Re: THEORY: Case stacking; was: Re: THEORY: genitive vs. construct case/izafe

On 7/26/05, tomhchappell <tomhchappell@...> wrote:
> 
> Hello, everyone, and thanks for writing.
> 
> TERMINOLOGY QUESTION: Is "Suffixaufnahme" the same as "case-
> stacking"?

I am unaware of "case-stacking" as a widely-used item of terminology, but as 
it's used here, it appears to refer to constructions in which a noun may 
receive more than one case morpheme -- a group of phenomena which includes 
the various Aufnahme-phenomena. (That is, Suffixaufnahme, the even rarer 
Präfixaufnahme, Tone-Aufnahme, etc. For convenience, sometime *all* of these 
are referred to as Suffixaufnahme.) But there are other case-stacking 
phenomena than Suffixaufnahme.

The definitive work on the subject (Double Case, Planck ed., 1995) gives a 
prototype definition of Suffixaufnahme roughly as follows:

"In what can be recognized as the prototype of Suffixaufnahme there is a 
nominal consisting of a noun or a personal pronoun in a relationship of 
attribution to another nominal, in the basic form that attributive 
constructions take in the language concerned, with the head nominal 
morphologically marked by a case suffix for its external syntactic relation, 
with the attribute carrying the inflectional marking of genitive case, and 
-- crucially -- with the attribute itself in addition separately marked for 
the same case, plus perhaps further categories expressed by suffixes, as the 
head." (p. 50)
 
That is, possessions take both the genitive case suffix and the case suffix 
of their possessors. (This also, in many languages, works with other cases, 
such as the dative or the locative. The prototype specifies genitive because 
it's the most common.) Here's an example from Hurrian:

shen-iffu-ve ashti
brother-my-GEN wife
"my brother's wife"

shen-iffu-ve-ne-zh ashti-zh
brother-my-GEN-SING-ERG wife-ERG
"my brother's wife"

As for recursive attribution -- phrases like "the dowry of the wife of my 
brother" -- different languages behave differently. Hurrian appears only to 
accept one Suffixaufnahme construction per recursive genitive phrase, and 
it's the deepest one. Schematically:

brother-GEN-GEN wife-GEN dowry-DIR

This is a bit odd -- perhaps I shouldn't have chosen Hurrian but it's the 
one I know the best -- but the canonical Suffixaufnahme construction would 
be 

brother-GEN-GEN-DIR wife GEN-DIR dowry-DIR.

When I say "canonical" I really mean "most consistent" -- I don't believe 
that the "canonical" recursive construction is used in the majority of 
Suffixaufnahme languages.

----------------------

Sumerian, on the other hand, exhibits a different sort of case-stacking, 
"Suffixhäufung". This occurs when suffixes are "postponed" until the end of 
the entire phrase -- that is, when the case suffix occurs after the phrase 
rather than the head. This resembles Suffixaufnahme but is really a 
completely different pattern. Here's a Sumerian example:

é dnin Girsu-k-ak-e
house lord Girsu-GEN-GEN-DEM
"this house of the lord of Girsu."

In my analysis, this is perhaps best understood as clitic case postpositions 
(morphological but not phonological words, which must attach to a nearby 
word to be pronounced) attaching to the final word of their phrase. When 
combined with head-initial genitive phrase structures, as above, we get the 
Sumerian pattern of Suffixhäufung.

------------------------

There are other construction that might qualify as case-stacking, as well. 
For example, in Jiwarli (a Mantharta language of Australia, distantly 
related to Warlpiri), the ablative case can only be added to a 
locative-case-marked noun. More strangely, all demonstratives must be marked 
dative before any other suffix (case or otherwise) may be added to them.

----------------------

Of course, if we stretch the definition of Suffixaufnahme to include 
non-case suffixes (which seems reasonable), then some sorts of 
Suffixaufnahme would not be case stacking. Take Elamite, which shows 
aufnahme of definiteness suffixes:

puhu-ri sijan Inshushinak-mi-ra
issue-DEF=ANIM temple Inshushinak-DEF=INANIM-DEF=ANIM

(The DEF=INANIM suffix -mi can be omitted from the noun (temple) itself; 
maybe this then appears to prevent further suffix accumulation on this noun. 
Perhaps; I don't know much about Elamite.)

--------------------

Anyway, hope this clears things up, and inspires some interesting ideas,
-- 
Patrick Littell
PHIL205: MWF 2:00-3:00, M 6:00-9:00
Voice Mail: ext 744
Spring 05 Office Hours: M 3:00-6:00