10 Jun 2009

Dialectal loss of PIE voiced aspirated stops via Para-MIE dialect merger?

This is too big an idea for me to translate into words without writing a 20-page essay. I think I can spare you readers most of my inevitable pedantics with the following animated picture of what's on my mind lately, shown here in six basic steps as a colourful summary (click on image to see it animate):

My new brainstorm is a masala of my previous online thought experiments (cf. Proto-Semitic as a second language and Winter's Law in Balto-Slavic, "Hybrid Theory" and phonation - Part 2) that points me to a location of the Mid IE stage centered in the Balkans, rather than in the NW Pontic (surely the location of later PIE, by the way), and the most recent input from savvy reader Kiwehtin (read here) concerning a means of explaining PIE 'stop harmony' using breathy vowels as a vehicle for the phenomenon.

Having not thought deeply about PIE's curious phonotactic constraint that barred the tautosyllabic cooccurence of both a voiced aspirated stop such as *dh with a voiceless stop such as *t in a root, I've had no good explanation for it up to now. That is quite lazy of me and as I learn more I realize that every detail is a cerebral gem unto itself. I can only lament in futility that the day is not 48 hours long.

To the point, now. Kiwehtin (Christopher Miller) has brought up the issue of the possibility of breathy vowels in some stage of Pre-IE or PIE itself and while I would personally hesitate to reconstruct it for the finalmost PIE stage, I have to admit that breathy vowels solve the source of stop harmony without too much fuss. In fact, upon looking up 'stop harmony' as a keyword search, I found this gem from Google Books that precisely gives us a real-life language with the same phenomenon called Jabêm (Lynch/Ross/Crowley, The Oceanic Languages (2002), p.274)! Neato!

If it seems like I've been gone for a while, it's partly because this piece of the puzzle has been co-mingling with other thoughts in my brain that together were weaving a crazy tapestry of detailed mental images that has been fascinating me for days. Namely, if we combine the idea that Mid IE arose out of the Balkans and spread to the North-West Pontic from which the next stage of the language in turn spread outwards, and if Kiwehtin is pointing to a delicious possibility that breathy vowels had developped in some stage of Pre-IE (say, the middle of the Late IE period), could we perhaps further entertain ourselves with a wild speculation that as the Late IE dialect began to form from the NW Pontic, it percolated through "para-MIE dialects" (see graphic above) situated around it and caused some of the familiar features of PIE dialects such as those of Anatolian and Tocharian that both coincidentally merged the breathy stops into modal ones?

I know this is a flamboyant thought, but stay with me for a moment, readers. For this idea to work, I would have to presume that PIE stop harmony is a common artifact of breathy vowels in a shared past, regardless of dialect. In other words, breathy stops must have formed much earlier, regardless of what I previously suggested about some dialects never forming breathy stops at all in my 'Hybrid' proposal. Furthermore, given my views on MIE syllable structure which lacks consonant clustering altogether, I can only conclude that any such stop harmony due to breathy vowels is likeliest to have developped in the Late IE period. (Consider as an example *dheubh- that would have been two syllables prior to the event of Syncope occurring in the earliest point of Late IE.)

So, presuming at least a momentary adoption of breathy vowels within the middle of the Late IE period to induce stop harmony, it would then be a perfect time for Late IE to start spreading outwards to form the dialect area familiar to Indoeuropeanists. However, if Mid IE was in the Balkans originally (to best explain apparent Semitic areal influence) and Late IE spread from the NW Pontic, then it seems logically inevitable that the new dialects forged from Late IE innovations would collide back into the sister dialects formed from the bygone Mid IE period which lacked such innovations, i.e. my so-called "para-MIE" dialects.

To add, if breathy vowels only formed in Late IE, we should expect that living speakers of such theoretical paradialects were lacking these sounds in their speech inventory and therefore found it a little difficult to pronounce them when Late IE dialects began to popularize in these former para-MIE territories. As regional bilingualism would eventually lead to a single dominant dialect, it seems to me that this would produce new Late IE dialects in those regions located outside of the "Late IE epicenter" within which breathy stops or vowels were replaced with locally more familiar modal phonation (ie. substratal influence). Ergo, voiced aspirated stops would return to plain voiced stops, yet at a price. The price being an increased likelihood of merger of former voiced aspirated stops like *dh (now modal again) and *d (creaky) due to a reduction in perceptual distinction between the two series.

In a nutshell, I'm suggesting that the merger of plain voiced and voiced aspirated stops may in effect be the result of a clash between two different stages of the same language, in a manner of speaking, by way of the preservation of archaicisms in surrounding para-IE dialects which surely existed but which are otherwise undetectable to the historian or archaeolinguist. Insanely complex? Sure. Nifty? You bet! All I can hope is that I explained my odd mental musings well enough for at least one other person on this planet to understand. Lol.


  1. The "slide show" is fantastic - much clearer and instantly comprehensible than any attempt to describe your hypothesis using only words could have been. Also more memorable for most people, because of the simple visuals.

  2. Thanks. They say a picture is worth a thousand words. An animated gif is worth ten thousand more. :o) If I can solve my "flash program storage problem", my slideshows would be a lot more fantastic but so far Web 2.0 is really disappointing me bigtime.

  3. Interesting thoughts. By the way, for some reason, the animation did not play for me on your blog page, but *did* play when I clicked on it and it opened in a new window... Don't know why...

    You would probably be interested in this information on Central Igbo breathy/aspirated consonant harmony by Rose-Juliet Anyanwu:


    The abstract is from a conference several years ago:


    It could be very interesting to follow up on what corresponds to harmonic morphemes in other Igbo varieties; in other words, what CVC sequences correspond to harmonic CVC sequences in Central Igbo, and plausibly correspond to a pre-harmonic stage of this variety. This might give some useful clues to possible pre-harmonic root structure in early PIE or pre-PIE that might help explain any apparent exceptions to the generalisation of C-harmony in voiced/aspirate C roots.

    I haven't seen any reference to anywhere Anyanwu's paper might have been published subsequently. I have to say I'm not as skeptical as she is about the possibility that nasal vowels might be an effect of breathiness on the consonants given the relatively widespread phenomenon of "rhinoglottophilia". That said, RGP can work in either direction. I can imagine, for example, that in Central Igbo, originally breathy vowels could have subsequently evolved into nasal vowels, after either being engendered by breathy consonants or spreading their breathiness onto the consonants.