Re: Sound changes causing divergence of ordinals from cardinals

Peter Bleackley wrote:
> This idea came up in the discussion of numerals. Below is a set of
> cardinal numbers, with the ordinals formed regularly from them by a
> suffix -ke. Let's take turns proposing sound changes (one at a time
> please) until the ordinals can no longer be regularly formed regularly
> from the cardinals, and indeed form a separate set of roots. Everything
> is in CXS.

(1) Labial harmony: After rounded elements, subsequent vowels become
rounded, unless there is a subsequent velar, which prevents it (cf.
English "what" vs "whack"). Thus:

(I assume the "ao" is two separate vowels. If it's a diphthong, then
20th should remain "naronaoke", 30th "mikonaoke")

  Value Cardinal       Ordinal
  1     iwQ            iwake
  2     naro           naroke
  3     miko           mikoke
  4     toko           tokoke
  5     ken            kenke
  6     jasu           jasuke
  7     Sigu           Siguke
  8     dZoz2          dZozeke
  9     hija           hijake
  10    nao            naoke
  11    haro           haroke
  12    otsu           otsuke
  13    naomiko        naomikoke
  20    naronQo        naronQoke
  21    naronQywQ      naronQywake
  22    naronQonQro    naronQonQroke
  30    mikonQo        mikonQoke
  31    naronQohQro    naronQohQroke
  32    naronQotsu     naronQotsuke
  33    mikonQomiko    mikonQomikoke