I had a look at that thread and decided that I'm not able to do all the reading to meaningfully participate.
Just to come back to zomp's original point, namely, that there may be a way to indicate syntactic roles by an "omission / presence" hierarchy, i.e., that in English
1. in a sentence with one argument this will always be interprted as a subject and
2. in a sentence with two arguments this will always be interpreted as subject and direct object (correct me if I got you wrong, Mark!).
Now, after thinking about this a while, I came to the conclusion that English doesn't seem to work this way:
1) Does not really work for transitive verbs if the sole argument follows the verb, i.e. when it is in the object slot:
- Normal interpretation: Mary (subj.) performs the action of kissing. Works.
- Normal interpretation: (Omitted subject) kisses Mary (object).
That is, the slot position overrides the putative omission / presence hierarchy.
Even if you include clearly nominative forms, that leads just to ungrammatical sentences (I'm not a native speaker, so I may be wrong here - and things like this may, of course, be admissible in poetry, but then, what isn't?): *Kisses she.
The rule seems to work, though, for some intransitive verba dicendi and cogitandi: Says / Thinks Mary.
But this is a special case, and AFAIK, these constructions require that the content of what was said or thought needs to float somewhere in the near context. For other intransitive
verbs, it doesn't seem to work: *Walks / sits / prevaricates Mary.
The same seems to be true for two arguments:
Gives Peter the book.
will be interpreted as (Omitted subject) gives Peter (IO) the book (O).
So, it seems to me, in English, the slot in the word order is decisive, not any omission / presence hierarchy.