5.1 General remarks
The Udi lexicon has experienced a rather broad documentation. If we cumulate the different soruces (such as Schiefner 1863, Bezhanov & Bezhanov 1902, Dirr 1904, Dzheiranishvili 1971, Gukasjan 1974, and Fähnrich 1999), we arrive at roughly 8.000 lexemes that are attested for this language. Yet, some lexemes (especially in the Gospels (Bezhanov & Bezhanov 1902)) seem to be based on idiosyncratic lexical knowledge - it is not always clear whether such words have a general distribution among Udi speakers. Moreover it should be born in mind that most Udi speakers are bi- or trilingual (Udi + Azeri, Udi + Georgian, Udi + Azeri + Armenian etc.). In consequence, it cannot always be decided whether a speaker treats a word from one of the contact languages as being part of the Udi lexicon, as part of another knowledge system, or simply as being part of his/her general lexical knowledge (note that code switching is a rather common phenomenon among Udi speakers).
The production of ‘new’ lexical structures is based on
three procedures:
a) Metaphorization and metonymy with resepct to
existing lexical structures.
b) Derivational procedures (word formation);
c) Borrowing;
Metaphorization and metonymy may be autochthonous or
contact-induced (calque). Derivational procedures are of two kinds: Referential
entities and qualifying stuctures (adjectives, adverbs) are derived with the
help of word formation affixes, among them:
-ba ~ -bu:
derives qualifying adjectives from nouns (no longer productive);
-la ~ -lu:
derives adjectives from nouns with a strong instrumental or possessive notion;
-en: Historically speaking, this
element represents the ergative-instrumental. It derives adverbial (in parts
adjectival) lexemes with a functional domain that covers the scale
{locative< instrumental < modal < causal}.
-un: this element represents a
metaphrouation of the possessive function of the genetive case marker. In
produces relational adjectives;
-lug/: a loan from Azeri which
denotes abstract referential structures;
-c^’i: a loan from Azeri to
produce nomina agentis.
-suz: a loan from Azeri
functioning as an alpha privativum (not-);
-o: the most productive means
to produce nouns from adjectives, participles etc., cf. 3.2.1.6.
nut’-: the Udi correlate to
the azeri loan -suz, see above.
‘New’ relational structures (verbs) are analytically
formed with the help of light verbs (see 3.3). Azeri verb forms are often
marked by the Azeri ‘perfect’ (inferential) participles (-mis^) to which the
light verb is added.
Borrowing probably
is the most important means to enrich (or to vary) the Udi lexicon. Borrowings
can be both substantial or structural (calques) or a combination thereof.
Today, the most significant and productive layers are represented by Azeri (in Azerbaijan) and Georgian (in Georgia). The actual lexicon Udi
speakers is based on a number of additional but older loan layers of which two
play a major in the formation of the lexicon:
a) Classical Armenian, Middle (East) Armenian,
and Modern East Armenian (all of them as represented
by ‘Azerbaijani’ dialectal variants).
Armenian covers a part of the religious terminology, but is not confined to
this domain. Armenian has been the source for at least two morphemes: plural -(u)x
< Armenian (dialectal) -x < -kc (plural)); te
(subordinator, quotation marker) < Classical Armenian etce
(note that Udi differs from its neighboring languages which have normally
integrated the Iranian (> Oriental) particle ke to cover these
functions).
b) ‘Oriental’: I use the term ‘Oriental’
to describe a lexical (in parts also structural) layer that represents the
common communicative and (in parts) cultural routines people use in societies
that have come into contact with islamic traditions. Most Oriental words (and
structures) stem from Arabic and may be transmitted by either a Turkic or an
Iranian language. Oghuz Turkic and Iranian (esp. Persian) have additionally
contributed to the formulation of the Oriental lexicon. In Udi, Oriental words
that have entered its lexicon since the 8th century are also used to
express religious affairs and religious objects. It is difficult to explain
this fact if we bear in mind that the Udi people are Christians. However, e.g.
the Karaim language (Pontic Kipchak-Turkic still spoken by a Jewish community
in Lithuania) clearly shows that such an intergration of the Oriental
(religious) terminology in the lexicon of a non-Islamic communicaty may well
take place. Perhaps, this process had been supported by Jewish (in parts Tat
speaking) communities in Vartashen which - just as the Karaims - belonged to
the Jewish Karaite tradition: it was just this tradition (founded by Anan Ben
David in the 8th century in Mesopotamia) which promoted the use of
Oriental religious terms in a non-Islamic context.
A loan layer yet not fully understood is represented
by Udi words like k’al-pesun ‘call’, pak ‘garden’, basta
(exclamation in the sense of ‘OK!’, ‘that’s all!’ etc.), port-besun ‘to
carry’ etc. Most probably, we have to deal with remnants of a venecalur related
to the lingua franca of the crussadors that was based in Kilikia:
k’alpesun: ~ 13th
century ‘calare’ (Italian)
pak ~ 13th
century ‘parc’ (French)
basta ~ 13th
century ‘basta’ (Italian)
portbesun ~ 13th
century ‘portare’ (Italien/French)
Other loan layers concern Old Iranian (Medic) and
Northwest Iranian languages (esp. Tat (Sothwest Iranian), perhaps another
Middle Northwest Iranian language).
About 55 % of the Udi lexicon can be related to other
Lezgian or other East Caucasian languages. Lezgian words cover all domains of
the Udi lexicon but do not show a particular (usage-based) distribution. In
fact, all types of economic and social life sho about the same distribution of
loans and autochthonous words. The percentage of Lezgian words in the actual
use of the language is higher because of the fact that most basic deictic and
functional elements are of Lezgian origin.
5.2 A lexical analysis of the sample text
Below I list the distribution of loans and
autochthonous words as present in the sample text (cf. 6.).:
|
Total tokens |
147 |
Lexical basis |
Frequ. |
Meaning (68 concepts) |
Origin |
Etymology |
|
Total types |
102 |
|
|
|
|
|
|
ägänä |
2 |
ägänä |
2 |
if |
Pers. |
agar ‘if’ |
|
aitpesax |
1 |
aitpesun |
1 |
to say |
Pers. < Ar. |
’āyat ‘sign’ |
|
amma |
1 |
amma |
1 |
but |
Pers. < Ar. |
’amma ‘but’ |
|
aq’sa |
1 |
aq’sun |
1 |
to take |
Lezgian |
*aq:’- |
|
aruxne |
1 |
arux |
1 |
fire |
Lezgian |
*c’ay- (pl.tant.) |
|
badallebaki |
1 |
badalbaksun |
1 |
to change |
Pers. < Ar. |
badal ‘change’ |
|
bakaltene |
1 |
baksun |
1 |
to be(come) |
Lezgian |
*-@k- |
|
barsa |
1 |
baresun |
1 |
to let, allow |
Lezgian ? |
|
|
barexa |
1 |
barpesun |
1 |
|||
|
barta |
2 |
bartesun |
3 |
|||
|
bartes |
1 |
bartesun |
||||
|
besai |
1 |
besun |
1 |
to make |
Lezgian |
*-i(’)- |
|
biq’sax |
1 |
biq’sun |
1 |
to seize |
Lezgian |
*-iq:’- |
|
bui |
1 |
bu |
1 |
COP |
Lezgian |
*-’u- |
|
bex |
1 |
bul |
1 |
head |
Lezgian |
*Butl’ul |
|
but’uq’sa |
1 |
buq’sun |
2 |
to love |
Lezgian |
*-utl’- |
|
buvaq’sa |
1 |
buq’sun |
||||
|
burqalle |
1 |
burqesun |
3 |
to start, begin |
Lezgian ? |
*-uRq- |
|
burreqsa |
2 |
burqesun |
||||
|
c^a?xk’axun |
1 |
c^’a?xk’esun |
1 |
to step on s.th. |
Lezgian ? |
? |
|
c^’axpi |
1 |
c^’axpesun |
1 |
to catch |
Pers. |
c^ax ‘fight’ |
|
cipa |
1 |
cipsun |
2 |
to pour our |
Lezgian ? |
*@ciw- ? |
|
cipit’uxo |
1 |
cipsun |
||||
|
dava |
1 |
dava |
1 |
war |
Ar. |
dacwā |
|
dog/rine |
1 |
dog/ri |
1 |
really |
Az. |
dog/ru ‘right’ |
|
dollug/axal |
1 |
dollug/ |
1 |
payment |
Ar. ? |
dawla + lug/ ? |
|
efsa |
1 |
efsun |
2 |
to hold |
Lezgian |
*@Lw- |
|
enefsa |
1 |
efsun |
||||
|
ek’an |
1 |
ek’a |
1 |
what |
Early Udi |
(h)e-k’a |
|
ec^esnest’a |
1 |
esc^un |
3 |
to bring |
Lezgian |
*@c^- |
|
ec^est’a |
1 |
esc^un |
|
|||
|
enesc^a |
1 |
esc^un |
||||
|
ari |
1 |
esun |
1 |
to come, go |
Lezgian |
*eg/- ? |
|
fikirbesax |
1 |
fikirbesun |
1 |
to think |
Ar. |
fikir ‘thought’ |
|
furuk’axun |
1 |
furuk’esun |
3 |
to walk, examine |
Lezgian ? |
* Lw@r@- ? |
|
furuk’az |
1 |
furuk’esun |
||||
|
furuk’san |
1 |
furuk’esun |
||||
|
furuq’unexa |
2 |
furupesun |
2 |
|||
|
g/enax |
2 |
g/e |
2 |
day |
Lezgian |
*yiq:(i) |
|
gölö |
1 |
gölö |
1 |
much, very |
Iranian |
ge:le:k (Kurd.) |
|
haq’ullu |
1 |
haq’ullu |
1 |
clever |
Ar. |
h.aqullu |
|
hödz^ätbesax |
1 |
hödz^ätbesun |
1 |
to quarrel |
Pers. < Ar. |
hudz^dz^at |
|
hörmätenne |
1 |
hörmät |
1 |
honor |
Pers. < Ar. |
h.urmat |
|
ic^ |
9 |
ic^ |
9 |
REFL |
Lezgian |
-@c:^- |
|
ixbaft’e |
1 |
ixbaft’esun |
1 |
to remember |
Early Udi |
*i-x ‘ear-dat2’-baf- ‘fall in ear’ |
|
k’ua |
1 |
k’odz^ |
1 |
house |
Early Udi |
*k’w@-dz^ |
|
k’ul |
2 |
k’ul |
4 |
earth |
Lezgian ? |
? |
|
k’ullul |
2 |
k’ul |
||||
|
ma |
1 |
ma |
1 |
PROH |
Lezgian |
*ma |
|
maate |
1 |
maate |
2 |
where |
Lezgian |
*ma |
|
maslahatbesa |
1 |
maslahatbesun |
1 |
to advise |
Az. < Ar. |
maslahat (+ c^i) |
|
maslahatc^’inen |
1 |
maslahatc^’i |
2 |
advisor |
||
|
maslahatc^’it’a |
1 |
maslahatc^’i |
||||
|
me |
4 |
me |
8 |
PROX |
Lezgian |
*mi |
|
met’a |
1 |
me |
||||
|
met’in |
2 |
me |
||||
|
met’uxo |
1 |
me |
||||
|
met’abaxt’in |
1 |
met’abaxt’in |
1 |
thus |
Pers. + Udi |
baxt ‘sake’ |
|
metärrebsa |
1 |
metärbesun |
1 |
to do so |
Pers. + Udi |
prox + tär ‘so’ |
|
ölkin |
1 |
ölkä |
7 |
land |
Az. |
ölkä |
|
ölkina |
3 |
ölkä |
||||
|
ölkinax |
2 |
ölkä |
||||
|
ölkinaxo |
1 |
ölkä |
||||
|
os/a |
1 |
os/a |
1 |
after |
Early Udi |
*o-s/-a ? |
|
ostavar |
2 |
ostavar |
2 |
strong |
Pers. |
ostovār ‘strong’ |
|
p’a?len |
1 |
p’a?len |
2 |
the two |
Early Udi |
p’a? ‘two’ + sa(?)-erg>com |
|
p’a?lenal |
1 |
p’a?len |
||||
|
paki |
2 |
pak |
2 |
garden |
LF crussad. |
*parc |
|
pasc^’ag/a |
1 |
pasc^’ag/ |
13 |
king |
Pers. |
pās-i-shāh |
|
pasc^’ag/ax |
3 |
pasc^’ag/ |
||||
|
pasc^’ag/axol |
1 |
pasc^’ag/ |
||||
|
pasc^’ag/en |
7 |
pasc^’ag/ |
||||
|
pasc^’ag/un |
1 |
pasc^’ag/ |
||||
|
pasc^’ag/lug/axal |
1 |
pasc^’ag/lug/ |
1 |
kingdom |
||
|
exne |
1 |
pesun |
2 |
pesun |
Lezgian |
*p@- / *@x- |
|
uk’ante |
1 |
pesun |
Early Udi |
uk’-a-n-te |
||
|
q’a |
1 |
q’a |
1 |
twenty |
Lezgian |
*q’@- |
|
q’odz^a |
1 |
q’odz^a |
1 |
old |
Pers. < Ar. |
hodz^a ‘old’ |
|
q’ullug/ |
1 |
q’ullug/ |
1 |
service |
Az. |
gullug ‘service’ |
|
s^et’a |
1 |
s^e |
2 |
DIST |
Lezgian |
*s^i- |
|
s^et’ux |
1 |
s^e |
||||
|
sa |
3 |
sa |
3 |
one |
Lezgian |
*sa- |
|
saganu |
1 |
saganu |
1 |
together |
Pers. + Udi |
sa-ga-n-u ‘in one
place’ |
|
sinamis^ba |
1 |
sinamis^besun |
1 |
to search |
Az. |
si.nami.s^- |
|
t’esahat |
1 |
t’esahat |
1 |
immediately |
Ar. + Udi |
t’e-sahat ‘this hour’ |
|
tag/a |
1 |
taisun |
5 |
to go (to) |
Early Udi |
ta(i)-eg/- |
|
tag/ain |
1 |
taisun |
||||
|
tag/ane |
2 |
taisun |
||||
|
taisan |
1 |
taisun |
||||
|
tanest’a |
1 |
tast’un |
1 |
to give |
Early Udi |
ta-desun ‘put to’ |
|
te |
3 |
te |
3 |
SUB, QUOTE |
Arm. |
etce ‘that’ |
|
tene |
1 |
te |
1 |
NEG |
Lezgian |
*t:e- |
|
un |
1 |
un |
1 |
you (sg.) |
Lezgian |
*g/w@-n |
|
usenaxo |
1 |
usen |
1 |
year |
Ar. |
sinnat (u-?) |
|
va?n |
1 |
va?n |
1 |
you (pl.) |
Lezgian |
*z^w@n |
|
vatan |
1 |
vatan |
1 |
homeland |
Ar. |
wat.an |
|
xabarreaq’sa |
1 |
xabaraq’sun |
1 |
to ask |
Ar. |
xabar ‘news’ (calque) |
|
xois/nebsa |
1 |
xois/besun |
1 |
to wish |
Pers. |
xwāhes^ ‘whish’ |
|
yesir |
4 |
yesir |
5 |
imprisoned |
Ar. |
’asīr ‘imprisoned’ |
|
yesirre |
1 |
yesir |
||||
|
z/omoxo |
1 |
z/omo |
1 |
mouth |
Lezgian |
*dz/om- |
|
zaft’ebsa |
1 |
zaf(t’)besun |
1 |
to rule |
Loan ? |
--- |
|
bez |
2 |
zu |
2 |
I |
Lezgian |
*zw@- |
Note that the Protolezgian reconstructions are only
preliminary. The overall distribution of loans vs. autochthonous lexemes draws
the following picture:
Overall statistics (sample text Yesir Pasc^’ag/):
Tokens 147
Types: 102
Lexial types: 68
Loan layers and autochthonous lexical types:
|
Arabic |
15 |
|
Armenian |
1 |
|
Azeri |
5 |
|
Crussadors |
1 |
|
Early Udi |
8 |
|
Iranian |
6 |
|
Lezgian |
25 |
|
Lezgian (unsure) |
6 |
|
Loan ? |
1 |
|
Total |
68 |
Loan layers and autochthonous tokens:
|
Arabic |
22 |
|
Armenian |
3 |
|
Azeri |
10 |
|
Crussadors |
2 |
|
Early Udi |
12 |
|
Iranian |
24 |
|
Lezgian |
52 |
|
Lezgian ? |
21 |
|
Loan ? |
1 |
|
Total |
147 |
The lexical knowledge of Udi as presented in this text
shows the following general distribution (I disregard the problematic items):
Loans: 29 42.64 %
Lezgian: 39 57.36 %
Loans 62 42.17 %
Lezgian 85 57.83 %
It should benoted that such a calculation is highly text and usage sensitive. In Udi based conversation, the loan layer rises significantly (esp. with respect to Azeri resp. Georgian). For the distribution of loans vs. autochthonous words in the Gospels see Schulze (forthcoming 2001).