Genetic Descent/Haitian Creole

[Larry Trask]

--On Tuesday, June 12, 2001 8:18 am +0000 X99Lynx at aol.com wrote:

I'm still marking, so I'll again have to be brief.

[on certain hypothetical examples, and replying to David White]

> So, a large part of Language A goes abruptly from "genetic" to
> "non-genetic" depending on the presence of certain verb morphology.

This is a response to David White's position, not to mine.

> I think any good scientist would, in the example above, call the
> distinction between "genetic" and  "non-genetic" operationally useless.

I'll leave David White to reply to this from his point of view, if he wants
to.  From my point of view, the distinction is very far from "operationally
useless".  We do not recognize genetic links on the basis of miscellaneous
elements: instead, as always, we rely entirely on patterns.  It is patterns
that demonstrate common ancestry, not the presence of any number of
miscellaneous common elements.  Recall the case of Tlingit and
Eyak-Athabaskan: loads of common elements; no patterns; hence no
demonstrable common ancestry.

> The road that Dr. White goes down in the quote above is one inevitable
> result of the assumption that languages can be "genetically" descended
> from ONLY one ancestor.  Sooner or later, you will have one "genetic"
> element coming up against another.  The contest then becomes which
> "genetic" element is the winning "genetic" element.  Everything else
> loses and becomes non-genetic.

No.  David White aside, this is simply not the way comparative linguistics
is done.  We do not count up elements of varying origins and then base any
genetic conclusions on our totals.

In English, we now have far more elements of French, Latin and Greek origin
than we have of native Germanic origin.  But no comparative linguist wants
to conclude, therefore, that English is no longer a Germanic language.
English is Germanic because we can identify the correspondences linking it
to the other Germanic languages, and because -- as a result of these
correspondences -- we can trace the ancestry of English back to
Proto-Germanic.  No matter how many Latino-Romance elements we find in
English, we cannot move English into the Romance family: the required
patterns are just not there.

In the same vein, Basque is not a Romance language, in spite of the vast
number of Latino-Romance elements it contains, and Romanian is not a Slavic
language, in spite of the very large number of Slavic elements in it.  And
Albanian is not Greek, Romance, Slavic, Hungarian or Turkic, even though
elements of these origins greatly outnumber the inherited elements.  It is
a distinct branch of IE, because this is the only conclusion permitted by
the patterns we observe.

[on a hypothetical mixed language]

> Common sense and half-way decent science would suggest that it is a form
> of both source languages.  Why should one genetic element make another
> non-genetic?

It doesn't.  When we borrowed the word 'kangaroo' from Guugu-Yimidhirr,
English acquired no Guugu-Yimidhirr ancestry at all, and the inherited
Germanic elements became no less inherited.

> Words and morphs are not like biological cells.  Each word or piece of
> morphology does not contain a hologram of the entire language in its DNA.
> We cannot clone an entire language out of a single word or verb
> morphology.  A language is made up of many totally independent parts.
> Why should the "genetics" of one part affect the "genetics" of another
> part.

This is not the way we use 'genetic' in linguistics.  We do not apply the
term to individual elements within a language.  We can no more ask whether
the English word 'pity' is genetic or not than we can ask whether it is
green and squishy.

[on Haitian Creole]

> Common sense and half-way decent science would suggest that there is
> something wrong here.  Obviously there are different elements in Haitian
> Creole and they obviously come from different sources.  The older source
> among the generations of speakers who passed on the language was West
> African.  The later French element may be "more genetic."  But why would
> that make the West African elements "non-genetic"?  They can lay claim to
> being original, continuous and native with those speakers.  How could
> they suddenly become totally non-genetic?

I don't know enough about Haitian Creole to make any pronouncements about
it -- though I have the impression that HC has a rather more complex origin
than a standard average creole.  Still, the position among those who have
studied it appears to be that HC is not the direct descendant of any
ancestor -- in other words, it is what we now call a 'non-genetic language'.
Whether this conclusion is correct is an empirical matter, but it least
it's coherent.

> Isolating systematic elements of two languages to find a common ancestor
> is a powerful methodology.  But why would we conclude from that process
> that either of those languages as a whole can only descend from one
> ancestor? What makes one element genetic and a similar element borrowed?

Patterns, patterns, patterns.

Larry Trask
COGS
University of Sussex
Brighton BN1 9QH
UK

larryt at cogs.susx.ac.uk

Tel: (01273)-678693 (from UK); +44-1273-678693 (from abroad)
Fax: (01273)-671320 (from UK); +44-1273-671320 (from abroad)