Demonstratives

[Koontz John E]

>>From rlarson at unlnotes01.unl.edu Mon Apr  9 09:34:24 2001
> In English, we normally restrict ourselves to a two-level system:
> "this" = "look toward the speaker"; and "that" = "look away from
> the speaker".  The Siouan languages have a three-level system,
> e.g. Lakhota:  le / he / ka, or Omaha:  dhe / she / ga.  Could we
> reasonably gloss these as:  le / dhe = "look toward the speaker";
> he / she = "look toward the listener, or to an item the listener is
> familiar with";  ka / ga = "look away from both the speaker and
> the listener" ?

The 'located relative to you' sense of s^e seems very strong in OP, at
least to judge from the glosses in Dorsey.  Use of the related s^u form as
proclitic with motion verbs to form an intermediate set 'toward you'
opposed to the regular 'to here' and 'to there' forms is also fairly
pronounced.  There are a lot of examples of this for OP because of the
number of letters recorded in OP.  Note that in the context of these
letters s^u definitely can refer to someone out of sight and far away.
It is possible to use du and gu with motion verbs, too, but this is much
less common, and most tokens of these are as conversational pronominals
and directionals, du=akha 'hither-the', du=adi 'this direction', etc.
There's some waffling on whether du/gu need -a- before =di.  I've also
seen gu=di ga hau!  '(go) that way IMPERATIVE'.

For the comparativists, the expected form dhu corresponding to dhe occurs
as a locative suffix, mostly with dhe, e.g., dhedhu=di (often dheu(=di)).
The form used in analogy with dhe in the "u" series is du.  The regular
correspondent of Lakota le would be *ne, not found (dhe occurs instead),
and the analogical *nu is also not found (du occurs instead), so du
doesn't seem to revert to a more regular or less reduced initial than dhe.
There are traces of "o" or "u" vowel demonstratives in other Siouan
languages, e.g., Winnebago.

The main s^u-motion verb instance that I noticed in fieldword (s^u=dhe 'to
go toward you') occurred when there was a power failure at the school that
left the interior room full of elders where I was working completely
blacked out.  The supervisor (WW) of the group stood up and maneuvered
through the darkened room to open the door, announcing as he came
"s^u'=bdhe!  s^u'=bdhe!" or 'I am going to you, I am going to you'.

I've also noticed that the s^e demonstrative is common as story enders,
e.g., s^e'=naN 'so many wrt you', s^e'=thaN 'so far wrt you'.  I take this
is a sort of listener orientation or involvements strategy with the
implicit references being to the words or incidents of the story and to
the progression of the plot or story cycle.  This is from Dorsey,
primarily, not personal experience.

I'm inclined to suggest that in the OP context s^e/s^u is the added
axis, as it were, and that the basic spatial references are dhe and ga.

I noticed in eliciting demonstratives that Omaha speakers felt no need to
map dhe/ga to the English set in my stylized way.  I was told flatly that
ga was this and dhe was that.  I fell back in disarray for the moment and
later concluded without having a chance to investigate further that the
pragmatics of using the OP demonstratives were sufficiently distinct from
those of English that speakers did not see the analogical mapping that
linguists did.  I had already noticed a similar disjunction between
linguistic glossing of motion verbs and speaker glossing in translations.
Putting it another way, linguists gloss these terms the way they do
because they see some general analogy between this/that and dhe/ga (etc.),
but I hypothesize that speakers know that in practical use dhe maps to
both this and that and ga to both this and that, depending on the context
and lack any presumption, for example, that dhe is primarily or
prototypically equivalent to this.

Or, on the other hand, it might have been a shifter problem:  my 'this'
was the speaker's ga.

> If so, would this be the general rule for three-level demonstrative
> systems?  (If this has all just been discussed, I apologize-- I just
> got on the list and haven't gone through the archives yet.)

It's been alluded to in passing only.  Actually the basis distinctions
among terms in demonstrative sets with more than one terms is fairly
variable, though proximal vs. distal is common for two terms.  A third,
more distal term might involve person, distance, visibility, time, death,
or other factors.  I'm afraid I'm not actually able to suggest any
typological studies.  Anyone?