A Tour of Proto-Nostratic grammar
=================================

Up to now, Nostratic linguistics has been mainly occupied with the
compilation of vast amounts of lexical cognates. The best known
efforts are Illich-Svitych's "Opyt srvnenija nostraticheskix jazykov"
and Bomhard and Kerns' "The Nostratic Macro-Family".

In the following, I'm going to ignore the lexicon almost completely.
Instead, I'll try to reconstruct the possible grammatical system of
Proto-Nostratic, as it can be inferred from the languages presumed to
be Nostratic. I will be focusing on Afro-Asiatic, Kartvelian,
Indo-European, Uralic and Eskimo-Aleut (with some excursions into
Basque, Sumerian and Etruscan), not because I think that for instance
Altaic (or "Altaic") and Dravidian are not Nostratic, but because I
don't have the necessary material and knowledge to go into these
language families with confidence. Which is not to say that I
consider myself to be a specialist in the langauges and families
that I *will* be discussing.

We'll start our tour in Kano, Nigeria, work our way through Eurasia,
to end up in Ultima Thule.

HAUSA

Hausa is the most important member of the Chadic subdivision of
Afro-Asiatic. It is spoken by over 20 million people in Northern
Nigeria and surrounding areas.

HAUSA PHONOLOGY

p b 'b f m w
t d 'd s z ts' n l r rr
c j sh y 'y
k g k'
? h

a a: e e: i i: o o: u u:

tone: high (unmarked), low `, falling ^


HAUSA PERSONAL PRONOUNS

The independent (subject) forms of the Hausa personal pronoun are:

1. ni:
2m. kai
2f. ke:
3m. shi:
3f. ita
1p. mu:
2p. ku:
3p. su:

Hausa uses a large set of so-called "person aspect pronouns" to
express the subject and the aspect (completive, continuative,
subjunctive, etc.) of the following verb. These forms are the result
of merger of the personal pronoun and aspectual/mood markers.
It would be too space-consuming to list all the different forms, so
I'll only list the perfective person-aspect pronouns:

1. na:
2m. ka:
2f. kin
3m. ya:
3f. ta:
1p. mun
2p. kun
3p. sun

Finally, the direct object and the separable possessive forms (masc.
possessed):

1. ni nà:-wa
2m. ka na:-kà
2f. ki na:-kì
3m. shi na:-sà
3f. ta na:-tà
1p. mu na:-mù
2p. ku na:-kù
3p. su na:-sù

A prelimnary analysis yields the following inventory:

1. *ni *wa
2m. *ka
2f. *ki *kin
3m. *si *sa *ya
3f. *ta
1p. *mu *mun
2p. *ku *kun
3p. *su *sun

The independent pronouns seem to contain an extra element *-i (sg.) or
*-u (pl.)

1. ni: < *ni-i
2m. kai < *ka-i
2f. ke: < *ki-i or *kin-i (?)
3m. shi: < *si-i
3f. ita < *i-ta (?)
1p. mu: < *mu-u
2p. ku: < *ku-u
3p. su: < *su-u

The second and third person pronouns have the same root in singular
and plural (*k-, *s-), but the first person pronoun has sg. *n and
plural *m. It would be nice if we were able to reconcile the two, and
there are to equally likely possibilities to do so:

(A)- *n is original and plural /mu/ arose by assimilation from *nu;
(B)- *m is original and sg. /ni/ arose by assimilation from *mi.

Within the Afro-Asiatic context, it makes sense to opt for possibilty
(A), but I will opt for possibility (B). This may perhaps also
explain the 1sg. possessive /wa/ < *mu-a.

The third person pronouns *si/*sa pl. *su(n) offer no problems from an
Afro-Asiatic point of view. The 3f. *(i)ta of course reflects the
feminine marker *-at(a). The 3rd. person sg. marker /ya/ is
remarkable within Chadic, but is a clear parallel of the ya- found
elsewhere in the Afro-Asiatic (Berber, Semitic, Cushitic)
3rd.p. prefix conjugation form *ya-.

The next question concerns the -n that is seen in the 2sg. feminine
and in the plural forms /mun/, /kun/, /sun/. On comparative grounds
(Berber 2f. indep. pn. /k&m/, Egyptian 2f. encl. /-t_m/ < *-kim) it is
clear that in the 2f., the -n goes back to *-m. It is equally clear
that in the plural, the -n is identical to the plural morpheme -n
which appears in Egyptian -t_n, -sn or Semitic -kun, -s^un.

Whether the plural forms without -n are secondary or they represent a
very old plural pattern based on u-vocalism alone is an open question.
In Semitic, or Berber, the -u (pl. -u:) is a subject marker (thought
to derive from an ergative), and this -u is lacking in Hausa, which
may imply that it was lost in the singular, and remained as a
secondary plural marker. On the other hand, the Basque personal
pronouns show the same i/u pattern as Hausa, which may imply that the
pattern is indeed very old.

EXCURSION ON THE BASQUE PERSONAL PRONOUNS

indep. absol. absol.past ergat./dat. erg.past 3rd.p.obj.
1. ni na- nin-/nen- -t, -da- n(e)-
2. hi ha- hin-/hen- -k, -ga- h(e)-
2f. -n, -na-
1p. gu ga- gin- -gu gen(e)-
2p. zu za- zin- -zu zen(e)-

Note the pattern sg. ni/hi vs. pl. gu/zu. Note also the plural -n in
the past ergative (with 3rd. person object) prefixes.

The forms can be reconstructed as follows:

1sg. *ni ~ *na. The ergative/dative suffix was denasalized to /da/.

2sg. *ki ~ *ka. The initial forms with h- can in principle point to
any initial voiceless consonant (*p-, *t-, *k- all yield 0-/h- in
initial position), but only *k would have given medial -g-.

2sg.fem. (dative/ergative only). Medial -n- can here come from *n or
*m. The most interesting option is *-m- (cf. Afro-Asiatic 2f. *ki-m,
Berber suffix -m), as it explains the denasalization of 1sg. *-na-.
When *-ma- became -na-, the first person form *-na- was denasalized.

1pl. *gu ~ *ga ~ *gen. As in Hausa, both the -u and the -n plural
patterns can be seen. The relationship between singular *n- and
plural *g- is unclear, but one can think of a compund 1pl. inclusive
pronoun *n-ku > gu.

2pl. *su ~ *sa ~ *sen. Initial z- (pronounced /s-/) can reflect
initial *s- or *d-, but only *s would have yielded medial -z-. Again,
the relation between 2sg. *k- and 2pl. *s- remains unclear. In the
light of Afro-Asiatic 3pl. *su(n) one might consider another compound
pronoun *k-su "you and them".


HAUSA NOUNS

Nouns are not inflected for case in Hausa. There is a feminine gender
with marker -a (-a:).

Hausa is an SVO language, unlike VSO Berber, Egyptian and Semitic. As
expected, in Hausa we find mostly prefixes instead of suffixes.

Pronouns have a prefix ma- to denote the indirect object (dative).
The indirect object prefix for nouns is mà- or wà-.

Location is expressed by the prefix à (à Kanò: "in Kano").
The comitative prefix is dà (dà shi: "with him"), the ablative dàgà
(dàgà Kadunà: "from Kaduna"). The allative prefix is gà (gàre: before
pronouns).

The most important particle is the relational particle na (fem. ta),
which is used for expressing possession (gida: na sarki: > gidan
sarki: "the house of the chief"), to connect adjectives to their head
noun (babba-n gida: "the big house"), to make relative clauses, and to
link a verbal noun with its object (mu-nà: harbì-n na:mà: "we are
hunting wild animals", literally: "we-CONTINUOUS hunting-of
wild-animals").

The feminine form ta (< *-t-na) provides evidence that the Hausa
feminine marker -a was originally *-at (sa:niyar Audù "Audu' cow" <
sa:niya: ta Audù < *sa:niya:t na Audù.

HAUSA PLURALS

Hausa has a large number of ways to make plural nouns. We can
basically distinguish between internal plurals and external plurals.

Some examples of internal plurals:
Class 1. [-o:Ci:] awà: -> awo:wi: "hour"; dabbà: -> dabbo:bi:
"animal"
Class 3.B. [-à:Ce:] bak'i: -> bak'à:k'e: "something black"
Class 5.A. [-à:Ca:] k'arfè: -> k'arà:fa: "metal"
Class 5.B. [-à:Cu:] k'afà: -> k'afà:fu: "leg"

The patterns mainly consist of insertion of a long vowel (-o:-, -a:-)
after the last or second root consonant, followed by reduplication of
the final consonant or by the third consonant, followed by a long
vowel (-i:, -e:, -a:, -u:).

Some examples of external plurals:
Class 2.A [-unà:] bà:ki: -> ba:kunà: "mouth"
Class 2.B. [-ukà:] kâi -> ka:yukà: "head"
Class 2.C. [-uwà:] kâi -> ka:yuwà: "head"
Class 3. [-à:ye:] 'be:ra: -> 'be:rà:ye: "mouse"
Class 4.A. [-ai] àbo:ki: -> àbò:kai "friend"
Class 4.B. [-u:] na:mà: -> na:mu: "wild animal"
Class 4.C.1. [-i:] bàk'e: -> bà:k'i: "guest"
Class 4.C.2. [-ki:] kwa:na: -> kwà:nàki: "day"
Class 4.C.3. [-nni:] ùba: -> ùbànni: "father"
Class 4.D. [-(n)au] kàre: -> karnau "dog"
Class 6. [-a:] màcè -> ma:ta: "woman"



=======================
Miguel Carrasquer Vidal
mcv@...