Wykładniki negacji w polskim języku migowym

The objective of my PhD thesis “Negation markers in Polish Sign Language” is to provide a large-scale and thorough description of mechanisms of negation in Polish Sign Language (polski język migowy, PJM). Polish Sign Language is a young, visual-spatial natural language that is used in every day communication by the Deaf in Poland. It first emerged around 1817 and has been an object of scientific research only since the 1990s. Most of the grammatical features of this language remain understudied and my PhD aims to help fill this gap. I begin my thesis with some theoretical considerations. I explain that two levels of articulation exist in sign languages: the first comprised by manual signs articulated by the hands, the second composed of non-manual signals understood as movements of the body, head and face used as meaningful parts of articulated utterances. It is widely believed that both of these levels play equally important grammatical roles in sign language discourse. In the literature, one of the most commonly mentioned non-manual signals is a headshake (horizontal movement of the head) articulated in negative contexts. This suprasegmental signal is reported to serve as a grammatical marker of negation in a variety of sign languages of the world. One of the main goals of my thesis is to check the extent to which this holds for PJM. In order to answer this main research question (and several others) I carried out and present a large-scale corpus based study of negation in PJM. I base my analyses on the PJM Corpus that is currently being compiled at the University of Warsaw by the Section for Sign Linguistics. I chose samples of annotated linguistic material and tagged them for all occurrences of negative lexical signs, headshakes and negative clauses. Analysis of the material so prepared leads me to interesting findings about nature of negation in PJM and non-manuals in sign languages. First of all, headshakes in PJM turn out to occur in a variety of different contexts. Most commonly, of course, they appear in negative utterances but they may also accompany questions, alternatives, classifier constructions or serve as emphasis markers. As for the negative lexical items, there are three main groups: signs negated by a negation morpheme (which is articulated as a circular movement of the wrist preceding the sign that is negated), negative particles, and signs negated by those particles. Interestingly, negative lexical signs turn out to be co-articulated with non- 247 manual activity as often as they are articulated without any head movements. What is more, there are many cases of negative head movements that are co-articulated with lexical signs but cannot be seen as negating the latter. These two findings lead me to believe that non-manual activity might serve more of a gestural than a grammatical role in Polish Sign Language discourse. However, it is possible (albeit rather rare) for a verb to be negated only by means of a headshake. In the analyzed material I found 203 such constructions. In these cases the headshake seems to be grammatically constrained: it cannot spread beyond the negated element in the clause (be it either a verb or an object). In clauses in which negative polarity is added by a lexical element, on the other hand, the length of the headshake does not seem to be constrained. It can spread freely over clause constituents. None of these findings have been previously noted in the literature concerning PJM negation. I end my thesis by positing some conclusions that follow from these observations: 1. Negative PJM constructions exhibit much more variation than initially predicted – it seems impossible to classify PJM according to the existing typology of sign languages into manual dominant and non-manual dominant with respect to negation. 2. Headshaking in PJM seems to have more of a gestural than a grammatical nature – it appears not to have been incorporated into the linguistic systems in any unexpected way. 3. Grammaticalization of the headshake might be an on-going process (taking into consideration the approx. 200 constructions in which it negates positive signs). 4. Corpus methodology seems to be the most suitable for conducting sign language research, as it helps to avoid the difficulties that typological studies face (such as an insufficient research sample that fails to contain all variants of the analyzed phenomena).