'Critical' Age

Mike Salovesh t20mxs1 at CORN.CSO.NIU.EDU
Fri Mar 10 15:08:18 UTC 2000


"Dennis R. Preston" wrote:
>
> Aaron,
>
> A never-ending problem for us sociolinguists (who usually work with
> apparent-time rather than real-time data). How do we distinguish between
> age-grading (the fact that one age group uses one form, another another)
> and language change (in which the younger age-groups' forms are the "wave
> of the future")? Indeed, quite a lot of work has been done on just this
> problem. For a good start, read pp. 76-81 in Chambers and Trudgill's
> "Dialectology," 2nd ed., Cambridge , 1998.

Speaking from my dotage, I remember when I was a brash young graduate
student and met Morris Swadesh for the first time. Norman McQuown had
invited him to the U.S. to give a lecture at the University of Chicago.
(Mac just didn't believe in such things as blacklisting.)  As Mac's
captive graduate student, I joined him and Swadesh at dinner -- and
proceeded to demonstrate just how brash I was.

I decided I'd never have a better chance to learn what Swadesh's
glottochronology was about.  As is typical of discussions around the U
of C, I approached the idea by questioning what I thought of as the
basic assumptions of glottochron. (I'm no longer embarrassed at having
gotten the assumptions wrong.  That had more to do with being young than
with being brash.) I asked "What makes you think that there is a
constant and universal rate of change in the first place? How do you
come up with a figure for that rate that sounds to me like spurious
quantification?  Couldn't it be 85%, or 59%, or some other rate of
change per millenium -- with different rates in different languages or
language families?"

Don Mauricio, as I came to call him after working with him on a joint
project in Mexico, must have been feeling gentle that day.  He suggested
that maybe my question was upside down.  "You might get farther trying
to explain  the rate of retention rather than the rate of change."  (He
didn't say that was the whole point, perhaps because he saw no need to
crush me.)

Then he proposed an idea that fits the direction of the present thread.

"Suppose we posit that languages are inherently centrifugal: they all
change as fast as they possibly can and still remain intelligible. Under
that assumption, there would have to be a very strong restraining force
if a language were to retain any recognizable continuity after a
thousand years. I think there is such a force, and that it operates in
all languages. It's very simple:  if language is to work at all,
grandparents have to be able to talk to their grandchildren. Retention
has to span three generations because that's how many generations have
always been in communicative contact.  I think the basic rate we use in
glottochronology is simply the thousand-year result of cumulative
three-generation continuities, or changes if you must put it that way."

At that point, another student at that dinner objected that three
generations of interaction really wouldn't have been possible in times
or places where life expectancy was 35 or 40 years: grandparents would
die off before their grandchildren were ready for adult conversation.
Swadesh and I were in full agreement in rejecting that notion; we almost
spoke in alternating paragraphs.  But that turned the conversation to
considering several kinds of fallacies that seem to make sense until you
return to the basic data.  We didn't return to talking about
glottochronology that night.

The short version of why every society has known a fairly high degree of
direct communication between grandparents and grandchildren is found in
the distribution of the ages when death is likely to occur. Average life
expectancies at birth that are in the 35 to 40 years range are
invariably the result of very high rates of infant and childhood
mortality rather than the premature deaths of adults. When comparatively
few people survive to adulthood, those few can live very long lives
without appreciably raising the average life expectancy at birth.

When I began fieldwork in the Highland Maya community I studied longest,
the local health authorities told me that life expectancies at birth
were somewhere between 35 and 40 years. I also was told that at least a
third of the children born alive died in infancy; half of those born
alive died before reaching the fifth birthday. Any man who survived long
enough to become a father was very likely to live until his 60s or 70s.
Women faced the additional complications of death in childbirth or by
perinatal infections, but any woman who survived her childbearing years
was very likely to outlive her husband(s) and other male relatives.  I
met many men who were still actively farming in their late 70s or early
80s, and plenty of women who were socially and economically active in
their 90s.  In a community of around 5000 adults, I knew around a dozen
women and a few men who were older than 100. (For reasons that aren't
relevant here, it was easier to establish the ages of very old women
than those of very old men.)

Okay, them young'uns out there don't talk like God meant them to, but us
older folks understand more of their talk than they think.

-- mike salovesh    <salovesh at niu.edu>       PEACE !!!



More information about the Ads-l mailing list