Re: vowel harmony extension?

Nik Taylor wrote:

>Kristian Jensen wrote:
>>    wati  'dandruff'        w~a~ti    'demon'
>>    bipi  'swollen'         mi~pi~    'badger'
>>    diti  'to lose'         ni~ti~    'coal'
>>    aka   'give food'       a~ka~     'choke on a bone'
>
>Is that first one a typo?  Should it be w~a~ti~?
Is it enough if I just write, "Ooops!"

Actually, this reminds me of other nasal harmony systems that I
purposely left out to keep things simple. There are nasal harmony
systems where certain consonants are opaque to nasal spread. That
means that the long distance spreading of the feature [+nasal] is
blocked by certain types of consonants. There is also a natural
heirarchy of what types of consonants are prone to be blockers of
nasal harmony. Basically vowels are least prone and voiceless stops
are most prone. Schematically this is:

   High <----- Compatability with Nasal Harmony -----> Low
   VOWELS     GLIDES      LIQUIDS     FRICATIVES     STOPS

Malay (Johore dialect), for instance, has the type where only vowels
and glides are involved in the spread of nasal harmony, while liquids,
fricatives, and stops are opaque to nasal harmony -- that is, they block
it. In Malay, nasality spreads to the right of a nasal consonant and
continues until an opaque consonant blocks the spread of the nasal
feature. E.g.:

   [mi~no~m]      'drink'
   [baNo~n]       'rise'
   [ma~j~a~N]     'palm stalk'
   [m@~na~w~a~n]  'to capture'
   [m@~ratappi]   'to cause to cry'
   [p@Na~w~a~san] 'supervision'

(Examples taken from: "Reinterpreting Transparency in Nasal Harmony" by
Rachel Walker)

In languages where stops and fricatives are also involved in the spread of
nasal harmony, usually these stops and fricatives (especially the voiceless
ones) are transparent to nasal harmony. That is, they do not block nasal
spread but at the same time remain oral.

There is also the issue regarding direction of spread. In the Malay
example above, nasality is progressive. But there also exists those that
are regressive. And then there are those that are bidirectional where it
is possible for each direction to differ in the kinds of consonants that
are opaque to nasal spreading.

Spreading can also be limited to certain prosodical units like syllables
and foot. In Guarani, for instance, nasality is blocked by a stressed
oral syllable. In my conlang, Boreanesian, the foot is the domain of
nasal/oral spans.

BTW, the reason why I have all this info handy is because I have done tons
of research to figure out how nasal harmony and other harmony systems work
in order to better understand how I can implement the feature in Boreanesian.
I discovered that there was a great deal to consider before I could implement
nasal harmony in Boreanesian phonology.

-kristian- 8)