Nicholas Ostler is one of LH’s favorite authors of language books; I enthused about his Empires of the Word: A Language History of the World in this post and about his Ad Infinitum in this one. Now Walker, the publisher, has sent me his latest, The Last Lingua Franca: English Until the Return of Babel, and I’m enjoying it just as much. If you’re still looking for a Christmas gift for a language lover (who might be yourself), add it to your list.

I should start by pointing out that the Big Idea of the book—the question of whether English will continue to dominate the world as it does now and if not, what will replace it—is not of great interest to me. I mean, I’m curious, of course, just as I am about how the world will end, but we’ll find out about each when the time arrives, and in the meantime, speculation is basically navel-gazing. It’s great for late-night bull sessions, but is not to be taken seriously, any more than the many “Iran [or other Currently Prominent Phenomenon]: Threat or Menace?” op-eds that fill the broadsheets. Fortunately, Ostler is incapable of writing an extended op-ed; his interest, like mine, is in the details… or, as Deborah Cameron puts it from the other side of the fence in her review in The Guardian: “The Last Lingua Franca is not the easiest of reads: Ostler does not have the populariser’s gift for uncluttered storytelling, and is apt to pile up details without much regard for what the non-specialist either needs to know or is capable of retaining.” Different strokes, as they say. So let’s get to the details.

Ostler starts by distinguishing the use of languages, and English in particular, as native tongues from their use as lingua francas (or, as he irritatingly insists on spelling it, “lingua-francas”—punctuation is not his strong suit). This is an important distinction that is not often taken as seriously as Ostler takes it, and one of the things he does throughout the book is hammer home the implications in various situations. In the Preface he writes, “But when English is seen as a lingua franca [N.b.: he’s also not consistent], it becomes possible to compare it with many languages that had this function to a greater or lesser extent in the past. What forces spread them, and how did they fare in the long term?” This (to me) is the meat of the book and the most interesting part, especially the chapters on Persian, though of course others will skim it in hopes of getting an answer about the future of English. He then goes on to look at “language politics in the modern world” and try to figure out where things are going from here.

Since what delights me about Ostler’s books are the piquant and unpredictable details, let me cite a few of them here. On page 11, in discussing the language policies of Malaysia, he quotes a Malay sentence: “Bahasa jiwa bangsa, as they liked to claim: ‘Language is the soul of the nation’.” It’s always nice to get a sentence in the local language, but the great thing is the footnote: “This is a simple sentence in Malay, but all its words are derived from Sanskrit: bhāṣā jīvanam vaṃśaḥ ‘speech-life-stock’. This reflects the influence of Indian traders and adventurers in the early first millennium AD. Malay’s roots are ancient, but they are not purely indigenous.” He can’t resist that sort of sidelight, and neither can I.
On page 40, a footnote to a discussion of the divergence of Nahuatl dialects into different languages after Spanish was made the only official language in 1770 gives some striking examples: “‘He is not at home’ is expressed as x-aak in Guerrero, amo-hka in Tezcoco, am-iga in Morelos, amo yetok in Puebla Sierra, and mach nikaan kah in Vera Cruz. In Classical Nahuatl, this would have been expressed as amo i-chaan-ko (kah), literally ‘not his-home-at (is)’.” (However, he doesn’t actually call them different languages, saying rather “at least nineteen quite different forms, according to systematic tests of mutual intelligibility” and adding, “Some doubt whether [such tests] should be followed in total oblivion of Nahuatl’s common cultural past”; he takes the same, to me reprehensibly unscientific, squishy approach to the question of whether Chinese and Arabic are one language or many. Politicians have to let politics bend their view of reality, but it’s a sad business when non-politicians do the same thing.)

On page 49 he casually mentions Russenorsk, “used in the eighteenth and nineteenth centuries among Russian and Norwegian traders on the arctic shores of Siberia”; I had never heard of it, and it’s so obscure there’s only one sentence on it in the Encyclopedia of the Languages of Europe, but it turns out somebody’s written a nice little Wikipedia article on it, with bibliography. And on page 98 he tosses in a reference to the Tajik family name Shansabani, “which seems to be derived (consider the consonants!) from the Pahlavi name Wišnasb“: hey, cool!

But even better are the careful explanations of difficult or unexpected situations. On page 86, for example, he quotes Ibn al-Muqaffa’ as distinguishing “five languages (besides Arabic) in Iran,” of which three are varieties of Persian: Pahlavi is said to be “the language of Isfahan, Hamadan, Nehavand, and Azerbaijan,” Dari “of the court … but also of the Sassanian capital city (Ctesiphon, on the Tigris), ‘though in Khorasan … the best Dari is heard in Balkh,” and Parsi “the language of Fars … but also of the mowhed Zoroastrian priests.” He says this tripartite division “is somewhat confusing since the accepted analysis of historical Persian only distinguished two (similar) main dialects over the preceding millennium: viz, Median (later, Parthian) in the northeast, and Old Persian (later, Pārsīg or Middle Persian) in the southwest. … But Gilbert Lazard (1993) has quite convincingly [explained that Parsi] is understood as traditional Persian…, Pahlavī had originally meant ‘Parthian’, and this may be what it means here, since the cities are about right for ancient Media…. Meanwhile, Darī refers in effect to New Persian: despite its title, this is the new, predominantly colloquial, version of the language, coming through in the capital but also out to the east.”

And his explanation of how Elamite was used by Persians after they had conquered Elam (now Khuzistan) was a real eye-opener for me. I had known they used it officially for proclamations and such, but I was thinking of it as comparable to the Romans using Greek, or everybody in nineteenth-century Europe using French—languages they themselves knew. Not so: “Persians indeed early made a virtue of illiteracy…. When a Persian needed to send written communication or to store records, it was done for him in Elamite, this being the beginning of alloglottography (writing down Persian by the use of some other language), which was to continue in Persia … for the next millennium… This did not mean that Elamite was widely learned, or even heard, by Persians. Rather Elamite-Persian bilinguals would be employed to use their writing skills, protecting their employers from the disconcerting experience of an unknown language by routinely translating the texts on request into, or out of, spoken Persian.”

Well, I could go on, but you probably have a sense by now of whether you’d like the book. If you want to know the future of English, you’ll have to read it—no spoilers here! Me, I think it’s a blast, and I look forward to Ostler’s next excursion into the multifarious history of the world’s languages.


  1. The same alloglottography was used by Persians later with Aramaic, as indeed Ostler explained in his previous book.

  2. Yes, and of course he explains it here as well.

  3. aqilluqqaaq says

    There’s something decidedly amiss with the notion of alloglottography between Old Persian and Elamite as conceived in Gershevitch’ original article (though maybe Ostler avoids this). You dictate a letter in Old Persian; you document it in Elamite; you re-translate into Old Persian (or whatever) as you read it. Alloglottography (in this case) is simply writing in another language – there’s no sense in which the text is Old Persian ‘ideographically’ documented in Elamite (in contrast with the case of Middle Iranian and Aramaic?), which is what Gershevitch wants to say (why, then, the determinatives? why do we find the Elamite readings attested elsewhere? and why no OP grammatical additions as apparently are found in Aramaicizing MI?).

  4. Yes, Ostler also talks about “Old Persian written in Elamite,” which I found odd.

  5. Trond Engen says

    I haven’t read any of the Norwegian literature on it, so I obviously should. Anyway, here‘s Fredrik Kortlandt’s take on it, linked to from the Norwegian Wikipedia article. (I’d seen it before, but a while ago.)
    I think it’s clear that it had no real grammar or stable lexicon, rather established routines for making oneself understood by someone speaking a different language.
    Both the Wikipedia article and Kortlandt mentions other origins of some words. Dutch (or Low German) and English are obvious, with examples like krank, sprek, slipom, verrigod.
    I think similar pidgins have been spoken elsewhere in Europe without much notice. E.g., the myth says that the fishermen of the Danish west coast and the people in the harbours of Scotland and Northern England understood eachother, each speaking his own language.

  6. Trond, I was told when young that almost all English (and presumably Scottish) maritime vocabulary comes from Scandinavia, so that yarn isn’t impossible. I once received a postcard from Sweden, in Swedish, and read it easily because the short message so resembled Scots.

  7. I had trouble with alloglottography, but if you look around you’ll find this paper gives an interesting explanation:
    Alloglottography and scribal antiquarianism in the Ancient Near East
    “The term alloglotography was coined by Ilya Gershevitch in his study of how the inscription of Darius at Bisitun was written. The Elamite version was most likely the first to be engraved, then the Babylonian one, and finally the Old Persian. If the Persian king used Old Persian as his language, one may wonder why Elamite figures so prominently on the rock.
    “According to Gershevitch, the Elamite version is the true original and represents the actual words of Darius, whereas the Old Persian on the inscription is a retranslation or back translation. This means that the Great king uttered the words in Old Persian, but the scribes wrote them down in Elamite and read them back to him (as the inscription says) in Old Persian.”
    Early Japanese was apparently also a similar situation: Early Japanese and Early Akkadian Writing Systems
    —A Contrastive Survey of “Kunogenesis”—

  8. That last spam is a keeper!

  9. Trond Engen says

    Is Wedding Gowns related to Bathrobe?

  10. aqilluqqaaq says

    I think the phenomenon of alloglottography is clear enough, it’s just not what we have here. You get it between Babylonian and Sumerian. The cuneiform reads: da-ri-ia-a-muš LUGAL GAL-ú (where lowercase is Babylonian and uppercase, Sumerian), but which is read: dariamuš šarru rabû, ‘Darius, the great king’. This is alloglottography: it’s (partly) written in Sumerian, but read as Babylonian. How do we know? Because the Sumerian is inflected as Babylonian (-ú affixed to GAL).
    But this is not what happens with Old Persian and Elamite. The same text in Old Persian is da-a-ra-ya-va-u-ša xa-ša-a-ya-θa-i-ya va-za-ra-ka, and in the Elamite text is da-ri-ya-ma-u-iš sunki ir-ša-ir-ra. This is writing in another language: it’s written in Elamite, and read as Elamite. How do we know? Because it is written without any Old Persian inflection to indicate, e.g., that the logogram – sunki – is read xšâyaθiya, let alone that phonetic ir-ša-ir-ra is read vazraka.

  11. That last spam is a keeper!
    Damn, I deleted it without even looking at it. Ah well, I preserved a magnificent specimen in the Much? thread.
    it’s just not what we have here.
    True, and your explanation is crystal-clear. It’s unfortunate that Ostler messes up details like that, but it’s inevitable when someone is trying to cover so much ground. (I used to be much more intolerant of error until I tried writing my own book.)

  12. “the divergence of Nahuatl dialects into different languages after Spanish was made the only official language in 1770”
    I don’t know anything about this, but I wonder whether the array of dialects pre-dated 1770–and “classical” Nahuatl simply lost its status as a prestige dialect.

  13. Bruno van Wayenburg says

    @dearieme I always thought a lot of English maritime vocabulary is Dutch (from skipper to yaught). My Frisian (minority language in the Netherlands) grandfather always claimed that they could understand him just fine in Denmark. Now he was the sort person that would make himself clear forcefully, without taking much note of problems at the receiving end. May he rest in peace.
    But seriously, I suspect that the ‘my local dialect is remarkably like a faraway cousin language’-feeling could easily be an illusion. When I’m reading Swedish or Norwegian, which is really not so hard to do after a while, many words can be understood from their Dutch cognates, or otherwise from German or English, which has a fair share of Old Norse roots anyway. All of this is to be expected, considering the germanic family tree.
    But occasionally, a word without clear English, German or Dutch cognate pops up, which does have one in Frisian (and even one I remember, since my Frisian is not that great). Now that will feel as something remarkable, considering the relative obscurity of the language and a slight tinge of backwardness about it. And after a few of these occurrences, I would almost be convinced that rural Frisian is inexplicably closer to Swedish than English, German or Dutch, almost mystically even (nation of hardy sea travellers, bone headed cold sufferers). You’d have to use some advanced statistics to prove this, of course, so instead, I’m offering this hunch as a LL comment.

  14. Two things: clicking on your Rubio abstract link got this: The requested URL /mt/ was not found on this server.
    And in all seriousness, shouldn’t it be linguas franca, and likewise with similar constructions? Is there a rule? There’s “passers by” but “bank robbers”…

  15. All seriousness aside, the plural is “linguini-franks”.

  16. I’ve been an enthusiast for Ostler’s books too. He does explain things that are not generally made very clear in specialist literature, e.g. why the Persians would have used Aramaic so extensively, or to my mind the following: I always thought it striking that the Greeks would have preferred different dialects for different genres. Why exactly? I’m not a classicist, but Ostler makes the simple comparison with British pop singers use of American (southern) dialect. Artists in their imitations don’t just borrow the words or meters of the work of art, but also linguistic features.
    Also, I must say in the areas where I know something about the languages, Ostler makes hardly any mistakes, and given the ground he covers in his books, that’s impressive.

  17. With beans or cheese?
    The links you people post here don’t work when clicked on, at least not in Firefox. E.g., gets me a 404 Not Found.
    But shortening it to works fine.
    Is this my problem or yours?
    Another question: how many of you actually read and speak the languages you write about around here? I can see Nahuatl, that’s being brought back like Hebrew and Gaelic, but Old Persian and Elamite would surprise me outside very highly cultivated academic environments.
    Not that I compare mind you, my English obviously has problems and the only other thing I can do is read & write just enough Spanish to keep from getting terminally lost or starving to death.

  18. On topic:
    I’ve always thought Aramaic spread so widely because it’s alphabetic and because the original speakers were very involved in commerce. A bigger empire meant more options and often safer routes, and since you can’t run an empire without commerce its use in government was obvious. My hunch is it had nothing to do with how easy the language was to master or any features particular to Aramaic itself, except that you only had to memorize a couple dozen letters instead of hundreds of cuneiforms or thousands of Chinese characters. The Canaanites just got lucky.

  19. And in all seriousness, shouldn’t it be linguas franca, and likewise with similar constructions?
    No. In some other language, maybe (though it would be linguas francas), but in English it’s simply a compound word like bank robber or language maven. One plural marker per customer, and it goes at the end. (Attorneys general me no attorneys general; that’s a peculiar construction that hardly anyone uses anyway.)

  20. Attorneys general, courts martial and the like are remnants of Law French. Law French is “the corrupted form of the Norman French language that arose in England in the centuries after William the Conqueror invaded England in 1066 and that was used for several centuries as the primary language of the English legal system.” — Black’s Law Dictionary, Seventh Edition, 1999.

  21. Yes, there are these “noun adjective” phrases of French origin, whose pluralization leads to awkwardness in English.
    I want to know about “Postmaster General”. A quick Google search suggests that the term dates back to 17th century England, so maybe modeled on some older names of positions.
    And what about “major general”? Is that noun adjective, or adjective noun, or what? In practice nowadays it is more or less adjective noun, but what’s the origin?

  22. The way I heard it “major general” is a shortening of “sergeant-major general”, which is why she is lower ranking than a lieutenant general. In Europe (Germany, Scandinavia, Russia anyway) they have Colonel Generals. But obviously “general” is the noun part, unlike General Motors or Gen. Foods, or attorney g., or in England also “Solicitor General”. In Britain, they abolished the postmaster general’s job (I can’t write his “post”) — it was a political appointment, he was a government minister — possibly at the same time they abolished the GPO, or General Post Office, somewhere around 1970.
    On a completely different subject (my mind wandered to Trollope) on television yesterday, the solicitor for Julian Assange described his living conditions in Wandsworth Prison first as “Dickensian”, then “Orwellian” before finally settling for “Victorian”.

  23. “before finally settling for “Victorian”.
    Well, Shagger Assange is an Aussie, so “Victorian” should be fine.

  24. And they probably don’t treat prisoners from Queensland very well in Victoria.

Speak Your Mind