tgivon at OREGON.UOREGON.EDU
Thu Mar 8 04:49:21 UTC 2001
Dear Brian, I'm forwarding your note to Scott, together with some
There is a similar passive case in English, the GET passive. Unlike the
BE passive, it tends to (in older working-class people, where it is most
entrenched) to carry a strong sense that the patiend did something
deliberately in order to deserve their misfortune. There's a paper by
Young & Givon tracking the history of this passive (Hopper & Fox eds
1994), but also showing why its history (as an ex reflexive- causative)
is related to its saynchronic characteristics. The 'agentivity' of the
patient of this passive was noted earlier by Robin Lakoff & Dwight
Bolinger, no doubt *inter alia*.
The reason why Scott is down on agents probably has to do wiyth is "best
theory" approach (the 'localistic case hypothesis'), which I myself find
less-than-useful in accounting for case-governed grammatical phenomena
in most languages I've worked with. But given that Scott works on
Tibettan, where the notion of intentionality is rather important in both
the drift to active/stative case-marking and the interaction with the
aspectual/modal system ('mirativity'), I am not sure why he says what he
says. If you ask people who woirk on other active-stative languages
(Kartvalian, Chictaw, Lakota, Iroquois) if they need 'agent' to account
for what they see in their languages, you'll get a resounding 'yes'
(there's a recent paper by Mithun & Chafe on this subject). So I see no
way of getting rid of it.
But of course, both cognition and folk psychology (not to mention the
study of animal behavior) suggest that a notion such as "action under
one's own motivation" is a crucial feature for categorizing moving
objects in an adaptive, predictive way. So one does not need grammar to
arrive at this.
Finally, many, perhaps most, cultures have a concept of culpability that
depends crucially on intentional action and having control. So the
category is certainly all over the place.
Brian MacWhinney wrote:
> Dear Tom,
> I tried posting this, but I am subscribed to Funknet from CMU and not Hong
> Kong, so it bounced. Anyway, maybe it is fine to just have you and Scott
> read it without a full posting. Unfortunately, I also don't have Scott's
> email, so maybe at least you can read my thoughts on this.
> Dear Tom and Scott,
> Could you please be more specific? I assume that Scott is saying that we
> choose Agents solely on the basis of whether or not we perceive them as
> having instigated an action, whether that action be intentional or not.
> However, Hopper and Thompson and many others have claimed that marking of
> transitivity prototypically codes volition. But perhaps Scott is saying
> that this only affects the nature of the object, not the Agent.
> I would have wanted to just jump in agreeing with Tom, but then I realized
> that I didn't really understand what he was saying either. Why shift ground
> from Scott's focus on the Agent role to the issue of "rules of grammar". If
> we just keep the focus on the linguistic marking of agency, can't we still
> safely conclude that intentionality or volition is still crucial for
> activation of the role. Consider this example from the news yesterday.
> Dick Cheney's doctor was discussing his case and asserted that "Patients who
> narrow their arteries end up with repeated chest pains." Now I am forced to
> imagine Dick Cheney sitting in bed focusing on trying to narrow his
> arteries. Perhaps he does this through some form of Tai Chi meditation.
> This misinterpretation of volition is induced by the grammatical requirement
> to package sentences with Agent roles. However, the opposite also occurs.
> Insurance company reports are full of sentences like, "Suddenly this
> confused old man appeared on my windshield."
> I would love to see a discussion of at least some of the detailed ways in
> which perception of intentionality or lack of intentionality ends up shaping
> the nature of language. Not being able to speak an ergative language, I
> have always worried about that too. And don't the two Finnish passives
> involve relative levels of attribution of intention? Can't linguistics
> articulate a detailed view of the specific processes of encoding of
> intention that is more grounded in discourse and everyday life than the
> rather ethereal statements we find in Searle, Dennett, and the others. If
> so, what might this vision look like? I know that Len Talmy and Ron
> Langacker include intentional arrows in many of their diagrams, but don't we
> need to wonder where the arrows come from? What induces us to infer
> volition or avoid the inference in our linguistic packagings of experience?
> --Brian MacWhinney
More information about the Funknet