Trigeriness, or something very much like it

Here's something I culled from Fillmore's "Case for Case" the other night--
it might ring some bells. (If I can do the graphics right).  He is
discussing Sapir's typological classification of Amerind langs. but the
application is wider.

(Keep in mind that Case for Case first appeard around 1965/6 or so, and
circulated in xerox form for quite some time before actual publication)
(Keep in mind too, that "case" as he uses it refers to the "arguments" of a
verb/predicate A(gent) and O(bject = Patient), plus their relationship to
the surface realizations (nominative, accusative, etc.) )

--Paraphrase--
With these two roles, we can have 3 sentence types:
V + A intrans. sentences with active "subject"
V + A + O transitive sentences with agents
V + O intrans. sentences with inactive "subjects"

Since the V element is constant, we can represent these 3 sentence types by
presenting the case frames in three lines, as follows:

     A
O       A
     O

...there are languages which, like Yana, have only one
form for pronouns in all four of these positions ---
-----------
|     A        |
| O      A   |
|___O___ |

There are languages ...that have a separate form for the O [pronoun] in the
trans.sentence, all others being the same. The two forms are traditionally
called "nominative" and "accusative"---.
      _____
      | A     |
  O \   A  |
      /_O_|

There are langs. ...which give one form to the A  of transitive sentences
and another to the remaining cases. The terms "ergative" and "nominative
[absolutive]" are often given to this--
______
|    A   |
|O      /  A
|__O_|

There are languages like Dakota which have separate forms for A and O; here
the terms are usually "active" and "inactive"--
   ___
   \  A \
\O \_ A_\
 \_ O\          (This isn't clear, the two A's should be circled, likewise
the two O's.)

Lastly, there is...Takelma, which has one form for the pronominal NP of
intrans.sentences, and two separate forms for the A and O of transitive
sentences---
             ___
             | A|
[O] [A] |    |
             | O|  (Here the top A and bottom O are enclosed in a circle,
the middle A and O are in their own circles.)
--end paraphrase--

A couple of additions, meae culpae and a confessio-- when I first brought
Fillmore's system up here on the list, I was drawing on sainted memory of
him as a refreshing alternative to Chomsky.  Upon rereading him, I find (1)
he was, at least initially, merely trying to amplify Transformational
Grammar-- he still has all the transformations etc.  But he must have had a
sneaking feeling that his system didn't really conform to TG. Moreover,
people like McCawley drew on him, in particular the resemblance of his "case
frames"-- V[A O etc]-- to predicates in formal logic --x {y,z}-- and took
semantic analysis off  into some quite non-Chomskian directions.
(2) Because he was working within the TG framework, his system IMO didn't go
quite as far as it might have; it strikes me now as just a bit....primitive.
(3) I thought he assigned experiential verbs a case frame [O A]-- he did
not, they are [O D(ative)] where "Dative" includes "Experiencer"-- though of
course in Engl. and many other langs., the "subject" form of D is indeed the
"nominative" or unmarked case.

Nevertheless, I hope you find his approach interesting, and will look into
"The Case for Case" on your own.