'Language,' I said many years ago in my Lectures1 on the Science of Language (1861), 'is our Language the Rubicon, and no brute will dare to cross it.'

Rubicon.

It might seem indeed as if in the concessions which I have made, I had myself done all that could well be done to help the brute across the river, by showing on different occasions how easy is the transition from sensuous vibrations to concepts, from shouts to roots. An attempt to minimise a difference is often supposed to arise from a wish to remove it. A clever pleader might say, 'Why, does not the very theory you propose on the origin of roots prove that Darwin is right? Have you not shown yourself that animals possess the materials of language in interjections; that they imitate the cries of other animals; that they communicate with each other, and give warnings by shrill cries; that they know their own names, and understand the commands of their masters? Have you not blessed us altogether by showing how in some cases at least interjections and imitations can be filed down, lose their sharp corners, become general, become, in fact, roots? Surely, after this, Darwin will be justified more than ever in saying that the language of man is the result of mere development, and that there must have been one or several generations of mere animals who had not yet generalised their intuitions and not yet filed down the sharp corners of their interjections?'

Few philosophers have studied animals so closely, and expressed their love and admiration Schopenhauer for them so strongly, as Schopenhauer. on animals. 'Those,' he says in one place, 'who deny understanding to the higher animals, can have very little

1 Vol. i. p. 403.

N

themselves.' 'It is true,' he says elsewhere, 'that animals cannot speak and laugh. But the dog, the only real friend of man, has something analogous-his own peculiar, expressive, good-natured, and thoroughly honest wagging of the tail. How far better is this natural greeting than the bows and scrapings and grinnings of men! How much does it surpass in sincerity, for the present at least, all other assurances of friendship and devotion! How could we endure the endless deceits, tricks, and frauds of men, if there were not dogs into whose honest faces we may look without mistrust!' The same philosopher assigns to animals both memory and imagination (Phantasie). He quotes the case of a puppy, unwilling to jump from a table, as a proof that the category of causality belongs to animals also. But with all this he is far too expert a philosopher to allow himself to be carried away by fanciful interpretations of doubtful appearances into ignoring the barrier which separates all animals from the animal homo. When explaining the formation of general concepts as the peculiar work of reason, he states without any hesitation or qualification, that it is this function of forming general concepts which explains all the facts that distinguish the life of men from the life of animals'.'

We have now to see in what this function of forming general concepts really consists, and if the conclusions which we arrived at in our first chapter be right, there is but one way which can lead us to a solution of this problem, namely to study the growth of thought in the history of language.

1

Schopenhauer, Welt als Wille und Vorstellung, 3 ed., i. 7. 46;,

ii. 72.

CHAPTER V.

THE CONSTITUENT ELEMENTS OF LANGUAGE.

I TRIED to prove in the first chapter that the whole of what we call the human mind is realised in language, and in language only. Our next task would have been to try to discover the constituent elements of language, and watch in their development the true historical development of the human mind. But before we could safely approach this task, it seemed necessary to remove a preliminary objection which arose from the theory of evolution, as interpreted by Darwin and some of his followers, namely, that as man was the descendant of an animal, the human mind could not differ in kind from the animal mind, and language therefore could only be a higher evolution of those sounds which animals utter, such as the roaring of lions, the barking of dogs, or the singing of birds.

Having shown, as I hope, in Chapters ii and iv, that the theory of evolution, as held by Darwin himself, does by no means necessitate the historical descent of the animal man from some other kind of animal, we now find ourselves free to undertake the analysis of language without any limitations as to the elements we ought to look for, and we therefore proceed to our task without any let or hindrance. If we should find that the ultimate elements in our

The Residua.

analysis of language turn out to be the cries of animals, more or less successfully imitated by men, or something like the sounds which men utter themselves when suffering from pain or joy or any joy or any other powerful emotion, this would prove a strong support of the opinion of certain followers of Darwin, stronger than any, I should venture to say, which has been produced by themselves. If on the contrary our analysis should lead us to a different result, we know at all events that our rear is safe, and we need no longer fear that the supposed descent of man from some other animal can again be appealed to as proving, without any further arguments, the utter futility of our researches. When we analyse any language and separate all that is merely formal in words, we always arrive in the end at certain residua, which resist further analysis. It matters little how we call these stubborn residua, whether roots, or elements, or phonetic types. What is important is the fact that, after we have removed the whole crust of historical growth, when we have broken up every compound, and separated every suffix, prefix, or infix, there remain certain simple substances which will yield to no solvent. This applies not only to the Aryan, but to the Semitic and Turanian languages likewise, nay, to every language which does not consist of roots only, such as the ancient Chinese. These simple substances being granted, we can understand the whole structure of language, we can make again the language which we have unmade. It was this simple process of etymological analysis and synthesis which I tried to represent in as clear a light as possible in my Lectures on the Science of Language, first published in 1861.

Roots ultimate

elements in the Science of

Language.

Those who have read these lectures will remember how strongly I opposed any attempt on the part of the students of language to go beyond roots, such as we actually find them at the end of a most careful analysis. It was thought at the time that my protests against all attempts either to go beyond roots or to ignore them as the types from which all words must be derived were too vehement. But I believe it is now generally admitted, even by some of my former opponents, that the slightest concession to what, not ironically, but simply descriptively, I called the Bow-wow and Pooh-pooh theories in the etymological analysis of words, would have been utter ruin to the character of the Science of Language. It is pleasant to find, as one grows older, how certain dangerous tendencies, which one had to oppose with all one's might, simply vanish and are seen no more.

Cells.

These roots of language have often been compared to cells, the last elements of all organic Roots and beings. Whatever differences of opinion there may be between different schools of physiologists, this one result seems permanently established, that the primary elements of all living organisms are the simple cells, so that the problem of creation has assumed a new form, and has become the problem of the origin and nature of these cells.

So far there is a certain similarity between the discoveries of physiologists and philologists. The most important result which has been obtained by a truly scientific study of language is this, that, after accounting for all that is purely formal as the result of juxtaposition, agglutination, and inflection, there remain in the end certain simple elements of human