This is the second part of the article dedicated to the discussion of some methodological problems related to the history of Khinalug, published in the current volume of Iran and the Caucasus (see Schulze 2018). Whereas the first part analyses some basic data on Khinalug in its genetic problem and addressed some questions of loans and cognates, the second one turns to grammatical issues. Khinalug, a minority language spoken by some 1500 people mainly in the village of Khinalug in the north of Azerbaijan Republic, is generally regarded as the most divergent East Caucasian language. Its exact genealogical place within the world of the roughly 30 East Caucasian languages has been debated since long. Still, at least some of the relevant contributions to this debate ground their arguments in just a rather small piece of evidence, usually taken from a handful of assumed lexical correspondences and typological analogies. The same holds for grammar. As for morphosyntax, the problem is complicated by the fact that hitherto it is virtually impossible to safely reconstruct a more systematic inventory of Proto-East Caucasian morphemes together with their function values. A closer look at the morphosyntax of Khinalug may lead to just the same conclusion that seems to emerge from a more comprehensive analysis of the lexicon: I suggest, to consider the possibility that Khinalug is not an East Caucasian language from a genetic point of view, but a non-East Caucasian language that has become ‘Caucasianized’ over times.