The following is my personal annotated translation of Wesen und Eintheilung der Sprachen (Essence and Classification of Languages), which is the first section of the Einleitung (Introduction) to the book Sprachen Europas in systematischer Übersicht (Europe’s Languages in Systematic Overview) (pp.5-10) by August Schleicher; published in 1850 by H.B. König, Bonn, Germany.

E s s e n c e
a n d
C l a s s i f i c a t i o n
o f
L a n g u a g e s

by August Schleicher

    All mental life, as far as it appears in the form of thought, needs language for its expression, for its actual coming into life, as does the mind the body. It’s only in a language that one can think.  The more the phonetically articulated expression of thought  - language - is able to represent all mental activities in phonetic form, the more perfect it is; the more sound stays back behind thought and yields abbreviations of it, as it were, the more imperfect the language will be. With all thinking, however, concepts,  [mental] representations are taken taken hold of in a certain [grammatical] relation; we ourselves can dissect thought into these two elements. Concepts, [mental] representations form as it were its substance; the [grammatical] relation - in which those [mental] representations are taken hold of - the form. However, in actual thinking both sides conceivably never turn up separated. A perfect language would have to reproduce both phonetically exact; imperfect ones make do with the more or less clear denotation of the [grammatical] relationship in which concepts and [mental] representations are taken hold of. However, this [grammatical]  relationship itself is impossible to leave unexpressed; without its phonetic expression a language is not thinkable. As far as they are thought to be phonetically represented, one is given to calling these mental representations and concepts [lexical] meaning. The essence of language is thus based on the way its [lexical] meaning and [grammatical] relation is expressed phonetically. There isn’t a third element to language; language comes out even with [lexical] meaning and [grammatical] relation. The phonetic expression of [lexical] meaning is the root, and it is fit to be represented in languages purely, i.e, to be kept separate from all grammatical morphemes [Beziehungslauten, literally “relation-sounds”]. [Grammatical] relation and [lexical] meaning together yield the word; the shaping of the word, word formation depends on the phonetic expression of both - and from here, in turn, the structure of the sentence and the entire character of the language which, indeed, is made up of words. It is only the phonetic expression of [grammatical] relation that enables the root to appear in a particular way - as adjective, noun, verb, so on; in a particular case, tense, mood, so on. ‘Word formation’ we take here in the broadest sense, not just in the sense familiar from grammars, of stem and theme formation [i.e, derivational morphology]; declination and conjugation [i.e, inflectional morphology] belong along with them, in this sense, to word formation; for they, too, are based on the tying together and phonetic expression of [lexical] meaning and [grammatical] relation. [In other words, by ‘word formation’ is meant what today we would call ‘morphology.’]

    Now, it is possible for only the [lexical] meaning to be expressed and the [grammatical] relation to be totally passed over phonetically. The [grammatical] relation itself is never missing: even though it may go unexpressed phonetically, the lexical morpheme [Bedeutungslaut, literally “meaning-sound”] must then be placed in a particular relation chiefly through its position in the sentence, through emphasis in speaking, through gestures, etc. And indeed there are languages of which it is a principle not to express [grammatical] relation phonetically. To this belong the monosyllabic languages, above all Chinese. A monosyllabic language is made up of only roots, of lexical morphemes [Bedeutungslauten] which contain the implicit [grammatical] relation in the abstract; the word-categories are not differentiated phonetically; the same sound can be a noun, verb, particle, nominative, genitive, present, preterit, subjunctive, indicative, so on; it’s chiefly the position in the sentence that indicates the current [grammatical] relation. These languages are monosyllabic since the roots (lexical morphemes [Bedeutungslauten]) of perhaps all languages, at least of all accurately known, are monosyllabic: also, the unity [Einheit, literally “one-ness”] of the concept, the [mental] representation is reflected in a unit [Einheit] of sound (a syllable). The word here is still by no means in parts; it is still an undifferentiated strict unit [Einheit], like crystal of Nature’s mineral kingdom. These languages which express phonetically only the [lexical] meaning, but not the [grammatical] relation, form the first class of languages, perhaps most fittingly termed monosyllabic languages. [We could more accurately term these languages isolative monosyllabic languages, ‘isolating’ meaning, basically, no morphology. Conceivably, or at least logically, a language could be monosyllabic and yet not isolating by having monosyllabic lexemes that go through processes of non-concatenative morphology, just like the word sing~sang~sung~song in English.]

    However, it is quite possible that between the strict application of this principle and the closer regulation of a morpheme [Laut, literally “sound”] through another placed nearby - the latter expressing the [grammatical] relation - there is but an imperceptible transition taking place. Proving to be excellently suitable for this way of denoting [grammatical] relation are those morphemes [Laute, plural of Laut] which have a general meaning (i.e, man, woman to denote gender) or the so-called ‘grammatical roots’ [Beziehungswurzeln, literally “relation-roots”], pronouns, i.e, roots which either already had, originally, a very general meaning, or else received such a one later on.

    Such assemblages are already to be found in the monosyllabic [and simultaneously isolating] languages; they are rampant; thus, there comes into being a totally different linguistic character set in total opposition to the previous. That is to say, the [grammatical] relation is expressed through morphemes [Laute] that are loosely appended to the unaltered lexical morpheme [Bedeutungslaute, dative singular of Bedeutungslaut]; thus, precisely that which characterizes the first linguistic class - the [grammatical] relation being implicitly contained in the lexical morpheme [Bedeutungslaute] but not phonetically expressed - is switched over into its opposite: the [grammatical] relation has attained a very manifest existence. These appended grammatical morphemes [Beziehungslaute] were originally lexical morphemes [Bedeutungslaute]; but most are to be found now only as grammatical morphemes [Beziehungslaute] and often distorted beyond recognition. [This is a process known today as grammaticalization.] This second class, that of the agglutinating (‘gluing together’) languages, comes in many gradations according to how more or less narrowly tied together the appended morphemes [Laute] are to the roots or among themselves; often these morphemes [Laute] are still written in a special way and perhaps even are felt to be separate words; often, additionally, the fusion is so intimate that it touches upon the third linguistic class. Just like all intermediate stages in nature, this linguistic class, too, is represented in extraordinarily great numbers; to it belong by far the most languages. [Oh?] In all of these languages, the word is divided in parts (compare to the first linguistic class); but these parts are not tightly fused into one totality (compare to the following class), but, rather, the word is still a combination of several individual words - “the one whole individual is more only the base (for more individual words), than a subjective unity [Einheit, literally “one-ness”] of the members” as in the natural organism of plants. Later on, we will have an opportunity to consider a unique sort of agglutination as a sub-category of this linguistic class.

    In the first class we found strict word-unity, but missed the phonetic denotation of the [grammatical] relation; in the second class the [grammatical] relation - and, at that, often a most highly complicated [grammatical] relation - is denoted through appended morphemes [Laute] quite distinctly, but at the cost of word-unity; only a third is still left: [lexical] meaning and [grammatical] relation receive their phonetic expression and the unity [Einheit] of the word is spared nonetheless. This stage is the highest one; it projects the most faithful picture of the mental process, of thought, in which, indeed, [lexical] meaning and [grammatical] relation penetrate themselves intimately. [Is this not ethnocentrism in the sense of the traditional bias of the West for the “Classical” languages, i.e, those of the Romans and Ancient Greeks, which were of this third linguistic class?] We encountered in the first linguistic class the undifferentiated identity of [lexical] meaning and [grammatical] relation;  with an undiluted view of [grammatical] relation, the second differentiated between grammatical and lexical morphemes, the [grammatical] relation having stepped out into a separate phonetic existence of its own; then that discrepancy closes back up in unity in the third linguistic class, but an infinitely higher unity because - having arisen out of this discrepancy - is based on it and subsumes it as dissolved within itself. The flexional languages thereby stand the highest on the scale of languages: it is first here in the organism of the word that a veritable anatomy developed; the word is the unity [Einheit] in the manifoldness of its members, corresponding to the animal organism, of which this same set-up applies. Additionally, this linguistic class is combined with the previous one through transitional forms; that is to say, it is not infrequently that essentially agglutinating languages present isolated flectional-type phenomena. There will be given a subcategory for this class, too.

    In the system of which we have just given a general outline are to be found all languages  - known up till now or yet unknown - for in that threefold relationship in which the phonetic expressions of [grammatical] relation and [lexical] meaning have the capacity to interact, are all possibilities exhausted. [But what about polysynthetic/incorporating as a fourth?] It is clear that to a certain extent the two last classes stand closer to one another than to the first, which totally lacks one of the two factors of language formation.