Plain layout version :: Printable version

A scheme for transcribing Proto-Indo-European with ASCII characters.

For use in email and Unicode-challenged milieux I've invented my own ASCII-friendly transcription of Proto-Indo- European. It is IMHO both easily typed, and easily readable, and it is also reasonably easy to convert automatically into a Unicode transcription.

  • Long vowels are indicated by a following colon e: o: (and a: i: u: if you are of the non-'laryngealist' persuasion)1.
  • The palatal series is *kj *gj *gjh *y,
  • the labiovelar series is *kv *gv *gvh *w and
  • the 'laryngeals' are *h1 *h2 *h3 (*h4), and an indeterminate 'laryngeal' is h h0 H.2
  • Acute accent is an apostrophe ' after the vowel and circumflex accent is a ^ after the vowel.
  • Syllabic sonorants are indicated if needed with a period . before the *r *l *m *n: *.r *.l *.m *.n.
  • 'Schwa' (*ə) can, if desired, be written with *@.

It is easily typed, unambiguous and mostly decipherable on the spot for someone familiar with the traditional transcription with macrons and superscript h w ^. To be sure one could use *k^ etc. but

  • superscript w (or u̯ — superscript u with inverted breve below!) isn't as easily simulated by any of the ASCII non- alphanumerics,
  • ' ^ will be needed to indicate the PIE tone accents, and
  • it is less ambiguous not to use ^ for two different things, especially for automatic conversion.
  • Also I think *kv *kj make a nicer pair3.

Actually there were hardly any minimal pairs between labiovelar and velar + w or palatal and velar + y, but

  1. there were words like *h1ekjwos,
  2. the tradition is to keep 'secondary' articulation markers and semivowel symbols separate,
  3. the transcription *h1ekywos looks a good deal less nice than *h1ekjwos to my eye and
  4. automatic conversion is easier if different letters are used for 'diacritics' and for semivowels.

Footnotes

  1. In a 'laryngealist' transcription long vowels could probably be written with a doubled vowel symbol ee oo, since any apparent hiatus would in fact be a 'laryngeal', but the colon is safer, is less ambiguous when capital *E *A *O are used for the 'laryngeals', and also allows easier automatic conversion. [back]

  2. I'd prefer to use *E *A *O for the 'laryngeals' because they are non- committal as to their phonetic nature[^phonetics], showing only their influence or lack thereof on an adjacent *e, as was de Saussure's original intention, but that would create problems with indicating an indeterminate 'laryngeal' — although one could conceivably use *@ for that — and h4 if one believes in that. Actually I'm disturbed by *h1 *h2 *h3 with non-subscript digits, as X-SAMPA has conditioned me to see mid-word digits as phonetic symbols in their own right. [back]

  3. One could go for something more radical like *c *j *jh *y for the palatals and *q *w *wh *v for the labiovelars, but only at the price of greater ambiguity and less on-sight familiarity. [back]

To top of page