Re: [LogoForum] Digest Number 1662

>>>>> "Brian" == Brian Harvey <bh@xxxxxxxxxxxxxxxxxxxxxxx> writes:

Brian> Andru Luvisi <luvisi@xxxxxxxxxxxxxxxx> writes:
>> Current graphical interfaces prevent kids from understanding
>> how to use computers, and instead kids should use Unix commands
>> to get a better understanding.

Brian> Okay, if *you* argue in favor of teaching arithmetic
Brian> algorithms, I'll take your argument seriously, because
Brian> you're consistent. :-) (Sort of like the way I feel about
Brian> anti-abortion people depending on whether they're also
Brian> anti-death-penalty and anti-war.)

Oddly enough, in places where calculators are cheaply available, I'm
not all that in favor of teaching arithmetic algorithms (at least as
they're currently taught) because, as you say, they usually aren't
taught in a way that helps the kids understand what's really going on.
They're these little black boxes that don't enlighten any more than a
calculator does. My father was a decent teacher, so when I was
learning fraction division or multiplication of multi-digit numbers in
school, he was able to help me understand what was actually going on,
but I suspect that most kids never had any idea why you "flip over"
one of the fractions in division or why you "move to the left one
space" for each digit in multi-digit multiplication.

I still remember coming home in the fifth grade confused about
fraction division. They had told me something like:

A C A*D
.
- - - = ---
.
B D B*C

I had no idea what on earth was going on.

My dad wrote down:

A
-
B
---
C
-
D

Then he multiplied the top and bottom by D and multiplied the top and
bottom by 1/C, and it all made sense.

Although I could do long division right away, it took me years to
understand how it works, and even now it would probably take me a
while to construct a proof that it actually gives the right answer. I
can't remember the last time I used it. I normally round off and do
approximate division in my head. If it's something I need an exact
answer on, I use my computer.

I think that teaching the arithmetic algorithms would be useful if
children were taught what is actually going on. However, I'm not sure
if most teachers understand what is actually going on, which would
make teaching it difficult.

I know several adults who just carry a calculator around with them all
the time, including my wife and mother-in-law. When we're out
shopping, they can work out price-per-pound and things like that much
faster than I can. I used to feel like they were cheating, but I
don't feel like I'm cheating when I use a one-line bash command to
lower case a thousand filenames while a coworker spends two hours
doing it by hand. Now I think it's just intelligent use of tools.

Brian> I completely agree that a kid who has gotten into computers
Brian> and wants to understand them through and through should
Brian> learn Unix. Our CS majors, for example, use Unix shell
Brian> commands. But I think you have to admit that many, many
Brian> people who would not be computer users at all if they had
Brian> only the Unix shell as an entry point have been able to
Brian> enter the computer-using world because of the WIMP
Brian> interface.

That's a question I've wondered about and argued with people about for
years. Unfortunately, there's no way to tell for sure. I've known
several people over the years who used DOS programs or VMS programs
and weren't "computer people". They knew how to do the tasks
necessary for their job, and that was it. They knew nothing about
what those commands were actually doing. I'm not sure how much
different the current state of affairs is where so many people just
know which buttons to push. I remember some advertisements on
television for computer training videos that would teach MS Word and
things like that where they said "It's easy. We just tell you where
to click and what to type."

I do find it interesting that the GUI was developed as a way to make
it easier for children to create their own programs, and was deployed
in a way that ended up making it harder for anyone to create their own
programs. I can't help feeling that something went terribly wrong
somewhere.

I watched the webcasts of your SICP lectures a few semesters back and
I really liked the three graphs you drew on the board comparing
"sophistication of the task" with "wizardry required" for Unix,
Mac/Windows, and the Alto or Lisp Machine. As I recall, they looked
something like this (sophistication on X axis, wizardry on Y):

Unix (Initial learning curve, but it's pretty gradual after that):


|
| ..........
| ....//
|../
|
|
------------------------------

Mac/Windows (Easy at the start, but then you run into this huge wall):

|
| .........
| ....../
| /
| |
| |
| |
| |
| |
| |
|............/
------------------------------

Alto/Lisp Machine (Starts out easy and grows gradually):

|
| ..........
| ....//
| .../
| ....//
|../
------------------------------


I've done some playing around with Squeak (http://www.squeak.org/) and
now that a CADR emulator and source code are available
(http://www.heeltoe.com/retro/cadr/index.html
http://www.heeltoe.com/retro/mit/mit_cadr_lmss.html) I may play with
that some day. Were you involved with the Lisp Machines at all or had
you left by the time all that happened?

Andru
--
Andru Luvisi

Quote Of The Moment:
"If you give someone Fortran, he has Fortran.
If you give someone Lisp, he has any language he pleases."
-- Guy Steele
.

Relevant Pages