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

[Ray Hendon]

Earlier, Glen Gordon wrote:

"However what defines a particular language as unambiguously unique from
another? How do we define "English"? There are many dialects of English. Do
we consider all the English dialects as part of a single
language or consider them seperate? How many dialects should there be
of English? At what point do the differences between one speech form
and another become insignificant or significant in these definitions?
Should Proto-Indo-European, Germanic or the descendant of English
500 years from now be considered "English"? What number of people need
to speak a particular speech form before it is considered a seperate
and statistically significant speech form. Is Eyak a real language?
What about artificial languages like Esparanto (some million or so
strong so I hear) or Klingon (you know how prevalent Trekkies are)?
Etc...."

The definition of a language can be as simple or as complex as you wish.
Since you work in the field of linguistics, don't you have a definition that
is workable for you professionally?  The same can be done for mathematical
uses.  Dictionaries are compiled, limited and incomplete as they are.  But,
larger and more comprehensive dictionaries can be compiled when needed.

My first guess about the nature of a linguistic dictionary would be a simple
data base of words-the most elemental unit that I can think of for a
language.  If I compiled a dictionary of spoken and written words, I could
then record a conversation between people and compare each word used in the
conversation with the data base of words.  Using a simple "hit"--"no-hit"
measuring device, I could then measure the percentage of the words used in
the conversation with words in the data base.  I could measure the
compliance of south Australian with Bristol English or Brooklyn, NY English.
Other measure like this "compliance ratio" could also be developed,
according to the needs of the researcher.  Thess types of statistics could
be used as a consistent measureing device for the degree of compliance and
of dispersion of a given population to a given linguistic dictionary.

For example: if I recorded 1500 random conversations in English in San
Antonio, Texas and  did the same in London, England on the same date, I
could then use these conversations  to complie a spoken dictionary for each
location and time.  I would have a vocabulary list of words likely to be
spoken in south Texas and in London.  These two data bases could then be
compared in an almost infinite number of ways, depending on what you were
looking for.

Admittedly, a simple word list leaves out all the rules of usage, non-verbal
clues, etc.  It is certainly not a perfect measurement of what constitutes a
language, but is could be valuable as a tool for measuring differences-which
it seems to me is an important issue in linguistics.  It would be a workable
starting point from which a more thorough or specialized definition may be
worked out as experience is gained.

What I am suggesting, at bottom, is the application of mathematical
techniques to the problems of linguistic analysis.  I guess such a field of
study would be called linguistimetrics, similar to econometrics as the
application of quantitative techniques to economic problems.  I haven't
found this word in any of my dictionaries, but I would place a bet that it,
or some other word meaning the same thing, will eventually be part of your
vocabulary.  The reason I am so confident about this prediction is because I
know how much the use of quantitative techniques has meant to other fields
of study, and I have not seen anything yet that convinces me that some of
the same techniques could not be used by linguists.

I am posting soon another note that addresses other questions that have been
raised about the efficacy of a medical model to linguistic change.  I wish I
could give more concise responses, but I am mushing around in a field whose
vocabulary is alien to me.  But, I am confident that I know the scientific
method, and it is this vocabulary that unites us.

To confess my meager credentials in the lingusitic field, I have studied
only two languages other than English-Russian and French.  And to call them
studies is an insult to anyone who has undertaken a serious study of a
foreign language.  My study of these languages was superficial at best.  A
slight bolstering of my case rests on other, non-spoken languages I have
experiences with: mathematics-algebra, calculus, statistics and geometry,
and computer languages-FORTRAN, BASIC,  C and COBOL, plus a few data base
manipulation languages, again, on a generally superficial level.  If the
field of linguistimterics does develop, these are some of the tools that
would probably be needed by a budding linguistimetrician.

Best Regards,
Ray Hendon