The syntax of nominal constructions has so far attracted relatively little attention in the rapidly growing literature on sign languages. In Poland, there have been virtually no studies addressing the topic. The goal of the present paper is to offer an overview of the nominal syntax of PJM (polski język migowy), the visual-spatial language of the Polish Deaf, which is diachronically and synchronically independent of spoken/written Polish. A key, and novel, aspect of the present proposal is that we base our descriptive model on a detailed investigation of extensive empirical data. For the purposes of this study, we have carefully inspected a sample of video material extracted from the first-ever corpus of PJM that is currently being compiled at the University of Warsaw. An in-depth examination of the data has allowed us to produce a typology of PJM nominal constructions involving adjectives and other adnominal modifiers. The present paper outlines the word-order generalizations that emerge from the analyzed data. We observe that PJM adjectives show a clear tendency to appear in postposition with respect to the head noun, whereas other adnominal modifiers (such as numerals or possessives) most often precede the noun. Additionally, we confront these findings with data on spoken Polish nominals extracted from the National Corpus of Polish.

The paper offers an autosegmental approach to Polish palatalizations whereby the presence of palatalizing features is the result of the translation of morpho-­syntactic features into phonological features. In the first part I present an analysis of the structural change of the relevant palatalizations, which boils down to the account of how floating autosegments are integrated into the underlying structures of the stem-­final segments. The second part is preoccupied with how relevant autosegments are inserted into representation: the palatalizing floating features are phonological ‘halves’ of vocabulary items matching the feature sets marking inflectional categories in Polish. The paper finishes with the discussion of the distribution of the endings i/y/i~ɨ/ and e /ɛ/ as the markers of the Dative and Locative in one of the declension classes in Polish. I show that the approach advocated here fares better at predicting the distribution of the said endings than the better established approach presented in Gussmann (2007).