THE PHONETIC TRANSCRIPTION OF RUSSIAN SPEECH SOUNDS
For the linguistic investigation of the sound features of speech, phonetic transcription is indispensable because orthography usually expresses sound through complicated transmissions.
This is so in Russian, too, where the basic principle of orthography is morpho- phonematic and not phonetic. Thus e.g. the written form and pronunciation of the word полоток differ considerably. The letter о denotes a different vowel sound in each syllable : [ъ] — [л] — [о]. The root morpheme of the word голова has an unchanged orthographic form in the different inflected forms: голов-, while its pronunciation keeps changing: [гьллв — гллъв — гллоф — гъллв’].For the precise phonetic analysis of the sound string we need a faithful taking down of what we hear. Of course the accuracy of the notation has certain limits and it is unnecessary to note every sound feature which is not made use of in language or speech and which in any case belongs to the sphere of individual peculiarities. (Unless of course it is these which we want to examine.) The depth and accuracy of phonetic transcription is determined by orthoepic rules and by the demands of correct pronunciation in investigations of the standard language.
For the phonetic notation of Russian speech there are several transcription systems in use. We use two of them: a system based on the Cyrillic script, which is traditionally employed in the Russian phonetic literature, and a system based on the Latin script, which I have compiled from the notational system of the APhl. However, I have deemed it advisable to introduce certain changes and additions, since in languages other than Russian such collateral features as e.g. the stress- and position-dependent duration of sounds are not at all natural and self-evident. Therefore, for distinguishing three degrees of duration (stressed long, unstressed short and unstressed reduced) I use zero marking, superscript semicircle [“] and subscript semicircle [~]> respectively. This is all the more necessary because in Hungarian, stress and duration are linguistically independent of each other. An unstressed sound can be long, just as a stressed one can be short. (Cp. :[va:roj’ і —orsago j ]). Beside the Cyrillic symbol I also provide in the sound plates the Latin symbol, as well as the orthographic form of the word in which I examined the sound in question. For typographical and practical considerations I have made some changes in the symbols suggested by the APhl, too, but only in cases where they do not disturb uniform interpretation. For instance I use the apostrophe [’] instead of [•] or [,] for marking palatalization. In some cases—for lack of a suitable symbol—I use new symbols or symbol combinations, e.g.: [*o*], [u].