On Saturday 29 March 2003 09:59 pm, Sally Caves wrote:

> So in other words, it's linked with intransitivity.   In an active
> language, which is a subset of an ergative language, the subject of an
> intransitive verb can be treated as an agent if it is performing that
> intransitive action in some form of volitionality.  "I looked and listened
> eagerly."  As opposed to "me looked and listened eagerly."
        Not necessarily. As your second quote from Trask indicated, there are
languages that "are rigidly divided into those taking agent subjects and
those taking non-agent subjects." In other words, there are some languages
where volitionality is not even considered. "Him fell" would be the only
correct form, regardless of whether he slipped on the ice or hit the deck
when the bullets started flying over his head. And in other cases, there are
some languages that have fixed agents or non-agents for certain verbs, while
other verbs can take an agent or patient as a subject, depending on volition.
Got that straight? :)
        Also, I don't think it would be correct to call an active system a "subset"
of an ergative system. It's really more of a third option, as in
nominative-ergative-active. After all, the chief distinguisher of the three
is how they treat the subject of an intransitive verb. In a nominative
system, the subject matches the case of the agent in a transitive sentence.
In a ergative system, the subject matches the case of the patient in a
transitive sentence. And in an active system, the subject can be either
agentive or patientive, depending on volition, semantics, or which side of
the bed you got out of this morning.

> I guess I would call that "agentive."  Or as Teoh suggests, "volitional."
> That's the distinction made in Teonaht in the "Split Nominative":
> volitional and non-volitional subjects, regardless of whether they govern
> transitive or intransitive verbs.
        "Agentive" and "patientive" are the terms I use, but "volitional" works as
well (and is clearer as to what semantic principles are at work.)

> "Lexical (or "full") words" are defined in Trask as:  "A word with real
> semantic content, such as "green," "kitchen," and "swim."  Again this
> brings up the question of what is not real semantic content.  So I look up
> "grammatical word," also known as "empty word," "form word," or the
> familiar "function word."  "A word with little or no intrinsic semantic
> content which primarily serves some grammatical purpose: of, the."   My
> confusion has always been that I considered even function words as serving
> some semantic purpose.  I have trouble seeing semantic and grammatical (or
> syntactical) as distinct, but I can understand how these terms function
> within THIS context.
        I am guessing that "non-lexical" verbs would by like auxillary verbs. In
which case, that makes sense: "Him was falling" vs. "Him fell" do not effect
the agentivity or patientivity of the subject. "Was Xing" only adds an aspect

> No.

        Ok, that sounds much like a semantically based active system (as opposed to a
volitionally based one), despite your terminology. It's just a natural
extension from the intransitive to the transitive. That is, while it is
proper to say "the girl (E) heard the sound," it is wrong to say "the girl
(A) heard the sound," right? In other words, the subject of "hear,"
regardless of whether it is in a transitive or intransitive sentence, is
alwways E, yes? If so, then that's semantic, because the meaning of the verb
is what determines the case of the subject, not the volition of the subject
        Enamyn, by way of contrast, is a (mostly) volitionally based active system.
In other words, "the girl (E) heard the sound," would mean much the same as
in Teonaht, but "the girl (A) heard the sound," would mean that she was
actively listening for the sound and heard it.
        Ah! But reading further, I see that you have a "ambivolitional" verb, which
seems to depend on the volition of the subject. Of course, it's perfectly
natural to mix the two. Let me quote Payne from the Conlang Bible
("Describing Morphosyntax," for the neophytes):

# As might be expected, split-S and fluid-S languages do not constitute two
# mutually exclusive language types. Typically, a given language will have
# some intransitive verbs that require S(A) subjects, others that require S(P)
# subjects, and still others that allow either S(A) or S(P) subjects.

(BTW, "split-S" refers to semantic systems, "fluid-S" to volitional systems.
Apparently, no one can agree on the terms. :)

> It's interesting to see Daniel's chart on page 9, with Dixon's terminology.
> Dixon makes a distinction between A and S (a subject that governs a
> transitive verb and a subject that governs a non-transitive verb).  This is
> a distinction I associate with an ergative language, but Teonaht has no
> such distinction.
        Payne does the same: "A" and "P" for a transitive sentence (regardless of
system) and "S" for intransitive systems. The chart is almost identical to
the one in D.M. I suspect that's because Payne is heavily influenced by

> Its agent and experiencer are not related to
> transitivity. Only volitionality.  So it doesn't really fit any of the
> categories on page nine.  I don't even think it matches Daniel's chart on
> page 10, because it hasn't up till now associated the experiencer with the
> patient.
        I suppose that's the key point. I would still call it an active system, even
if it doesn't associate E with P, mostly because there's still a split within
S: it's either A or E, rather than only A or only E (or P). Just out of
curiousity, is there an intransitive Teonaht sentence where S is P, and *not*
E? If yes, then you would have a three-way split between A, P, and E. If not,
then you've still got the two way split between A and P of an active system,
except that P is in a different case. :)

> In making the change I suggested, dropping lorfa to lorf in
> non-agentive subjects and in patients, I'm moving T. towards a known
> linguistic type.  I can't tell whether this is an advantage or not.  Part
> of what I want T. to suggest is an overlay of two different language types.
>  It is accusative first, and volitional second, and moving further towards
> a more developed sense of volitionality.
        This seems like a perfectly natural progression to me, especially if you are
moving from a semantic-based system to a volitional system.

Oh what a tangled web they weave who try a new word to conceive!