
direction, where we have to use the method of description
by independent words.

Once more, man in its distinctive sense indicates a
male animal, and we have a different word, woman, for
a female of the same kind ; and so all through the list
of animals in which sex is a conspicuous or an important
distinction : as brother and sister, bull and cow,
ram and ewe : nor is there a language in the world
which does not do the same. Only, as we have already
seen, our own family of languages (along with two or
three others) has erected this distinction of sex into a
universal one, like number, making it a test to be applied
in the use of every word ; breaking away from
the actual limits of sex, and sexualizing, as it were, all
objects of thought, on grounds which no mortal has
yet been wise enough to discover and point out in detail.
And, though we in English have abandoned the
artificial part of the system, we retain its fundamental
distinction by our use of he, she, and it ; the test of sex
is to us a real and ever-present one. The modern Persian
215has lost from his language even that degree of generic
distinction ; and to him, as to the Turk or the Finn,
whose ancestors never acknowledged any grammatical
gender, it seems no less strange to use one pronoun for
a male being and another for a female than it would
seem to us to use one for a small, or a young, or a near,
or a white object, and another for a large, or an old, or
a remote, or a black object. And he has really reason
on his side ; it is our usage that is the exceptional one,
and needs justification. There is in the nature of
things no necessity for our choosing among the various
accidents of a conception any particular ones, to the
exclusion of the rest, as subjects of grammatical distinction —
although, of course, there may be reason
enough why one is practically better worth distinguishing
than another. There is a second, somewhat analogous
yet not identical, distinction made by us, also
solely by the use of pronouns — namely of who and
which or what — between persons and non-persons ; and
the American Indians have one between things animate
and things inanimate, with (as in the case of our gender)
abundant figurative and personifying transfer :
either of these is perhaps as valuable in itself, and as
capable of higher uses, as is the Indo-European distinction
of the three genders.

We will notice only one more item in connection
with the noun, its cases. Our language has preserved
to most of its nouns their old genitive case, though not
without restriction of the limits of its former uses.
And in the pronouns we distinguish the object from
the subject or nominative case : he him, they them, etc.
By this difference, the distinction of subject and object
relation is kept so clearly before us that we transfer it
in apprehension to the whole class of nouns, and reckon
216them also as possessing objective cases, though there is
really none such in the language. We do not recognize
a dative, though we have some really dative constructions — as
in “I give him, the book” — because there is
not in use even one dative of different form from the
accusative. Just so, the Latin and Greek reckon accusatives
neuter, though these are not in a single instance
different from the nominatives, because the two
cases are usually unlike in other words ; so the Latin
reckons an ablative plural different from the dative,
because there is in a part of its words an ablative singular
different from the dative. This transfer of a
formal distinction only partially made to the words in
which it is not made at all is an important feature in
the history of forms. Our two or three cases seem to
compare but ill with the Sanskrit seven ; yet these
compare as ill, in one sense, with the Scythian fifteen
or twenty : and, on the one hand, we are able, by the
help of another instrumentality, to express all that is
expressed by either Sanskrit or Scythian ; while, on the
other hand, we imply a great deal more than we or
they distinctly express ; if we were to use different
signs for all the shades of case-relation which we can
recognize by analysis in our speech, we should have to
multiply our list of prepositions many times.

For a part of our adjectives of quality, we have
forms (strictly, derivative rather than inflectional) denoting
two “degrees” of increment : high, higher,
highest ; they seem to have been at the beginning
rather intensive than strictly comparative. But, as
means of comparison, they cover only a small part
of the conceivable ground, and cover it only rudely.
The possible degrees of a quality are indefinitely numerous,
and there are descending as well as ascending
217grades, which have in theory an equal right to notice :
many of them we clearly mark by our analytic substitutes
for the old derivatives ; and we frame such kindred
means of expression as are exemplified by reddish
and bluish, German röthlich and bläulich (‘redlike,’
etc. : resembling the quality, but not quite it), French
rougeâtre and bleuâtre. Most of the later tongues of
our family still retain that adaptedness of the qualifying
adjective, in gender and number and case, to the
noun qualified, which, inherited from the time when
adjective and substantive were not separated, was charasteristic
of their ancestors ; to this we preserve nothing
whatever that is correspondent ; that an adjective
should change its form on account of the character of
the noun it belongs to is as strange to us as to many
languages it is that the verb should change its form on
account of the character of the subject of which it
predicates something.

In fact, we have almost reduced to a nullity also the
concord of the verb and its subject. How there came
to be such, we have seen in the foregoing chapter : the
endings were the actual subject-pronouns themselves ;
and the distinction of person and number in the verb
was the necessary concomitant and result of that in the
pronouns and nouns. Nor is it yet quite a nullity :
while we say I love, but thou lovest and he loves, and
while they love stands over against he loves, so long shall
we continue, by an apprehended extension of these
clearly-felt distinctions, to reckon three persons and two
numbers in all our verbal inflection. But our triple
distinction of persons is far from exhausting the possibilities
of personal relation ; many tongues have a
double first person plural, one inclusive and one exclusive
of the person or persons addressed : one we which
218means ‘I and my party’ as opposed to you ; and one
that means ‘my party and yours,’ as opposed to all
third persons. Others, again, distinguish genders in
verbal inflection : ‘he loves’ has one ending, ‘she loves’
has another. We have seen that some older languages
of our family have a dual number ; and it would be
quite as proper in theory, only not so manageable in
practice, to have a whole decimal system of numbers,
just as of numerals.

But the attendant circumstances which present themselves
for inclusion in verbal expression, and in one or
another language find expression, are simply numberless ;
and the richest verbal scheme that was ever put
together takes account of only a part of them, even
when supplemented by the resources of analytic phraseology.
To us, the element of time is the conspicuous
and pressing one ; the denoting of an action appears
almost to require an implication of tense-relation. Yet
many languages do not regard this element as calling
for inclusion in the fundamental structure of the verb
rather than others ; and they leave it to be inferred
from the connection, or intimated by external means,
particles, auxiliaries, as we on our part treat other elements
which they weave into the verbal structure. To
any given act of speaking, for example, there cleaves
some qualification of time ; but so also of place, of
manner, of purpose. Equally modifications of the indefinite
act of speaking are speaking repeatedly or
habitually, rapidly, with violence, under compulsion,
for another, or causing, ceasing, appearing to speak,
declaring another to speak, speaking to one's self — and
so on, indefinitely : and these, or many of them, are
actually incorporated in derivative verbal forms by
races who treat the tense-element less elaborately than
219we. And our tense-making is on the smallest scale, as
compared with the infinite possibilities of tense-distinction.
We hare not even, as some languages have,
a nearer and remoter past, a nearer and remoter future.
That a thing was done long ago is as true a temporal
relation as that it happened in past time at all ; but we
intimate only the latter by an inflection, and the former
by relational words ; and therefore, to our way of thinking,
he who wants the inflection has too little, and he
who converts the other into an inflection has too much.
Our triple forms for each tense — I love, I do love, I am
loving — by their incessant use, and the necessity constantly
imposed on us of choosing among them, keep
before our minds certain distinctions which are comparatively
unnoticed in French or German ; yet they
are in the French and German minds also, and if any
of them rises to prominent importance, those languages
have sufficient means of intimating them. It is good
English or German to say “I picked up the book that
lay there ;” but to the Frenchman it would be a gross
blunder to use the same tense for the instantaneous act
of picking up and the continuous condition of lying ;
the difference is clearly involved in our thought as well
as his ; only our language does not compel our attention
to it. The case is quite the same with our moods,
those means of defining the contemplated relation between
subject and predicate, or modifications of the
copula. There are infinite shades of doubt and contingency,
of hope and fear, of supplication and exaction,
in our mental acts and cognitions, which all the synthetic
resources of Greek moods, with added particles
and adverbs, which all the analytic phraseology of
English, are but rude and coarse means of signifying.
And an Algonkin verb makes a host of distinctions
220which are so strange to us that we can hardly learn to
appreciate them when defined.

There is one other mode of formal distinction which
demands a moment's notice from us : namely, position.
In “you love your enemies, but your enemies hate
you,” the distinction of subject and object is dependent
solely on position, and is given by that means with all
necessary clearness. In a language of which the inflections
are so much worn out as are ours, this method
counts for much ; and there are tongues in which it is
of even superior importance. Those, on the other
hand, which have a greater abundance of inflections
possess a freedom of arrangement which, to us is surprising,
and almost puzzling.

The principal conclusions intended to be suggested
by this brief exposition, and to be made of use in comparing
the structure of various languages, are, it is believed,
sufficiently clear. In the first place, the realm
of formal relation is infinite, unexhausted by the formal
resources of even the richest language, or of all languages :
however much, may be expressed, there is
vastly more of the same kind left unexpressed, to be
inferred by the intelligent mind from the perceived
conditions of the particular case, or passed over as unessential
to the ordinary purposes of communication — which
is, at the best, only a rude and fragmentary
means of putting one mind, or heart, into communion
with another. There are no relations to which a language
must necessarily give expression ; there are only
certain ones which are more naturally suggested, of
which the expression is more practically valuable, than
others : and what these are, we can learn only from the
general study of languages ; our own educated preferences
are no trustworthy guide to them. In the second
221place, there is no absolute dividing-line between what
is material and what is formal in a language ; material
and form are relative words only, names for degrees,
for poles of a continuous series, of which the members
shade into one another. And, as we saw in the fifth
chapter, the grandest internal movement in a growing
and improving language is that from more material to
more formal uses, whereby both words and phrases take
on a less gross and physical meaning, even to the extent
of being attenuated into form-words, or, in combination
with other elements, into formative elements
— both alike indicators of relation. Hence, in the
third place, the means of formal expression are of the
utmost variety ; they are not to be sought in one department
of a language only, but in all ; they are scattered
through the whole vocabulary, as well as concentrated
in the grammatical apparatus. Deficiency in one
department may be compensated, or more than compensated,
by provision of resources in another. There
is no human tongue which is destitute of the expression
of form. ; and to call certain languages, and them alone,
“form-languages,” is indefensible, except as the term
may be meant to describe them as possessing in a higher
or exceptional degree a quality which they really share
with all the rest.

In judging other languages, then, we have to try to
rid ourselves of the prejudices generated by our own
acquired habits of expression, and to be prepared to
find other peoples making a very different selection
from our own of those qualifications and relations of
the more material substance of expression which they
shall distinctly represent in speech, and also sharing
these out very differently among the different modes
of formal expression. It is a common error of uncultivated,
222and of narrowly though highly cultivated peoples,
to regard themselves alone as speakers, and all
others as babblers, “barbarians,” unintelligent because
to them unintelligible talkers. We are in no danger
of doing that ; but we are in danger still of over-estimating
the peculiar traits of our speech, and depreciating
those of others' speech. Nothing is harder than to
be perfectly impartial here ; to judge the comparative
merit of one's own and of another language requires a
grasp of all the particulars involved, a power of analysis
and comparison, and a freedom from both national
and individual prejudice, of which only exceptionally
endowed and exceptionally trained minds will be capable.
Even great scholars are liable here to great errors.
There are eminent English-speaking philologists who
regard English analysis as the only reasonable or “logical”
mode of expression, and look down on Greek
synthesis as something characteristic of a rude and undeveloped
intellectual condition ; there are many more,
doubtless, of various nationality, who undervalue the
resources of English, and are loath to assign a high
rank to a tongue which has lost or thrown away so
much of its inherited structure.

On the whole, perhaps the best and most trustworthy
test of the value of a language is, what its
speakers have made it do. Language is but the instrument
for the expression of thought. If a people has
looked at the world without and within us with a penetrating
and discerning eye, has observed successfully the
resemblances and differences of things, has distinguished
well and combined well and reasoned well, its language,
of however apparently imperfect structure, in the technical
sense of that term, enjoys all the advantage which
comes from such use ; it is the fitting instrument of an
223enlightened mind. There is nothing in the grammatical
form of either Greek or English that may not be degraded
to serve only base uses.

In another sense also a language is what its speakers
make it : its structure, of whatever character, represents
their collective capacity in that particular direction of
effort. It is, not less than every other part of their civilization,
the work of the race ; every generation, every
individual, has borne a part in shaping it. Whether,
however, the language-making capacity can be correlated
with any other, so that we may say, a highly-organized
speech could not be expected from a historical
community whose work in this or that other respect
shows a deficiency of excellence, is extremely doubtful ;
thus far, at any rate, nothing of value has been done in
that direction. The Chinese is, as we shall see in the
next chapter, a most striking example of how a community
of a very high grade of general ability may exhibit
an extreme inaptitude for fertile linguistic development.
We may suitably compare this with the grades of aptitude
shown by various races for plastic or pictorial or
musical art, which by no means measure their capacity
for other intellectual or spiritual products. Nο uncultured
people ever spends consciously any time or effort
upon its speech ; this cannot be thought over and worked
up into better shape ; it must come by the way, as incident
to the work of thought, as result of unreflective
effort at communication. That race which possesses
most of the right kind of regulative force will turn out
a product that is admirable ; and the contrary.

Only, also, the possibility of a radical change of history,
a new turn of development, is different at different
periods of growth. After a certain stage of advance in
definite and established expression is reached, the conservative
224forces, depending on acquired habits of speech,
are too strong to be overcome, and the language goes on
forever on the course which the directing hands of the
earlier generations have determined. This is a point
upon which we have no right yet to speak with definiteness ;
we may hope some day to understand it better :
to be able, for example, to lay down exactly what conditions
the stagnation of Chinese speech. There are other
departments of civilization in which a race does not
always show itself able to develop unaided its own best
capacities. The Celtic and Germanic tribes, which have
proved themselves equal to taking leading places in the
world's history, might have remained comparative barbarians
to the present time, if they had not received
Greek civilization, as shaped over and reorganized by
Rome. But though a nation may borrow culture from
its neighbors, it does not in the same way borrow linguistic
development ; no race ever adopted a new mode
of structural growth for its native speech by imitation
of another ; though many a community has, under sufficient
external inducement, exchanged its native speech
for another ; and borrowing, as we have already seen,
especially accompanies transfer of culture, and is capable
of going on to such an extent as vastly to enrich the
borrowing speech, and fit it for higher uses.

While a people's capacities and acquirements make
its language, we must not fail to notice also the contrary
truth, that its language helps to determine its intellectual
character and progress. The powerful reflex
influence of language on mental action is a universally
admitted fact in linguistics ; to allow it is only to allow
that rooted habits, learned by each generation from its
predecessor, have a controlling influence on action —
which is axiomatic. But the subject belongs to a much
225more advanced and elaborate discussion of language
than this work makes any pretense of being ; and it has
never yet been worked out fruitfully.

On the analogy of Indo-European speech alone we
have a right to assume, at least provisionally, that whatever
of inflective structure may be possessed also by
other languages, whatever of formal and formative
apparatus they may contain, of any kind, has been
wrought out by somewhat similar methods, from a
similar initial stage of rude and gross material. If
there shall be found languages in which this is demonstrably
not the case, we can modify or abandon the
assumption hereafter ; but it will require very definite
and cogent evidence to make such, demonstration. For
language is an instrumentality ; and the law of simplicity
of beginnings applies to it not less naturally and
necessarily than to other instrumentalities. Some seem
to imagine that to regard men as having begun to talk
with formless roots, which we now arrive at “by abstraction”
from the material of living languages, is like
regarding them as having begun the use of physical
instruments with the bare abstract motive powers — the
inclined plane, the wheel, the pulley. But such a parallel
is as absolutely erroneous as anything can be : the
analogues of the motive powers, rather, would be the
attributive and predicative relations, the assertive, interrogative,
and imperative modes, and their like. The
analogue of the root is the stick or the stone which
was indubitably man's first instrument : a crude tool or
weapon, used for a variety of purposes to which we
now adapt a corresponding variety of much more intricate
and shapely tools. And to hold that formed words,
divisible into radical and formative elements, were first
in the uses of speech, is just as defensible as to hold
226that men began to labor with hammers and saws and
planes and nails, and to fight with iron-headed lances
and bows and catapults. In each single root was present
at the outset — as may be present in a single interjectional
monosyllable now — a whole assertion, or inquiry,
or command, to which the tone and accompanying
gesture, or the mere circumstances of its utterance, furnished
the sufficient interpretation : just as in the stick
or stone was present — and may, on an emergency, be
made present still — a variety of instruments or weapons.

Again, to maintain, for the purpose of explaining
the variety of later languages, that the expressions of
the earliest men must have been potentially different in
the different races, as the seeds or germs which develop
into different animals or plants are different ; that a formative
principle must have been present in the material
of one language and not of another ; that in the elements
which came afterward to be put to formative
uses there was from the beginning a form-making function
inherent, and so on — all this is sheer mythology.
One might as well claim that in the stick or stone, as
used by some races, there was lying perdu a well-membered
instrument or machine, which somehow developed
out of it in the hands of its users, and that in the wood
and metal of certain regions were inherent machine-making
functions, not possessed elsewhere. Language
comes to be just what its users make it ; its offices correspond
to their capacities ; if there is a higher degree
of formative structure in one language than in another,
the reason lies in the difference of quality of the two
races, their different capacity of education and growth ;
not at all in the character of the beginnings from which
both alike started, nor of the materials which both alike
have ever since had at command.227

Chapter XII.
Other families of language : their locality, age,
and structure.

Classification by families. Scythian or Ural-Altaic or Turanian family ;
doubtful members of it. Monosyllabic family : Chinese, Farther
Indian, etc. Japanese. Malay-Polynesian ; other insular families :
Papuan, Australian. Dravidian. Caucasian languages. Semitic
family ; question of its relationships. Hamitic : Egyptian, etc.
South African or Bantu. Middle African languages. Basque.
American Indian languages.

We have called a certain body of languages a family,
the Indo-European. The name “family,” we saw,
was applied to it by strict analogy with the use of the
same term elsewhere : the languages in question had
been found, on competent examination, to show good
evidence of descent from a common ancestor. We had,
however, to confess that the limits, even of this best-known
of families, cannot be traced with absolute precision ;
one or another tongue, not now thought of, or
else doubtfully regarded, as Indo-European, may one
day make good its title to a place with the rest. We
have also seen that, by the operation of completely comprehensible
causes, no language on earth exists in a state
of absolute accordance through the whole community
that speaks it ; it is a group, even if a very limited one,
of related dialects. This being the case, it is the first
228task of the comparative study of languages to divide all
human speech into families, by recognizable signs of
relationship : only thus can there be made any such
examination of their character and history as shall lead
the way to the other results which the science seeks to
attain. And such a classification has in fact been made.
It is, of course, in parts only a tentative and provisional
arrangement, held liable to rectification, both by addition
and by the giving up of what is now held even
with a fair degree of confidence : for it not seldom happens
that lines which in a half-light appear definite and
fixed dissolve away when full illumination is turned
upon them. The cautious philologist combines only so
far as trustworthy evidences take him, leaving the rest
to be settled when more knowledge is won.

As a matter of fact, moreover, linguistic scholars
have hitherto been able to put together into families
only those languages which have a common structure.
That is to say, only tongues which have snared at least
a part of their growth out of the original radical stage
(provided they have left it) have yet been found to
exhibit reliable evidence of relationship. Nο one, it is
evident, has a right to declare à priori that there cannot
remain even from the initial stage sufficient signs of
common descent, in branches whose whole structural
development has been separate : in fact, philologists are
feeling about among the roots of certain families for
such signs, and may one day succeed in bringing them
to light ; but thus far no definite results have been
reached. We shall have occasion to note in the next
chapter the difficulties which environ the inquiry, and
to point out the reasons why, on a large scale, it is
likely to fail of success.

The first family, then, which we take up is that of
229which the leading branches occupy more or less of
European soil, alongside those of our own kindred.
Of these branches there are three. The first, the
Finno-Hungarian, or Ugrian, is chiefly European : it
includes the Finnish, with the nearly related Esthonian
and Livonian, and the remoter Lappish in the Scandinavian
peninsula ; the Hungarian, an isolated dialect in
the south, wholly environed by Indo-European tongues,
but of which the intrusion into its present place, by immigration
from near the southern Ural, has taken place
within the historic period ; the dialects from which the
Hungarian separated itself, the Ostiak and Wogul, in
and beyond the Ural ; and the tongues of other related
tribes in eastern Russia, as the Ziryanians, Wotiaks,
Mordwins, etc. The Finns and Hungarians are the
only cultivated peoples of the branch : there are fragments
of Hungarian language from the end of the
twelfth century, but the literature begins only four
centuries later, and scantily, the people formerly using
the Latin much more than their own speech for literary
purposes ; the earliest Finnish records are of the sixteenth
century ; the language has a mythic poem, the
Kalevala, written down in this century from the mouths
of popular singers, of especial originality and interest.

The second branch, quite nearly related with this
one, is the Samoyed, belonging to a Hyperborean race,
which stretches from the North Sea to beyond the
Yenisei, and up the course of this river into the central
mountains of the continent, the Altai range, probably
the starting-point of its migrations. It has no culture,
nor importance of any kind.

The third branch, the Turkish or Tartar (more
properly Tatar), only touches and overlaps the European
frontier at the south. The race to which it belongs,
230after having been long the restless foe of the
Iranians on their northeastern frontier, finally, after the
Mohammedanizing of Persia, forced its way through,
worked on westward, captured Constantinople in the
fifteenth century, and was arrested there only by the
combined and long-continued efforts of the powers of
central Europe. It is stretched out at present from
European Turkey (in which it nowhere forms the mass
of the population) over a great part of central Asia,
and even, in its Yakut branch, to the mouth of the distant
Lena. The Yakuts, Bashkirs, and Kirghiz, the
Uigurs, Usbeks, and Turkomans, and the Osmanlis of
Asiatic and European Turkey, are some of the principal
divisions of the race. The Uigurs, getting their
alphabet and culture from Nestorian missionaries, were
the first to produce a scanty literature, as far back as
the eighth to the tenth centuries ; the southeastern peoples
have records (“Jagataic”) of the fourteenth to the
sixteenth ; the abundant and varied but little original
literature of the Osmanlis dates from the time of their
European conquests ; it is full of Persian and Arabic
materials.

Respecting the family relationship of these three
branches there is no question. As to the common name
by which they shall be called, usage is very diverse.
“Turanian” is perhaps more frequent than any other,
but there are grave objections to its genesis and application,
and, till use shall pronounce more definitely in
its favor, it is hardly fit to be employed in scientific
description. “Ural-Altaic,” “Scythian,” “Tartaric”
are others, employed by various authors : the first has
its advantages, but is unwieldy, and implies rather more
knowledge as to the movements of the family than we
actually possess ; we may use here “Scythian,” provisionally,
231and disclaiming for it any marked or partisan
preference.

Scythian language is the type of what is called an
“agglutinative” structure, as distinguished from the
“inflective” Indo-European. By this is meant that the
elements of various origin which make up Scythian
words and forms are more loosely aggregated, preserve
more independence, than do the Indo-European ; there
is far less integration of the parts, with disguise and
obliteration of their separate entity. All our own formations,
as has been seen, begin with being agglutinations ;
and such words as un-tru-th-ful-ly preserve an
agglutinative character ; if all our words were like it,
there would be no marked difference between the two
families as to this fundamental item. For the Scythian
formative elements are also only in small part traceable
to the independent words out of which they have
grown ; they are, like the Indo-European affixes, mere
signs of relation and of modification of meaning. But
Scythian formations do not go on to fuse root and ending,
even to the replacing of an external by an internal
flection. As a rule, the root maintains itself unaltered
in the whole group of derivatives and inflection, and
each suffix has an unchanged form and office : whence,
on the one hand, a great regularity of formation, and,
on the other hand, a great intricacy. Thus, in Turkish,
for example, lar (or ler) forms plurals everywhere ; to
it are added the same case-endings which alone make
the singular cases ; and pronominal elements indicating
possession may be yet further interposed between the
two : so ev, ‘house,’ ev-den, ‘from a house,’ ev-üm-den,
‘from my house,’ ev-ler-üm-den, ‘from my houses.’
The case-relations indicated by these endings or suffixed
particles are numerous, in some dialects rising to twenty.
232The verb exemplifies the same peculiarity still more
strikingly : there are half a dozen modifying elements
capable of insertion, singly or in variously combined
groups, between root and endings, to express passive,
reflexive, reciprocal, causative, negative, and impossible
action ; so that from the simple root sev, for example,
we may make the intricate derivative sev-ish-dir-il-e-me-mek,
‘not to be capable of being made to love one another,’
which is then conjugated with the various forms
of the simple verb ; thus bringing the possible inflective
forms from one root up to a number which is immense
as compared with any Indo-European verb.

But the distinction of verb and noun in these languages
is much less original, fundamental, and sharply
drawn than with us. The verbally used forms are,
rather, but one step removed from nouns used predicatively,
with subjective or possessive pronominal elements
appended. The types of verbal forms are, for
example, (Turkish) dogur-um, ‘striking I,’ i. e. ‘I
strike,’ and dogd-um, ‘act of striking mine,’ i. e. ‘I
have struck ;’ and the third person is without ending :
dogdi, ‘he has struck,’ dogdi-ler, ‘they have struck,’
literally ‘striking,’ ‘strikings.’ To say this is not to
say that these languages have no real verb ; since to
make a verb it needs only that certain forms be set
apart and strictly devoted by usage to the expression of
the predicative relation ; but it does imply a decided
inferiority in the grade of clearness of this most fruitful
of formal distinctions, and may shade off into a
total absence of it. Of tenses and moods such as those
instanced above, and others made with auxiliaries, these
languages have a plenty ; and their variety of resource
in derivatives is very great ; so that all the formal apparatus
is provided which is needed for shaping by the
233light usage into a sufficient instrument of thought ; and
the most cultivated of the dialects do indeed come so
near to “inflection” that their falling short of it is
hardly more than nominal.

The Scythian adjective is as bare of inflection as the
English ; and there is an utter absence of gender as
one of the categories of noun-inflection or of pronominal
distinction, just as in Persian. Relatives and conjunctions
are also nearly unknown, the combinations
of dependent clauses being, as is natural in languages
where the verb is a less definite part of speech, rather
by case-forms of verbal nouns. These constructions
make upon us the impression of great intricacy, and invert
that order of the members of the sentence to which
we are accustomed.

In the phonetic structure of these languages, the
most striking trait is the so-called “harmonic sequence
of vowels.” There are, namely, two classes of vowels,
light and heavy, or palatal (e, i, ü, ö) and other (a, o,
u) ; and it is the general law that the vowels of the
various endings shall be of the class of that in the root,
or in its last syllable — thus marking the appurtenance
and dependency of the endings in their relation to the
root in a manner which, though undoubtedly at first
euphonic only (like the Germanic umlaut), has lent itself
usefully to the purposes of formal distinction.
Every suffix, then, has two forms, a light and a heavy :
we have al-mak, but sev-mek ; ev-ler, but agha-lar, and
so on. In some dialects this assimilative process is of a
wonderful degree of intricacy.

There is field and scope in these languages for a
comparative grammar of the highest interest and importance ;
but no one has yet taken up the work seriously
and comprehensively ; the science of language has advanced
234far enough to demand its execution, which, it is
to be hoped, will not be long deferred. One obstacle
in its way, the lack of really ancient records, from a
time comparable to that of the early Indo-European
documents, is likely to be removed, if recent claims
shall prove well-founded. There is, namely, in the
Mesopotamian and Persian records, a third language,
the so-called Accadian, of greatly disputed character
and connections, but which has been for some time past
persistently declared by one party of its students to be
Ugrian, an ancient dialect of the Finno-Hungarian
stock, and a grammar of it has lately been written (by
M. Lenormant) on that understanding. This is a point
of very high importance, but we have no right yet to
consider it fairly settled ; it is doubtful whether so exact
and comprehensive knowledge and so sound method
have yet been applied as to yield a trustworthy result.
What adds greatly to the interest of the matter is that
this language and its community are demonstrably the
original owners of the cuneiform mode of writing,
which has been borrowed and adapted by both Semitic
and Indo-European peoples : it would follow, then, that
the original basis of culture in that great and important
centre of the world's civilization was Scythian. We
have no right to deny the possibility of this ; at the
same time, it is so inconsistent with what we know of
the activity of the race elsewhere that we have a right
to regard it with provisional incredulity, and to demand
a full demonstration before yielding it our belief.

Along with the three branches we have been considering
are generally ranked, as belonging to the same
family, two others, the Mongolian and the Tungusic :
but the evidence for their inclusion with the rest is
confessedly less positive, and we are justified in holding
235a doubtful position as regards them. Their languages
are of a much lower grade of development, verging
even upon monosyllabic poverty, having nothing which
can be called a verb, possessing even no distinction of
number and person in their predicative words. This
may well enough be the result of arrested growth, but
whether it demonstrably is so is another question, to
which we demand a more competent and satisfactory
reply than has yet been given. An opposing consideration
of no slight weight is the different physical type
(“Mongolian”) of these races, which connects them
rather with the extreme eastern Asiatics than with the
Europeans. Another is their possession of a “classificatory”
system of estimation and designation of relationship
(Mr. L. H. Morgan), as opposed to the analytic
or “descriptive” one of the other branches. It is not,
then, undue skepticism that leads us to limit the Scythian
family for the present to its three demonstrated
branches. Just in this direction there has been such an
excess of unscientific and wholesale grouping, the classification
of ignorance, that a little even of overstrained
conservatism ought to have a wholesome effect.

The Mongol territory occupies a great space on the
inhospitable plateau of central Asia ; and, as a consequence
of the great movement by which, in the twelfth
and thirteenth centuries, the race became the conquerors
and devastators of almost the whole world, fragments
of it are scattered far westward, one even occupying
a considerable tract astride the Volga, near its
mouth. The Mongols reach eastward along a great part
of the northern frontier of China, and are there succeeded
by the Tungusic tribes, who range still farther east and
north, almost to the coasts. Of these tribes, the only
one of note is the Manchu, whose great deed and title
236to historic fame is its conquest and administration of
China during the past two centuries. Both Mongols
and Manchus have alphabets, their usual ones derived
through the Uigur Turkish from the Syriac ; their literatures
are quite modern only, and reflections of Chinese
originals.

If in Mongol and Manchu we are close upon the
absence of all inflective structure, in the Chinese we
actually reach that condition. The Chinese is a tongue
composed of about five hundred separate words, as we
should reckon them, each a monosyllable. But in this
language tone is pressed into the service of ordinary intellectual
distinction, and the words are multiplied to
over fifteen hundred by the significant variety of intonation.
Nor are these words, like English monosyllables,
worn-out relics of a formerly inflected condition
of speech ; there is no good reason to doubt their being
the actual undeveloped roots of the language, analogous
with the Indo-European roots except in the results of
use by an enlightened community for communication
and thought during thousands of years. They have
been crowded with meanings of every kind, and of
various degrees of formality ; they have been combined
into standing phrases, with balance of parts and unity
of emphasis, as in our I shall have gone, by the way,
and so on ; many of them have become auxiliaries,
signs of relation, indicators of special uses analogous
with those of our parts of speech ; but yet they have
never been made into actual parts of speech, nor united
into inflectional systems. If they had gone through
any such process as this, the present speech would show
plainly the results of it : there would be a much greater
number and variety of words ; they would fall into related
groups ; and they would be more sharply defined
237and discriminated in their uses. The Chinese word
admits of employment indifferently as one and another
part of speech, and plainly by an inherent non-distinction
of their various offices.

The Chinese language is therefore, in one most important
and fundamental respect, of the very lowest
grade of structure and poverty of resource. But it is
also the most remarkable example in the world of a
weak instrumentality which is made the means of accomplishing
great things ; it illustrates, in a manner
which the student of language cannot too carefully
heed, the truth that language is only an instrumentality,
and the mind the force that uses it ; that the mind,
which in all its employment of speech implies a great
deal more than it expresses, is able to do a high quality
of work with only the scantiest hints of expression,
catching from the connection and from position the
shades of meaning and the modes of relation which it
needs. It is but a difference of degree between Chinese
inexpressiveness and the frequent overloading of
distinctions which in our view characterizes some of the
agglutinative idioms : for example, the American Indian ;
and, with a right view of language, one is as
explainable as the other. A few scratches on a board
with a bit of charcoal by a skilled artist may be more
full of meaning, may speak more strongly to the imagination
and feeling, than a picture elaborated by an
inferior hand with, all the resources of a modern art-school.

The abundant and varied literature of China goes
back in its beginnings to nearly 2000 B. C., an antiquity
exceeded in only two or three other countries of the
world. Though a tongue of so bald structure is comparatively
little liable to disguising alteration, the Chinese
238of to-day is quite unlike what it was so long ago —
to what extent and how, learned men are now making
effort to determine. A still more obvious measure of
the progress of alteration is given by the dialectic varieties
of the existing language, which are so great that
almost every hundred miles along the southern coast
brings one to a new speech, nearly or quite unintelligible
to dwellers in other districts. The literary dialect
is one in its written character, but somewhat discordant
in its spoken form, through the whole empire. Some
hold that here and there, in the dialects, the line which
separates utter uninflectedness from a rude agglutination
has been overstepped.

The various languages of Farther India, — as the Annamese
or Cochin-Chinese, the Siamese, and the Burmese,
with the tongues of numerous other wilder and
less important tribes or races — are sufficiently unlike to
Chinese and to each other in material to pass for wholly
unrelated. But they are all alike in the capital point
that they are uninflected ; and this cannot but be regarded
as a strong indication of ultimate relationship
between them. We can point out, indeed, no reason
why one race more than another should exhibit an incapacity
for linguistic development ; and if we met
with monosyllabic tongues in different parts of the
earth, we should have no right to infer their connection ;
but that the dialects of one corner of Asia should
share a peculiarity so exceptional can hardly be other
than the result of a common fixation of the monosyllabic
type. At any rate provisionally, therefore, we class
all these together as the southeastern Asiatic, or monosyllabic
family. The Farther Indian tongues are inferior
to the Chinese in just that manner and degree
which was to be expected in dialects of inferior races
239and lower culture. They abound in such means of
definition as auxiliaries and indicative particles.

How far the limits of the family thus constructed
extend, is a question which only further research can
determine. Running up the southern border of the
Asiatic plateau, from northern Farther India westward,
is a region occupied by a great and far from homogeneous
mass of dialects, generally called Himalayan, of a
low type of structure, which are at any rate not sufficiently
known to be classified as distinct from the family
we have been considering. With them goes the
Tibetan, though this has an alphabet, of Indian origin,
and a Buddhist literature, from the seventh century
down.

Among all these peoples, the position of the Chinese
is a striking and exceptional one, as that of the only
race possessing a wholly independent and highly-developed
civilization, with attendant literature. It is somewhat
like the position of the Accadians — if they be
proved Scythian — among the other Scythian peoples.
China has been as grand a centre of light to all its
neighbors as Mesopotamia ; but with this marked difference :
by a persistency which is one of the most striking
facts in the history of the world, it has maintained
its own institutions, political and religious and linguistic,
substantially unchanged from the very dawn of the
historic period.

The nation which has profited most by Chinese
teaching, which has alone shown the capacity to assimilate
and continue the Chinese culture, with adaptations
to its own peculiar character, is the Japanese. It is of
the same pronounced physical type which we are accustomed
to call Mongolian. Attempts have been made
to connect its language with those of the Mongols and
240Manchus, but they have not met with approved success,
and the Japanese still stands alone. It is by no
means monosyllabic, but rather an agglutinative dialect
of extremely simple structure, with hardly an established
distinction between noun and verb, and with no
determinate flexion ; the relations of case and number
and person are indicated by analytic means, by separate
particles or auxiliary words ; number in part by duplication.
Variations of the radical verbal idea akin with
those exemplified above from the Turkish are also
made, by various compounded elements. Combination
of separate root-words, often with considerable contraction
or mutilation, is very common ; but it does not tend,
as with us, to the production of formative elements and
of forms, except coarsely and restrictedly. Relatives
and subordinating conjunctions are wanting. The
language is burdened with the over-elaborate recognition
of degrees of dignity in the speaker and the persons
addressed or spoken of, almost to the disuse of
simple pronouns. The Chinese vocabulary is imported
en masse into the more learned styles, especially of
writing. The phonetic structure of the language is
very simple and euphonious. The oldest literary remains
are from the seventh and eighth centuries.

The shores and peninsulas and islands of the northeastern
corner of Asia are occupied by a variety of races
and languages, which are too little known, and of too
little interest, to demand attention from us in this hasty
review.

On the islands, however, which lie off the southeastern
part of the continent, and through most of the
groups and isolated islets that dot the Pacific, north to
Formosa, east to Easter island, south to New Zealand,
and west even to Madagascar, on the very border of
241Africa, are found the scattered members of a vast and
perfectly well-developed family, the Malay-Polynesian.
From what central point the migrations of the tribes
and their dialects took place, it is not possible to tell :
the family is strictly an insular one, the hold which a
part of the Malays have on the mainland in Malacca
being only recently gained (since the twelfth century).
The Malays proper have adopted Mohammedanism, and
taken for use the Arabic alphabet ; and they have a
tolerably abundant literature, reaching up into the
fourteenth century. Some of the other less conspicuous
tribes — as the Battaks, Mancassars, and Bugis, and
the Tagalas of the Philippines — have alphabets, which
are believed to come ultimately from India, but nothing
that can fairly be called a literature. But in Java and
its dependencies, especially Bali, the introduction of
culture and writing from India dates back even to the
first century of our era, with a considerable literature,
founded on the Sanskrit. Elsewhere in the family,
record begins only with the labors of Christian missionaries
in the most recent period.

The family is divided (Friedrich Müller) into three
great branches : 1. The Malayan, filling on the one
hand the great islands nearest to Asia, and on the other
hand the Philippine and Ladrone groups ; 2. The
Polynesian, in most of the smaller groups, with New
Zealand and Madagascar ; 3. The Melanesian, of the
Fijian and other archipelagos off the northeastern corner
of Australia. The various Polynesian dialects are
clearly and closely related ; the Melanesian show the
extreme of dialectic division, with, other peculiarities —
which, along with the darker hue and other physical
differences of their speakers, have been plausibly explained
as due to an imposition of Polynesian speech
242upon a population chiefly Papuan. The Malayan dialects
are farthest developed, making most approach
toward something like a rude flexion. For, in general,
the languages of the family are almost as bare of derivative
and inflectional combinations as is the Chinese
itself ; their grammatical relations are indicated by pronouns
and particles, which only in the Malayan group,
and in derivation rather than inflection, take on the
aspect of affixes : gender, case, number, mood, tense,
person, are wanting ; nor is there any distinction of
noun from verb ; the verb is a substantive or adjective
used predicatively without copula. The roots, if we
may call them so, the most ultimate elements accessible
to our analysis, are prevailingly dissyllabic ; and their
reduplication, either complete or by abbreviation, is a
means of variation of which great use is made, and for
very various purposes. Only the pronouns have distinct
numeral forms, and the first person has the double
plural, inclusive or exclusive of the person addressed,
referred to above (pp. 218, 219). The determinative
particles are more often prefixed than suffixed.

The Malay-Polynesian languages are more simple in
regard to their phonetic structure than any others in
the world. Hardly any of them have more than ten
consonants ; many only seven. And they do not allow
a syllable to begin with more than one consonant, or to
close with a consonant.

Not the whole population of the Pacific islands
belongs to this family. The mass of the great islands
Borneo and New Guinea, with the more inaccessible
parts of the Philippines and others, are inhabited by a
black and woolly-haired race, the Papuans or Negritos,
resembling the Africans though not related with them,
and quite distinct from the Malay-Polynesians, by whose
243incursions they nave been exterminated or crowded
back from parts of their ancient possessions. Their
languages are almost utterly unknown.

Australia, again, and the neighboring Tasmania,
were inhabited, when discovered, by a third island-race,
of dark color but straight-haired, and of nearly or quite
the lowest known grade of endowment. Their greatly
varying dialects are polysyllabic and agglutinative, of
simple phonetic character, and especially different from
the Polynesian in using exclusively suffixed instead of
prefixed particles.

In reviewing the Indian branch of the Indo-European
family, we saw that the tribes of our kindred had
worked their way in through the passes of the northwest,
driving out or subjecting a more aboriginal population.
This primitive race still holds in possession
most of the great southern peninsula, beyond the chain
of mountains and wild highlands which cuts it off from
the wide valleys of Hindustan proper. The so-called
“Dravidians” number thirty to forty millions : their
principal languages are the Tamil, Telugu, Canarese,
and Malayâlam or Malabar ; there are several others, of
inferior importance ; and the Brahuî, of Beluchistan,
outside the Indian border, is believed to belong to the
group. The Dravidian tongues have some peculiar
phonetic elements, are richly polysyllabic, of general
agglutinative structure, with prefixes only, and very
soft and harmonious in their utterance ; they are of a
very high type of agglutination, like the Finnish and
Hungarian ; and the author has been informed by an
American who was born in southern India and grew
up to speak its language vernacularly along with his
English, a man of high education and unusual gifts as
a preacher and writer, that he esteemed the Tamil a
244finer language to think and speak in than any European
tongue known to him.

Excepting that they show no trace of the harmonic
sequence of vowels, these languages are not in their
structure so different from the Scythian that they might
not belong to one family with them, if only sufficient
correspondences of material were found between the
two groups. And some have been ready, though on
grounds not to be accepted as sufficient, to declare them
related. The comparative grammar of the Scythian
languages has not yet been so reduced to form that it
should be possible to define the boundaries of the family,
either on the west or in the south.

Among the less familiar languages of Asia we have
occasion to notice further only that intricate and problematical
group known as the Caucasian. As the name
denotes, its locality is the region between the Caspian
and Black Seas, filled by the Caucasus range and its
dependent hills and valleys. The chief dialects on the
south of the main crest are the Georgian, Suanian, Mingrelian,
and Lazian, all plainly related to one another,
and the first having an alphabet, derived along with its
religion from Armenia, and a literature of some antiquity.
The principal groups on the north are the
Circassian, Mitsjeghian, and Lesghian, the first bordering
the Black Sea, the last the Caspian. The variety of
sub-dialects, especially of the Lesghian, is very great.
There is no demonstrated affinity between the southern
and northern divisions, nor between the members of the
northern ; how many independent groups there may be
is yet undetermined ; and also, whether there is any
tie of analogical structure to bind them together into a
family, or whether they are the relics of ultimately
separate families, left stranded, as it were, on the
245mountains, and defended by them and by the great seas
in front and behind from the movements of migration
which have swept the families elsewhere out of existence.

Last among the Asiatic languages, we come to the
Semitic, so called because in the genealogies of the
Genesis the communities which speak them are mostly
described as descendants of Shem. They fill the immense,
but barren and thinly-populated peninsula of
Arabia, with its northern border-lands, of Mesopotamia
and Syria and Palestine, and with a district in Abyssinia,
lying opposite its southwestern corner. The various
dialects of the Arabic, with its African outlier,
constitute one branch of the family ; the Canaanitic
dialects, chief among which are Hebrew and Phœnician,
with the Syrian or Aramaic, a second ; and the Assyrian
and Babylonian a third. This is their ancient territory :
the Phœnician was carried to its colonies, and, as Carthaginian,
might perhaps have become the tongue of
Mediterranean civilization, but that the long struggle
for supremacy ended with the complete overthrow of
Carthage by Rome ; the Hebrew, replaced in vernacular
use, even in its own home, four centuries before
Christ, by the Syrian (Chaldee, Aramaic), has led ever
since the artificial life of a learned language, scattered
among the civilized nations ; the Arabic, as the sacred
dialect of a conquering people and religion, has been
carried, since the seventh century, over a part of the
world comparable with that which the Latin came finally
to occupy : it is the speech of the whole northern
border of Africa ; it has crowded out the other Semitic
branches, and has filled with its words the Persian,
Turkish, and Hindustani, and to a less extent the Malay
and Spanish vocabularies. It has given birth, however,
246to no such group of independent derived languages as
the Latin can show.

The ancient Hebrew literature is familiar to us far
beyond the rest, being our “Bible ;” its earliest parts
go back into the second thousand years before Christ.
The Phœnician has left no literature, and the inscribed
coffin of a king of Sidon (probably 500 B. C.) is its chief
monument ; a very recently discovered Moabite tablet
(of 900 B. C.) gives us a specimen of another ancient
Canaanitic dialect, almost identical with Hebrew. The
Aramaic has an abundant Greco-Christian literature,
beginning from the second century, besides its share in
the Talmudic writings. The Assyrian has a fragmentary
literature in the inscriptions and tablets of Nineveh
and Babylon, from a period beyond that of the earliest
Hebrew. The Arabic begins its record mainly with
the rise of Islam ; since that time it is one of the richest
literatures in the world. In southwest Arabia prevailed
a very different body of dialects, usually styled
Himyaritic, now preserved only in the jealously-guarded
remains of an earlier civilization. With the Himyaritic
is most nearly akin the Abyssinian group, which,
in two principal literary dialects, the earlier Geëz or
Ethiopic and the later Amharic, has a considerable literature,
beginning in the fourth century.

The Semitic family of languages and races is, after
the Indo-European, by far the most prominent in the
history of the world. None but the Semites have, since
the dawn of the historic period, seriously disputed with
our family the headship of the human race ; and, of the
three great conquering religions, two, Christianity and
Mohammedanism, are of Semitic birth — although the
former won its world-wide dominion in connection with
its transfer to the hands of Indo-Europeans, the Greeks
247and Romans. That we have put off, then, our examination
of Semitic language to this point is mainly owing
to its exceptional and anomalous character. Semitic
speech stands more alone in the world than any other,
than even the nakedly isolating Chinese or the indefinitely
synthetic American. For, as regards all other
tongues, the basis of radical elements and the principle
of their combination being given, it is easy enough in
theory to explain their various structures, as products
of one general method of development. But no such
thing is at present practicable for the Semitic ; this
contains two characteristics — the triliterality of the roots
and their inflection by internal change, by variation of
vowel — which belong to it alone.

What we call the Semitic root, namely, is (except in
the pronouns and a wholly insignificant number of
other cases) a conglomerate of three consonants, no
more and no less : thus, for example, q-t-l represents
the conception of ‘killing,’ k-t-b that of ‘writing.’ By
this is not meant, of course, that such conglomerates
were, like the Indo-European roots, the historical germs
of a body of derivative forms ; but, as we arrive at the
root in Indo-European by taking off the variously accreted
formative elements, we arrive at such a Semitic
root by removing its formative elements. The latter
includes no vowel that has an identity to preserve ; the
addition of any vowel makes a form. Thus, in Arabic
(the best preserved and most transparent in structure of
the various dialects), qatala is a verbal third singular,
‘he killed ;’ as it were, the base of a system of personal
forms, made, like ours, by pronominal endings :
thus, qataltu, ‘I killed,’ qatalat, ‘she killed,’ qataltumâ,
‘ye two killed,’ qatalnâ,‘we killed.’ A change of vowels,
to qutila, makes of it a passive, ‘he was killed ;’
248and from this we have by a like process qutiltu, qutilat,
qutiltumâ, qutilnâ, etc. Another change, to aqtala,
signifies ‘he caused to kill,’ with its passive uqtila ; and
so on. Then (u)qtul is imperative, ‘kill !’ and something
like this is base of another set of persons, formed
partly by prefixes, partly by suffixes : as yaqtulu, ‘he
kills,’ taqtulu, ‘she kills,’ yaqtulûna, ‘they (men) kill,’
naqtulu, ‘we kill,’ etc. Then, qâtil is present participle,
‘killing,’ and qatl infinitive, ‘act of killing ;’ while
iqtâl is ‘causing to kill’ as noun, and muqtil the same
as adjective. And qitl, ‘enemy,’ and qutl, ‘murderous,’
are specimens of derivative noun and adjective.
These forms at once suggest our sing, sang, etc., already
often used as illustrations ; yet there is an immense difference
between the two cases : the Semitic phenomena
are infinitely more intricate and various ; and then they
are the very life and soul of the inflection of the language,
not in a single item reducible to anything more
original, out of which they should be seen to grow, by
an “inorganic” process. If we could conceive that, at
some peculiarly plastic period in the history of a Germanic
dialect, by an abnormal extension of the analogy
of sing, sang, etc., the popular taste taking a sudden
bent toward such formations, all the rest of the language
should come to be patterned after that model,
with consequent complete oblivion of the state of things
out of which sing, sang, etc. proceeded — that would be
something analogous with the present condition of Semitic.

The other peculiarities of the language are trifling
as compared with these, not different in kind or degree
from such as are variously found in other tongues. The
structure of the verb is quite unlike ours. The element
of time does not enter distinctly into it ; the (only) two
249sο-called tenses are explained as indicating primarily
complete and incomplete action, and each fills various
offices of tense. In Assyrian, the tense of complete
action has gone almost entirely out of use. Of forms
analogous with our moods, too, there is great poverty.
But, as we have found the case in more than one other
family, there is a disposition to the formation of numerous
conjugations from one root, representing the radical
idea in a causative, a reflexive, an intensive, a conative
form, and so on. In Arabic, where these changes are
fullest, there are some fifteen such conjugations ; and
about a dozen of them, each with its passive, are in tolerably
frequent use. The tense of incomplete action
(yaqtulu, etc.) has the aspect of being younger than
the other, and of standing at only one remove from a
noun ; since its endings of number are mainly coincident
with those of ordinary noun inflection, and it denotes
person by prefixes, while the other (qatala, etc.)
indicates person and number together by added endings,
evidently of pronominal origin. Both tenses distinguish
masculine from feminine subject, except in the
first person. We find the distinction of gender (masculine
and feminine only) here again for the first time
since we left the Indo-European family. The nouns
have the same three numbers as the verb, but of case
distinction there is almost nothing. Derived nouns are
formed by the help both of internal flexion and of external
additions, both prefixes and suffixes ; but only
directly from the root : those successive derivations, by
ending added to ending, in which the Indo-European
abounds (as true, tru-th, truth-ful, un-truthful-ly) are
quite unknown. Nor are compounds formed, save in
exceptional cases. Finally, connecting particles, as
means of the intertwining and subordination of clauses,
250their conversion into a period, are almost wanting :
Semitic style is bald and simple, proceeding from assertion
to assertion. Another marked peculiarity is the
persistency of radical meaning in derivative and figurative
expression : the metaphorical or other transfer by
which a new term is won, instead of soon passing out
of memory, as in Indo-European, lets the old meaning
continue to show through. Picturesqueness, pictorial
vividness, therefore, are leading characteristics of Semitic
language.

The scale of dialectic differences is much less in
Semitic than in Indo-European ; all the great branches,
even, are as it were the closely related members of a
single branch. This is not necessarily because their
separation has been more recent than that of the
branches of our family ; for Semitic speech has shown
itself much more rigid and changeless than Indo-European — or,
it is believed, than any other variety of human
speech. The ground of this difference doubtless
lies partly in the character of the speakers ; but it is
also in part to be plainly read in the character of the
language itself, with its rigid framework of three consonants
appearing in the whole body of derivatives of
each root, with its significant and therefore more carefully
maintained variations of vowel, and with its incapacity
of new formations by composition. Its primitive
development, if development it was, was into so
individual and sharply defined a type that it has since
been comparatively exempt from variation.

There are two ways of looking at the peculiarities
of Semitic structure. One, by far the simpler and.
more comfortable, is to pronounce them original and
inexplicable, an indefeasible part of the appanage of
the Semitic mind, to be taken as presented, and no
251questions asked. This, however, is virtually to declare
them outside the pale of science, to abnegate with regard
to them the right of the linguistic student to ask
after the why of what he finds anywhere in language.
The other way is to put this question and pursue it, not
daunted by the acknowledged difficulties of the case.
If all other languages have had a history of development
into their present shape, then doubtless the Semitic
also ; if all the rest have started from pronounceable
roots, composed of a combination of consonant and
vowel, and have grown by external accretion of other
similar elements to these, then it is not lightly to be
believed that the Semitic has not done the same. That
is to say, there must probably lie behind the consonantal
triple roots and the internal flexion of the Semitic
something more analogous with what is seen to lie at
the basis of all other human speech ; and there must
have been a history of change from the one of these
conditions to the other — whether we shall or shall not
prove able to retrace the history and restore the primitive
condition. Most linguistic scholars, as might be
expected, take the latter view ; and the attempt has
been repeatedly made to reduce the roots to a more
primitive form ; but no definite and solid results have
been yet attained. The most plausible conjectural account
of the matter, probably, yet suggested has been,
that the universality of the three root-consonants is due
(as in our hypothetical case above) to the inorganic extension
of an analogy which had in some way become a
dominant one ; and that a stage of dissyllabic or trisyllabic
derivative nouns lies between the primitive roots
and their present shape. But to offer a plausible conjecture
is one thing, and to demonstrate its value as a
true explanation is another ; and until something like
252a demonstration is reached (which possibly may never
be), there will doubtless continue to be those who
will look upon Semitic triliterality and internal flexion
as original, as not only inaccessible to explanation but
calling for none.

It must, however, be admitted that with the retracing
of Semitic root-history is indissolubly bound up the
historical connection of Semitic language with any other
form of human speech. So long as Semitic flexion remains
what it is, it cannot be identified with that of
any other language ; so long as Semitic roots remain
what they are, no resemblances which may be traced
between them and those of any other language can have
real value. It has been a favorite subject of effort with,
scholars, ever since the beginning of linguistic study, to
connect the germs of Semitic and Indo-European speech,
and to prove the two families and the races that speak
them branches of an ultimately common stock. There
are many things which tempt to this : the two peoples
are, at the beginning of their cultural history, near
neighbors and mutual helpers ; they are the two great
conquering and civilizing white races, exchanging influence
and institutions with one another through the
ages : how natural to connect them more closely with
one another than with mankind in general ! This consideration
goes all the way back to the representation
of Shem and Japhet as sons of one father. But here,
again, plausible theory is one thing, and scientific demonstration
another. If the items of apparent agreement
which great scholars have hunted up between
Semitic and Indo-European had been pointed out as
existing between Indo-European and Zulu or Papuan,
no one would think them of any account ; and they are
really worth no more where they are, as scientific evidence.
253It cannot be too strongly insisted on that, until
the anomalies of Semitic language are at least measurably
explained, it is too soon to say anything about a
relationship between it and any other tongue.

The same rule is to be applied to the current assertions
of Semitic relationship in the opposite direction,
with the tongues which are grouped together to form
the “Hamitic” family. In this family, the Egyptian
occupies the same commanding position as the Chinese
among the monosyllabic tongues of southeastern Asia.
Egypt is the home of by far the oldest civilization of
which we have any records. The question as to the
chronology of its earliest monuments is not, to be sure,
settled beyond dispute ; but the present tendency of
scientific inquiry seems decidedly toward recognizing as
well founded even the extreme claims put forth respecting
them, and fixing the reign of the first historical
king at nearly 4000 B. C. ; and even at that time the
race must have been a powerful one, with a highly developed
civilization. The knowledge of Egyptian language
has been recovered in our own century, after
being utterly lost for near two thousand years, and remarkable
discoveries of new material in the country
itself, and advances in Egyptian learning in Europe,
are at this very time going on ; so that many of the
historical and chronological questions about which we
are disputing will be fully settled for the generation
that succeeds us.

The key to the decipherment of the ancient Egyptian
was furnished in its descendant, the modern Coptic.
The Coptic records are Christian only, written in
an alphabet derived from the Greek, and dating back
to the early centuries of our era. But the language
was extinguished in vernacular use by the Arabic, three
254or four centuries ago. Several slightly different dialects
are to be recognized in its literary remains.

The Egyptian language, old and new, was of the
utmost simplicity of structure. It hardly knew a distinction
between root and word ; its fundamental elements
(not always monosyllabic) were brought directly
into the combinations of the sentence, without formal
means of distinction of one part of speech from another.
Nor even in inflection is such distinction clearly
made ; noun and verb are separated in part by the connection
only : ran-i, for example, is literally ‘naming-mine,’
and means either ‘my name’ or ‘I name or call.’
The personal inflection of the verb is by means of affixed
pronouns, loosely agglutinated to it, that of the
third person being omissible when a subject noun is
expressed. Mood and tense are marked, within narrow
limits, by prefixed auxiliary words. The noun has no
declension : relations of case are denoted by connectives ;
its use as noun is generally marked by a prefixed
“article.” And in this article, as in the pronominal
elements generally, is made in the singular a distinction
of masculine and feminine gender — a marked peculiarity
of the language, putting it so far into one class with
the Semitic and Indo-European. This particular, however,
is one of which the reach and importance are wont
to be greatly exaggerated ; in its general character, the
language can sustain no comparison at all with the other
two mentioned ; it is little richer or more developed
than the lowest tongues of the eastern Asiatic races.

It must be clearly apparent from this description
how venturesome is the assertion of a relationship between
the Egyptian and Semitic. There are, to be sure,
certain remarkable resemblances between the pronouns
of the two languages ; but to rely on these as sufficient
255proof of connection is not an acceptable proceeding.
In many languages, signs of relationship, abundantly
traceable through their whole material, are especially
conspicuous in the pronouns ; of connection proved by
pronominal evidence solely, or chiefly, there are no examples.
And the question is, whether pronominal
words could possibly retain an almost undisguised identity
while the rest of the language was undergoing such
a tremendous revolution as should alone be able to convert
Egyptian poverty of inflection and fixity of root
and freedom of radical form into the strictly regulated
wealth and internal flexion of the Semitic. And the
provisional answer must be in the negative. We do
not need to deny the possibility of ultimately proving
the Semitic related with the Hamitic, any more than
with the Indo-European ; we have only to see that no
sufficient evidence of it has yet been brought forward,
nor is likely to be so until the riddle of Semitic structure
is solved.

It is held by students of African language that a
considerable body of other tongues show signs of ultimate
connection with the Egyptian, forming with it the
Hamitic family. There is the Libyan or Berber of
northern Africa, and a considerable group south of
Egypt, having the Galla as its most prominent member,
and known as the Ethiopian.

Nearly the whole of the narrower southern peninsula
of Africa is occupied by the branches of a single
very distinct family, best called the South-African
(known also as Bantu, Chuana, Zingian). It has no
culture and no literature, except what it has produced
by the aid of Christian missionaries in the most recent
time. It is strikingly characterized by its extensive use
of prefixes : a word without a formative prefix being
256here nearly as unknown as, in the synthetic period of
Indo-European, a word without a formative suffix.
Different prefixes distinguish various classes of nouns,
and numbers in those classes : thus, in Zulu, um-fana
is ‘boy,’ and aba-fana ‘boys ;’ in-komo is ‘cow,’ and
izin-komo ‘cows ;’ ili-zwe is ‘country,’ and ama-zwe
‘countries,’ and so on. Then, in the clauses into which
any one of these words enters as dominant member,
other members relating to them — as adjectives, possessives,
verbs — take into their structure representative
parts of the same prefix : e. g. aba-fana b-ami aba-kulu,
ha tanda, ‘my large boys, they love ;’ but izin-komo
z-ami izin-kulu, zi tanda, ‘my large cows, they love.’
This is like Latin or Greek inverted ; an alliterative
instead of a rhyming congruence. Verbal mood and
tense are signified in part by suffixes, as are also conjugational
distinctions analogous with those made in
Scythian and Semitic language : thus, from bona, ‘see,’
come bonisa ‘show,’ bonana,‘see each other,’ bonisana,
‘show each other,’ and so on. Case-relations are signified
by prefixed prepositions. The South-African
languages are thus by no means unprovided with the
formal means of sufficiently various distinction. Those
of them which border on the Hottentot dialects have in
their alphabets peculiar sounds called “clicks,” made
by sharp separation of the tongue from the roof of the
mouth, with suction.

The clicks are a marked feature of the Hottentot,
and look as if they had been introduced into the South-African
from thence, perhaps along with mixture of
blood. There is no relationship whatever between the
two families ; nor, probably, between the Hottentot and
the Bushman. Of the last mentioned, the scientific investigation
is now just beginning (Bleek) ; the other,
257chiefly on the ground of its partial distinction of genders,
has been by some accounted a branch of the
Hamitic family, strayed away into the far south and
greatly degraded in type ; but the connection is confidently
denied by others.

Between the South-African and Hamitic domains,
in a broad band extending across the widest part of the
African continent, is found an intricate and heterogeneous
mass of dialects, of which the classification is a
matter of much difference of opinion among even the
latest investigators, and which are of too little importance
to be dwelt on by us. The region is that of the
typical negro ; yet there are also in it races of a lighter
tint : the variety of physical characteristics in Africa,
among races which we in our ignorance lump together
as one, is very great.

Before leaving the eastern continent, we must return
to Europe for a word or two upon one language
which has as yet found no place for notice — the Basque,
now spoken, in four principal dialects and a number of
minor varieties, in a very limited mountain-district at
the angle of the Bay of Biscay, astride the frontier, but
chiefly on the Spanish side. It is believed to be the
modern representative of the ancient Iberian, and to
have belonged to the older population of the peninsula,
before the irruption of the Indo-European Celts.
Traces of local nomenclature show it to have occupied
also at least the southern part of France. The Basque
may then be the sole surviving relic and witness of an
aboriginal western European population, dispossessed
by the intrusive Indo-European tribes. It stands entirely
alone, no kindred having yet been found for it in
any part of the world. It is of an exaggeratedly agglutinative
type, incorporating into its verb a variety of
258relations which are almost everywhere else expressed by
independent words.

The Basque forms a suitable stepping-stone from
which to enter the peculiar linguistic domain of the
New World, since there is no other dialect of the Old
World which so much resembles in structure the American
languages. Not that the latter are all of accordant
form. Although it is usual among philologists to account
them as making together but a single great family,
this is in no small part a classification of ignorance,
and should be held only provisionally, ready to be
changed, if necessary, when additional knowledge is
won. As regards the material of expression, it is fully
confessed that there is irreconcilable diversity among
them. There are a very considerable number of groups,
between whose significant signs exist no more apparent
correspondences than between those of English,
Hungarian, and Malay : none, namely, which may not
be merely fortuitous. So, for example, between the
neighboring tongues of the Algonkin, Iroquois, and
Dakota groups, the speakers of which we have every
reason to regard as ultimately related, on the ground of
common physical characteristics, gifts, and institutions.
Indeed, there is even linguistic evidence to the same
effect. The case seems to be clearly one where the
style of structure of a language is more permanent than
the material, constituting of itself a satisfactory proof
of relationship. That is to say, while the material elements
of these tongues have been highly variable since
their separation from one another, till identities in this
department are no longer traceable — a feature in their
history which we shall understand and judge more truly
when the special laws of their growth and change shall
be much better comprehended — there still remains, unaltered
259in its main features, their common mode of
managing and combining the linguistic material, of
apprehending the relations which are to be expressed
in language, and the way in which they shall be expressed.

And this common mode of structure, which, in its
various aspects and degrees, is at least generally characteristic
of American language, is called the polysynthetic
or incorporating. Its marked tendency is toward
the absorbing of the other parts of the sentence into
the verb. Not the subject alone, as in Indo-European,
enters into combination with the root for predicative
expression, but the objects also, of every kind of relation,
and the signs of time and place and manner and
degree, and a host of modifiers of the verbal action,
for purposes unknown to any grammatical system with
which, we are ordinarily familiar. It has been deliberately
calculated, by one long versed in the chief Algonkin
dialects (Rev. T. Hurlbut), that 17,000,000 verbal
forms may be made from an Algonkin root ; and even
if our credence were to extend to only the thousandth
part of this, enough would be left to be very characteristic
of a structural style. Everything tends to verbal
expression : nouns, and adjectives, and even adverbs
and prepositions, are regularly conjugated ; nouns are
to a great extent verbal forms : e. g. ‘home’ is ‘they
live there,’ or ‘where they live.’ Or, to express it
more accurately, our grammatical terminology does not
at all suit these languages ; we are involved in contradictions
and absurdities as soon as we attempt to apply
it to them. Of course, the tendency is toward the
formation of words of immense length, and of an intricate
structure that gives expression to a host of things
left by us to be understood. The longest word in Eliot's
260Massachusetts Bible, however, is of eleven syllables :
wut-appesituqussun-nooweht-unk-quoh, which renders
“kneeling down to him” in our version ; but it really
means ‘he came to a state of rest upon the bended
knees, doing reverence unto him’ (J. H. Trumbull).
All the parts of such combinations must be recognized
in their separateness ; the word must be in all its members
significant and self-explaining. And the separate
elements are not, as is often represented, a reduction to
manageable fragments of long words for which they
stand ; they are rather the desired significant element
among those which compose the other word. Of
course, there are infinite possibilities of expressiveness
in such a structure ; and it would only need that some
native-American Greek race should arise, to fill it full
of thought and fancy, and put it to the uses of a noble
literature, and it would be rightly admired as rich and
flexible, perhaps, beyond anything else that the world
knew. As it is, it makes upon us the impression of as
much exceeding the due medium of formal expressiveness
as the Chinese comes short of it ; it is cumbrous
and time-wasting in its immense polysyllabism. Partly
as a result of its multiplicity of accessory details, it
seems to us deficient in simple abstract terms : as having,
for instance, separate roots for washing all kinds of
objects, in all kinds of ways, but none for ‘washing’
pure and simple. There is something of our prejudice
in this, however ; so a Chinaman or Englishman might
criticise a Latin adjective unfavorably, saying : “The
Latin is deficient in the power of abstraction, of considering
a quality apart from its accidental accessories :
so magnus, for example, does not signify simply ta,
‘great,’ but a quality of great of a first degree, and as
belonging to only one object, and to one that is (for
261some unassignable reason) regarded as masculine and
can be only the subject of a verb ; magnas indicates in
like manner an objective and feminine and plural greatness ;
but for the bare idea of ta, ‘great,’ the Latin has
no expression.”

There are other characteristics of American speech,
of universal or general prevalence, like the distinction
of animate and inanimate gender (which would seem to
be quite as significant, and as capable of being applied
to higher formative uses, as is our own sexual gender),
the possession of the inclusive and exclusive first persons
plural, the classificatory system of designation of
relationships, and so on ; but they are of only minor
importance, as compared with the general style of
structure.

The polysynthetic structure does not belong in the
same degree to all the American languages ; on the
contrary, it seems to be altogether effaced or originally
wanting in some. So, for example, a monosyllabic or
uninflective character has been claimed for the Otomi
in Mexico, and for one or two dialects in South America ;
and all sign of polysynthetism has been denied (C.
F, Hartt) to the great Tupi-Guarani stock, on the
eastern side of the South American continent. It remains
yet to be determined how far such exceptions are
real, and how far apparent only. But the common
character is recognizable in so large a part of American
tongues, from the Eskimo of the extreme north to the
Antarctic Ocean, that the linguist regards them, with
considerable confidence, as members of a family, descendants
of one original speech, of unknown age, locality,
and derivation. Attempts have been made to
connect them with some dialect or family of the Old
World, but with obviously unavoidable ill-success. If,
262for example, there is not left in Algonkin, Iroquois, and
Dakota enough of the material once common to the
ancestors of all to furnish ground for trustworthy
identifications, much less are they to be identified with
tongues from which they have been so much longer
separated that even their structure is of a different character.
It is not proper, perhaps, to limit the possibilities
of the future ; but there appears to be no tolerable
prospect that, even supposing the American languages
derived from the Old World, they can ever be proved
so, or traced to their parentage.

An exhaustive classification of the American languages
is at present impossible ; and to give what can
already be given would demand much more space than
can be afforded here. There are many great groups,
and a host of lesser knots of idioms, or of dialects
isolated or unclassified. The Eskimos line the whole
northern coast, and the northeastern down to Newfound-land.
The Athabaskan or Tinné occupies a great region
in the far northwest (the Apache and Navajo in
the south also belong to it), and is flanked on the west
by the Selish and other smaller groups. The Algonkin
had in possession the northeastern and middle United
States, and stretched westward to the Rocky Mountains ;
within its territory was included that of the
Iroquois. The Dakota (Sioux) is the largest of the
families occupying the great prairies and plains of the
far West. The Muskokee group filled the States of the
southeast. In Colorado and Utah commence the towns
of the settled and comparatively civilized “pueblo
Indians,” rising to the more advanced culture of the
Mexican peoples, attaining its height in the Mayas of
Central America, and continued in the empire of the
Incas of Peru. The Quichua of the latter, with the
263related Aymara, are still the native dialects of a considerable
part of South America ; with the Tupi-Guarani,
already referred to, on the east, in the valleys of the
Amazons and its tributaries.

The condition of American languages is thus an
epitome of that of the language of the world in general.
Great and wide-spread families, limited groups,
isolated and perishing dialects, touch and jostle one
another. Such, in the vicissitudes of human affairs,
must be the history of races and of their dialects.
What families, once covering great tracts of the earth's
surface, have been wiped out without a trace, what
others have been reduced to mere fragments, what have
started from a narrow beginning, and, by prosperous
growth and by working in parts of other races, have
risen to prominence — on such points as these we must
remain forever only imperfectly informed. We need
to guard against supposing that, when we have succeeded
in classifying all existing languages and determining
their relations, we shall have gained a complete outline
of the history of human language : the darkness of the
past may hide a great deal of which we do not even
catch a glimpse.

Some of the questions bearing on this point will
engage our attention in the next chapter.264

Chapter XIII.
Language and ethnology.

Limitations to the scope of linguistic science : materials of speech not
analyzable to the end ; annihilation, transmutation, new creation,
possible in it ; cumulative character of evidences of relationship.
Impossibility that language can prove either unity or variety of the
human race. Relation of language to race, as transmitted institution
only ; exchange of language accompanying mixture of blood.
Insolubility of the ethnological problem. Contributions to it of
archaeology and linguistics ; merits of the latter ; importance of the
testimony of language to race. Reconciliation of the various lines of
ethnological evidence. Inferior value of other classifications of language
as compared with the genetic.

The classification of languages given in the preceding
chapter has confessedly represented only the present
state of knowledge, and is liable to amendment hereafter,
as further investigation shall bring more light.
But its main features will probably stand unaltered.
The leading independent families will continue separate
to the end. One and another of those now recognized,
it is true, may hereafter assume a dependent place, as
branches of a wider and more comprehensive family,
but there is no reasonable ground for anticipating that
such will ever be the case with them all. To maintain
this is not so much to limit the future of linguistic science,
as, rather, to recognize the limits which in the
265nature of things are set to its progress ; as a brief and
simple exposition will show.

We must not fail to appreciate the essential difference
between the material of the physical sciences and
that of our subject ; that we have to deal with the
usages of men, in all of which intervenes that indefinite
element, the human will as determined by circumstance,
by habit, by individual character ; and that these do not
admit an analysis penetrating to the ultimate elements.
There is no natural substance which the chemist may
not aspire to analyze ; into whatever new forms and
combinations an element may enter, he has tests which
will detect its presence ; neither new creation nor annihilation
is possible ; all change is but recombination
of material always existing ; there is no transmutation
of one element into another. But it is altogether different
with speech. A word, a whole family of words,
perishes by simple disuse, and is as if it had never been,
unless civilization is there to make a record of its departed
worth. A whole language, or family of languages,
is annihilated by the destruction of the community
that spoke it, or by the adoption of another language
by that community. When the Gauls learned
Latin, there was nothing saved which, without the aid
of external evidences, should show what their primitive
speech had been ; when the Etruscans were Latinized,
but for the scattering words which they had written
down, their speech passed out of all reach of knowledge :
and many a dialect has doubtless gone out in a
like way, leaving no such telltale records. The actual
creation of the new in speech is, as we have seen, very
rare ; yet there is nothing whatever to prevent it save
men's preferences. And it amounts, for all purposes of
analysis, to a new creation, when a derivative word gets
266so far from its primitive, in form and meaning, that the
tie between them is traceable only by external, historical
evidence : and of such cases all language is full. A
formative element is annihilated when it is worn off
from every form which it once made ; such a one is
created when it is fully established in its derived and
subordinate use : no process of analysis that we have or
of which we can conceive would ever find the lost masi
of our first persons plural, or detect the presence of did
in loved : there is wanted the historical support, for
lack of which a host of other like cases cannot be accounted
for.

The changes of linguistic usage are all the time separating
in appearance what really belongs together :
bishop and évêque are historically one word ; so are eye
and auge ; so are I and je and ik and ἐγών and aham ;
though not one of them has an audible element which
is found in any other. And then, the same changes
are bringing together what really belongs apart : the
Latin locus and the Sanskrit lokas, ‘place, room,’ have
really nothing to do with one another, though so nearly
identical and in closely-related languages ; likewise
Greek ὅλος (holos) and English whole ; and so on. We
may take the English language (as too many do), and
compare it with every unrelated dialect in existence,
and find a liberal list of apparent correspondences ;
which then a little study of the English words will
prove unreal and fallacious. This is, above all others,
the decisive fact which stands in the way of a comparison
that shall penetrate to the bottom of the matter.
If there were no resemblances in either the material or
the structure of language save such as have a historical
basis, we might let them be swept away as much as
they would ; what was left, if anything were left, would
267suffice to prove relationship. As it is, the process of
proof is not direct and absolute, but cumulative ; the
result comes from a sufficient number of particulars of
which each, taken by itself, would prove nothing. We
have had expressly to allow that two dialects may diverge
from a common original so far that all sign of
their kinship shall be lost ; there may be a plenty of
the altered products of common material in them both ;
but if it have gotten into the condition of bishop and
évêque, it is of no use to the linguist. Accidental correspondences
are capable of rising to a certain percentage ;
if all that appear stand at or near this figure, the
case is one hopeless of settlement.

This cumulative character of the signs of relation-ship,
the uncertain value of any single item, and the
need of historical evidence to support their interpretation,
set limits to the reach and competence of linguistic
investigation. Thus far, the recognized families are
such as have had a common development. There are
even some of which the sole uniting tie is a common
style of structure. If we cannot prove the American
languages related except by the characteristic of polysynthetism,
nor the southeastern Asiatic except by that
of monosyllabism, it is obviously impossible to prove
American and Chinese related by the material correspondences
of their roots. In the present stage of linguistic
science, root-comparisons are surrounded with
too many uncertainties and dangers to have any value.
All that have been made thus far are worthless ; whether
the future will show anything better, we may leave
for the future to determine. There is no harm in any
one's rating even too highly the possibilities of a progressive
science like linguistics, provided he do not let
his sanguineness warp his judgment as to what shall
268have been at any given time already accomplished, and
lead him to take plausible fancies for tried and approved
facts. He who realizes the immense difficulty
of arriving at the ultimate roots even of a family like
the Indo-European, despite the exceptional antiquity
and conservation of its oldest dialects, will be likely to
be saved from hanging his expectations on root-comparisons.

It is, then, impossible that linguistic science should
ever be able to prove, by the evidence of community
of the first germs of expression, that the human race
in the beginning formed one society together. Even if
the number of families be lessened by future research,
it will never be reduced to one.

But it is even far more demonstrable that linguistic
science can never prove the variety of human races and
origins. As we have repeatedly seen, there are no limits
to the diversity which may arise by discordant
growth between languages originally one. Given any
angle of divergence, and the law of increasing divergence
(p. 165), and the distance of the ends of two lines
may be made, by their production far enough, to exceed
any assignable quantity ; and in linguistics, as has been
just pointed out, there comes, far short of infinite prolongation,
a distance across which the historical scholar,
with his limited vision, cannot see : and that is, for all
practical purposes, infinity. The understanding now
won of the methods of growth and change in speech
has taken away all possibility of a dogmatic assertion
on the part of the linguistic scholar that language has
a various origin. If every tongue had from the beginning
its own structure and material complete, then language-history
would run back only in parallel lines,
with no indication of convergence. But the difference
269of English, and German and Danish comes by divergent
growth from a common centre ; that of English and
Russian and Armenian and Persian is by similar divergence
from a more distant centre : and we cannot say
that English, and Turkish and Circassian and Japanese
may not owe their difference to the same cause. The
lines of development of all families of language do point
back to one original common condition of formless
roots ; and precisely what these roots were, in shape
and meaning, we cannot in most families even begin to
trace out ; we cannot, then, deny that they may have
been the same for all. We may talk of probabilities as
much as we please ; but of impossibility there is actually
nothing in the assumption of identity of origins.

This, again, implies that linguistic science cannot
assume to prove the diversity of human races. But it
deserves to be pointed out that there is an additional
difficulty in the way of the same proof. If we must
regard it as at least possible (whether we admit it as an
established conclusion or not) that men made the beginnings
of their own speech, as well as created all its
after-development, then we shall be obliged also to allow
that a period of some length may have elapsed
before any so settled store of expression had been won
that it should show itself in the later forms of language ;
and during this period the race, though one,
might have spread and separated, so that the abiding
germs of the speech of each part should be independent.
As a general conclusion, the incompetence of linguistic
science to pass any decisive judgment as to the unity
or diversity of the human race, or even as to that of
human speech, appears to be completely and irrevocably
demonstrated.

Another highly important anthropological question,
270connected with and suggested by our classification of
languages, concerns its relation to the ethnologist's classification
of races. And here we have to make at the
outset the unreserved confession that the two do not by
any means correspond and agree : wholly discordant
languages are spoken by communities whom the ethnologist
would not separate in race from one another ; and
related languages are spoken by men of apparently different
race. And the view we have taken of language
is entirely consistent with this. We have seen that
there is no necessary tie between race and language ;
that every man speaks the language he has learned, being
born into the possession of no one rather than another ;
and that, as any individual may learn a language
different from that of his parents or of his remoter
ancestors, so a community (which is only an aggregate
of individuals) may do the same thing, not retaining
the slightest trace of its ancestral speech. The world,
past and present, is full of examples of this, of every
class and kind, and sundry of them have been already
noticed by us in passing — as the combination of heterogeneous
elements, now using only English as their
native tongue, found in the American community ; the
Celts of Gaul, the Normans of France, the Celts of Ireland
and Cornwall, the Etruscans of Italy, and all the
other communities whose idioms have been crowded
out and replaced by the Latin, the English, the Arabic.
There are conquering languages which are always encroaching
upon the territory of their neighbors, as there
are others which are always losing ground.

The testimony of language to race is thus not that
of a physical characteristic, nor of anything founded on
and representing such ; but only that of a transmitted
institution, which, under sufficient inducement, is capable
271of being abandoned by its proper inheritors, or assumed
by men of strange blood. And the inducement
lies in external circumstances, not in the nature of the
language abandoned or assumed. Political control,
social superiority, superiority of culture — these are the
leading causes which bring about change of speech.
Or rather, these are the added circumstances which, in
the case of a mixture of communities, decide which
element of population shall give, chiefly or wholly, its
tongue to the resulting community. If there were no
such thing as mixture of blood, then there would at
least be next to nothing of the shifting of speech.
Borrowing there would still be, but not substitution.

It is mixture of communities which creates the
great intricacy of the ethnological problem, on its linguistic
side as on its physical ; which renders it, in fact,
insoluble except approximately ; and which, so far as
the history of races is concerned, makes the linguist as
glad of the help of the physicist as vice versâ. The
ethnologist has to confess the same possibility which
was admitted on the part of the linguists at the end of
the preceding chapter. During the long past, there
have been indefinite encroachments, superpositions,
mixtures, displacements, destructions, among human
races (or derived branches of a unitary race), as among
human languages (or derived branches of the unitary
human language). In neither department is it likely
that the history will ever be unraveled with anything
approaching to completeness : especially, since the great
extension which the generally-admitted period of man's
existence on the earth has lately received. Opinions
are by no means as yet agreed upon this point ; but
even those who still refuse to accept the new doctrine
are preparing themselves to believe by-and-by, if the
272evidence to that effect shall turn out irresistible, that the
life of man has lasted for scores, if not for hundreds,
of thousands of years. This is a doctrine of the highest
interest to the ethnologist ; but it balks his hopes of
being able to trace more than a little way into the thick
darkness of early time the lines of race-history ; it gives
the precedence to anthropology as the science of man's
development as a whole race, or a congeries of undistinguishable
races, as yet not sufficiently differentiated
in their capacities and products to be held apart from
one another ; and to zoölogy as alone capable of answering
the question as to his origin.

The records of the earliest and rudest period of
man's activity are of two kinds : the products of their
art and industry, wrought by their hands ; and the
primitive materials and forms of their speech, wrought
for the uses of their minds ; the latter the instrument
of sociality, as the former of individual subsistence and
defense ; both turning, each in its own way and measure,
to the education and equipment of the higher
capacities of the race, and its advance toward self-control,
the control of Nature, and civilization. Both kinds
of record are eagerly sought and carefully scanned by
historical students, as evidences of a remoter past than
the pen of history or the voice of legend reports. But,
of the two, the linguistic remains are infinitely the more
important and instructive ; and it is almost they alone
which can serve the purpose of the ethnologist, since
the others are indicative rather of a grade of development
than of the special endowments or habits of a
race. The linguistic evidence has over even the physical
the advantage that it is far more abundant and
varied, and therefore manageable. The differences in
the kingdom of language are not like those which prevail
273within the limits of a single species of animals ;
they are equal, rather, in range to those which belong
to the whole animal kingdom. It is, to the other, like
a microscopic image thrown up by optical means upon
a wall, where its parts may be examined and measured
and described and compared by even the unskilled student.
Breadth of knowledge and competent judgment
are to be won in physical ethnology only by rare
opportunities, peculiar gifts, and prolonged training.
Though languages are traditional institutions, they are
of a special kind, capable of application to ethnological
purposes far beyond any other, as being so various and
so distinct as they are, capable of being looked at objectively,
and handled and compared with accuracy.
They are persistent, also, at least to a degree far beyond
other institutions.

To admit that a language can be exchanged, therefore,
is by no means to deny its value as a record of
human history, even of race-history ; it is only to put
that value upon its proper basis, and confess those limitations
which can in no manner be avoided, and of which
a due consideration is needful to the proper use of linguistic
evidence. It still remains true that, upon the
whole, language is determined by race, since each human
being usually learns to speak from his parents and others
of the same blood. And the marked exceptions to this
rule take place in the full light of historical record.
Civilization facilitates mixture, as it does communication.
It is not the wild and obscure races which are, or
have ever been, mixing blood and mixing or shifting
speech upon a grand scale ; it is the cultivated ones. If
one barbarous tribe overcomes another, unless the conquerors
absorb the conquered into their own community,
there is not usually a change of speech : but nations
274like the Romans and Arabs, who come with the force of
an organized polity and a literature, extend their speech
widely over strange peoples. Where the information
derivable from language, therefore, is most needed, there
it comes with the greatest presumption of accuracy.

Hence, when the ethnological relations of a community
or of a group of communities are to be settled, the
first question is as to the affinities of its speech. This
does not necessarily decide the case ; the linguistic evidence
may be overborne by some other ; but nothing
can be determined without it ; it lays the basis for further
discussion. We need only to quote an example or
two in illustration of this. The Basques are a white,
“Caucasian” race ; there is nothing in their other ethnological
characteristics which should forbid our connecting
them with any great division of the white race ;
but their speech at once cuts them off from every other,
and we accept its decision as authoritative. Out of what
mixtures the original Iberians may have grown, we cannot
tell ; nor can we ever absolutely know that the
Basques did not borrow their Euskarian dialect, as the
French their Romanic dialect ; there are indefinite possibilities
lying behind ; but the language tells us a great
deal, and probably all that will ever be within our reach.
Again, of the Etruscans there are records and descriptions
and pictures, and products, art and industrial ; but
to settle the relationship of the race the ethnologists
with one consent appeal to the infinitesimal remnants
of Etruscan speech : a single page of connected Etruscan
text, with but a hint of its meaning, would in the
briefest time settle the question whether the race is to
be connected with any other on earth, or whether, like
the Basque, it is an isolated fragment. There lies before
us a vast and complicated problem in the American
275races ; and here, again, it is their language that must
do by far the greatest part of the work in solving it.
American ethnology depends primarily and in bulk on
the classifications and connections of dialects ; till that
foundation is laid, all is uncertain ; although there are
points involved which may not yield even to the combination
of all attainable evidence, from every quarter.

We are to look for no real reconciliation between
the results won by the two great branches of ethnological
study until their methods are more fully established
than at present ; nor is it worth while to hurry the process — least
of all, to attempt prematurely an artificial
and superficial scheme of combination. All that will
come in good time, if we only have patience. Within
its own domain, each is supreme. The classifications
and relations of speech are what they are, without any
reference to underlying questions of race ; and yet,
those questions cannot be kept down and ignored by
the linguist : his study is too thoroughly a historical
one, it involves too much of the element of race in the
later periods, to allow of our leaving that element out
of account for the earlier. As one of the leading
branches of historical investigation, as claiming to make
its contribution to the elucidation of the past, it must
offer its results to be criticised by every other concurrent
branch. And to exaggerate its claims, or to put
them upon a false basis, is both needless and harmful.
If any one is not content with the degree of dignity and
authority that belongs to the science of language when
kept within the very strictest limits which a sound and
impartial criticism is impelled to draw, there are other
departments in which his aid will be welcomed, and he
had better turn to them.

There is one more point calling for brief notice in
276connection with, our classification of the dialects of the
world. That classification aimed at being a strictly
genetical one, each family embracing those tongues
which, by the sum of all available evidences, were
deemed traceable to a common ancestor. To the historical
philologist, still deep in the labor of determining
relations and tracing out the course of structural
development, this is by far the most important of all ;
indeed, the value of any other at present is so small as
to be hardly worthy of notice. The wider distinction
of languages as isolating, agglutinative, and inflective,
which has a degree of currency and familiarity, offers a
convenient, but far from exact or absolute, test by which
the character of linguistic structure may be tried ; the
three degrees lie in a certain line of progress, but, as in
all such cases, pass into one another. To lay any stress
upon this as a basis of classification is like making the
character of the hair or the color of the skin a basis of
classification, in physical ethnology, or the number of stamens
or the combination of leaves in botany : it ignores
and overrides other distinctions of an equal or of greater
importance. If the naturalist had the actual certainty
which the linguist has of the common descent of related
species, he would care little for any other classification,
but would spend his strength upon the elaboration and
perfection of this one. The linguist has enough of this
still left to do ; and till it is all accomplished, at any
rate, any other is of small account to him.277

Chapter XIV.
Nature and origin of language.

Language an acquisition, a part of culture. Its universality among men ;
limitation to man ; difference between human and other means of expression.
Communication the direct motive to the production of
speech ; this the conscious and determining element in all language history.
Natural cries as basis of the development ; question as to
their nature and range ; postulation of instinctive articulate utterances
uncalled for. Use of the voice as principal means of expression.
Imitative element in the beginnings of speech ; range and
limits of onomatopœic expression. The doctrine of roots. Sufficiency
of this view of the origin of language ; the opposing miraculous
theory. Capacity involved in language-making ; difference in
this respect between men and lower animals. Relation of language
to development of man ; rate and manner of its growth.

Our examination of the history of language, of its
mode of transmission, preservation, and alteration, has
shown us clearly enough, what we are to hold respecting
its nature. It is not a faculty, a capacity ; it is not an immediate
exertion of the thinking power ; it is a mediate
product and an instrumentality. To many, superficial
or prejudiced, inquirers this seems an unsatisfactory,
even a low, view ; but it is because they confound together
two very different senses of the word language.
Man possesses, as one of his most marked and distinctive
characteristics, a faculty or capacity of speech — or, more
278accurately, various faculties and capacities which lead
inevitably to the production of speech : but the faculties
are one thing, and their elaborated products are
another and very different one. So man has a capacity
for art, for the invention of instruments, for finding out
and applying the resources of mathematics, for many
other great and noble things ; but no man is born an
artist, an engineer, or a calculist, any more than he is
born a speaker. In regard to these various exercises of
our activities our condition is the same. In all alike, the
race has been undergoing almost from the beginning a
training of its capacities, step by step, each step being embodied
in a product. The growth of art implies a period
of rude shapings, and a rise to higher and higher production
by improving on former models and processes.
Mechanics still more clearly has the same history ; it
was by the use of ruder instruments, by the dexterity
acquired in that use and the consequent suggestion of
improvements, that men came finally to locomotives and
power-looms. Mathematics began with the apprehension
that one and one are two, and its development has
been like that of the others. And every new individual
of the race has to go through the same series of
steps, from the same humble beginnings. Only, he
takes them at lightning-speed, as compared with their
first elaboration ; because he is led onward by others
over a beaten and smoothed track. The half-grown boy
now is often a more advanced mathematician or mechanician
than the wisest of the Greeks : not because his
gifts are superior to theirs, but because he has only to
receive and assimilate what they and their successors
have wrought out for him. Though possessing the
endowments of a Homer or a Demosthenes, no man
can speak any language until he has learned it, as truly
279learned it as he learns the multiplication-table, or the
demonstrations of Euclid.

Now these collected products of the exercise of
man's developing powers, which, are passed on from
one generation to another, increasing and changing as
they go, we call institutions, constituents of our culture.
Something of them is possessed by every section of
humanity. There is no member of any community,
however barbarous, who is not raised vastly above what
he would otherwise be by learning what his fellows
have to teach him, acquiring their fragments of knowledge,
however scanty, and their arts — including the art
of speech. Doubtless the most degraded community
has more to teach the most gifted individual than he
would have learned, to the end of his life, by the use of
his own faculties unaided ; certainly this is so as regards
speech. Every one acquires that which the accident of
birth places within his reach, exercising his faculties
upon that foundation, expanded and at the same time
constrained by it, making to it his individual contribution,
if he have one to make : just as truly in the case
of language as of any other part. Language is in no
way to be separated from the rest : it is in some respects
very unlike them ; but so are they unlike one
another ; if it be the one most fundamentally important,
most highly characteristic, most obviously the
product and expression of reason, that is only a difference
of degree.

We regard every language, then, as an institution,
one of those which, in each community, make up its
culture. Like all the constituent elements of culture,
it is various in every community, even in the different
individuals composing each. There are communities in
which, it has come down within the strict limits of race ;
280in others it has been, partly or wholly, taken from
strange races ; for, like the rest, it is capable of being
transferred or shifted. Race-characteristics can only go
down by blood ; but race-acquisitions — language not
less than religion, or science — can be borrowed and
lent.

The universality of language, we may remark in
passing, is thus due to nothing more profound or mysterious
than that every division of the human race has
been long enough in existence for its language-capacities
to work themselves out to some manner of result.
Precisely so, there is a universal possession by men of
some body of instruments, to help the hands in providing
for human needs. This universality does not at all
prove that, if we could see coming into being a new
race, by whatever means brought the existing race into
being, we should find it within any definite assignable
period possessed of instruments — or of speech.

But, as things are, every community of men has a
common language, while none of the lower animals are
possessed of such ; their means of communication being
of so different a character that it has no right to be
called by the same name. No special obligation rests
upon the linguist to explain this difference, any more
than upon the historian of art or of mechanics to explain
why the lower animals are neither artists nor
machine-makers. It is enough for him to point out
that, the gifts of man being such as they are, he invariably
comes to the possession of this as well as of the
other elements of culture, while not one of the lower
races has shown itself capable of originating a civilization,
in any element, linguistic or other ; their utmost
capacity being that of being trained by the higher race
to the exercise of activities which in their own keeping
281had remained undeveloped, of being taught various arts
and acts, performed partly mechanically, partly with a
certain hardly determinable degree of intelligence. But
the subject is one upon which erroneous views are so
prevalent that we can hardly help giving it a brief consideration.

The essential difference, which separates man's
means of communication in kind as well as degree from
that of the other animals, is that, while the latter is
instinctive, the former is, in all its parts, arbitrary and
conventional. That this is so, the whole course of our
exposition has sufficiently shown. It is fully proved by
the single circumstance that for each object, or act, or
quality, there are as many names as there are languages
in the world, each answering as good a purpose as any
other, and capable of being substituted for another in
the usage of any individual. There is not in a known
language a single item which can be truly claimed to
exist φύσει, ‘by nature ;’ each stands in its accepted use
θέσει, ‘by an act of attribution,’ in which men's circumstances,
habits, preferences, will, are the determining
force. Even where the onomatopœic or imitative element
is most conspicuous — as in cuckoo and pewee, in
crack and whiz — there is no tie of necessity, but only
of convenience : if there were a necessity, it would extend
equally to other animals and other noises ; and
also to all tongues ; while in fact these conceptions have
elsewhere wholly other names. No man can become
possessed of any existing language without learning it ;
no animal (that we know of) has any expression which
he learns, which is not the direct gift of nature to him.
We are not less generously treated in this latter respect
than the animals ; we have also our “natural” expression,
in grimace, gesture, and tone ; and we make
282use of it : on the one hand, for communication where
the usual conventional means is made of no avail — as
between men of different tongue, or those who by deafness
are cut off from the use of speech — and, on the
other hand, for embellishing and explaining and enforcing
our ordinary language : where it is of a power and
value that no student of language can afford to overlook.
In the domain of feeling and persuasion, in all
that is intended to impress the personality of the communicator
upon the recipient, it possesses the highest
consequence. We say with literal truth that a look, a
tone, a gesture, is often more eloquent than elaborate
speech. Language is harmed for some uses by its conventionality.
Words of sympathy or affection can be
repeated parrot-like by one whose heartless tone takes
all value from them ; there is no persuasion in a discourse
which is given as if from a mere animated
speaking-machine. And herein comes clearly to light
the true sphere of natural expression ; it indicates feeling,
and feeling only. From the cry and groan and
laugh and smile up to the lightest variations of tone
and feature which the skilled elocutionist uses, it is
emotional, subjective. Not a tittle of evidence has
ever been brought forward to show that there is such
a thing as the natural expression of an intellectual
conception, of a judgment, of a cognition. It is
where expression quits its emotional natural basis, and
turns to intellectual uses, that the history of language
begins.

Nor is it less plain what inaugurates the conversion,
and becomes the main determining element in the
whole history of production of speech ; it is the desire
of communication. This turns the instinctive into the
intentional. As itself becomes more distinct and conscious,
283it lifts expression of all kinds above its natural
basis, and makes of it an instrumentality ; capable, as
such, of indefinite extension and improvement. He
who (as many do) leaves this force out of account, cannot
but make utter shipwreck of his whole linguistic
philosophy. Where the impulse to communication is
wanting, no speech comes into being. Here, again, the
parallelism between language and the other departments
of culture is close and instructive. The man growing
up in solitude would initiate no culture. He would
never come to a knowledge of any of the higher things
of which he was capable. It needs not only the inward
power, but also the outward occasion, to make man
what he is capable of becoming. This is characteristic
of his whole historical attitude. Races and generations
of men have passed away in barbarism and ignorance
who were as capable of civilization as the mass of the
present civilized communities : indeed, there are such
actually passing away around us. It is in no wise to
deny the grand endowments of human nature that we
ascribe the acquisition of speech to an external inducement.
We may illustrate the case by a comparison.
A stone has lain motionless for ages on the verge of a
precipice, and may lie there for ages longer ; all the
cosmic forces of gravity will not stir it. But a chance
thrust from some passing animal jostles it from its
equilibrium, and it goes crashing down. Which, shall
we say, caused the fall ? gravity, or the thrust ? Each,
in its way ; the great force would not have wrought
this particular effect but for the aid of the petty one ;
and there is nothing derogatory to the dignity of
gravitation in admitting the fact. Just so in language :
the great and wonderful powers of the human soul
would never move in this particular direction but for
284the added push given by the desire of communication ;
when this leads the way, all the rest follows.

Our recognition of the determining force of this
element is far from implying that communication is the
sole end, or the highest end, of speech. We have sufficiently
noticed, in the second chapter, the infinite value
of expression to the operations of each individual mind
and soul, and its fundamental value as an element in
the progress of the race. But it is here as elsewhere ;
men strive after that which is nearest and most obvious
to them, and attain thereby a vast deal more than they
foresaw. In the devising and constructing of instruments,
of all kinds, men have had directly in view only
what may be called the lower uses of them, their immediate
contributions to comfort and safety and sensuous
enjoyment ; but the result has been a calling-out of
many of the higher powers which could find appropriate
exercise in no other way, a reduction of Nature to service
in a manner that allows a part of the race to engage
in the more elevated and elevating occupations ; and a
discovery of truths in bewildering abundance. A yet
closer parallel is afforded by the closely kindred art of
writing, which adds to and enhances all the advantages
belonging to the art of speech, and is as indispensable
to the highest culture as is speech to the lower ; but,
like speech, it came into being by a process in which
the only conscious motive was communication ; all its
superior uses followed in the train of that, and were
unthought of until experience disclosed them ; indeed,
they are even yet unthought of by the greater part of
those who derive advantage from them. And this last
is true, to a degree which we must not fail to observe,
of spoken language also : its higher uses are not conscious
ones. Not one in a hundred, or a thousand, of
285those who speak realizes that he “uses language ;” but
there is no one who does not know well enough that he
can talk. That is to say, language, to the general apprehension
of its users, is simply a means of receiving
from others and giving to them : what it is to the individual
soul, what it is to the race, few have reach of
vision to see. And least of all is such penetration to be
credited to primitive man : he, especially, needs some
motive right before his eyes, and of which he can feel
every moment the impelling force ; and the desire to
communicate with his fellows is that motive, the sole
and the wholly sufficient one. He has no thoughts
swelling in his soul and demanding utterance ; he has
no foreboding of high capacities which only need education
to make him a little lower than the angels ; he
feels nothing but the nearest and most urgent needs.
If language broke out from within, driven by the
wants of the soul, it ought to come forth fastest and
most fully in the solitary ; since he, cut on : from other
means of improvement, is thrown back upon this as his
only resource : but the solitary man is as speechless as
the lower animals.

There might be ground for questioning this conclusion
as to the decisive value of the impulse to communication
in the initiation of language-history, if the
after-course of that history showed entire independence
of it. That is no acceptable scientific explanation
which calls in a special force at the beginning, like a
deus ex machina, to accomplish what we cannot see to
be otherwise feasible, and then to retire and act no
more. But communication is the leading determinative
force throughout. This it is for which and by which
we make our first acquisitions ; this leads us, when
circumstances change, to lay our old acquisitions aside
286and make new ; this determines the unity of a language,
and puts a restraint upon its dialectic variation ;
this is, both consciously and unconsciously, recognized
by every individual as the regulator : we speak so as to
be intelligible to others ; we hear and learn that we
may understand them ; we do not speak simply as we
ourselves choose, letting others understand us if they
can and will.

If this be so, then we have virtually solved, so far
as it admits of solution, the problem of the origin of
language ; we have ascertained what was the original
basis, and what the character of its development. The
basis was the natural cries of human beings, expressive
of their feelings, and capable of being understood as
such by their fellows. That is to say, the basis so far
as audible speech is concerned ; for it is not to be maintained
that this was the only, or even the principal,
means of primitive expression. Gesture and grimace
are every whit as natural and as immediately intelligible ;
and in the undeveloped condition of expression
every available means will unquestionably have been
resorted to, perhaps with a long predominance of the
visible over the audible. But it cannot be that the use
of the voice for expression should not have been suggested
and initiated by Nature's own endowments in
this direction.

Here, however, comes in a question respecting which
even the most recent opinions, and among those who in
general accept the view of language here taken, are
divided. How wide was this basis, and of what and
how definite character ? Did it consist of articulate
sounds instinctively attached to certain conceptions ?
Was there a limited natural vocabulary of actual words
or roots, of the same kind with later language, and
287needing only to be extended into the latter ? There
are those who would answer these questions in the affimative,
and who hold, therefore, that the fruitful way
to investigate concretely the problem of the origin of
language is to study the means of expression of the
lower animals, especially of those which stand nearest
to man, in order to find there something analogous with,
the roots of our speech. But this view has its basis in
the clinging impression, which many of those who reason
and write about language cannot possibly get rid
of, that there is somehow a real internal connection between
at least a part of our words and the ideas which
these represent — if one could only find out what it is.
If we recognize the truths that all existing human
speech is in every part and particle conventional, that
all of which there is record in the past was of the same
character, and that there is an utter absence of evidence
going to show that any uttered sound, any combination
of articulations, comes or ever came into existence as
the natural sign of an intellectual conception — we shall
be led to look with extreme disfavor upon any suggestion
of this kind. Beyond all question, it is wholly
uncalled for by necessity : the tones significant of feeling,
of which no one can deny the existence because
they are still an important part of our expression, are
fully capable of becoming the effective initiators of
language. Spoken language began, we may say, when
a cry of pain, formerly wrung out by real suffering, and
seen to be understood and sympathized with, was repeated
in imitation, no longer as a mere instinctive
utterance, but for the purpose of intimating to another,
“I am (was, shall be) suffering ;” when an angry growl,
formerly the direct expression of passion, was reproduced
to signify disapprobation and threatening ; and
288the like. This was enough to serve as foundation for
all that should be built upon it.

It is further to be considered, in judging this point,
that, as we approach man, the general capacities increase,
but the specific instincts, the already formed
and as it were educated capacities, decrease. It is
among the insects that we find those wonderful arts
which seem like the perfected results of training of a
limited intellect ; it is among birds that we find specific
modes of nest-building and a highly art-like, almost
artistic, song. Man is capable of acquiring everything,
but lie begins in. the actual possession of next to nothing.
Except suckling, he can hardly be said to be born
with an instinct. His long helpless infancy, while the
chicken and the calf run about and help themselves
from the very day of their birth, is characteristic of
Nature's whole mode of treatment of him. There is
no plausibility in the suggestion that he should have
begun social life with a naturally implanted capital of
the means of social communication — and any more in
the form of words than in that of gestures. It is a
blunder of our educated habit to regard the voice as
the specific instrument of expression ; it is only one of
several instruments. We might just as hopefully look
among the higher animals for the particular and definite
beginnings out of which our clothes, our buildings, our
instruments, are a development. In these departments
of human production, we see clearly enough what the
natural beginning should have been. No animal save
man is known to make any attempt at dressing ; but if
any did, it would amount to nothing ; for there are
tribes of men that go utterly, or almost utterly, naked ;
and no one, probably, would think of suggesting that
the rudiments of dress are not a turning to account, for
289perceived purposes of comfort or decency, just such
materials as Nature placed in man's way. The earliest
shelters were of the same sort : it would be of high interest
to find the animals nearest to man showing that
kind of capacity which he possesses, of putting to use
freely, simply as directed by circumstances, the varied
resources of Nature ; but probably the idea has never
come into any one's head that man, as an animal uneducated,
would be found building a particular style of
shelter (as the beaver its dam, the oriole its hanging
nest, the wasp its cells), out of which have grown, by
a process showing nowhere a saltus or lacuna, the
huts and palaces and temples of the more educated
races. And the same thing is true of instruments :
clubs and stones we allow to have been the first, only
because Nature offers such most conveniently within
reach of the beings who were gifted with mind enough
to see how they could be made available for perceived
needs.

Now it is only an unclear or a false view of the nature
of speech that prevents any from seeing that its
case is entirely analogous with these others, and that to
postulate, and then seek for traces of, a primitive basis
for language in the form of specific articulate signs for
ideas is an uncalled-for, even a necessarily vain and futile,
proceeding. It is, indeed, a matter of high interest,
and promising of valuable instruction, to investigate as
closely as possible the means of communication of the
lower animals, so as to determine its character and
scope ; but the point calling for special attention is,
how far the natural tones and utterances and postures
and movements are used secondarily and mediately, for
the purpose of signifying something, in rudimentary
correspondence with what we have seen to be the inferable
290beginnings of human language-making. We need
not be surprised to find, in more than one quarter, such
methods of communication in use, only limited, and,
for lack of the right kind and degree of capacity in
their users, incapable of development ; and these would
be the real analogues of speech, and would bridge the
saltus of which some are so afraid. If the Darwinian
theory is true, and man a development out of some
lower animal, it is at any rate conceded that the last
and nearest transition-forms have perished, perhaps exterminated
by him in the struggle for existence, as his
special rivals, during his prehistoric ages of wildness ;
if they could be restored, we should find the
transition-forms toward our speech to be, not at all a
minor provision of natural articulate signs, but an inferior
system of conventional signs, in tone, gesture,
and grimace.

As between the three natural means of expression
just mentioned, and constantly had in view by us in
this discussion, it is simply by a kind of process of
natural selection and survival of the fittest that the
voice has gained the upper hand, and come to be so
much, the most prominent that we give the name of
language (‘tonguiness’) to all expression. There is no
mysterious connection between the thinking apparatus
and the articulating apparatus, whereby the action that
forms a thought sets the tongue swinging to utter it.
Apart from the emotional (and non-articulate) natural
cries and tones, the muscles of the larynx and mouth
are no nearer to the soul than those of voluntary motion,
by which, among other things, gestures are produced.
Besides the lack of all evidence in language,
rightly understood, to indicate such connection, it is
sufficiently disproved, in a positive way, by the absence
291of vocal expression in the deaf, whose thinking and articulating
apparatus is all in normal order, but who, by
the numbing of the single nerve of audition, are removed
from the disturbing infection of conventional
speech ; it ought to be many times more instructive to
watch the “natural utterances” of a person thus affected
than to study the jabberings of monkeys. The analogy
between gesture and speech here is in the highest degree
instructive. The hands and arms are muscular
instruments under control of the same mind which produces
conceptions and judgments. Among their manifold
capacities, they are able to make gestures, of infinite
variety, all of which are reported by the vibrations
of the luminiferous ether to a certain apprehending organ,
the eye, both of the maker and of others. There
is a natural basis of instinctive gesture, which to the
human intellect is capable of suggesting a method of
intimation of intended meaning, developable into a
complete system of expression ; and it is so developed
for the use of those who by lack of power to hear are
cut off from the superior advantages of the other means
of expression. In the same manner, the larynx and
the parts which lie between it and the outer world are
muscular organs, movable by the same will which moves
the arms and hands. The parts have other offices to
perform besides that of shaping tone ; and the tone
which it is the sole office of the vocal chords to generate
is for other purposes as well as that of utterance : yet,
along with other things, they can produce an indefinite
variety of modified vibrations, reported through the
sympathetic vibrations of the air to another apprehending
organ, the ear, both of the producer and of others ;
and the sounds so reported are capable of combination
into groups practically infinite in number. There is a
292natural basis of tonic expression ; and on this and by
its suggestion human intelligence has worked out a great
number of diverse systems of expression, used, one or
other of them, by all ordinarily endowed men.

There is nothing here to require the admission of a
peculiar connection between thought and articulate utterance.
In a certain sense, it is true, the voice may
fairly be said to have been given us for the purpose of
speech ; but it is only as the hands have been given us
to write with ; our speaking organs do also our tasting,
breathing, eating. So iron has been given us to make
rails with for fast traveling : that is to say, among the
various substances provided in the world for man's various
uses, iron is the one best suited to this use ; its
qualities had only to be discovered by men, in the
course of their experience of Nature, and, when the
time for the use came, the perception of its adaptedness,
and the application, necessarily followed. In the course
of man's experience, it has come to light that the voice
is, on the whole, the most available means of communication,
for reasons which are not hard to understand :
it acts with least expenditure of effort ; it leaves the
hands, much more variously efficient and hard-worked
members, at leisure for other work at the same time ;
and it most easily compels attention from any direction.
Only the smallest part of its capacities are laid under
contribution for the uses of speech ; of the indefinite
number of distinguishable sounds which it can produce,
only a fraction, of twelve to fifty, are put to
use in any one language ; and there is nothing in the
selection to characterize a race, or to be used (except
in the same historical way as language in general) for
ethnological distinction : from among the many possibles,
these have chanced to be taken ; mainly the sounds
293easiest to make, and broadly distinguished from one another.

Under these determining considerations, vocal utterance
has become everywhere the leading means of expression,
and has so multiplied its resources that tone,
and still more gesture, has assumed the subordinate
office of aiding the effectiveness of what is uttered.
And the lower the intellectual condition of the speaker
and the spoken-to, the more indispensable is the addition
of tone and gesture. It belongs to the highest
development of speech that the word written and
read should have something like the same power as
the word spoken and heard ; that the personality of
the writer, even his frame of mind, should be felt,
and should move the sympathetic feeling of the reader.
And yet, it should also be noted here that, as we saw
in the twelfth chapter, there are languages (e. g. Chinese)
in which tone and inflection come to be used, in a
secondary and conventional way, to eke out the too
scanty resources of intellectual designation.

If we thus accept the impulse to communicate as
the governing principle of speech-development, and the
voice as the agent whose action we have especially to
trace, it will not be difficult to establish other points in
the earliest history. Whatever offered itself as the
most feasible means of arriving at mutual understanding
would be soonest turned to account. We have regarded
the reproduction, with intent to signify something,
of the natural tones and cries, as the positively
earliest speech ; but this would so immediately and certainly
come to be combined with imitative or onomatopoetic
utterances, that the distinction in time between
the two is rather theoretical than actual. Indeed, the
reproduction itself is in a certain way onomatopoetic :
294it imitates, so to speak, the cries of the human animal,
in order to intimate secondarily what those cries in their
primary use signified directly. Just as soon, at any
rate, as an inkling of the value of communication was
gained, and the process began to be performed a little
more consciously, the range of imitation would be extended.
This is a direct corollary to the principles laid
down above. Mutual intelligence being aimed at, and
audible utterance the means employed, audible sounds
will be the matter most readily represented and conveyed ;
just as something else would come easiest to
one who used a different means. To repeat once more
the old and well-worn, but telling, illustration : if we
had the conception of a dog to signify, and the instrumentality
were pictorial, we should draw the outline
figure of a dog ; if the means were gesture, we should
imitate some characteristic visible act of the animal —
for example, its bite, or the wagging of its tail ; if it
were voice, we should say “bow-wow.” This is the
simple explanation of the importance which is and must
be attributed to the onomatopoetic principle in the early
stages of language-making. We have no need of appealing
to any special tendency toward imitation. Man
is, to be sure, an imitative animal, as we may fairly say ;
but not in an instinctive or mechanical way ; he is imitative
because he has the capacity to notice and appreciate
what he sees, in other animals or in nature, and to
reproduce it in imitative show, if anything is to be
gained thereby — whether amusement, or artistic pleasure,
or communication. He is an imitator just as he is
an artist ; the latter is only the higher development of
the former.

The scope of the imitative principle is by no means
restricted to the sounds which occur in nature, although
295these are the most obvious and easiest subjects of significative
reproduction. What it is, may be seen in part
from the range of onomatopoetic words in known languages.
There is a figurative use of imitation, whereby
rapid, slow, abrupt, repetitive motions are capable of
being signified by combinations of sounds which make
something such an impression on the mind through
the ear as the motions in question do through the eye.
And we can well conceive that, while this was the chief
efficient suggestion of expression, men's minds may have
been sharpened to catch and incorporate analogies which
now escape our notice, because, having a plentiful provision
of expression from other sources, we no longer
have our attention keenly directed to them. Our judgments
on such points as this can only be partially trusted,
and must be tested with extreme caution, because
we are all of us now the creatures of educated habit, and
cannot look at things as men uneducated and with no
formed habits would do. We can safely investigate and
combine and speculate in this direction, if we keep fully
in mind the governing principle that mutual intelligence
is the end, and that whatever conduces to mutual
intelligence, and that alone, is the acceptable means.
We shall thus be saved from running off into, or toward,
that most absurd doctrine, the absolute natural significance
of articulate sounds, and the successful intimation
of complex ideas by a process of piecing these elements
together.

There are one or two further points connected with
this theory of the imitative origin of language which
call for a few words of explanation. In the first place,
it does not rest on a discovery of the signs of onomatopœia
as predominant in the early traceable stages of
language. Those stages are still too far from the beginning
296to furnish any such discovery. The intent was to
find means of mutual intelligence ; and when this was
won, the way it came was a matter of small consequence,
and might be left to be covered up. This has been, as
we abundantly saw above, a governing tendency in the
growth of speech down to the present time. Speakers
know not and care not whence their words came ; they
know simply what they mean ; even the wisest of us
can trace the history of only a small part of his vocabulary,
and only a little way. The very earliest dialects
are as exclusively conventional as the latest ; the savage
has no keener sense of etymological connection than the
man of higher civilization. Nothing has done so much
to discredit the imitative theory with sound and sober
linguistic scholars as the way in which some pass beyond
the bounds of true science in their attempts to
trace our living vocabularies to mimetic originals. The
theory does, indeed, rest in part on the undeniable presence
of a considerable onomatopœic element in later
speech, and on the fact that new material is actually
won in this way through the whole history of language ;
onomatopœia is thus raised to the rank of a vera causa,
attested by familiar fact ; and nothing that is not so
attested — for example, the assumed immediate intellectual
significance of articulate combinations — has the
right to stand as a causa at all ; but it rests also in
part, and in the main part, on the necessities of the case,
as inferred from the whole traceable history of speech
and its relation to thought, its use and its value.
Here is just the other support which it needs : no account
of the origin of language is scientific which does
not join directly on to the later history of language without
a break, being of one piece with that history.

But, in the second place, it may at first sight seem
297to some that there is a break in the history : for why do
we not still go on to make words abundantly by onomatopœia ?
A moment's thought will show the baselessness
of this objection. The office of onomatopœia was
the provision, by the easiest attainable method, of the
means of mutual intelligence ; in proportion, then, as it
became easier to make the same provision by another
method, the differentiation and new application of signs
already existing, the primitive method went into comparative
disuse — as it has ever since continued, though
never absolutely unused.

Once more, our theory furnishes the satisfactory
solution of a difficulty which has had influence with
some minds. Why should the germs of speech be what
we have called roots, elements indicative of such abstract
things as acts and qualities ? surely concrete
objects are soonest and most easily apprehended by
the mind. Without stopping to dispute on more philosophical
grounds this last assertion, claiming instead
that we apprehend only the concreted qualities and acts
of objects, it will be more to the point with those who
feel the difficulty to note that the process of speech is
one of signifying, and that only the separate qualities
of objects, at any rate, are capable of being signified.
To revert to our former example : there may be a state
of mind in which there should exist a confused concrete
impression of a dog, just sufficient to make it possible
to recognize another as agreeing with one already seen,
but without any distinct sense of its various attributes.
But so long as that is the case, no production of a sign
is possible : it is only when one has so clear a conception
of its form that he can signify it by a rude outline picture,
or of its characteristic acts that he can reproduce
the bite, or wag, or bark, in imitation of them, that he
298is ready for an act of language-making of which the dog
shall be the subject. And so with every other case ;
the first acts of comparing and abstracting must precede,
and the first signs must follow ; even as we have
before seen that it is through the whole history of
speech : the conception first, then the nomenclative
act. And bow-wow is a type, a normal example, of the
whole genus “root.” It is a sign, a hint, that calls
before the properly prepared mind a certain conception,
or set of related conceptions : the animal itself, the act,
the time and other circumstances of hearing it, and what
followed. It does not mean any one of these things
exclusively ; it comprehends them all. It is not a verb,
for that adds the idea of predication ; nor is it a name :
it may be put to use in either of these two senses.
What it comes nearest in itself to meaning is ‘the action
of barking’ — just that form of abstraction into which
we now most naturally and properly cast the sense of a
“root.” And so with both the other suggested signs.
Only, the outline figure has a decidedly more concrete
character than either of the others, and is in a certain
way their antithesis. It is a curious fact, and one tellingly
illustrative of how the character of the sign depends
on the instrumentality by which it is made, that
hieroglyphic systems of representation of thought (which
are in their origin independent systems, parallel with
speech, though they are wont finally to come into servitude
to speech) begin with the signs for concrete objects,
and arrive from these, and secondarily, at the designation
of acts and qualities. In Chinese, a combination
of the hieroglyphs of sun and moon makes the character
for ‘light’ and ‘shine ;’ in speech, on the contrary,
both luminaries are apt to be named from their shining
(see above, p. 83). In Egyptian, a picture of a pair
299of legs in motion means ‘walk ;’ while, with us, the
foot is so named as being the ‘walker.’

That by the methods thus described it was possible
to make a provision of signs capable of development, by
processes not different from those traceable in the historic
period of language, into such vocabularies as we
find actually existing, it does not seem as if any one
could reasonably deny. If this is true, and if the
methods are not only not inconsistent, but even in complete
harmony, with the whole traceable course of human
action on language, then we have found an acceptable
solution of that part of the problem we are seeking
to solve which is at present within our reach. A
scientific solution requires that we take man as he is,
with no other gifts than those we see him to possess,
but also with all those that constitute his endowment as
man, and examine whether and how he would possess
himself of the beginnings of speech, analogous with
those which our historical analysis shows to have been
the germs of the after-development, but beyond which
historical research will not carry us. As he would, if
need were, make the acquisition now, so may he, or
must he, have made it of old. This is not a part of
the historical science of language, but a corollary to it,
a subject for the anthropologist who is also a linguistic
scholar, who knows what language is to man, and how.
He is not prepared to deal with it who is merely master
of the facts of many languages.

Of course, a language thus produced would be a rude
and rudimentary means of expression. But that constitutes,
in the mind of the modern anthropologist, no
bar to the acceptance of the theory. If we deny to
primitive man the possession of the other elements of
civilization, and hold him to have gradually developed
300them out of scanty beginnings made by himself, then
there is no reason why we should not hold the same
view in respect to language, which is only such an element.
Even in existing languages the differences of
degree are great, as in existing states of culture in general.
An infinity of things can be said in English
which cannot be said in Fijian or Hottentot ; a vast
deal, doubtless, can be said in Fijian or Hottentot
which could not be said in the first human languages.
For what can be done in the way of distinct, even cultivated
and elaborate, expression, by only a few hundred
formless roots, we have a brilliant, almost a startling,
example in the Chinese. Of how sentences can
be made of roots alone, with the relations left to be supplied
by the intelligently apprehending mind, the same
tongue is a sufficient illustration. The Greek, or German,
or English, can elaborate a thought in a period
half a page long, determining by proper connectives
the relation of each of its clauses to the central idea,
and also, in widely varying degree and method, that of
the members of each clause to one another. This is a
capacity which belongs only to languages of high cultivation,
working on a richly inflective basis. Many
another tongue can form only simple clauses, possessing
no more intricate apparatus of connection than ‘ands’
and ‘buts,’ though having form enough in its words
to construct a clause of defined parts. Yet others lack
this definition of parts ; they strike only at the leading
ideas, presenting them in such order that the hearer
supplies the missing relations out of his general comprehension
of what must be the intended meaning. And
it is but another step backward to the primitive root-condition
of speech, where an utterance or two had to
do the duty of a whole clause. Men thus began, not
301with parts of speech which they afterward learned to
piece together into sentences, but with comprehensive
utterances in which the parts of speech lay as yet undeveloped,
sentences in the germ ; a single word signifying
a whole statement, as even yet sometimes with us :
only then from poverty, as now from economy. To
demand that “sentences,” in the present sense of that
term, with subject and predicate, with adjuncts and
modifiers, should have been the first speech, is precisely
analogous with demanding that the first human abodes
should have contained at least two stories and a cellar ;
or that the earliest garments should not have lacked
buttons and braces ; or that the first instruments should,
have had handles, and been put together with screws.
These conditions, in the last three cases, are at once
recognized as possible only to a miraculous endowment
of humanity, a gifting of man, at his birth, not with
capacities alone, but also with their elaborated results,
with the fruits of education ; and the assumption in regard
to language is really precisely the same, a proper
part of a miraculous theory of the origin of speech, but
of no other.

The word “miraculous,” rather than “divine,” is
here used to characterize the theory in question, because
it is the only truly descriptive one. One may
hold the views advocated in this chapter without any
detriment to his belief in the divine origin of language ;
since he may be persuaded that the capacities and tendencies
which lead man universally and inevitably to
the acquisition of speech were implanted in him by the
Creator for that end, and only work themselves out to
a foreseen and intended result. If language itself were
a gift, a faculty, a capacity, it might admit of being
regarded as the subject of direct bestowal ; being only
302a result, a historical result, to assert that it sprang into
developed being along with man is to assert a miracle ;
the doctrine has no right to make its appearance except
in company with a general miraculous account of the
beginnings of human existence. That view of the
nature of language which linguistic science establishes
takes entirely away the foundation on which the doctrine
of divine origin, in its form as once held, reposed.

The human capacity to which the production of
language is most directly due is, as has been seen, the
power of intelligently, and not by blind instinct alone,
adapting means to ends. This is by no means a unitary
capacity ; on the contrary, it is a highly composite and
intricate one. But it does not belong to the linguistic
student to unravel and explain, any more than to the
student of the history of civilization in its other departments ;
it falls, rather, to the student of the human
mind and its powers, to the psychologist. So also with,
all the mental capacities involved in language, the
psychic forces which underlie that practical faculty, and
which, being by it brought to conscious action, are
drawn out and trained and developed. The psychologist
has a work of highest interest and importance to
do, in analyzing and exhibiting this ultimate groundwork,
on which have grown up the great institutions
that make man what he is : language, society, the arts
of life, machinery, art, and so on ; and in tracing the
history of education of the human powers in connection
with them ; and his aid and criticism must be everywhere
of great value to their student. And this is most
of all the case with regard to language ; for language
is in an especial manner the incorporation and revelation
of the acts of the soul. Out of this relation has grown
the error of those who look upon linguistic science as a
303branch of psychology, would force it into a psychologic
mould and conduct it by psychologic methods : an error
which is so refuted by the whole view we have taken
of language and its history, that we do not need to
spend any more words upon it here. Language is
merely that product and instrumentality of the inner
powers which exhibits them most directly and most
fully in their various modes of action ; by which, so far
as the case admits, our inner consciousness is externized,
turned up to the light for ourselves and others to see
and study.

Out of the same close relation grows another and a
far grosser error, that of actually identifying speech
with thought and reason. This, too, we may take as
sufficiently refuted by our whole argument ; nothing
but the most imperfect comprehension of language can
account for a blunder so radical. The word reason, to
be sure, is used so loosely, in such a variety of senses,
that an unclear thinker and illogical arguer can comparatively
easily become confused by it ; but no one
who attempts to enlighten his fellow-men on this class
of subjects is excusable for such inability to grasp their
most fundamental principles. Language is, upon the
whole, the most conspicuous of the manifestations of
man's higher endowments, and the one of widest and
deepest influence on every other ; and the superiority
of man's endowments is vaguely known as reason — and
that is the whole ground of the assertion of identity.
There are many faculties which go to the production of
speech ; and they have other modes of manifestation
besides speech. And we have only to take the most
normally endowed human being and cut off artificially
the avenue of a single class of sensuous impressions,
those of hearing, and he will never have any speech.
304If speech, then, is reason, reason will have to be defined
as a function of the auditory nerve.

Whether, among the powers that contribute to the
production of language, there is one, or more than one,
not belonging in any degree to a single animal below
man, is a point which must be left to the psychologist
to decide. It may fairly be claimed, however, that
none such has yet been demonstrated ; and also, that
none such is necessary : a simple difference of degree in
the capacities common to both is amply sufficient to
account for the possession and the lack, on the one side
and the other. A heightened power of comparison, of
the general perception of resemblances and differences ;
an accompanying higher power of abstraction, or of
viewing the resemblances and differences as attributes,
characteristic of the objects compared ; and, above all
else, a heightened command of consciousness, a power
of looking upon one's self also as acting and feeling, of
studying one's own mental movements — these, it is believed,
are the directions in which the decisive superiority
is to be looked for. It is the height of injustice
to maintain that there is not an approach, and a very
marked approach, made by some of the lower animals
to the capacity of language. In the ratio of what we
call their “intelligence,” they are able distinctly and
fruitfully to associate conceptions with signs — signs,
namely, which we make for them, and by which we
guide and govern them. But, as an actual fact, their
capacity, though rising thus far, stops short of the
native production of such a sign, even of its acquisition
from the higher race and its independent use among
themselves. There is a long interval, incapable of
being crossed by the lower animals, between their endowments
and ours ; and he is a coward who, out of
305fear for the preservation of man's supremacy, attempts
to stretch it out, or to set up barriers upon it.

There is yet another important corollary from our
established view of language as a constituent element
of human civilization. Its production had nothing to
do, as a cause, with the development of man out of any
other and lower race. Its province was to raise man
from a savage state to the plane which he was capable
of reaching. The only development in which it was
concerned is the historical development of man's faculties.
Except, of course, that minor and limited change
which, falls within the sphere of ordinary heredity. The
descendant of a cultivated race is more cultivable than
the descendant of a wild one. The capacity of a yet
higher cultivation grows with the slow increase of cultivation ;
and if a people is suddenly brought in contact
with a civilization too far in advance of it, it is rather
deteriorated and wasted than elevated. The power of
brain, the capacity of thought, is enhanced by speech ;
but no such differences are produced as separate one
animal species from another. All men speak, each race
in accordance with its gift and culture ; but all together
are only one species. To the zoölogist, man was
what he is now when the first beginnings of speech
were made ; it is to the historian that he was infinitely
different. “Man could not become man except by
language ; but in order to possess language, lie needed
already to be man,” is one of those Orphic sayings
which, if taken for what they are meant to be, poetic
expressions whose apparently paradoxical character shall
compel attention and suggest thought and inquiry, are
admirable enough. To make them the foundation or
test of scientific views is simply ridiculous ; it is as if
one were to say : “A pig is not a pig without being
306fattened ; but in order to be fattened he must first be
a pig.” The trick of the aphorism in question lies in
its play upon the double sense of the word man
properly interpreted, it becomes an acceptable expression
of our own view : ‘Man could not rise from what
lie was by nature to what lie was able end intended to
become, and ought to become, except by the aid of
speech. ; but he could never have produced speech had
he not been at the outset gifted with just those powers
of which we still see him in possession, and which
make him man.’

We have already noted the linguist's inability at
present to form even any valuable conjectures as to
the precise point in the history of man at which the
germs of speech should have appeared, and the time
which they should have occupied in the successive steps
of their development. Men's views are greatly at variance
as to this, and with no prospect of reconciliation at
present, because there is no criterion by which they can
be tested. That the process was a slow one, all our
knowledge of the history of later speech gives us reason
to believe. As to the precise degree of slowness, that
is an unessential point, which we may well enough leave
for future knowledge to settle — if it can. What we
have to guard especially against is the tendency to look
upon language-making as a task in which men engage,
to which they direct their attention, which absorbs a
part of their nervous energy, so that they are thereby
prevented from working as effectively in other directions
of effort. Language-making is a mere incident of
social life and of cultural growth ; its every act is suggested
or called forth by an occasion which is by comparison
the engrossing thing, to which the nomenclative
act is wholly subordinate. It is as great an error to hold
307that at some period men are engaged in making and
laying up expressions for their own future use and that
of their descendants, as that, at another period, men are
packing away conceptions and judgments for which their
successors shall find expression. Each period provides
just what it has occasion for ; nothing more. A generation
or period may, indeed, by a successful incorporation
in speech of an exceptionally fertile distinction,
start a train of development which shall lead to immense
consequences in the future, and lay a foundation on
which a great deal shall admit of being built : such, for
example (as we thought to see above), was the early
Indo-European establishment of a special predicative
form, a verb. This is truly analogous with those fortunate
inventions or discoveries (like that of treating iron,
of domesticating useful animals) which appear now and
then to have given a happy turn to the history of a race,
initiating an upward career of growth which would
have seemed à priori equally within the reach of any
other race. Such occurrences we are in the habit of
calling accidental ; and properly enough,, if we are careful
to understand by this only that they are the product
of forces and circumstances so numerous and so
indeterminable that we cannot estimate them, and could
not have predicted their result. But, slower or more
rapid, the production of language is a continuous process ;
it varies in rate and kind with the circumstances
and habits of the speaking community ; but it never
ceases ; there was never a time when, it was more truly
going on than at present.

What term we shall apply to the process and its
result is a matter of very inferior consequence. Invention,
fabrication, devisal, production, generation — all
these are terms which have their favorers and also their
308violent opposers. Provided we understand what the
thing in reality is, we need care little about the phraseology
used in characterizing it. Each word may be
not unfitly compared to an invention ; it has its own
place, mode, and circumstances of devisal, its preparation
in the previous habits of speech, its influence in
determining the after-progress of speech-development ;
but every language in the gross is an institution, on
which scores or hundreds of generations and unnumbered
thousands of individual workers have labored.309

Chapter XV.
The science of language : conclusion.

Character of the study of language ; its analogies with the physical
sciences. Its historical methods ; etymology ; rules of its successful
pursuit. Comparative philology and linguistic science. History
of the scientific study of language.

What we have to observe here in conclusion with
regard to the study of language must be very brief, and
mainly in the way of more or less obvious corollary to
what has been already said. With any one who accepts
the views of language set forth above, the rest will follow
as a matter of course ; with one who does not, it is
too late here to argue.

Whether, in the first place, men be willing to allow
to the study the name of a science or not, is a matter of
the smallest moment. It has its own character, its own
sphere, its own importance of bearing on other departments
of knowledge. If there are those whose definition
of a science excludes it, let it be so ; the point is
one on which no student of language need insist.

What he does need to insist upon is that the character
of his department of study be not misrepresented, in
order to arrogate to it a kind and degree of consequence
to which it is not entitled — by declaring it, for example,
a physical or natural science, in these days when the
310physical sciences are filling men's minds with wonder at
their achievements, and almost presuming to claim the
title of science as belonging to themselves alone. It is
curiously indicative of the present as an early and formative
period in the history of this study, that there
should exist a difference of opinion among its conspicuous
followers as to whether it be a branch of physical or
of historical science. The difference may be now regarded
as pretty conclusively settled : certainly, it is
high time that any one who takes the wrong view be
read out of the ranks, as one who has the alphabet of
the science still to learn. No study into which the acts
and circumstances and habits of men enter, not only as
an important, but even as the predominant and determining
element, can possibly be otherwise than a historical
or moral science. Not one item of any existing
tongue is ever uttered except by the will of the utterer ;
not one is produced, not one that has been produced or
acquired is changed, except by causes residing in the
human will, consisting in human needs and preferences
and economies. There is no way of claiming a physical
character for the study of such phenomena except by a
thorough misapprehension of their nature, a perversion
of their analogies with, the facts of physical science.

These analogies are real and striking, and are often
fitly used as instructive illustrations. There is no
branch of historical study which is so like a physical
science as is linguistics, none which deals with such an
infinite multiplicity of separate facts, capable of being
observed, recorded, turned over, estimated in their various
relations. A combination of articulate sounds forming
a word is almost as objective an entity as a polyp or
a fossil ; it can be laid away on a sheet of paper, like a
plant in a herbarium, for future leisurely examination.
311Though a product of voluntary action, it is not an artificiality ;
what the producer consciously willed it to be is
but the smallest part of what we seek to discover in it :
we seek to read the circumstances which, unconsciously
to himself, guided his will, and made the act what it
was ; we regard it as a part of a system, as a link in a
historical series, as an indicator of capacity, of culture,
of ethnological connection. So a flint-chip, a scratched
outline of an animal, an ornament, is a product of intention ;
but it is also, as a historical record, pure of all
intention ; a fact as objectively trustworthy as is a fossil
bone or footmark. The material of archaeology is
even more physical than that of linguistics ; but no
one has ever thought of calling archaeology a physical
science.

As linguistics is a historical science, so its evidences
are historical, and its methods of proof of the same
character. There is no absolute demonstration about
it ; there is only probability, in the same varying degree
as elsewhere in historical inquiry. There are no rules
the strict application of which will lead to infallible
results. Nothing will make dispensable the wide gathering-in
of evidence, the careful sifting of it, so as to
determine what bears upon the case in hand and how
directly, the judicial balancing of apparently conflicting
testimony, the refraining from pushing conclusions beyond
what the evidences warrant, the willingness to
rest, when necessary, in a merely negative conclusion,
which should characterize the historical investigator in
all departments.

The whole process of linguistic research begins in
and depends upon etymology, the tracing out of the
histories of individual words and elements. From
words the investigation rises higher, to classes, to parts
312of speech, to whole languages. On accuracy in etymological
processes, then, depends the success of the whole ;
and the perfecting of the methods of etymologizing is
what especially distinguishes the new linguistic science
from the old. The old worked upon the same basis on
which the new now works : namely, on the tracing of
resemblances or analogies between words, in regard to
form and meaning. But the former was hopelessly
superficial. It was guided by surface likenesses, without
regard to the essential diversity which might underlie
them — as if the naturalist were to compare and class
together green leaves, green paper, green wings of insects,
and green laminæ of minerals ; it was heedless of
the sources whence its material came ; it did not, in
short, command its subject sufficiently to have a method.
A wider knowledge of facts, and a consequent better
comprehension of their relations, changed all this.
Especially, the separation of languages into families,
with their divisions and subdivisions, the recognition
of non-relationships and relationships and degrees of
relationship, effected the great revolution, by changing
the principles on which the probable value of particular
evidences is estimated. It was seen that, whereas a
close verbal resemblance between two nearly related
tongues has the balance of probabilities in its favor,
one between only distantly related tongues, or those
regarded as unrelated, has the probabilities against it ;
and hence, that, in order to be successful, comparative
investigation must be carried on with strict regard to
demonstrated affinities. While affinities are unsettled,
of course, all comparisons are tentative only, and may
be made in any direction, with due caution as to overestimate
of the results reached. But when a family
like the Indo-European is constituted, with its branches
313and sub-branches and dialects, all founded on the collection
and thorough examination of a vast body of evidence,
and by its side another like the Semitic and yet
another like the Scythian, then even cross-comparisons
between the branches are to be held in strict subordination
to the general comparison of branch with branch,
and cross-comparisons between families not less so : indeed,
they are not to be admitted at all, except as possible
evidences bearing on the question whether the
families are not, after all, ultimately akin — a question
which is ever theoretically an open one, but of which
the extreme difficulty has been sufficiently pointed out
in previous chapters. It is, at any rate, only when the
structure and material of the families shall have become
understood with equal thoroughness, by the bringing
to bear of all the evidences lying within the boundaries
of each, that apparent resemblances between them can
be deemed genuine, or used as signs of original connection.
It is not enough that such preparatory work
be done in one family ; all the subjects of comparison
must be reduced to the same value before they can be
treated as commensurable.

There are, in short, two fundamental rules, under
the government of which all comparative processes
must be carried on : 1. comparisons must have in view
the established lines of genetic connection ; and 2. the
comparer must be thoroughly and equally versed in the
materials of both sides of the comparison. For want of
regard to them, men are even yet filling volumes with
linguistic rubbish, drawing wide and worthless conclusions
from unsound and insufficient premises. On the
other hand, if they be duly heeded, there is no limit to
the scale on which the comparative process may be carried
on, and made fruitful of valuable results. We
314have already noticed that no fact in any language is
completely understood until there has been brought to
bear upon it the evidence of every other analogous fact,
related or unrelated ; and doubtless, to the end, so long
as any corner of the earth remains unransacked, some
of the views which we hold with confidence will be
liable to modification or overthrow.

The comparative method is really no more characteristic
of the study of language than of the other
branches of modern inquiry. But it was sufficiently
conspicuous in connection with the new start taken by
the study early in this century to make the name of
“comparative philology,” like the earlier “comparative
anatomy” and the later “comparative mythology,”
familiar and favored, for a time, beyond any other.
And the title is still accurate enough, as applied to that
aspect of the study in which it is engaged in collecting
and sifting its material, in order to determine correspondences
and relationships and penetrate the secrets
of structure and historic growth ; but it is insufficient
as applied to the whole study — the science of language,
or linguistic science, or glottology. Comparative philology
and linguistic science, we may say, are two sides
of the same study : the former deals primarily with the
individual facts of a certain body of languages, classifying
them, tracing out their relations, and arriving at the
conclusions they suggest ; the latter makes the laws and
general principles of speech its main subject, and uses
particular facts rather as illustrations. The one is the
working phase, the other the regulative and critical and
teaching phase of the science. The one is more important
as a part of special training, the other as an element
of general culture — if, indeed, it be proper to raise
any question as to their relative importance, even to
315the special student of language ; for the lack of either
will equally unfit him for doing the soundest and best
service.

Yet the two are certainly different enough to make
it possible that a scholar should excel in the one and not
in the other. The science of language runs out, on its
comparative side, into an infinity of details, like chemistry
or zoölogy ; and one may be extremely well versed
in the manipulation of its special processes while wholly
wrong as regards its grander generalizations : just as
one may be a skillful analyst while knowing little or
nothing of the philosophy of chemistry, or eminent in
the comparative anatomy of animals with no sound
knowledge or judgment as to the principles of biology.
To illustrate this, it would be easy to cite remarkable
examples of men of the present generation, enjoying
high distinction as comparative philologists, who, as
soon as they attempt to reason on the wider truths of
linguistic science, fall into incongruities and absurdities ;
or, in matters of minor consequence, they show in manifold
ways the lack of a sound and defensible basis of
general theoretical views. Comparative work of the
broadest scope and greatest value has long been done
and is still doing ; but the science of language is only
in the most recent period taking shape ; and its principles
are still subjects of great diversity of opinion and
of lively controversy. It is high time that this state of
things, tolerable only in the growing and shaping period
of a study, should come to an end, and that, as in other
sciences of observation and deduction — for example, in
chemistry, zoölogy, geology — there should be acknowledged
to exist a body, not of facts only, but of truths,
so well established that he who rejects them shall have
no claim to be considered a man of science.316

To review the history of the study is a task for
which we have no room remaining, and which may well
enough be left here unattempted ; it is a subject by itself,
and has been treated in independent works. 16 The
beginnings of the science lie as far back in the past as
the time when men first began to inquire and to speculate
concerning the facts which they observed in themselves
and in the world about them. The germs of all
the most important modern doctrines are to be found in
the reasonings of the Greek philosophers, for example ;
but unclearly apprehended, and mixed with much that
is erroneous. Their basis of knowledge was almost entirely
limited to the facts of their own language, and
hence insufficient for sound generalization. In the
great progress which has taken place during the last
century, resulting in the elaboration of a whole sisterhood
of new sciences, it was in the nature of things
impossible that linguistics should not come into being
with the rest ; and it came. The movement toward it
was well initiated in the last century, by the suggestive
and inciting deductions and speculations of men like
Leibnitz and Herder, by the wide assemblage of facts
and first classifications of language by the Russians
under Catherine and by Adelung and Vater and their
like, and by the introduction of the Sanskrit to the
knowledge of Europe, and the intimation of its connections
and importance, by Jones and Colebrooke. No
317one thing was so decisive of the rapid success of the
movement as this last ; the long-gathering facts at once
fell into their proper places, with clearly exhibited relations,
and on the basis of Indo-European philology was
built up the science of comparative philology. Frederick
Schlegel was a forerunner of the study ; more than
any other man, Francis Bopp was its leader. Parallel
with Bopp's great Comparative Grammar of Indo-European
tongues came forth Jacob Grimm's Comparative
Grammar of the Germanic branch of the family, each
in its own way a masterpiece, and both together raising
the historical study of language at once to the rank of
a science.

Almost all these names, it will be observed, are
German ; and, in truth, to Germany belongs nearly the
whole credit of the development of comparative philology ;
the contributions made to it from other countries
are of only subordinate value. In Germany, the
names of George Curtius, Pott, Benfey, Schleicher,
Kuhn, Leo Meyer, are perhaps the most conspicuous,
in the generation still mainly upon the stage ; but they
have so many fellows of nearly equal eminence that
it is almost invidious to begin specification and to stop
anywhere, without going on to include as many more.
Outside of Germany, Rask in Denmark, Burnouf in
France, and Ascoli in Italy, have most right to be mentioned
on the same page with the great German masters.

But while Germany is the home of comparative
philology, the scholars of that country have, as was
hinted above, distinguished themselves much less in
that which we have called the science of language.
There is among them (not less than elsewhere) such
discordance on points of fundamental importance, such
uncertainly of view, such carelessness of consistency,
318that a German science of language cannot be said yet
to have an existence. And, accustomed as the world is
to look to Germany for guidance in all matters pertaining
to this subject, until they shall come to something
like agreement it will hardly be possible to claim that
there exists a world's science of language. In the present
condition, however, of linguistic study on the one
side and of anthropology on the other, it cannot be that
the period of chaos will endure much longer ; if men
will begin with learning to understand those facts in
the life and growth of language which lie nearest to
them, they will surely be guided to consistent and sensible
views as to the past history, the origin, and the
nature of this most ancient and valuable of man's social
institutions.319