22 Jun 2008

Rasmussen's consonantal *o and laryngeal deletion

Here's a link to Jens Rasmussen's theory, the Consonant Infix Theory: Selected Papers on Indo-European Linguistics: With a Section on Comparative Eskimo Linguistics (1991). Essentially, Rasmussen proposes that there is an underlying Pre-IE uvular fricative /ʁ/ which has produced instances of *o in words that he deems should not have had this vocalism. I have strong doubts about his interpretation of the data but the patterns he analysed seem to be real nonetheless and require explanation.

Overall, the thing that I object to here is the less-than-plausible sound changes that he claims to have occurred. I like theories to be optimally efficient and maximally plausible however I'm unaware as yet of any realworld examples or credible reasons behind the linguistic processes that are invoked. This is an example of his version of Pre-IE sound changes:
*h₃lOig-é'- > *lOigé'- > *loig-ó-s
I want to talk more on this but essentially my reinterpretation of this is that the "consonant" that Rasmussen perceives was never a phoneme at all but rather the absence of one. So in other words, I reinterpret the above to the following:


MIE *hʷalaig̰ása
> *hʷᵊlᵊig̰ásᵊ (Reduction)
> *lᵊig̰ásᵊ (Laryngeal Deletion)
> *laig̰ásᵊ (a-Epenthesis)
> *laig̰ás (Syncope)
> PIE *loigós
By changing things like this, we can explain an apparent addition of *o (from earlier unstressed *a) as the simple result of phonotactic readjustment during the stage when Reduction (i.e. the reduction of most unstressed schwas to supershort schwas) happened. Here, a-Epenthesis is caused when a resulting onset of CCC- (i.e. **lyg̰-) threatens to rear its awkward head during Syncope. So before Syncope can kick in, *a is inserted into the first available position from the lefthand side of the word in order to optimize syllabics and avoid this icky clustering. Clever, no? I thought so. [Glen pats himself on the back.] The Laryngeal Deletion rule preceding a-Epenthesis is yet another phonotactically motivated rule to avoid complex clusters which Rasmussen also explains was first discovered by Monsieur de Saussure. And so far, I've found little reason to think otherwise but this account seems to make more sense to me than what Rasmussen has published.

Unlike a completely assumed uvular-fricative-turned-vowel, I figure that this idea would treat the attested evidence more as it is and gives it linguistic motivations that are far better attested in other languages in the world. I think I might talk about this further later[1] but enjoy these brief thoughts for now. As always, you're all welcome to share your constructive criticisms to what I've suggested because maybe I've overlooked something.


NOTES
[1] This link may be of further interest concerning Rasmussen's theories: click here.

6 comments:

  1. Rasmussen's theory seems far out there. First assuming a consonantal *o, then assuming an infixation of this consonantal *o as a derivation.

    Seems like he's trying really hard to fit something into a theory he has, rather than letting the theory develop from the examples he has.

    Back to your idea though. The way you propose it, these words behave like neuter s-stems (seeing as sa is not being ignored by the penultimate accentuation rule).

    Problematically though, these words with their o-grades are generally masculine o-stems. λοιγος and μοιχος at least are. Sure we could assume analogy, but this would not be my first choice.

    Also I'm not sure why you consider the cluster **lyg- to be illegal, wouldn't be be perfectly normal for that cluster to develop into [lig]? After all, it isn't all that much different from the stem *leikʷ- which freely alternates with *likʷ-. And *woid-/wid- seems to function perfectly normal as well.

    I'm not really sure if the environment for the deletion of the Laryngeal is valid, but it probably is. It seems likely anyway.

    I really feel like I'm missing a vital point here, feel free to point it out to me :D

    ReplyDelete
  2. Phoenix: "First assuming a consonantal *o, then assuming an infixation of this consonantal *o as a derivation."

    Yes, he's not only assuming a new sound but also assuming that it's a morpheme. Yet he coincidentally can't ascertain its meaning or usage. I have great trouble with his formulation of Pre-IE but yet I feel it's clever enough to be useful somehow.

    Phoenix: "The way you propose it, these words behave like neuter s-stems (seeing as sa is not being ignored by the penultimate accentuation rule)."

    The ending is meant to be the genitive case ending *-ása which would have been useful in forming adjectives in Mid IE. In early Late IE after Syncope, the ending would be *-ós. I believe that in the midst of the Late IE period, these genitive-derived adjectives became misanalysed as thematic nominatives ending in *-z because this gives us a source of PIE's neuter thematic adjective in *-ó-m (i.e. the genitive plural was reanalysed as a neuter thematic). Adjectives were freely formed from nouns and vice versa, and the declensional system of both adjectives and nouns clearly share the same, single source.

    Phoenix: "Also I'm not sure why you consider the cluster **lyg- to be illegal,[..."

    In the early Late IE (eLIE) nominative and accusative cases (the "strong cases"), a noun can only begin with a strong syllable (fullgrade), not a weak one (zerograde). The latter type of syllable was only allowed in the remaining "weak cases". So for the subject and direct object, a-Epenthesis was employed to assert a strong initial syllable. This strong/weak syllable thing may seem crazy to you but if you think about it, it's a great way to distinguish nominatives from genitive singulars when they end up with similar endings: eLIE thematic nominative singular *-əz versus nominative plural *-es versus genitive *-ós. Even more handy when the accent shifts to the initial syllable in acrostatic nouns during this period (but perhaps the strong form was levelled throughout the paradigm already by this time).

    ReplyDelete
  3. Whoops, correction... I meant to say "In early Late IE after Syncope, the ending would be _*-ás_." since this precedes Vowel Shift (i.e. *a > PIE *o).

    ReplyDelete
  4. Another correction of above: "eLIE thematic nominative singular *-əz versus nominative plural *-es versus genitive _*-ás_." Clearly I can't function on five hours of sleep. Sorry again.

    ReplyDelete
  5. Disallowance of zero grade in 'strong stems' definitely isn't a crazy idea. Though it indeed does sound a bit crazy at first, when you think about it, over all throughout the language this is a usual pattern so a 'fake' full grade is not that unlikely to appear.

    On a different, but not wholly unrelated note, the other day I found myself pondering on the zero grade in *uĺkʷos.

    Personally I'd expect an earlier form like:
    *wálkʷa=sa But after syncope we'd then expect a form **uólkʷos which we clearly don't have.

    To explain the zero grade in the root we could go for
    *walkʷá-sa a reanalysed genitive, resulting in *ulkʷós, but that wouldn't explain the accent on the *l. And even then you'd sort of want to insert an *a like the epenthetic a you have just described.

    So my question is, any ideas where the wolf gets his zero grade from?

    ReplyDelete
  6. Phoenix: "So my question is, any ideas where the wolf gets his zero grade from?"

    I think the short answer is that *uĺkʷos was based on an adjective *ulkʷós at the very end of the Late IE period, long after Syncope.

    The long answer is that PIE likely had different grammatical rules in different eras. My theory so far tells me that the productive rules of ablaut in Mid IE (MIE) were different in early Late IE (eLIE) and different yet again in PIE proper. Meanwhile, along with these ablaut rules still productive in the language, there were inherited forms of ablaut that no longer were productive.

    So... I think that in eLIE, the inherited ablaut from Syncope was /zero but the more *productive* ablaut in this era was the more transparent / alternation. Later still, even this latter alternation became passé as accent became tonal. Once tonal, accent shifted in newly coined words without concommitant changes in vowel quality.

    So since a zerograde is given accent in *uĺkʷos, it follows the very last rules of ablaut just before PIE, giving us a means to date precisely when the word was first coined.

    ReplyDelete