(Serbo-)Croatian: A Tale of Two Languages—Or Three? Or Four?
With all the media brouhaha about Croatia’s ascension, one of our key issues at GeoCurrents has been largely ignored: the issue of the Croatian language. Multilingualism is central to the European Union’s cultural diversity. The European Commission employs a permanent staff of around 1,750 linguists, 600 staff interpreters, 3,000 freelance interpreters, and 600 support staff, making it one of the largest translation and interpretation services in the world. Still, this only amounts to some 25 staff interpreters per language, as the EU now has 24 official languages; their website allows one to read and/or hear a short text in Bulgarian, Czech, Danish, Dutch, English, Estonian, Finnish, French, German, Greek, Hungarian, Italian, Irish, Latvian, Lithuanian, Maltese, Polish, Portuguese, Romanian, Slovak, Slovene, Spanish, Swedish—and now Croatian as well.* But basic issues about what constitutes the Croatian language are far from settled.
As Linda B. Glaser of the Cornell Chronicle put it, “nowhere has linguistic research involved more discord than in the Balkans”. I recall that in the late 1990s a newly established online forum dedicated to Slavic languages and linguistics quickly devolved into a fierce argument about what the language(s) of former Yugoslavia should be called.** The online forum was then promptly discontinued by the moderators. Caution in the use of language labels in the Balkans is indeed called for, as they may fuel tensions in this volatile region where the linguistic situation is especially complex.
Much of the following exposition draws heavily on my discussions with my former colleague, Professor Wayles Browne of Cornell University, an expert in “the Serbo-Croatian area (Bosnian, Croatian, Serbian)”, as he carefully phrased it on his webpage. Browne received his Ph.D. from the University of Zagreb in 1981 and has visited the area on numerous occasions since then, working on a wide spectrum of issues in South Slavic languages ranging from accent rules to clitic placement, and from loanwords to linguistic identity. “In the Balkans, whatever country you’re in the language serves as a national symbol. It’s a big part of their identity,” Browne said in an interview for the Cornell Chronicle. In a recent email exchange with me, he framed the issue as follows: If you ask people in post-Yugoslavian countries what language they speak, “Serbs say “We speak Serbian” (“Mi govorimo srpski”), Croatians say “We speak Croatian” (“Mi govorimo hrvatski”), and Bosniacs (and some other people from Bosnia who do not want to get into ethnic politics) say “We speak Bosnian” (“Mi govorimo bosanski”)”. It is hard to miss the fact that apart from the labels, the languages themselves are very similar. The same point is made by Geoffrey Pullum in a Chronicle of Higher Education article “A Trinity of Languages”: although each pack of cigarettes sold in Bosnia and Herzegovina carries the same “Smoking kills” label thrice—in Bosnian, Croatian, and Serbian—each warning sounds exactly the same. In Croatian and Bosnian, moreover, it is spelled identically. (Ironically, he adds, the triple warning hardly makes any difference, as the Serbs still smoke more than any other nation in the world, according to the data published in The Economist.) As Pullum writes:
“The risk of death doesn’t bother the people here; what bothers them is the possibility that any ethnic group might miss out on being treated exactly the same as the others. So Bosniaks, Croats, and Serbs must all be explicitly warned in their own separate languages.”
So how many languages are spoken in the “Serbo-Croatian area” (to use Browne’s term)? One? If so, what should it be called? “Serbo-Croatian”, as it was deemed until the 1990s (see the older map on the left)? “BCS” for Bosnian-Croatian-Serbian, as linguists now typically write, always using alphabetical order, lest one linguistic group takes offense? “BCMS”, lest we forget Montenegrin? Or even “Illyrian”, a term used by some scholars in the 1830s and 1840s for a unified literary language? (Not to be confused with ancient Illyrian, a poorly known non-Slavic language.) Or is the “Serbo-Croatian area” home to two distinct languages, Serbian and Croatian, as the Novi Sad Agreement of 1954 stated? (More on the agreement below.) Or are there three separate languages, with Bosnian added to the mix? Or four, adding also Montenegrin? As we shall see in the remainder of this post, each of these positions has some arguments in its favor, so no definitive answer can be given.
Contrary to the common perception, this search for linguistic identity began long before the dissolution of Yugoslavia, dating back to before the creation of the country in the wake of World War I. Already in the second half of the 19th century, different labels were used for the emerging literary language of Croatia: “Illyrian”, “Croatian”, “Croatian or Serbian”, and so on, as can be seen from the images on the left (reproduced from Milan Moguš’s A History of the Croatian Language: Toward a Common Standard, pp. 172, 174, 202). The term “Serbo-Croatian” (alongside “Croato-Serbian”, both spelled without a hyphen in the language itself) was officially approved by the Novi Sad Agreement, which resulted from a meeting of Serb and Croat linguists in December 1954 and formally established equality of the two constituent tongues. However, this unity and equality was short-lived. In March 1967, a number of Croatian cultural and scientific institutions issued the Declaration on the name and status of the Croatian literary language, which called for the use of four official languages in Yugoslavia: Slovenian, Croatian, Serbian, and Macedonian. State authorities launched a merciless attack on the Declaration and its signatories, which only intensified after Josip Broz Tito, president of Yugoslavia, intervened in 1971. Work on the Croatian dictionary was then brought to a halt. An orthographic manual, Hrvatski pravopis, published in Zagreb in 1971, was destroyed by the authorities (a surviving copy, however, was smuggled out of the country and published in London in 1972). Members of the related Croatian Spring Movement, which sought enhanced autonomy from Belgrade, were persecuted as “counter-revolutionaries” and some were imprisoned. It was not until the dissolution of Yugoslavia in the 1990s that the seeds of the Croatian Spring Movement were allowed to blossom—bringing the issue of language identity back into focus.
Curiously, Google Translate offers translation between Bosnian, Croatian, and Serbian as if they were distinct languages—unsurprisingly, it does a fine job of it: after all, for the most part all it has to do is map strings of words to themselves, or at most respell them. When it comes to writing, a strong split differentiates Croatian and Serbian, as Serbian is officially written in the Cyrillic alphabet, similar to that of Russian, owing largely to the influence of the Orthodox Church. Yet despite the symbolic and historical significance of Cyrillic, some Serbians advocate switching to the Roman alphabet. Unlike Serbian, Croatian is written in the Roman script due to the influence of the Roman Catholic Church, which began when the ancestors of modern Croats were converted to (Western) Christianity by their Frankish rulers.***
In historical terms, however, the situation is considerably more complicated. Although the Croats adopted Latin as their ritual language after their conversion to Christianity, they subsequently turned to Old Church Slavonic, a language much closer to the their vernacular than either Latin or the indigenous Dalmatian Romance tongues. Old Church Slavonic was originally written in the Glagolitic alphabet, invented ca. 863 CE by Saint Cyril, for whom the Cyrillic alphabet is named, and his brother Methodius. In accordance with the Byzantine tradition of autonomy and equality for all the languages of eastern Christianity (Georgian, Armenian, Syriac, Coptic, etc.), Glagolitic was created as “a unique and homogeneous graphic system” (Horace Lunt, Old Church Slavonic Grammar, 2001, p. 15); its letters are markedly different from the corresponding letters of the Greek or Roman alphabets. Until the 12th century, Glagolitic was the only script used for Croatian (Roman alphabet was used to write Latin and Italian). As late as the 17th century, Glagolitic was still widely used in Croatia; according to Lunt, “a few priests in northern Dalmatia still use glagolitic missals to this day” (ibid, p. 16). Eventually, Glagolitic was replaced in most of its former range by Cyrillic, whose letters look more like those of the Greek, Roman, and in some cases Hebrew alphabets. (The image on the left shows Glagolitic and corresponding Cyrillic letters.)
Many linguists brush such distinctions in the writing system aside, as they are only interested in the spoken language. “But when you look at how people use language in practice, writing is very important and standardization is important and prestige factors matter,” explains Wayles Browne in his Cornell Chronicle interview. In the former Yugoslavia, the distinctions in the way people speak—and even more so, in the way people think they speak—are very subtle indeed.
Let’s begin by looking at the labels people use for their language in the “Serbo-Croatian area”. It has been claimed—even by linguists—that the name of a language is always derived from the name of the people who speak it, not from the country in which they live. This generalization does seem to work for most languages in Europe and elsewhere. For example, Danes, though they are from Denmark, speak Danish rather than “Denmarkish”; Poles speak Polish, not “Polandish”; and there was never a “Czechoslovakian language”. Similarly, Austrians speak German, not “Austrian”, as do most of the Swiss.**** Canadians speak either English or French, not “Canadian”, and so on. But in the Balkans, most names for peoples, languages, and countries are indistinguishable: Albanians speak Albanian in Albania, Bulgarians speak Bulgarian in Bulgaria, Slovenes speak Slovenian in Slovenia, and so on. The same is true in the “Serbo-Croatian area”, as Browne’s description above (“Mi govorimo…”) shows. The only exception is the Montenegrins who typically refer to their language either as “Serbian” or as “Montenegrin”. Note that the label “Bosnian” is typically used to designate the language of Bosniaks (defined largely by their Muslim faith) in Bosnia. The term Bošnjak (‘Bosniak’, adjective: bošnjački) is rarely applied to their language (the Wikipedia map reposted on the left is a rare exception). However, as an ethnic designation, this term has existed for centuries; since independence, Bosniaks themselves have shown a clear preference for using this term instead of Musliman ‘Muslim’. However, Bošnjak/bošnjački should not be confused with Bosanac (‘Bosnian’, adjective: bosanski), which can apply to anyone originating from Bosnia. Here we see the Eastern European fixation on the distinction between “nationality” in the sense of country of origin and “nationality” in the sense of ethnic (read, “religious or linguistic”) belonging; Russians are equally keen to distinguish rossijane (anyone from Russia) and russkie (ethnic Russians).
The country-language correlation, however, falls apart if we consider the census data (here, from 2006) more closely. For example, as can be seen from the map posted above, not everybody in Bosnia speaks “Bosnian” (here and below, I use the terms in quotes to designate labels that people assign to their own language). “Bosnian” is what people claim to speak in the core area of Bosnia and Herzegovina, as well as in the northwestern area around Bosanska Krupa. The official state of Bosnia and Herzegovina, however, is politically divided into two ethnically defined units, the Federation of Bosnia and Herzegovina and the Republika Srpska; in the former area, some people label their language “Croatian”, while in the latter the term “Serbian” is used. The language labels used in Bosnia thus correspond most closely to the ethno-religious designations: Muslims Bosniacs call their language “Bosnian”, Eastern Orthodox Serbs refer to it as “Serbian”, and Catholic Croats use the label “Croatian”. However, this equation of ethnic and religious labels is a relatively recent one. It emerged effectively during Tito’s rule, when “uneducated political commissars and party functionaries decided the fate” of languages and peoples (Robert Greenberg, Language and Identity in the Balkans: Serbo-Croatian and its Disintegration, p. 115), much like their “colleagues” in the Soviet Union did. According to Greenberg (ibid, p. 32), “by the late 1960s, it was anomalous for a Serb to self-identify as a Catholic or Muslim, just as it was most unlikely for a Croat to self-identify as Orthodox or Muslim”. In 1971, the Yugoslav authorities elevated the Muslim Slav population to the status of a constituent nation in Yugoslavia, thus formally equating religious and ethnic identity and creating “the forerunner to the post-1992 Bosniac people” (ibid, p. 32).
The correlation between ethno-religious and linguistic labels is also evident in other areas of the former Yugoslavia. For example, the two maps of Montenegro posted here—ethnic (on the left) and linguistic (on the right)—show that Bosnians in the country speak “Bosnian” or “Bosniak”, Albanians speak “Albanian”, Serbs speak “Serbian”, while Montenegrins speak either “Serbian” or “Montenegrin”. According to this map, the latter label is only used in the Cetinje region.
Nor is Serbia uniformly “Serbian”-speaking. Linguistic heterogeneity is particularly marked in the northern autonomous province of Vojvodina. The juxtaposition of ethno-linguistic and religious maps (from Wikipedia) shows once again a near-perfect correlation: areas of Eastern Orthodox majority or plurality are Serbian-speaking, the two areas with a Protestant majority/plurality are Slovak-speaking, and areas of Catholic majority/plurality are Hungarian-speaking.
As for Croatia, according to the map posted above, most people speak “Croatian” (with the exception of several small “Serbian”-speaking pockets). Yet if we consider closely how people actually speak, as opposed to what they think they speak, a more complex picture emerges, with three distinct dialects spoken across Croatia, named after their pronunciation of the word for ‘what’: što or šta in Shtokavian, ča or ca in Chakavian, kaj or kej in Kajkavian. Chakavian is spoken by the inhabitants of the Dalmatian coastal area and on some of the islands in the Adriatic, as well as on the Istrian Peninsula. The northern area near Zagreb is home to Kajkavian, and Shtokavian is spoken elsewhere (see map on the left).
Note that different maps vary as to their depiction of dialectal areas, with Shtokavian being much less prominent on maps reflecting the linguistic situation in the past, such as the Wikipedia map on the left, which purports to depict “Serbo-Croatian dialects prior to the 16th-century migrations”. As a result of these migrations, which were due mostly to the pressure from the Ottoman Empire, the distribution of dialects was considerably changed; Chakavian dialect in particular lost much ground and was reduced to a narrow coastal strip.
While the Chakavian and Kajkavian dialects are spoken only in Croatia, different forms of Shtokavian are found in other former Yugoslavian countries. Different scholars distinguish two, three, or more forms of Shtokavian, as shown on the map posted above and the two maps on the left. The breakup of Shtokavian into subdialects is based mostly on the reflex of a certain old vowel, “jat”. Thus, the main split is into Eastern, or Ekavian, and Western, or Ikavian/Ijekavian, subdialects. Eastern Shtokavian is spoken in most of Serbia, as well as in Montenegro. Western Shtokavian (which some scholars further subdivide into separate Ikavian and Ijekavian varieties) is found in parts of Croatia, as well as in most of Bosnia and Herzegovina, and a western portion of Serbia. The dialectal picture is completed by Torlakian, spoken in southeastern Serbia; some dialectologists consider it a separate dialect, while others regard it as merely a transitional form linking Shtokavian to Macedonian and hence Bulgarian. The upshot of the preceding discussion is that the “Serbo-Croatian area” exhibits great variety when it comes to pronunciation, spelling, word choice, and even grammar. Moreover, these dialectal differentiations are important for national standard languages, which go back at least to the mid-1800s. Unlike in many other parts of the world, standardized forms really matter in the Balkans. “People get to criticize each other for not obeying the standard, and they can draw some unwarranted conclusions, like ‘anyone who speaks like that must be lazy’,” explains Wayles Browne.
But the differences between the standard Bosnian, Croatian, and Serbian are even more subtle than those between regional dialects because all three groups use a form of Shtokavian for their national languages. In particular, most Serbs in Serbia, including speakers of Torlakian in southeastern Serbia, use as their standard the Eastern (Ekavian) variety of Shtokavian. Bosnians, Croats, and many Serbs who live outside Serbia, however, use as their standard the Western (Ikavian/Ijekavian) form of Shtokavian. For example, almost everyone in Bosnia speaks Shtokavian, in either Western or Eastern form, and almost everyone regards Western Shtokavian as their standard, although some refer to it as “Serbian”, some as “Croatian”, and some as “Bosnian”. But if these geopolitical labels are set aside, one can now speak of the Bosnian national standard language, in as much as dictionaries and orthography books (pravopisi) from neither Croatia nor Serbia are regarded as authoritative in Bosnia. Thus, Bosnians use the same standard language, but apply different labels to it, depending on their religious affiliation.
The situation in Croatia is the opposite of that in Bosnia: while it is dialectally highly differentiated (with three major dialects, as discussed above), and “all three dialects were used by Croats as the ‘raw material’ upon which to base their literary language” (Moguš, p. 12), yet all the inhabitants of Croatia share the same standard language, based primarily on Western Shtokavian (even more precisely, on the Eastern Herzegovinian variety of Shtokavian). This situation is particularly unusual because Western Shtokavian is not (and was not) the dialect of the capital, Zagreb. Unlike the French, who chose the dialect of Paris (Île-de-France) as the basis for their standard, or the Russians, who picked the dialect of Moscow and St. Petersburg, the Croatians did not use Kajkavian as the basis for their national standard. Until the standardization of Croatian began in the early 1800s, some Croatian writers wrote and published in Kajkavian and some wrote and published in Chakavian. However, in the mid-1800s standardization efforts converged on Western Shtokavian because it had the largest number of native speakers at the time (Moguš, p. 169).
In fact, the creation of standard Serbian and standard Croatian went along largely parallel lines. As Browne writes in “What is a standard language good for, and who gets to have one?”, the two national standards were “based on very similar material, to some extent the same material (as when Vuk Karadžić’s Serbian Dictionary was taken as part of the Croatian word stock in the late 1800’s).” But the two national standards still differ in their pronunciations, vocabulary choices, and sentence structures. For example, “the Croats followed a policy of purism in issues of vocabulary, while the Serbs were largely concerned with remaining true to the vernacular language” (Greenberg, p. 47). However, the purist policy of eliminating internationalisms and perceived Serbianisms, which characterizes the standard Croatian today, goes back centuries; it was embraced with particular fervor at the time of the Croatian Fascist state (1941-1945). As a result of this purism, standard Croatian relies heavily on “native Croatian” words, including archaic or newly coined ones, turning occasionally to words found in the Kajkavian or Chakavian dialects (ibid, p. 121). Examples include the Slavic-derived zračna luka (literally ‘air harbor’) rather than aerodrome, munjovoz (literally ‘lightning vehicle’) instead of tramvaj ‘tram’, osposoba instead of kvalifikacija ‘qualification’, and so on. Standard Croatian retained native names for calendar months, whereas standard Serbian uses the Gregorian januar, februar, mart, and so on. The Croats play nogomet (literally ‘foot-throwing’), while the Serbs play fudbal ‘soccer’. Some of the differences between the two national standards are morphological: our readers in Zagreb use preglednik ‘browser’ to read GeoCurrents, while those in Belgrade use pregledač. Although standard Croatian has loanwords from Latin and Hungarian, prescriptivist linguists in Zagreb tend to view Turkish and Russian borrowings negatively, while displaying “tolerance towards borrowings from lending languages of nations for which the Croats have felt cultural affinity”: French, Italian, Hungarian (ibid, p. 123). In contrast, their colleagues in Belgrade have made known their bias against German loanwords and acceptance of Russian loanwords (ibid, pp. 53-54, 124). These prescriptivist efforts purposefully result in reduced mutual intelligibility in the “Serbo-Croatian area”. As for the future, it remains to be seen whether “after the next generation of Croats, Bosniacs, Serbs, and Montenegrins assume positions of power, they truly will not be able to understand one another any longer” (ibid, p. 167) or if the integration of Croatia—and possibly of other former Yugoslavian countries—into the European Union will reverse this process.
* Five other languages—Catalan, Galician, Basque, Scottish Gaelic, and Welsh—are not official languages of the EU but have a semi-official status: treaties are officially translated into those languages and citizens of the EU have the right to correspond with the Union’s institutions using them.
** Slovenian and Macedonian, however, are exceptions here. Though their respective countries used to be constituent parts of Yugoslavia, Slovenians and Macedonians speak distinct languages, related to yet not mutually understandable with the “Serbo-Croatian”. Macedonian is most closely related to Bulgarian; some Bulgarians indeed consider it a dialect of Bulgarian.
*** Like Croatian, Bosnian is typically written in the Roman alphabet.
**** Other Swiss citizens speak French, Italian, or Romansh.
« Who Serves in the U.S. Military? Mapping...
A Conservative Core and a Reformist... »