[Lingtyp] Lingtyp Digest, New Issue: Nominal Classifiers in IE languages

Paul Hopper hopper at cmu.edu
Tue Jun 18 10:31:28 EDT 2019

Research on the lexical and cognitive sources of classifiers and noun categorization in general continues to be a significant theme of typology, and Edmond Cane's discussion of Albanian is bound to engage much interest. However, I'm hesitant about Edmond's remark "The main function appears to have been unitizing for the purpose of quantification." Classifier research needs to be supplemented, or even preceded, by actual studies of the role played by classifiers in discourse (e.g., why do speakers unitize/quantify?) There are a number of such studies from the 1980s on; one of them is my paper <https://www.researchgate.net/publication/247583502_Some_Discourse_Functions_of_Classifiers_in_Malay>

-- a study based on monologic written texts. Especially valuable these days would be research from the spoken dialogic perspective of Conversational Analysis/Interactional Linguistics.

Paul Hopper

Paul J. Hopper
Paul Mellon Distinguished Professor Emeritus of Humanities
Department of English
Carnegie Mellon University
Pittsburgh PA 15213, USA

From: Lingtyp <lingtyp-bounces at listserv.linguistlist.org> on behalf of Edmond Cane <ecane2000 at yahoo.com>
Sent: Tuesday, June 18, 2019 9:38 AM
To: lingtyp at listserv.linguistlist.org
Subject: [Lingtyp] Lingtyp Digest, New Issue: Nominal Classifiers in IE languages

The following is a short summary of an argument I am going to present in a conference on Language Typology, in July. As I have read, the numeral classifiers are common among languages in Asia as well as  some other areas of the world. Indo-European languages are not evaluated as rich in such pattern. My knowledge can be limited, so I would appreciate any comments, remarks about the matter. My main concern is how widespread this pattern is among IE languages, how (if any) these compare to other languages generally described as numeral classifier languages - and such relevant issues

Albanian with regard to nominal classifiers

Several authors have analyzed the numeral classifiers (NC) in typological or case studies (Greenberg 1977, Craig 1986, Dixon 1986, Croft 1994, Seifart 2010, etc.). A large number of Asian languages (with an isolating typological profile), but also some Caucasian, West African and American Indian languages, have been evaluated and argued as nominal classifier ones based on certain features (Greenberg 1977, Aikhenvald 2000, Nichols 1992, Bisang 1996, 1999, Seifart 2010, etc.).  However, even Albanian, an IE language, is abundant with such constructions, with the features as defined in the cited literature. Their use is not strictly obligatory. Many enumerated nouns may also choose two or more alternative NCs. The choice is basically semantic and they constitute an open lexical class.

The main function appears to have been unitizing for the purpose of quantification. Then, because of the means employed, there has been a nominal classification imposed on the enumerated nouns. There is a clear classification of human versus nonhumans

Num + frymë (souls) – for people

Num + krerë/kokë (heads) – for animals mainly for domestic animals, but also expanded to other animals (less common).

As for inanimates, there is a wide range of NCs, generally having emerged from usage, needs and utilitarian purposes. There is structuring based on how speakers interact with objects:

5 barrë dru/shkarpa/qymyr, etc. – 5 loads wood/kindling/coal

2 krahë dru – 2 arms  wood

Një dorë miell/kripë/para, etc – one hand flour/salt/money

Një grusht miell, ujë, para, etc  - a fist flour/water/money

This use could be similar to the English: a handful of.., an armful of…, a fistful of….

Shape and size are often the underlying cognitive factors to the emergence of some NCs

3 kokrra mollë – 3 round-shaped apples

2 pëllëmbë mustaqe – 2 palms mustache (measuring 2 palms in length)

Liquids have their own NCs

Një bardhak/kupë ujë/qumësht – one bowl/cup of water/milk

Një kotruve ujë – one bucket water

Very often liquids and hard matter like sand, flour, sugar etc. have the same NC, their proper container vessels. For the same noun or set of nouns, there can be more than one alternative, which can be a different container and generally provides for different size or amount.

The NCs impose numerability on mass nouns but also on countables too (3 kokrra mollë/3 kokrra fiq/10 kokrra domate etc.). It is interesting to note that they convert even some abstract nouns to numerability

një vorbull ere (a swirl of wind);

një tis avulli (a soft breeze/cover of steam),

një shkundje ere (a shake of wind) etc.

It has been observed even in other languages that the NCs do not simply serve the purpose of providing for unitizing and counting of the nouns. The numeral NPs very often involve the numeral ‘one’. It can also be seen that in some cases the opposition underlying the presence of the NC is that of none vs. any.

Metonymy is a very common source: breath/soul for people, heads for cattle and some other animals, roots for trees, etc. The most common mechanism is the metaphorical projection of certain schemas, lexical items, which already ‘carry numerability’, onto new scenes and schemas. From idiosyncratic lexical items, they spread and may become more or less generic. Hence, the NCs can be considered as ‘vessels’ of numerability.

Most of the usage underlying the NCs has almost faded away: either such vessels are not used nowadays, or, such work is not carried out in the present. However, the pattern remains strong and large. The historical emergence and use must have provided for the deeply embedded cognitive patterns, which accounts for the productive spread of the NCs. The presence of several alternatives for the same elements supports this view too. The NCs are idiosyncratic lexemes, and their lexical content imposes constraint on their use. However, the strength of the pattern seems to have encouraged the spread beyond the expected line. This accounts for their expansion to abstract nouns, attempting to unitize or individualize elements like wind, sea water, or abstract ones like pleasure, anger, work, time, etc. Some NCs have even provided for figurative collocations.

Albanian has a rich system of NCs, but very few collective classifiers (a series of, a heap of, a pack of, etc.) – comparative to languages like English. While the NCs are highly constrained, the collective classifiers apply to large sets. The few ones open to humans, other animates as well as inanimates, (një grumbull …– a collection of …; një mori … - a series of …) or, for animals and inanimates (tufë – also used for human, when implying brutes, primitives, uneducated), or for a wide range of inanimates (pirg – a heap of). Even the existing (low-level) constraint seems to be basically due to the pre-emptive effect of the few constructions in use.

The typical NCs appear in a morpho-syntactic blend, a straight juxtaposition, just like in the numeral constructions (Num + N). There are some NCs which can appear in both, the straight blend with simple form juxtaposition or having the enumerated noun in ablative (indefinite), which is actually a simple form, and very common in many NPs. The NCs have two forms: singular for numeral ‘one’ and plural for ‘two’ or higher. Both forms are indefinite.

The numeral constructions can also develop into a full inflected NP, in which case it is not a numeral classifier NP:

Një/dy barrë dru – one/two load woods

Barra e druve – the load of woods

Hence, Albanian appears as a numeral classifier language, and this requires an explanation considering inheritance (as an IE language) as well as the areal environment. In addition, Greenberg argues that the numeral classifier system developed in languages with a prior mass/unit nouns distinction. Albanian obviously does not qualify as such.

The strength of the pattern in Albanian can be regarded as a deeply embedded behavioral pattern, supported/supplied /provided for by the large number of available schemas, the wide usage and the already functional needs. The large number of the NC constructions represent a particular combination of semantic structure and information packaging. This evidence may contribute to what can be learned from classifier systems with regard to the nature of language organization.


Aikhenvald, Alexandra Y. 2000. Classifiers. A typology of noun categorization devices. Oxford: Oxford University Press.

Bisang, Walter. 1996. Areal typology and grammaticalization: processes of grammaticalization based on nouns and verbs in East and Mainland South East Asian languages. Studies in Language 20. 519–97.

Bisang, Walter . 1999. Classifiers in East and Southeast Asian languages. Counting and beyond. Numeral types and changes worldwide, ed. by Jadranka Gvozdanovı´c, 113–85. Berlin and New York: Mouton de Gruyter, Trends in

Linguistics. Studies and Monographs 118.

Croft, William. 1990. Typology and universals. Cambridge: Cambridge University Press.

Croft, William (1994) Semantic universals in classifier systems, Word, 45:2,145-171,

Croft, William. 2014. Grammatical categories, semantic classes and information packaging. Draft chapter of book project Morphosyntax: Constructions of the world’s languages. http://www.unm.edu/~wcroft/Papers/Morphosyntax-ch1-Jan14.pdf.

Greenberg, J.  (1972) Numeral Classifiers and Substantival Number: Problems in the Genesis of a Linguistic Type. Working papers on Language universals, No. 9..1-40,

Haspelmath, Martin. 2010a. Comparative concepts and descriptive categories in crosslinguistic studies. Language 86. 663–687.

Nichols, Johanna. 1992. Linguistic diversity in space and time. Chicago: University of Chicago Press.

Seifart, F. (2010). Nominal classification. Language and linguistics compass, 4(8), 719-736.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/lingtyp/attachments/20190618/ee3653aa/attachment.html>

More information about the Lingtyp mailing list