Edited by  H. B. Paksoy

Table of Contents:

H. B. Paksoy "Ibadinov's Kuyas Ham Alav"
Peter B. Golden (Rutgers) "Codex Comanicus"
Richard Frye (Harvard) "Narshaki's The History of Bukhara"
Robert Dankoff (Chicago) "Adab Literature"
Uli Schamiloglu (Wisconsin-Madison) "Umdet ul Ahbar"
Kevin Krisciunas (Joint Astronomy Centre) "Ulug Beg's Zij"
Audrey Altstadt (UMass-Amherst) "Bakikhanli's Nasihatlar"
Edward J. Lazzerini (New Orleans) "Gaspirali's Tercuman"
David S. Thomas (Rhode Island) "Akcura's Uc Tarz-i Siyaset"

ISBN: 975-428-033-9
Library of Congress Card Catalog: DS329.4 .C46 1992
173 Pp. (paperback)  US$20 

ISIS Press

Isis Press
Semsibey Sokak 10
81210 Beylerbeyi Istanbul
Phone: +90 + 216 321 38 51
FAX: +90 + 216 321 86 66


Please refer to the printed version for the footnotes

The Codex Cumanicus
Peter B. Golden

     From the time of the appearance of the "European" Huns until
the collapse of the Cinggisid khanates, the Ponto-Caspian steppe
zone and as a consequence, to varying degrees, the neighboring
sedentary societies, have been dominated by or compelled to
interact intimately with a series of nomadic peoples. Although
Scythian and Sarmatian tribes of Iranian stock had held sway here
for nearly a millenium before the coming of the Huns and Iranian
elements both in their own right and as substratal influences
continued to have an important role in the  ethnogenesis of the
peoples of this region, the majority, or at least politically
dominant element, of the nomads who became masters of these rich
steppelands were Turkic. In the period after the Turk conquest of
Western Eurasia in the late 560's, until the Cinggisid invasions,
the Turkic polities of the area all derived, in one form or
another, from the Turk Qaganate.
     Of these peoples, only the Khazars, the direct political
successors of the Turks, produced a qaganate in the classical
Turkic mold. The others remained essentially tribal
confederations which, for a variety of reasons, did not feel the
impetus to create a sturdier political entity, i.e. a state.
Those that were driven from the area into sedentary or semi-
sedentary zones, such as the Hungarians ( a mixed Turkic and
Ugrian grouping under strong Khazar influence ) and parts of the
Oguz, under Seljuq leadership, did create states but along
largely Christian (Hungary, Danubian Bulgaria) or Islamic (the
Seljuqs) lines. These polities, whether full-blown nomadic
states, such as  Khazaria, or tribal unions, such as the
Pecenegs, Western Oguz (Torks of the Rus' sources) or Cuman-
Qipcaqs, however great their military prowess and commercial
interests, have passed on little in the way of literary monuments
stemming directly from them in their own tongues. Khazaria, for
example, which as a genuine state had a need for literacy, has
left us only documents in Hebrew, reflecting the Judaization of
the ruling elements. Indeed, their language about which there
are still many unanswered questions, is known, such as it is,
almost exclusively from the titles and names of prominent Khazars
recorded in the historical records of neighboring sedentary
states.  The Balkan Bulgars who, living in close physical
propinquity to and cultural contact with Byzantium and ruling
over a Slavic majority to which they eventually assimilated, have
left somewhat more in the way of scattered inscriptions in mixed
Bulgaro-Greek (in Greek letters) and in mixed Slavo-Bulgaric. 
Their kinsmen on the Volga who adopted Islam in the 10th century,
have left a number of tomb-inscriptions (dating largely from the
Cinggisid era, 13th-14th centuries) in a highly stylized, mixed
Arabo-Bulgaric language in Arabic script. Volga Bulgaria, as an
Islamic center, used, of course, Arabic as its principal language
of communication with   the larger world. The inscriptional
material, it might be argued, bespeaks a long-standing Bulgaric
literary tradition. But, in this respect, as in a number of
others, Volga Bulgaria, which did form a state, in the forest-
steppe zone ruling over a largely Finnic population and in which
denomadization was well-advanced, was atypical.
     What is interesting to note here is that unlike the Turkic
peoples of Central Eurasia and Inner Asia ( the Turks, Uygurs,
Qarakhanids ), the Western Eurasian Turkic tribes did not create
significant literary monuments either in Turkic runic script,
several variants of which were in use among many of them or in
any of the other script systems that were available to them
(Greek, Arabic, Hebrew and even Georgian).  This seeming lack
of literary ambition ( which may yet be disproved by archaeology)
is probably to be attributed to the weak articulation of
political organization among peoples such as the Pecenegs,
Western Oguz and Cuman-Qipcaqs. Thus, it should come as no great
surprise that one of the most significant literary monuments
connected with the language of one of the dominant tribal
confederations of the region, the Codex Cumanicus, was largely
the work of non-Cumans. Before turning to the Codex itself, we
must say something about the people whose language it describes.
     The tangled knot of problems that revolves around the
question of Cuman-Qipcaq ethnogenesis has yet to be completely
unraveled. Even the name for this tribal confederation is by no
means entirely clear. Western ( Greek and Latin ) and
infrequently Rus' sources called them  Comani, Cumani, Kumani.
Medieval Hungarians, who had close relations with them and to
whose land elements of the Cumans fled in the 13th century
seeking sanctuary from the Mongols, knew them as Kun. This name
is certainly to be identified with the Qun of Islamic authors
(such as al-BÔrönÔ and al-MarwazÔ, the notices in Y°qöt and al-
B°köwÔ clearly derive from al-BÔrönÔ  ) who, according to al-
MarwazÔ, figured prominently in the migration of the Cuman-
Qipcaqs to the west. Whether the Qun are, in turn, to be
associated with the Hun ( < *  u n) = Xun/Qun people affiliated
with the T`ieh-le/Toquz Oguz confederation is not clear. 
     Old Turkic sources knew elements of what would become the
Cuman-Qipcaq tribal union as Qibcaq and perhaps other names. 
The ethnonym Qibcaq was picked up by Islamic authors (e.g. in the
forms  Xifj°x, Qifj°q, Qipc°q etc.) and Transcaucasian sources
(cf. Georgian Qivc`aq-, Armenian Xbsax). These Altaic names were
loan-translated into some of the languages of their sedentary
neighbors. Thus, Rus' Polovcin, Polovci ( > Polish, Czech 
Plauci, Hung. Palocz), Latin Pallidi, German and Germano-Latin 
Falones, Phalagi, Valvi, Valewen etc. Armenian Xartes. These
terms are usually viewed as renderings of Turkic qu  <  *qub or
similar forms meaning "bleich, gelblich, gelbraun, fahl."  A
variety of sources equate them, in turn, with the Qangli, one of
the names by which the easternmost, Central Eurasian branch of
the Cuman-Qipcaq confederation was known. 
     These tribes included Turkic, MOngol and Iranian elements or
antecedents. The inter-tribal lingua franca of the confederation,
however, became a distinct dialect of Turkic that we term Qipcaq,
a language reflected in several dialects in the Codex Cumanicus.
The Cuman-Qipcaqs held sway over the steppe zone stretching from
the Ukraine to Central Eurasia where they constituted an
important element, closely associated with the Xw°razmian royal
house via marital alliances. They had equally close relations
with Rus' (with whom they often warred), Georgia (where elements
of them settled and Christianized ), Hungary and the Balkans
where later, under Mongol auspices,the Cuman Terterids
established a dynasty.
     Cuman-Qipcaq hegemony extended to much of the Crimea as
well. Here their interests were, as in many other areas,
commercial. In the pre-Cinggisid period, the Cumans took tribute
from the Crimean cities. The city of Sudaq, an ancient commercial
emporium, was viewed by Ibn al-AÔr (early 13th century) as the
"city of the Qifj°q from which (flow) their material possessions.
It is on the Khazar Sea. Ships come to it bearing clothes. The
QifjÔqs buy (from) them and sell them slave-girls and slaves,
Burtas furs, beaver, squirrels..."  By virtue of their
political hegemony, Cuman became the lingua franca of this area.
 It spread to the other communities resident there as well.
Thus,the Crimean Armenian and Karaite Jewish communities adopted
this language and preserved it for centuries afterwards in
milieus far removed from the Crimea  With the Mongol conquest
of the Qipcaq lands completed by the late 1230's, some Qipcaq
tribes (most notably those under Kten) fled to Hungary. The
majority, however, were incorporated into the Mongol Empire. The
pan-nomadic empire of the Turks was thus recreated on an even
larger scale. The Qipcaq language, far from receding into the
background, established itself as a lingua franca in the Western
Eurasian zone of the Cinggisid state within a century of the
Conquest. Thus, a Mamlök scholar, al-`UmarÔ (d.1348), observed
that the "Tatars," whose numbers, in any event, were not great
and whose ranks already included numerous Turkic elements from
Inner and Central Asia, had intermarried extensively with the
local Turkic population and had, in effect, become Qipcaqicized.
 In the latter half of the 13th century (beginning in the
1260's), as the Cinggisid khanates began to squabble over
territory, the Jocids of Saray in their struggle with the Hlegids
of Iran, found a useful ally in the Qipcaq Mamlöks of Egypt-Syria
to whom they continued to supply mamlöks from their Crimean
ports. The spread of Islam to the Mongols beginning with Berke
(1257-1266) and culminating with Ozbeg (1313-1341) helped to
strengthen this tie. 
     The Codex Cumanicus,  which is presently housed in the
Library of St. Mark, in Venice, Cod. Mar. Lat. DXLIX, is not one
but several unrelated (except in the broadest sense) works which
were ultimately combined under one cover. The Codex may be
divided into two distinct and independent parts : I) a practical
handbook of the Cuman language with glossaries in Italo-Latin,
Persian and Cuman II) a mixed collection of religious texts,
linguistic data and folkloric materials ( the Cuman Riddles),
stemming from a number of hands, with translations into Latin and
a dialect of Eastern Middle High German. It is also clear that a
number of subsequent hands made contributions to both sections.
Many scholars have simply termed these two, distinct works, the
"Italian" part and the "German" part. This is undoubtedly true
with respect to the ethno-linguistic origins or milieus of the
authors. But, Ligeti is probably closer to the mark in calling
the first part, the "Interpretor's Book" and the second part the
"Missionaries' Book." 
     The Codex was first mentioned in the 17th century and was
believed to have come from the library of the great Italian
Humanist Petrarch (1304-1375). This attribution, however, has
been shown to be incorrect.  The dating and place of origin of
the Codex's different sections have long been in dispute. Bazin,
who has closely studied the calendrical entries (CC, 72/80-81)
concluded that the "Interpretor's Book" was probably composed
between 1293-1295. Drll, however, would place it as early as
1292-1295.   The date found in the Venice ms. "MCCCIII die XI
Iuly" (CC, 1/1) should be viewed as the date of the first copy or
the beginning of the first copy. The copy preserved in the
Venice ms., as an examination of the paper has demonstrated,
stems from, or was at least copied on, paper made in the mid-13th
century.  The "Missionaries' Book" comes from a variety of
sources and was put together ca. 1330-1340. Other elements were
perhaps added later. The authors are unknown, although it seems
likely that they were part of the Franciscan community. The
German Francsicans who played an important role in the creation
of the "Missionaries' Book," came from an Eastern  High German-
speaking background. The "Interpretor's Book" was compiled by
Italian men of commerce (Venetians or Genoese) or their scribes
in Solxat (Eski Krim) or Kaffa (Feodosija). There is evidence to
indicate that different individuals ( perhaps many ) were
involved in preparing/translating the Persian and Cuman sections
of the tri-lingual glossary. The first copy (1303), it has been
suggested, was done in the monastery of St. John near Saray. The
later copy which is preserved in Venice, dating to ca. 1330-1340,
probably came from some Franciscan monastery. Here too, it seems
likely, is where the different sections of the Codex were
combined. Somehow, these various parts came again into Italian
hands and thus to Venice. The work, then, is a pastiche of
larger and smaller pieces which were composed/compiled with
different intentions. The "Interpretor's Book" was largely, but
not exclusively, practical and commercial in nature. The
"Missionaries' Book," in addition to its purely linguistic goals,
contains sermons, psalms and other religious texts as well as a
sampling of Cuman riddles.
     The Venetians and Genoese were actively involved (as well as
competitors) in trade in the Crimea. This trade, as we know from
contemporary accounts, such as Pegolotti, went by stages from
Tana (Azov, a major unloading site for goods coming from Asia to
the Crimea ) to the Lower Volga (Astraxan-Saray) and thence to
the Urals and Xw°razm and ultimately to China. It dealt with a
wide variety of items, e.g. wax, metals (including precious
metals), spices and other foodstuffs, silk and other fabrics,
pelts of valuable furs etc.  The Italian commercial colonies
in the Crimea, had, of course, regular contact with Tana. There
was also contact with Ilkhanid Iran via Trapezunt. Indeed, Drll
argues that the author(s) of the Latin-Persian-Cuman glossary of
the "Interpretor's Book" must have been Genoese, operating from
Kaffa, as the Genoese were the only ones who had contact with
merchants from both the Ilkhanid and Jocid realms.  ALthough
the Italian merchants were not involved in the slave or mamlök
trade with Egypt, the Crimea had a long history of involvement in
this activity. There is a Modern Kazax proverb that reflects this
: uli irimga, qizi Qirimga ketti "the son went as a hostage and
the daughter went off to Crimea (i.e. to slavery)." The
trilingual glossary reflects this trade orientation and as we
shall see has extensive lists of consumer goods.

     The Latin of the Codex is found in two variants, indicating
the ethno-linguistic affiliations of the authors and their
educational level. The Latin of the "Interpretor's Book" is a
Vulgar Italo-Latin, while that of the "Missionaries' Book" is
more "correct," reflecting the ecclesiastical training of its
Franciscan authors. The Persian material has been the subject
of two recent studies. Daoud Monchi-Zadeh has argued that the
Persian material came through Cuman intermediaries, a kind of
Cuman filter, and was translated by them. Andras
Bodgrogligeti, on the other hand, suggests that this Persian was
rather a lingua franca of the Eastern trade. As a consequence, it
had undergone, to varying degrees, standardization, back
formation and simplification. Some words are archaic, others
unusual. In short, what we see reflected is not the living
language of a native speaker, but rather a kind of simplified
     The Cuman of the CC also represents some kind of lingua
franca, one that was understood throughout Central Asia. This
language, however, was not perfectly reflected in the CC. The
latter, we must remember, was compiled by non-Turkic-speakers
with varying levels of command of the language. There are a
number of "incorrect" syntactical constructions as well as
mistakes in grammar, phonetics and translation. Some of these are
simply the result of faulty knowledge or scribal errors. Other
deviations from the "norms" of Turkic are probably to be
attributed to the word for word, literal translations. These
types of translations in the Middle Ages, were well-known,
especially when translating sacred, religious texts. Thus, in
Karaim, one of the closest linguistic relatives of the Cuman
mirrored in the CC, we find sentences such as :  kisi edi
yerind'a Ucnun, Iyov semi anin, da edi ol kisi ol t'g'l da t'z,
qorxuvcu t'enrid'n ("There was a man in the Land of Uz whose name
was Job and that man was perfect and upright  and one that feared
God," Job,1 ) , a word for word rendering of the Hebrew. Some
of the forms which have an "unturkic" character about them may
almost certainly be attributed to the influence of the compilers'
native Italian/Italo-Latin and German. Many of these forms,
however, are ambiguous in origin as similar phenomena can be
found in other Turkic langauges as well and may here also
reflect the influence of Indo-European languages.
     Of greater interest is the fact, hardly unexpected in a work
in which so many different hands were involved, that the CC
lexical material is comprised of several Qipcaq dialects. Some of
these can be most clearly seen in the different sections :
"Interpretor's            Missionaries' 
 Book"                    Book"
kendi                     kensi      "self"
tizgi                     tiz        "Knee"
bitik                     bitiv      "book,writing" 
berkit-                   berk et-   "to strengthe"
ipek                      yibek      "silk"
ekki                      eki        "two"
todaq                     totaq      "lip"
etmek                     tmek       "bread"        
yag                       yav        "fat"
tag                       tav        "mountain"
kyeg                      kyv        "bridegroom"
igit                      yegit      "youth" 
sag                       sav        "healthy"
abusqa                    abisqa     "old, aged"
qadav                     xadaq      "nail"
agirla-                   avurla-    "to honor" 
     In some instances, one of the sections indicates several
dialects, e.g. "Interpretor's Book" (CC, 52/57, 57/61)   Lat.
similo Pers. chomana mecunem  (hom°n° mÃkunm "I resemble" )
Cum. oscarmen  (osqarmen), (CC, 76/86) Lat. similtudo Pers.
manenda Cum. oasamac ( or  oosamac which Grnbech reads as 
oqsamaq) and the "Missionaries' Book" (CC, 141/199) ovsadi
(ovsadi "resembled, was like"), (CC,162/226) ovsar  (ovsar)
"enlich;" (CC, 131/183) job sngnc ( ypsengenca ) "sin quod tu
approbas," (CC, 140/195), iopsinip  ( ypsinip ) :  ypsen- /
ypsin- "billigen, genehmigen, gutheissen."
     The well-known shift in Qipcaq g > w/v is clearly indicated
in the "Missionaries' Book." The latter also has greater evidence
of the q > x shift (e.g. yoqsul  > yoxsul  "arm, mettellos"). The
"Interpretor's Book" appears to represent an older or more
conservative dialect.
     We may also note that whereas the "Missionaries Book"
clearly renders j with g in non-Turkic words, e.g. gahan =jah°n
"World," gan = j°n "Soul,"  gomard = jomard "generous" ( all
borrowings from Persian), the "Interpretor's Book" renders this
with j or y. This might indicate a pronunciation with y 
(although the Persian forms with j  are also regularly rendered
with i), cf. jaghan = yahan or jahan, jomard, jomart = yomard  or
jomard, joap = yowap or jowap ( < Ar. jawab "answer") and yanauar
= yanawar or janavar. This shift in initial j > y is known to
some Qipcaq dialects, especially in loan-words, cf. Baskir  yawap 
"answer," yemeyt "society, community" (Ar. jam`iyat), yihan
"universe" (Pers. jihan, jahan).
     Finally, we might note that intervocalic v/w which Grnbech
regularly transcribes as v, may just as easily represent w, e.g.
(CC, 65/72) youac = yovac or yowac "opposite," (CC, 102/121)
culgau = culgav or culgaw "foot-wrappings," (CC, 90/105) carauas
=  qaravas  or qarawas "maid, slave," (CC, 139/192) koat = qovat
or qowat ( < Arab. quwwat ) "might" (CC, 109, 113/130,134) tauc,
taoh = tavuq or tawuq, tavox or tawox "hen."
     The numerous orthographic peculiarities (e.g. s is
transcribed by s, s, z, x, sch , thus bas "head" in the
"Interpretor's Book" is rendered as (CC, 29,86, 94,/30,99,109)
bas, bax and in the "Missionaries Book" (CC,
121,126,128/161,171,175) as bas, basch, baz; basqa "besides,
apart from" the "Interpretors's Book" (CC, 64/70) bascha and in
the "Missionaries' Book" (CC, 121,123,138/158,163,189) baska,
baschka, bazka) clearly indicate that there were many
contributors to the CC and little attempt was made at
regularization. This, of course, makes many readings conditional.

     The "Interpretor's Book" consists of 110 pages (CC,1-110/1-
131). Pages l-58/1-63 contain a series of alphabetically arranged
(by Latin) verbs in Latin, Persian and Cuman. The first entry is
audio. A sampling of some of the forms is given below: audio
"I hear" mesnoem  (mÃsnowm) eziturmen (esitrmen), audimus "we
hear"  mesnam  (mÃsnowÔm) esiturbis (esitrbiz), audiebam "I was
hearing" mesin(.)dem (mÃsinÔdm) esituredim (esitredim), audiebant
"they were hearing" mesinident (mÃsinÔdnt) esiturlaredj
(esitrleredi), audiui "I heard".sinide (= sinÔdm) esitum (esitm),
audiueratis "you had heard" sindabudit (sinada bödÔt)
esitungusedi (esitnguzedi), audiam "I will hear" bisnoem
(bisnowm) esitcaymen (esitqaymen or esitkeymen), audiemus "we
will hear" besnoym (besnowÔm) esitqaybiz/esitkeybiz, audi "hear!"
bisnã (bisno) esit (esit), audirem "were I to hear" ysalla
mes(i)nde (is°ll° mÃsinÔdm "if I should only hear" ) chescha
esitkaedim  (keske esitqayedim/esitkeyedim) audiuisse(m) "If had
heard" y sinada budim  (is°ll° sina budÔm "if I had only heard")
c esitmis bolgayedim (keske esitmis bolgayedim), audiam "if I
should hear" y besnoem  (is°ll° besnowm "if I should only hear")
c esitchaymen (keske esitqaymen/esitkeymen "would that I hear"),
audire(m) "were I to hear" zonchi mesnide(m) ( conki mÃsinÔdm
"since I hear") esittim essa  (esittim ese), audires "were you to
hear" z mesnidi  (conki mÃsinÔdÔ  "since you hear" nezic chi
esiti(n)gassa ( necik ki esiting ese "lorsque tu as entendu" ,
audiueim  (=audiverim ) "were I to have heard" z s(.)ndidem
(conki sinÔdm "since I heard") esittim ersa (esittim erse),
audire "to listen"  sanadae(n) (sanadn)  esitmaga, yzitmaga
(esitmege, isitmege),audiens "one who hears, hearer" sanoenda
(sanownda "he who hears") esattan (for esatgan = esitgen),
auditurus  "one who will hear, is about to hear"  ghoet sinidn 
(xoht sinÔdan "he who wants to hear")  esitmaga cuyga  (esitmege
kyge "one who expects to hear").
     No other verb is given such detailed treatment. Most have 3-
5 entries, e.g. (CC, 5/6)  adiuuo "I help"  yari medehem  (y°rÔ
mÃdehm) boluzurmen  (bolusurmen), adiuuaui  "I helped" yari dadem 
(y°rÔ dadm)  boluztum  (bolustum), adiuua "help!"
yari bide  (y°rÔ bideh)  bolus (bolus) adiutorium  "help, aid"
yari  (y°rÔ) bolusmac  (bolusmaq).
     Some Latin terms are translated by two verbs in Cuman, eg.
(CC, 6/7) albergo hospito  "I lodge" ghana cabul mecunem  (x°na
qaböl mÃkunm) conaclarmen vel condururmen (qonaqlarmen  or
qondururmen, (CC, 9/10) balneo aliquid " bathe something" 
tarmecunem  (tar mÃkunm) "I wet" us etarmen vel iuunurmen ( us
etermen  or  yuvunurmen ). In a number of instances, we are given
deverbal nouns as well as the verbs, e.g. (CC, 12/13) coquo  "I
cook" mepaxem  (mÃpazm) bisuturmen (bistrmen) coqui "I cooked"
pohten  (poxtm) bisurdum  (bisrdm)  coque "cook!" bepoh  (bepox)
bisur  (bisr) motbahi  (motbaxÔ) bagerzi  (bagirci  <  baqir
"copper," cf. Nogay baqirsi bala "junosa obsluzivajuscij ljudej v
roli povara u kotla iz medi") coquina "kitchen" muthagh  (=
mutbax "kitchen") as bisurgan eu  (as bisrgen ew (lit. "house
where food is cooked").
Compound Verbs (henceforth, unless needed to further explicate
the Cuman forms, the Persian entries will be omitted and the
Cuman forms will be given only in transcription) :  yk tsrrmen 
"I unload," tinimdan kecermen  "I despair," (CC, 19/21 eligo "I
pick, I choose") kngl icinde ayturmen "say what is in my heart,"
eygirek etermen "I make better," (CC, 35/37, nauigo "I sail" dar
driy° mÃrowm "I go on the sea") tengizda yrrmen  ("I go on the
sea"), qulluq etermen "I serve."      

Compound Verbs with Arabic Elements are fairly well represented.
The Arabic element does not always correspond to the that found
in similar compound verbs in the Persian entries : (CC,20/21)
denpingo (sic) "I paint" naqs mekunm  naqslarmen  ( < naqs 
"painting"), (CC, 23/25)  expendo  "I spend" xarj mÃkunm, xarj
etermen  etc. But, cf.(CC,44/47-48) quito  "I quit" rah° mÃkunm
tafs etermen  < Arab.  tafs "flight, run away, escape").

Compound Verbs with Persian Elements. In many instances it may be
presumed that  the Arabic elements entered Cuman via Persian. The
words considered here are only those that are etymologically
Persian. (CC,23/26) estimo  "I estimate, value" bah° mÃkunm "I
consider the value" bacha ussurmen  (baha usurmen "I consider the
value," KWb., p.266 reads it as baha ur- "schtzen bewerten," <
paha "price." (CC, 42/454) penito "I repent" pesm°n m, pesman
bolurmen  <  pesm°n  "penitent."
     The verb "to have" is expressed using three different forms:
(CC, 29/30-313)  habeo  "I have"  mende bar, habui "I had" tegdi 
( < teg- "treffen, berhren, erreichen, gelangen, zuteil werden")
habeas  "you have!"  dar "have!" saga/sanga bolsun "may you
     The section of verbs is followed ( CC, 59-65/64-72) by one
on adverbs ( many of which are expressed by postpositioned
forms), e.g. ( CC, 54/61) ante "before" eng borun or  ilgeri    
ab "from"   idan, aput  "at, near, by, with" qatinda  ( < qat 
"Seite, der Raum neben oder bei etwas"), brevitur  "soon"  
terklep, bene "good, well"  yaqsi or  eygi, benigne "benignly,
heartily"  xos kngl bile ( "with a good heart"), com "with"
birle, bile,  (CC, 61/66) hodie "today"  bu kn, (CC, 61/67) ideo
"on that account, therefore"  aning cn, jam "now, already" saat
digar "immediately" bir anca  or imdi, (CC, 62/68) multum "much" 
kp,  malicioxe  "maliciously" knavishly, wickedly" yaman kngl
bile, non "no"  yoq, nihil "nothing" hec-neme-tagi, (CC, 62/69)
postea "afterwards" songra  (CC, 63/70) quid  "what?" ne, (CC,
64/70) sane "healthily" sagliq bile.

Personal Pronouns (CC, 66-68/72-74) follow the listing of
adverbs, examples are :  ego  "I"  men, mei "of me"  mening,
michi "to me" manga, me "me" meni, ame "from me" menden, nos "we"
biz etc.(CC, 68/74) ipse met  "himself" anlar ox (anlar z  ?)
"they themselves." This same section contains a series of
indeclinable nouns, e.g. :  alius "other (than)" zge, (CC, 69/74)
omnis "all" tegme  or barca, solus  "alone" yalguz, talis "of
such a kind, such" falan, qualis "of what kind?" qaysi and basic
adjectives, e.g. : ulu "big," kici "little," yaqsi or eygi
"good," yaman "bad," yngl  "light," agir "heavy." 

Vocabulary Pertaining to Religion (CC, 70/77) Tengri  "God,"
Maryam qaton "The Queen (Virgin) Mary" mater dey, friste "angel,"
peygambar "prophet," ari, algisli  "holy, saint" santus, xac
"cross," bapas "priest," tre "law,"yarligamaq "mercy," bazliq
"peace," tengri svmeklig "love of God"
(caritas, dostÔ-i xud°).

The Elements (CC, 71/78-79): hawa "wind" and salqon "wind" (cf.
Mong. salkin "wind"  and Old Turkic salqim  "cold, hoar-frost,"
Siberian Turkic salqin  "violent (cold) wind"), su "water,"
yer "earth, land," ot "fire."

Humours of the body (CC, 71/79): qan "blood," balgam "phlegm (

Arab. maqdönis/baqdönis, cf. also Osm. maydanoz  > Mod. Gr.
           . Marul "lettuce"  <                     < Lat.
amarula (lactuca), cf. Osm. marul, Mamluq Qipcaq marul.
Timean  "incense"  <             possibly via Eastern Slavic
timian.  Trapes  "table"  < 
Eastern Slavic : izba  "room, chamber" (CC, 100/119 camera,
hujra)  < izba  "house, bath." Ovus "rye"  < Old Rus' ov's,
Russ. ovs  "oats," cf. Qaraim uvus. Pec "stove"  < pec', cf.
Qaraim pec. There are also more recent borrowings of this word
into other Turkic languges from Modern Russian. Samala "pitch"
< smola "soot," cf. Mamlök Qipcaq samala, samla, salama. Salam
"straw"  < soloma, cf. Mamlök Qipcaq salam, kk salam - saman,
found also in Qaraim, Qaracay-Balqar, Qazan Tatar salam and in
Hungarian szalma. The connection of Turkic saman "straw" with
this term is unclear. Some terms are problematic, e.g. terem 
"tabernacle, shrine," cf. Old Rus' terem "high house, court,
cupola, watch-tower," Russ. "room, tower-chamber" - Gr.           
         "room, chamber." But Sagay Turkic has trb "yurt,"  cf.
Mong. terme  "wall."   Similarly, bulov "some kind of weapon,
probably a club ( cf. Mamlök Qipcaq bulav, bula'u) may be taken
from Eastern Slavic bulava. The reverse may also be true.
Mongol : The CC contains a number of Mongol loanwords. Given the
historical contacts of the Turkic and Mongolian peoples, not to
mention the much-debated Altaic question, the dating and nature
of these words pose many problems. Our task is further
complicated by the fact that Mongol-speaking, or bilingual,
Mongol and Turkic-speaking (i.e. Mongol tribes that were becoming
Turkicized) joined the Cuman-Qipcaq confederation before the 13th
century. Other Mongol influences undoubtedly stem from the era of
Cinggisid hegemony. Thus, there are many layers of Cumano-Qipcaq-
Mongol interaction, some very old, which cannot be easily
differentiated. Poppe has done a very thorough study of these
words. As a consequence, we shall give here only a representative
sampling :
Codex Cumanicus                  Mongol
abaga                            abaga "uncle"
abra- "to defend"                abura- "to save"
bilev "grindstone"               bileg, bile', bile-, bili-       
                          "to stroke, stripe, streak"
ceber "pleasant, amiable"        ceber "pure, sober"
egeci "father's sister"          egeci < *ekeci  "older sister"
elbek  "richly"                  elbeg  "richly"
kenete "suddenly"                genete, genedte "suddenly"
maxta- "to praise"               magta-, maxta-, maqta- "to       
nger "friend, comrade"          nker  "companion"
olja "war booty"                 olja "booty"
bge "grandfather"              ebge < *ebke "grandfather"
qaburga  "rib"                   qabirga  "rib"
silevsn "lynx"                  silegsn  "lynx" etc.

     Among some of the problematic words, we may note Cuman
bagatur, Pers. bahadur, Mong. bagatur  "hero" which Poppe
considered a Mongol loanword. Clauson, however, suggested that
this very old, Inner Asian culture word went back to the language
of the Hsiung-nu. Cuman qarav, qarov "recompense, reward,
retribution" ( CC, 43/46 premium, jaza ) and qarav berrmen "I
forgive, absolve" ( retribuo, miamorzm ), cf.  Qaraim qaruv
"answer" --Mong. qarigu, xarigu  "answer, response, return,
retribution." Cuman tepsi "plate, dish" ( in numerous Turkic
dialects ) -- Mong. tebsi  "large oblong plate, platter or tray,
trough"  < Chin. tieh-tzu  Middle Chin.  dep tsi. Of uncertain
origin is  (CC, 90/105) bogavul/bogawul  "officer of the court"
placerius, tatawul, cf. the Ilxanid functionary bukawul/buqawul 
"Vorkoster, vielleicht General Zahlmeister." 

Arabic: Arabic elements, as we have seen, are quite numerous in
all the socio-linguistic categories noted in the "Interpretor's
Book" and elsewhere. This reflects the important Muslim
political, commercial and religio-cultural influences in the
Crimea. That these words were not limited to the Muslim
population can be seen by their presence, without sectarian
connotations, in Qaraim and Armeno-Coman. Elsewhere in this
study, frequent reference has been made to words of Arabic
origin, many of which entered Cuman via Persian. We shall cite
here only a few examples : alam  "banner"  <  Arab. `alam, albet
"certainly, of course" < Arab. albatta, azam "man"  < adam,
seriat "judge" < Arab. sar`iyyah "Muslim law." This use of a
specific Muslim term for a broader category is also a feature of
the Tolkovanie jazyka poloveckogo  (13th century ?, discussed
below), cf. alkoran  "zakon" < al-qur'°n, elfokaz "uciteli i
velikie tolkovnici" < al-fuquh° "jurists of religious law." 
Xukm "judgement, decision" < Arab. hukm, hakim, xakim  "doctor" 
<  Arab. hakÔm, aziz, haziz "rare, costly, pilgrim, holy, sacred"
<  Arab. `azÔz, nur  "light"  < Arab. nör, safar "journey" <
Arab. safar, seir  "poet" < Arab. si`r  "poetry,"  s°`ir  "poet,"
tafariq  (CC, 132/184, tafsanyt ) "difference"  < Arab. tafrÔq,
pl. taf°riq "separation, differentiation."
Persian: The principal Muslim lingua franca of the East, Persian,
is also well-represented in the CC. As these words have been
pointed out in much of the foregoing, the following is only a
very brief sampling : daru "medicine" < Pers. d°rö, drust  "true"
< Pers. drust, durust, bazar "bazaar,market" <  Pers. b°z°r,
bazargan "merchant" <  Pers. b°zarg°n, hergiz, herkiz "never" <
Pers. hargiz  "ever, always, continuously," jahan, jehan  "world"
< Pers. jah°n, jih°n, jigar "liver  < Pers. jigar,  piyala 
"goblet"  < Pers. piy°la etc.
Hebrew, Syriac and Others Elements :  as was noted earlier, 
Cuman sabat kn  "Saturday" derives ultimately from Hebrew sabbat
via a probable Khazar intermediary. The name  (CC, 143/202)
Hawa/Hava  "Eve" also appears in its Hebrew form (Hava) rather
than the expected Eva. Interestingly, the word for "Messiah"
appears in its Syriac form, or a form derived from it : (CC,
138/189) misixa   <  Syr. MesÔha. There are a number of words of
undetermined origin. Among them is (CC, 160/222) kesene
"grave mound," which is preserved in Qaracay and Balqar k`esene,
kesene  "Friedhoff, grobnica." Ligeti suggested a Caucasian
provenance without adducing further evidence.  Zajaczkowski noted
Pelliot's earlier Persian etymology, kasana "a small house." 
But, it is not quite clear how the Cuman form could have emerged
from the Persian.

     The authors of the "Missionaries' Book" had to create or
elaborate a special Christian Vocabulary. Certain religious terms
were already known to Cuman, as part of the Inner Asian Turkic
legacy of long-standing contacts with a variety of religions.
Thus, terms such as tamu,tamuq, tamux "Hell," ucmaq "Paradise,"
both loanwords from Sogdian ( tamw,  'wstmg ) or some  other
Iranian language , were already familiar concepts and not
necessarily in a Christian form. These and other Old Turkic terms
were now given a specific Christian nuance,e.g. bitik ( < biti-
"to write" < Middle Chin. piet "brush" ) "anything written, book"
now became "The Book," i.e. the Bible. Other terms were loan-
translated into Turkic, e.g. Bey(imiz) Tengri  "Dominus Deus,"
clk "the Trinity," ari tin  "the Holy Ghost," kktegi xanliq " the
Kingdom of Heaven," etc. An interesting usage (if not original in
Cuman) is yix v ( < iduq ev  "holy, sacred house") "church"
(found in Qaraim as yeg'v  "church," a semantic parallel can be
seen in Hung. egyhaz "church, " lit. "holy house"). The notion
of "saviour" was directly translated into Turkic : (CC, 122/160)
"Yesus Christus bitik tilince, tatarca qutqardaci, ol kertirir
barca elni qutqardaci" "Jesus Christ, in the language of the
Book, in Tatar, is the Saviour, that means the Saviour of all
     The Cuman calendar ( see above) shows neither specific
Christian influences nor any trace of the Sino-Turkic 12 year
animal cycle. This appears to be an archaic system, typical,
perhaps, of the Northern Turkic milieu from which the Qipcaqs
     Other examples of this older Turkic culture can be seen in
words such as qam "sorceress"  < qam "shaman, sorcerer,
soothsayer, magician." 

Cuman Documents Contemporary to the Codex Cumanicus
     A number of Qipcaq-Arabic grammar/glossaries (sometimes
containing other languages as well ) appeared in Mamluk lands in
the 14th and 15th century. Close in content to the CC, although
very different in format, are the Kit°b al-Idr°k li'l-Lis°n al-
Atr°k (ca. 1313 or 1320) of Abu Hayy°n (1286-1344), the  Kit°b
Majmö` Tarjum°n TurkÔ wa `AjamÔ wa MugalÔ wa F°rsÔ (now dated to
1343), the Kit°b Bulgat al-Must°q fi Lugat at-Turk wa'l-Qifj°q 
of Jal°l ad-DÔn Abu Muhammad `Abdall°h at-TurkÔ (which may date
to the late 14th century, but certainly before the mid-15th
century), the At-Tuhfah az-Zakiyyah fi'l-lugat at-Turkiyyah of as
yet undetermined authorship (written before 1425) and the al-
Qaw°nÔn al-Kulliyyah li-Dabt al-Lugat at-Turkiyyah written in
Egypt at the time of Timur. To this list may perhaps be added
the  thus far partialy published six-languge Rasölid Hexaglot
(dating to the 1360's) which contains vocabularies in Arabic,
Persian, two dialects of Turkic (one of which is clearly Oguz,
the other may be viewed as Qipcaq or a mixed Eastern Oguz-Qipcaq
dialect), Greek, Armenian and Mongol.
     There are also fragments of Cuman-Rus' glossaries such as Se
tatarsky jazyk  which is found in a 15th century sbornik from
Novgorod and the Tolkovanie jazyka poloveckago found in a 16th
century menologium. These undoubtedly date from an earlier
     Finally, mention should be made of the Qipcaq translation of
Sa`di's Gulist°n done by Sayf-i SarayÔ in Cairo in 793/1390-

This counter has been placed here on 04 March 1999