20 Aug 2007

More on thematic vowels and the voicing of Pre-IE word-final *-s

I received a very intriguing comment from Chris Miller under my previous post the other day. (His full comments can be see under Pre-IE alternating thematic vowels.) I felt that it deserved to be talked about with another blog entry since this is a complex topic. He offers a clever alternative view of things:

"Based on the data you give in your entry, I think there might be an alternative explanation, again phonetically based, albeit slightly more complex. I notice that in the data you give, /o/ appears before three kinds of consonants:
  • 1) h2 (the "a-coloring" laryngeal, often hypothesized to be a pharyngeal)
  • 2) r
  • 3) nasals m and n.
It so happens that all three of these classes of consonants often have the effect of lowering adjacent vowels cross-linguistically. [...] It should be useful to follow up this hypothesis and compare it to your original voicing-based hypothesis. "

Excellent idea. Be careful what you wish for, hehe.

I already attributed in the previous post the subjunctive 1ps *-oh₂ to an earlier Pre-IE subjunctive 1ps in *-om which, because of the semantics of certainty/uncertainty, acquired the so-called "perfect" ending. I remind everyone that only the 1ps acquired the perfect ending so it's an interesting oddity that I felt needed explanation. Also, in this period of time, the perfect would have been also doubling as a mediopassive. IEists already blame the perfect for the origin of the mediopassive so this isn't a stretch[1]. To boot, the matter of the subjunctive 1ps ending might tie in with the origin and dating of the pronoun *egoh₂ (see The origin of Indo-European ego). However, there is an added reason for my choices. Wherever else we find word-final *-h₂ with a preceding vowel, that vowel is invariably *e (pronounced [a] due to vowel colouring), and never *o. Note the ancient "animate collective" suffix *-eh₂, later becoming feminine in most but not all IE languages.

The *o in the lonely subjunctive 1ps is really a case that sticks out like a sore thumb and cries out for a special explanation as I have provided it, whereby an original *-om explains *o by way of a following voiced *m just as we find in the 1pp, *-o-mes. The co-existence of both 1ps subjunctive *-oh₂ and nominal ending *-eh₂ shows that pharyngeals aren't any better at explaining thematic vowels.

Then there is the matter of word-final *-s in Pre-IE. To me, Szemerenyi's Law suggests that in order for *-s to have disappeared the way it did (e.g. Pre-IE *dʰǵʰom-s > PIE *dʰǵʰōm, nominative form of "earth"), it would have to have been voiced. A voiced word-final allophone [z] following a voiced resonant would be far less salient and would be especially prone to erosion over and above a voiceless [s]. Immediately following vowels, *-s apparently was preserved (as in *n̥mr̥tos). The genitive demonstrative *tesyo "of that" (from the stem *to-) shows me that *s had always been voiceless in medial position, and *tod (inanimate nomino-accusative form) shows me that *-d was voiced during this Pre-IE period when alternation in thematic vowels first arose.

While the pair *tesyo/*tod can be explained by my voicing theory, pharyngeals simply cannot explain it.

Now, I didn't mention the voicing of word-final *-s in my previous post because it would only obfuscate the larger pattern I wanted to discuss. Afterall, an inquisitive mind might hit upon seeming contradictions in my own theory:
1. If *-s were voiced at the time, why don't we see 2ps non-indicative **bʰeros instead of *bʰeres then? (Because *bʰeresi reinforced *e in *bʰeres by simple analogy.)

2. If *-s were voiced at the time, why plural *-es instead of **-os? (Because of the PIE speaker's need to keep athematic nominative plural *-es distinct from singular nominative thematic *-o-s and there might have been already a distinction between Pre-IE *-es and *-əs. Keeping singular/plural contrast was at stake.)

[1] Originally proposed as far back as 1932 by both Kurylowicz (note Filip Baldi, The Foundations of Latin, p.364) and Stang. Further information on the similarities and relationship of the perfect and mediopassive conjugations may be gleaned from Benjamin Fortson's Indo-European Language and Culture, 2004 on p.93.


Post a Comment