Re: ergative/accusative

MorphemeAddict@WMCONNECT.COM wrote:
 > In a message dated 1/27/2007 8:10:36 PM Central Standard Time,
 > schlaier@SBCGLOBAL.NET writes:
 >
 >
 >
 >>are there any languages (natural or otherwise) that treat agent and
object
 >>the same?
 >>maybe i just havent been paying attention and missed it
 >
 >
 > What kind of object?

Quite so. The category 'object' is somewhat vague, being "a generic term
for any noun phrase other than subject" [Trask].*

Therefore, yes, ergative languages treat agents as objects.

*Subject, however, is far less easy to define   :)

[snip]

 > So, is your object an indirect object, a patient, a focus, or what?

My _guess_ is that schlaier@SBCGLOBAL.NET meant 'direct object', but I
am not certain. As the direct object is "[t]he *grammatical relation*
borne by an NP which occurs inside a verb phrase and which is the second
obligatory argument of a transitive verb, most typically expressing a
patient which undergoes the action of a the verb....." [Trask], the
question would then apply only to transitive verbs. So I am not really
certain what the original question meant.
================================================

MorphemeAddict@WMCONNECT.COM wrote:
> In a message dated 1/27/2007 8:29:59 PM Central Standard Time,
> tepples@SPAMCOP.NET writes:
>
>
>
>>J. B. Rye wrote a "Ranto", or a rant against the promotion of Esperanto.
>>One appendix to the Ranto discusses the case marking systems in various
>>languages, and he mentions what he calls "the monster raving loony candidate
>>(some Iranian sightings)", but he does not identify the specific language.
>>It's not Farsi, is it?
>>http://www.xibalba.demon.co.uk/jbr/ranto/r.html
>>
>>
>
>
>>From that site:
>
> Agent = "Subject" of a transitive verb ("we saw Sam")
> Experiencer = Argument of an intransitive verb ("we waited")
> Patient = "Object" of a transitive verb ("Sam saw us")
>
> I think Rye is using the labels wrong.
IMO you are quite correct.

Agent, Experiencer, Patient etc are *semantic roles*; this is to do with
_meaning_ and denotes the semantic relation that a NP bears to its clause.

But 'subject' and 'object' refer to the *grammatical relation* a NP has
with its clause. It know of no natlang where the categories of semantic
role and grammatical relation have 1 to 1 correspondence as Rye
apparently gives above.

The category 'semantic role' may also be known as 'semantic function',
'theta role' or, in Case Grammar theory, as 'deep case.' But it is quite
clear that Rye is using case to denote grammatical surface cases and not
'deep case'.

> An agent is the subject of a transitive *action* verb ("we ate the food")
Probably so in _English_. The Latin verb 'uapulare' certainly conveys
the notion of action, one of a beating or flogging taking place. But the
grammatical subject is certainly not the agent, e.g. Marcus uapulat =
Marcus is getting a flogging.

Certainly in Rye's example "we saw Sam", 'we' is not the agent. The
exact role of the English subject of verbs of perception like 'see' and
'hear' is IIRC one of the questions the Voksigid developers couldn't
agree on and, therefore, one of the reasons the language never got
completed.

> The experiencer and the patient are the same thing,
I think not. Let me quote Trask.

Experiencer
"The *semantic role* borne by a NP which expresses the animate NP who is
the passive recipient of a sensation or a mental experience, such as
_Lisa_ in _Lisa has a head ache" and "Lisa is happy."

Patient
"The *semantic role* borne by an NP which expresses the entity
undergoing an action, such as _the roof_ in _I've repaired the roof_ or
_The roof collapsed_."

[snip]
> His example of a patient is wrong too.  The subject "Sam" in "Sam saw us" is
> the patient, and "us" is the focus.
I do not think "Sam" is the patient - looks more like experiencer, to
me. But I'll not argue here what the semantic roles of the English
subject & direct objects are of the verb "to see".

It is possible 'us' is the focus (i.e. new information') - but 'focus'
is neither a semantic role nor a grammatical relation and, without the
sentence being in context, we cannot be certain whether it is the focus
or not. Focus is normally understood to mean: "Special prominence given
to some element in a sentence which represents the most important new
information in the sentence or which is explicitly contrasted with
something else. In English, focused elements are frequently marked only
by stress, though *cleft* constructions are sometimes use." [Trask]

I would expect 'focus' to be discussed in connexion with 'topic' (theme)
and 'comment' (rheme). It is seem to me that introducing it in a
discussion about semantic roles and grammatical relations is confusing.

[snip]
> object), or state verbs with patient as subject and focus as object.  No second
> objects with state verbs.
> Possible verb argument structures are:
> active verbs:
> Agent + Patient
> Agent + Patient + Focus
>
> state verbs
> Patient
> Patient + Focus
Surely other arguments are possible with both verbs, and 'focus' is not
a separate argument. It is possible (quite common, I think) to focus an
agent or a patient; indeed, the verb itself could be the focus.

I agree that Rye is using labels incorrectly. He is IMO confusing
semantic roles and grammatical relations. But, similarly, the label
'focus' is surely being used incorrectly by MorphemeAddict.

--
Ray
==================================
ray@carolandray.plus.com
http://www.carolandray.plus.com
==================================
Nid rhy hen neb i ddysgu.
There's none too old to learn.
[WELSH PROVERB}

Reply