W. K. M at (hem s 
T H E P H O N E T I C V A L U E O F JAT' 
IN O L D R U S S I A N 
i 
This brief study1 of a very small subject must be prefaced by a qualification 
and a delimitation. The first bears on the frequently forgotten difference between 
conjecture and fact: in dealing with problems of historical phonetics we are obviously 
never sure that our conjectures, however plausible they may seem, represent the 
actual state of things, and we should therefore not allow those conjectures to become 
synonymous in our thinking with facts, which after all are the result of observation 
and capable of empirical proof. The second attempts to fix our conjectures in place 
and time. The purpose of this is to limit their application and simultaneously to 
increase their probability. Here jOld Russian* then is taken to mean the literary 
language used all over the East Slavonic territory from the middle of the llti> to 
the end of the 12'h century (1056—1199). 
Old Russian, thus defined, was cultivated as a literary language in two more 
or less divergent forms. The more literary of the two was not widely different from 
Old Church Slavonic (Old Bulgarian), which had come to Russia with its mainly 
translated literature in the 10Ul century; the other used the Old Church Slavonic 
alphabet to represent the language of affairs as illustrated by treaties, deeds, and 
other legal documents. It has been stated, notably by S. P. Obnorskij,2 that this 
language of affairs was a purely Russian product uninfluenced by Old Church 
Slavonic example, but such a view cannot be sustained, because it does less than 
justice to existing facts. The difference between the literary and non-literary styles 
is no more than a difference of degree, dependent'on the varying proportion of Old 
Church Slavonic elements, and not of kind, implying an opposition between two 
distinct literary forms of Slavonic. Such an opposition would have been possible if 
the literary application of Slavonic had originally had more that one focus. But as 
1 Dedicated as a token of esteem to Professor Fran Ramovš. 
2 Oterki po istorii russkogo lileralurnogo jazyka staršego perioda (Moscow-
Leningrad, 1946). 
a matter of historical fact the Slavonic literary language, as a stylised form of an 
Old Bulgarian dialect, had been worked out, in both vocabulary and syntax, long 
before it became a stimulus and a model to Russian and the other Slavonic languages. 
II 
The acceptance of the Old Bulgarian (Old Church Slavonic) church books in 
Russia meant the acceptance of their language and the alphabets devised to repre-
sent it. These alphabets — Glagolitic and Cyrillic — were made to fit the broadly 
analysed phonetic, or phonological system of Macedonian Bulgarian, and the basis 
of both was the adaptation of the Greek symbols with their Byzantine (in effect 
largely modern)3 phonetic values to figure the contemporary pronunciation of an 
adjacent type of Slavonic. We need not discuss here the relations and the relative 
antiquity of the two alphabets.4 Suffice it to say that they agree in having a separate 
symbol for jat' (ë). The Glagolitic symbol in its triangular Bulgarian aspect (д), 
which does not differ much from its later Croatian variant, resembles one form of 
the minor jus (A) as it appears, for instance, in the Book of Savva (Savvina kniga). 
The Cyrillic representation of jat' (f.) shows it to be a letter of the non-Greek series 
and to resemble the jery, viz. ъ and ь in design. In both alphabets jat' as a letter 
is distinct from the other vowel-symbols, but this in itself is not enough to prove 
that it represented a distinct vowel sound, because we have, following Greek pre-
cedent, two distinct letters for the sounds [i] and [o], viz. Cyrillic и/i and o/u) 
respectively. Proof of its independence as a sound however is forthcoming: jat' (è) 
in the oldest Old Church Slavonic monuments is kept separate, say, from e by 
appearing in words which show it to correspond to I. E. ê, ai (ai), and oi (ôi) 
(cf. О. С. S. et,«*, л-ккъ, цкнл with Lat. sêmen, Gk /.awç, Lat. poena), whereas e 
reproduces I. E. e (cf. O. C. S. tip* with Gk <pbQi»). 
If we attempt to draw up a scheme of the Old Church Slavonic (here mainly 
Old Bulgarian) vowel-system we shall find that it assumes the triangular shape 
pivoting on a single type of a which characterises so many phonetic systems.5 In 
this scheme all the Old Bulgarian vowels, oral and nasal, except jat', take their 
3 Sec G. N. Hatzidakis, Einleitung in die neugriechische Grammatik (Leipzig, 
1892); H. Pernot, D'Homère à nos jours. Histoire, écriture, prononciation du grec 
(Paris, 1921); A. C. Juret, Phonétique grecque (Paris, 1938); R. M. Dawkins, The 
Greek Language in the Byzantine Period* (N. H. Baynes and H. St. L. B. Moss, 
Byzantium, Oxford, 1948, pp. 252—267). 
4 See my article »Sources of Old Church Slavonic« (The Slavonic and East 
European Review, XXVIII, 71, London, 1950). 
5 See my article »The Old Bulgarian Language-Type« (Archivum Linguisticum, 
I, 2, Glasgow, 1949). 
appropriate places without raising any serious doubt in the classifier's mind. The 
scheme appears as follows in Latin transliteration: 
and as follows in Cyrillic: 
elf ojg 
•k 
«/л О, U'/ж 
It will be observed from this that the Cyrillic letters with prefixed iota, viz. н, ь». 
и, к»., ю, do not appear, but then, as they stand, they symbolise two sounds (/ + 
vowel), and when they follow a consonant symbol, they are merely a device for 
representing palatalisation. The above scheme however lacks the presence of jat', 
whose phonetic value has been variously defined, but without unanimity on the part 
of those who have attempted to do this. Both quantity (length) and a diphthongal 
pronunciation (e. g. Leskien's ea or Mikkola's ей) have been attributed to it, while 
the other vowels have been interpreted as simple and short. The correspondence 
between the Olagolitic symbol * on the one hand and the Cyrillic k." on the other 
(cf. the transiternted ^ко in Cod. Zogr. with in the Book of Savva) suggests 
that the sound was short and possessed an e-like quality, but that O. C. S. 'fc (è) could 
not have been и (ja) is shown by the coexistence of the two symbols in Cyrillic. 
By the same train of reasoning it could not have been < (г), with which it coincided 
in Russian in the course of the development of that language. That it was a front 
and not a back vowel may be seen in its equivalents in the modern Slavonic languages, 
whose phonetic systems are known empirically. Here it varies from a close [i] (e.g. 
Ukr. and N. Russ. lis) to an open front [a] (e. g. Pol. las). Where then are we to 
place jat' along the line of front vowels from [i] to [a]? We are of course considering 
the Old Bulgarian sound of the 911' and 101'11 centuries as it may have been pro-
nounced in a word like ккл-i.. If this sound could not have been i, i,, e, or a, it must 
necessarily have come between the last two in the scheme and resembled the [ее] 
of Eng. man, Slovak «, or the Russian value of я in пять. Accordingly we may 
complete our scheme of Old Bulgarian front vowels by inserting ё between elf 
and a, thus:" 
tit 
0 Cf. J. Ruduyčkyj, Narys gramatyky staro-cerkovno-slov'janškoji movy (Mu-
nich, 1947). 
But probably this é was not a uniform sound any more than was, say, the sound 
represented by the discarded Bulgarian symbol b, which is now replaced, according 
to phonetic context, by я and e (г). We can imagine that the * of O. C. S. к-клити 
may have been a closer sound than the "k of с-клу exactly as it is in Bulgarian to-day. 
Ill 
The Cyrillic alphabet with its 10"> century Bulgarian values was adapted to 
Russian use, and although Old Russian resembled Old Bulgarian more than the 
modern languages resemble each other, there can be no doubt, a priori as well as 
a posteriori, that there were discrepancies between sound and spelling from the 
outset. This is most strikingly illustrated by the occasional confusion of the symbols 
of nasal and oral vowels in the Ostromir Gospels (e. g. * and «у/1®! л/w and л/и : 
глаголю for глагола, RLCA for ккси). This earliest dated Old Church Slavonic monu-
ment — it was copied in North Russia in 1055—57 — also illustrates an early 
Russian pronunciation of the symbol i as « (e. g. UP<K» for чрки»). The substitution 
of -k for < and the reverse occurs in monuments of North Russian (e. g. Novgorodian) 
origin. But even South Russian (Kievite) monuments have « for -k in the period 
under consideration.7 The existence of these substitutions points to the prevalence 
of a general Russian pronunciation of the symbol i . which differed appreciably 
from the conjectured Bulgarian pronunciation. 
What was this Old Russian pronunciation of and was it as uniform over 
the Old Russian area as the pronunciation of historic jat' is in the modern period? 
To answer these questions we must review a variety of facts provided by historical 
records and the present-day phonetic study of Russian and its dialects and of its 
Slavonic cognates. 
IV 
The symbols k and <, we have seen, are interchangeable in a number of in-
stances in the Ostromir Gospels. Other 11"' century monuments contain inter alia 
the following examples: » K for HI (Svjatoslav Miscellany, 1073), ESMVI for с«д*т< 
(Turov Gospels), иждк for -кждк (Archangel Gospels, 1092). In the next century 
the number of such interchanges becomes larger and the examples more obvious 
(e g. жк for Ж1, Galič Gospels, 1144; (клм-кмч for млинм, Inscription on the Cross 
of Jevfrosinija of Polock, 1161; кт.стликть. for вт.сстлм1тк, Dobrilo Gospels, 1164; 
л'кнк for Aniw, Thirteen Homilies of Gregory of Nazianzus). This phenomenon inclines 
us to infer that perhaps as early as the 11century the Eastern Slavs, both northern 
7 V. Vondrâk, Cirlicvnèslovanskà chrestomatie (Brno, 1925), p. 172. 
and southern, pronounced -k and < identically, and the sound, except between pala-
talised consonants, was e, as in modern Russian. But there are other interchanges 
of vowel-letters involving jat', viz. н/'к, л/к, and и/к. The first two are character-
istically Russian, the last is assumed to have existed in Common Slavonic, and occurs 
in the oldest records of Old Church Slavonic wherever the preceding sound is a 
palatal or palatalised consonant (e. g. ствитн, слыимти, дроиити, молчлти). Russian 
examples of it occur in the Ostromir Gospels (e. g. C M ^ S , НДИЛШ«, - К Д И ^ Т , ) . Examples 
of the interchange of и/к may be seen in «и for «k (Svjatoslav Miscellany, 1076), 
В И Д А Т Ь for в-кдлтк (Turov Gospels) and of л/k in кдпл-к for клплл (Ostromir Gospels), 
въдскнц-к for К-ЬДОКНЦЛ (Svjatoslav Miscellany, 1076). The last interchange indeed is 
regarded as a correspondence: O. C. S. л is paralleled by O. R. "k, as in the acc. 
plur. of the masculine jo-type nouns (cf. О. С. S. KOHIA with O. R. квн-k) or the gen. 
sing, of feminine /a-types (cf. O. C. S. з<мл1л with O. R. зшл-fc). The parallelism 
emerges best by comparing the A-forms of Old Church Slavonic monuments with 
the corresponding -k-forms of Old Russian legal documents (cf. О. С. S. acc. plur. 
лкмА with О. R. ме-к, Mstislav Volodimirovič's Deed of Gift, 1130). It occurs ex-
clusively in the nominal paradigms and is absent in the verb, where we find the 
correlation O. C. S. л/О. R. IA (e. g. дадлтъ/дадмтк, мвлишл/молнши, чнтльл/чнтлм). 
The adjectives, as a nominal category, run parallel to the nouns in this respect, 
except that a soft long type like the O. C. S. fem. gen. sing, сшили» has the O. R. 
parallel ciim-k as well as the more »regular« снн-kk. Here again we come up against 
the phonetic parallelism of k/f in Old Russian. 
The collation of »pure« Old Church Slavonic on the one hand and Old Church 
Slavonic in Russian recension on the other yields, as we have noted, the following 
sporadic ratios, viz. the symbol k occurs as a substitute for и. i, A, and и, and may 
be replaced by any of these symbols. The inference from this may be that the pro-
nunciation of k varied between the phonetic »maxima« of [i] and [a] in Old Russian. 
We may assume a priori, after studying the present-day Russian sound-system, that 
the phonetic value of k was anything but uniform and tended to vary with dialect 
and stressing. Some modern Russian dialects of the Northern type still distinguish 
a closer e from a more open e, and the first generally corresponds to historic k (£), 
whereas the literary language identifies 'b (è) and о (e) phonetically. Stress tends 
to preserve the »unreduced« pronunciation, as in modern Russian, and its absence 
favours »reduction«, or what, in phonetic terms, is a change of sound. It appears 
likely that an unstressed k, as in cimkk, would prefer a »raised« vowel, viz. e, 
while the corresponding stressed symbol, as in яшлк, would incline to be lowered, 
perhaps, to [ а л ] . This seems to be supported by the' evidence of Polish parallels 
such as siano: с-кно and wodzie: код-к. Here the second Polish form, incidentally, 
shows a shift of stress. 
But the evidence of purely Slavonic parallels is not enough. Old Russian was 
in contact in the early Middle Ages not merely with Old Church Slavonic, but with 
West Finnic (West Somian), Baltic, Scandinavian, Byzantine Greek, and Arabic, 
besides a great many other languages of both related and totally alien stocks. Let 
us consider a number of phonetic correspondences between these languages. West 
Finnic8 offers such examples as Est. nädal (cf. O. R. тд-кли), lääva, rnäära (cf. O. R. 
Хл-кБ-ъ, лгкра), Car. reähkä (O. R. гр-к̂ -к). The last three examples are paralleled by 
Latv. klêvs, mers, and grëks, where ë is phonetically [эе]. Later Carelian loans 
from Russian show ie for к (cf. miela with O. R. м-кл-к), which may reflect a changed 
pronunciation of к in Russian. This ie indeed is generally accepted by Russian 
scholars as the phonetic equivalent of * in the medieval language.9 Some Russian 
loans from West Finnic seem to support this (e. g. п̂ р-к < ; Finn, purje), whereas 
others, for instance the tribal names Häme and Karjala, figure in Old Russian as 
•ћмк/М.ик and Кор-клл respectively, which suggests an open value for k.10 Old Russian 
loans from Scandinavian sources prepare us for a closer pronunciation of к (cf. the 
personal names Ол-ккъ: ôleifr, Рогн-кдк : Ragn(h)eid, Свкпъ: Sveinn, Сгкндлд-к: 
Sveinaldr).11 Byzantine Greek in the 10th century records of Constantine Porphyro-
genitus15 offers no transcriptions of Russian names with k, except possibly rwr 
Sepßlwv, which is thought to stand for С-Ьыриш, but here there has been obvious 
contamination with Sf(>/2?.la and Sepßf.oi. On the other hand the much older river-
name ddranpi{ has the O. R. equivalent Д-кн-кпр-к, where к seems to be very open. 
Here Arabic comes to our aid with Dulabe and its variants Ditlaba, Dulaja, Dulana, 
Dulavana, for O. R. Дсул-кс-ki,13 if this is the right correlation. But then in Arabic 
we also have the form Kujabe for Ккнв-к (cf. Constantine's Ktoäßa). 
The testimony of non-Slavonic languages which were in contact with Old 
Russian leads therefore to the same conclusion as that drawn from ancient and 
8 J. J. Mikkola, Die ältesten Berührungen zwischen Ostseefinnisch und Russisch 
(Helsinki, 1938). 
9 Cf. A. A. Šachmatov, »Očerki drevnejšego perioda istorii russkogo jazyka« 
(Ene. Slav. Eil., 11/1, Petrograd, 1915); A. D. Grigor'jev, Russia j jazyk (Warsaw, 
1915); N. N. Durnovo, Olerh istorii russkogo jazyka (Moscow-Leningrad, 1924); 
L. A. Bulachovskij, Istorileskij kommentarij k Ijteraturnomu russkomu jazyku (Khar'-
kov-Kiev, 1937). 
10 A. A. Šachmatov, Povest' vremennych let, I (Petrograd, 1916). 
11 V.Thomson, The Relations between Ancient Russia and Scandinavia and the 
Origin of the Russian State (Oxford-London, 1877). 
" J. P. Migne, Patrologiae cursus completus. Patrologiae graecae torn. CXII1, 
Paris, 1864). 
13 A. Ja. Harkavi, Skazanija musuVmanskich pisatelej о slavjanach i russkich 
(St. Petersburg, 1870). 
modem Slavonic parallels, viz. that O. R. * appears to have been a phonological 
complex, or variable sound poised normally between the phonetic points« [e] 
and [ее], with a tendency to be raised above [e] or lowered below [ее] according 
to the phonetic context as determined by the distribution of sound and stress, and 
according to the vagaries of dialectal practice. 
University of London. 
P o v z e t e k 
Pri historični fonetiki je treba dobro ločiti konjekture od resničnih podatkov; 
vse konjekture pa je treba postavljati v določen čas in prostor. Avtorju gre za 
določitev glasovne vrednosti ê v stari ruščini, t. j. v knjižnem jeziku, kakor je bil 
v rabi na vsem slovanskem vzhodu 1056—1199. V skrbni obliki ta jezik ni bil daleč 
od stare cerkvene slovanščine, v poslovni pa se je bolj odmikal, vendar je to isti 
jezik v dveh stopnjah. 
S staro cerkveno slovanščino so Rusi prevzeli pisavo, ki fonološko predstavlja 
makedonsko bolgarščino, a je oprta na grške črke takratne izgovarjave. V glagolici 
in cirilici imamo posebno znamenje za jat, a to še ni dokaz za samosvoj glas, ker 
iniamo več črkovnih dublet brez glasovne razlike v grški in stcslov. pisavi. Glasovno 
vrednost za è proti etimološkemu e navadno opredeljujemo s tem, da zastopa è čisto 
druge indoevropske glasove (ë, ai-äi, oi-öi) kakor e. Toda v vokalnem sistemu 
stare cerkvene slovanščine iniamo en sam tip a in v tem sistemu ni prostora za ê 
(gl. trikotno razvrstitev). Pripisovali so mu dolžino in dvoglasnost (n. pr. Leskien ea, 
Mikkola ea) zlasti zaradi zamenjave ê in [a posebno na začetku besed. Moral je biti 
sprednji vokal, a tu niha med ozkim in širokim glasom i-a (ukr. lis — polj. las 
za stcsl. 1ёэъ) ; v vrsti sprednjih vokalov ga je treba postaviti med elf in a. Verjetno 
pa ê ni imel zmeraj enake glasovne vrednosti. 
Zamenjave v prepisovanju Ostromirovega evangelija nam pričajo, da se je 
glasovno stara ruščina že zgodaj odmaknila od stare cerkvene slovanščine; poleg 
zamenjave nosnikov ? in p z ustnima г in o imamo že tudi zamenjavo e in ê. 
V 11. stoletju so zahodni Slovani na severu in na jugu enako izgovarjali e in i. 
Srečujemo pa tudi zamenjave ê z i, с in (я; zadnja zamenjava je utemeljena v stari 
cerkveni slovenščini za palatali, tu jo srečujemo tudi v drugačnih pozicijah. Za-
menjava ( in ê je pogostna pri osnovah na la in io; imamo pa hkrati tudi zamenjavo 
C in (a. Po vsem tem bi morali sklepati, da je glasovna vrednost ê nihala nekje 
med i in a. Verjetno je bila različna po narečjih in naglasu. Današnja ruska severna 
narečja ločijo zaprti p od bolj odprtega e in prvi predstavlja historični ê, medtem 
ko sta v knjižnem jeziku sovpadla e in Poudarek varuje e pred redukcijo, v ne-
poudarjeni poziciji pa se nagibi je k redukciji; zdi se, da je bil nepoudarjeni ? v 
primerih kakor sind »visok« samoglasnik, medtem ko je 'bil poudarjeni v primerili 
kakor zeml'ê nižji, usmerjen proti ее, knkor bi kazala poljščina s siano za sêno in 
wodzie za vod?. 
Isto nam potrjujejo tudi neslovanski jeziki v substituiranju starega ruskega <'• 
(finščina, baltščina, skandinavščina, bizantinska grščina in arabščina). Avtorjeva 
izvajanja potrjujejo sklepe, do katerih je prišel Fr. Ramovš v članku Fonetična 
vrednost psi. t. Razprave AZU H, 111—124. Ljubljana 1944.