Чтение онлайн

This paper will not present new data on Slavic languages, but aims at highlighting some correlations between aspect and voice (particularly the passive voice), and at pointing out some leads for an analysis of their interaction. I shall attempt to clarify several points by analysing certain examples where context is important to the semantic interpretation:

1. the reflexive passive voice usually expresses a process (present, iterative or habitual according to context) or an event and encodes a potential/generic agent (explicit or implicit);

2. the periphrastic passive voice having a perfective participle usually expresses a resulting state with stative or dynamic meaning according to the context, whereas an imperfective participle expresses either a process or a complex state, represented as an open class of events.

The analysis of genuine corpora and a thorough study of the Slavic lexicon should help circumscribe the distribution of forms and the meanings encoded in such forms.

1. The structure to be + participle with — n/-t

Diachronically, it is well known that the passive past participle is based on the Indo-European adjective with *-to or *-no, which was originally attached to a root or to a nominal stem. Therefore, these forms were initially independent of the verbal system and served to indicate «a state resulting from the possession of the notion indicated by the noun or of the process expressed by the root» [Meillet 1965: 268]. Their integration into the verbal system is therefore an innovation in Indo-European languages. Concerning the earliest period of common Slavic, forms with *-to or *-no are no longer nominal derivatives but participles, by virtue of their integration in the verbal system. This is why constructions with an — n/ – tparticiple and the auxiliary «to be» first functioned as denoting a state, their actional meaning only appearing later [Maslov 1988: 77].

It is therefore unsurprising that in Slavic languages the interrelation of constructions with — n/-t is organised around a stative reference which, depending on several factors, may or may not imply a preceding event. In fact, certain constructions are conceived as purely stative for two raisons:

1. the participles may sporadically acquire this meaning because, diachronically, they are linked to a root or a nominal theme or because they are semantically removed from the verb (in this case, Khrakovskij [1991:151] speaks of lexicalisation):

2. the verbs are polysemic and the participles are used in their true sense (3a) or may take on psychological connotations (3b):

Only (3a) is open to discussion because, out of context, it may be analyzed either as an objective resultative, as defined by Nedjalkov and Jaxontov [Nedjalkov, Jaxontov 1988: 9] [18] , or as a passive form, because the perfective participle is bound to the base verb privar-zvam/privarza «to tie/attach» and the meaning of the construction is thus linked to transitivity and passivization (4a). But this perfective participle is mostly used in a psychological sense (3b) and it is thus semantically bound to the reflexive intransitive verb privarzvam se/privarza se and to (4b):

The literature provides many examples in which the constructions «to be» + — n/-t participles are used with a purely adjectival meaning. They then predicate a property of the entity in the same manner as an adjective:

Removed from any context, one may consider iznoseni «worn-out» and izpomackani «wrinkled» as resultative participles, since they may be associated with transitive verbs and allow the characterisation of the objects «coats» and «kepis» as affected and changed by a preceding event, but the coordination of the adjective tanki«light» leads to the elimination of the resultative interpretation in favour of an adjectival interpretation. Therefore (5) denotes a state. Syntactically, the construction is predicative for the two following reasons: 1. «to be» does not function as a voice auxiliary because it operates on the past passive participle as it operates on the coordinated adjective tanki«light»; 2. the past passive participle fills the syntactic function of an attribute, and, being a verbal adjective, cannot be treated as a participle oriented toward the patient of the basic transitive verb. More complex is the following Bulgarian example (quoted by [Barakova 1980: 141]):

If one admits that the form bjaxa obrani «were harvested» is resultative because it is formally derived from the verb obiram/obera «harvest, pick fruit» and because it implies a resulting state evidenced by the adverb vece «already», it would be difficult to explain the occurrence of the coordinated adjective pusti «empty». Just as in example (5), the participle is part of the paradigm of adjectives and the utterance denotes a state. As a result, it is impossible to give it either a corresponding active counterpart (6b) or to introduce an agent (6c):