We know that Aboriginal languages, despite their sophisticated grammar and ability to express abstract ideas, tend to have restricted numeral systems. Claire Bowern, from Yale University, has comprised a study of numeral systems of Australian languages, which usually escape a linguist’s attention. The research paper examines the diversity of numeral systems found in Australia and suggests they have an important place in world’s numeral typologies. Here’s the summary of the article, and a few interesting points I’ve gathered from it.
When talking about numbers in Indigenous languages, we think of a stereotypical ‘one, two, many’ system. Contrary to this widespread view, only 5% of 189 sample languages examined in the article have their numeral term limit of ‘two’. Seventy four percent have the upper limit of three and four, and 11% have an upper limit of five. The remaining 10% have limits that vary from six to, surprisingly, one hundred (Tyapwurrung, a Kulin language). Here’s the statistics illustrated:
Most of the languages surveyed combine smaller numerals to create larger ones. For numbers such as three, four, and five, some languages have both atomic (indivisible) and compositional form (derived from smaller units) for the same number. For instance, in Warlpiri: compositional jirrama-(kari)-jinta ‘three’ (literally, ‘two (and) one’); atomic marnkurrpa ‘three’. It’s fascinating that most languages surveyed use some kind of base, whether additive, multiplicative, or additive-multiplicative, to form larger numbers. In this respect, they exhibit similarity to unrestricted numeral systems.
Etymology of numerals may also be quite compelling. Rembarrnga has numerals between ‘two’ and ‘five’ that are counted on the fingers (from little finger to thumb) and have the names of wallaby species. For instance, ngabutj ‘female black red-eyed rock wallaroo’ signifies ‘two’, garndalppurru ‘female antilopine wallaby’ stands for ‘three’. Similarly, the Karnic term *kulpari for ‘three’ is based on the emu’s distinctive three-toed footprint.
Also, it is common to use the word ‘hand’ to denote numerals such as ‘five’ or ‘ten’. Examples include mara in Wangkumara and Yandruwandha; in Worrorra, ‘five’ is ŋanuŋalja ‘my wrist’. ‘Foot’ is used for numerals between ‘ten’ and ‘twenty’ in languages such as Gamilaraay, where balar-ya dinna is ‘twenty’ (lit., ‘two-ya foot’) and balar-ya marra is ‘ten’ (lit., ‘two-ya hand’).
There are forms in the data that are clearly borrowed from English. For example, a loan from English exists in Martu Wangka, where ‘six’ is jikaki, derived from English ‘six card’, the six in a deck of cards. The Yolŋu languages in Arnhem Land use terms based on English words for money in counting currency and other financial transactions. These include dudala ‘two dollars’ and dindala ‘ten dollars’, as well as tjilba ‘small change’ (from ‘silver’).
The author, Claire Bowern, has published several papers on numeral systems of Australian languages. Her research is, of course, not restricted to this subject. According to her profile on The Conversation, “her research focuses on the Indigenous languages of Australia, and is concerned with language documentation/description and prehistory. This includes fieldwork in Northern Australia with speakers of endangered languages, as well as archival work, shedding light on the linguistic history of Pama-Nyungan. With colleagues in linguistics, anthropology, and evolutionary biology, she is currently comparing features of hunter-gatherer languages in different areas of the world.”
Claire Bowern’s profile: http://theconversation.com/profiles/claire-bowern-1098
“Diversity in the Numeral Systems of Australian Languages” – Jstor link:
Published by Sonya (Sofia) Mamonova