Qualitative ablaut and prefixing

From: Glen Gordon
Message: 14199
Date: 2002-08-02

>The traditional term "o-grade" of IE linguistics covers quite a few
>different things that cannot all have the same causal background.


>1. Alternation � / zero depending on accent. This is practically
>only seen in reduplicated verbal categories,

Practically but not entirely by any means. The *o-grade is
associated with stative formations and must have a long prehistory.
From what I can tell, it arises in IndoTyrrhenian c.8500 BCE
from vowel harmony between the vowel stem and the two different
sets of personal endings ([*-em, *-ec, *-e] versus [*-xa, *-ta,

The *o-grade can't just be explained away as senseless
reduplication since a simple question arises: Why isn't the
perfect **kWe-kWer-e or *kWo-kWer-e instead of *kWe-kWor-e?
It's tough to explain all *o-grade forms as recent innovations.

>2. Alternation o / zero, both unaccented, further alternating
>with �; in case of lengthening we have unaccented /o:/ and
>accented /�:/.

In the case of *xakmo:n, the last syllable was originally accented,
later altered (a little after the Acrostatic Regularisation) in
Late IE. The ending here is *-mon-, an animatized variant of
inanimate *-mn. Knowing that *-mon- was originally accented, we
see that it alternated with *-men- when unaccented. This ablaut
change of *o to *e derives from Mid IE Vocalic Constraints where
unaccented MIE *a (> *o) was simply reduced to schwa (> *e) when

As for *wr.h�:n 'lamb', the ending there is the agentive *-�:n
(also *-�:r), quite a different ending which doesn't alternate
with *o.

>3. A related regularity has produced nom.sg. with /o:/ from stems
>with underlyingly long /e:/, as *pe:d- 'foot' => nom. *p�:d-s.

Erh... no. The nominative is *po:d-s, there's no doubt. It becomes
*ped-os in the genitive only because of Mid IE Vocalic Constraint
which explained the same ablaut phenomenon above.

The word for house could very well have been also *de:m- in order
to explain genitive *dems but I fail to see this as unquestionable
proof that we necessitate a ridiculous "superlong *e:". We merely
need explain the *e/*o problem here as a case of variant forms
of one word. Thus, athematic *dem- using an *e-grade, versus
the thematic *dom-o- using the *o-grade stative *dom-.

Concerning *nokWt- & *nekWt-, again there are two different stems
being used at the same time. The reason why nominative *-s does
not lengthen *nokWt- is because *-s can only lengthen an
immediately preceding syllable (which would in fact be the vowel
lost between *kW and *t).

It's plain to see in this third case that there is no true
qualitative ablaut simply because we are dealing with alternating
stems, not the paradigmatic alternations of a single stem! Again,
*o alternates with *e within the paradigm merely via MIE Vocalic

>4. There is the "thematic vowel", i.e. vowels in stem-final position which
>regularly behave in a way of their own. Being
>independent of the accent, they have no zero-grade alternants, but show up
>as /o/ before voiced segments

I now understand this phenomenon. It formed within the Late IE
period. The thematic vowel was originally schwa (and this is
_after_ the loss of unaccented schwa of Mid IE). Many of these
schwas developed after the Acrostatic Regularisation had altered
the accents of some stems. Naturally, the new unaccented vowel that
was produced was subsequently reduced to mid-central schwa.
However, next to unvoiced segments the schwa was fronted. Schwas
temporarily remained next to voiced segments but they were later
dissimilated to the nearest vowel *o (hence the unexpected
"rounding" of the vowel that couldn't be merely the product of
voicing or frequency alone).

The reason for *o in the nominative does not necessarily imply
voicing of nominative *-s. Rather, since the nominative marker
tends to lengthen preceding syllables in other stems, it's
natural to presume that the same phenomenon is to be blamed for
the *o in thematic stems. In other words:

NOM. *ekw&-s > *ekw&:-s > *ekwo-s
ACC. *ekw&-m > *ekw&-m > *ekwo-m
ACC. *ekw&-syo > *ekwe-syo > *ekwe-syo

>5. Then there is the funny -o- of the causative *mon-�ye-ti.
>6. Further Support.

I'll think on that one but I so far am resisting your basis for
the conclusion of some consonant *-O-. It smells of linguistic

>[...] its antonym based on *r�gW-os 'darkness' forms
>*�-rgWsnaH2 with prefixed /o-/;

I've already mentioned "*a-insertion" and was added to my website
monthes ago to explain the relationship between *kWetwores "four"
and *okto:u "eight". I already mentioned previously that *a- was
inserted in a very early stage of Late IE to avoid complex
consonant clustering (three or more initial consonants) after the
loss of unaccented Mid IE schwa. That would explain *okto:u
(< *a-kWtwax) and *osdo- (< *a-sdwa-). It would appear that it could
also explain *orgWsnax (*a-rgWs-nax). Also *oxgax (< *a-xgC-ax)
assuming some consonant *C (such as a semivowel or laryngeal).
This has nothing to do with "infixing" since there is no true infix
to speak of. It's rather a coping mechanism in IE syllabics.

>8. The Opposition. "Last time [Glen] checked, IndoEuropean tended towards
>suffixation, not prefixation."

Then where are the prefixes in PIE? There are none. Plus things
like the question marker *ne placed at the end of the sentence
and the overwhelming predominance of suffixation suggest an
originally SOV language. What you're intending on doing is
reconstructing a Pre-IE with prefixation. You're swimming
against the stream. The infixes are merely late formations,
originally suffixes. They formed due to misanalysis of atomic
stems like *leikW- as **lei- + some obscure modal suffix that
never existed **-kW-. Hence *-n- was placed after the "perceived"
root, causing infixing.

Interesting as always, but I'm not yet convinced, although it's
nice to at least see that we're noticing the same curiosities and
both attempting to address them in our own ways.

- gLeN

Chat with friends online, try MSN Messenger: http://messenger.msn.com