Abbreviations and phonetic conventions

Tindale's phonetic transcription conventions are set out in the table on page 147 of his paper 'Results of the Harvard-Adelaide Universities Anthropological Expedition, 1938-1939' Trans Roy Soc SA 64.1, 26 July 1940.
In particular, a vertical stroke beneath a consonant means interdental, and beneath a vowel "a" changes it from that in "cat" to that in "cart". (The table also allows a vertical bar to change an "e".)
 

5.1MB high-resolution GIF here

Tindale was certainly aware of the fact that there were commonly two rhotic sounds contrasting in Australian languages.  In his outline of the phonetic transcription he employed in his first publication on the distribution of Australian Aboriginal tribes (Tindale 1940:147) he has an ‘Alveolar Rolled’ phoneme written [r] and a ‘Cerebral Fricative’ written [r], i.e. plain r and italic r respectively.  He states in a footnote that “In the accompanying map, letters with a vertical stroke beneath them correspond to those shown in black letters here; those with a dot beneath them are indicated in this text by italics.”  In a fairly extensive but not complete search, I was not able to find any letters with a dot beneath them on the map, nor could I find any italics in the language names in the text.

Tindale did use Gerhardt Laves’s method of distinguishing the two rhotics in his 1938-39 collection of ‘parallel vocabularies’.  However, he did it so badly that at first I guessed which symbol represented which rhotic wrongly. Having heard of a certain tape recording in which he can be heard repeatedly failing to hear the difference between an Aboriginal interviewee’s waru and warru, I thought it might be worth making a quantitative assessment of his success in making the distinction in his vocabularies.

To do this I chose the vocabularies of fourteen languages for which spellings of the words, or at least a fair proportion of them, as transcribed by a competent linguist or by me were available.   Some of the lists had very few r’s: Wik-Mungkan had only three, Lardil and Nyawaygi (both short lists) ten and eleven respectively.  Most had between the mid twenties and the mid thirties.  Warrungu had 44 and Yukulta was well ahead of the field with 54.  In some cases there were a fair number of words that were missing from the linguists’ vocabularies, however.

There were only two words with initial rhotics (and a third that Tindale wrongly wrote with initial rhotic).  He got one right.

Of words where the rhotic was intervocalic, Tindale got 70 of 84 right when the rhotic was a trill, and 14 out of 57 when it was a glide.  So the appearance of comparative reliability for the sequence VrrV is vitiated by the many occurrences of VrV as VrrV, not to mention the many of VrrC as VrrVC.  His apparent intervocalic V-rhotic-V sequences were wrong more than half the time, whichever rhotic it was.

Of 44 instances of a rhotic-initial consonant cluster, Tindale got two right. He had the right rhotic about half the time, but usually interpolated a vowel.

Word-final rhotics are correct in eight of 34 cases.

I conclude that there is no point in even noticing which of the two r symbols Tindale used.

© 2006 Gavan Breen

 URL: http://www.anu.edu.au/linguistics/nash/aust/nbt/symbols.html