On 01.12.2000 17:41 Uhr, "Richard F. Smit" <r.f.smit@...> wrote:

> If some element is unclear, and isn't covered by the one known
> grammar, I would tend to favour a Latin-like construction (not that I've
> had time to practise this). Is this a sensible approach?
To be frank -I don't think so-. There are a lot of elements in Q. which are
quite un-Latin. Also it can be gathered from J.R.R.T.'s writings that the
classical languages weren't exactly prominent in his modelling Q. and
Sindarin. If you are looking for parallels look for e.g. Finnish and for
some constructions also (Proto-) Germanic. There is a wealth of relations
there and I wdn't completely exclude Latin (or Greek), but to give them a
prominent role wd IMHO be quite inadequate.

-David Kiltz