On Mar 16, 8:56 pm, John Doty <j...@[EMAIL PROTECTED]
> wrote:
> But to teach what you want to teach you must demand the student avoid
> using his most powerful mental "muscle".
You are still trapped in the negative. If verbal reasoning is the
student's strongest mental "muscle", they really ought to work on
strengthening the others. And in any event, forcing Forth into the
mainstream tendency to over-emphasize one learning style on the
grounds that its been labelled as using the "strongest mental muscle"
makes no strategic sense.
> It's a programmer's idea of teaching: you're programming the human to
> serve the computer.
You are arguing to your preconceived conclusion rather than arguing
from the point you are addressing ... forcing people to work around
the limitations of a specialized tool when they are trying to
accomplish a task for which it was not designed is putting the first
priority on the tool.
> It's better to program the computer to serve the human.
And so you want to take the language where I have been most effective
in getting the computer to do what I want, and make it more like
languages where I have been less effective in getting the computer to
do what I want and have been forced to devote more time to working
around the way that the programming language wants to do things.
This is an examples of trying to force all learning styles into a
single framework based on an _a priorit_ prejudice in favor of one
learning style.
> > Learning the limits of something comes after learning its
> > possibilities.
> > And further, a computer programming language that gives your the
> > freedom to create a programming language with the syntax that you wish
> > it to have, rather than forces you to work within the syntax that the
> > implementer of the language decided you would work under ... that is
> > tapping a dimension of our human language instinct that no language
> > with a fixed syntax can. If humans were merely syntax users, and never
> > syntax creators, our natural languages would not be in a state of
> > constant evolution and flux, and all of our languages would exhibit
> > syntax close to our normal underlying language instincts, instead of
> > our learned languages displaying a quite remarkable diversity and
> > variety of quite extreme departures from our natural language
> > instinct.
> Yes, but that evolution is very slow and conservative.
This is a counterfactual statement. The speed of that evolution varies
widely depending on social setting, size of language group, and
factors such as literacy in the language ... literate languages spoken
among a very wide group tend to evolve slowly and conservatively, and
to a certain extent seem to be evolving by losing complexity and
structure ... non-literate languages spoken among small social groups
with an incentive to not be widely understood by eavesdroppers from
groups can evolve with dizzying speed.
> Few individual speakers ever change a natural language: changing English
> willy-nilly for every new situation that arises would be pure Babel.
Yes, in settings where languages tend to evolve more rapidly, there
also tends to be a larger number of them each spoken by fewer native
speakers. However, innovations certainly occur, even in a very slowly
evolving (some would say decaying) language like English or Han
Chinese.
> But to claim Forth applications are application-specific languages is an
> exaggeration:
This is a mis-characterization of the argument. To say that Forth
gives the ability to write an application specific language is quite
evidently not to say that every Forth application is the writing of a
new language.
> Forth generally creates application-specific *vocabularies*, not full
> languages.
Forth general creates nothing whatsoever. Its only people using Forth
that create things. And further, much of the time the Forth programmer
is not creating the language, but only implementing the language in
Forth. JenX, for example, did not create XML ... someone else did
that. What JenX provides a framework for executing XML as source.
> That's not so dangerous, and wrapping a truly linguistic framework
around
> the process would be a good thing.
This is just a claim. I have made the counter-claim. I have seen no
evidence on your part that the problem of interference between the
syntax of a programming language with restrictive syntax and the
syntax of the target language is going to disappear if the "right"
syntactic framework is found. You seem to be begging the question as
to whether such a "right" design of a substantially larger set of
restrictions that is not, in fact, constraining, in fact exists.
> > It should be evident that any language that is going to excel in
> > giving programmers the ability to write languages where "how you say
> > it" can be tailored to the problem ... is itself going to be far more
> > focused on "how you do it" than "how you say it".
> It should be evident that this is a recipe for confusion. "Write-only",
> "unreadable", ...
The freedom to do something well is the freedom to do something badly.
That's what we in the profession call a "trade-off".
The solution for the Forth position on that continuum is to provide
better sup****t for more people to write more Forth code and share it
more widely, so that the small number of excellent designs that result
have a better op****tunity to propagate.
|
