The article further deals with the subject of the reference of zero subjects on the basis of the main theoretical premises discussed earlier (see Akamauskaitė, Kardelis 2014, 1–18). The analysis of individual examples in which the position opened up by the verb is occupied by a zero article has demonstrated that zero subjects tend to prevail. However, it is also important to note that a considerable part of the examples are subject to a dual interpretation where a phonologically unexpressed zero element may be considered both as a zero, and as a zero anaphoric subject. As a result, the most reliable criterion in examining such cases is the analysis of the wider context. Discourse is a domain in which it is possible to observe the real picture of the usage of zero and zero anaphoric subjects in spoken language. The analysis of individual examples has revealed that in order to clearly distinguish between zero and zero anaphoric subjects, and to determine the referents of zero subjects as well as describe their features, the analysis of one sentence is not sufficient. The investigation of the reference of zero subjects requires the analysis of a more extensive discourse and the consideration of the formal and the semantic consistency of the text. In other words, it requires the analysis of coherence and cohesion, their interaction and the relationship between them. The surface structure of a sentence does not always reflect the deep (semantic) coherence, whereas a discourse, which enables a researcher to reveal specific means of cohesion, an important factor is the concrete situation, around which centres the discourse that is (being) created. In addition, a significant role is also played by the cognitive aspect, the general background knowledge, and the relationship between the addressee and the addresser.
The main objectives of this article are as follows: 1) to identify zero anaphoric (and cataphoric) and zero subjects that are present in discourse; 2) to analyse the reference of zero subjects and define the features of their referents. In this article the theoretical principles of the analysis are not discussed separately since they correspond to the ones presented in the analysis of individual examples.
The empirical data for the analysis of the reference of zero subjects have been obtained from collections of dialectal texts of the Šakiai and Griškabūdis sub-dialects of the Kauniškiai variety of Western Aukštaitija. The sources include the following texts: Rima Bacevičiūtė’s “Texts of the Šakiai sub-dialect” (ŠT) (2006) and Rima Bacevičiūtė and Vilija Sakalauskienė’s “Texts of the surroundings of Griškabūdis” (GT) (2008). In order to see a more extensive distribution of zero subjects, three texts that differ in length and structure have been purposefully selected for the analysis. To maintain the visual representation and clarity the same text in the article is presented twice. At first, all zero subjects are marked in the text using symbol Ø1 (zero subjects are marked with number; if the number repeats it means that the subject of the zero referent (or antecedent) is the same). Having analysed the discourse, i.e. having identified zero anaphoric (cataphoric) subjects and zero subjects, having established the referents of zero subjects and having described their features, the text is presented one more time. It includes the marking of:
a) zero subjects (and the features of their referents); Øunspecified, Øgeneralised, Øx (this marking is used to indicate a zero subject when the features of its referent are subject to a dual interpretation or are unclear);
b) zero anaphoric subjects (and their antecedents) [ØA farmers];
a) zero cataphoric subjects (and their antecedents) [ØK they both].
The analysis of three different texts has revealed important evidence, i.e. zero anaphoric subjects have absolutely no prevalence in discourse. In addition, an important role is played by the structure of a text. Zero anaphoric subjects are dominant in the second text, whereas they no longer prevail in the third one, although their subject-matter is rather similar. As a result, it is possible to conclude that the distribution of zero subjects in a text is determined by more factors than just the syntactic structure. Most probably it is determined by the semantic structure created by means of cohesion and coherence. In addition, it is important to highlight that the above-mentioned mechanism and the distribution of zero subjects as well as syntactic structures related to them are most probably governed by certain cognitive schemas created by the addressee and the addresser.
In order to get a clearer picture of incompleteness, it would be interesting to research two more issues such as: a) the distribution of constructions with anaphoric zero subjects and constructions with anaphoric pronouns (Pro3 (he, she, they[masculine] , they[feminine])) and the relationship between such constructions in discourse; b) to compare the discourse of a similar topic in different sub-dialects.