The subjunctive in Polish

Basing on M. Achard's (1998) analysis of complementation in French and incorporating T. Givon's (1994) insights into the nature of the subjunctive, the paper aims at developing a cognitive analysis of the subjunctive clauses in Polish. It is claimed that what distinguishes the subjunctives from the finite complements is that, in contrast to the latter, the complement process of a subjunctive is not located with respect to elaborated reality (hence it is not grounded), but it represents an arbitrary instance of a process type in the sense of Langacker (1991). Thus whereas the complement process of a clauses such as Marysia powiedziala S[ze Jan przyszedl] is grounded and has a putative address in reality, the complement process of the subjunctive expressing wishes and desires of the speaker such as Marysia chcialaby [zeby Jan przyszedl] is situated in its own local space set up by the subjunctive inflection. This local space is provided by the basic epistemic model (of reality), and to be precise-- by its component called irreality (Langacker 1991). Events such as wishes and desires expressed by the subjunctive should be distinguished from propositions expressed by the indicative: propositions are grounded in reality, whereas wishes and desires are not grounded at all. The grounded/ungrounded contrast in the case of the subjunctive/indicative distinction relates to another contrast--namely, that between the finite and the infinitive clause. Finite (indicative) clauses are grounded, while infinitivals are not grounded (cf. Langacker 1991).

This, in turn relates to Achard'sanalysis of the two types of complements, couched by him in terms ofthe Optimal Viewing Arrangement (OVA) (in which the distance between the conceptualising subject of the main clause and the conceptualised scene is the largest) and the the Ecocentric Viewing Arrangement (EVA) (where the conceptualising subject of the main is on stage, i.e. it is part of the conceptualised scene of the embedded process). The finite clauses represent the OVA, while the infinitival clauses conform to the EVA configuration. Whereas the OVA/EVA contrast in the case of the finite-infinitive distinction holds along the so-called objective axis, the contrast between the finite/infinitive clauses and modal constructions is established along the so-called subjective axis. In the latter case the asymmetry holds between the speaker/conceptualiser (not the conceptualising subject) and the process expressed by the clause which is followed by the modal verb.

Now, if we assume, following the Polish linguistic tradition, contrary to what was claimed in Kardela (2000), that so-called infinitival clauses in Polish, especially those of the type Marysia chce pic (although perhaps not Marysia wie, co robia), are in fact modal constructions (see, for example Polanski (ed.) Encyklopedia jezykoznawstwa Ogolnego, Wroclaw 1993), then, analogously to modals, we will have here the asymmetry between the scene of the infinitival process established along the subjective axis as well. It is along the subjective axis, we wish to claim, that the two foci of the subjunctive in the sense of Givon (1994) are established: the epistemic and the deontic focus of the subjunctive. In Givon's theory these two foci form a scale based on the strength of manipulation exerted by the subject/agent of the main clause on the on the complement clause. Because, as demonstrated by Achard, and assumed in Kardela, manipulation translates easily into the OVA/EVA configuration (in the case of EVA the manipulation is stronger), we have a "weaker" case of the subjunctive which is centred around the epistemic focus (the speake/scene asymmetry is the largest) and the "stronger" cases of the subjunctive clustered around the deontic focus. The epistemic / deontic scale for Polish can be envisagesd as follows (see Givo 1994: 280) for a similar scale for English verbs):

deontic side
-----
weak intended manipulation: kazac, sugerowac, prosic, etc.
preference: chciec, zyczyc sobie, wolec, oczekiwac, etc.
epistemic anciety: miec nadziej, obawiac sie, etc.
Low epistemic certainty: nie byc pewnym, watpic, etc.

-----
epistemic side

The theory being developed can account, in a principled way, for a distinction between modal constructions introduced by a modal verb and the subjunctive clauses, including the "infinitival clauses" in Polish such as Marysia chce pic. In the case of a modal construction the designated process is grounded by the modal eitherin irreality (which is the case of deontic modality) or in potential reality of the dynamic evolutionary model (which is the case of epistemic modality) (Langacker 1991). In contrast, the process of the subjunctive is not grounded at all: wishes and desires are anchored in local spaces "conjured up for specific purposes of the speaker" (Langacker 1991; Achard 1998). The theory proposed here, incorporating Achard's theory of complemetation, langacker's models of reality and Givon's insights into the nature of the subjunctive can offer a unitary account of complemetation and all all other types of modality, the modality of the subjunctive included.

REFERENCES

--Achard, M. 1998, Representation of Cognitive Structures . Syntax and Semantics of French Sentential Complements, Berlin, Mouton de Gruyter.

-- Givon, T. 1994 "Irrealis and the Subjunctive," Studies in Language 18-2, 265-337

--Kardela, H. 2000 Parameters and Dimensions in Grammar. A/D Asymmetries and Subjectivity Relations in Polish, UMCS Press, Lublin

--Langacker, R. 1991. Foundations of Cognitive Grammar. Vol 2. Descriptive Application, Stanford, Cal: Stanfrod University Press.

--Polanski, K. (1993) Encyklopedia Jezykoznawstwa Ogolnego, Wroclaw Ossolineum