Page 1 of 1

New study to distinguish between languages and dialects

Posted: Sat Jan 12, 2019 9:08 pm
by yong321
https://aeon.co/ideas/why-languages-and ... nt-animals

Danish linguist Søren Wichmann evaluated different approaches to distinguishing languages from dialects, and proposed a new method based on Levenshtein distance between words in two speech varieties. Here's the last part of his article (the conclusions are in the parentheses):

"Some pairs of speech varieties that are considered national languages, such as Bosnian and Croatian, fall way below the cut-off of LDN = 0.48 (the same language, regardless of Yugoslavia’s existence). Some fall not far below it, such as Hindi and Urdu (different languages, barely). And varieties of Arabic and Chinese, both of which are often thought of as single languages, soar above LDN = 0.48 (the varieties are themselves different languages). Indeed, there are a few pairs of varieties that are normally considered distinct languages but which are on the borderline: Danish and Swedish, for instance, score LDN = 0.4921.
...
the amount of time it takes for dialects to drift far enough apart to qualify as separate languages ... is 1,059 years."

One of the research articles he published is at
https://www.academia.edu/37621464/How_t ... er_review_

Re: New study to distinguish between languages and dialects

Posted: Sat Jan 12, 2019 9:17 pm
by zenmonkey
It's interesting but it does seem rather limited to focus on only ...

Since each wordlist contains a fixed set of 40 concepts and are transcribed in a uniform manner, they can easily be compared, and a measure of difference can be obtained


... to complete the differentiation and nothing about grammar?

Is that right?

You can look at the database here:
https://asjp.clld.org

Re: New study to distinguish between languages and dialects

Posted: Sun Jan 13, 2019 3:43 am
by yong321
zenmonkey wrote:It's interesting but it does seem rather limited to focus on only ...

Since each wordlist contains a fixed set of 40 concepts and are transcribed in a uniform manner, they can easily be compared, and a measure of difference can be obtained


... to complete the differentiation and nothing about grammar?

Is that right?


I think you're correct. We can consider his research to be a starting point in scientifically dealing with this problem. Another thing I can think of but apparently missing in his research is that even mutual intelligibility is dependent on the mode (spoken vs. written) of communication. In case of Chinese, oral mutual intelligibility is what speech varieties are about. But people just write words down when they don't understand each verbally, and mutual intelligibility is instantly achieved, turning languages into dialects.

Re: New study to distinguish between languages and dialects

Posted: Sun Jan 13, 2019 4:17 am
by Axon
I like this a lot. I wasn't able to find out anything about the transcription protocol - are they narrow IPA transcriptions? As yong321 says, although it has flaws it is a solid piece of research that can be used for other things later on.

Another thing that can affect intelligibility is prosody and sandhi. Even closely-related dialects can sometimes have different prosody that totally throws you off for a moment before you adjust. That's part of the big exposure problem, that as someone gets exposed to more and more languages they get better and better at figuring things out and achieving intelligibility.

Re: New study to distinguish between languages and dialects

Posted: Sun Jan 13, 2019 5:39 am
by zenmonkey
Just leaving this...

The ASJP lexical items are transcribed using a broad phonetic transcription called ASJP Code (Brown et al., 2008). The ASJP Code collapses distinctions in vowel length, stress, tone and reduces all click sounds to a single click symbol.


from http://aclweb.org/anthology/C12-2095

Re: New study to distinguish between languages and dialects

Posted: Sun Jan 13, 2019 8:50 am
by Saim
Even if comparing wordlists with forty basic terms was a sound measure of linguistic difference, I don't know what this would accomplish. Once you've measured the linguistic distance why put the cut-off for languages and dialects at any given point? The fact that linguistic distance doesn't tend to hover around a midway point is interesting but I'm not sure what conclusions I'd draw from it.

I can't see what sort of use this theoretical frame would be in actual research. Now that we've established these categories what do we do with them? It seems like categorisation for categorisation's sake.

And from a political (i.e. supporting minority languages) perspective: I'd be happy to see Chinese and Arabic 'scientifically' abolished, but I'd be worried what the implications for Basque (five languages?), Asturian (Spanish dialect?), Silesian (Polish dialect?) would be... unless we finally divorce the language/dialect division from the issue of what political status different varieties should get.

Re: New study to distinguish between languages and dialects

Posted: Sun Jan 13, 2019 12:23 pm
by zenmonkey
Saim wrote:Even if comparing wordlists with forty basic terms was a sound measure of linguistic difference, I don't know what this would accomplish. Once you've measured the linguistic distance why put the cut-off for languages and dialects at any given point? The fact that linguistic distance doesn't tend to hover around a midway point is interesting but I'm not sure what conclusions I'd draw from it.

I can't see what sort of use this theoretical frame would be in actual research. Now that we've established these categories what do we do with them? It seems like categorisation for categorisation's sake.

And from a political (i.e. supporting minority languages) perspective: I'd be happy to see Chinese and Arabic 'scientifically' abolished, but I'd be worried what the implications for Basque (five languages?), Asturian (Spanish dialect?), Silesian (Polish dialect?) would be... unless we finally divorce the language/dialect division from the issue of what political status different varieties should get.


I strongly agree with Saim here. Some of the value on this type of classifications of linguistic distance that I could see are related to learning - linguistic distance may aid in approaching language like Ladakhi from Tibetan or Newar (Nepali) or anthropological - historical evolution of language separation, geographic isolation or migration and resulting shifts in language (how the Kurdish diaspora has affected pronunciation with certain vowel shifts if fascinating)... But this requires more in depth work than just classification.