Talk:Indo-European sound laws/Draft


 * I've created this page as a draft for a better sound laws page. --Nathan M. Swan (talk) 22:41, 14 December 2010 (UTC)

Stop consonants
The fifteen stop consonants reconstructed for PIE form a 3×5 matrix when sorted according to their phonation and place of articulation. For the most part, individual consonants did not evolve independantly; instead, changes affected either a phonation set or a POA set in its entirety, and an individual consonant's development would be the sum of these.

Voiceless stops
The class of voiceless stops remains in most IE languages. The major exceptions are Germanic, where they turned (by Grimm's Law) into voiceless fricatives in most contexts (with the exception of the position following another voiceless consonant, where they remained, and stressed contexts where they merged with the voiced aspirated series (Verner's Law; see below for the output)), and Armenian, where they turned into aspirated stops. Two large-scale conditional developments took place within the Indo-Iranian languages: in Indo-Aryan, sequences of voiceless stops and laryngeals coalesced to create a series of voiceless aspirated stops, while in Iranian, voiceless stops were fricativized before consonants.

Voiced stops
The class of voiced stops remains in most IE languages. They are devoiced in some groups: Anatolian, Tocharian, Germanic and Armenian, in the former two merging with the voiceless stops but in the latter two creating new voiceless stops after the shift of the original voiceless stops to new values (see above).

Voiced aspirated stops
Voiced aspirated stops survive only in the Indo-Aryan branch of the Indo-Iranian languages. In most other languages they are reflected as plain voiced consonants. In Germanic, they were additionally fricativized by Grimm's Law, creating voiced fricatives.

Greek and Italic show a different path of development: in the former these were devoiced to yield voiceless aspirated stops, in the later additionally fricativized to yield voiceless fricatives.

Anatolian and Tocharian merge these, too, into the voiceless stops. Stop phonation distinctions are thus lost entirely in these groups.

Labial stops
The class of labial stops remains unchanged in all IE languages.

Dental stops
The class of dental stops remains in all IE languages. However, in Tocharian and Greek, palatalization applied (…)

Palatovelar stops
The centum-satem isogloss. The basic Satem reflexes are (see below for exceptions):
 * II (…)
 * Plain alveoalr fricatives /s z/ in most Iranian languages, except /θ d/ in Old Persian.
 * In the Indo-Iranian languages, the voiced members merged with the reflexes of labialized velars
 * In Balto-Slavic, initially postalveolar fricatives /ʃ ʒ/, further simplified to /s z/ in Latvian and the Slavic languages.
 * In Albanian, interdental fricatives /θ ð/.
 * In Armenian, alveolar affricates /ts dz/. *ḱ yielded /s/ instead of the expected aspirated member.

Plain velar stops
The class of plain velar stops survives in all IE languages, fortified by either the palatovelar stops (in the Centum languages) or the labialized velar stops (in the Satem languages).

Secondary palatalization, preceding PIE front vowels, occurred in several Satem groups.
 * In Indo-Iranian,
 * In the Slavic languages, these were shifted to // in the so-called "first" palatalization.
 * In Albanian, these become alveolar fricatives /s z/.

Even later palatalizations have occurred in several IE subgroups, including the Romance languages (in Vulgar Latin), English (in the Old English period), Albanian, Latvian, and

*p
Spirantization of *p, but not of similar consonants, occurred in Celtic and Armenian.

*gʷ
In the Celtic languages,

*s
PIE *s, the only sibilant consonant in the language, is widely reflected as /s/, but has several conditional developments. The RUKI change (…). In the Germanic languages, it developed to *z by Verner's Law, which later merged with *r during the development of all Germanic languages other than Gothic. A similar change also occurred in Latin, but affecting all instances of intervocalic *s.

In Greek and the Iranian languages, *s preceded by a vowel generally developed to *h.

*m
PIE *m usually stayed the same, with these exceptions:
 * Word-final *m was frequently changed. It became /n/ in Lithuanian, Albanian, Tocharian, Greek, and Old Irish, like in EXAMPLE. It was retained in the Italic languages (including Latin), and the Indo-Aryan languages (including Sanskrit).
 * In Armenian, Hittite, Gothic, and English, it disappeared at the end of words, like in EXAMPLE.
 * In Old Church Slavonic, it nasalized the previous vowel and disappeared at the end of a word, like in EXAMPLE.
 * In Greek, *my had developed into a palatalized consonant *mʲ. This merged with *nʲ, and was later broken to /in/. An example is __EXAMPLE.

Liquid consonants
The Indo-Aryan languages merge *l into *r. Aside from this, liquids are largely unchanged in IE languages. A distinction between a tapped and trilled rhotic has arisen in Albanian and some scattered examples elsewhere, e.g. Iberian Romance. A distinction between plain and velarized laterals has arised in Albanian and Armenian, later on also in several other languages including French and Dutch.