Re: [TIED] PIE ablaut

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

From: Mark Odegard

To: cybalist@egroups.com

Sent: Saturday, June 10, 2000 2:08 AM

Subject: Re: [TIED] PIE ablaut

Mark writes:

 

The original poster asked for an 'exact' definition. I've asked that question myself. A major problem with the definition of 'ablaut' is that Indo-Europeanists tend to use it more broadly than the dictionary definition allows. It's not just used to describe vowel-variation in inflections and reflexes of a root word, but is frequently a shorthand term to describe the whole host of phonological changes that lead to the daughtering of proto-Indo-European, and within the daughter-languages themselves.

 

Technically, the term "ablaut" is reserved for morphophonologically relevant vowel alternations in PIE, OR to their systematic reflexes in the daughter languages. So it's legitimate to speak of "Germanic ablaut" in the sing/sang/sung series (PGmc *in/*an/*un < PIE *en/*on/*n), of "Greek ablaut" in leípO/léloipa/élipon (a faithful continuation of PIE *ei/*oi/*i), etc. It should not be used loosely of ANY vowel (or, worse, consonant) alternations; those which are historically important usually have their own names (e.g. umlaut, as in man/men). If you need a general term for such phenomena, "vowel alternation" does the job well enough.

 

Of course there must have been a time when (any subtype of) PIE ablaut was a transparently motivated, unremarkable phonetic process. In some cases the conditions determining the alternation (stress contrast, compensatory lengthening, etc.) are easily reconstructable, sometimes less so (e.g., there are several competing "explanations" of qualitative ablaut). The problem is that once an alternation begins to play a morphological function, it spreads far beyond its original domain and covers its tracks, so to speak.

The vowel represented by 'a' in the words photograph and photographer are different (as is the second orthographic O in 'photographer'). In 'photographer', the vowel represented by orthographic A (ipa script a in my dialect) has reduced to schwa, while the second orthographic O has gone from schwa to ipa-script-a.

 

This change in vowel quality is directly caused by the shift in stress between two words: FOE-tuh-graf, foe-TAW-gruh-fer (I'm too impatient right now to construct the proper IPA form in Unipad).

 

Of course (according to the uniformitarian principle in historical linguistics) currently observable phonetic processes in English or whatever other language must be similar to those operating under similar conditions in languages spoken a few thousand years ago. In linguistics almost nothing happens without known precedent. By studying modern sound changes we can get a lot of insight into the working of PIE phonology and morphophonology.

 

Take the Slavic vowel/zero alternations caused by the loss of alternate years: Old Polish ps-ek 'doggie', gen. pies-k-a does NOT continue an IE pattern but nonetheless looks exactly like PIE *xój-u/*xj-óu-.

 

The loss of postvocalic /r/ in non-rhotic English accents has produced effects similar to those postulated by the laryngeal theory (including vowel colouring!): Middle E /hert-/ 'heart' > early Modern E /hart/ > RP /ha:t/; note also the schwa-vocalisation of syllabic /r/ in /wintr/ > /wint@/.

 

The lengthening of the stressed final-syllable vowel in divine or serene (upon their borrowing in Middle E times) is reminiscent of the lengthening in *pxte:r or *kuo:n and may have been similarly motivated. Such examples could be multiplied ad infinitum.

When discussing our strong verbs, however, most native-speakers of English are at a loss to explain the logic behind strong verbs, e.g., sing/sang/sung. It is 100% non-transparent, the precise converse of the situation for words such as photograph/photographer. What it comes down to is that the rules that generated strong verbs have been long defunct, and our strong verbs are fossils of the previous regime. Combined with other phonological shifts as well as 'erosion' (so I gather) of certain endings, the system that gave us strong verbs is unintelligible -- and making these changes intelligible requires a rather lengthy lecture in historical linguistics and phonetics.

 

As I understand it, the grammatico-phonological process that generated the strong verbs found in Germanic is apparently *directly* inherited from proto-Indo-European, or at least, at the level of an immediate daughter.

 

In IE, we need to mention that ablaut also worked with nouns, especially the case endings. IE had eight noun cases, most of them with a distinctive endings (nominative and vocative were mostly naked, if I remember my reading correctly). At one point in the history of IE, these ablaut changes were as transparently understandable to the IE-speakers as the ablaut changes in the words photograph/photographer are to native-speakers of English. At a later point, as IE daughtered into different languages, for some IE-groups, these ablaut changes were as inexplicable to them as the ablaut changes in the verb ring/rang/rung are to native-speakers of English.

 

Note, however, that language recycles a lot of historically accumulated junk, and even an "inexplicable" alternation may be utilised in morphology. For example, ring was originally a weak verb (OE hring-an, pret. hring-d-e), but the sing/sang/sung pattern, though unmotivated phonologically, was salient enough to exert analogical pressure. Speakers of Modern English don't know or care a damn about the origin of the cónvert (n.)/convért (v.) alternation, but are able to apply it analogically (the number of such pairs has been rising slowly but surely over the centuries). Similarly, speakers of PIE utilised the *e/*o alternation very productively in their conjugations despite the fact that they no longer understood its origin (whatever it was).

 

You've got it all right, on the whole ;)

 

Piotr