Suppose you hear that a colleague is working on a language called “@t~q^M#%”. What is your reaction? What’s wrong with the language name “@t~q^M#%”? It’s perfectly unique, it consists only of ASCII characters so is eminently typable, and it has a certain beauty. But of course it lacks pronounceability, so is it good as a name? In general, we expect a language to have a name that we can use in speaking about the language, not only in writing. For this reason, nobody has suggested that traditional language names should be abolished entirely and replaced by codes such as “uppe1395” (Glottocode for Upper Sorbian) or “xtq” (ISO 639-3 code for Tumshuqese).
Now what about a hypothetical language name such as “QmncpwuX”? This name is pronounceable, at least by people with special training in phonetics and IPA. So is pronounceability everything? I would say that the requirements on English language names should be even more stringent: They should be pronounceable in English, just like all other well-formed English words (or pronounceable in Russian, if we are in a Russian-speaking context, etc.). Because language names, when used in English, are part of the English language. We talk about the Greek language, the Polish language, and the Basque language, using the English words for these languages, rather than the Elinika language, the Polski language, or the Euskara language. Thus, language names are English words, like city names (we talk about Rome, Moscow and Cairo, not Roma, Moskva or Al-Qâhira), and like names of monarchs (Pope Francis, not Franciscus; or King Frederick the Great, not Friedrich the Große). (I have elaborated on this in a recent paper called “Some principles for language names”, where I talk about naming practices discussed in the content of Glottolog.)
Of course, the speaker communities may have an opinion on how to call their language in English, so they may prefer the name Inuktitut to Eskimo, or the name Oromo to Galla. This is just as with city names: The inhabitants of Mumbai may prefer Mumbai to Bombay, and the inhabitants of Beijing may prefer Beijing to Peking in English. In such cases, English speakers outside these cities typically follow these requests, for straightforward reasons of politeness. And likewise, linguists have changed their habits and now use the language names Inuktitut and Oromo. So now the English word for China’s capital is Beijing.
But can we imagine that someone would ask English speakers to use a name that is unpronounceable in English? For instance, inhabitants of Mexico insisting on the pronunciation [mexiko] with a velar fricative? Or inhabitants of Seoul asking for the mid unrounded vowel [ʌ] in the first syllable? This would be a public relations disaster. A city with an unpronounceable name would scare away tourists or investors. (The Danish city Århus changed its name (back) to Aarhus fairly recently, presumably to facilitate international recognizability.)
So is it reasonable that speaker communities should ask outsiders to call their language by some unpronunceable name? Maybe it is comprehensible from the perspective of minorities that feel under cultural siege and want to make a clear statement of their cultural and linguistic uniqueness. Thus, we find some indigenous groups in North America using names such as “Skwxwú7mesh” (e.g. here). I take such usages to be assertions of identity, and I find it hard to take them seriously as requests for replacing older English names. English can easily accommodate Inuktitut and Oromo, Mumbai and Beijing, but one cannot expect English speakers to really use names like “Tłı̨chǫ Yatıì” (for Dogrib), or “Kanien’kéha” (for Mohawk).
Of course, linguists who work on these languages know how to pronounce the language names, and one would not surprised to hear an expert of Nuuchahnulth actually pronounce the language name [nu:tʃa:nuɬ] in a scholarly presentation. But what about their colleagues? If I want to refer to that paper in my own oral presentation, how should I pronounce the language name? Should I learn this pronunciation as well? Should I expect the same from my first-year students, as soon as they have learned the IPA? I think the idea that we should all use the native pronunciation in English has the following two problems: (i) It assumes that languages of small communities that do not have a long tradition of writing and of state power should be treated differently from languages such as Greek, Polish and Basque, and (ii) it is based on the idea that language names should not be English words, but that when talking about them we should code-switch into these languages, even if we know very little about them otherwise. Since linguists agree that all languages have the same status and the same value, I find it hard to accept that small languages should be treated differently than large and powerful languages.
So while I am all for politeness with regard to language-choice requests by minorities, I don’t think that it’s an expression of such politeness to do away completely with English words for languages. If we don’t have pronounceable English words, the strategy that people will use is not to mention the language name at all (“as you can see in example (12) from this Athabaskan language (Dënesųłiné)”, rather than “as you can see in the Chipewyan example”). And maybe the next step is that they will replace the example by an example from another, more pronounceable language. Choosing an unsayable language name is a public relations disaster, as with city names.
(On the other hand, I accept that some scholars see it differently – they may not care so much about the visibility and recognizability of the smaller languages, but may put more of an emphasis on the association of traditional English names with earlier bad behaviour of English speakers toward the speakers of the smaller North American languages, as seen strikingly in this blog post by anthropologist Sarah Shulist, who ends up denying the principle that language names are English words – but of course not consistently, so she has to treat Mohawk/Kanienkéha’ speakers differently from Greek/Elinika speakers.)
This blog post is primarily addressed to my fellow linguists, and I can’t blame them of course if the speaker community (or at least an influential part of it) insists on a difficult name. (The Indian city formerly known as “Trivandrum” has surely not made life easy for its board of tourism when it decided to be called Thiruvananthapuram, and has apparently put parochial pride before international public relations.) However, there are also many cases where the speaker communities have no particular opinion on how to call their language in English. In quite a few cases they don’t have a traditional name of their own, but call their language simply ‘our language’, or ‘the people’s language’. In such cases, where it’s the linguists that choose which name to use in English (or Spanish, ot Portuguese, or Chinese, etc.) shouldn’t they use a name that is easy to pronounce, to read and to write? Overexoticized language names, and especially unmotivated replacement of established names, achieve the opposite of what the linguists want to achieve: They make it difficult for people to recognize and talk about the language, so they make the language less visible.
Thus, I would hope, for example, that linguists working on Khoisan languages will accept that they cannot expect their colleagues (let alone nonlinguists) to correctly pronounce (or even write) language names such as Nǁng. At the very least, if they insist on writing the language names with these highly special characters (which at least Wikipedia has accepted, even though they are not real letters and do not exist in most fonts), they should tell us how to pronounce them in English, i.e. without clicks, not only how to pronounce them in the language itself.
Haspelmath, Martin. 2017. Some principles for language names. Language Documentation & Conservation 11: 81-93. [publisher]