Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site mit-athena.ARPA
Path: utzoo!watmath!clyde!cbosgd!ihnp4!zehntel!hplabs!hpda!fortune!amdcad!decwrl!decvax!mit-athena!jc
From: jc@mit-athena.ARPA (John Chambers)
Newsgroups: net.lang
Subject: Re: "high-level" (some thoughts)
Message-ID: <17@mit-athena.ARPA>
Date: Wed, 9-Jan-85 13:30:55 EST
Article-I.D.: mit-athe.17
Posted: Wed Jan  9 13:30:55 1985
Date-Received: Mon, 14-Jan-85 04:07:41 EST
References: <331@wu1.UUCP>
Lines: 41

sing the term, and their motive is clear:
they want to sound "technical".  Physicists are quite justified in using
the term with its correct technical meaning, and disregarding the public's
misuse of the term.

Mathematicians as well as Computer Scientists have been annoyed by the
public misuse of the term "parameter", which the media, bureaucrats, and
so on are saying when they mean "limit".  We aren't forced to adopt the
public's misuse; we are quite justified in continuing to use "parameter"
with its standard technical meaning. 

For some years, the public has been misusing the term "doctor" to mean only
medical doctors.  This doesn't mean a person with a Ph.D. in Chemistry or
Paleontology or Basket Weaving is forbidden the use of the term.  It just
means that the general public is using a precise academic title incorrectly.

Similarly, I think that the term "high-level", misused though it may be
by admen and journalists, does nonetheless refer to a useful concept, and
one for which there doesn't at the moment seem to be a better term.  Calling
one language "high-level" and another "low-level" is not praising either,
nor is it an insult.  C is called a "low-level high-level language" by
its original developers, and we know what they meant by that.  It is a
good description, and I think it is a useful one.  It says that C was
designed to deal more directly with hardware issues than other languages
that are superficially similar; it was not designed to deal well with
abstract data independent of its representation.  This is good in some
applications (like writing a device driver), bad in others (like writing
a portable symbolic math package).  So C is well-designed for some tasks,
and poorly designed for others.  This is useful information to those who
need to make decisions about how to implement things.  Please don't try
to take this away from us!

I wish we could find a way to stop (or at least seriously impede) the misuse
of technical terms by people who are just trying to sound impressive.  But I
suspect that this is impossible as long as science and technology have such
high regard among the general the population.  We just have to learn to live
with the fact that many technical terms also have a "media" meaning that is
often a garbled or distorted parody of the "real" meaning.  It's part of the
price we pay for our success.

				John Chambers