Re: The Promise of Forth




John Doty wrote:

Jonah Thomas wrote:

OK, what's the confirmed hypothesis here? I'll say it in my own words
and see if John agrees that this is his meaning. He may not agree; I may
have misunderstood, and if he says he means something different then
I'll have a chance to understand better.

John says the only testable hypothesis that has been proposed, the
confirmed hypothesis, is:

Forth as it is used is inadequate for professional programming because
it doesn't well enough fit the way real people think.

It is not adequate
as a *language*. This is confirmed by the fact that not many people use
Forth, the fact that Forth has lost the footholds it had in some niches
that John knows about, and the fact that many people have heard that
Forth is unreadable,

and they have heard that Forth programmers can't
read their own code.

No matter how many alternative hypotheses have been put forth to explain
these three facts, John says there is only one hypothesis on the table.

Consider the alternatives that have been proposed:

1. The powerful institutions hypothesis.

The claim is that powerful institutions were behind Forth's lack of
success. But what languages did the powerful institutions favor? PL/I
and Ada. Not exactly dominant. C? C and Unix made AT&T powerful in
general-purpose computing, not the other way around.

2. The academic NIH hypothesis.

This is more historical revisionism. Forth was nurtured at NRAO, an
academic research institution. Forth was widely appreciated in academia
from ~1975-1985 (I was there). If there was any NIH syndrome, it worked
the other way: the Forth establishment rejected later improvements
(STOIC, MAGIC/L) that originated in academia.

3. The nothing hypothesis

This assumes that, despite Forth's advantages, the users who have
abandoned it had no real reason to, and are simply idiots. Sorry, some
of them are extremely smart people.

I find it interesting that when I did an analysis of your logic
a while back you got real quiet all of a sudden, but here you
are once again trying to make the same points that I refuted.

I also find it interesting that you have once again ignored
the null hypothesis, and instead substituted a strawman that
you call "the nothing hypothesis."

At first I found it strange that someone who works with
scientists would be ignorant of the null hypothesis, but
now that I see you pretending that nobody mentioned the
null hypothesis after I explained it to you in detail
(Null Hypothesis, meet John Doty. John Doty, Meet Null
Hypothesis...) I can only conclude that the omission is
deliberate.

Let's review, shall we?

So far, at least *THREE* people have told you:

"The effects you cite are not *necessarily* the
result of the cause to which you attribute them." -ER

"This conclusion does not follow from the premises." -AH

"you simply assume without evidence that A leads to B,
never ever examining whether that assumption is warranted."-GM

....and yet the only argument that you can come up with is
"there is no alternative hypothesis."

According to standard scientific method, when considering the
merits of the "failings in the language" hypothesis, the null
hypothesis is all other explanations, known and unknown,
combined.

Thus the "historical accident" hypothesis (the hypothesis
that some languages become popular by sheer luck) -- which
you conveniently left out of your list of "the alternatives
that have been proposed" above -- is a proper subset of the
null hypothesis and a clear alternative to your "failings
in the language" hypothesis.

This is why your argument that there exists no other hypothesis
and therefor the "failings in the language" hypothesis must be
correct is an invalid argument. You can't simply pit your
favorite hypothesis against all the others that you can think
of. You have to pit it against *all* possible reasons (known
or unknown), including the hypothesis that there is no cause
and effect relationship and that the correlation is simply the
product of random chance. You need to show some sort of
evidence that the effect is really due to the cause and not
just due to chance.


References to ignore:
http://en.wikipedia.org/wiki/Null_hypothesis
http://www.stats.gla.ac.uk/steps/glossary/hypothesis_testing.html#hypothtest
http://www.stats.gla.ac.uk/steps/glossary/hypothesis_testing.html#h0
http://davidmlane.com/hyperstat/A29337.html
http://davidmlane.com/hyperstat/A73664.htmlhttp://davidmlane.com/hyperstat/A72620.html
http://davidmlane.com/hyperstat/A73079.html
http://www.britannica.com/eb/topic-422222/null-hypothesis
http://sociologyindex.com/null_hypothesis.htm


--
Guy Macon
<http://www.guymacon.com/>

.

Relevant Pages

  • Re: The Promise of Forth
    ... John says the only testable hypothesis that has been proposed, ... If there was any NIH syndrome, it worked the other way: the Forth establishment rejected later improvements that originated in academia. ... According to standard scientific method, when considering the merits of the "failings in the language" hypothesis, the null hypothesis is all other explanations, known and unknown, combined. ... Thus the "historical accident" hypothesis -- which you conveniently left out of your list of "the alternatives that have been proposed" above -- is a proper subset of the null hypothesis and a clear alternative to your "failings in the language" hypothesis. ...
    (comp.lang.forth)
  • Re: Esperanto and Interlingua
    ... John> Esperanto isn't badly weighted down with English yet. ... Zamenhof put the language together from those he knew. ... the latter joining the Dutch Kingdom. ... French speaking part, with a few "joint" regions, mostly ...
    (sci.lang)
  • Re: Esperanto and Interlingua
    ... John> Esperanto isn't badly weighted down with English yet. ... John> was disagreeing with itself in lots of mutually ... virtually the same language. ...
    (sci.lang)
  • Re: UserLinux chooses Python as "interpretive language" of choice
    ... And that's a big plus for python. ... > John> point. ... > he regretted adding functional constructs to the language. ... lambdas seem quite ok with the anonymity of code blocks (given how ...
    (comp.lang.python)
  • Re: mine and I and me
    ... no one correct syntax. ... Audience, formality, perception, those and other notions ... come into play making for a difficult language task. ... sentence could be viewed as two unique tasks; John has his ...
    (alt.usage.english)