Taboos on personal names can produce changes in grammar

Grammar is shaped by constraints on language use. This is the key finding of a study by Joe Blythe from the University of Melbourne, Australia and the Max Planck Institute for Psycholinguistics in Nijmegen, The Netherlands. The paper demonstrating this was published in the December edition of the journal Language. Here is the press release. Download the article here.

Blythe, Joe (2013). Preference organization driving structuration: Evidence from Australian Aboriginal interaction for pragmatically motivated grammaticalizationLanguage 89, no. 4: 883–919. doi:10.1353/lan.2013.0057.

This interdisciplinary study of language use is partly historical and partly interactional. If you want to know more about the research, I recommend reading the paper itself. On this page I provide a synopsis of the paper and explain some of the research methods. On the remaining pages I discuss evolutionary convergence, the conversational pressures that give rise to changes in grammar, and cause and effect relationships.


Family relationships are expressed in the grammars of many Australian Aboriginal languages. These kinship inflections have independently evolved many times because they are useful in dealing with taboos that limit the use of personal names. The research identifies patterns of language use that get around the name avoidance issues. When we speak in a group about someone known to all present, we need to ensure that we can all recognize who is being referred to. Names are excellent ‘recognitionals’ – words that can secure that recognition. When a particular name becomes ‘unavailable’ for cultural reasons, speakers of the Aboriginal language Murrinh-Patha can use their kin-inflected grammar to encourage those present to recognize the person being spoken about. For example, a woman avoids the name of her late husband by initially stating, ‘We two who were not brother and sister set off’ (ngan’gungintha ngunungamnginthadurr). She then singles out her late husband by adding, ‘He put bullets in that big rifle as he came along this way’ (thunggu banurdurditharragathu thunggu ngalla nyiniyu). Making the initial reference as ‘dual’ (‘we two’) and subsequent reference as ‘singular’ (‘he’) is a routinized pattern of language use. It places the burden of inference on the kin-inflected grammar, sidestepping the need to specify individuals by name. Widespread taboos on personal names have led to the development of kin-based inflections in languages across the Australian continent because these structures can secure recognition when names are not appropriate.

Although structural changes to words and grammar happen very slowly, these ongoing processes are happening everyday. Snapshots of grammatical change can be witnessed as the language is used in social interaction. The paper firstly describes grammatical changes occurring at various stages in the prehistory of the Murrinh-Patha language. The historical account is then supplemented by analyses of informal conversations that reveal the social motivations driving these structural changes. Together these provide zoomed-out and zoomed-in perspectives on the same processes of grammatical change. The conversational pressures motivating the emergence of kin-based inflections in Murrinh-Patha seem to be comparable with other accounts of language use in Aboriginal communities across the continent. These pressures are argued to be driving the emergence of similar structures in other Australian languages.


What did I do?

  • Noting that Murrinh-Patha’s kinship inflections do not occur in closely related Ngan’gityemerri (the only other member of the Southern Daly language family), I realized they must have emerged within the language that was to eventually become modern Murrinh-Patha, and not in the ancestor of both languages.
  • I found historical evidence showing that certain Murrinh-Patha grammatical markers had taken on new meanings – they began to mark a “siblinghood” contrast. These markers then permeated much of the language’s grammar.
  • I recorded, transcribed and annotated a corpus of informal face-to-face conversations.
  • I tracked various initial references to persons to see how those individuals are referred to subsequently.
  • When encountering kin terms or kin-inflected verbal references, I established the relationships between the individuals being referred to in the conversations, and their relationships to the people having the conversations. I then cross-referenced these individuals against a genealogical database to see whether the relationships called for circumspection about using the particular person’s name.
  • Using conversation analytic techniques, I looked for interactional evidence that the people being spoken to were able to recognize who was being spoken about, even though those people were not mentioned by name.

What I did not do.

  • I did not instruct people what to talk about before recording them.
  • I did not participate in the conversations. The conversations took place in my absence.
  • I did not ask people how best to avoid personal names.
  • I did not study only kin terms, nor only kin-based structures. The research formed part of a larger study of person reference strategies within informal Murrinh-Patha conversation.

Next, Convergent Evolution

Joe Blythe is a research fellow at the School of Language Studies, University of Melbourne, Australia and an affiliate of the Language and Cognition Department at the Max Planck Institute for Psycholinguistics, Nijmegen in the Netherlands. The research was funded by the Australian Research Council (DP110100961 and DE130100399) and the European Research Council (240853).