Austronesian Numbers 'Project'

Richard Parker richardparker01 at YAHOO.COM
Thu Apr 3 15:24:11 UTC 2008

  My current Austronesian/Papuan numbers chart is on the web:  which anybody is welcome to utilise (with a hat-tip, of course, or perhaps a dime or two, to me). It’s a working document, and will be updated, perhaps monthly.
  I have attempted to code each ‘critical numeral’ to show its structure as simply as possible, in a code column beside it. It is purely a structural study, and I have simplified the phonetics and orthography more than many of you will approve. There is a key to the code within the workbook. 
  I depended very heavily on the research carried out by Dr Glendon Lean  (I have incorporated most of his data and notes into the chart), but there are huge gaps:
  I have next to no numbers over 10 for the whole area west of New Guinea (ie 95% of the Austronesian-speaking world) – simply because I can’t find them on the internet.
  So I appeal to any of you who are studying particular languages to help me out. All I need are about half a dozen extra numeral forms in any particular language to be able to analyse the overall numeral structures:
  11 – 15 – 20 – 30 - 40 -50 - 100 (maybe 1000, etc)
  Also please advise if the language uses numeral classifiers, or has special numeration for certain things like coconuts, fish, people, or little round objects.
  Even if you just jot down those few numbers next time you have a particular dictionary to hand, it would be a great help.
  I would be especially grateful for contributions from anyone studying Borneo, the CEMP area or Vanuatu/New Caledonian languages, where some interesting things seem to have happened to numerals. 
  For example: 
  *sa puluq   = 1 – 10,   so PAn 20 should be *dua puluq
  And that rule seems to apply throughout the WMP area.
  But 20 changes word order in Flores (to mbulu zhua (10 - 2) in Rongga)
  And first seems* to become tautological in West New Guinea:  
  samfur-di-suru in Biak (1 - 10 - 2)
  (*I have very little data on teens or twenties west of New Guinea, hence my appeal above).
  Then mostly ‘vigesimal’ forms that might include reflexes of *sa puluq only in the teens and in odd-numbered decades, until (and beyond): - sangaul ru in Arop-Lukep (opposite the Bismarcks). 
  Most of the very little data I have on Vanuatu higher numbers show the 10-2 construction. Tryon did a great job, collecting numbers 1-10, but didn’t go further.
  and then on to: rua-sagavulu in Fiji (Bau) and lua sefulu in Samoa (with another change in word order (2-10), but still tautological). 
  In between, there are more than a few complications in mainland New Guinea, the Admiralties, Bismarcks, Solomons, and Vanuatu. 
  Manus, for instance, has almost the ‘true’ PAn constructions, but greatly abbreviated:  (Andra-Hus sanguh -10, lunguh – 20), as does New Ireland (Barok – sangaun, dura sangaun). So this may be the ‘pass-through’ area for Polynesians.
  But New Britain does not (Gimi – sungul, sungul kiep), and nor does most of mainland New Guinea. 
  Something that becomes immediately apparent from the chart is that, however ‘primitive’ the system, innate arithmetical/grammatical ‘rules’ seem to apply quite regularly. Even in so-called ‘primitive binary’ systems, 3 is always 2-1, never 1-2. It is exceptional when 11 is 1-10, not 10-1, in either An or non-An languages.
  So, to have an anomaly like the ‘tautological 10’ is ‘unruly’, and worth further investigation.  It implies that the reconstructed POc *rua na puluq = 2 – 10 = 20 might be a little shaky.
  If, as I now believe, (apparent from the data) the evolution of higher number systems, and some of the words used, long post-dated the time when PAn and POc might have been actually spoken, then the combination of the number systems and the words used could be very useful for relative dating of group migrations and settlements.   
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <>
-------------- next part --------------
An-lang mailing list
An-lang at

More information about the An-lang mailing list