Andrew Byrd reading Schleicher's Fable

« previous post | next post »

Eric Powell, "Telling Tales in Proto-Indo-European", Archaeology Magazine:

In 1868, German linguist August Schleicher used reconstructed Proto-Indo-European vocabulary to create a fable in order to hear some approximation of PIE. Called “The Sheep and the Horses,” and also known today as Schleicher’s Fable, the short parable tells the story of a shorn sheep who encounters a group of unpleasant horses. As linguists have continued to discover more about PIE, this sonic experiment continues and the fable is periodically updated to reflect the most current understanding of how this extinct language would have sounded when it was spoken some six thousand years ago. Since there is considerable disagreement among scholars about PIE, no one version can be considered definitive. Here, University of Kentucky linguist Andrew Byrd recites his version of the fable using pronunciation informed by the latest insights into reconstructed PIE.

You may want to compare this version with the one from the movie Prometheus.

The Archeology article also include Andrew Byrd reciting "his version of the 'The King and the God' in PIE, based on the work of linguists Eric Hamp and the late Subhadra Kumar Sen".


  1. William Steed said,

    September 23, 2013 @ 7:29 pm

    It says a lot about accent that it's clear that Andrew Byrd is a native speaker of English. Here I am, listening to a (reconstructed) language I've never heard, and I can pick that he's an English speaker. Interesting.

  2. Rod Johnson said,

    September 23, 2013 @ 7:35 pm

    I don't hear any differentiation between the laryngeals. Is h1 supposed to be different from h2?

  3. Jonathan Gress-Wright said,

    September 23, 2013 @ 11:16 pm

    h1 is supposed to be a velar or fronted velar fricative, while h2 is a back velar or uvular fricative. This is probably not a distinction that comes naturally to a native English speaker, though a velar/uvular fricative contrast does occur in several languages, according to Ladefoged & Maddieson 1996. Interestingly, such a contrast is reported for some Caucasian languages, and I remember Don Ringe noting in class that the sound inventory of Proto-Indo-European was not unlike many Caucasian languages today.

    If I had to pronounce PIE, I would probably articulate h1 towards the front and h2 towards the back to be sure of getting that contrast.

  4. Jake Nelson said,

    September 24, 2013 @ 1:23 am

    Always happy to see more PIE material. Etymology and tracing cognates in other languages have always been my main linguistic interests, so PIE has naturally fascinated me. Would like to see more attempts at large-scale reconstruction.

    Regarding the laryngeals, however: there's good reason those sounds were some of the first lost in nearly all descendant languages.

  5. Colin Fine said,

    September 24, 2013 @ 2:04 am

    I think he's aspirating his unvoiced plosives in the manner of English. IIRC IE didn't have a distinctset of aspirated unvoiced plosives, so he's not confusing phonemes as somebody would who did that in classical Greek. But I think he would sound less English if he didn't do it.

  6. Belles Lettres said,

    September 24, 2013 @ 6:59 am

    The prosody is (like always) the real problem. The speaker simply puts the stress on those syllables which have an accent sign on the top of the vowel in the reconstructed word, as every modern European speaker would do. As a consequence, the quantities of the vowels are lengthened as we do today in English and German. If the prosody is not right, the hole grammar is collapsing in PIE.

    The laryngeals are not so important. You can get them right by colouring the vowels towards a, e, o automatically as German does in 'ich' versus 'ach'. In CHC-positions like ph2ter it is harder to get the schwa right.

  7. Bill Benzon said,

    September 24, 2013 @ 7:58 am

    So, you get yourself a helicopter with some good loudspeakers (like in the famous air assault scene from Apocalypse Now), and broadest those stories to the folks below. What do you call it? Pie in the sky.

  8. GeorgeW said,

    September 24, 2013 @ 9:34 am

    @Bill Benzon :)

  9. Rube said,

    September 24, 2013 @ 9:50 am

    I suppose that, in terms of good uses of a time machine, this wouldn't be up there with killing Hitler, but I'd love it if we could go back and find out if Proto-Europeans could make head or tail of this.

  10. dw said,

    September 24, 2013 @ 12:04 pm

    His voiced aspirates don't sound voiced to me. They are pretty tough buggers to pronounce, unless you have them in your native tongue. I'd love to hear a native speaker of an Indic language recite this passage.

  11. dw said,

    September 24, 2013 @ 12:30 pm

    He sounds more British than American to me, mainly because of his short o, which sounds like a typical Southern England LOT vowel.

  12. Nick Zair said,

    September 24, 2013 @ 12:40 pm

    @ Jonathan Gress- Wright
    (Hi Jonathan.) Depends who's doing the supposing. There have been so many poposals over the years. Martin Kuemmel in Konsonantenwandel. Bausteine zu einer Typologie des Lautwandels und ihre Konsequenzen für die vergleichende Rekonstruktion. Wiesbaden: Dr Ludwig Reichert Verlag, 2007, 327-336 has rather a good discusson and plumps for h1 [h], h2 voiceless uvular fricative (and h3 as a voiced uvular fricative, although in my view the evidence for voicing of h3 is very limited).

  13. Pete said,

    September 24, 2013 @ 1:09 pm

    William Steed/dw: He should have got a Lithuanian peasant to read it.

  14. Jonathan Gress-Wright said,

    September 24, 2013 @ 1:17 pm

    Hi Nick,

    I can see h1 as [h]. It would explain why it doesn't survive even in Anatolian, and why it doesn't color adjacent /e/. I'd be interested to hear how Kuemmel explains the reflex in Greek, where Ch1C > CeC. ChC > CeC could make sense if [e] was the epenthetic vowel in that particular dialect of PIE or pre-Greek, though usually epenthetic vowels are [high].

  15. Sally Thomason said,

    September 24, 2013 @ 2:32 pm

    About the phonetics of the Proto-Indo-European laryngeals — positing fricatives, much less voiceless fricatives, makes it hard to account for the morphophonemic patterns in which the laryngeals behaved like (voiced) resonant consonants. They behaved a lot like the pharyngeal resonant consonants in some Salishan languages — vocalizing when not next to a vowel; disappearing (later on) with compensatory lengthening of a preceding vowel; etc. In relevant Salishan languages, for instance Montana Salish, the plain pharyngeal causes lowering of a neighboring unrounded vowel, e.g. e to a, and vocalizes to [a] in unstressed syllables (when no full vowel occurs because most vowels get deleted in unstressed syllables); and the labialized pharyngeal causes lowering of a neighboring vowel u to o and vocalizes to [o] in unstressed syllables. It also causes rounding of neighboring vowels sometimes. Salish has glottalized pharyngeals too, both plain and labialized, and they have the same effects on neighboring vowels; when they vocalize, the resulting vowel is followed by a glottal stop. This last phenomenon, as far as I know, is unlike anything found in IE.

  16. Nick Zair said,

    September 24, 2013 @ 3:03 pm

    Sally (if I may),
    – part of the reason (some – plenty don't, and I don't remember if Kuemmel in particular makes this argument) people tend to think of them as fricatives is because they behave very much like PIE *s in having great freedom to appear more-or-less anywhere in the root, which otherwise tends to follow a rather strict sonority hierarchy.

    – I don't know anything about Salish – the developments sound very interesting and indeed very similar – but I think that mainstream PIE-ists are loth to accept vocalisation of laryngeals (as opposed to epenthetic vowels followed by loss of laryngeals) because there is already a group of sonorants (*r, l, m, n) which are syllabic when surounded by non-syllabic consonants. And if a laryngeal is next to them, it is clearly the sonorant which syllabifies, not the laryngeal. So if we want syllabic laryngeals, we have to have two rounds of syllabification, which I guess is seen as a bit inelegant, although some people have accepted it. There is also some evidence that in *CVCHC- clusters, which develop to CVCiC- in Vedic, the first syllable scans heavy, i.e. as if the laryngeal were still there, which has been taken to be evidence for epenthesis rather than vocalisation (although it could be seen as similar to the glottal stop creation of Salish, I suppose).

    Personally, I'm pretty sceptical about attempts to reconstruct the phonetics of the laryngeals very precisely (or at least I don't think it's very important for the kind of IE stuff I do, when I do IE stuff).

  17. Rubrick said,

    September 24, 2013 @ 11:37 pm

    Is there a Geekiest Comment Thread of the Year award?

    Man I love this blog.

  18. mollymooly said,

    September 27, 2013 @ 4:54 pm


    Rube mentioned it once but I think he got away with it all right.

RSS feed for comments on this post