You are on page 1of 22

The System of Russian Verb


Oppositional Properties of Russian Accentual Paradigms

I. Main goal: Provide lexemes with the markings AA, BB, CC, BC, CB
(total of 5 regular types).
The double letter designation indicates the stress type in the two
subparadigms that constitute a full paradigm (e.g. singular and
plural for nouns, non-past and past for verbs).
By definition, type A only occurs across two subparadigms and does
not normally combine with other types. (E.g. , .)

1. Type A stress (i.e. lexically marked stress) remains on

the same stem syllable throughout the paradigm.
2. In the case of stress types B and C, the
morphophonemic environment determines the stress of
the given lexeme and each type can occur either in both
subparadigms (BB and CC) or in only one of the two
subparadigms (BC and CB).

Type B admits either predesinential stress or stress

on the first desinential syllable in both
subparadigms of the noun and verb. I.e. stress
occurs immediately adjacent to the stemdesinence boundary (marked plus (+) on the
diagram to follow).
Type C admits word-initial stress or stress on the
first desinential syllable in both subparadigms of
the noun and verb. I.e. stress occurs on the first
syllable of either the stem or the ending. (In the
special and limited case of reflexives, stress
can be adjacent to the word-final boundary.)

Invariants of the distinctive Russian stress types

(applies to distinctive environments)

___ = stress can occur

...... = stress cannot occur
# = word boundary
+ = stem-desinence

II. Brief summary of stress rules for Russian nouns.

Some noun examples before getting to the verb.
Each type (A, B, C) has its own particular rules about how stress is
Type A: lexical and fixed. Simply marked on any stem syllable. This
rule applies to nouns and verbs alike.
Type B: based on the phonetic shape of genitive case endings.
(End-stress if genitive is non-zero in the subparadigm, but final
stem-stress if genitive is zero in the subparadigm. E.g., B stress of
singular / ; plural / .
Type C: a single deviating (mobile) form is possible in the
subparadigm if either a direct case (nom./acc.) or locative uses a
high vowel (i/u). E.g. /; /. If no such
case exists, the nominative form is correlated to stress in the
subparadigm (low-vowel nominative ~ end-stress, e.g. ;
non-low nominative ~ initial stress, e.g. ).

Sample Set of Nouns with Stress Types

III. Some major differences between Russian noun and verb stress.
1. The Russian noun is divided into singular and plural subparadigms. The verb
has the two basic subparadigms of non-past and past, but is more complex,
with other forms that can be derived from the basic subparadigms (past
passive participle, other participial forms, imperative infinitive.
2. As shown above, type B and C Russian noun stress is correlated to the
phonological shape of grammatical endings (desinences). (E.g. genitive
zero/non-zero, nominative low/non-low.)
3. Verb stress is predicted on the basis of three possible criteria: the
phonological shape of endings, as in the noun; the phonological shape of
the verb stem; and the presence or absence of a vocalic verbalizing suffix.
The following flow charts and examples will show how this works in both
the non-past (present-future) and past subparadigms.
4. One of the most important rules is that the presence of a verbal suffix in
the form prevents the B vs. C opposition; it doesnt matter if the suffix was
never there or was deleted in the Jakobsonian system.

Model of Russian Verb Stress

Comments on non-past (present-future) tense stress.

1. Verbs such as - are not assigned immobile type A stem stress,
due to the retraction of stress in the past passive participle, which does
not happen to type A verbs: e.g. , , etc. vs.
, .
2. In order to demonstrate initial vs. predesinential position of stress,
non-past forms are cited in the prefixed perfective form, e.g. .
3. The curious case of vs. . The stem-final aj- sequence
behaves both as a suffix and non-suffix in different forms. The B vs. C
opposition can occur due to the fact that znaj- is really an unsuffixed
root. Yet, we get the participial retraction (, ), which
treats the vowel a- as suffixal. Similar cases of ostensible B-C
opposition of suffixal forms occur when the vowel of a root is confused
with a similar phonetic sequence that is suffixal. Cf. vs. .

The forms , , are ambiguous and could be

desinential stress on , etc., or retracted from (-s).

Note that Ivanovas treatment of allows for 3 possible

stress types:

Note that Ivanova allows for the use of the non- type C
stress for , but that Eskova et al do not
recommend it.
Ivanova (2005):
Eskova et al (1988):

A book on Russian stress by Robert Lagerberg contained an interesting criticism

of my system, which I will illustrate and explain. Lagerberg raised the following
Why do I consider forms such as , to have initial stress, but forms
such as , to have predesinential, when they look the same?
These forms have the same stress on the surface, but they all have
monosyllabic stems, and such stems are accentually ambiguous, since you
cannot distinguish initial, medial, or predesinential position.
But, if you look at the equivalent polysyllabic morphological type in each given
subparadigm, you can draw conclusions about the ambiguous monosyllabic
I.e., is analogous to and called type C; likewise, the plural stress
of is analogous to , and is assigned to type B.
Generally, an ambiguous environment never should be used as evidence
without a supporting unambiguous environment.

Some other differences of this system in comparison to others:

1. There is no constant stress position implied by types B and C. Both types admit
stress on the first desinential syllable (end-stress), but differ in their use of
predesinential and initial stress.
For example, in the singular opposition of ~, the end-stressed type is
assigned to type B, since the opposed stress is initial and the invariant of type C.
Conversely, in the plural opposition of ~, end-stress is assigned
to type C, since the predesinential stress throughout the plural of is the
invariant of type B.
2. This system does not use a constant symbol (B or C) to represent
subparadigmatic mobile stress: if the mobility involves the word-initial syllable, it
belongs to type C (e.g. ~); in like manner, if the mobility uses
the predesinential syllable, it is assigned to type B (e.g. ~).
An example of the need for two types of mobility can be seen in the opposition of
type B ~ drink something after swallowing food, as contrasted to
type C ~ start drinking excessively. If all mobility had to be marked
as a single type C (as in many extant systems, such as that of Fedjanina), the
system could not accurately describe the different kinds of mobility just illustrated.

3. Some or all of the accentual oppositions (A, B, and C) can be suspended (or
neutralized) in specified morphophonemically environments. This can be
illustrated with the following two examples:
A. There is no B vs. C opposition when a syllabic verbalizing suffix is present in the
word, either absent in the basic form or deleted. In this case, the neutralized
form is realized as type B in the non-past and past subparadigms. E.g. the verbs
and differ in the non-past as type B vs. type C, where the
syllabic suffix is absent (and deleted, according to the Jakobsonian one-stem
system), but in the past tense they merge as suffix-stressed type B, due to the
presence of the verbal suffix i- in all forms. When a syllabic suffix is present in
the past passive participle, the opposition between types B and C is also
neutralized, but the form is realized as neither B nor C stress, but stress that is
retracted to one syllable preceding the suffix: e.g. , .
B. There is no opposition of stress types when the stem is non-syllabic. This occurs
in the non-past tense, where all stress types merge and are realized as type C endstress. E.g. the non-past forms of and merge as type C (, ),
but the past tense has a syllabic stem and opposes these lexemes as type A vs.
type C: e.g. , ; , .

4. It was noted above that the past passive participle merges stress types B and C
when a syllabic suffix is present. The realization does not match either that of B or
C, since it is pre-predesinential (e.g. , ). This allows us to
make the point that the type A, B, and C invariants only are valid in distinctive
environments. Neutralized environments can take on the accent of either of the
merged types, or a third type, such as in the past passive participle. In the latter
case, the stress also becomes a redundant grammatical marker.
5. Observe that when a masculine noun has the nominative plural ending a, and
non-zero stress in the genitive case, the resulting plural stress fits the definition of
both types B and C (e.g. , ). This means that in this restricted subset
of nouns, there is a neutralization of stress types B and C in the plural. Taken
together with the singular, the choice of stress for such words is either CC or CB.
Arbitrarily, the simpler type CC would seem to be indicated.