User:Kisseran/Proto-Indo-European phonology

Labial Coronal Palatal Velar Labiovelar Laryngeal
Stop Voiceless *p *t *ḱ *k *kʷ
Voiced (*b) *d *g *gʷ
Voiced aspirate *bʰ *dʰ *ǵʰ *gʰ *gʷʰ
Fricative *s *h₁, *h₂, *h₃
Nasal *m *n
Liquid *r, *l
Glide *w *y

Stops

edit

Stops in Proto-Indo-European are generally reconstructed as follows:[1]

Labial Dental Palatal Velar Labiovelar
Voiceless *p *t *ḱ *k *kʷ
Voiced (*b) *d *g *gʷ
Voiced aspirate *bʰ *dʰ *ǵʰ *gʰ *gʷʰ

The status of *b is controversial. If it does exist, it only exists in an extremely limited number of roots. Putative examples include *dʰewb- ('deep'), as in English deep and Old Church Slavonic дъно (dŭno, 'bottom'), and *h₂ebōl ('apple'), as in English apple and Old Irish ubull.[2]

Thorn clusters

edit

Glottalic theory

edit

Centum and satem

edit

The descendants of Proto-Indo-European have traditionally been divided into two camps based on sound changes to the three velar series: the centum group and the satem group. The centum languages, named for the Latin term for 'one hundred', are those in which the palatovelars (* *ǵ *ǵʰ) merged with the plain velars (*k *g *). These include the Italic, Celtic, Germanic, Hellenic, and Tocharian subfamilies.[3] By contrast, the satem languages, named for the Avestan term for 'one hundred' (𐬯𐬀𐬙𐬆𐬨, satəm), are defined by the merger of the labiovelars (* * *gʷʰ) with the plain velars and the assibilation of the palatovelars into fricatives.[4] This category includes the Balto-Slavic and Indo-Iranian subfamilies.[5]

The status of Albanian, Armenian, and the Anatolian languages is the subject of some debate.

s-mobile

edit


Laryngeals

edit

Proto-Indo-European had a set of laryngeal consonants of debated values, represented with *h₁, *h₂, *h₃, and possibly *h₄.[6]

Sonorants

edit

Ablaut grades

edit

Vowels

edit

Only two short vowels – *e and *o – are universally accepted by linguists for Proto-Indo-European.[7] Similarly, their long vowel equivalents – *ē and *ō – are also accepted.[8] While *i and *u were present in the language, their status is not phonemic; they were allophones of *y and *w, respectively, especially in zero-grade positions.[7] The status of an *a phoneme is controversial.[7][additional citation(s) needed] Some linguists, particularly those of the Leiden school, view *a as an outgrowth of laryngeal coloring. Thus, etymologies like *sals ('salt'), with *a, are often reconstructed as *sh₂els, with laryngeal–vowel sequences.[7]

Certain etymologies are difficult to square without *a; for example Hittite 𒀀𒀊𒉺 (appa; 'behind, afterwards') and Greek ἀπό (apó; 'away from') cannot be derived from *h₂epó.[9] Alternative theories have been proposed to account for these handful of apparent counterexamples. One hypothesis is that *h₄ was an a-coloring laryngeal wherein the original consonant sound was not preserved as it would have been in Hittite. This would explain the a-vocalization and lack of a largyneal consonant at the beginning of appa.[10] Other examples include alternative etymologies for problem words, such as deriving appa from the etymon of Greek ἐπί (epí; 'on, upon') rather than of ἀπό, positing sound laws that explain the disappearance of laryngeals in Hittite, or attributing the changes to analogical leveling.[11][8] Despite some notable holdouts such as Frederik Kortlandt and Alexander Lubotsky, most mainstream Indo-European linguists accept *a and *ā as legitimate – if relatively rare – phonemes.[12][additional citation(s) needed]

There are four accepted diphthongs for Proto-Indo-European: *ey, *oy, *ew, *ow. The *a equivalents – *ay and *aw, respectively – are also sometimes included, though they may be likewise attributed to the effects of laryngeal coloring. For example, while some linguists reconstruct *kaykos for 'blind', the Leiden school argues that the term should be reconstructed as *kh₂eykos with the a-coloring laryngeal abutting the *ey diphthong.[8]

In cases where two vowels bordered each other and did not become diphthongs, the second vowel was typically lost and the first vowel lengthened to compensate for the loss. Examples of this include the suffix *-ōs, which derived from the earlier form *-o-es.[8][a] This also occurred in some instances where a vowel abutted a diphthong; the first element of the diphthong was deleted and the diphthong was made long with the abutting vowel, as in the suffix *-ōy from an earlier form *-o-ey.[8]

Accent

edit

Phonotactics

edit

Development

edit

History

edit

Neogrammarians

edit
Photograph of Karl Brugmann
Karl Brugmann's Grundriß was the authoritative Neogrammarian treatment of Proto-Indo-European phonology.

Early 20th century

edit

Typological turn

edit

Generative approaches

edit

See also

edit

References

edit

Notes

edit
  1. This process is sometimes known as Wackernagel's law.[13]

Citations

edit
  1. Kapović 2008, p. 135.
  2. Kapović 2008, p. 137.
  3. Trask 2000, p. 53.
  4. Trask 2000, p. 297.
  5. Zair 2012, p. 3.
  6. 1 2 3 4 Byrd 2018, p. 2057.
  7. 1 2 3 4 5 Byrd 2018, p. 2058.
  8. Byrd 2018, pp. 2057–2058.
  9. Zair 2012, pp. 3–4.
  10. Zair 2012, p. 4.
  11. Zair 2012, pp. 10–11.
  12. Trask 2000, p. 365.

Sources

edit