1. Vowels
  2. Consonants



PIE-0 had a five-vowel system. Quantitative differences played no distinctive (i.e. phonological) role. The vowels were:

i - - - - u

e - - - - o


Examples of these vowels are:


The mid and low vowels could combine with the glides y and w, to form the diphthongs ey, ew, oy, ow, ay and aw. Examples:


In some unstressed contexts, namely in the neighbourhood of liquids, nasals and laryngeals, the reduced vowel ə could be present. Some examples, next to the different possible consonants:



Each consonant in PIE-0 can be placed in a slot in a 9 x 5 matrix, places of articulation along one axis and type of consonant along the other. Some slots in the matrix are, however, empty, while others contain very rare phonemes (within normal brackets in the table) or sounds without phonemic status (in square brackets). Here is the full matrix:

  Labial Dental / Apical Velar / Postvelar Labiovelar Glottal
Voiceless stops p t k kw  
Voiced stops (b) d g gw  
Voiced aspirated stops bh dh gh gwh  
Voiceless fricatives   s x xw (h)
Voiced fricatives   [z] ɣ ɣw  
Nasal consonants m n [ŋ]    
Lateral consonant     l    
Trill consonant     r    
Semivowels w   y    


PIE-0 had four voiceless stops: p t k kw. These were all very frequent, and could occur at the beginning, middle or end of a syllable. Phonetically, they were simple, unaspirated, voiceless stops. Examples:


PIE-0 had four voiced stops: b d g gw (the fifth, the glottal stop ', will be considered under the Fricatives). However, the voiced bilabial stop b was very rare. It has been suggested that the rarity of this phoneme, as well as certain restrictions on the co-occurrence of stops, can be best understood if the series of voiced stops is re-interpreted as glottalized (or ejective) stops, similar to those found, for example, in Georgian, Hausa and Quechua. Without categorically rejecting this hypothesis, I do not think that it is necessary to use different symbols in what is, essentially, a phonemic treatment.

Voiced stops occurred in somewhat more restricted environments than voiceless stops: they did not normally occur before other stops or fricatives (except across morpheme boundaries, where they may have developed by forward assimilation to another voiced consonant). Examples of occurrence:


Although this third series of stops is firmly established through the comparative method, it has become fashionable to doubt its existence as such. The main reason for this is that languages that have voiced but no voiceless aspirate stops are thought to be "typologically unusual". Let us note, however, that Sanskrit, one of the most important languages used for the reconstruction of PIE, had a clear preponderance of voiced over voiceless aspirate stops. It is thus clearly possible to have a sound system in which voiced aspirates are very common, while their voiceless counterparts are marginal at best.

I, for one, will keep the voiced aspirates and leave it for later speculation as to the original nature of these consonants. Examples of their occurrence in PIE-0:


The classic reconstruction of PIE allowed for a single fricative: that of the dental-alveolar *s. With the emergence of the laryngeal theory it became evident that other fricatives might have existed in PIE-0, all of them pronounced in the velar or glottal region. Different linguists have developed different sets of "laryngeals", while some have stuck to algebraic formulations, claiming that it is not possible to reconstruct the exact nature of these consonants. I shall stick my neck out, and reconstruct five laryngeal fricatives (*x *xw *h *ɣ *ɣw), of which one (*h) is rare. For convenience, I shall include the glottal stop (*' ) here as well, as it corresponds to one of the usually reconstructed laryngeals.

In the list of examples below, I shall also include (between slashes) the corresponding notation H1, H2 or H3, as presented in standard works on laryngeal theory (such as Lindeman (1987)).


The widespread occurrence of forward assimilation in IE languages suggests that /n/ = [ŋ] when immediately followed by a velar or labiovelar consonant: *yuwənkos = [juwəŋkos] 'young', *penkwe = [peŋkwe] 'five'.



N.B. The notation of the second semivowel as *y, rather than +j, is a conventional one in IE studies. It recalls usage in English/French/Spanish and the transcription of Sanskrit, rather than in German/Scandinavian/East European languages.

Copyright Gabor Sandi 1998-2019

Disclaimer: Everything on this site was written and/or prepared at the author's discretion. The author has tried to be accurate where facts are presented, but this does not mean that these facts are necessarily accurate. People needing more reliable information should refer to appropriate sources presented, on the Internet or elsewhere, by authoritative professional or academic bodies.

To get in touch with the webmaster, send an e-mail to g_sandi at

Flags on this site are from the FOTW Flags Of The World website.

Valid HTML 4.01!