The core idea of Nanosyntax (Starke 2009 et seq.) is that phrase structure trees have just single features at their terminals (rather than feature bundles). Morphemes are inserted late, and they may pronounce several such terminals – more specifically, a syntactic phrase. In this system, morphemes or words are not the input to syntax, but they are all the products of Merge, operating over single features. I show how we can use this theory to shed syntactic light on traditional murky areas of morphology, such as arbitrary word classes (declensions), root suppletion, syncretism, allomorphy and other apparently morphological phenomena. I assume the basic knowledge of generative syntax: you have to be familiar with terms such as head, complement, phrasal movement, Merge.
You can find Pavel’s slides here.