Re: Natural Order of Events

2009/1/26 David J. Peterson <dedalvs@...>

>
>  (1) Agent = Boy; Patient = Glass; Action = Tilts to mouth
>> (2) Agent = Captain; Patient = Pail; Action = Swings
>> (3) Agent = Woman; Patient = Knob; Action = Twists
>> (4) Agent = Girl; Patient = Hat; Action = Puts on
>>
>
> Now consider the action, and consider the viewer.  In order for
> the action to make sense to the viewer (e.g., why is he holding
> his hand that way?), the viewer should have some idea of what
> the gesturer is holding.  The gesturer is probably going to try
> to make that obvious first so that it will make literal sense that
> he's shaping his hand as if he's holding a glass, not not something
> else.
>
Basically, the gesturer is setting the stage before going on to describe the
action.


>
> This leaves a couple things up to interpretation.  For example,
> what about the subject?  It seems to me that the position of the
> subject with respect to the OV complex is entirely non-crucial,
> and that one might just as well expect OVS order as SOV order.
> This, however, one would *not* expect from an English speaker
> who is accustomed to putting the subject first.  What of a speaker
> of Malagasy?
>
I'm not sure. When it comes to gesturing, I think the idea that one first
"sets the stage" before describing the action would be relatively universal.
Even mimes always do that. I'd expect it is because many active gestures are
actually rather ambiguous when taken out of context (a source of hilarity in
many comedies ;) ). So one needs to "set the stage" first, including all the
actors, before one attempts to describe the action itself, so that the
existing context remove as much ambiguity as possible from the action.


>
> Now about that OV complex.  It should come as no surprise
> that the object has a greater tie to the verb than the subject does
> (semantics has taken that as a given for pretty much ever).
> Certain verbs do so more obviously than others, though, and
> each of the four verbs above--if you consider the *gestures*
> and not the English forms--do so greatly.  To drink something,
> you have to be holding it in your hand, which will affect the
> shape of your hand.  To swing a pail is different from swinging
> a bat (in fact, I bet you couldn't convey "swing" just on its own
> in gestures).  And turning a knob?  Putting on a hat (especially
> if it was a hat with a brim)?  These four processes are so closely
> associated with a *specific* handshape that it's any wonder
> that a speaker could successfully separate the verb and the
> object at all!
>
Actually, I would expect that the gesturer didn't really separate them as
much as first putting their hand or arm in a specific position and then do
the gesture, which the viewer interpreted as presenting the object before
the action.

It's true that in many cases the action cannot be separated from the object
as such. Drinking from a bottle would require a different gesture from
drinking from a glass.


>
> This is not the case with all verbs, though.  As such, I think all
> this experiment says (in so far as the examples I found I
> representative of all the tokens in the experiment) is that certain
> verbs or actions, when gestured, require rather specific actions.
> This should have been obvious before they started.  Just because
> English only has one word for "carry" that doesn't vary based
> on the object carried doesn't mean all languages do!
>
That is a very interesting observation here. In fact, what about spoken
languages? Do people have examples of languages that use different verbs
depending on the nature of the object, while English uses a single one?


>
> In order to really test it, the study should, of course, be cross-
> linguistic, but it should also throw in some verbs that are not
> so obviously and crucially affected by the nature of the object.
> Some ideas:
>
> *send an e-mail (object would be recipient here; might not work...)
> *upload a photo
> *to tell
> *to shout at
> *to fire (as in, to fire an employee)
> *to imagine
>
> This will require a bit more creativity on the part of the gesturer,
> but I bet it'd turn up some SVO gestures.
>
>
>
I'm not sure. I still think that the rule "set the stage, then describe the
action" would still be enforced in most cases, whether the object can be
gestured separately from the verb or not. I would expect that a sentence
like "The boy sends an e-mail" would actually end up being gestured as a
small act, with the gesturer indicating the subject first, motioning to
create a make-believe computer screen and keyboard in front of them, type on
the keyboard for a moment, and then make a hand movement from the
make-believe computer and away from it to indicate the fact that the e-mail
"leaves" the computer.

In fact, if you look at how actions actually happen in real life (not in
mimicry), real life is pretty much SOV: some "actors" are around, and then
an action happens (you can't tell something to someone unless that someone
is already available for being told something. You can't drink from a glass
unless that glass is already there). That is to say, in temporal terms
(which are the way we experience the world), the actors are always present
before the action is taking place. The only exception is when the action
actually creates something, in which case reality can be considered SVO (in
temporal terms). Given this, I'd expect mimicry (as opposed to actual sign
language) to try to stick to the way we experience reality rather than to
the word order of whatever language we speak. The reason is that with
mimicry one tries to represent to the viewer a scene as the viewer would
normally witness it in real life, the best strategy to get yourself
understood if you have no other way to do it.

In other words, I'd expect SVO gestures only if the action actually creates
the object. In fact, I'd say that gestures would relatively universally
follow a temporal-causal type of gesturing order:
Setting-Action-Consequence. The subject (or rather the agent, as a patient
subject might fit more in the Action or the Consequence depending on the
type of action) would most likely be part of the Setting, while the object
could be part of the Setting, the Action itself or the Consequence,
depending on its relationship to the Action itself.

Of course, this is pure speculation on my part. But it doesn make sense,
IMHO.

Now to go back to conlanging, what about a language with a temporal-causal
word order? It might be a candidate for a "primitive" conlang, i.e. a
conlang for a species that has just reached sapience and is only starting to
use abstractions, and thus would likely describe events according to the
temporal-causal order in which they happen. What do you think?
--
Christophe Grandsire-Koevoets.

http://christophoronomicon.blogspot.com/
http://www.christophoronomicon.nl/

Reply