Re: IE athematics

----- Original Message -----

From: Glen Gordon

To: cybalist@egroups.com

Sent: Monday, March 27, 2000 5:45 AM

Subject: [cybalist] Re: Odp: IE athematics, the Semitic w-verb and Akkadian /ala:ku/ vs.*leikW-

Finally, the range [*leigheti/*leighti/*lighéti] is most certainly not
reconstructable for the most ancient stages of IE, unless you can justify
reconstructing this range of variation with valid daughter examples of this
or parts of this range. Perhaps you were just tired at the time you wrote
this. The verb *pelh- is reconstructed. Many verbs like *linékWti with modal
affixes *-n- or *-ye- are clearly derivatives and different in both form and
function, Piotr, so there is nothing to argue.

---

 

Glen,

 

I argued mainly against identifying ROOTS with STEMS. English morphology is "word-based" in the sense that productive derivatives are predominantly based on forms which may occur independently as words. E.g. loving, lover = /l@.../ + /-IN/, /-r/ (and, with more nesting, lovingly = [l@... + -IN] + -li). Also, regularly inflected forms consist of a free stem (existing as an independent word) plus inflections, e.g. loved, loves = /l@.../ + /-d/, /-z/. But what works for English needn't be generally valid. Even Old English was different, though evolving towards the modern system: the infinitive was lufian, the inflected forms of the present tense were lufath, lufiath, etc., the imperative (sg.) was lufa the preterite was lufode, lufodon, the corresponding noun was lufu, etc. However, the root /luf-/ underlying it all was a "bound" morpheme which didn't occur in isolation as a word or even as a stem.

 

Likewise, many reconstructed PIE "verbs" are not actual verb stems but bound roots yielding various derivatives (I agree that different derivatives had different aspectual overtones and were functionally not the same word). The fact that a verb is "derived" doesn't automatically make it younger than the root to which it is related. The bare root (with or without a thematic vowel) may have never existed as an autonomous stem in the first place.

 

The thematic/athematic distinction is clearcut when it refers to conjugation types, not to individual verb-yielding roots, which could produce stems of either or both types. We say (very inaccurately) that *bher- is a "thematic verb", because the most common derivative of this root is *bhere-ti, but it's wise not to ignore *bhi-bher-ti and to have a place for it in the reconstructed system.

 

You seem to assume (if you don't, correct me) that morphologically complex forms are DIACHRONICALLY reducible to simpler ones, and that today's suffixes are yesterday's clitics. Of course this is often true, but the question is, is it true often enough for your analysis of the thematic vowel to be the most reasonable one (let alone the only imaginable one)? New morphemes may also come into existence through reanalysis, back-formation, etc. The hypothesis that *-e/o- represents an old pronoun has been discussed in the literature (Specht argued for it the 1940s, and I don't think he was the first one). Knobloch (1953, Lingua 3) analysed it as an object marker. But other analyses are also possible, and have also been put forward and discussed ad nauseam without reaching a consensus.

 

The fact that a speculative argument is internally consistent doesn't make it less speculative. Other people might produce equally clever alternative interpretations. A statement of your position -- especially when it concerns unattested languages -- does not automatically become a "theory" until you show that it respects the constraints imposed by historical linguistic methodology. In particular you commit a methodological error if you segregate the evidence on which your theory is based, ignoring the part which demands a different conclusion. IMHO, the way in which you dismiss the problems caused by IE athematic conjugations is an unacceptable simplification and that your scenario needs to be revised in this respect. You may shrug off my reservations (this is what your Semitic-loan proposal amounts to), but that won't help your theory. Other linguists will probably tell you the same things.

 

 

Piotr