Technology and Grammar

Richard Allan Stauch rstauch at charter.net
Fri Aug 16 18:34:32 EDT 2002


Hi Mark, Carl, et al.,

As a Systems Engineer, working with computers daily for the last 12
years, and a hobbyist for almost 20 years, I can safely say that Carl is
exactly correct. One programs in what one expects to get out. Especially
when one is dealing with data that depends on subjective value
judgments.

In the case of medical data analysis (for instance), one can program a
computer to look for various cholesterol levels in a blood sample, and
expect it will report when the ratios reach a set standard of
"unhealthy." What that means is that someone has to first decide what is
"unhealthy" for a broad group, and then someone else has to decide
whether the source of any given sample fits into that group. Then, after
the tests are run, someone has to decide if this particular person has
any mitigating circumstances before any decisions are made about a
course of treatment. And this is a hard science!

With language, however, it is a very different matter. It is all about
the subjective decisions one makes before the program can be written,
much less run. What happens when more than one noun exists in a block of
text that could relate to a verb in the same block? What if there are
two verbs? How does one decide where a block begins and ends?

A computer program is only as good as the one programming it. That's
because a program knows nothing -- it doesn't even know that 1=yes and
0=no unless someone tells it so. It is purely an extension of the
programmer's thinking. One cannot begin to write a program, of any kind,
without first defining exactly what it will do: where it will get its
input, how it will store the data, what it will do with the data, and
how to format its output. Every step must be calculated, and all
possible flows of logic have to be thought through. What happens if a
returned value is neither 0 nor 1? One even has to anticipate the
unexpected!

So where does that leave us with technology and grammar? Back to square
one. Only, faster.

EN XRISTWi,
Richard Allan Stauch
Long Beach, CA
> -----Original Message-----
> From: Mark Wilson [mailto:emory2oo2 at hotmail.com]
> Sent: Friday, August 16, 2002 11:37 AM
> To: Biblical Greek
> Cc: b-greek at franklin.oit.unc.edu
> Subject: [b-greek] Re: Technology and Grammar
> 
> Carl:
> 
> You wrote:
> 
> --------
> >So far as I can tell, this little exercise tells you what the
analyzers
> who
> >do the converting think the English counterparts of these participles
> are.
> >It is essentially circular, revealing essentially what the analyst's
> >judgment of these aorist participles is. But what is sought, I
thought,
> is
> >a basis for the analyst to make such a judgment.
> ------
> 
> We're getting bogged down in the non-essentials here. You have to
> remember that judgments are unavoidable, and because of that, you
> have to understand your results! This is a non-issue for linguists
> or statisticians.
> 
> Because what you are looking for are not one or two anomalies, you
> are looking for patterns that transcend translator judgment. That
> is partly the dynamics of such a search... it thrives on allowing
> translator judgments. Findings like, "in 94 percent of the time,
> you have...." is what you are after. The work of the researcher
> is to manually look at the other 6 percent. And statistically, he
> does not need to examine all 6 percent.
> 
> The only way to skew the whole process if ALL TRANSLATORS were
> working in concert to sabotage all translations from Greek
> to English. On a realistic or practical basis, that is impossible.
> Or as a linguist would say, the probably of that happening is
> not zero, but the closest number to zero.
> 
> The mindset still seems to be looking for ways for this NOT to work.
> 
> I really had not intended this discussion to get into these kinds
> of side issues. I realize some may think they are critical, but when
> you look for trends, translator decisions are not a show stopper by
> virtue of the fact that you have so many (hundreds of them), dispersed
> all over the world, and spanning over a hundred years of independent
> work.
> 
> Mark Wilson





More information about the B-Greek mailing list