Augment (was Re: German ge- ptcpl cognates?)

[Robert Whiting]

On Thu, 10 Feb 2000, Miguel Carrasquer Vidal wrote:

> "Vidhyanath Rao" <rao.3 at osu.edu> wrote:

>> From: "Miguel Carrasquer Vidal" <mcv at wxs.nl>
>>> Still, the unmarked form is a simple past, while the marked forms
>>> are the imperfective ("durative", "present-future") with
>>> geminated C2, and the perfect (CtCC [iptaras], with infix -t-).
>>> Such a system is potentially very close to one with unmarked past
>>> vs. marked present (all it takes is the loss of the perfect).

>> Is it s a simple past or narrative past? [zero forms do survive as
>> subsequent forms even when they have been ousted from isolated
>> sentences, conversation etc.]

> I don't know much about Akkadian syntax, but what I gather is:

> The preterite (iprus) is the unmarked narrative past.

This is generally true.

> The perfect (iptaras) is less frequent.  According to Lipin'ski
> it denotes "that a state is produced in someone or in something,
> whether it be caused by another or by himself/itself".  The -t-
> infix in other Semitic languages (as well as in Akkadian modal
> forms) denotes a reflexive (Ugaritic yr-t-HS "he washed himself",
> preterite with t-infix).

Avoid Lipin'ski.  At least as far as Akkadian is concerned (I don't
have the competence to judge his treatment of the other languages).
Lipin'ski gives no footnotes so you do not know whose opinion he is
basing his description on and in most cases, what the evidence is
for the position he takes.  But even worse, he glides over controversial
points without even indicating that there is a controversy.  Only one
point of view is ever expressed and there is no way for the uninformed
reader to know this or to know what the relative merits of the
unmentioned positions are.  For Akkadian use the most recent edition
of von Soden's grammar.

In classical Old Babylonian the perfect expresses an action that is
subsequent to some previous action.  In narrating the past, an action
that took place in the past is normally expressed with the preterite;
subsequent actions (usually ones that are dependent on the first action
or result from it) are expressed by the perfect (coordinate
actions will still be expressed by the preterite).  Since the perfect is
used to express subsequent action, it can also be used after a present/
future as a future perfect.

In later periods (Neo-Assyrian and Neo-Babylonian) the distinction
disappears.  In main clauses only the perfect is used to express the
past.  The preterite is now restricted to subordinate clauses.

The use of iptaras quoted from Lipin'ski is not really the perfect (it
doesn't have to do with time), but a t-stem form used as a middle verb.
There is both a stem-forming -ta- infix and a tense-forming -ta- infix.
The former results in reflexives and reciprocals to the basic stem (G
stem) and in passives to the derived stems (D and $ stems).  Sometimes
it is used to deny the participation of an agent in the action (i.e.,
it happened all by itself; cf. R. Reagan:  "Mistakes were made").

> I don't know to what extent the imperfective (iparras,
> "present/future") was used in past tense contexts.  Judging by
> its traditional name, not often.

The present/future is often used to express a durative in the past.

> There is also the Akkadian stative (paris), which is is the
> normal perfective / past tense in other Semitic languages (having
> ousted the preterite), but which in Akkadian is a true stative,
> i.e. a verbal adjective (paris "he is separate"(?)).

The Akkadian stative is indeed a verbal adjective and in the basic stem is
the least marked form (in derived stems the stative/verbal adjective has
the same form as the infinitive) in the tense system.  But the form is not
unmarked as the vocalization (CaCiC) is a marking (the unmarked form of
the verb is the basic stem imperative).  I always explain the stative as
the absolute form of the verbal adjective to which the bound forms of the
nominative personal pronouns are added (the bound form of the third
person masculine singular being 0).

The stative makes no reference to time and is often used as a permansive
(particularly in geographical descriptions:  $umma a:lu ina me:le $akin
'if a city is situated on a hill').  In stative verbs the stative simply
expresses the existence of the state (damiq 'it is good'); coming into the
state at some point in time (inchoative/ingressive) is expressed by the
present/future or preterite.  In intransitive action verbs, the function
of the stative gets blurred; it often expresses an action that was going
on or a state that existed when some other action took place or an action
that took place over a certain period of time (often translated by a past
or present progressive tense).  In transitive verbs, the stative is often
translated as a passive (the verbal adjective functions as a passive
participle); thus paris 'it is decided', Sabit 'he is captured'.

> Campbell ("Compendium of the World's Languages") says:
> "Instead of the typical Semitic division into perfective and
> imperfective aspects, Akkadian has an idiosyncratic quadruple
> segmentation which corresponds broadly to a present/
> preterite/perfect system, with the fourth memeber acting as a
> kind of stative".

Formally, the Akkadian stative corresponds to the West Semitic perfective
(and the Egyptian so-called "old perfective") and the Akkadian preterite
corresponds to the West Semitic imperfective.  The Akkadian present/
future is not represented in West Semitic (although Ethiopic has a similar
form) and West Semitic has stem-forming -ta- forms but of course no tense-
forming ones.

> Diakonov (in EB), contrary to Lipin'ski, seems to say that the
> preterite was in origin a perfective (opposed to the iparras
> imperfective).  "Later a new "perfect" with an infixed -ta- in
> the stem developed".

It is hard to say whether iparras is an innovation of East Semitic or was
original and lost in West Semitic (the Ethiopic form makes a decision
difficult).  A plausible case could be made for iprus and iparras once
having been the same form with the outcomes being the result of
differences in stress.  If so, then the 'old perfective' became the
stative and simply dropped out of the tense system (it is not specific
with regard to time).  But the sequence presented by Diakonov is correct.
Old Akkadian has iparras as does Eblaite, but there are no clear examples
of tense-forming -ta- (although stem-forming -ta- is present).
Unquestionable examples of tense-forming -ta- do not appear until archaic
Old Babylonian (around 2000 BC).

Bob Whiting
whiting at cc.helsinki.fi