Nights and Time

[Koontz John E]

On Thu, 22 Mar 2001, Wablenica wrote:

> I wonder  what is the etymology of root haN- "night", suffix -haN "time" and
> continuative/imperfective enclitic haN~he~hiN in Dakotan. Are these the one
> and the same morpheme, do they all originate from haN' "for a tall object to
> be located" ?
> Examples:
> haNhe'pi (L.) / haNye'tu (D.) night
> haNcho'kayan - midnight

I think that the combining stem here is haN- 'night'.  But it looks like
the independent stem is either haNh(A) (as in haNhe=pi), or haNy(A) as in
haNye=tu.  The first alternative looks like an an ablauting epenthetic a
with a C-final stem haNh-. There is a real problem here with =pi, however,
as I think this always conditions the a-grade, right?  This looks a
nominalizing =pi, perhaps, as in thi=pi, but I've never heard that this
behaved any different from the pluralizing =pi.  In fact, I've always
assumed they were the same, though I don't necessarily understand the
connection.  So maybe haNhepi is haN + he + =pi with he being some
morpheme that doesn't ablaut.

For comparison with haNyetu consider on the "extension" wiN- vs. wiNyaN
(with -yA nasalized by iN?), iN- ~ iNyaN, and he- ~ heya.  This sort of
looks like an abluating "epenthetic" a attached to a monosyllable and
separated from it by an epenthetic y.  For the -e=tu part of the second
alternative consider aNpe=tu 'day'.

In OP haN (invariant) is 'night', but there is one old compound
haNhewac^hi 'night dance', that might suggest *haNhe, which is why I
wonder about the haNhepi alternant in Dakotan.  Still the -e=pi in Dakota
is a bit of a problem for this comparison.  Does anyone know of another
-he element in temporal adverbs, etc.?

> haNblA' (haNwa'ble)- to fast and dream or attain vision

This stem presumably incorporates haN- 'night' and occurs pretty
consistantly across at least Mississippi Valley Siouan.

> he'haN, hehaN'l - at that time
> hehaN'taNhaN - from that time on; therefore
> hetaNhaN - from that place/time

In these elements I think =haN is something else; basically a sort of
postposition.  It corresponds to Omaha-Ponca =thaN, which tends to be
associated with extents, e.g., e=di=thaN and e=tta=thaN 'from there' (in
various senses).  The correspondence of OP aspirated t to Da h is regular.
The classic example is OP (a)thi 'arrive here' vs. Da (a)hi 'arrive here'.
Another example is OP (wa)c^hi vs. Da (Santee?) hu, but I can't gloss this
for fear of being banned in Switzerland.

=haNl is from probably from =haN=t(u), a sort of reversal of the *tu=thaN
underlying OP =di=thaN.  =taNhaN might be a direct match for OP =tta=thaN.
The tta is probably from *-k-ta-, as I think I mentioned recently.  So
*-k-ta=thaN => OP =tta=thaN, and just *-ta-thaN => Da -ta-haN with =taNhaN
involving nasalization across h, perhaps.  -haNtahaN looks like a piling
up of =haN + =tahaN.

JEK