http://www.orinst.ox.ac.uk/sites/jap-ling/files/files/matsumoto.talk_.pdf http://pkdas.in/JNU/typo/cv.pdf <~ this one is actually about hindi, but written in English. http://ir.lib.hiroshima-u.ac.jp/files/public/14164/20141016120714597161/J-Quantitative-Linguistics_11_3_233-250_2004.pdf (page 3) https://en.wikipedia.org/wiki/Compound_(linguistics) I'm not sure if you ran across these in your search or not, but I figured I'd share them, just in case. I think the last one was the most clearly explanatory. So far as I can understand it, a lexical-compound is one that takes it's definition from a pairing of two specific words together. So, the new verb doesn't necessarily come from the meaning of either word individually, but rather the words used to form it. The example given in the third link are "tabe+kuraberu (‘eat+compare’), nomi+aruku (‘drink+walk’) and kaki+toru (‘write+take’)". These sorts of verbs seem to have very strict rules about how they are formed. Syntactic compounds, however, takes their definition from the meaning of each of the paired verbs. The examples, again from the third link, are tabe+hazimeru (‘eat+begin’), nomi+oeru (‘drink+finish’) and kaki+naosu (‘write+fix’) (which I can only assume mean "start eating, finished drinking, and to correct writing, possibly editing?) These verbs are supposed to be more transparent in their meaning, and so the pairing rules seem to be a little more lenient. Long story short: Lexical compound verbs seem to be a fixed set of v1+v2 combinations based purely on the actual words, absent their literal definition. Syntactic compounds rely completely on the definitons of v1+v2 to provide meaning. I also found "exocentric and copulative" as descriptors for these kinds of words under the "compound" wiki. Is this what you were looking for?