Print

Print


On Tuesday 29 October 2002 08:58 pm, Ian Maxwell wrote:
> So, I was just randomly wondering: How possible is it to reconstruct an
> ancestor language based on only one descendant? I was under the
> impression that reconstruction was basically done by comparing the
> grammars of various descendants. If it's not possible, theoretically
> there could be older versions of the reconstructed proto-languages that
> we don't (and can't) know about, ne?

        Of course. The classic example is Latin. If you take French, Spanish,
Romanian, Portugeuese, et al and throw them into a pot, out comes something
akin to Late Vulgar Latin, which is not the same as Classical Latin, which
they teach in schools. Classical Latin is essentially a frozen form of an
older version of the language, created by poets and orators, while Vulgar
Latin is just the natural progression of the language as spoken by the common
folks and garrison soldiers. We would have a difficult time reconstructing
Classical Latin were it not for the number of documents written in it, not to
mention its preservation throughout the middle ages as the lingua franca.
        As for how close the reconstruction is to the real deal--well, no one claims
that if you learn PIE and then enter a time machine that you'll be able to
fluently converse with the natives...but there's good reason to think that
you'll probably be able to understand and make yourself understood to a
higher degree than just grunts and hand gestures. Of course, the farther back
you go, the lower this degree goes. 5,000-7,000 years is probably the upper
limit for IE; beyond that, it's just guess work. Different language families
also will give you different results, but much of that is due to the
historical record. We can go farther into the past with IE because we have
three well-documented ancient languages (Sanskrit, Greek, and Latin) and some
lesser documented, but still helpful, languages like Hittite and Tocharian.
That gives us about 3,000 years right off the bat.
        :Peter