The article deals with the issue of the reference of zero subjects which has been very scarcely examined in Lithuanian linguistics. The first part of the article presents a review of the theoretical tools to analyse zero elements and its key theoretical tenets. The main framework that the study relies on is Igor Mel‘čuk’s theory of incompleteness as well as other studies that supplemented and verified Mel‘čuk’s theory. In the discussion of theoretical premises and relying on other linguists’ argumentation, a crucial distinction is made in the article between a zero element and ellipsis. The second part of the article analyses authentic empirical linguistic data of Kauniškiai sub-dialect of Western Aukštaitija obtained from a collection of dialectal texts of the Šakiai and Griškabūdis sub-dialects. The main goal of the article is to determine referential peculiarities of zero subjects and to encourage further motivation to differentiate between the zero subject and the zero anaphoric subject. It should be emphasised that the study is based on individual examples which were analysed. Such a choice in terms of examples was made in order to demonstrate that the distinction of the zero subject and the zero anaphoric subject without a detailed examination of the wider context may prove complicated. In some cases when no adequate context is available, there are no reliable criteria to distinguish these two elements.
The main conclusions of the article are as follows:
a) The analysis of individual examples shows that zero subjects could be dominant. However, in many cases a dual interpretation is possible, i.e. a zero element may be treated as a zero subject or as a zero anaphoric subject. In such cases it is necessary to examine the wider context, since neither the syntactic structure, nor the semantics of the sentence can provide clear criteria in order to distinguish a zero subject from a zero anaphoric subject.
b) Referents of zero subjects may be different. According to the features that are typical to those referents, it is possible to distinguish at least two groups. The procedure requires paying special attention to the features of the referent, which are determined by a concrete sentence and its semantics: ‘±animate’, ‘±person’, ‘±concrete’ etc.
c) Referents of zero subjects in the first group typically share the feature ‘unspecified’. Usually these are referents that have the feature ‘some person; a person or people’. In this case the speaker includes a zero subject into the sentence structure yet does not try to name the subject most probably because s/he does not think that it (the subject) needs to be specified. The subject may also be obvious, for example, it may be clear from the general background knowledge or the mutual experience of the interlocutors.
d) The second group consists of referents of zero subject, which may be assigned the feature ‘generalised’. It includes cases in which the speaker cannot clearly name the referents for different reasons or the referent is difficult to name using one word only.
e) Referents of zero anaphoric subjects, i.e. referents of the ØPro3 antecedents may only include concrete ‘persons, people’ mentioned in the text that only possess concrete features ant not subjects that have the feature ‘natural forces’.
Presumably, the most important conclusion of this article is the following: in order to have a real and full picture of zero subjects and zero anaphoric subjects in spoken language, in order to clearly distinguish zero and zero anaphoric subjects and in order to reliably determine the peculiarities of the referents of zero subjects, it is necessary to undertake an analysis of extensive coherent texts.