The Slavonic Tongue Is One.

I’ve been reading Simon Franklin’s Writing, Society and Culture in Early Rus, c.950-1300, and I found the following passage so sensible and interesting I thought I’d quote it here (it’s near the start of chapter 2, “Scripts and Languages”):

For the present purposes the relevant variants are East Slavonic and Church Slavonic. In Rus, East Slavonic and Church Slavonic were both written and spoken, though in the relationships between written and spoken they are like mirror images of each other. In origin East Slavonic was the spoken vernacular of the East Slavs of the lands of the Rus, while Church Slavonic was the written medium which was developed in the process of translating Scripture for West and (eventually, mainly) South Slavs. In usage, spoken Church Slavonic was a means to disseminate the written forms (e.g. through liturgy, through sermons, through recitation and reading aloud), while written East Slavonic derives from (which is not to say that it is identical to) speech. Initially distinct both geographically and culturally, East Slavonic and Church Slavonic form the outer parameters of most discussion about language in Rus. Like Common Slavonic, they are abstractions, and hence contentious: should they, for example, be regarded as distinct languages, or as variants of the same language? If viewed separately, is either of them in fact an entity, or do they, too, dissolve into their own subvariants? If viewed together, what is their interrelationship?

Pristine Church Slavonic (the Cyrillo-Methodian translations) and pristine East Slavonic (the speech of the East Slavs) are irrecoverable from direct contemporary evidence. ‘Old’ Church Slavonic is normally deduced from a more or less agreed ‘canon’ of somewhat later manuscripts (mostly from the eleventh century) which are deemed to reflect it most accurately. In practice, however, virtually all Church Slavonic manuscripts already contain hints of their own linguistic milieu, and over time Church Slavonic divides into increasingly pronounced regional variants, or dialects, dubbed ‘Russian Church Slavonic’, ‘Bulgarian Church Slavonic’ (or ‘Middle Bulgarian’), and so on. Nevertheless, the differences in written convention are too trivial to be interpreted as disintegration, and throughout the Middle Ages, Church Slavonic – in the legitimately capacious sense of the term – continued to serve as a written lingua franca for Orthodox Slavs. Spoken forms of East Slavonic (aka. ‘Old Russian’, ‘Old Ukrainian’, or the ingenious coinage ‘Rusian’) are of course unrecorded, because our only evidence is filtered through the selective and more or less conventionalised medium of writing. We see spoken East Slavonic as if through occasional gauze-covered holes in a screen. There is no way in which we can reconstruct, for example, the rhythms and nuances of an authentic domestic conversation. Nevertheless, the glimpses are sufficient to reveal certain general features, as well as certain distinctive regional features. Accidents of survival mean that we are particularly well informed about the Novgorodian version.

How different from one another were Church Slavonic and East Slavonic? Linguistic comparisons cover four main categories: sounds (phonology), word-forms (morphology), sentence structures (syntax), and vocabulary and meaning (lexis and semantics). The sounds of pristine Church Slavonic, reflecting South Slav pronunciation, would have been somewhat strange to the ear of an untutored East Slav at the time of the official Conversion of the Rus; and stranger still by the late twelfth or early thirteenth century as the loss of reduced vowels brought about major changes in the sound structures (and word-forms) of East Slavonic. However, this is to some extent a false contrast, since the sounds of pristine Church Slavonic are unlikely to have been imported intact together with the writing system. Church Slavonic is a written language, but this does not mean that the conventions of writing ‘are’ the language. For readers and their listeners in Rus, Church Slavonic probably had a strong local accent. Our notional untutored East Slav might have been even less struck by the morphological contrasts between his spoken vernacular and Church Slavonic. Inflected word-endings, for example, were broadly similar, and one could quite easily get used to the consistent alternatives in word-formation. More exotic was the way in which words were strung together in clauses and sentences. Devised for the purpose of translating from Greek, Church Slavonic was apt to mirror Greek rhetorical structures unfamiliar to spoken East Slavonic: complex structures of subordination, or the widespread use of participles. But perhaps most alien of all were many of the words themselves, and their meanings. Although Church Slavonic and East Slavonic shared a common core of vocabulary, Church Slavonic brought a mass of concepts which were wholly new to the East Slavs. It was saturated with words and expressions which had no precise precedent in any pre-literate variety of spoken Slavonic: words borrowed or calqued from Greek, or familiar Slavonic words imbued with unfamiliar connotations.

What, then, does our East Slav make of Church Slavonic? If he listens to a catalogue of his debts read from a piece of birch-bark, and then to the Lord’s Prayer, is he experiencing two languages, or one? Compare the following assertions by modern linguists: (i) ‘all the evidence says that Old Church Slavonic and Rusian belong to a single language’; (ii) ‘the most striking feature of East Slav writing is the juxtaposition of two languages’; (iii) ‘[the language reflected in early Novgorodian sources is] simply a dialect of the Late Common Slavonic language’; (iv) ‘we must accept that there were . . . two types of Early Russian literary language’. Some of these assertions relate to language in general, others specifically to written language, but the underlying question is the same.

Linguistic argument alone cannot produce an adequate answer. There is no purely quantitative measure – a particular number of distinct phonological, or morphological, or syntactic, or lexicographic features – which determines whether the Rus version of Church Slavonic and the written derivatives of East Slavonic should be regarded as separate languages. Linguists can plausibly assert that substantial elements of Church Slavonic might have been incomprehensible to an audience of ordinary East Slavs, but comprehensibility is not the paramount criterion. I find great difficulty in comprehending some varieties of writing and speaking in English (such as computer manuals, or specialist discourses on literary theory, or the Statutes and Ordinances of the University of Cambridge, or the dialogue in a Newcastle pub), but I have no difficulty accepting that the language is English. What matters is perception: the perception of those who see themselves as within – or outside – the linguistic community.To put it crudely: language does not define community; community defines language. The wider community of early Rus – the silent majority – cannot tell us what it thought, but for the community of articulate Christians there was no doubt: the whole point of Church Slavonic lay in its affinity with the native tongue, in the fact that it was not Latin or Greek or Hebrew, in the fact that it was therefore, in principle at least, accessible. No source ever suggests that there might be two written languages, or even that there might be different languages for writing and speaking: ‘The Slavonic tongue is one.’

I wish all my readers a happy new year — удачи всем и здоровья!


  1. A fascinating post. This conversation brings my mind around to the transition during Pushkin’s time from one literary language to another. In Pushkin, one can find elements of what one might consider “other” Slavic languages. He was a sort of unifying force between the spoken and written language.

  2. Like Common Slavonic, they are abstractions, and hence contentious: should they, for example, be regarded as distinct languages, or as variants of the same language? If viewed separately, is either of them in fact an entity, or do they, too, dissolve into their own subvariants? If viewed together, what is their interrelationship?

    Well, of course the resounding answer to all these questions is “Yes”. The whole essentialist mess reminds me of Engineer Scott’s definition of Hilbert spaces (legitimately enough) as “having as many dimensions as ye need for the solving of the particular problem.”

  3. La Horde Listener says

    How cool. On a tangent, I heard Lemko language for the first time, sung by the tribal folk band Rastaban, the song “Hore Dolom”. What a great language, perfect for romance, lullabies…plus Rastaban’s musicians look and sound like a dream. Happy New Year! Melanii.

  4. Happy New Year! And yes, thanks for the most interesting post. Is the final quotation (and title of the post) attributable to some specific source, or just a general restatement of the conclusion?

  5. It’s a quotation from the Primary Chronicle, а язык Словенскї єдин.

  6. David Marjanović says

    See also: species problem.

  7. J.W. Brewer says

    Not a perfect parallel, but I just got back from a short tropical trip with wife and youngest child to Antigua, and you can ask similar questions about the relationship between the local variety of written standard English (accentless, but with Br spelling conventions and perhaps a few local lexical items), the local variety of spoken standard English (West Indian phonology which is quite distinctive from other major world varieties of spoken English), and the local creole (English-based rather than French-based as on some other islands, but far from completely comprehensible to Anglophone outsiders esp when spoken quickly). Although maybe the ability to code-switch into the local-phonology version of Church Slavonic was less common among inhabitants of early Rus, not least due to lack of universal primary education in the prestige variant?

  8. Happy New Year!

    By the way, that video of representatives of many of Russia’s minority nationalities saying “Happy New Year” in their languages is well worth watching if you’ve ever been curious about what Evenki, Yukaghir, and so on sound like.

  9. David Eddyshaw says

    There are yet more extreme cases than Church Slavonic/Russian, where at the very least the two are closely related, even if you decide that they aren’t the same language.

    Arabic diglossia often involves a much greater gap; and the ultimate example would be Kanbun – which is Japanese. I suppose. It isn’t Chinese, anyway. Is it?

    The matter shades into the question of heterograms that we discussed a while back, like the Achaemenids writing their Persian as Aramaic.

  10. David Eddyshaw says

    It reminds me a bit of

    Roger Wright’s take on Mediaeval Latin, which is essentially that people wrote what looks to us like Latin but actually pronounced it as whatever their local Romance vernacular was until the Carolingians introduced some hardcore spelling pronunciation, and – viola!- Mediaeval Latin.

    As far as I can make out from a position of complete lack of relevant specialist expertise, the idea seems not to have caught on. There’ll be LH-niks who know all about it, though …

  11. David Eddyshaw says

    Voilà, not viola. What *was* I doing last night?

  12. David Eddyshaw says

    I am inspired by the wonderful Youtube video to say to everyone:

    Barika ne ya yuumpaalig!

  13. Viola and walaa are becoming very common alternative spellings, in English anyway.

  14. Yeah, I just assumed you were making a standard-issue joke. My wife and I occasionally say “Viola!” to each other.

  15. Ein var þá tunga á Englandi sem í Noregi ok í Danmörku. ‘One [a unity] was then [before 1066] the tongue in England as in Norway and in Denmark’. Well, not really, but not really not, either.

  16. David Eddyshaw says

    In Gordon’s “Introduction to Old Norse”, selection 17 is an Old Norwegian account of the Battle of Stamford Bridge; it represents an English peasant as saying to a Norwegian fugitive (in Norse) that he recognises him to be a foreigner from his accent. (It doesn’t end well for the peasant.) The commentary doubts the historicity, alas.

  17. David Eddyshaw says

    It’s no more bizarre, really, than the PRC insisting that Cantonese is the same language as Mandarin.

  18. he recognises him to be a foreigner from his accent

    Of course he might have said the same to a man of Kent, or even merely one from Lancashire (the Wars of the Roses were a late echo of what was once a perpetual quarrel two separate nations).

  19. David Marjanović says

    the Achaemenids writing their Persian as Aramaic

    Not the Achaemenids, but the Sassanids pining for Achaemenid glory – hilarity ensues.

    Well, not really, but not really not, either.

    I’m gonna steal this.

  20. David Eddyshaw says

    @David Marjanović:

    I meant Achaemenids. You’d dictate your letter in Zarathustra’s good Persian to the the scribe, who’d write it in Aramaic (with the words backwards, but hey); at the far end, your correspondent’s scribe would read out the Aramaic as Persian.

    The Pahlavi scribal atrocities arose from this later.

  21. I’m gonna steal this.

    Do, by all means. Though it seems perfectly commonplace to me, Dr. Google insists that I am its onlie begetter, at least so far as writing is concerned. A good example of linguistic creativity, really: “the effortless ability to communicate and comprehend novel expressions and new ideas” (Birnbaum). This is usually mixed up with silly notions about infinity (as if linguistic expressions were countably infinite, or even uncountably infinite!) but really has to do with how the behavior of complex devices is unforeseeable, something probably first articulated by Maurice Wilkes in the 1940s: “[T]he realization came over me with full force that a good part of the remainder of my life was going to be spent in finding errors in my own programs”; that is, understanding their unpredictable consequences.

    I just repeated the substance of the above to Gale, and said “All the time we utter sentences that no one has ever said before” — which indeed, on checking it, has no Google hits either. And yet I said it and she understood it at once. “What a piece of work is man! […] In apprehension, how like a god!”

  22. David Eddyshaw says

    Reminds me also of the computer science riddle: what language is this code in?

    #!/bin/csh -f

    cat >! foo.c <<QQ
    main(){printf("Hello, world!\n");}
    cc -w foo.c
    /bin/rm foo.c a.out

  23. Perl, obviously. (Perl will respect a shebang that doesn’t contain the string “perl” and will exec() the specified interpreter in place of itself.) No QQ.

    The WP article on polyglots shows an example that is simultaneously valid C, PHP, and bash, and has links to many other such programs, including one that is simultaneously valid zsh, bash, (da)sh, Ruby, Perl, Tcl, C++, C, Haskell, Python, HTML, JavaScript, Perl6, and a Makefile — at least.

    By the way, that “in apprehension how like a god” phrase may be due to a punctuation error after all. The Second Quarto says “What a piece of work is a man, how noble in reason, how infinite in faculties, in form and moving, how express and admirable in action, how like an angel in apprehension, how like a god!” This fits with Elizabethan thinking better, since it was a cliche of the Renaissance that angels could apprehend but not act. Also, “express and admirable” goes better with “in action” than with “[in] moving”, which is in this version associated with “infinite” along with “in faculties” and “in form”.

    Punctuation as a plot point in Ralph Roister Doister.

  24. It should run in csh, tcsh or zsh — >! breaks bash, ksh and Bourne shell. (If treated as an executable file on a ‘standard’, ‘modern’ UNIX or Linux system, it will of course run under csh).

    It does define an external subroutine in C89 or so, but that’s just one way of skinning the cat.

    It illustrates, however, that the concept of the ‘language programmed in’ is quite non-restrictive on most modern OS’es. Calling out to scripts / binaries in other languages, or even defining them dynamically, is just using the toolbox as designed.

    The exception being, of course, Java whose community delights in coding up alternative implementations of everything and shipping them with products instead of linking to system libraries, so that you have 10 or 12 exciting combinations of security misfeatures to keep track of instead of one. Not to mention that each copy of the JRE has its own copy of the root certificates list, to be found and updated when things break. But even a Java programmer could conceivably break through their brainwashing and call out to better suited domain specific languages if they wanted.

    (This is a blog about languages, right?)

  25. David Eddyshaw says

    On my system, I would say if pressed to answer what is in fact a scarcely meaningful question, that it runs in bash (which hands it off to csh, which compiles and runs the C program and then deletes it …)

    It occurred to me as an analogy of the main thing we’ve been discussing; thus even if you think that Kanbun “is” Classical Chinese, it’s still Japanese when it’s (as it were) “invoked” from Japanese. Likewise Church Slavonic from Russian …

    I wrote my favourite snippet from the Analects of Confucius (君子不器) on a whiteboard in an idle moment the other day and somebody asked me what language it was in, to which I replied “When I write it, it’s in Japanese.”

  26. David Eddyshaw says

    (like Pierre Menard’s Quixote)

    JC’s answer to the riddle is obviously correct, in retrospect. *All* programs are in Perl until proven otherwise.

  27. It’s not Java programmers’ fault that JNI, the escape from Java, is only a square centimeter wide and slow as molasses/treacle in January (though to be sure the high today is 41F/5C).

  28. Note that JC was the only one to answer the actual question (“what language is it ‘in’?”).

    David and I restricted ourselves to the question of which interpreters/compilers would actually in some sense interpret the whole file successfully, if only to parse out literal substrings for other interpreters/compilers.

    And, well, all language interpreters/compilers operate on the assumption that the input is in that language until the opposite is proved (called a syntax error). It’s just that Perl is designed so that’s it’s unusually hard to prove that something won’t run in Perl — the opposite of a minimalist language 🙂

  29. It occurred to me as an analogy of the main thing we’ve been discussing; thus even if you think that Kanbun “is” Classical Chinese, it’s still Japanese when it’s (as it were) “invoked” from Japanese. Likewise Church Slavonic from Russian …

    Admittedly I wouldn’t want to call kambun necessarily Classical Chinese, but I do hesitate to call it Japanese, though clearly some forms of it, like the epistolary style, seem to have been written and read fluently enough. The fact that Russian and Old Church Slavonic were related is to me the key difference that makes the comparison doubtful.

    On the contrary the case of the Achaemenid scribes seems a very good match (though the far greater literacy of Medieval Japan means the latter case is much more complicated). Even if Persian speech could be dictated directly into, and later more-or-less directly re-extracted from, a written Aramaic that neither involved native Aramaic speakers nor may indeed have been 100% comprehensible to them, would we really want to say the language of such surviving tablets was a form of Persian?

  30. Kanbun is of course Japanese. Most early-Meiji writers basically wrote spelt-out Kanbun, when they considered themselves as writing (serious) Japanese.

  31. Granting up front that this is an issue of ongoing debate, I’m at least very comfortable saying that the certainty of “of course Japanese” isn’t justified by any criterion more refined than “was written by and for Japanese people.”

    It’s really a difficult issue, complicated by reading practices counterintuitive to the dominant romanographic paradigm. But the important role of kambun-style diction in the Meiji experiments doesn’t cut it. From a post-mid-twentieth-century perspective it might well seem that such a style was just “kambun spelled-out,” but this is only because by that time, kambun was the only visible remaining habitat for many features of style which had earlier also flourished in Japanese expository prose most definitely *not* of the kambun variety (coexisting, in fact, in contradistinction to writing practices that were). It’s a species of recency illusion. Many of these features were indeed from kambun ultimately, but had long been naturalized in written Japanese.

  32. 2+2=4

    What language is that?

  33. From a post-mid-twentieth-century perspective it might well seem that such a style was just “kambun spelled-out”

    So what would the 19th century perception have been?

  34. “whole phrases in Russian text may have the Church Slavonic appearance, in other words, within the framework of modern Russian, an entire Church Slavonic phrase can be constructed.

    A good example would be a phrase “Da zdravstvuet sovetskaya vlast!” (Long live Soviet power!)

    This phrase unquestionably belongs to the Russian literary language, and it is obvious that it is a relatively new addition. However, from a formal point of view this phrase can be considered as Church Slavonic: indeed, the vocabulary and grammar of this phrase meets the standards of the Church Slavonic language.

    Strictly speaking, it doesn’t contain a single Russian word (in the sense of the genetic origin):

    Here, zdravstvuet and vlast are typical Slavonicisms (with nepolnoglasie), the same can be said about the word Soviet (with “clarification” of reduced vowels); Likewise, the syntax (da + indicative) is Church Slavonic in origin.

    So, we have, in fact, Church Slavonic phrase. However, we can not express the same content without resorting to Slavism, ie, by using only native Russian forms. We may consider this phrase as Church Slavonic, but we can not translate it into Russian.

    We could translate this phrase into Old Russian language (since the old language, in contrast to the modern, had not yet developed organic synthesis of Church Slavonic and native Russian forms) and it would look something like this: A свѣтскѣй волости здоров* быти. “(C)

    B.A.Uspensky” A Short History of Russian literary language

  35. David Eddyshaw says


    What language is that?”


  36. This is tangential – but looking at traditional Cyrillic numerals just now, it dawned on me that you could completely Cyrillify the Greek alphabet if you wanted to.

    Ξεσκεπάζω την ψυχοφθόρα βδελυγμία.
    Ѯескепа́зѡ тин фухофѳо́ра вделумнíа.

  37. (ѱ rather than ф, sorry.)

  38. However, we can not express the same content without resorting to Slavism, ie, by using only native Russian forms. We may consider this phrase as Church Slavonic, but we can not translate it into Russian.

    Is this really true? Borrowings are pervasive in English, and when we try to write without them, it is very hard (and we have to invent new technical vocabulary), but we can always paraphrase away non-native syntax (indeed, some people claim there is no non-native syntax in English, but I think that is a by-product of a false theory of syntactic change).

  39. Michael Everson’s demonstration that Cyrillic and Coptic are more alike than either is like Greek, their common ancestor. (The unification of Coptic and Greek in Unicode was eventually undone as a soul-destroying abomination.)

  40. David Marjanović says

    demonstration that Cyrillic and Coptic are more alike than either is like Greek, their common ancestor

    Their common innovations are shared by the Greek of late Antiquity and the early Middle Ages, while modern Greek fonts are thoroughly classicist…

  41. Ксёнѕ Фаўст says

    The relationship between Old/Middle Polish and Old/Middle Czech somewhat resembles the one between older forms of Russian and Church Slavonic. There was a lot of lexical borrowing and a lot of phonetic interference as well (people imitated certain Czech phonetic features almost as if it were a prestigious dialect of the same language; some of them stuck: serce instead of the regular sierce, wesoły rather than wiesioły, hardy rather than gardy, hańba in place of gańba; the regular forms can still or could still until recently be heard in some dialects).

  42. @SF Reader: And of course “Long live” isn’t Modern English syntactically either, except in that phrase.


  1. […] Hat considers what a language […]

Speak Your Mind