[quote="J.M.D."]
That's...a lot of spacetime postpositions.
:D
[/quote]

Yep.  Some of them are rarely useful, but you might be surprised how often I find use for some highly specific postposition with no close English equivalent.

[quote="J.M.D."]
[quote]
Since it's an engelang of sorts, I'm trying to make the argument structures consistent within semantic domains, e.g. all verbs of thinking or feeling have the same argument structures; but I'm finding it tricky.  Different verbs of saying/speaking seem to want different argument structures.[/quote]
As currently envisioned, speak and say (or the approximate translations) would indeed be different, though they would have the same root.  "Say" (or more aptly, "utter") would be an affectual transitive noun, with words, statements and utterances as the typical direct objects.  That is, you [i]say words[/i], 'affecting' the words in the sense that you bring them into existence, much as one [i]writes[/i] a book, [i]bakes[/i] a cake, [i]builds[/i] a cabinet, [i]conceives[/i] a child, etc.  But "speak" would be a non-affective verb, which can be either transitive or intransitive.  When the transitive version of it takes a direct object, X, then this would mean "to speak [i]about[/i] X".  So while "P says 'hello'" means P affected the word 'hello' by uttering it, "P speaks 'hello'" would mean that P talks about the word hello.[/quote]

There's another argument structure for {tw-z} I've overlooked; if you say that someone says a single word, or name, or whatever, it might be marked with the performative case rather than the direct quotation subordinating conjunction.  (gzb has various postpositions for objects of result and object of performance, that aren't preexisting things affected by the action of the verb but come into existence either momentarily or for a while afterward because of the action of the verb; e.g. "to write an email", "email" would be marked as creative/abstract object of result; "to build a house", "house" is tangible/constructive object of result; "to say 'ni' / play a polka / sing a ballad / play Glossotechnia", all the objects of those verbs would be object of performance.)

I've been thinking about how to render the concept "to mention" which appeared several times in the examples from  [i]Simpler Syntax[/i] earlier in the thread; I think I could do it by marking the object of the "say" verb with the topic postposition instead of the object-of-performance postposition.   Not sure yet how generalizable that is.  I might still need some kind of quotative particle, or maybe a quotative stress/intonation pattern, to distinguish words from the things they refer to in ambiguous cases.

{&#934;&#293;u} &#265;ul-i tw-&#421;-z.
"elephant" performance-at say-3-V.ACT
She said (the word) "elephant".

{&#934;&#293;u} m&#301;-i tw-&#421;-z.
"elephant" TOP-at say-3-V.ACT
She mentioned (the word) "elephant".

&#934;&#293;u m&#301;-i gju-&#421;-z.
elephant TOP-at speak-3-V.ACT
She talked about elephants / discussed elephants.

??? &#934;&#293;u &#265;ul-i gju-&#421;-z.

Not sure what this last one (performative case and "speak" verb) would mean, if anything.


