Australia - absence of sibilants

Question

Are there any sciencific/linguistic/historical theories about reasons of absence of sibilants in some Australian languages? As far as I know, sibilants are common accross world languages. Since sibilants are common, it would be nice to hear/read any theory about lacking of sibilant sounds in Australia.

Answer 1

I remember reading about a hypothesis that Australian languages may have been influenced by high-frequency hearing loss caused by otitis media. High frequencies are supposed to be more important for accurately perceiving certain kinds of fricatives, perhaps sibilants in particular (as described in this answer--I can't give an explanation myself because I don't know phonetics).

You can see that idea mentioned by Andrew Butcher on the following web pages:

• "The special nature of Australian phonologies: Why auditory constraints on the sound systems of human languages are not universal", The Journal of the Acoustical Society of America 144, 1939 (2018) (abstract).

  Unfortunately, I haven't been able to read more than the abstract, and I also can't find any other work that has cited this article.

• "Australian phonologies and Aboriginal hearing", Seminar by Andy Butcher, Professor at Flinders University, Adelaide and Visiting Fellow of The Cairns Institute, March 5th 2012 ("About the Seminar")

Answer 2

The problems with attributing this fact to otitis media are (a) that given this study, the causes of the condition may be modern environmental and not genetic and (b) it also afflicts populations that don't have the sibilant gap. One could conjecture that OM is a minor contributing factor. A founder effect combined with areal diffusion seems the most plausible explanation. The weakness of the founder effect is that there are dozens of language families and isolates in Australia, though again that may be because the original language was spoken so long ago that we can't prove monogenesis.

Areal diffusion refers to the fact that people in contact who speak different languages tend to adopt features of each other's languages (perhaps asymmetrically depending on social relationships). This is evident in southern Africa, where the "Khoisan" language hypothesis seems to have been based on typological similarity. An example of an areally-connected gap is the lack of otherwise linguisticaly-universal nasals in Twana and Lushootseed (Salishan), Quileute (Chemakuan) and Makah and Ditidaht (Southern Wakasah). It is also reported that one finds lack of [s] but presence of [θ/ð] is Dinka, Lango, Didinga and Kikuyu – the latter is Bantu, the former are Nilo-Saharan. However, these languages are in reasonable proximity and this fact might be attributed to an areal phonetic property.

I should point out that the factual generalization is a little bit overstated for Australia at least at the phonetic level. I noticed a number of papers in the O'Grady festschrift (Boundary rider) report fricatives as allophones of stops in a number of Australian languages. Sibilantization of θ,ð is possible albeit not common. I should also point out that sibilants are not common in the world's languages. Stop are common. It is correct that most languages have a sibilant consonant, and sibilant-free languages are uncommon. It's good practice to state the generalization correctly.

Answer 3

I think that a phenomenon known in biology as founder effect is responsible here: When the proto-language lacks fricatives, its descendants probably continue with this particular feature. In addition, on the isolated continent there was no contact with other languages that could give raise to fricatives.