Last Chapter on Conjugation

February 24, 2010

Russian Conjugation -- Finale

I’ve omitted quite a few details, all of which in the long run are important one way or another, but this has gone on long enough. We need a demonstration of what all this does and what its descriptive power is like.

Let’s take some of the base forms for verbs I discussed at the beginning of this discussion. The morphophemic stem p’ok is a velar stem, so it will have substitutive softening in the present endings 2nd sg to 2nd pl. The vocalic endings will have round vowels u and o, according to the rule of acuteness attraction, since the stem is itself unstressed, so that all endings bear a stress. So here is the present tense: p’okU, p’očOš, p’očOt, p’očOm, p’očOt’e, p’okUt. In Cyrillic this is пеку печёшь печёт печём печёте пекут. (The phonemic transcription will reveal the reductions in the vowels: p’iku p’ičoš.) The infinitive will be печь, as velar stems have infinitives in чь and o is not permitted between two soft or palatal consonants. The past tense will be, phonemically, p’ok, p’ikla, p’ikl’i, with end-stress everywhere, as the velar stems have, and with the masc ending l suppressed. The imperative, also stressed, will have i, causing bare softening: пеки.

This shows us that in producing the forms of the present, infinitive, and past tenses the stem category must be specified, since it spells out lexically within its entry the movement of stress and the interplay of consonants and vowels at morpheme boundaries. These interplays may be strongly conditioned by the category of the verb, namely, the type of the last C and the stress pattern.

The general rules, for example, V(1) plus V (2) > V(2), C(1) plus C(2) > C(2), apply across all stem categories. To see how widespread this is, let’s look at the verbs ending in j (spelled й, or as a soft vowel letter after a vowel). These are suffixed stems, unlike p’ok, and embrace a variety of derived stems, like спрашивать, опазывать, советовать, делать, уметь. Note that, of course, there is no jot (j) in any of these infinitives. Why? Because the j, seen as a C, is deleted before a C ending (an ending beginning in C), and so jot disappears before infinitive and past, but is maintained in the present and imperative with their V endings.

So: делаю делаешь делает делаем делаете делают. Note that to the learner the endings seem to be –ю, –ешь, –ет, еtc., which are analyzed into j-u, j-iš, etc. here. This analysis is hard to teach because the very Cyrillic spelling seems to belie it.

In Jakobson’s analysis, the endings are not ю, ешь. ет, etc., but rather u, iš, it, im it’i, ut, in phonemic transcription with reduced vowels. The phonemic spelling is crucial, because the rule of intensity attraction predicts unstressed endings will begin with a high (diffuse) vowel: u, i. And so it is. The morphophonemic base stem is d’él-aj—, showing the interior root and stem, with stem-stress througout. This predicts all of the forms of the verb. The jod is everywhere and in every case deleted before any C.

Most handbooks give the endings as the naive student sees them, and conveniently contrast them with stressed endings spelled like иду идёшь идёт идём идёте идут. The problem with jod remains, however. What was it doing in the endings of делать and then disappearing in идти? When you look at these stems in the wider context of Russian verbs, you have to concede that the jod is part of the verb stem that appears only before V, just like the n in denu, the v in živu.

In verbs like давать даЮ даёшь даёт, where the vowel of the ending bears stress (this is the so-called deep truncation before va), the vowel changes to o, as a stressed V in a hard stem (rule of acuteness attraction).

It is interesting that in Russian spelling we see e for phonemic i, and ё for phonemic o. The o developed from stressed e in Old Russian. Nonetheless Jakobson points out that the vowel of плАчешь ‘you weep’ and and that of вИдишь ‘you see’ are phonemically the same. In a more abstract analysis, you might want to derive the former from an underlying o, and the latter from an i. This, however, goes outside of Jakobson’s model, which in itself captures an important generalization.
Some more examples, to see what generalizations they conceal.

Жить живу живёшь живут, жил, жилА, жили. This is a resonant stem in v that behaves very much like jot stems, also resonants. The v is lost before the C-endings of infinitive and past. Endings of the present are stressed, and so bear the rounded vowels if not soft (rule of acuteness attraction). The morphophonemic base živ lacks a stress, so that the present endings are stressed, but in the past, there is mobile stress — only the feminine ending bears the stress, otherwise it stays on the stem.


Compare the stress of unstressed suffixed stems, where the mobility appears in the present, but the past has stressed fixed on the theme vowel. There is only one suffixed stems that has past mobility: родитьс(ся) родился, родилась 'to be born'. These may be very important verbs, e.g. l’ubi, kup’i, p’isa, poluči. Example of mobility: куплЮ кУпишь кУпит кУпим кУпите кУпят. купил, past tense купила, купили. The unstressed endings show a diffuse (high) vowel u, i (rule of intensity attraction — weaker position is unstressed, so weaker vowel (diffuse).

This neatly described the Russian of Jakobson’s Moscow generation, where people said видют, учут though they wrote видят, учат. Now this is archaic and the rule is less attractive today. But still serviceable!
Looking briefly at the consonantal unsuffixed stems, we see an unproductive but important group. There are eight C’s, after the resonants j n m l r, that verbs may end with, and they are the very symmetrical group p b t d s z k g. Four voiced, four voiceless, two dental fricatives, four stops, two labial, two dental, two velar. They share an important trait: they maintain the group C(1) -C (2) at least some of the time. They are the exception that proves the rule!

Those in labials and dental plosives change those stops to s before the infinitive ending -ти. Dental stop stems lose their stop only before the past tense: в’ед– > вести вёл велА велИ. Velars lost the stop only in the infinitive (печь). Dental fricatives preserve their final consonant everywhere: несу несёшь несут, нести, нёс, несла, несло, несли, нёсший, неси.

One of my favorite C-stems is the velar žg– ‘burn’, present žgu, žžoš, žžot (жгу жжёшь жжёт), where the long жж is pronounced soft (palatalized), as it is both across morpheme boundaries and in roots, as in извозчик ‘coachman, driver’, вожжи ‘reins’, дрожжи ‘yeast’. Past tense of the verb is жёг (the masculine past drops the ending l, but keeps the velar; a fill vowel is necessary so the word can be pronounced) жгла жгли жгли, infinitive (с)жечь like печь, and past passive participle (со)жжён with that beautiful palatal.
That’s all for this topic,
gmc