Evolution of Language?

["Martin F. G. Schaffernicht"]

Dear Colleagues,

A friend of mine is a linguist; he is interested in the evolution of
language. The kind of questions he asks is "how come that french people
in 1900 used different words and different grammatical constructions to
express an idea <I> that french people of the year 2000 do? (Even though
they still understand what the year 1900 frenchmen tried to say.)"

When he came to my office he saw "Industrial Dynamics" and asked "what
is it"; I tried to explain and he asked "could I try to answer my kind
of questions with this?"

I am not a linguist; I would believe that this involves dynamics in
society, and language would be but one element in a chain, so why not
give it a try. On the other side, "evolution" seems to be approached
more by "genetic algorithms" or "evolutionary programming" and the like.

But maybe someone in the System Dynamics Community has worked on this
subjet before? In this case I would be grateful if you can give me some
hints.

Thanks,

Martin Schaffernicht
From: "Martin F. G. Schaffernicht" <martin@utalca.cl>
Facultad de Ciencias Empresariales
Universidad de Talca

http://dig.utalca.cl/carpeta/Martin_Schaffernicht

["Jack Homer"]

Martin Schaffernicht asks about models of language evolution. An amazing
piece of work on this topic, combined with an analysis of global human
migration patterns over thousands of years, is by Harald Sverdrup at Lund
University, Sweden. His paper at the 2002 conference in Palermo was
entitled: "System Dynamics Applied to Reconstruct the Dispersal of Modern
Man on Earth and Language Patterns During the Last 120,000 Years." No
kidding. It was a highlight of the 2002 conference for me.

- Jack Homer
From: "Jack Homer" <jhomer@comcast.net>

["Michael Bean"]

Martin,

Regarding your query about using a model to explain the evolution of language, I
believe you might be able to use the work of Jason Wittenberg and John Sterman
in their paper, "Self-Organization, Competition, and Succession in the Dynamics
of Scientific Revolution"

The model and paper are available at:
http://web.mit.edu/jsterman/www/SDG/self.html

Try substituting "New Scientific Theories" with "New Grammatical Constructions"
and you might have something that could serve as a framework for developing a
linguistic evolution model.

Best regards, Michael
_______________________________________
Forio Business Simulations

Michael Bean
mbean@forio.com
www.forio.com

[GBHirsch@aol.com]

Martin,

There was a paper presented at the 2002 International System Dynamics
Conference in Palermo on the evolution of language over a much longer time-scale and
broader geographical area. It was presented by Harald Sverdrup of Lund
University and was titled ""System Dynamics Applied to Reconstruct the Dispersal of
Modern Man on Earth and Language Patterns During the Last 120,000 Years". The
full paper should be on the conference CD.

Gary Hirsch
GBHirsch@comcast.net

[klausv@stud.ntnu.no]

For computer modelling of evolution of languages,
"Simulation for the Social Scientist" by N Gilbert and KG Troitzsc
provides some relevant examples and references using agent-based models.


The book is also available on web, see
http://www.uni-koblenz.de/~kgt/Learn/Te ... wBook.html
and in particular for the evolution of languages,
http://www.uni-koblenz.de/~kgt/Learn/Te ... de112.html

Klaus
From: klausv@stud.ntnu.no

["Philip Keogh"]

Hi All

A very interesting book I came across, relating to language is...

Le Ton Beau De Marot by Douglas Hofstadter, Some of you may be familiar with his work Godel, Escher Bach...

It examins some of the issues around translation and meaning. It might be worth a look.


Regards


Philip Keogh
From: "Philip Keogh" <Philip.Keogh@leedsth.nhs.uk>
Pathology Information Officer
(see our website at www.leedsteachinghospitals.com)

["geoff coyle"]

This sounds like a fascinating question but Id hate to rush into it without
finding out more about the domain knowledge. Can your friend point you to
what Id guess would be an extensive literature?

Geoff
From: "geoff coyle" <geoff.coyle@btinternet.com>

["Prof. Dr. Niall Palfreyman"]

"Martin F. G. Schaffernicht" schrieb:

> But maybe someone in the System Dynamics Community has worked on this
> subjet before? In this case I would be grateful if you can give me some
> hints.

Im afraid I cant give any hints, because my thinking is not yet far enough
advanced, but I can tell you that the issue interests me greatly. I am
concerned with the flow of information in biochemical systems, and there I
notice immediately on the one hand a need for the tools of system dynamics,
and on the other hand the need for a concept of structure.

The quantities which flow around a stock and flow diagram are scalars - in
the language of Petri nets they are uncoloured. The contents of a stock is
just a number, but the things which flow around a chemical system are
objects (molecules, complexes, ions) with internal structure which interact
with each other in predictable ways - just like grammatical structures in a
language. Now one obvious way of incorporating these interactions into S&F
diagrams is to represent the interactions between these objects as S&F
structures, and provided your friends evolving linguistic objects are
sufficiently simple, that would certainly work (I have a number of simple
models representing evolution of traits in genetics).

This solution regards the internal structure of an object as merely an
elegant way of specifiying the possible interactions between the objects.
However this solution lacks the ability to make generic statements about
interactions between the evolving objects (e.g.: any indefinite article must
be followed by some noun). Once you start trying to express such
relationships as S&F structures, the complexity of the models skyrockets
exponentially because each possible juxtaposition of article and noun must
be explicitly modelled as a small S&F structure.

So do I have any solutions? No, sorry. Im working on it myself, and Id be
very interested in exploring the matter further in discussion, but I dont
expect to reach any useful conclusions myself for the next two years or so.
I can only say that in my opinion this is a very important meeting point
between SD and mainstream scientific modelling.

Best wishes,
Niall Palfreyman.
From: "Prof. Dr. Niall Palfreyman" <niall.palfreyman@fh-weihenstephan.de>