humanswhoreadgrammars:jespru:mapsontheweb:Mutual Intelligibility Map vmpfc1:This is a map of
humanswhoreadgrammars: jespru: mapsontheweb: Mutual Intelligibility Map vmpfc1: This is a map of languages that are mutually intelligible (based on Wikipedia and testimony from native speakers at UC Berkeley). This map is not meant to be precise, as I have had to make choices about where to exclude languages in dialect continuums (i.e. Romance languages like Occitan). Also, my knowledge of Austronesian and Bantu languages is very limited, so suggestions for improvement are greatly appreciated. Thanks! I think Farsi/Dari/Tajik could be on thereCorrect me if I’m wrong though What’s going on here? Who did this? Is there a publication behind this? Could we all try and help track down where this is from? The reddit discussion and comment about native speakers at UC Berkeley and wikipedia is NOT satisfactory (and obviously a HIGHLY biased sample).Oh, btw: you would probably all enjoy reading chapter 5 of Hammarström’s PhD thesis from 2009 It’s called Counting Languages in Dialect Continua Using the Criterion of Mutual Intelligibility. You can find it for free here. Here’s an excerpt: In trying to answer the question “how many languages are there in the world?”, linguists have had a hard time coming up with a satisfactory answer. Even when explicitly disregarding non-linguistic criteria (such as ethno-socio-economicopolitico- cultural ones), they say that defining languages by the mutual intelligibility criterion (MI) is not possible (e.g. Anderson 2005). Firstly, mutual intelligibility is not a strict yes/no distinction but a matter of degree. Subsumed hereunder are also cases where the degree of intelligibility is not enough to enable communication immediately, but high enough to enable communication after, say, only a few days of exposure, such as among the Mekeo languages (Jones 1998:19). Also, there may exist cases where intelligibility is not symmetric, i.e., A understands B but not vice versa, although I have yet to see a genuine well-documented example. Secondly, even if it were simplified into being yes/no and symmetric, counting languages by the MI, would lead to contradictions in dialect-chain situations. E.g. if A is MI with B, B is MI with C but A is not MI with C – a completely realistic situation – then setting A and B to be the same language and B and C as the same language is contradictory because A and C are not the same language by the MI criterion.In this paper we will show that the second objection is premature. There is a perfectly consistent way to count languages using a symmetric strict yes/no mutual intelligibility criterion that preserves intuitive properties about languages and numbers of languages. /h -- source link