jak Ote wrote: a2) jan Kikamesi li lape li lukin -- Gilgamesh was sleeping and looking/seeing.
i like this because in this construction the reader would imply the causative relationship between lape and lukin. it is easy to get the meaning because the both are predicates. "lukin lape" is also good as a complex verb, but (to me) requires a little bit more time to analyse
I don't see a causal connection between the two in 'li lape li lukin,' merely something like simultaneity. But I do think its a good way to deal with dreams where the content gets mentioned somehow. 'lukin lape' (and the others) do take a minute and may get misread (see my bollixes often above and elsewhere on these fora), but once established as an idiom, it goes smoothly.
I don't see seeing a dream as clearly passive in an abstract sense: the agent is the seer, the dream the seen -- "a dream was seen" actually seems odd to me (not to criticize Russian, mind you; every language has quirks of grammar that make no logical sense -- we are just starting to learn tp's) even with "by me" thrown in.
I'm not too sure that the absence of passive forms is really a part of tp philosophy (really Sonja's philosophy which shares some features with her language) or is just a matter of keeping the grammar as simple as possible (which is, of course, one of those common features).