The Indo-European Hypothesis [was Re: The Neolithic Hypothesis]

[Ray Hendon]

Thanks for taking the time to send your remarks on the credibility of the IE
hypothesis.  I have no problem with believing that at some time in the
ancient past, a group of IE speakers migrated from some part of the Eurasian
land mass to another, and took their language with them.  And the tendency
for languages to diverge because of the lack of contact between them, seems
imminently sensible to me.

Larry Trask was quite eloquent in defending IE, but he was also explicit in
pointing out the failing of the IE model at a practical level.  He affirmed
that the IE model does not explain how the languages that are used in Europe
today, got that way.  And others have pointed out the failings of IE to
explain how other languages developed.  In some instances, it seems that
convergence of languages is more explanatory of the current linguistic state
than a divergence model.

But, the similarities of vocabulary among the IE languages, at least to the
extent that we know about them, is too strong to ignore.  And the IE family
model of language divergence does fit the European experience.  However, it
is not a relaible predictor of specific languages, and the family model upon
which it is based is not a reliable predictor of the divergence/convergence
dicotomy of all languistic groups.  These are the conclusions I draw from
what I have read so far.

Given the absence of written records, it is remarkable that we can know as
much as we do about ancient languages.  Along with mathematics, archeology
and geology, linguistics has entered the era where the further away we get
from an event, the more we know about it.  In the future, we will know more
of how these things happen.

My academic training is in Economics and econometrics, where we build models
of economic relationships that explain choices people make when a trade-off
confronts them.  My first intensive study was on the spread of the use of
heroin, and in my investigation I found a dissertation from a graduate
student in mathematics at  MIT who had adapted an epidemiological model from
the medical community to predict the spread of an epidemic of heroin
addiction.  He showed that the mathematical model used by epidemiologists
was applicable to the spread of addiction to herion and herion use, and that
the exact same parameters that defined the spread of infectious disease
explained the spread of heroin addiction.  The parameters were different in
magnitude, but the model was accurate for either event.

The reason I bring this up is that it seems to me that an over-all
theoretical framework of the spread of language has not been developed.
Not only the IE model, but all other linguistic models that deal with the
development, spread, divergence and convergence of languages, need a
theoretical framework if they are to be successful in predicting linguistic
adoptions.   It seems to me, that the theorectival framework of an
epidemicological model of the spread of disease would easily supply this
need of the linguistic community.  It would not be a herculian task to
attempt to specify such a model, and the results could be quite
enlightening.

Here are the basic parameters of an epidemiological model:  You begin with a
population which is divided into those who are "susceptible" to the
infection vs those who are "unsusceptible".  Once that ratio is known, or
estimated, the rate of contact between "infecteds" and "susceptibles" is
calculated or estimated.  From these simple parameters then, the rate of the
spread of the infection and the saturation level can be estimated and
predicted.

Susceptibles, in the linguistic world, would be those who were exposed to
the new tongue and those who found a need to learn it: a child attended by a
care giver would find it advantageous to learn the language of the
care-giver:  a merchant who wished to conduct business with a different
linguistic group.  Both these examples actually share an underlying
interest--an advantage seen, in learning a new language.  This is the kind
of parameter that "susceptiblity" deals with.

The exposure side (rate of contact between infecteds and susceptibles) of
the equation is easily identified in the linguistic community:  The
relatively few Romans sent to govern England was not sufficient to generate
a critical mass of exposure units for Latin to predominate.  And the Romans
left local courts and laws stand, lessening the need for everyone to know
Latin in order to get along. In this case, the exposure rate was too law for
Latin to make the transition of being adopted.

This model, if the parameters were known, could be used to explain why the
Normans adopted the local French language after they conquered that part of
France and why the Saxons kept their language when the conquered England.

Do you see any applicability of this kind of mathematical approach to the
spread of languages?  Has this type of approach been undertaken by anyone in
the linguistics community?

Ray Hendon