Origin of nominalising morphology.

[Vit Bubenik]

----------------------------Original message----------------------------
Re:

On Sun, 15 Jun 2003, Guy Deutscher wrote:

> ----------------------------Original message----------------------------
> Dear Histling-ers,
>
> I have a question about the origin of nominalizing morphology, which I
> hope someone can help with. What prompts the question is the wish to
> understand how infinitives develop in language.

Akkadian -- the earliest recorded Semitic language -- with its plethora of
infinitives (10 according to von Soden) and verbal adjectives might be a
good starting point. The basic infintive para:sa-um and the factitive
infinitive purrus-um/parrus-um share their vocalic patterns with primary
nouns and adjectives: ata:n-um "donkey (female)" (but in Akkadian this
nominal pattern is isolated) and dunnun-um "fortified", kubbur-um "thick,
fat", kuSSur-um "crippled", sukkuk-um/tummum-um "deaf", respectively.
Here the question of priority, i.e. whether the nominal pattern is imposed
on the verbal one or vice versa, will bring us to the chicken-or-egg
argument. But the story is different with non 'non-basic' infinitives:
the causative infinitive s'u-prus-um/s'a-prus-um (Assyrian) has no
adjectival counterpart (there is only an isolated  Old Babylonian/poetic
s'a-lbub-um "wild");
the passive infinitive na-prus-um has been related to the vocalic pattern
of nomina loci as ma-s'kan-um "place" (and other nouns such as na-ra:m-um
< na-r?am-um "beloved"); its -u- can be traced back to words such as
ma-qlu:-um "burning". Assuming some satisfactory solution of the
initial n- (infinitive) vs. m- (nomen loci) one might argue for the
priority of the nominal patter.

In Biblical Hebrew the basic infinitive qa:To:l (Proto-Semitic a-a:, cf.
Akkadian para:s-um) shares its vocalic pattern with primary adjectives:
?a:do:m "red", ka:Ho:l "blue", ya:ro:q "green" (and some other nouns);
here again the question of priority is elusive.


 But there is another half. In order for those allative or
> purposive markers to appear before a verb, the verb usually has to have
> some nominalizing affix. In other words, to get the infinitive, we need to
> start with some 'action nominal' or participle of some kind. But where
> does the morphology which takes a verb and turns it into an 'action
> nominal' come from in the first place? One would expect that it arises by
> grammaticalization, and ultimately from some lexical source. But what are
> these sources? ... In theory, therefore, one way for
> nominalizing morphology to reach verbs would be by extension of such
> affixes from nouns to verbs. But again, are there examples where one can
> actually see such a process in action ?

Here are some observations on Hittite which possesses two infinitives (in
-anzi and -anna) and the verbal noun in -war (e.g. esu-war "being",
sesu-war "sleeping"). This suffix is also found with two primary nouns
asa-war "fence" and parta-war "war"; the difference, however, appears in
their oblique cases (e.g. the genitive of the infinitive is -mas [esu-mas
"essendi" in Latin] vs. the nominal suffix -nas [partau-nas "of the
wing"]). I am not sure what to make out of this. The other two
infinitives have solid IE counterparts in *-ti > zi [tsi] and PP
adjectival/partcipial *-no.

Vit Bubenik