Thursday, July 02, 2015

US Census Bureau and African languages

The number of people speaking an African language at home in the United States* increased by 111% in the decade before the last census. Which African languages are most common among them? Looking at a map of most spoken African languages by state in the US from an article by Ben Blatt in Slate last year kind of answers that, but then raises more questions. To begin with, the names listed in the map include two languages (Amharic and Swahili), two branches of language families (Bantu and Cushitic), one language group (Nilotic), and a seemingly eclectic grouping of "Kru, Ibo, and Yoruba."

According to the Census Bureau site, it tabulates non-English speaking households by 328 languages or language categories (see the full list of codes). A list of African language categories excerpted from the Census Bureau's list raises more questions (the below includes headings and subheadings from the list; those categories marked with "" are listed in the Slate article as the "choices"; those highlighted in yellow are listed in one or more states on the map):

Source: "Tagalog in California, Cherokee in Arkansas: What language
does your state speak?" by Ben Blatt, Slate, 13 May 2014
Germanic languages (607-618)
611    Afrikaans

Pacific Islands languages (730-776)
   West Indonesian (732-741)
738    Malagasy

OTHER LANGUAGES (679-683, 696-697, 777-999)
Semitic languages (777-780)
777    Arabic
780    Amharic

Afro-Asiatic languages (781-783)
781    Berber
782    Chadic
783    Cushite

Nilo-Saharan languages (784-789)
784    Sudanic
785    Nilotic
786    Nilo-hamitic
787    Nubian
788    Saharan
789    Nilo-saharan
790    Khoisan

Niger-Congo Languages (791-798)
791    Swahili
792    Bantu
793    Mande
794    Fulani
795    Gur
796    Kru, Ibo, Yoruba
797    Efik
798    Mbum (and Related)
799    African (not further specified)

Time for the Census Bureau to review its African language categories?

It is hard to know what to say about this list. It is obvious that some thought has gone into it (despite the puzzling combination of two major and distinct Volta-Niger languages - Yoruba and Igbo - with the linguistically and geographically distant Kru languages). And indeed the Bureau indicates it has based the categories "generally on Classification and Index of the World's Languages (Voegelin, C.F. and F.M., 1977)" and that it updates them "constantly using linguistic books and online resources." But without further  background it is hard to get away from the feeling that it is a bit random.

Another unknown is what the process is for determining who is counted under what category. The Census questions include an open-ended one for people who speak a language other than English at home: "What is this language?" Only the few languages listed directly on the list would be simple to tabulate - such as Amharic and Swahili. But if someone answers, say "Hausa," they would at some point (presumably not that of the enumerator) have their answer tabulated under "Chadic," since Hausa is in the Chadic branch of the Afro-Asiatic family of languages. If "Zulu," then "Bantu," and so on. The reason for tabulating under broader categories, as I understand it, has to do with confidentiality concerns (when there are few people speaking a language in an area).

However, if "Wolof" or "Akan" is the answer, there doesn't seem to be a category under Niger-Congo that applies, other than the catch-all 799. How many cases are there like this?

Another issue is that a couple of categories may not be current - Sudanic and Nilo-Hamitic.

Altogether, it would seem that an expert review and revision of the categories would be timely, as we are still five years from the next census. That would make the results more useful.
* According to U.S. Census Bureau, Statistical Abstract of the United States: 2012, "Table 52. Population by Selected Ancestry Group and Region: 2009," a total of 777,553 people speak African languages (not specified) at home. This total does not include Arabic, which was counted separately at 845,396.


Ted Johnson said...

How do these speakers of African languages in the US influence business decisions -- marketing, content, user interfaces? Not at all, is my guess.

Don said...

Thanks Ted. I think you're right, but a few thoughts. First the numbers of speakers of African languages in the US are low - so much that figures for most languages are aggregated by group or family. That wouldn't help much for anyone interested in localizing for specific demographics. Second, my impression is that aside from a very few major groups - notably Hispanophones, but also some others like Chinese - language is not a major consideration in such business decisions within the US. Third, on more local levels, and where public education and information is involved, languages that are important in specific communities might enter into consideration. How is Somali used in the Twin Cities area for example?