3. The constantly used term “language” has a vague meaning when applied to everyday speech in all its variety. Thus, for example, the term “Russian language” refers to the entire complex of Russian dialects (e.g., Olonec, Rjazan’, or even the various Ukrainian dialects). The term “language,” however, must of necessity designate a finite, definite whole. Consequently, the term “Russian language” cannot apply to the speech of all speakers of Russian, as the latter presents an almost infinite variety. It is correct to speak of the “Russian language” only with respect to its literary language, a language which has been normalized and subjected to certain rules. The most suitable and objective term to be used with respect to any living speech community would be “linguistic area” or “linguistic territory.” Thus we could speak of a Russian lingustic area, Slavic linguistic area, etc. (pp. 8-10).*
. . . Since “grammar,” in the strict sense, deals only with one language, it would be appropriate to replace the term “comparative grammar” with the terms “comparative survey” or “comparative study” of a certain group of languages or language areas.
6. Comparative surveys, and comparison in general, of linguistic phenomena can be of three types:
I) Linguistic processes can be examined without regard to linguistic kinship, in order to establish the degree of similarity or difference between the structures of two languages. For example, Slavic forms exhibit side by side a k or a č in the root: peku pečeš’. This alternation is a clear case of coexistence which was determined by historical facts: the presence of k in peku was brought about by conditions favoring its retention, whereas č appeared at a certain moment under conditions which caused the “softening” (palatalization or centering of the tongue position) of k. Three stages can be singled out in this process:
. . . 1) There is a k which is independent of conditions, and a k which is palatalized under certain conditions.
2) k and palatalized (softened) k’ exist side-by-side.
3) k and č (from the earlier k’) exist side-by-side.
A similar coexistence of k and č can be found in almost all languages: in Romance, Germanic, Indie, Semitic, Finnic, etc. From this one may conclude that the alternation of k and č under certain conditions is a universal linguistic phenomenon.
II) Two or more linguistic areas may be compared without regard to their historical origin, if they show similar linguistic phenomena as a result of their territorial proximity. Thus, for example, Armenian, one of the Indo-European languages, can be compared with other Caucasian languages with which it shares phonetic and other peculiarities as a result of geographic contiguity. We observe a similar phenomenon on the Baltic shores: here too, some fundamental linguistic similarities between genetically unrelated languages, such as Latvian (an Aistian or “Baltic” language) and Estonian (a Finnic language) can be explained only by the geographic proximity of the peoples speaking these languages and who (judging from historical data) came into mutual contact and lived together as far back as the 13th century. As a result of this mixing, Latvian acquired several features peculiar to Estonian: stress on the first syllable (instead of a mobile stress), the distinction of long and short syllables (quantitas). The role of geographic proximity on languages of diverse origin is also apparent in the Balkan peninsula where, along with the descendants of ancient Greeks, we find Albanians (a tribe belonging to a separate Indo-European branch), Slavs, and Roumanians (of Romance origin). Bulgarian, Roumanian and Albanian show, for example, the same use of the postpositional article. The similarity of linguistic phenomena enables us to give a comparative survey of such linguistic areas.
The cultural influence of the literature of a representative of one or another linguistic area can also serve as a basis of comparative surveys. Good examples of this type are the Romance and other West European languages which underwent the influence of Latin literature, or the languages of the Russian area which experienced the impact of Church Slavonic.
III) Finally, there is the comparative study of linguistic areas which are assumed to stem from a common historical source and which can, therefore, be viewed as variants of one originally common state that has subsequently broken up into the very same variants that are being compared. This kind of comparison constitutes so-called comparative grammar in the strict sense of the word (§§ 14-19).
26. “Comparative grammar” or the comparative survey of a definite group of languages consists of the following main parts:
1) Phonetics, or the study of phonetic representations.
2) Morphology, in the broad sense, or the study of the concepts of a given language. Morphology encompasses two parts: (a) the study of concepts pertaining to the structure of words and phrases (morphology in the narrow sense of the word); and (b) the study of concepts pertaining to the structure of sentences, their parts and combinations (syntax).
3) Lexicology, or the study of words as parts of speech and of the various categories of words, such as the grammatical genders and verbal aspects.
1) Because of their common morphological origin we can compare identical morphemes occurring in different words of the same language, as for example, the radical pt- in (Russian) pt-ic-a and pt-aš-k-a.
2) Because of their assumed etymological or historical relationship, we can compare morphemes occurring in different languages, as, for example, the Russian boloto, the South Slavic blato, and the Polish błoto.
3) We can compare phonemes on the basis of their pronunciation. Take, for example, the representation of č: the Polish cz (č) is pronounced differently than the Russian č; in Russian it is mediopalatal, whereas in Polish it is apical. In “comparative grammars” such comparisons have almost no place; their main attention is focused on the kinship of phonemes that form a part of genetically related morphemes.
To shorten this exposition we should (only add that we shall) not compare individual sounds, but the various elements of sounds, their phonic components. Take, for example, the sound b, which combines the representations of several specific articulatory activities: labial closure, nasal closure, vibration of the vocal chords, etc. We can, in turn, consider the representations of the labial activity which characterize the pronunciation of this phoneme. In our comparative approach to phonetic problems we shall ask: does this activity of the lips occur in other languages? And if not, how does it differ? Etc. (pp. 94-96).
* (Page numbers refer to the Russian original.)