The nanosyntax of the Northwest Germanic reinforced demonstrative

Start - End 
2011 - 2016 (ongoing)



This dissertation is a detailed study of the reinforced demonstrative pronoun (RDem) of the oldest Northwest Germanic (NWGmc) languages: Runic Norse (RN) súsi, sási, þatsi; Old Norse (ON) sjá/þessi, sjá/þessi, þetta; Old English (OE) þēos, þe(:)s, þis; Old Frisian (OF) thius, this, thit; Old Saxon (OS) thius, *these, thit; and Old High German (OHG) dësiu, dësēr, diz

A fine-grained morphological decomposition of RDem in ON leads to the identification of five distinct morphological ingredients, which combine in three different structures within the paradigm. I treat these facts within the formal framework of Cinque (2005). Working within his U20 program, I demonstrate how the correct functional sequence of RDem can be deduced, how the three RDem structures are derived, and why only these three structures (of the 24 possible structures) are attested. Interestingly, support for Cinque’s system is found in the allomorphy between the reinforcer morphemes -s (found in the cyclic-type derivations) and -a (found in the roll-up derivation).

Since this allomorphy seems to be predictable only on the basis of the syntactic derivation rather than lexical content, however, a more nuanced view of the RDem system is needed, in the form of a nanosyntactic analysis. Making the functional sequence more fine-grained, in line with the nanosyntactic approach, allows us to capture the -s ~ -a alternation in terms of lexical structure. A number of other facts also fall out from this more fine-grained approach. Finally, I bring the WGmc facts into the scope of the analysis. With a complete understanding of ON under our belts, we are able to capture the various points of morphological variation in the RDem paradigms of OE, OF, OS, OHG (and even RN) in a very simple way.

All of the variation observed across the RDem paradigms can be explained in terms of the way lexical entries are structured. The functional sequence happens to be packaged slightly differently in each language, which then leads to different derivations during lexicalization. 



Phd Student(s)


Michal Starke

University of Tromsø