Countering the claims about Australia's Aboriginal number systems

Ray Norris wrote recently that he thought some academic researchers were unaware of the diversity of Aboriginal ways of counting.

As part of his argument he examined a paper I wrote with a former student, Kevin Zhou, that was published last year in the Proceedings of the Royal Society (Biology). The paper studied change across time in numeral systems in Australia’s largest language family, Pama-Nyungan.

Norris claimed that our paper was about the largest numbers in Australian Aboriginal languages that could extend beyond ten, but still didn’t extend past 20. He gave some examples of other Aboriginal counting systems with limits higher than those we talked about in the paper, leading him to ask:

As a physicist, I am fascinated by the fact that the authors of this paper didn’t engage with the contrary evidence. They simply didn’t mention it. Why?

The short answer to Norris' question is that I had already published two papers on the full diversity of Australia numeral systems, and this paper he referred to was about a different question.

The first paper (with then student, Jason Zentz) looked at how counting word systems vary across the continent.

It included the Tjapwurrung case Norris mentioned, as well Young People’s Tiwi and systems from the Western Desert and elsewhere, such as Kukatja’s or Warlpiri’s, where the words for numbers above five are loans from English, or based on the shapes of Arabic numerals. For example:

6 = jika (from English “six”)

7 = wirlki (a boomerang with arms of uneven length, often called a “number 7 boomerang”)

8 = milpa (“eyes”)

9 = kartaku (“cup, billycan” – referring to the shape of the cup plus handle viewed from above)

Body-tallying

We also discussed other systems which involve counting, including body tallying, where speakers point at different parts of the body to refer to different quantities. Counting on fingers is a small example of body-tallying.

The Australian examples come from three languages in the Torres Strait, Arnhem Land and Victoria. While body-tallying is a way of counting, it’s not strictly a numeral system (since the term applies to words).

We showed there was a lot more variation than the stereotypical idea that Aboriginal languages only have numbers up to four, and we discussed ways of counting (such as birth order names) that do not involve numerals.

For example, the Kaurna language of the Adelaide Plains has words for “first born child”, “second born child”, up to the tenth child. The oldest child is kartammeru for a male child and kartanya for a girl, the fourth is munnaitya (male) or munato (female), the fifth is midlaitya (male) or midlato (female).

No limits

The paper Norris referred to was about how the small number systems change over time. There are systems with 2, 3, 4, 5, 6, or 7 or more numbers in languages that are related to one another.

Can languages lose numerals over time, or only gain them? When languages gain numerals, do they do so slowly, and stepwise, or do they add many numbers at a time?

We never hypothesised an upper limit. We found that languages did lose numerals over time as well as gain them, and when they gained them, the systems tended to increase rapidly.

In short, we didn’t talk about larger systems in the paper because that wasn’t what the paper was about. Norris' comment on that paper is somewhat like criticising a number theorist who publishes a paper on prime numbers for failing to discuss the composite number 6.

I agree with Norris that a great deal of misinformation continues to circulate about Aboriginal languages and the people who speak (and spoke) them.

As a gardiya (an Aboriginal word for a white person in many languages of northern Western Australia, where I do a lot of my work) who unfortunately spends a fair amount of time with that literature, I can only imagine what it must be like as an Aboriginal person to read such material.

The portrayal of Aboriginal people in scholarship historically is, in the aggregate, horrible. There’s no way of glossing over it. But the fact that people have said uninformed things in the past about Aboriginal people does not implicate current researchers by association.

Language preservation

So let’s talk about what we are actually doing to help preserve Aboriginal languages.

Since 2007, I have been compiling a database of Australian language words.

The Chirila database now contains more than 750,000 words and continues to grow. Where I have permission, I have made portions of it freely available online, and I continue to work informally with communities and Aboriginal language organisations across Australia, such as the Resource Network for Linguistic Diversity.

That archival and curatorial work is complemented by my own ethnographic and linguistic fieldwork, going back to my time as an undergraduate at the ANU.

The Chirila database is not just a research resource, it’s a source for Aboriginal people looking to reclaim their languages, a testament to their survival and their speakers’ perseverance, and a reminder of the brilliant diversity of Indigenous Australia.

As a scientist and historical linguist, my job is to figure out how languages have changed. It’s about engaging with the evidence, asking questions, figuring out how to test my hypotheses and puzzling over what the results might mean. It’s not that different from physics, in the end.