Re: Japanese Long Consonants

Okay. At the moment.... the structure of each syllable is:

(C)V(C)

So a word is like:

(C)V(C)(C)V(C).....

Sorry... I'm not sure how to write it, but I was using brackets to show
that something is optional. So basically each syllable has to contain a
vowel, but both initial and coda consonants are optional. :) The
consonants that can occur in coda position are limited though: mostly
fricatives, nasals, l, r and (under the current system) the glottal
stop. I'm not sure how realistic it is allow a glottal stop as a coda
consonant but not any other stops, especially since the glottal stop
can't occur as an initial consonant (it's always coda).... I suppose I
might be able to come up with an explanation if I thought hard enough.
:)  So for instance:

dùnen
/du?nen/

is possible, but

dùsnen
/du?snen/

isn't, since each syllable can have only one coda and one initial
consonant. Initially I was thinking about not counting the glottal stop
(so syllables could have a coda consonant and a glottal stop), so
"dùsnen" would be legal, but so far I've decided against having any
words like that. I could change it though, since adding flexibility
doesn't make any of the words I've come up with so far illegal.
 As for sound changes involving the glottal stop... I haven't really
finished designing the proto-language, but I've already sketched some of
the glottal stop influenced sound changes that take place in the
daughter languages (again, subject to change):

Language One: Intervocalic softening (like lenition) happens, blocked by
the glottal stop. The glottal stops then disappear (I haven't decided if
this causes vowel lengthening, but I don't want it to lengthen the
consonants), giving rise to a system where the accents on the vowels
actually indicate whether the following consonant is softened or not, eg:

dùden
/duden/
duden
/duDen/

note the dental fricative in the second. Such a system could also give
rise to welsh style initial consonant mutations, although I'm still
working on arranging it properly so that I can work my way to the
changes being essentially grammatically determined. :)

Language Two: Softening doesn't occur. Glottal stops merge with
following consonants to produce long consonants. Where a glottal stop
occurs between vowels, it vanishes but the two vowels remain separate
rather than becoming a diphthong. I'm then planning a series of vowel
demolishing sound changes to produce more interesting consonant clusters
(I haven't decided what changes yet though.... possibly something along
the lines of vowel centralization similar to english followed by
dropping of @ in certain circumstances). The vowel demolishing and
consonant assimilation afterwards in this language will probably destroy
the distinctiveness of a lot of the grammatical (especially verbal)
endings, thus leading to more pronoun use and the end of the language
being pro-drop.

This is all still being planned, or subject to revision... after the
glottal stops intervocally vanish, I may then merge consecutive vowels
to give new vowel sounds like /y/ as you suggest (I think /y/ is the
right letter for the sound ü isn't it?)
 BTW, so far the long and short vowels have the same quality, unlike in
English. :)

Chris.

>What actually is the syllable structure of your language? (sorry if I've
>missed a previous discussion).  Let's say:
>
>.VCV..  and ...V:Cv..
>.VCCV... (? and ..V:CCV.. -- that might sort-of violate a sort-of
>universal, but IIRC Latin, Greek and Sanskrit allowed long V before 2
>consonants, but universals are there to be violated :-) )
>
>Now, if one of the C's is /?/, there's no intrinsic reason you couldn't have
>V?V and V:?V
>V?CV and V:?CV
>
>Possible sound changes:
>
>1. Loss of /?/ after short V--
>V?V > new single V combining features of both, e.g. a?i > e, u?i > y; (in my
>Gwr, any sequence of i/u > 1) Like V would presumably > long, e.g. a?a > a:
> Many possibilities, including diphthongs.
>
>V?CV > compensatory length of V1: V:CV OR of the C: VC:V
>
>2. Loss of /?/ after long V:--
>V:?V > distinct V in separate syllables, a:?i > a(:).i etc. (If your
>language doesn't allow vowels in hiatus, then introduce a glide of some
>sort.)
>V:?CV > V:CV  or perhaps V:C:V ???
>
>Another question/possibility:  are the long V different in quality, or
>simply lengthened versions of the short V?  Perhaps /a/ = [6], /a:/ = A, /e/
>= [E], /e:/ = [e]???
>
>
>
>

Reply