Germanic Lexicon Project
Message Board







[ Main Message Index ]     [ Previous | Next ] [ Reply ]

Author: Sean Crist (Swarthmore College)
Email: kurisuto at unagi dot cis dot upenn dot edu
Date: 2004-09-01 07:08:04
Subject: Torp, page b0322, entry 3
This is a margin comment on: Torp, page b0322, entry 3

This post demonstrates the capability to write notes "in the
margins" of the online dictionaries. You can add margin notes to
the dictionaries too; just click Comment on this entry for
words you find using the Search system.

This post is also a legitimate criticism of the content of a
particular dictionary entry in Torp. Click the link above to see the
original entry in Torp which I am responding to.

In entry tp_b0322:3, Torp lists two Proto-Germanic reconstructions for
the word for "payment, reward": *mizdô- and *mzdô-.
Torp posits the second form to explain the long vowel in forms such as
Old English mêd. I claim that the second form never existed and
is an erroneous reconstruction. All of the cited forms can be derived
by regular sound change from the first form, which contains a short

As a point of background, many older handbooks of Germanic claim that
there were two non-high long front vowels in Proto-Germanic,
* and *2 (Torp's reconstruction *mzdô- contains *2).
* uncontroversially continues Proto-Indo-European *.
*2, by contrast, is said to be "of obscure origin" (some have
speculated that it arises from *i, a putative lengthened grade
of the PIE diphthong *ei). These two front long vowels are said to
have merged in Gothic, but they do not merge in North or West
Germanic, where * is backed and merges with *.

It has been argued since at least the 1980's that *2 is a
mirage and never actually existed. Ringe (1984) examine each of the
dozen or two words for which *2 has been reconstructed. With
one exception, none of these words appears to actually reflect a long
front vowel in Proto-Germanic. Some are late loans from Latin. Some
have no cognates outside the Ingvaeonic dialects (OE, OSax, OFris) and
therefore cannot be securely projected back to Proto-Germanic. (Ringe
treats the individual words in much more detail than I give here.)

The only form with *2 which cannot be impugned is *hr
"here"; this word is attested with a long front vowel both in Gothic
and in Northwest Germanic. However, the non-Germanic cognates for
this word point to a PIE form with a short vowel. Why is this vowel
long in Germanic, then? Ringe argues that it is a matter emphatic
lengthening (if you want to emphasize that something is HERE,
you might pronounce the word with a longer vowel than usual, and this
length could eventually become phonemicized). If the vowel had
already been lengthened in Proto-Germanic, we would expect the
* to become by regular sound change in North and West
Germanic, but it does not. Ringe therefore suggests that this
emphatic lengthening occurred as a post-Proto-Germanic parallel
innovation in East and Northwest Germanic. It is unfortunate that
this parallel innovation is required, but the alternative is to posit
an entire phoneme to explain a single Germanic word. In sum,
we can reject that 2 actually existed at the Proto-Germanic

If the word *mizdô did not have a long vowel in Proto-Germanic,
then why is the vowel long in the OE, OSax, and OFris reflexes of this
word? In Crist (2002), I argue that there was a regular sound change
in Ingvaeonic where *z is deleted after *i or *e when a syllable
boundary immediately follows. When *z is deleted, there is regular
compensatory lengthening of the preceding vowel, as in e.g. *miz > OE
m "me (acc.)" and in other words. There is therefore no need
to posit a long vowel for the "reward" word (*mizdô-) in

The 2002 paper which I reference is available in PDF form:
Ringe's 1984 paper appears in Die Sprache, 30:138-155.

No other messages in this thread