Composed number (also called "double plural")

Bernhard Waelchli bernhard.waelchli at ISW.UNIBE.CH
Mon Oct 12 15:46:35 UTC 2009

Dear typologists,

Does anybody know of more languages with double nominal plural (or 
dual/trial/paucal) marking where both plural (or dual/trial/paucal) 
markers added to the same noun express different kinds of 
plurality/duality? This is called "composed numbers" in Corbett (2000: 
36, "number built as it were on another"). I mean cases such as, e.g.:

(a) X-PL 'Xs',
X-PL-PL 'different groups/sets of Xs, large quantities of X'

(b) X-PL 'a plurale tantum X',
X-PL-PL 'a plural of a plurale tantum X'

Paraphrasing the question for the specific (b)-kind:
Does anybody know languages where pluralia/dualia tantum can be 
pluralized just by adding another plural/dual marker?

I know of the following languages with composed numbers (or groups of 
languages): Ful, Burushaski, Breton, Warekena, and Mongolian. It has 
also been reported for Khamtanga (Chamir, Agaw, Cushitic), and Arabic. 
According to Corbett (2000: 36) composed number is rare, but I am rather 
confident that somebody of you knows about some more examples of a 
similiar kind as those:

Ful (Adamawa) (Klingenheben 1963: 139)
nagge 'cow[SG] > cow'
na'i 'cow[PL] > cows'
na'i-ije 'cow[PL]-PL > groups/flocks of cows'

Burushaski (Lorimer 1935: 46)
tsəriš SG 'one part of a door frame'
tsərʌŋ PL 'one complete doorframe'
tsərʌŋ-čiŋ DOUBLE-PL 'door frames'

Breton (Ternes 1992: 415; Corbett 2000: 36)
bugel 'child > child'
bugal-e 'child-PL > children'
bugal-e-où 'child-PL-PL > groups of children'

Warekena (Aikhenvald 1998: 300-4; Corbett 2000: 37)
abida-pe 'pig-PL > pigs'
abida-nawi 'pig-GREATER.PL > very many pigs'
abida-pe-nawi 'pig-PL-GREATER.PL > very many pigs indeed, too many to count'

Khalkha Mongolian (Poppe 1970; Plank, 1985: Linguistics 23.45-82)
lamə [SG] 'lama'
lamə-nər PL 'lamas'
lamə-nər-uud PL-PL 'highly esteemed lamas'

(As far as I know, no triple plural with three different kinds of plural 
meanings has been reported. Or does somebody have any?)

(I am NOT interested in this query in the more common case of two or 
more distinct plural suffixes expressing just a single kind of 
plurality, such as in Dutch _kind-er-en_ 'children' [or etymologically 
English _child-r-en_] or the more spectacular case of Breton 
diminutives, Stump 1990, Corbett 2000: 152)

Many thanks in advance!

Bernhard Waelchli
University of Bern

More information about the Lingtyp mailing list