Has anyone thought about how dialectal variance patterns geographically?
If we could find some general patterns, archaeology and historical linguistics may become mutually influential. For example, historical reconstructions show that there were several proto-dialect groups; archaeologists could then extrapulate population distribution from assumptions supported by geo-dialectal studies. Or, archaeological evidence uncovers ancient population distribution, from this linguists could estimate how many dialects were operating. Since language change is related to time, temporal estimations would also come into play. Breakthroughs in one field could spark breakthroughs in the other.
Some interesting questions:
What landscape features influence linguistic change?
At what point geographically do dialects begin to become unintelligible?
How does language variation correlate to geography? Is distance a primary factor? If not, what could be???
Any thoughts?
Of course this would take years and would need to be tested against other data.
But without written records, there is no way to know what languages were in use. (And even then, the popular language might be something else. Based on numismatic and epigriphal evidence, future archeologists may conclude that Americans spoke Latin!)
How on earth could they do that? And, why are you speaking of “dialects” in particular, rather than whole languages?
I understand that this is already done for fields like reconstructing PIE. They must have some rules of thumb.
Anything that would serve to group people together, and/or influence their intercourse with other groups. Mountains, rivers, oceans…the suitability of a region for agriculture, pastoral nomadism, or whatnot (which would take us into questions of climate and ecology)…
For example, some scholars think that China’s geography makes it more likely to be ruled as an empire, since its central plains lack natural barriers such as mountains which would have protected smaller polities (as happened in Europe).
Geographic distance is only one element of geography, and geography only one influence on language. (How far apart are you and I?)
Does there have to be one primary factor, which is true of most cases? Maybe linguistic drift is just natural, like genetic drift, and would occur regardless of geography.
Firstly, written records are not always needed to reconstruct past languages. Comparative analysis of cognates can go a long ways.
As for a reconstructed population distribution shedding light on proto-linguistic variance, I was thinking the following:
(1) Archaeologist begin to understand the geographical extent of a culture – people showing some material, ideological, linguistic connectedness.
(2) Through models based on trends in the patterning of dialectal variance over landscapes, linguists could roughly extrapolate the degree of variance.
When I said primary factor I meant that instead of citing geographical isolation as a reason for dialectal variation in cases where isolation may not have occurred, we could consider other possible reasons: ideological differences, economic differences, differences in contact with other cultures and languages, ext.
I think it would be interesting to look at the morphology of landscapes in areas where there is a high degree of linguistic variance. One could even try to see how variance patterns over different features. Conclusions could be compared to archaeological data to see if linguistic/dialectal similarity correlates to cultural similarity or trade. If there is a lot of variance in how linguistic variance patterns of landscapes, perhaps other conditions were more influential .
Of course determining whether two or more proto-languages were dialects or mutually unintelligible but related forms must be done on a case by case basis.
I just like thinking about how different disciplines can provide new insight or questions. All of this would take enormous amounts of data.
They are in some areas such as the history of the colonisation of the pacific. Linguistic evidence as well as the archaeological, anthropological and botanical, is used to make assumptions about the timing and direction of the spread of peoples. Apparently one generalisation is that language variance is more pronounced at a source of migration and less so at its extremities. This is used as support for the theory that the origin point of South Pacific migration may have been Taiwan.
Landscape is more important than distance. Most of the plains of Northern China share similar dialects, whereas in hillier areas of the south it can vary a lot form town to town. In Papua New Guinea there’s a vast number of differences in the mountainous highlands.
But without written records, there is no way to know what languages were in use. (And even then, the popular language might be something else. Based on numismatic and epigriphal evidence, future archeologists may conclude that Americans spoke Latin!)
[/quote]
Historical linguistics is a reconstructive science, like forensics, archeology and evolution: researchers piece together clues and build testable hypotheses based on the preponderance of evidence.
As such, techniques such as the comparative method and internal reconstruction are well-developed means of reconstructing ancient languages.
[quote]
How on earth could they do that? And, why are you speaking of “dialects” in particular, rather than whole languages? [/quote]
There are many clues at their disposal, including comparing modern dialects, looking at migration patterns, etc.
We know, for example, that Old Chinese split into a northern and southern dialect during the Han Dynasty, with Cantonese and Mandarin developing from the northern dialect (i.e. Middle Chinese) and Wu and Min developing from the southern dialect. This can be deduced even without written records, by examining historical migration patterns and comparing linguistic features of the dialects (voiced initial stops preserved in the southern dialects; the replacement of voiced initials with Yang tones and the development of the “f” sound in Mandarin and Cantonese), as well as borrowings into Japanese (Go-on and Kan-on readings of words).
It is my understanding that the voiced initial obstruents of Middle Chinese are only preserved in Wu dialects. In Mandarin, they became aspirated and unaspirated obstruents according to their Middle Chinese tone category; observe:
Similar splits of the MC tone categories (平 上 去 入) into Mandarin characterize the development of the other MC obstruents. None of the voiced initials were preserved, which holds true for the southern dialects – there was a large-scale process of devoicing.
MC tone categories are shown in various rhyme books, but we don’t know what the pitch variations were. (guǎngyùn; yùnjìng)
Here is a an interesting quote from my course notes:
“The kind of tone splits and mergers that you see in the development of tone categories from Middle Chinese to Modern Standard Mandarin is typical of most Chinese dialects. While the exact patterns differ from dialect to dialect, the tendency of Middle Chinese tones to split in two, and of resulting tones to merge, is common.”
If you want more information regarding regular sound correspondences between Mandarin obstruents and MC, let me know and I will post it.
It is my understanding that the voiced initial obstruents of Middle Chinese are only preserved in Wu dialects. In Mandarin, they became aspirated and unaspirated obstruents according to their Middle Chinese tone category; observe:
[/quote]
Taiwanese certainly has voiced initial b- and g- (and l- where d- would be expected). But they seem to have developed from older m-, ng- and n-.
Taiwanese certainly has voiced initial b- and g- (and l- where d- would be expected). But they seem to have developed from older m-, ng- and n-.[/quote]
The example I was thinking of was PIE. Yeah, some of it can be reconstructed (though we’ll never know if we got it right!), but it’s hard to match this to anything on the ground. The closest it comes is telltale vocabulary for certain plants and animals. The geographical distribution of IE speaking peoples is obviously crucial, but doesn’t solve the problem of how they got to where they are, or whether their ancestors might have switched languages at some point. Older generations supposed the “Aryans” to have been light-skinned, but this doesn’t perfectly match the IE speaking groups in India.
In the Chinese case (for which written records exist), there must have been other languages and dialects which are no longer extant, and about which we can know nothing–thereby frustrating the task of historical reconstruction (of the linguistic situation “on the ground,” not of the surviving Chinese languages).
[quote=“Screaming Jesus”]
In the Chinese case (for which written records exist), there must have been other languages and dialects which are no longer extant, and about which we can know nothing–thereby frustrating the task of historical reconstruction (of the linguistic situation “on the ground,” not of the surviving Chinese languages).[/quote]
Good point. But, analysis of the Chinese languages, many of which have not been studied, can shed light onto this matter, especially because we know something about MC.
Also,…
There is variation within the roots of reconstructed languages; example: free variation in the actual language that reconstructions attempt to describe. Consider either and neither: variation between the pronunciation /ij/ and /aj/ for the first vowel is common in America. Free variation between the two forms is widespread – myself included. This may complicate future reconstructions of contemporary English.
