The (Ir)relevance of Language

Those accustomed to understanding and evaluating culture through linguistic material — from the sociolinguist to the most textually fetishizing post-modernist — have a vested interested in language remaining important to understand society. In many ways it is. But it is equally important to assess just how staggeringly irrelevant the fascinating phenomenon of language can often be.

The incredible weakness of language on its own as a force for cultural or political unity is exhaustively documented. Throughout history, ideology, religious authority, religious schism, class-interests, commercial gain, secular power, imperial interests, or simple geography have had a far greater power to unite and divide ethnic groups and polities. At most, language can help to cement pre-existing cleavages and groupings, as in the manner of a regional lingua franca, but barring extremely unusual circumstances it cannot by itself create cultural or political links where they are not already felt for other reasons.

Even cultures and nations famous for linguistic conservatism and for preserving extraterritorial languages, like Jews and Roma, are in fact characterized by high levels of multilingualism and linguistic variety. In both cases ethnic commonality can surface in common language (e.g. liturgical and inter-communal use of Hebrew, or the Balkan-influenced register of of Finnish Romani which uses more conservative inflectional patterns and fewer Finnish loans.) But it can just as easily surface not in a common language so much as in retention of common lexical items. Jews have, until the 20th century, traditionally maintained a variety of different vernaculars based on one or another gentile vernacular, though often with preference for different dialectal variants than those more current in gentile speech. In cases where the speakers become cut off (whether via expulsion or emigration) from the earlier gentile linguistic community, they may take their vernacular with them into "compound exile" as it were — thus Eastern Yiddish winding up spoken in, and influenced by, Slavic territory (and subsequently everywhere from the US to Israel.) Or Ladino brought to, and influenced by, the Balkans. The one and only thing Jewish languages, and distinctly Jewish speech varieties, have in common is a layer of loanwords from Aramaic and Hebrew (though some may have a further layer of loans from other Jewish languages, such as the Yiddish element in some Jewish sociolects of American English, and the considerable Ladino element in many forms of Judeo-Italian.) This layer too, though, varies greatly in size and character. Sometimes it is restricted to religiously salient or otherwise culturally Jewish vocabulary. At the other extreme, the semitic loan component may encompass words from all sorts of registers and semantic spheres including words for "face", "answer", "really", "even", "moment, sec" and "orphan." It may be almost trivially small, as in Baghdadi Judeo-Arabic or spoken varieties of Ladino. It may be sizable, as in Judeo-Piedmontese and Judeo-Occitan. It may also be staggeringly enormous, occupying a large portion even of the core lexicon, with Semitic derivational morphology remaining productive, as in Eastern Yiddish. (In the latter case it is just possible, though I've yet to be convinced, that the ancestral population at some point absorbed, or was even founded by, Judeo-Aramaic speakers who underwent a language shift but retained a large North Semitic lexical component — which they then would have been able to control and augment through access to Hebrew literacy.)

It tends to be a prior sense of ethnic or religious commonality which creates the links which linguistic commonalities may, or may not, spread along. Not the other way around. Because ethnicity or sectarian distinctions may closely track with whom one most associates with (or is allowed to associate with) linguistic variations may often develop along almost entirely ethnic or sectarian lines, as with specifically non-Muslim dialects of Baghdadi Arabic, or Sunni vs. Shi'ite dialects in Bahrain, or Black English in the US (with a verbal system that is in some ways radically different from any other variety of English.) Where common sense of ethnicity is lacking, or where religious or sectarian cleavages are particularly animous, separate languageness may well be asserted or codified. Witness Macedonian's separate status from Bulgarian, whereas Tajik, in the post-Soviet era, is treated as one of a number of varieties of Persian. The hotbutton tizzy surrounding Serbo-Bosno-Montenegro-Croatian is another case in point. It is for cultural reasons that Yiddish is not considered to be German (and would not be considered German even without the large Semitic and Slavic loan vocabulary), that Galician is not considered to be Portuguese, and that Zazaki is often insisted by (Kurmanji and Sorani) Kurdish speakers to be a Kurdish dialect. Likewise, in a counter-factual universe where Roms in the Baltic and South Italy agreed that they had nothing ethnically in common, it seems to me likely that dialects such as Abruzzese Romani and North Russian Romani would be though of as different but related languages.

This is all a matter, if not exactly of public record, then at least of generally available record. Yet language remains a point of unsubstantiated primordialist fascination, and has been the locus of a great deal of nationalist (and anti-nationalist) polemic and experimentation. We like to talk about a language as "the bedrock of a culture" and a thousand other such things, despite the mountain of empirical evidence that it is, in many ways, no such thing. It might be nearer to the truth to say the reverse: culture is the bedrock of language.

Language often feels to people like far more than it really is. It often reflects culture, but the relationship does not go both ways. One need only look at the Middle Ages for an abundance of examples.


  1. Your point is taken that a shared language does not guarantee unity, but it is equally fair to say that lack of a shared language makes unity very, very challenging.
    So "irrelevant" may be too strong a position.

    1. Roman Catholic attitudes to the Vatican may be quite orthogonal to their knowledge of ecclesiastical latin or of Italian.

    2. Lack of a shared language has not prevented Romani groups that no longer speak dialects of Romani from feeling a close cultural connection to those groups that do still speak it. A Rom from Southern Italy and a Rom from Northern England may have no language in common. (If anything, it is likely to be English.) This fact alone wouldn't bar either from seeing themselves as having something ethnically in common.

  2. A lot of what is said here about Jewish and Romani linguistic conservatism is not at all true. Both Romani languages and Medieval Hebrew have had lots of semantic shifts, innovations and borrowings on a huge scale

    1. What exactly is it that you think I'm saying that you disagree with?

      "Both Romani languages and Medieval Hebrew have had lots of semantic shifts, innovations and borrowings on a huge scale"

      Where did I say anything that would even remotely imply that I thought otherwise?

      Did you actually read what I wrote in its entirety, or did you just seize on the phrase "linguistic conservatism" and fill in the rest yourself?