HURRIAN-URARTIAN PHONOLOGY
c-HURRIAN-URARTIAN-9_phonology.htm
Tlazoltéotl Urartian divinity (restored), wood and bronze. Karmir-Blur. 8th
century BPE

PROTO-LANGUAGE PHONEMES

in IE and Hurrian-Urartian


by Patrick C. Ryan

(2/12/2001)

Copyright 1998 Patrick C.
Ryan

SUMMARY OF PHONOLOGICAL CHANGES

FROM PL TO HURRIAN-URARTIAN





Assumptions

    1. Nostratic, from which IE and Hurrian- Urartian developed, had gone through a Pontic stage, during which the earlier contrasts of CE, CA, and CO, were supplemened by glides: CyE, CA, and CwO.

      a. E, A, and O were leveled to A, yielding: CyA, CA, and CwA, which became phonemic.

        1). A (open central vowel) may have had a stress-accented allophone of 6 (schwa = close central vowel).

      b. This vocalic development enabled Ablaut gradations to take place in IE, and vowel patterning in Semitic (but not Egyptian).

    2. In all Nostratic branches except Hurrian-Urartian, PL (P)F/FH was voiced to /w/hw/.

    3. In all Nostratic branches, PL ¿(y)E became /j/.

    4. Hurrian-Urartian

      a.Glottalized stops were indicated in Hurrian by the corresponding voiceless stop but in Urartian, by the corresponding voiced stops; aspirated stops were deaspirated, and indicated in Hurrian-Urartian by voiceless stops.

      b.Labial and dorsal glottalized affricates were de-affricated, and voiced in Hurrian-Urartian, in which they were indicated by voiced stops (Urartian) or voiceless stops and non-doubled stops medially (Hurrian); aspirated labial and dorsal aspirated affricates were de-affricated, and indicated by voiceless stops (Hurrian-Urartian) or doubled stops medially (Hurrian).

        1).The apical glottalized affricate was deglottalized and voiced to /dz/ in Hurrian-Urartian, and was written s in Hurrian and S (emphatic, dotted s) in Urartian; the apical aspirated affricate was de-aspirated to /ts/ , and was written s in Hurrian-Urartian.



      c.Unaspirated spirants were voiced in Hurrian-Urartian; aspirated spirants were deaspirated and W[H] was devoiced; both were interdentalized, resulting in v and f, written u/w and u/w+u/w.

      d.The dorsal nasal was denasalized, and backed to uvular /g/, i.e. /G/, written q. Aspiration was lost in all nasals and R([H]). The phoneme rw became lw.

      e.The voiced pharyngal fricative became /j/, written i. The voiceless pharyngal fricative becomes /x/ medially, written h, but /h/ initially, which is unindicated. The voiceless laryngal (glottal) stop was maintained though not unindicated in Hurrian but written ' in Urartian. The voiceless laryngal (glottal) fricative was maintained but unindicated in Hurrian or Urartian.



      f).These developments resulted in the following consonantal inventory for Hurrian-Urartian:

      {press to see}


      Equivalencies
      I. PP glottalized stops and affricates became Hurrian unvoiced stops except PP t?sy becomes Hurrian-Urartian z (presumably ts(h)); possibly, the former affricates (except t?sy) were aspirated and lengthened the following vowel before aspiration disappeared:

      p?, t?, k? -> pIa, Ib, tIc, Id, Ie, If, kIg; p?f, t?s, k?x -> p(h)(V:?)Ih, t(h)(V:?)Ii, [zIj], k(h)(V:?)Ik, Il.
      Exx.:
      ______________________
      II. PP aspirated stops and affricates became Hurrian unvoiced stops; possibly, the former affricates were aspirated and lengthened the following vowel before aspiration disappeared:
      ph, th, kh -> pIIa, tIIb, kIIc,IId,IIe,IIf; pfh, tsh, kxh -> p(h)(V:?)IIg, t(h)(V:?)IIh, k(h)(V:?)IIi.
      Exx.:
      ______________________
      III. PP unaspirated spirants remained but f was voiced to v:
      f, s, x -> v [written w/u]IIIa, IIIb, s [written š]IIIc, x [written h]IIId, IIIe.
      Exx.: ______________________
      IV. PP aspirated spirants remained but fh was voiced to v; while becoming unaspirated, aspirated spirants may have lengthened the following vowel.
      fh, sh, xh -> v(h)(V:?) (written w[w])IVa, s(h)(V:) (written š?[š])IVb, IVc, IVd, x(h)(V:?) (written h[h])IVe.
      1)) A reluctance to realize the sequence wo/wu modified o/u after w to a/a:IVa;
      Exx.: ______________________ V. PP unaspirated nasals remained; however, q in Hurrian-Urartian probably represents not ng, a voiced dorsal nasal, but unvoiced 'back (uvular) K':
      m, n, q -> mVa, nVb, q [written q but probably K]Vc .
      Exx.: ______________________
      VI. PP aspirated nasals remained and may have lengthened the following vowel before de-aspiratization. PP q was devoiced to 'back (uvular) K' (written q but occasionally k).
      mh, nh, qh -> m(V:?)VIa, VIb, n(V:?)VIc, VId, q(V:?) [written q but pronounced k@.]VIe (see above) / kVIf.
      1) there is a fluctuation and confusion among n, r, l8 which reminds us of the Egyptian n, which is either n or l(y), and Egyptian r, which is either r or l(w);
      Exx.:
      ______________________ VII, PP unaspirated and aspirated r remained but aspirated rhw became l(w). Any r (except rhw) was devoiced in initial position and disappeared if not inverted after a vowel. In addition, aspirated r may have lengthened the following vowel:
      r -> rVIIa rhy/-(6) [rhw(6)] -> rVIIIb, VIIIc/Ø (initial) [lVIId].
      Exx.:
      ______________________ VIII. ?, the PP glottal stop, becomes Ø VIIIa, VIIIb,
      VIIIc; however, the sequence ?6¿ irregularly becomes haiIj.
      Exx.:
      ______________________ IX. h, the PP glottal (laryngal) fricative, probably lengthens the following or preceding vowel (if closing the syllable), and becomes (V:?)Ø(V:?)IXa.
      Exx.:
      ______________________
      X. ¿, the PP voiced pharyngal fricative, becomes yXa [i (in a closed syllable), initially/finally]: initial ya -> i:Xa, Xb, Xc.
      Exx.: ______________________
      XI. hh, the PP voiceless pharyngal fricative, becomes h. (written h/hhXIa, XIb).
      Exx.: ______________________
      XII. The Grundvokal, representing PP 6, is a (å); ay (from V+¿) becomes e/i and aw (from V+f/fh) becomes o [written u]. Lengthened vowels retain the "coloring" (or lack of it) occasioned by the PP glides: y6:, -6:, w6: become e:/i:, a:, o: (written u(-u).







the latest revision of this document can be found at

HTTP://WWW.GEOCITIES.COM/Athens/Forum/2803/c-HURRIAN-URARTIAN-9_phonology.htm

Patrick C. Ryan * 9115 West 34th Street - Little Rock, AR 72204-4441 * (501)227-9947

PROTO-LANGUAGE@email.msn.com





Nostratic Phonological Inventory at Pontic Stage

A (6)
P[?], P[?y], P[?w], P[H], P[Hy], P[Hw],
P[?]W, P[?]W[y], (P[?]W), P[H]W, P[H]W[y], (P[H]W),
W, W[y], (W), [H]W, [H]W[y], ([H]W)
T[?], T[?y], T[?w], T[H], T[Hy], T[Hw],
T[?]S, T[?]S[y], T[?]S[w], TS[H], TS[Hy], TS[Hw],
S, S[y], S[w], S[H], S[Hy], S[Hw],
K[?], K[?y], K[?w], K[H], K[Hy], K[Hw],
K[?]X, K[?]X[y], K[?]X[w], KX[H], KX[Hy], KX[Hw],
X, X[y], X[w], X[H], X[Hy], X[Hw],
M, M[y], M[w], M[H], M[Hy], M[Hw], N, N[y], N[w], N[H], N[Hy], N[Hw], Q, Q[y], Q[w], Q[H], Q[Hy], Q[Hw],
R, R[y], R[w], R[H], R[Hy], R[Hw],
?, ?[y], ?[w], H, H[y], H[w], ¿, J, ¿[w], HH, HH[y], HH[w]
[?] indicates glottalization; [H] indicates aspiration; [y] indicates palatal glide; [w] indicates velar glide; Q = /ng/, a voiced dorsal nasal; ? = voiceless laryngal (glottal) stop (/?/); ¿ is a voiced pharyngal fricative (/¿/); HH is a voiceless pharyngal fricative (/bar-h/); J = /j/; W = /w/; parentheses ( ) indicate a phoneme merged with another.






Hurrian-Urartian Phonological Inventory at Change a

{} indicates Hurrian variant response

a
p[?], p[?y], p[?w] (Hurrian: initial /p?/, written p; medial /p?/, written p)(Urartian /p?/, written b)
p, p[y], p[w] (Hurrian: initial /p/, written p; medial /p/, written pp)(Urartian /p/, written p)
P[?]F, P[?]F[y], P[?]F[w], PF[H], PF[Hy], PF[Hw],
W, W[y], (W), [H]W, [H]W[y], ([H]W)
(t[?y]), t[?y], t[?w] (Hurrian: initial /t?/, written t; medial /t?/, written t)(Urartian /t?/, written d)
t, t[y], t[w] (Hurrian: initial /t/, written t; medial /t/, written tt)(Urartian /t/, written t)
T[?]S, T[?]S[y], T[?]S[w], TS[H], TS[Hy], TS[Hw],
S, S[y], S[w], S[H], S[Hy], S[Hw],
(k[?y]), k[?y], k[?w] (Hurrian: initial /k?/, written k; medial /k?/, written k)(Urartian /k?/, written g)
k, k[y], k[w] (Hurrian: initial /k/, written k; medial /k/, written kk)(Urartian /k/, written k)
K[?]X, K[?]X[y], K[?]X[w], KX[H], KX[Hy], KX[Hw],
X, X[y], X[w], X[H], X[Hy], X[Hw],
M, M[y], M[w], M[H], M[Hy], M[Hw], N, N[y], N[w], N[H], N[Hy], N[Hw],
Q, Q[y], Q[w], Q[H], Q[Hy], Q[Hw],
R, R[y], R[w], R[H], R[Hy], R[Hw],
?, ?[y], ?[w], H, H[y], H[w], ¿, j, ¿[w], HH, HH[y], HH[w]

EXPLANATION OF NOTATION





Hurrian-Urartian Phonological Inventory at Change b

{} indicates Hurrian variant response

a
p[?], p[?y], p[?w] (Hurrian: initial /p?/, written p; medial /p?/, written p)(Urartian /p?/, written b)
p, p[y], p[w] (Hurrian: initial /p/, written p; medial /p/, written pp)(Urartian /p/, written p)
b, b[y], b[w] (Hurrian: initial /b/, written p; medial /b/, written p)(Urartian /b/, written b)
p[h], p[hy], p[hw] (Hurrian: initial /ph/, written p; medial /ph/, written pp)(Urartian /ph/, written p)
W, W[y], (W), [H]W, [H]W[y], ([H]W)
(t[?y]), t[?y], t[?w] (Hurrian: initial /t?/, written t; medial /t?/, written t)(Urartian /t?/, written d)
t, t[y], t[w] (Hurrian: initial /t/, written t; medial /t/, written tt)(Urartian /t/, written t)
dz, dz[y], dz[w] (Hurrian: initial /dz/, written s; medial /dz/, written s)(Urartian /dz/, written S [emphatic, dotted s])
ts, ts[y], ts[w] (Hurrian: initial /ts/, written s; medial /ts/, written s)(Urartian /ts/, written s)
S, S[y], S[w], S[H], S[Hy], S[Hw],
(k[?y]), k[?y], k[?w] (Hurrian: initial /k?/, written k; medial /k?/, written k)(Urartian /k?/, written g)
k, k[y], k[w] (Hurrian: initial /k/, written k; medial /k/, written kk)(Urartian /k/, written k)
g, g[y], g[w] (Hurrian: initial /g/, written k; medial /g/, written k)(Urartian /g/, written g)
k[h], k[hy], k[hw] (Hurrian: initial /kh/, written k; medial /kh/, written kk)(Urartian /kh/, written k)
X, X[y], X[w], X[H], X[Hy], X[Hw],
M, M[y], M[w], M[H], M[Hy], M[Hw], N, N[y], N[w], N[H], N[Hy], N[Hw],
Q, Q[y], Q[w], Q[H], Q[Hy], Q[Hw],
R, R[y], R[w], R[H], R[Hy], R[Hw],
?, ?[y], ?[w], H, H[y], H[w], ¿, j, ¿[w], HH, HH[y], HH[w]

EXPLANATION OF NOTATION





Hurrian-Urartian Phonological Inventory at Change c

{} indicates Hurrian variant response

a
p[?], p[?y], p[?w] (Hurrian: initial /p?/, written p; medial /p?/, written p)(Urartian /p?/, written b)
p, p[y], p[w] (Hurrian: initial /p/, written p; medial /p/, written pp)(Urartian /p/, written p)
b, b[y], b[w] (Hurrian: initial /b/, written p; medial /b/, written p)(Urartian /b/, written b)
p[h], p[hy], p[hw] (Hurrian: initial /ph/, written p; medial /ph/, written pp)(Urartian /ph/, written p)
v, v[y], (v) (Hurrian: initial /v/, written w/u; medial /v/, written w/u)(Urartian /v/, written w/u)
f, f[y], (f) (Hurrian: initial /f/, written w/u; medial /f/, written w/u+w/u)(Urartian /f/, written w/u)
(t[?y]), t[?y], t[?w] (Hurrian: initial /t?/, written t; medial /t?/, written t)(Urartian /t?/, written d)
t, t[y], t[w] (Hurrian: initial /t/, written t; medial /t/, written tt)(Urartian /t/, written t)
dz, dz[y], dz[w] (Hurrian: initial /dz/, written s; medial /dz/, written s)(Urartian /dz/, written S [emphatic, dotted s])
ts, ts[y], ts[w] (Hurrian: initial /ts/, written s; medial /ts/, written s)(Urartian /ts/, written s)
z, z[y], z[w] (Hurrian: initial /z/, written š; medial /z/, written š)(Urartian /z/, written š)
s, s[y], s[w] (Hurrian: initial /s/, written š; medial /s/, written šš)(Urartian /s/, written š)
(k[?y]), k[?y], k[?w] (Hurrian: initial /k?/, written k; medial /k?/, written k)(Urartian /k?/, written g)
k, k[y], k[w] (Hurrian: initial /k/, written k; medial /k/, written kk)(Urartian /k/, written k)
g, g[y], g[w] (Hurrian: initial /g/, written k; medial /g/, written k)(Urartian /g/, written g)
k[h], k[hy], k[hw] (Hurrian: initial /kh/, written k; medial /kh/, written kk)(Urartian /kh/, written k)
?, ?[y], ?[w] (Hurrian: initial /?/, written h; medial /?/, written h)(Urartian /?/, written h)
x, x[y], x[w] (Hurrian: initial /x/, written h; medial /x/, written hh)(Urartian /x/, written h)
M, M[y], M[w], M[H], M[Hy], M[Hw], N, N[y], N[w], N[H], N[Hy], N[Hw],
Q, Q[y], Q[w], Q[H], Q[Hy], Q[Hw],
R, R[y], R[w], R[H], R[Hy], R[Hw],
?, ?[y], ?[w], H, H[y], H[w], ¿, j, ¿[w], HH, HH[y], HH[w]

EXPLANATION OF NOTATION





Hurrian-Urartian Phonological Inventory at Change d

{} indicates Hurrian variant response

a
p[?], p[?y], p[?w] (Hurrian: initial /p?/, written p; medial /p?/, written p)(Urartian /p?/, written b)
p, p[y], p[w] (Hurrian: initial /p/, written p; medial /p/, written pp)(Urartian /p/, written p)
b, b[y], b[w] (Hurrian: initial /b/, written p; medial /b/, written p)(Urartian /b/, written b)
p[h], p[hy], p[hw] (Hurrian: initial /ph/, written p; medial /ph/, written pp)(Urartian /ph/, written p)
v, v[y], (v) (Hurrian: initial /v/, written w/u; medial /v/, written w/u)(Urartian /v/, written w/u)
f, f[y], (f) (Hurrian: initial /f/, written w/u; medial /f/, written w/u+w/u)(Urartian /f/, written w/u)
(t[?y]), t[?y], t[?w] (Hurrian: initial /t?/, written t; medial /t?/, written t)(Urartian /t?/, written d)
t, t[y], t[w] (Hurrian: initial /t/, written t; medial /t/, written tt)(Urartian /t/, written t)
dz, dz[y], dz[w] (Hurrian: initial /dz/, written s; medial /dz/, written s)(Urartian /dz/, written S [emphatic, dotted s])
ts, ts[y], ts[w] (Hurrian: initial /ts/, written s; medial /ts/, written s)(Urartian /ts/, written s)
z, z[y], z[w] (Hurrian: initial /z/, written š; medial /z/, written š)(Urartian /z/, written š)
s, s[y], s[w] (Hurrian: initial /s/, written š; medial /s/, written šš)(Urartian /s/, written š)
(k[?y]), k[?y], k[?w] (Hurrian: initial /k?/, written k; medial /k?/, written k)(Urartian /k?/, written g)
k, k[y], k[w] (Hurrian: initial /k/, written k; medial /k/, written kk)(Urartian /k/, written k)
g, g[y], g[w] (Hurrian: initial /g/, written k; medial /g/, written k)(Urartian /g/, written g)
k[h], k[hy], k[hw] (Hurrian: initial /kh/, written k; medial /kh/, written kk)(Urartian /kh/, written k)
?, ?[y], ?[w] (Hurrian: initial /?/, written h; medial /?/, written h)(Urartian /?/, written h)
x, x[y], x[w] (Hurrian: initial /x/, written h; medial /x/, written hh)(Urartian /x/, written h)
m, m[y], m[w], (m), (m[y]), (m[w]), n, n[y], n[w], (n), (n[y]), (n[w]),
G, Gy, Gw, (G), (Gy), (Gw) (Hurrian: initial /G/, written q; medial /G/, written q)(Urartian /G/, written q)
r, ry, lw, (r), (ry), (lw) (Hurrian: initial /r/ and /l/, written r and l; medial /r/ and /l/, written r and l)(Urartian /r/ and /l/, written r and l
?, ?[y], ?[w], H, H[y], H[w], ¿, j, ¿[w], HH, HH[y], HH[w]

EXPLANATION OF NOTATION





Hurrian-Urartian Phonological Inventory at Change e

{} indicates Hurrian variant response

a
p[?], p[?y], p[?w] (Hurrian: initial /p?/, written p; medial /p?/, written p)(Urartian /p?/, written b)
p, p[y], p[w] (Hurrian: initial /p/, written p; medial /p/, written pp)(Urartian /p/, written p)
b, b[y], b[w] (Hurrian: initial /b/, written p; medial /b/, written p)(Urartian /b/, written b)
p[h], p[hy], p[hw] (Hurrian: initial /ph/, written p; medial /ph/, written pp)(Urartian /ph/, written p)
v, v[y], (v) (Hurrian: initial /v/, written w/u; medial /v/, written w/u)(Urartian /v/, written w/u)
f, f[y], (f) (Hurrian: initial /f/, written w/u; medial /f/, written w/u+w/u)(Urartian /f/, written w/u)
(t[?y]), t[?y], t[?w] (Hurrian: initial /t?/, written t; medial /t?/, written t)(Urartian /t?/, written d)
t, t[y], t[w] (Hurrian: initial /t/, written t; medial /t/, written tt)(Urartian /t/, written t)
dz, dz[y], dz[w] (Hurrian: initial /dz/, written s; medial /dz/, written s)(Urartian /dz/, written S [emphatic, dotted s])
ts, ts[y], ts[w] (Hurrian: initial /ts/, written s; medial /ts/, written s)(Urartian /ts/, written s)
z, z[y], z[w] (Hurrian: initial /z/, written š; medial /z/, written š)(Urartian /z/, written š)
s, s[y], s[w] (Hurrian: initial /s/, written š; medial /s/, written šš)(Urartian /s/, written š)
(k[?y]), k[?y], k[?w] (Hurrian: initial /k?/, written k; medial /k?/, written k)(Urartian /k?/, written g)
k, k[y], k[w] (Hurrian: initial /k/, written k; medial /k/, written kk)(Urartian /k/, written k)
g, g[y], g[w] (Hurrian: initial /g/, written k; medial /g/, written k)(Urartian /g/, written g)
k[h], k[hy], k[hw] (Hurrian: initial /kh/, written k; medial /kh/, written kk)(Urartian /kh/, written k)
?, ?[y], ?[w] (Hurrian: initial /?/, written h; medial /?/, written h)(Urartian /?/, written h)
x, x[y], x[w] (Hurrian: initial /x/, written h; medial /x/, written hh)(Urartian /x/, written h)
m, m[y], m[w], (m), (m[y]), (m[w]), n, n[y], n[w], (n), (n[y]), (n[w]),
G, Gy, Gw, (G), (Gy), (Gw) (Hurrian: initial /G/, written q; medial /G/, written q)(Urartian /G/, written q)
r, ry, lw, (r), (ry), (lw) (Hurrian: initial /r/ and /l/, written r and l; medial /r/ and /l/, written r and l)(Urartian /r/ and /l/, written r and l
?, ?[y], ?[w] (Hurrian: /?/, written Ø)(Urartian: /?/, written ')
h, h[y], h[w] (Hurrian: /h/, written Ø)(Urartian: /h/, written Ø)
j, j, j[w] (Hurrian: /j/, written i)(Urartian: /j/, written i).
(x), (x[y]), (x[w]) (Hurrian: initial /h/, written Ø; medial /x/, written hh)(Urartian: initial /h/, written Ø; medial /x/, written h)

EXPLANATION OF NOTATION





Final Hurrian-Urartian Phonological Inventory

{} indicates Hurrian complementary orthography for medial occurrence

a, i, u
p? [p{-pp-}], ph [p{-pp-}]
b [b/p]
v [w/u/b/p], f [w/u/b/p{w/u/b/p+w/u/b/p}]
t? [t{-tt-}], th [t{-tt-}]
d [d/t]
dz [z{z}], ts [s; U: S (dot-s)]
z [š], s [š{-šš-}; Ug: th]
k?[k{-kk-}], kh [k{-kk-}]
g [g/k]
g [h]; c [h{-hh-}]
palatal g [š/z], palatal c [š/z{-šš-/-zz-}]
m [m]; n [n]; ? ['/h; medial q]
r [r]; l [l]; L (dark /\/) [l{-ll-}]
j [i; U: i/g]; ? [Ø; U: ']; h [h].







==== Proto-Pontic The main root inventory of Proto-Pontic, therefore, consisted of Cy,-,w(6), supplemented by early compounds of the form Cy,-,w6C(y,-,w). It cannot be overemphasized that there was only one original vowel, virtually only the feature of syllabicity, in all these language-familes at the time of their formation. In PIE, 6 (schwa) was fronted only to e, fronted and lowered to a (i.e. a@v) in PAA, backed and lowered in Sumerian and Hurrian-Urartian to a (i.e. a@o). In PIE, i and u are late reflexes of avocalic PP ¿ and f/fh; o is only a late Ablaut variation of e. Diphthongs are combinations of stress-accented e/o + i/u from the same sources. Long vowels are reflexes of voiceless affricates which became aspirates before their aspiration disappeared after lengthening the following vowels, plus vowel. Long vowels also arose from a contiguous h, which, of course, is simple aspiration without a previous obstruction, and from contiguous h@-. In addition, h@- lowered PP 6, which would have become PIE e, to a; a should be regarded as an allophone of e in specified phonetic environments, and not a separate phoneme. The change of 6 to a has been studied in Kabardian (Eastern Adyghe), where it has been demonstrated that (j/w)a and a@- is a product of hy, w+6 and h+6 (Kuipers 1960). In PAA, i and u are Ablaut variations, on the one hand; but also stem from w and y, which is an allophone of ¿ under (un)certain phonetic conditions. Therefore, a grouping of PIE and PAA (and others) under any rubric such as Proto- Nostratic, which reconstructs more than one vowel with non-phonemic allophonic variations, is fundamentally wrong. It is compounding an error with a mistake to reconstruct a consonantal inventory with sporadic glides for some consonants but not for others in a system with full vocalism. Glides originate in only one way: from contiguity with front or back vowels. When palatal and velar glides are available, a distinction between front and back vowels becomes redundant. Glideless articulations identify former mid vowels as well. For earliest Proto-Nostratic, on the evidence of PIE and PAA, the only permissible vowel is 6, syllabicity, which simply becomes PIE e and PAA a without moderating influences. Even though PIE and PAA may not often show different reflexes for consonants without glides and for those with y or w, glides should be reconstructed to account for those consonants which do since glides were occasioned by former PL front and back vowels, which followed every consonant. In late PIE, glides were absorbed mostly without trace (but see below). Tellingly, the only two consonants with glides that properly should be reconstructed for late PIE are gw and kw; those that remained were not derived from PP k?w and khw but from PP x and xh so it is probable that they differed in some significant way in a transitional stage from gw and kw from k?w and khw. That gw and kw had a spirantal origin is assured by their reflexes in Sumerian (š and h); Egyptian (š and h); PAA (Arabic (š); and Hurrian-Urartian (h(h)). Equally impermissible is a reconstruction of Proto-Nostratic without phonemes representing reflexes of PP q and qh. Though the PIE reflexes may be considered ambiguous when taken alone, the PAA reflex of both PP phonemes is straightforward: q, representing a uvular k (in Egyptian as q and g). Apparently, its reflex in Hurrian-Urartian was also q. The Sumerian reflexes, however, are rather more complicated ((n)g/k/n). Though there are languages which differentiate between rounded and unrounded labials, dental and alveolar apicals, and palatal and velar dorsals -- all variations on a theme of front and back in articulatory positions -- PIE would be asymettrical and hence typological unusual if it had only distinguished between front and back articulations in the velar series. From their reflexes in Greek (b, d, g, p, t, k), it should be abundantly clear that something other than simple velarized dorsals were at work. Indo-European Phonological Changes PP glottalized stops and affricates became PIE voiced unaspirated and aspirated stops except initial p? (usually) and all p?y -> w: p? [p?y], t?, k? -> b (usually w initial33) [w], d, g36; p?f, t?s, k?x -> bh, dh, gh. PP aspirated stops and affricates became PIE unvoiced stops with sporadic aspiration for the former affricates and lengthening of the following vowel (tshw appears as tw): ph, th, kh -> p, t, k. pfh, tsh [tshw], kxh -> p(h)(V:), t(h)(V:) [tw(V:], k(h)(V:). PP unaspirated spirants were voiced; in addition, the resulting t was stopped and velarized: f, s, x -> w, z (later, s), gw. There is, almost certainly, a connection between the velarization of this phoneme and the almost universal tendency to round back articulations. PP aspirated spirants were de-aspirated; in addition, the resulting x was stopped and velarized; shw appears as sw. Aspirated spirants sporadically lengthened the following vowel: fh, sh [shw], xh -> f (later, w)(V:), s(V:) [sw(V:], kw(V:). PP unaspirated nasals remained but q was frequently denasalized to g, and always reduced to g in initial position unless an epenthetic vowel preceded; this, apparently, was governed by the stress-accent. The exception was ny6, which became l(y)e: m, n [ny], q -> m, n [l(y)], (Vn)(n)g. PP aspirated nasals developed similarly except qh was devoiced to (n)k, and aspirated nh, before any vowel, becomes l. Aspirated nasals sporadically lengthened the following vowel: mh, nh, qh -> m(V:), l(V:), (Vn)(n)k(V:). PP unaspirated r remained. Aspirated r remained but rhw became l(w). Aspirated r sporadically lengthened the following vowel: r -> r. rh [rhw] -> r(V:) [l(w)(V:)]. Among the PP tectals, the "coloring" discovered in connection with lost PIE "laryngeals", a singularly inappropriate designation, is due to the influence of the glides w and y principally. ?, the PP glottal stop, becomes 0@/e+, e+0@/e+, or e+0@/. h, the PP glottal (laryngal) fricative, becomes 0@/6@-/e+, 6@-/e+0@/e+, or 6@-/e+0@/. ¿, the PP voiced pharyngal fricative, becomes ye+, e+i/yC/e+, or e+y. h@-, the PP voiceless pharyngal fricative, becomes 0@/a@-/a+, a@-/a+0@/e+, or a@-/a+0@/. Lengthened schwa (6@-) retained the "coloring" occasioned by the preceding PP glides: y6@-, -6@-, w6@- become e@-, a@-, o@-. The Grundvokal, representing PP 6, was e (supra). After influencing the reflexes of certain consonants and vowels noted above, glides were absorbed by their consonants and disappeared, leaving faint traces like palatalization of g and k (perhaps n) in satem dialects. As stated above, velarized g and k are from a different source (PP x and xqh). Hurrian Phonological Changes General Stripped of awkward and imaginative theoretics, we see a vowel system in Hurrian- Urartian of a, i, u. Since Ablaut seems not to have been present, the most economical explanation of i and u is that they derive from earlier y and w just as they do in PIE. There is no really convincing evidence for either e or o except, perhaps, as allophones of i and u. Plene- spellings should be simply regarded as indications for clarity (materes lectionis) or as indications of lengthened vowels or stress-accent. For the stop series of Hurrian-Urartian, we see no distinction being made for voiced (glottalized?) and voiceless (aspirated?) articulations in initial position; these are written, seem- ingly whimsically, as b/p, d/t, g/k. If such a distinction operated phonemically and hence semantically, Hurrian-Urartian scribes would have followed consistent writing conventions or modified cuneiform to reflect them. That they did not, shows conclusively that they did not exist, at least at the time of our transcriptions. Scholars can entertain preconceptions with greater conviction than laypeople but a preconception remains unacceptable until it is proven. A preconception that mars Hurrian- Urartian studies is the idea that a language should distinguish voice from voiceless articulations phonemically. Languages of the immediate area (Elamite) show bare stop series (p, t, k) with gemination contrasts (lax/tense; McAlpin 1981: 65 ff.) perhaps analogous to the doubling of consonants in Hurrian (but not Urartian), which, has, however, here been intrepreted as indicating intervocalic voicelessness in Hurrian-Urartian, for which interprtetation there is again, meagre evidence. In Sumerian, also, we find the same reflex for C? and Ch, e.g. PP p? and ph both result in Sumerian b, p being reserved as a reflex for p?f and pfh. This is true also for Sumerian d and t as well as g and k with the exception that t(?)s(h)y has the special reflex z (phoneticallly ts(h)). s(h)y has the special reflex š while s(h)-,w becomes s. Similarly, x(h)y has the special reflex š (here, at least originally, phonetically ç) while x(h)-,w remains x (written h). As for the other spirant, PP f(h) only shows up as u (presumably, originally w) except that before a front vowel, just as in Armenian, it becomes g. PP ¿ similarly appears only as i (presumably, originally y). The glottal stop and h/hh, perhaps after lengthening the vowel, fall out. PP m(h) and n(h) remain but n(h)y or n(h)y becomes l. Somewhat analogously, r(h) remains but r(h)w or r(h)w becomes l. Finally, PP q(h) is seen as (n)g, (n)k, and n, with patterning yet to be fully clarified. This is similar to the case of Egyptian where the contrast between C? and Ch has been neutralized; the contrast between Egyptian d and t (both Coptic t) is based on the secondary con- trast between vowel glides: PP t(?/h)y/- and t(?/h)w -> th so that Egyptian d represents t phonetically, and Egyptian t is T (from th). This doublet is matched by p and f for p(?/h)y/- and p(?/h)w -> ph (f) and k and t for k(?/h)y/- and k(?/h)w -> kh (x). Interestingly, we see analogous developments from PP ?/h/¿/h@- where Egyptian i and h are ?/h/¿/h@-y/- and ?/h/¿/h@-w, which become phonetically y and h. That the y-glide formed the basis of one of these pairs is virtually made certain by z (phonetically, sh) and s as well as š (phonetically ç) and h (phonetically back x). As might be expected, the contrast of wy and ww was neutralized, or at least cannot be traced; both are Egyptian w. Similarly p(?)f(h)y and p(?)f(h)w both resulted in Egyptian b (phonetically v) while the affricates k(?)x(h)y and k(?)x(h)w were also voiced so that Egyptian H represents palatal t phonetically, and Egyptian h@v is velar t. Finally, we have t(?)s(h)y and t(?)s(h)w, written d and ', phonetically dzh and dz, the latter of which eventually developed into the glottal stop. There seems to be no discernible difference in the reflexes of the nasals m and n but the velar nasal had k@. for PP q(h)y (phonetically (n)gy) and g for PP q(h)w (phonetically (n)gw). To complete the description of Egyptian consonants, 3 was PP r(h)y and r was PP r(h)w. What motivated a voiced pronunciation of the affricates cannot be determined at this time. Accordingly, the simplest interpretation of the facts that we see is that the Hurrian- Urartian reflex of PP C?, h was unitary as it was in Sumerian and Egyptian, though admittedly not in PAA as a whole. In the stop series, stopped and probably spirantal articulations were positionally not phonemically determined: initial word-position (purely antevocalic) = stop; any other = spirantal; similar to Hebrew. That the articulation was spirantal rather than voiced is suggested by the interchange of y and g (t) in several Urartian words7. We may hypothesize but shall never be able to conclusively prove that aspiration was lost after producing lengthened vowels. Though this development might be expected, it is, by no means, certain. We may also speculate that voiceless consonants contiguous to voiced nasals (m and n) and "liquids" (r and l) and possibly "semi-vowels" (y and w) had an allophonic voiced variants but this is also immaterial phonemically and hence semantically. The relatively more important issue is what the significance of Hurrian (but not Urartian) intervocalic doubled consonants is. The safest path is to assume that if the distinction was not made in Urartian, it was neither phonemic nor semantically meaningful. To posit the maintenance of a distinction between voiced and voiceless phonemes only intervocalically through doubling seems unduly contrived and doubly improbable when it was not employed in Urartian. This is for stops -- but what of doubled non-obstruents: should we assume typologically rare voiceless m, n? The only reasonable course is to assume that doubling consonants had some other function than indicating voiceless stopped allophones of otherwise intervocalic voiced (or better, spirantized) stops. The easiest assumption is to adopt the lax/tense distinction which exists in Elamite (and related Dravidian) which seems not to be phonemic (semantic) but rather vaguely morphological, as, of course, gemination is, more clearly in PAA. While it may be that doubled intervocalic stops were voiceless (their normal initial articulation) and stopped (as opposed to a voiced or spirantal articulation), this has no semantic significance. At best, it may be indicating a secondary stress-accent to define morpheme boundaries desirable in Hurrian but found unnecessary in Urartian. In Hurrian-Urartian, affricates have additionally coalesced with the reflexes of the stops with the sole exception, as in Sumerian, of PP t(?)s(h)y, which is seen as z. We are far from understanding the nature of stress-accent in Hurrian-Urartian, but that it did exist is strongly suggested by Urartian bamni, 'the mountain`, seemingly a reduction of "bab(a)ni. If doubled consonants indicate stress-accented syllables, it might explain the equivalent Hurrian form pappini (pa"pini <- *"pap(a)+ini).