The official count of the world’s languages has recently increased by 96 languages. The steward of this figure, SIL International, released an annual report in January that reviewed changes in three-letter codes that are assigned to languages by the International Standards Organization; these codes are metadata used in information organization and software development.
In the 2005 edition of Ethnologue, SIL’s language atlas & gazetteer, there were 6912 languages; this represented 103 languages more than Ethnologue counted in 2000.
On the surface, these increases are interesting, since it appears to
contradict the notion that languages are disappearing, and the overall count of languages shrinking. But when you dig more deeply, you see it’s how languages are lumped and split.
Of the new codes, 47 were assigned because previously assigned codes were rearranged. The remaining 59 were assigned to languages that apparently had never been described before. Twenty of the new codes were proposed by Jamin Pelkey, a linguistics graduate student from La Trobe University in Melbourne, Australia, who “discovered” about half of of those languages in a 2006 language survey in southern China for his dissertation. The other half, more or less, had been previously documented by Chinese linguists or by Pelkey’s supervisor at La Trobe, named David Bradley. I went through the change requests to count Pelkey’s proposals. Pelkey submitted the highest number of proposals to ISO.
Two of the new language codes were given to artificial languages: Kotava, an artificial language invented in 1978 that has approximately 40 fluent speakers in France and Polynesia, and Lingua Franca Nova, spoken by an online community of 180 people. (I’m quoting from the ISO proposals.)
Three were sign languages, all in south Asia.
The remainder of new codes are “new” by virtue of being recategorized in broad “macrolanguage” families. (That is, they weren’t the product of two languages merging or two dialects splitting).
A couple of things are significant about this: one, because people are mapping linguistic diversity more thoroughly, more codes are needed to describe linguistic diversity even as that diversity decreases; two, we’ll never really know “how many” language there are (or were) on the planet, given varying ways to distinguish “language” from “dialect.”