Proto-Wanic phonology 1

Wanic was/is the supposed language family to which a bunch of Germanic-like
conlangs belonged, that I doodled on some years back.  I never developed
them beyond some handfuls of words and names, but have decided to (re)do it
all
from the bottom up.  Here goes:

    *******************************
    *                             *
    *   Proto-Wanic phonology 1   *
    *                             *
    *******************************

           LAB       DENT/ALV    VEL/PAL
                        SIB   LAT
VCL PLOS          t     ts    t$    k
VCD PLOS    b     d     dz    dL    g
VCL FRIC    f     T     s     $     x
VCD FRIC          D     z     L     Q
VCD NAS     m     n                [N]
VCD CONT    w        r        l     j


    i      [I]      u
        e  [@]  o
            a

Syllable-pattern:

(S) (P) (N) {(L)(R)(W)} V (W) {(L)(R)} (N) {(P)(S)}



"$" here denotes a voiceless lateral fricative, which I usually
would write with the UK "=A3" rather than the US currency symbol,
but I'm not sure you all could see that.  In print I would use
what would be input as \v\l in TeX -- i.e. a slashed "l" with
a "hacek".  You can probably guess, then, that "L" is a voiced
lateral fricative, which I would print as an unslashed "l" with
a "hacek".  I usually simplify "q" to "Q", which I will do here
henceforth, and write /T/, /D/ and [N] with "=FE" thorn, "=F0" edh
and "=F1" n-tilde.  I also simplify by "y" for [I] (should be barred-i)
and "v" for [@].

The <META> reason for including the lateral fricatives (apart from
them being cool sounds :-), is that </META> lateral fricatives
give an wide scope for possible changes in the daughter languages:

*$  >> x l r s S C sl Sl Sr Cr f fl xl xr

*t$ >> ts tS tr tj kl kr kj p pl ps ks

*L  >> x l r z Z j vl vr ql qr

*dL >> dz dZ dr dj zl Zl bl gl gj bz gz

(In each case with the loss of affrication as a further possi-
bility, so that all outcomes listed for *$ apply also to *t$, and
the same for *L vs. *dL.  The change *t$ > p implies *$ > f.
Another sorce of p is *nkw > mp and (heteromorphemic) *Fb *Fkw --
voiceless sibilant+b becoming voiceless sibilant+p but sibilant+velar+w
remaining, with *zgw/*zb > zw/zv/zb.

In a sense I wouldn't have had to list the affricates as separate
phonemes, since there occur numerous other CW and CR clusters like
*tj *kj *kL *gl *kr *gr *kw *gw *tr *dr *fL *fr *bl *br, but there
is e.g. no **kx **gq or **bw corresponding to them, and neither
are there *ks *gz or *fz *bz in initial position, although these
do occur medially.  It should be noted that *L and *l can be
distinguished from each other only initially and between vowels,
so that e.g. *bl or *nl could equally well be written *bL *nL.
Note also that NR and NF clusters do arise in composition (including
initial *ml *mr) but are not found within the same morpheme, and
they usually don't survive into the daughter languages, which
show insertion of a stop, or change of the nasal into a stop, or
*ml > wl/vl (the latter even initially), so that bl, vl etc. will
be found grammatically corresponding to mj.  Under similar
conditions a nasal prefix voices voiceless stops, affricates and
fricatives, but disappears itself.  Underlying *//bw// and
*//mw// *//nw// show up as b, m, m, while *//ns// etc. merge with the
reflexes of *dz etc.  Note that *//mn// becomes vn or fn,
sometimes medially bn.

As for vowels and stress the most crucial factor is the role of
the mora.  As could be expected of a mora-counting language there
exists vowels of one and two morae, but there are also "reduced"
vowels that count a half-mora (as do all consonants).  The
traditional transcription of Wanic scholars was complicated by the
fact that they didn't grasp the nature of the reduced vowels, and
posited phonemic **y and **v, since this is a common outcome of
the reduced vowels where they don't merge with short vowels or are
lost.  Reduced vowels are written with a diaeresis, or with a
following ` as *a` *e` *i` *o` *u`.  Note that traditional *y
corresponds to *a` *e` *o` and traditional *v to *i` *u`, except
that *e` *o` are affected by a high vowel in the following
syllable.  Syllables with reduced vowels have the pattern
(C)V{(W)(R)}(C).

Accentuation was variously affected by affixation, which occurred
as both prefixation and suffixation, but the former was much the
commoner, and prefixes tended to attract the stress.  If the
underlying, or "nominal", stress fell on a consonant or reduced
vowel (which it may, since the stress placement moved between
morae rather than between syllables) the quantity of that reduced
vowel and/or surrounding vowels was variously affected.  Wholly
unstressed reduced vowels were as a rule not preserved.

end


/BP

 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  B.Philip Jonsson  <bpj@...> <melroch@...>

        Solitudinem faciunt pacem appellant!
                                            (Tacitus)