Quoting Ian Spackman <[log in to unmask]>:

> >
> >Yes, and about simplifying an existing system much has been said.
> >But: What if somebody wants to create an artlang just for himself,
> >saying: I want more parts of speech! Many ways of minimizing and
> >combining parts of speech have been pointed out. What, on the
> >contrary, is the known, thought or perceived maximum for their
> >number and diversity? Sounds like a stupid question? Maybe. But
> >since its about artistic freedom of conlanging, this list
> >should be the right place to discuss it. ;-)
> >
> >Harald
> >:-))
> Well, I posted the other day indicating I was considering a language with
> more parts of speech, but I haven't made one yet.
> Notionally of course every word could be its own part of speech, showing
> idiosyncratic behaviours, but such a language would be very cumbersome in
> practice, which is no doubt why we have parts of speech.
> I tripped over an English word which I think is the only one-word member of
> its part of speech a few months back: _evens_.  It's a ratio: it can be
> used in all the same contexts as can _three-to-one_ or _seven-to-two
> against_, etc., and only those contexts, so far as I can make out.  I've
> seen it cited in dictionaries as a noun, and adjective and an adverb (the
> traditional catch-all), but it's clearly really none of those.  (I mention
> this mostly to point out that English, and I suspect most languages, have a
> lot of minor parts of speech that aren't usually recognised.)

A Swedish grammar I had in school listed the infinitival marker _att_ "to" as
a part of speech of its own. It had a total of ten or so listed. I'll, for the
benefit of those seeking inspiration for excessively large inventories, try to
list them all, but several could, little doubt, reasonably be collapsed.

infinitival marker

I'm thinking those were all, but I might've missed one or two.