Indo-European sound laws
As the Proto-Indo-European language (PIE) broke up, its sound system diverged as well, as evidenced in various sound laws associated with the daughter Indo-European languages.
Especially notable is the palatalization that produced the satem languages, along with the associated ruki sound law. Other notable changes include:
- Grimm's law and Verner's law in Proto-Germanic
- an independent change similar to Grimm's law in Armenian
- loss of prevocalic *p- in Proto-Celtic
- Brugmann's law in Proto-Indo-Iranian
- Winter's law and Hirt's law in Balto-Slavic
- merging of voiced and breathy-voiced stops, and /a/ and /o/, in various "northern" languages
Bartholomae's law in Indo-Iranian, and Sievers's law in Proto-Germanic and (to some extent) various other branches, may or may not have been common Indo-European features. A number of innovations, both phonological and morphological, represent areal features common to the Italic and Celtic languages; among them the development of labiovelars to labial consonants in some Italic and Celtic branches, producing "p-Celtic" and "q-Celtic" languages (likewise "p-Italic" and "q-Italic", although these terms are less used). Another grouping with many shared areal innovations comprises Greek, Indo-Iranian, and Armenian; among its common phonological innovations are Grassmann's law in Greek and Indo-Iranian, and weakening of pre-vocalic /s/ to /h/ in Greek, Iranian and Armenian.
Consonants
The following table shows the Proto-Indo-European consonants and their reflexes in selected Indo-European daughter languages. Background and further details can be found in various related articles, including Proto-Indo-European phonology, Centum and satem languages, the articles on the various sound laws referred to in the introduction, and the articles on the various IE proto-languages, language groups and language phonologies. For development of the laryngeals and syllabic consonants, see the vowels table below.
PIE | Indo-Iranian | Balto-Slavic | Alb. | Arm. | Anatol. | Toch. | Greek | Italic | Celtic | Germanic | |||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Sanskrit | Avestan | O.C.S. | Lith. | Hitt. | Latin | Old Irish | Gothic | English | |||||||
normal | C+[j] [C 1] | normal | -C- [C 2] [C 1] | ||||||||||||
*p | p; ph [pʰ] [C 3] | p; f [C 4] | p | h; w [C 5] |
p, pp | p | pt | p | ∅ | f; b [β] [C 6] |
f; v, f[C 2] | ||||
*t | t; th [t̪ʰ] [C 3] | t; θ[C 4] | t | tʿ [tʰ] | t, tt; z [ts] [C 7] |
t; c [c] [C 7] |
t | s; tt/ss[C 5] | t | t | th [θ] | þ [θ]; d [ð]; [C 6] |
th; d; [C 6] | ||
*ḱ | ś [ɕ] | s | š [ʃ] | th [θ]; k[C 8] |
s | k, kk | k; ś [ɕ][C 8] |
k | c [k] | c [k] | ch [x] | h; g [ɣ] [C 6] |
h; ∅;[C 2] y [C 6] | ||
*k | k; c [t͡ɕ]; [C 7] kh [kʰ] [C 3] |
k; c [tʃ]; [C 7] x[C 4] |
k; č [tʃ]; [C 7] c [ts][C 9] |
k | k; q [c][C 9] |
kʿ [kʰ] | |||||||||
*kʷ | k; s; [C 7] q [c][C 9] |
ku, kku | p; t; [C 7] k[C 10] |
qu [kʷ]; c [k] [C 11] |
ƕ [ʍ]; gw, w [C 6] |
wh; w [C 6] | |||||||||
*b | b; bh [C 3] | b; β [C 12] | b | p | b | pt | b | b [b] | -[β]- | p | |||||
*d | d; dh [C 3] | d; δ [C 12] | d | d; dh [ð][C 2] |
t | ts; ś [ɕ] [C 7] |
d | z [zd] > [z] | d | d [d] | -[ð]- | t | |||
*ǵ | j [d͡ʑ]; h [ɦ] [C 3] |
z | ž [ʒ] | dh [ð]; g[C 8] |
c [ts] | k | k; ś [ɕ][C 8] |
g | g | g [ɡ] | -[ɣ]- | k | c / k; ch[C 9] | ||
*g | g; j [d͡ʑ]; [C 7] gh; [C 3] h [ɦ] [C 3] |
g; j [dʒ]; [C 7] γ [C 12] |
g; ž [ʒ]; [C 7] dz[C 9] |
g | g | k | |||||||||
*gʷ | g; z; [C 7] gj [ɟ][C 9] |
ku | b; d; [C 7] g[C 10] |
u [w > v]; gu [ɡʷ] [C 13] |
b [b] | -[β]- | q [kʷ] | qu | |||||||
*bʰ | bh [bʱ] | b; β [C 12] | b | b; w[C 2] |
p | ph [pʰ] | pt | f;[C 14] b |
b [b]; b [β];[C 2] f [C 15] |
b; v / f[C 16] | |||||
*dʰ | dh [dʱ] | d; δ [C 12] | d | t | t; c [c] [C 7] |
th [tʰ] | tt/ss | f;[C 14] d; b [C 17] |
d [d] | -[ð]- | d; d [ð];[C 2] þ [C 15] |
d | |||
*ǵʰ | h [ɦ] | z | ž [ʒ] | dh [ð]; d[C 8] |
j [dz]; z[C 2] |
k | k; ś [ɕ] [C 7] |
kh [kʰ] | h; h / g[C 8] |
g [ɡ] | -[ɣ]- | g; g [ɣ];[C 2] g [x] [C 15] |
g; y / w[C 16] | ||
*gʰ | gh [ɡʱ]; h [ɦ] [C 7] |
g; j [dʒ]; [C 7] γ [C 12] |
g; ž [ʒ]; [C 7] dz[C 9] |
g | g | g; ǰ [dʒ] [C 7] | |||||||||
*gʷʰ | g; z; [C 7] gj [ɟ][C 9] |
ku | ph [pʰ]; th [tʰ]; [C 7] kh [kʰ][C 10] |
f;[C 14] g / u [w];[C 2] gu [ɡʷ] [C 13] |
g; b;[C 14] w;[C 2] gw [C 13] |
g; b;[C 14] w[C 2] | |||||||||
*s | s | h [h, x] | s | sh [ʃ]; gj [ɟ];[C 18] h[C 2] |
h; ∅[C 2] |
š [s] | s; ṣ [ʂ] |
h;[C 14] s;[C 19][C 15]/ ∅;[C 2] [¯] [C 20] |
i | s; r[C 2] |
s ʃ | -[h]- | s; z [C 6] |
s; r [C 6] | |
ṣ [ʂ][C 21] | š [ʃ][C 21] | x [x][C 21] | š [ʃ][C 21] | ||||||||||||
*m | m | in | m | m [m] | -[w̃]- | m | |||||||||
*-m [C 15] | m | ˛ [˜] | n | ∅ | n | n | -- | m [˜] | n | ∅ | |||||
*n | n | n; ˛ [˜] [C 15] |
n | n; ñ [ɲ] |
n | in | n | ||||||||
*l | r (dial. l) | r | l | l; ll [ɫ][C 2] |
l / ɫ [ɫ > ɣ] |
l | il | l | |||||||
*r | r/l[C 22] | r | r [ɾ]; rr [r][C 2] |
r | ir | r | |||||||||
*y | y [j] | j [j] | gj [ɟ]; ∅ |
∅ | y [j] | z [zd] > [z] / h; ∅ [C 2] |
?i | i [j]; ∅ [C 2] |
∅ | j | y | ||||
*w | v [ʋ] | v [w] | v | v [ʋ] | v | g / w | w | w > h / ∅ | i | u [w > v] | f | -∅- | w | ||
PIE | Skr. | Av. | O.C.S. | Lith. | Alb. | Arm. | Hitt. | Toch. | Greek | Greek+/j/ | Latin | Old Irish | Gothic | English |
Notes for table 1:
- A capital C stands for consonant in this table
- Between vowels
- Before an original h₂.
- Before a consonant or original laryngeal.
- After a vowel.
- Following an unstressed vowel (Verner's law).
- Before a (PIE) front vowel (*i, *e).
- Before a sonorant.
- Before secondary (post-PIE) front-vowels.
- Before or after a (PIE) u
- Before or after a (PIE) rounded vowel (*u, *o).
- In Younger Avestan, after a vowel.
- After n.
- At the beginning of a word
- At the end of a word.
- Between vowels, or between a vowel and r, l (on either side)
- After u, r or before r, l.
- Before a stressed vowel
- Before or after an obstruent (p, t, k, etc.; s)
- Before or after a resonant (r, l, m, n).
- After r, u, k, i (Ruki sound law).
- rare
Consonant clusters
Proto-Indo-European also had numerous consonant clusters, such as *st, *ḱs. In most cases in most languages, each consonant in a cluster develops according to the normal development given in the table above. Many consonant clusters however also show special developments in multiple languages. Some of these are given by the following table (with cases of otherwise predictable development in gray):
PIE | Indo-Iranian | Balto-Slavic | Alb. | Arm. | Anatol. | Toch. | Greek | Italic | Celtic | Germanic | ||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Skr. | Av. | O.C.S. | Lith. | Hitt. | Latin | Old Irish | Gothic | English | ||||||
normal | C+[j] | |||||||||||||
*sr | sr | r | str | sr | rr [r] | (a)r | š(ša)r ʃʃr | rh | n/a | fr-, -br- | sr | str | str | |
*tw | tv | θβ | tv | tv | t | kʿ [kʰ] | ttu, ddu | s-, -ss- | n/a | p? | t | þw | thw | |
*dʰw | dhv | ðβ | dv | dv | d | ? | tu, du | f | d | dw | dw | |||
*dw | dv | (e)rk | tu, du | b | tw | tw | ||||||||
*dl | ll | |||||||||||||
*dn | nn, ṇṇ | |||||||||||||
*ḱw | śv | sp | sv | šv | s | sk, š | n/a | qu [kʷ] | cu [kʷ] | ƕ [xʷ] | wh | |||
*ǵʰw | hv | zv | žv | z | ||||||||||
*ǵw | jv | q [kʷ] | qu [kw] | |||||||||||
*sw | sv | xuu [xʷ] | sv | sv | v | kʿ [kʰ] | normal dev. | h | n/a | su [sw] | s | sw | sw | |
*sp | sp | sp [CC 1] | sp | f | sp [CC 1] / pʿ [pʰ] |
normal dev. | sp [CC 1] | ? | sp | f | sp [CC 2] | |||
*sd | d; ḷ[CC 3] > ḍ, ḍḍ | zd | zd | zd | st | d | -t- [d] | st | st | |||||
*sdʰ | dh; ḷh[CC 3] > ḍh, ḍḍh | zd | zd | zd | sth | -t- [d] | zd | d | ||||||
*st | st; ṣṭ[CC 3] | st [CC 1] | st | sht [ʃt] | st [CC 1] | normal dev. | st [CC 1] | s; tt/ss[CC 4] | st | st [CC 2] | ||||
*sḱ | ch [t͡ɕʰ]; cch[CC 4] | s? | sk | š? | h | ?? č`; c`[CC 4] | normal dev. | sk; [CC 1] kh [kʰ];[CC 5] skh [skʰ] [CC 6] |
sc [sk] | sc [sk] | sk [CC 2] | sh [ʃ] | ||
*sk | sk, ch[CC 7] | sk, sč [CC 1] | normal dev. | ? | sk [CC 1] | normal dev. | ||||||||
*skʷ | norm. | squ [skʷ] | sq [CC 2] | |||||||||||
*t+t [tst] | tt; tth [CC 8] | st; sθ[CC 9]? | st | s | s? | zt, zzašt, zzazz [tst] | ss? | st | ss | ss / st | ||||
*sǵ | jj | |||||||||||||
*sgʰ | jj[CC 7] | |||||||||||||
*dt | tt | st | st | st | s | ss | ss [s] | |||||||
*ddʰ | ddʰ | zd | d | t | ||||||||||
*dʰt | ddʰ | zd, st | st | st | ||||||||||
*pt | pt | ft | t? | pt | t | pt | pt | cht [xt] | ft [CC 2] | |||||
*ḱt | ṣṭ [ʂʈ] | št [ʃt] | st | št [ʃt] | kt | ct [kt] | ht [CC 2] | ght [t] [CC 2] | ||||||
*kt | kt | xt | t? | kt | ||||||||||
*kʷt | pt | ct [kt] | ||||||||||||
*ps | ps | ps | ps | s, ss | fs | ps | ||||||||
*ts | ts | |||||||||||||
*ḱs | kṣ | š | sh [ʃ] | ks | x [ks] | hs | x [ks] | |||||||
*ks | kṣ[CC 3] | xš[CC 3] | (ks) | |||||||||||
*kʷs | kʷs | ps | x [ks] | |||||||||||
*ǵʰs | kṣ | |||||||||||||
*gs | kṣ | |||||||||||||
*gʷs | kṣ | |||||||||||||
*gʰs | kṣ | |||||||||||||
*gʷʰs | kṣ | |||||||||||||
*tḱ | kṣ | š | k | kt | s | |||||||||
*tk | kṣ | xš | kt | |||||||||||
*dʰgʰ | kṣ | |||||||||||||
*dʰgʷʰ | kṣ | γž | phth | s | ||||||||||
*dʰǵʰ | kṣ | z | ž | tk/k | khth | |||||||||
Notes to Table 2:
- In contrast to *s normally giving h in Avestan, Armenian and Greek.
- Germanic spirant law: *p, *t, *k, *kʷ remain stops when preceded by another stop or *s.
- After r, u, k, i (Ruki sound law).
- Between vowels.
- After r, l, m, n, t, d, possibly other consonants?
- After (Greek) th causes aspiration in the cluster and then disappears, typical example *πάθ-σκ-ω (*páth-sk-ō) gives πάσχω (páskhō).
- Before front vowels.
- Before an original laryngeal.
- Before a consonant or original laryngeal.
Vowels and syllabic consonants
This table shows the Proto-Indo-European vowels and syllabic consonants (as reconstructed both before and after the acceptance of laryngeal theory), and their reflexes in selected Indo-European daughter languages. Background and further details can be found in various related articles, including Proto-Indo-European phonology, the articles on the various sound laws referred to in the introduction, and the articles on the various IE proto-languages, language groups and language phonologies.
Trad. PIE | Laryng. PIE | Skr. | Av. | O.C.S. | Lith. | Arm. | Alb. | Toch. | Hitt. | Greek | Latin[V 1] | Proto-Celtic | Gothic[V 2] | Old English[V 1] | |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
normal | umlauted[V 3] | ||||||||||||||
*e | *e, *h₁e | a | e | je, ie, e, i; ja[V 4] | ä | e, ʔe, i | e | i; aí [ɛ][V 5] | e; eo[V 6] | i; ie[V 6] | |||||
*a | (*a[V 7]), *h₂e | o | a | a | ha, a | ā | ha, a | a | a | æ; a;[V 8] ea[V 6] | e; ie[V 6] | ||||
*o | *h₃e | o, a | a | a, e | a | o | |||||||||
*o | a; ā[V 9] | ||||||||||||||
*ə[V 10] | *h₁[V 10] | i | i, ∅ | ∅ | a, ∅ | ā | ∅ | e | a | a, ∅ | |||||
*h₂[V 10] | ḫ | a | |||||||||||||
*h₃[V 10] | ∅ | o | |||||||||||||
*- | *h₁-[V 11] | ∅ | e (a?) | ∅ | a, ∅ | e (o) | ∅ | ||||||||
*h₂-[V 11] | a | ḫa | a | ||||||||||||
*h₃-[V 11] | a, ha | o | |||||||||||||
*ē | *ē, *eh₁ | ā | ě | ė | i | o, ua | a/e?; ā?[V 12] | ē, e, i | ē | ī | ē | ǣ | |||
*ā | (*ā[V 7]), *eh₂ | a | o [oː] | a | a/o? | a, ah | ā > ē[V 13] | ā | ā | ō | ē | ||||
*ō | *ō, *eh₃ | uo | u | e | a/ā?; ū?[V 12] | a, ā | ō | ā; ū[V 12] | |||||||
*i | *i | i | ь | i | i | i; e[V 14] | ä | i, ī | i | i; aí [ɛ][V 5] | i | ||||
*ī | *ih₁ | ī | i | y [iː] | i | i | ī | ī | ei [iː] | ī | |||||
*ih₂ | i or (j)a?[V 15] |
yā | ī or (j)ā?[V 15] | ||||||||||||
*ih₃ | ī or (j)ō?[V 15] | ||||||||||||||
*ei | *ei, *h₁ei | ai > ē | ai > ōi, āi > aē[V 9] |
ei; ie[V 16] | i | e | ē | ei | ī | ē | |||||
*oi | *oi, *h₃ei | ě | ai; ie[V 16] | e | e, ai | ē, ai | oi | ū | oi | ái | ā | ǣ | |||
*ai | (*ai[V 7]), *h₂ei | ai | ae | ai | |||||||||||
*ēi | *ēi | āi > ai; ā[V 12] | āi; ā(i)[V 12] | i | i | ē | ēi | ī? | ei [iː] | ī | |||||
*ōi | *ōi (*oei) | y; u[V 12] | ai; ui[V 12] | e, ai | ai | ōi | ō | u[V 12] | ái | ā | ǣ | ||||
*āi | *eh₂ei | ě | ai | āi > ēi[V 13] | ae | ||||||||||
*u | *u | u | ъ | u | u | u; y[V 17] | ä | u | u | u | u; o[V 18] | u; aú [ɔ][V 5] | u; o[V 19] | y | |
*ū | *uh₁ | ū | y | ū | y; i[V 12] | u | ū | ū | ȳ | ||||||
*uh₂ | u or (w)a?[V 15] |
wā | ū or (w)ā?[V 15] | ||||||||||||
*uh₃ | ū or (w)ō?[V 15] | ||||||||||||||
*eu | *eu, *h₁eu | au > ō | ə̄u; ao[V 9] | ju | iau | oy | e | u | eu | ū | ou | iu | ēo | īe | |
*ou | *ou,*h₃eu | u | au | a | o, au | ou | áu | ēa | |||||||
*au | (*au[V 7]), *h₂eu | aw | au | au | |||||||||||
*ēu | *ēu | āu > au | āu | u | iau | e | ū? | iu | ēo | ||||||
*ōu | *ōu | a | au | ō | áu | ēa | |||||||||
*m̥ | *m̥ | a | ę | im̃; um̃[V 20] | am | a | äm | a, un | a | em | em, am | um | um | ym | |
*m̥̅ | *mh₁ | ā | ìm; ùm[V 20] | ama | mā | mē | mā | ||||||||
*mh₂ | mā > mē[V 13] | ||||||||||||||
*mh₃ | mō | ||||||||||||||
*m̥m | *m̥m | am | ьm/ъm | im; um[V 20] | am | am | em | am | |||||||
*n̥ | *n̥ | a | ę | iñ; uñ[V 20] | an | än | an | a | en | en, an | un | un | yn | ||
*n̥̄ | *nh₁ | ā | ìn; ùn[V 20] | ana | nā | nē | nā | ||||||||
*nh₂ | nā > nē [V 13] | ||||||||||||||
*nh₃ | nō | ||||||||||||||
*n̥n | *n̥n | an | ьn/ъn | iñ; uñ[V 20] | an | an | en | an | |||||||
*l̥ | *l̥ | r̥ | ərə | lь/lъ | il̃; ul̃[V 20] | al | il, li; ul, lu | äl | al | la | ol | li;[V 21] al | ul | ul; ol[V 19] | yl |
*l̥̄ | *lh₁ | īr; ūr[V 22] | arə | ìl; ùl[V 20] | ala | al | lā | lē | lā | ||||||
*lh₂ | lā > lē[V 13] | ||||||||||||||
*lh₃ | lō | ||||||||||||||
*l̥l | *l̥l | ir; ur[V 22] | ar | ьl/ъl | il; ul[V 20] | al, la | al | el | al | ||||||
*r̥ | *r̥ | r̥ | ərə | rь/rъ | ir̃; ur̃[V 20] | ar | ir, ri; ur, ru | är | ar, ur | ra | or | ri;[V 21] ar | aúr [ɔr] | ur; or[V 19] | yr |
*r̥̄ | *rh₁ | īr; ūr[V 22] | arə | ìr; ùr[V 20] | ara | ra | rā | rē | rā | ||||||
*rh₂ | rā > rē[V 13] | ||||||||||||||
*rh₃ | rō | ||||||||||||||
*r̥r | *r̥r | ir; ur[V 22] | ar | ьr/ъr | ir; ur[V 20] | ar | ar | ar | ar | ||||||
Trad. PIE | Laryng. PIE | Skr. | Av. | O.C.S. | Lith. | Arm. | Alb. | Toch. | Hitt. | Greek | Latin[V 1] | Proto-Celtic | Gothic[V 2] | normal | umlauted[V 3] |
Old English[V 1] |
Notes:
- In initial syllables only.
- In non-final syllables only.
- Before i, ī, or /j/ in the next syllable in Proto-Germanic (i-umlaut).
- In a closed syllable.
- Before r, h. Gothic, but not other Germanic languages, merges /e/ and /i/.
- Before h, w, or before r, l plus a consonant ("breaking").
- The existence of PIE non-allophonic a is disputed.
- Before a back vowel in the next syllable (a restoration).
- In open syllables (Brugmann's law).
- Between consonants, or at the end of a word after a consonant.
- At the beginning of a word, followed by a consonant.
- In a final syllable.
- ā > ē in Attic and Ionic dialects only.
- Before ā in the following syllable.
- The so-called breaking is disputed (typical examples are *proti-h₃kʷo- > Ved. prátīkam ~ Gk. πρόσωπον; *gʷih₃u̯o- > Ved. jīvá- ~ Arm. keank‘, Gk. ζωός; *duh₂ro- > Ved. dūrá- ~ Arm. erkar, Gk. δηρός)
- Under stress.
- Before i in the following syllable.
- Before wa.
- Before a non-high vowel in the next syllable (a-mutation).
- In the neighbourhood of labiovelars.
- Before a stop or m.
- In the neighbourhood of labials.
Examples
See the list of Proto-Indo-European roots hosted at Wiktionary.
*p
- Vedic Sanskrit: pā́t, pád-
- Avestan: paδa, pāδa
- Old Church Slavonic: pěšĭ, "on foot"
- Albanian: poshtë, "below"
- Lithuanian: pėda, "foot bottom"
- Armenian: otn
- Tocharian: A pe, B pai
- Luwian: pa-da-, pa-ta-
- Ancient Greek: poús, podós
- Latin: pēs, pedis
- Celtiberian: ozas
- Gothic: fotus (*p -> f by Grimm's Law)
*t
- Vedic Sanskrit: tráyas
- Avestan: θrāiiō
- Old Church Slavonic: trĭje
- Lithuanian: trỹs
- Albanian: tre
- Ancient Greek: treĩs
- Latin: trēs
- Old Irish: trí
- Welsh: tri
- Armenian: erekʿ
- Gothic: þreis (*t -> þ by Grimm's Law)
*ḱ
*ḱm̥tóm, "hundred" (from earlier *dḱm̥tóm)[1]
- Vedic Sanskrit: śatám
- Younger Avestan: satəm
- Old Church Slavonic: sŭto
- Lithuanian: šim̃tas
- Tocharian: A känt, B känte
- Ancient Greek: hekatón
- Latin: centum
- Old Irish: cét
- Welsh: cant
- Gothic: hund (from Proto-Germanic *hund-)[2]
*k
- Vedic Sanskrit: kravís-, "raw meat"
- Lithuanian: kraũjas, "blood"
- Old Church Slavonic: kry, "blood"
- Ancient Greek: kréas, "meat"
- Latin: cruor, "raw blood"
- Old Irish: crú, "blood, gore"
- Old English: hrǣw, "raw"
Sound laws within PIE
A few phonological laws can be reconstructed that may have been effective prior to the final breakup of PIE by internal reconstruction.
- Sievers' law (Edgerton's law, Lindeman's option)
- Bartholomae's law
- Szemerényi's law
- Stang's law
- Siebs' law
See also
Further reading
- "Voiceless high vowels and syncope in older Indo-European" (PDF). Martin Kümmel, department of Indo-European linguistics, University of Jena.
- "Uvular Stops or a Glottal Fricative? Theory and Data in Recent Reconstructions of PIE "Laryngeals"" (PDF). Martin Kümmel, department of Indo-European linguistics, University of Jena.
References
- Meier-Brügger, Michael; Gertmenian, Charles (translator) (2003). Indo-European linguistics. Berlin [u.a.]: de Gruyter. pp. 101–131. ISBN 3-11-017433-2.
{{cite book}}
:|first2=
has generic name (help) - Hock, Hans Heinrich (1986). Principles of historical linguistics. Berlin; New York: Walter de Gruyter. p. 132. ISBN 3-11-010600-0.