Inherent Circularity in Laryngeal Realism? Three Levels of Explanation of the Pre-sonorant Sandhi Patterns in Polish (Part 2)

In the first part of the article, two approaches to laryngeal phonology – the realist and the relativist – were introduced and compared with regard to their treatment of the sandhi patterns in two major varieties of Polish. The discussion revolved around three types of circularity that result from the blurring of the line between phonology and phonetics. The current part of the article examines the theoretical consequences of the privative approachcal led new laryngeal realism (e.g. van der Hulst 2015; Wojtkowiak and Schwartz 2018) with respect to the pre-sonorant sandhi effects in Polish, assuming the broader framework of Onset Prominence (Schwartz 2010). While new realism does suffer from circularity to some extent, it seems to capture not only the main phonetic and phonological intuitions about the sandhi phenomena, but it also makes strong claims about the diachronic development of the two major dialects of Polish. In comparison to laryngeal relativism, it leads to a similar structure of sound systems, with strict separation of phonetics and phonology, but it places the explanation of the sandhi phenomena at the interface rather than in the phonology itself. On the other hand, the general Onset Prominence framework appears to subvert these achievements by merging phonetics, phonology and the interface into one system. A solution to this problem may be to assume that the Onset Prominence representation should not be hierarchical, as it reflects the phonetic representation alone.