John McWhorter has an interesting discussion of why some languages lose their inflections and begin depending on word order and function words, from a global perspective an unusual development:

I ask these questions [why Native American, Australian, and Bantu languages have kept their "bristling paradigms of prefixes and suffixes"] because my research increasingly suggests to me that for a language to shed its inflections, rather than consistently replace or even retain them, is less business as usual than the unexpected case. From a global perspective, languages appear to usually do this as the result of widespread acquisition by adults, whose ossified language organs tend to clear away languages’ “junk.”
Thus the inflection-shy nature of Romance compared to Latin—and Romanian’s remnants of case-marking are dishwater compared to Polish, Greek or Lithuanian—was due to imperfect renditions of that language being passed on in the context of invasion and imposition from the outside. English is the only Indo-European language in Europe with no gender marking on articles or nouns—ever notice that?—because of Vikings’ approximation of Old English starting in the eighth century. It is presumably no accident that Persian, with its low inflection and gender-neutral third person pronoun, has been lingua franca par excellence throughout much of its history.

I’m pretty sure I’ve seen this argument before, but McWhorter puts it well, and it’s one of the few supposed historical explanations for language change that makes sense to me.


  1. I’ve just been thinking of this, or maybe cursing why Japanese is so inflected. They didn’t conquer of get conquered much compared to those people who live in the flood plains of Eastern China. It really has been so much harder to pick up Japanese just by listening to and trying to speak to people and hoping the frequency effects will kick in. On the other hand, in Mandarin you hear the same words again and again very often and so it just makes it so much easier to pick up.

  2. I believe that the non-inflected nature of Chinese may be for this same reason. One mystery of Chinese to me is that it is most differentiated in the South, where it is newest. As I understand, anyway, the area around Xi’an has been predominantly Chinese for well over three thousand years (maybe four), whereas much of China South of the Yangtze was still being Sinified ca. 300 A.D.
    Interference from other languages and geographical dissection could explain this. It still seems odd that there was almost no differentiation in the north though (all northern dialects are classified as Mandarin, and are really all dialects whereas the mutually-unintelligible southern “dialects” probably need to be called separate languages.)
    In the North, perhaps extensive churning caused by foreign conquest and population displacement prevented differentiation.
    At a conference last year a student of the isolate language Nivkh said that that language is becoming more complex rather than less. I asked him whether this might be the reverse process — a language never learned by adults, but only by children from parents. He thought it was an interesting idea. Wish I could remember his name (he’s at U. Washington I think).

  3. OK, I’ve been meaning to post on this for a while but since you bring it up I’ll ask it in the comments. This is something I’ve never understood. Chinese is, as you say, first attested in the north, in the Yellow River valley. But it’s a member of the Sino-Tibetan family, which means it must have come from the south; in fact, Dalby says:
    “Meanwhile [as speakers of early Tibeto-Burman dialects moved west from SE China] speakers of early Chinese dialects will have spread generally north-eastwards from their evident nucleus near the south-eastern coast.”
    Of course, it could simply have been written down first in the north while being spoken much farther south, but everyone seems to present it as you do, with the language spreading south (and with the consequent puzzlement over dialect differentiation). Wouldn’t it make sense to explain the very different spoken forms in the south by saying they’d been there since Proto-Chinese days?

  4. I’ve never thought of it that way. However, Tibetan extends pretty far north (Qinghai is mostly Tibetan IIRC) and I believe that Burmese came south from the north.
    I do always see it assumed that — whatever the ultimate origin of Chinese — the peoples of the south were originally speakers of Thai, Khmer, Vietnamese, or related languages, and that the period ?200 BC –> ??600 A.D. was a period of Sinification. This wouldn’t preclude Chinese having first come north from the south and then being replaced by the other languages in the south.
    I confess that as a sort of classicist, I’m really most aware of the spread of Chinese writing. I’ve seen others make this error — Hittite is the first IE language attested in writing, and you often see it described as “the oldest” IE language, as if the accident of transcription gave it a special place in the branching tree.
    As a general practitioner, though, I’d advise you to call in a specialist before I kill the patient. I’ve been guessing a bit.

  5. Tibet could be considered central in terms of latitude and you can still find Burmese speakers in Sichuan province. Their movement South was in historical times as was the Thai and Chinese.

  6. But they certainly didn’t come from as far north as the Yellow River. All indications are that Sino-Tibetan originated well south of where Chinese is first attested. Surely someone knows a Sino-Tibetanist they can pester?

  7. Does your morphology lose its flavour (“flavor”) on the bed-post overnight?
    If your mother says don’t chew it, do you swallow it in spite?
    Can you catch it on your tonsils, can you heave it left and right?
    Does your morphology lose its flavour on the bedpost overnight?

    This, and the discovery of the bedpost-holes of the Battle-Axe (“Battle-Ax”) people, is surely an adequate explanation? Anyway, don’t typologists (and Dixon) claim that these typological changes are cyclical?

  8. xiaolongnu says:

    Hi all — I’ll chime in but I’m going to have to tread carefully around my lack of historical-linguistic expertise. Zizka is right to distinguish between the spread of written Chinese and the spread of the Sino-Tibetan language group.
    For the sake of argument, let’s assume Dalby’s assertion about the spread of S-T from a nucleus in southeast China is correct. LH, you say that “all indications” are that the origin point was south and east of the Yellow River basin. Can you perhaps let us know where exactly this is? (In the conventional geography of China, places like Shanghai and Hangzhou — i.e. anything south of the Yangtze river — are considered “south,” so “southeast” could be as far north as the Yangtze delta). Also, what kind of information does Dalby cite (that is, what are “all indications”)?
    At any rate, I have to assume Dalby is using historical-linguistic reconstruction (reverse sound-change) to find this original core area. I’m a little dubious about any historical-linguistic work done on Chinese because we have so little access to historical pronunciation to begin with. Since Chinese characters aren’t purely phonetic, the historical pronunciations some people supply for them are already kind of conjectural. We can work a few things out (for example, treatises on the principles of poetics are what clued us in to the loss of the rusheng tone after the Tang), but there’s more than the usual amount of guesswork involved.
    That said, it’s not inconceivable that the language group originated in the southeast some time before the earliest appearance of proto-written Chinese (incised pictographs found on pottery vessels from the Dawenkou culture, c. 2800 BC). Southerners could well have migrated north especially since at that time, the climate was warmer and wetter (elephants and rhinoceros were known to range as far north as the Ordos) and so there wouldn’t have been the drastic difference in climate and rainfall that now exists between north and south China. Interestingly, Dawenkou pottery vessels have been found as far south as Zhejiang and Anhui provinces, which starts to get you into territory that could be called “southeast.” Typically, Chinese archaeologists interpret this as evidence for a southward migration, but who’s to say that it wasn’t a northward migration?
    The main argument for the ethnic Chinese originating in the north (besides the greater cultural prestige of the north) is that the Chinese are supposed to be descended from the Homo erectus known as Peking Man, whose remains were found at Zhoukoudian outside Beijing in the twenties. Generally the idea that PM’s descendants are the modern Chinese is not accepted in the West because of (a) lack of evidence of occupation during the intervening ice age and (b) the African Eve (mtDNA) hypothesis.
    At any rate, it is fair to say that written Chinese and institutionalized Chinese culture as we know it from the historical record did get started in the Yellow River Valley and spread outward, even though discoveries like the Sanxingdui bronze hoard suggest that Chinese culture was never as centralized as all that. But the truth is that even at the point when mature, grammatical Chinese is being written on oracle bones, we really don’t know what was being spoken in other parts of what is now China. The Chinese historical record has a notorious tendency to “barbarize” other peoples, but hey, maybe they were speaking Sino-Tibetan languages!
    The range of Chinese political power into south China is pretty quick though not that stable: in the Han dynasty (206 BC-221 AD) there was a Chinese governor in Hanoi. But there was also a Chinese governor in Bukhara, so I wouldn’t make too much out of that linguistically. From that point on there’s always a nominal Chinese presence in the south, though the political boundaries are often shifting. It sounds to me like the spread of the language and the spread of the writing system ought to be treated as two entirely separate categories.
    Now: as to why Chinese is so uninflected, I haven’t got the damnedest idea. Is it possible that it spread so widely because it was not inflected, rather than vice versa?

  9. Can you perhaps let us know where exactly this is?
    Hell, no! That’s quite specific enough for me, maybe too specific. The main point is that it didn’t originate in the north, where the historical record places it. It seemed to me the indications were pretty obvious: if you remove the Chinese element, every other member of the family is far to the south. Sure, Burmese moved south in historical times — but there’s no reason to think it came a couple thousand miles south. It looks to me like the family came from somewhere between Yunnan and the Yangtze, as for the Austro-Tai family; in fact, Dalby says “Chinese has changed drastically from its proto-Sino-Tibetan ancestor — and the changes may partly have been the result of interaction with early forms of Miao, Yao, Tai, and other languages now believed to belong to the Austro-Tai family.”
    I freely admit that I don’t really know what I’m talking about, and also that most specialists seem to think Chinese originated in the Yellow River area. I just think that’s a perverse interpretation of the linguistic data: that a dialectally uniform area on the fringe of the family should be taken as the point of origin.
    I suppose I should take a look at this book.

  10. Amazon text search got me a couple of relevant pages (384, 385) from Archaeology and Language II: Correlating Archaeological and Linguistic Hypotheses (One World Archaeology, 29), by R. Blench; it suggests that “The domestication of O. japonica [map on p. 385] may thus have occurred in an area which overlapped with the proposed Sino-Tibetan homeland.”

  11. xiaolongnu says:

    Hi guys — I am going to a conference in Philly this weekend where someone is giving a paper called “Origins of Sino-Tibetan and Prehistoric Linguistic Exchanges in Eurasia.” The speaker is a Chinese scholar called Xu Wenkan, so I expect it’s pretty current Chinese scholarship on the subject. If there is sufficient interest I will report on this when I return, probably Monday or Tuesday. Or should this maybe be a new thread? LH, your call.

  12. Oh goody! Please do report, and if you have a useful link on the subject let me know and I’ll start a new thread — otherwise we can continue it here.

  13. Looking forward to the new thread.
    As for the “northwestern origins of the Chinese”, various things I’ve read suggest that the **political** units of the early Chinese (Shang and Chou) originated in the northwest. The Shang seem very possibly to have been heavily influenced by chariot-riding invaders from the west, and the Chou seem to have been a NW border people, possibly of hybrid culture, allied to many non-Chinese peoples (esp. Qiang).
    This has nothing to do with the spread of Sino-Tibetan either. Possibly **all** the northern “non-Chinese” peoples spoke some sort of Sino-Tibetan or even Sinitic language. (Boodberg argued for a Sinitic-Iranian-Turkic progression on the steppe.)
    Some of the “shrike-tongued barbarians” of the South may have spoken Sino-Tibetan languages too, but I think that the presence of large numbers of Vietnamese- and Thai-related peoples there is well-attested.

  14. It’s annoying that so many scholars unquestioningly accept the Han Chinese viewpoint on things: “we civilized, they barbarian.” It’s like Classics a century ago. Few seem able to stand back from the Wonder of Civilization (oracle bones!) and consider such things as language families and who all those folks down south might have been.

  15. Try these links:
    Homeland for Sino-Tibetan placed somewhere on the Himalayan plateau.
    Homeland placed somewhere in eastern Tibet (geographical considerations) or ‘southeastern Tibet, northern Burma, western Yunnan, and the frontiers of Arunachal Pradesh’ (linguistic geography).
    Says that Sino-Tibetan originated in the loess plateau of NW China, then split by migrations to the SE (Chinese) and W and SW (Tibeto-Burmans).
    Urheimat for Tibeto-Burman placed in Sichuan-Yunnan-Tibet.

  16. Sorry,
    is talking about Tibet-Burman, not Sino-Tibetan.

Speak Your Mind