MathGroup Archive 1997

[Date Index] [Thread Index] [Author Index]

Search the Archive

Re: Not so Useful Dumb User Questions

  • To: mathgroup at
  • Subject: [mg9176] Re: [mg9148] Not so Useful Dumb User Questions
  • From: Mark Evans <evans at>
  • Date: Tue, 21 Oct 1997 02:02:49 -0400
  • Sender: owner-wri-mathgroup at

I understand the theoretical justification for complex numbers in
algebra and in mathematical software just fine, thanks.  While we are
all busy making huge assumptions, thinking in plenty, and being
astounded with each other's persistence, I will restate what I said in
another post.

> I am not
> talking about school girls and factory workers.  Since he asks, let me
> say that the people I have in mind are typically Ph.D.'s in technical
> fields who are simply not expert programmers.  I have tried to teach
> Mathematica to many of them.  The fact that they read the manual and
> still don't "get it" tells me that the program is hard to use.  I have
> been evangelizing Mathematica for years, but find that these people
> typically tinker with it and then fall back to their old tools because
> they encounter so many error messages.

I understand that your side is trying to argue from extremes, suggesting
that a bunch of illiterate dummies who don't know algebra are naturally
going to have problems with Mathematica.

Yet I am not anywhere near that extreme, not even in the same
hemisphere, maybe not on the same planet.  The people I am talking
about do know algebra.  They have (1) the need, (2) the technical
education, and (3) the desire to learn how to use Mathematica.  They
still fail, just like the true dummies that you have invented for
debating purposes.  They have no problems with a calculator, which -- I
remind everybody -- does return the value of "a" when you ask it for
"Sqrt[a^2]," as expected.

Right now, Mathematica assumes that any symbol is complex.  I think it
would equally legitimate to assume that any symbol is real unless
declared otherwise.  Suppose the program assumed that all your
undefined symbols were quaternions (generalizations of complex
numbers).  You would say "Egads!  When do I ever deal with quaternions
in my work?  Can't I declare these symbols as complex and get on with
life?"  If you replace "complex" with "real" and "quaternion" with
"complex," then you have a portrait of many everyday users of
mathematical software.  I know a few mechanical engineers who last saw
a complex number in college.  They build some astounding machines with
their calculators, you know, and reject complicated mathematical
software time after time, with good reasons.

It's wonderful that Mathematica has so much more power than a
calculator.  But allow me to argue from the other extreme.  If it is so
difficult to use that only crack programmers who spend all their time
writing LISP and don't have a social life can draw on this power with
efficiency, then let's use our calculators.  We'll get the answers


Bruce Alan Fast wrote:
>  Mark Evans persists,
>  > The analogy of the calculator has been misunderstood....
>  We understand just fine, thanks.  Most people who use a
>  calculator know something about basic arithmetic, so they
>  know what to expect doing a few basic operations on decimal
>  numbers.  Most people who use a word processor are literate,
>  they've read a book maybe, and much of word processing is
>  nothing beyond what you see painted on the keys.
>  I am astounded at the attempt to compare calculators and
>  word processors to advanced software.  Even in the case of
>  word processors, you ignore the great difficulty that many
>  people have as soon as they have to use the mouse, or menus.
>  > We exhibit technical condescension when we see one novice after
>  > another ask the question, "How come Sqrt[a^2] doesn't yield a?"
>  > and forever return the stock answer that We Must Accomodate the
>  > General Case without thinking about whether, indeed, the novice
>  > has a decent right to expect "a" as his answer, or should have
>  > the means to declare a in such a fashion that "a" becomes the
>  > answer to "Sqrt[a^2]."
>  People who don't know basic algebra -- and that is most people
>  -- perhaps should be using a calculator instead of Mathematica.
>  Think seriously for a minute about the consequences of having
>  any "mathematical" software actually return "a" for "Sqrt[a^2]".
>  When you say that we don't think "about whether, indeed, the
>  novice has a decent right to expect "a" as his answer", you are
>  making a huge assumption.  It has been thought about **plenty** --
>  but rejected in the end, time after time, with good reasons.
>   Bruce Fast                        Bruce.Fast at Colorado.EDU
>   Information Technology Services   (303)492-8995
>   University of Colorado            Boulder  CO  80309-0455

  • Prev by Date: Re: Help with findroot
  • Next by Date: Re: Problems with Limit, Log, E
  • Previous by thread: Not so Useful Dumb User Questions
  • Next by thread: Re: Re: Not so Useful Dumb User Questions