Comments on: The linguistic arrow of time

By: Rakko

Rakko — Fri, 20 Jun 2008 05:37:30 +0000

Katz: Deutscher doesn’t say that. It’s been about a year since I read it, but I seem to remember he says language structures simplify and become complex in a cyclical way.

By: brian

brian — Tue, 26 Feb 2008 16:37:57 +0000

@ Lord: Vocabulary items that imply a date are a crucial tool for reconstructing the chronology and evolution of languages. The best known example is that most of the Indo-European languages have cognate words for wheel and axle, suggesting that the major bifurcations in the language family took place sometime after the development of wheeled vehicles.

My question, though, is whether it’s possible to set aside such clues and still reconstruct the tree.

By: brian

brian — Tue, 26 Feb 2008 16:30:43 +0000

@ Seb: If we look only at a single set of words — such as those for the concept “100″ — it seems really hard to pin down the full structure of the evolutionary tree. Still, I would argue in a loosey-goosey way that “k’mtom” looks closer to “centum” than it does to either “cent” or “cento,” justifying the way I’ve arranged that branch of the tree. But you’re quite right that there’s nothing in the evidence presented to favor the tree I drew over your alternatives of centum->cento->cent or centum->cent->cento.

For what it’s worth, a paper by Russell D. Gray & Quentin D. Atkinson (Nature, Vol. 426, 27 November 2003, pp. 435-438) suggests this tree structure for some of the descendants of Latin:

(Sardinian (Italian (Romanian (French (Catalan (Spanish Portugese)))))).

By: baoilleach

baoilleach — Tue, 26 Feb 2008 09:27:28 +0000

Trees in evolutionary biology have the same problems. The only way to identify ‘the root’ is to use a known outgroup. That is, something which it is known split off from the other languages in the tree at an early period. (Disclaimer: I’m not a biologist :-) )

By: Lord

Lord — Mon, 25 Feb 2008 21:55:38 +0000

I would think vocabulary would be a better measure of advancement. People don’t create and use words for which they have no need. Thus much legal terminology was borrowed from latin since they already had created a set of ideas and reasoning for that purpose. While scientific terminology was fabricated rather than borrowed out of greek and latin roots, other languages incorporate rather than invent a parallel terminology.

By: Seb

Seb — Mon, 25 Feb 2008 21:34:32 +0000

You were just trying to be consistent with the topology, the order is not that important, right?. Sorry.

By: Seb

Seb — Mon, 25 Feb 2008 17:20:16 +0000

I don’t see why you can not interchange say French and Latin (or Italian and Latin) in your diagrams or even write French->Italian->Latin in series. What is that made you order them the way you did?

By: Jonathan Katz

Jonathan Katz — Mon, 25 Feb 2008 16:08:16 +0000

In “The Unfolding of Language…,” Guy Deutscher puts forth the thesis that there is a similar “arrow in time” for linguistics, and that the trend is generally to simplify words/syntax/grammar (as you point out in your post). You may want to take a quick look through his book.