Re: Language family trees

Doug Dee wrote:
>No one (or no one with any sense anyway) doubts that languages evolve from
others over time, but whether the tree model/metaphor describes this well,
and whether recosntructions based on that model are accurate in all cases,
is more disputable.

Yes, I think the tree-model is at best a first approximation, and clearly
undergoes revision as knowledge increases.  Even IE does not branch in neat
bifurcations, except at the very earliest (assumed) stages.  If you accept
the Indo-Hittite hypothesis, you have
IH > 1. Anatolian languages (an ill defined group at best)
         2. All other IE
Then IE > A.  Indo-Iranian (a well-defined group I think)
                B.  All remaining IE
Then "Remaining IE" divides into severeal first-order subgroups (i.e. no
clear relation between any of them)
a. Greek
b. Armenian (perhaps subsumable w/ Greek, or is it just "influence", or is
Arm. actually an Anatolian remnant??)
c. Tocharian (position somewhat uncertain I think)
d. Celtic
e. Italic
f. Germanic
g. Balto-Slavic
h. a few uncertain lgs. like Venetic, Albanian and other extinct and/or
poorly attested languages of Illyria.

d/e, and f/g, and perhaps d/f, apparently spent some time together in
contact situations, but it's not possible to reconstruct a
Proto-Italo-Celtic or Proto-Germano-Balto-Slavic subgroup.

R. Blust has proposed a widely accepted tree for the Austronesian languages,
which basically has 4 first-order groups:

AN > 1, 2, 3, three language groups of Taiwan (and some extinct) that cannot
(at present!) be united under a single node)
        4. All other AN = older "Malayo-Polynesian"
He makes a nice bifurcating tree for MP, but IMO it remains controversial.
The very last twigs and smaller branches may be clear
(e.g.Proto-Fiji-Polynesian; a Malayic group) but the very large branches
aren't -- is there in fact a Proto-Philippine node, is there a
Proto-Eastern-Indonesian node, where if at all is the dividing line between
"Indonesian languages" and "Oceanic" languages??  As to this last, it seems
more a continuum than a sharp division....

Then there is the problem, in the entire AN area, of constant ongoing
contact.  Islanders/sailors, as they say, have many neighbors.

>See for example RMW Dixon's book _The Rise and Fall of Languages_, in which
he says that attempting to construct a family tree for Australian languages
is likely to be useless, because the history of Australian languages is
unlike the history of IE and the other groups that gave rise to the tree
model.

I regret to say I've never read that (and must), but I can imagine some of
his arguments.  Compared with mainland areas, Australia enjoyed _relative_
isolation for a very long time; the climate for the most part was not
conducive to large agglomerations of people, but rather, necessitated the
constant splitting-up of small tribal groups once they became too numerous
to support themselves in a given area.
While that would tend to cause fairly rapid language diversification, it
could be mitigated by almost certain ongoing contact between groups.

(I wonder-- what is known about Austr. groups who lived in more fertile
areas?)

Nor should it be assumed that one busload of proto-Australians arrived in
40-50,000BC or whenever, and that was it.  Surely over the
centuries/millennia many groups arrived, not all necessarily speaking the
same/related language....It's likely IMO that most set foot first in New
Guinea, only later migrated south.  Aust. and NG probably became separate
entities only about 8-10,000 BC, when the glaciers melted.

All this by way of saying, pace Dixon, that the history of other groups is
after all _not_ so different from the Australian situation, if one just
thinks about it for a moment.

Reply