This is not
the case with a lot of the 'toy' programming languages
and environments where you do have to re-learn things when you move
on to other languages and systems.
Yes and no.
You have to relearn details. To pick a simple example, going from
Pascal to C, you have to relearn syntax ({ } instead of begin end, for
example).
Sure, but as you ho on to say, the concepts are the same.
If you learnt about valves (vacuum tubes), you would ahve to re-learn
soem details if you wanted to use bipolar transistors (transistors come
in both polarities, there are no valves that use positrons rather han
electronics ; bipolar transistors are current controleld, valves are
voltage controlled; etc). But the concepts of amplidiers, oscillators,
impedance, ohms law, krichoff's laws, etc do not have to br re-learnt.
But, if the toy language/environment has done its job,
that's all you
have to relearn. To continue with the Pascal->C example, you don't
need to relearn what a variable is, how an if statement works, etc.
What a procedutre is, what a function is (OK, C doesn't have procedures,
but I think a void function is essentially the same thing), what formal
parameters are, loops, recursion, etc, etc, etc. You don't have to relarn
anything like that.
Furthermore, you don't have to forget the old way. Learning C doesn't
necessarily involve forgetting Pascal.
My views are totally opposite to most teachers
(on the other hand
they have worked well for me) in that uyou should _never_ learn a
simplified viersion that had to be un-learned later..
As phrased, I agree with that. But the last six words are very
important. (Seven, if "un-learned" counts as two.)
Of course.
You ahve to start with a simplified version. No sane person tries to
learn all the keywords and syntax of their first computer language at the
sart. Jsut as nobody tried ot learn the entire vocabularly of a human
language at the start :-)
[...]
As a somewhat contrasting case, I learnt Newtonian mechanics before I
learnt relativistic mechanics. Newtonian mechanics _are_ incorrect,
yes, but they are not to be unlearnt; rather, they are to be recognized
as an approximation, useful in many cases, as long as you know the
limits of the approximation's applicability. (I was also taught from
the beginning that Newtonian mechanics _are_ just an approximation.)
And of course relativeity, even genraly relativity is an appoximation
too. It does not apply to very small objects. You need quantum mechanics
for that.
All phycial theopries are (IMHO) approximations. This does not make them
non-uyseful, or not suitable to be taught. Heck, fo jsut abotu everythign
I do now, newtonain mechanics is more than accurate enough.
All to say that I don't think learning simplified versions necessarily
means unlearning later. When it does, then yes, I agree with you - but
Nor do I, hence those 'last 6 words'.
I also think it usually is possible to teach a
simplified version
without necessitating unlearning upon graduating to something less
simplified.
Agreed. But this started from 'toy' programming environments/languages
that do not teach the concepts needed to go further. And where yuu
essentially do ahve to forget everything you learned when you want to do
something more advanced.
Case in point. Compare the classic Philips 'EE' series of kits with 'Snap
Circuits'.
The Philips EE kits contains 'real' loose compoentns. Individual
resisotrs, capacitors, audio transformers, etc. The later series of
transsitors and all ICs were pre-mounted on little PCBs but with nothing
else. Jsut hte termianl brought out to make them easier to conenct to.
And you joined them together with spring terminals. The manual included
real circuit diagrams and theory sections. With the trivial exception of
how to use the spring termiansl, everything you learnt from those kits
carried over to when you started buying loose components nad solderign
them togehter yourself.
Now Snap Circuits. The first problem is that there are ridiculous number
of 'black box' modules. They are undocuemtned in the manual (yes, there
is some docuemtnation o nthe web, but that has to be hunted for). And
compoennts are used in very odd ways. Almost nothing you learn can be
carried on.
The later Philips EE kits included a circuit for an FM radio receiver.
You got to wind the tunig coils, they were resoianted with a varicap
diode, etc. You got to understnad how such sets owrk. The Snap Circutis
one seems to consist of an 'FM radio module', an 'Audio Amplifier
module', a speacker, and not much else. Oh well...
-tony