Ethnologue

[Dr. John E. McLaughlin and Michelle R. Sutton]

Larry Trask wrote:

> The Ethnologue volume is compiled and published by a division of the
> Summer Institute of Linguistics (SIL).  As far as I know, it has no
> connection with the UN.  The information contained in it is derived from
> an almost limitless variety of sources, of quite variable nature and
> quality, and there are lots and lots of errors.

One problem is that they have a separate entry for each language in each
country in which it is found, so an entry for, say, Blackfoot in the USA
will differ slightly (or sometimes majorly) from the entry for Blackfoot in
Canada.  If one is looking for complete information on a language spoken in
the border region of five small west African countries, one must look at
five separate (and sometimes differing) entries.  There are a high number of
typos and since they use an electronic data base for creating their language
classification trees, there are a few places where the ancestry for a given
language has been entered (by typo) incorrectly and thus shows up at an odd
place in the classification.  There are more errors that one might expect to
find in a thoroughly edited work, but the level of error doesn't obviate its
exceptional usefulness.

> The book is updated every few years.  The most recent paper edition, I
> think, is still the 12th (1992), but a new edition is in preparation and
> should be out soon, or it may even be out already, though I haven't seen
> it yet.  I haven't checked the Website lately; maybe that has already
> been updated.

The 13th edition (1996) is the current one on the shelf and the web.  The
website is www.sil.org/ethnologue.  It's easier to use on the web.

> The editors of Ethnologue actively encourage specialists to write in
> with corrections, which they are happy to incorporate in their next
> edition.  I've done this myself, and anybody who finds misinformation on
> the site or in the paper edition should consider sending in corrections.
> Ethnologue is the most serious attempt I know of at assembling
> comprehensive information about the world's languages (signed as well as
> spoken) in one place, and it deserves support.

I completely agree.  I've sent in corrections and the editors are quite
pleased and friendly about it.  There is a form on the web site that makes
sending in corrections electronic and easy.  This is arguably the best
single site for defining and listing the languages of the world (including
population data).  The Summer Institute of Linguistics has been heavily
involved in the Americas and New Guinea for decades and has an exceptional
handle on these most difficult regions.  Some linguists never look at the
site (or at Ethnologue) because SIL's primary mission is to translate the
Bible and some scholars scoff at that, but that should not dissuade any
scholar from using it as a primary source for languages of the world
information.  (If you're offended, just don't read the last line of each
entry which often includes the history of Bible translation in that
language.)  Ethnologue (1996), Voegelin & Voegelin (1977), Ruhlen (1991),
and the International Encyclopedia of Linguistics (1992) tend to be our main
sources for worldwide coverage of languages.  I rely on the first three (the
fourth is a bit too pricy to have in my personal library).  Ethnologue's
coverage of extinct languages is very sketchy.  It excludes most of them,
but includes some that are still used in formal and literary communication,
such as Standard Arabic and Latin, and a few other odd ones (such as the
four extinct "Gulf" languages--Tunica, Natchez, Atakapa, and Chitimacha and
the extinct "Coahuiltecan" language Tonkawa).  I rely on Ruhlen and Voegelin
for extinct tongues (the Encyclopedia also excludes them from its
classifications--most of its living language information comes from
Ethnologue).  As far as the classification scheme goes, Ethnologue tends to
use the most recent classification for each group that seems to be accepted
by specialists.  "Indo-Pacific" is out, "Amerind" is out, "Nostratic" is
out, "Austric" is out.  But some groupings which haven't been completely
demonstrated, but are of longstanding interest, are included such as Hokan
and Penutian.  It is superior to Ruhlen's classification system and more
current than Voegelin & Voegelin's

> Incidentally, the 1992 edition reports just over 6500 spoken languages
> as currently in use, plus some dozens of sign languages.  This is
> probably the most accurate count we have.

The 13th edition raises that number to 6690.

That's probably more than most of you ever wanted to know about Ethnologue.

John McLaughlin
Utah State University