What is Relatedness?

[Sean Crist]

In an earlier post, I gave the following as the family tree for Germanic:

>>             PGmc
>>            /    \
>>        PNWGmc    \
>>        /   \      \
>>      WGmc  NGmc   EGmc

On Tue, 11 Jan 2000, Rick Mc Callister wrote:

> What is the basis of this tree in Germanic?
>
> 	How secure is East Germanic?
> 	Why is it not just an early form of N. Germanic?
> i.e. what are the points that distinguish it?
> 	Is there any merit to the idea that there was a separate branch of
> Germanic including Anglian, Jutish and pre-Frisian, intermediate between N
> & W Germanic, andf that modern English and Frisian are the result of a
> fusion between this and W Lowland Germanic?

Okay, let me answer your first two questions first.

The sound changes which characterize NWGmc, which are not shared by EGmc,
are as follows:

-Stressed */e:/ > */a:/
-Word final */-o:/ > */-u/
-Unstressed word-final */-am/ > */-um/

There's also a morphological character whose specific identity is slipping
my mind at the moment; there's some noun case or verb form where NWGmc has
an innovative form.

Now, it's true that some earlier authors grouped NGmc and EGmc together,
so that the tree would show PGmc branching into WGmc and "NEGmc".  The one
decent argument for doing this is the so-called "sharpening"
(Verscha"rfung).  PGmc *-ww- comes out in Gothic as something spelled
<ggw>, and in Old Icelandic as something spelled <ggv> or <gg>.  This does
superficially look like a shared innovation, but when you consider this
one character against the four listed above, it appears that the correct
analysis is that the Verscha"rfung is a parallel innovation
(Alternatively, the phonologist Rolf Noyer and I have discussed the
possibility that the Verscha"rfung was a _synchronic_ rule of PGmc which
was lost in WGmc, which would likewise be consistent with the tree I drew
above).

Now, for your second question.  You asked:

> 	Is there any merit to the idea that there was a separate branch of
> Germanic including Anglian, Jutish and pre-Frisian, intermediate between N
> & W Germanic, andf that modern English and Frisian are the result of a
> fusion between this and W Lowland Germanic?

Within West Germanic, it's not possible to draw a clean Stammbaum, because
the dialects developed in close contact with one another.  Still, we
recognize a major division within WGmc: 1) Ingvaeonic, also known as North
Sea Germanic, which includes Old English, Old Frisian, and, with some
complications, Old Saxon; versus 2) Old High German.  (You used the term
"Jutish"; we don't have any documents written in a language called
"Jutish", but given that there's a traditional claim that there was a
people called the Jutes who participated in the Germanic invasion of
Britain, we can take Jutish to mean some part of the Ingvaeonic dialect
continuum.  Anglian, of course, is a dialect of OE.)

There's sometimes been an idea floating around that "West Germanic" is an
artificial designation, i.e. that "WGmc" is really an amalgamation of
dialects which don't actually descend from any single Proto-West-Germanic
language. This idea is can be safely rejected; there are some seven secure
cases where the WGmc languages all share early innovations which are not
found outside WGmc (again, I can list them if there's interest, but it
would take a lot of time). The Ingvaeonic languages all show the common
WGmc developments, and none of the NGmc developments; Ingv. is thus
securely WGmc and not "intermediate" between WGmc and NGmc.

There are a few sound changes which distinguish Ingvaeonic from the other
WGmc languages.  I won't go thru them all here unless there's interest,
but one change which gives readily recognizable results to anyone who
knows English and German is this: in Ingv., but not in OHG, a nasal before
a voiceless fricative deletes with compensatory lengthening of the
preceding vowel (this had already happened before */x/ in PGmc, but Ingv.
extended it to all voiceless fricatives).  Examples:

	English	German
	other	ander
	tooth	Zahn
	soft	sanft
	five	fu"nf
	(etc. - many more examples)

OE and the other Ingv. languages all underwent this change; none of the
OHG dialects did.

So if I were going to further flesh out the tree I drew above, I'd put a
further branching under WGmc with a two-way split into Ingvaeonic and OHG.
I wouldn't care to draw a Stammbaum for WGmc with any further detail than
that, because there's just too much cross-contamination between dialects.

I can't tell you how refreshing it is to actually talk about something
with some real linguistic content for a change, rather than the
interminable and largely vacuous discussions about whether a parent can
coexist with its daughter, etc.

  \/ __ __    _\_     --Sean Crist  (kurisuto at unagi.cis.upenn.edu)
 ---  |  |    \ /     http://www.ling.upenn.edu/~kurisuto/
  _| ,| ,|   -----
  _| ,| ,|    [_]
   |  |  |    [_]