Skip to main content

Method and limitations


AIATSIS carried out NILS3 as a survey. The survey had two participant groups:

  1. Language centres and other organisations working on language projects
  2. Linguists who work with Aboriginal and Torres Strait Islander languages.

The process of developing the survey included the following steps:

  • The NILS3 Indigenous Advisory Group gave feedback on the survey design.
  • AIATSIS tested the survey with pilot participants.
  • The AIATSIS Research Ethics Committee approved the project.
  • Participation in the survey was voluntary and every question was optional.
  • NILS3 asked respondents to choose which AustLang code they were reporting on.

(AustLang is a dataset of over 1200 Aboriginal and Torres Strait Islander language varieties. AustLang includes languages, dialects, and group, clan and people names.)


Limitations of the data gathered include:

  • AIATSIS received responses for 141 language varieties. This means there are a number of language varieties with no data.
  • AIATSIS did not receive a response for languages known to have many speakers. These languages include Anmatyerr, Burrara, Tiwi and Aboriginal English.
  • In most cases, AIATSIS received one response per language variety. If there was more than one response, sometimes these answers conflicted. When this happened, the AIATSIS Languages team made an informed choice about which answers to accept.
  • Respondents interpreted ‘fluency’ differently. The survey tried to prevent this by giving definitions of fluency.
  • Some respondents counted only 'full' speakers. Other respondents counted all 'full', 'part' and 'little' speakers.
  • NILS3 only reported on spoken languages. It did not report on signed languages. There are many Indigenous Australian sign languages.
Last updated: 01 March 2022