L-complex
« previous post | next post »
From Peter Daniels:
Do the 7 or 8 (or whatever) “dialects” of Sinitic constitute what Hockett called an “L-complex,” like Romance, such that you could traverse the entire domain and never encounter neighboring villages that didn’t understand each other, with cultural centers where the language described in the regional grammar book and dictionary is spoken, or are they distinct languages as far back as one can look?
I enquire because in her new edition, Amalia Gnanadesikan inserts a description of the Rhenish Fan into her Chinese chapter as if using Dutch vs. German as a standard language / orthography is relevant to using Mandarin as the written language throughout China.
An additional thought is that the topography might be different enough that there are enough land barriers to interfere with regular communication? Like New Guinea.
"L-complex" is a synonym for a dialect continuum. It also refers to the study of how intricate a language's structure is.
Selected readings
- "Mutual intelligibility" (5/28/14)
- "Mutual Intelligibility of Sinitic Languages" (3/6/09)
- "Intelligibility and the language / dialect problem" (10/11/14)
- "Language, topolect, dialect, idiolect" (10/3/23)
- "English is a Dialect of Germanic; or, The Traitors to Our Common Heritage" (9/4/13)
JPL said,
October 20, 2025 @ 1:21 am
It's all about speech communities. What is the system (of categories and schemata, for both form and content) that enable equivalence of one person's propositions and another person's propositions? These systems vary continuously over geographical space, some variations more drastic than others. Another interesting case is what they call "linguistic areas" in the language contact field. There will be historical reasons for the differences between the relatively continuous and abrupt cases. The appropriateness of naming the systems will differ wrt these empirical differences. Does this make sense? (I think, if I'm not off the mark, that I've only expressed the problem differently from the usual way.)
John S. Rohsenow said,
October 20, 2025 @ 1:46 am
Is Peter Daniels referring to Amalia Gnanadesikan's "The Writing Revolution: Cuneiform to the Internet," an accessible introduction to the history of writing and the structure of writing systems?
wgj said,
October 20, 2025 @ 2:39 am
I'm certain that "never encounter neighboring villages that didn’t understand each other" is an utter pipedream. In my travels through various parts of China, I'd been told countless times that "when we talk in our native tongue, even people from neighboring villages couldn't understand us". Some of that may have been exaggerated, but it's improbable that all of it is untrue.
An extreme case of this is Nantong, Jiangsu, infamous for its strange topolect – to the point that the Chinese Red Army employed Nantong people as radio operators, similar to the Navajos in the movie Windtalkers. According to friends of mine from there, even within Nantong there are several mutually unintelligible topolects, with the coastal folk unable to understand the city folk and vice versa. It has been strongly argued by many scholars that Nantong has a highly complex population history with multiple waves of non-Nan people migrating in. And these people presumably settled different parts of Nantong, mixing very little among each other subsequently.
Jerry Packard said,
October 20, 2025 @ 7:28 am
Daniels’ question is interesting and makes many assumptions. To answer, first, in my opinion the ‘either-or’ nature of the posed opposites is overly extreme. Second, Chinese is well known for the fact that adjacent villages often don’t understand each other. Finally the question about distinct languages as far back as one can go is much harder to answer, because it is indeed difficult to determine the extent to which a. the 7-8 dialects were always different, or b. whether they started out as the (more or less) ‘same’ language and came to differ due to 1. divergences over time and space and 2. contact with languages that formed underlying strata.
Jonathan Smith said,
October 20, 2025 @ 7:50 am
Even assuming development from a proto-language, time+space are such that Sinitic couldn't constitute anything remotely like a single cline. Then add millennia of population movement and mutual influence (esp. outward from influential regional standards) due to close geographical+cultural contact.
One basic take is Norman 1988 e.g. 189-190: "it is possible […] to speak of strong and weak dialect boundaries. As an example of a strong boundary, we can cite the demarcation which separates Wu and Min […]. The boundary between Min and Kejia […] cannot be drawn so sharply [as] [i]n the far western and southwestern parts of Fujian province, there are dialects which are clearly transitional between the two groups […]. An example of a very weak dialect boundary is that dividing the Xiang dialects from Mandarin" etc. etc.
Some Wu languages share much with Min in the lexicon and beyond… while the Min/Kejia (=Hakka) boundary is arguably not exactly blurry (Branner has studied some aspects of this question closely in e.g. The Classification of Miin and Hakka…)
Chester Draws said,
October 20, 2025 @ 2:32 pm
The smoothness of the Romance "gradient" hasn't always been true.
When various parts were inhabited by Germanic invaders, for example, on the fall of the Roman Empire.
Or when the Vikings took over parts of Normandy.
Chris Button said,
October 20, 2025 @ 4:31 pm
Is the question about the spoken word (e.g., an Alsatian vs Lorraine Franconian overlap) or about the written word (e.g., a written Spanish vs written Portuguese overlap)?
Jarek Weckwerth said,
October 21, 2025 @ 4:33 am
In a continuum situation, mutual intelligibility is never 100% and never 0%. So the cutoff point will always be arbitrary.
wgj said,
October 21, 2025 @ 7:51 am
@Jarek: Arbitrary from an academic point of view, but in practical terms, the cutoff criteria is whenever the local populations themselves decide they don't understand each other – and they tend to have a consensus among themselves on that.
David Marjanović said,
October 22, 2025 @ 7:47 am
There was no ethnic cleansing involved there, however. Instead, you basically had an extended nobility that spoke Gothic or Frankish or Lombardic (for up to 200 years in some cases maybe; not in Normandy, where Norse died out quickly), while the peasants and the city folk continued to occupy their places in the Romance dialect continuum.
Today it's no longer a dialect continuum because, especially in France, the school system put a bizarre amount of effort into getting everyone to speak only the tightly standardized national language all the time.
Spoken – but what do you mean by "overlap"?
Philip Taylor said,
October 22, 2025 @ 9:26 am
"Today it's no longer a dialect continuum because, especially in France, the school system put a bizarre amount of effort into getting everyone to speak only the tightly standardized national language all the time" — I would substitute "commendable" for "bizarre", David, and drop the "only".
Jerry Packard said,
October 24, 2025 @ 6:40 am
For Chinese dialect mutual intelligibility the best source and well worth reading are the works by Tang and van Heuven.
David Marjanović said,
October 24, 2025 @ 10:31 am
But "only" is true. The entire Occitan language is seriously endangered, outside at least of one valley in Spain and one valley in Italy; and the rest of the diversity isn't doing much better in France, even inside the home.
So, no, not commendable. The policy seems to have been based on the manifestly false belief, still widespread in the 20th century, that there's only enough space for one language in one head.
Philip Taylor said,
October 24, 2025 @ 4:58 pm
Well, if it really is "only the tightly standardized national language all the time", then I willingly withdraw my comments. But perhaps "misguided" rather than "bizarre". Certainly I would commend them for teaching everyone to speak good standard French, but not at the expense of minority languages such as Occitan.