On Wed, 30 Jan 2019 at 15:00, Fog Flowerzone <[log in to unmask]> wrote:
> Hello, I'm also a member of the Sajem Tan tribe. I don't really know how to
> use the listserv, so if there's a formatting issue, I apologize.
> >Surely not, as verbs are not inherently passive or active; they may
> >however, be inherently non-agentive.
> > Now, if you can *also* say ?"kemönmon daţnycfê sê" to mean something
> >like "I found the deer (on purpose)", then maybe that experiencer
> >construction is best called a (syntactic) passive. But if not, if the
> >OSV construction is actually inherently required by that verb, then
> >"voice" is the wrong way to be thinking about this.
> I think the tribe is mistaken in declaring any verb inherently experiencer.
> We can indeed say "kemönmon daţnycfê sê". The verb itself doesn't trigger
> the OSV word order, it's just that in many cases a verb will be
> overwhelmingly more likely to be used one way or the other, though in each
> case you can change it to imply something about the volition or what have
> you about the subject.  For example, *Jidök ţê nëmfê* means "I saw a
> flower", but *Jidök nëmfê ţê* means "I looked at a flower". So it's not
> like OSV is the required word order for these verbs, but rather for
> indicating thematic role. We've always said experiencer-ism is inherent to
> the verb in Sajem Tan, but I'm starting to feel like it's more than that.
> >Is it possible to say ?"vazöt sê dëgemfêsun dac" to mean "the cat was
> >chosen by me"?
> We would say "sêžê vazöt dëgemfêsun dac". "žê" is the focus case marker. I
> feel like English might have had a bit of influence on this construction,
> since it seems to be built around the English syntax, where "cat" would be
> the subject (of the verb "was") and "me" would be the indirect object of
> the preposition "by". So according to this the ST sentence is SOV, but it
> really deserves more thinking. So far, it seems like active sentences are
> PVA (patient-verb-agent), passive sentences are APV (agent-patient-verb),
> and experiencer sentences are PAV (patient-agent-verb). So the word order
> changes in both cases.
> The reason we decided it might have to do with voice is this: we were
> proposing terms to be used in the field of linguistics, and we came up with
> _vycgyn_ ("to carry out an action" + agent suffix, "-er") for subject/agent
> and _vycžnu_ (vyc + patient suffix, "-ee") for object/patient. We weren't
> sure how to denote experiencer subjects, since they don't carry out an
> action, and the use of the agent suffix _gyn_, which literally means
> "worker", didn't make sense either, so we can't call it a vycgyn.

OK then, based on that evidence it does seem clear that you do not
have a distinct syntactic class of experience verbs--you have an
experience *construction*, which can be applied to transitive verbs.

Strictly speaking, you *can* call this a voice operation. The whole
thing with voice operations is that they alter the argument structure
of the verb; altering the semantic role of the subject technically
counts. It's clearly *not* passive; it is both morphosyntactically and
semantically distinct from the already-identified passive
construction, and it does not involve promoting a basic object into
subject position (the primary hallmark of a passive construction). So,
you can just go ahead and call it an "experiential voice" or something
like that.

However, personally I feel like "voice" is not the *best* category to
use to describe that. I would prefer to say that, in the active voice,
there is a syntactic split based on volitionality; OVS order signals a
highly agentive, volitional subject, which OSV order signals a
low-agentivity, non-volitional subject.

Either way, you have an internal terminology problem to solve, because
you have conflated the concepts of "subject" and "agent" in your
existing terminology, which do not always line up. One route out that
problem is to just say that etymology, or even derivation, does not
fully determine lexical semantics, and accept that despite its origins
that may not fully "make sense", as a technical grammar term "vycgyn"
just means "morphosyntactic subject", regardless of what semantic role
is associated with it. Another route it to just come up with a new,
different word for subject, and retrofit your existing documentation,
reserving "vycgyn" for actual semantic agents. And a third option
would be to just ditch the concept of "subject" entirely, and describe
your grammar entirely in semantic terms, after deriving a new word for
the "experiencer" role. Then you would have an active voice order of
PVA, a passive order of PV, and an experiential order of PEV (as long
as you continue to conflate Patients and Themes in a single
grammatical category, anyway--going all out on strict semantic
categorization would result in an explosion of supposedly-different
structures that don't actually add anything to ones understanding of
how the language practically works).