The question of the distinction between the æsthetic and the intellectual fact appears in Linguistic as that of the relations between Grammar and Logic. This problem has been solved in two partially true ways: the inseparability and the separability of Logic and Grammar. But the complete solution is this: if the logical form be inseparable from the grammatical (æsthetic), the grammatical is separable from the logical.

Grammatical kinds or parts of speech.

If we look at a picture which for instance portrays a man walking on a country road we may say: "This picture represents a fact of movement, which, if conceived as voluntary, is called action; and since every movement implies a material object, and every action a being that acts, this picture also represents a material object or being. But this movement takes place in a definite place, which is a piece of a definite heavenly body (the Earth), and precisely of a piece of it which is called terra-firma, and more precisely of a part of it that is wooded and covered with grass, which is called country, cut naturally or artificially into a form called road. Now, there is only one example of that star, which is called Earth: the earth is an individual. But terra-firma, country, road are genera or universals, because there are other terra-firmas, other countries, other roads." And it would be possible to continue for a while with similar considerations. By substituting a phrase for the picture that we have imagined, for example one to this effect: "Peter is walking on a country road," and by making the same remarks, we obtain the concepts of verb (motion or action), of noun (material object or agent), of proper noun, of common noun; and so on.

What have we done in both cases? Neither more nor less than submit to logical elaboration what first presented itself only æsthetically; that is to say, we have destroyed the æsthetic for the logical. But since in general Æsthetic error begins when we wish to return from the logical to the æsthetic and ask what is the expression of motion, action, matter, being, of the general, of the individual, etc.; so in the case of language, error begins when motion or action are called verb, being or matter, noun or substantive, and when linguistic categories, or parts of speech, are made of all these, noun and verb and so on. The theory of the parts of speech is really identical with that of artistic and literary kinds, already criticized in our Æsthetic.

It is false to say that the verb or noun is expressed in definite words, truly distinguishable from others. Expression is an indivisible whole. Noun and verb do not exist in it, but are abstractions made by us, destroying the sole linguistic reality, which is the sentence. This last is to be understood, not in the way common to grammars, but as an organism expressive of a complete meaning, which includes alike the simplest exclamation and a great poem. This sounds paradoxical, but is nevertheless the simplest truth.

And since in Æsthetic the artistic productions of certain peoples have been looked upon as imperfect, owing to the error above mentioned, because the supposed kinds have seemed not yet to have been discriminated, or to be in part wanting; so in Linguistic, the theory of the parts of speech has caused the analogous error of judging languages as formed and unformed, according to whether there appear in them or no some of those supposed parts of speech; for example, the verb.

The individuality of speech and the classification of languages.

Linguistic also discovered the irreducible individuality of the æsthetic fact, when it affirmed that the word is what is really spoken, and that two truly identical words do not exist. Thus were synonyms and homonyms destroyed, and thus was shown the impossibility of really translating one word into another, from so-called dialect into so-called language, or from the so-called mother-tongue into the so-called foreign tongue.

But the attempt to classify languages ill agrees with this just view. Languages have no reality beyond the propositions and complexes of propositions really written and pronounced by given peoples at definite periods; that is to say, they have no existence outside the works of art (whether little or great, oral or written, soon forgotten or long remembered, does not matter) in which they exist concretely. And what is the art of a given people but the whole of its artistic products? What is the character of an art (for example of Greek art or Provençal literature) but the whole physiognomy of those products? And how can such a question be answered, save by narrating in its particulars the history of the literature, that is to say, of the language in its actuality?

It may be thought that this argument, although possessing validity as against many of the usual classifications of languages, yet is without any as regards that queen of classifications, the historico-genealogical, that glory of comparative philology. And this it certainly is; but why? Precisely because that historico-genealogical method is not a mere classification. He who writes history does not classify, and the philologists themselves have hastened to say that languages which can be arranged in historical series (those whose series have hitherto been traced) are not distinct and separate species but a single whole of facts in the various phases of its development.