Scheme - Foundations of scientific knowledge (was Re: Final SRFI 49: Indentation-sensitive

Author

Foundations of scientific knowledge (was Re: Final SRFI 49: Indentation-sensitive

Anton van Straaten

2005-07-31, 10:02 pm

alex.gman@gmail.com wrote:
> If you had actually studied inductive reasoning, you'd know that you
> are making a classic mistake. It's so classic that it's explained in
> any text on the topic worth a dime.
>
> Instead of calling you an idiot, I'll illustrate where you are wrong by
> a simple example. If you measure the unique density of a sample of
> water, you'll find that it's about 1000 kg/m^3. Given sufficient
> knowledge of liquid matter, you can induce that all samples of water
> have this unique density. Now, I got news for you: Schemers are not a
> homogeneous liquid. By looking at a tiny sample of Schemers, you can
> not possibly know what all Schemers do or think.

You're correct that inductive reasoning alone is suspect as a way of
reaching conclusions. However, that's not what I'm doing. I'm not
concluding anything purely from looking at a tiny sample of Schemers.

Rather, my claim that no Schemers are using SRFI-49 rests on the premise
that SRFI-49 is effectively unusable, and is supported by evidence that
what little experimental use it ever had, even by its author, has been
discontinued. In addition, my claim is easily falsifiable (which makes
it a strong claim from a Popperian perspective), but no evidence has
been found to falsify it.

The falsification point bears a close relationship to the question of
inductive reasoning, which is why I raised the latter in my previous
post. Even though we can't use inductive reasoning alone to reach
conclusions "known" to be valid, inductive reasoning is still an
indispensable part of science and epistemology in general, in at least
two ways:

First, inductive reasoning is one of our most important strategies for
discovering hypotheses, or possible theories. Having noticed some
pattern, we may then s

reasons for it, and develop hypotheses to
account for it. In the present case, noticing a distinct lack of
discussion anywhere from users of SRFI-49, and negative comments from
those who have examined it, we may look for some hypothesis to account
for this.

We then need to test our hypotheses to determine which one(s) best fit
the evidence. This is the second way in which inductive reasoning is of
critical importance: even once an hypothesis has been elevated to the
status of a theory, all such scientific theories depend on induction in
a way that is ultimately uncertain. Once we are sufficiently confident
in a hypothesis to call it a theory, we believe that its conclusions
hold, but we cannot usually test that its conclusions hold in every
possible situation. Instead, we rely on inductive reasoning,
essentially saying that because no tests so far have contradicted the
theory, that the theory holds, and therefore it's likely that future
tests will also hold.

It's the combination of a theory, with testable consequences, and the
fact that tests of the theory do not falsify it, which allow us to
develop some degree of confidence in a theory.

You've essentially been telling me that despite lack of falsification so
far, I don't have enough information to conclude that my theory is
valid. Faced with such a situation, the reality is that social factors
come into play: what conclusions do other people examining the available
theories and evidence come to?

Over time, a combination of consensus and continued lack of
falsification may lead us to develop stronger and stronger faith in a
particular theory. But there is no certain guarantee against the
possibility that at some point in the future, a previously successful
theory will be falsified, in which case it will hopefully replaced by a
theory that better fits the available evidence.

Note that this is true even of the water density example quoted above:
the example refers to "sufficient knowledge of liquid matter", but that
knowledge is based on theories, and those theories depend on the same
sort of inductive reasoning related to non-falsification which I've
described above. It is possible that our theories about liquid matter,
or matter in general, could turn out to have missed some corner case,
and that the density of a sample of water could vary under some
circumstances which we haven't yet discovered or tested. This sort of
thing has happened repeatedly in the history of science[*], and we can
be quite sure it will happen again (inductive reasoning tells us that!)

Note that in this post, I've described the situation mostly in Popperian
falsificationist, hypothetico-deductive terms. There are other ways to
analyze it, but none of them put scientific theories on any firmer
footing. In the end, it all depends on inductive reasoning.

As David Hume apparently put it, someone who insists on sound deductive
justifications for everything would starve to death.

Anton

[*] See Thomas Kuhn's "The Structure of Scientific Revolutions"

Anton van Straaten

2005-07-31, 10:02 pm

Matthias Buelow wrote:
> Anton van Straaten <anton@appsolutions.com> writes:
>
>
>
>
> Why the agelong discussion about it? If it's been abandoned even by
> its author, why not just quietly delete the nonsense from the list of
> SRFIs and leave it at that, then?

The discussion I've been having with Alex isn't so much about SRFI-49
itself. Alex has disputed the validity of many of the claims I've made,
and as a result I've been explaining my position, quite patiently I hope.

Feel free to ignore these messages if you don't like them. I imagine
you won't be the only one. Along those lines, sorry if my change of
subject line tricked anyone into reading my previous post; that wasn't
the intent.

As for deleting SRFI-49, afaik the SRFI process doesn't allow for that.
The author could have withdrawn it, but due to diligent efforts on
the part of the SRFI editors and others, the author of the SRFI was
recently contacted and chose to finalize it instead.

Anton

Ray Dillinger

2005-08-01, 4:10 am

Anton van Straaten wrote:

> The discussion I've been having with Alex isn't so much about SRFI-49
> itself. Alex has disputed the validity of many of the claims I've made,
> and as a result I've been explaining my position, quite patiently I hope.

I suggest laughter instead of explanation. It feels much better,
and it doesn't give me the feeling of wasting effort.

Bear

Anton van Straaten

2005-08-01, 4:10 am

Ray Dillinger wrote:
> Anton van Straaten wrote:
>
>
>
> I suggest laughter instead of explanation. It feels much better,
> and it doesn't give me the feeling of wasting effort.

Well, I've been deriving some amusement out of this discussion as it is,
but perhaps that's not as apparent to onlookers as it could be. I'm
easily entertained, in some respects.

Anyway, for the sake of the other readers of this otherwise rather quiet
newsgroup, I will endeavor to restrain myself. I will just have to
contain my disappointment that I haven't yet had the opportunity to
introduce quantum mechanics into the discussion, perhaps linking it to
the probabilistic nature of the foundations of our knowledge.

Anton

Anton van Straaten

2005-08-02, 5:03 pm

alex.gman@gmail.com wrote:
> I don't know about a novel, but if I had to turn this thread into a
> play, it would go like this

Once again, you resort to a fictional world in order to achieve the
result you want. Reality is so inconvenient.

Chachie

2005-08-02, 10:02 pm

Why, making up fairy tales about people really doesn't seem "civil", alex.

I think we're all surprised that a characteristic and unhypocritical
Schemer like you wouldn't know that.

alex.gman@gmail.com wrote:
> I don't know about a novel, but if I had to turn this thread into a
> play, it would go like this
>
> (all characters are fictional and any resemblence is accidental)
>
> Dillinger: SRFI 49 is really stupid, Moller is an idiot, I will not use
> a Scheme that supports SRFI 49 as an option
> Goldman: It's not nice to call someone an idiot (who clearly is not)
> over different syntax tastes
> Dillinger: But I really really really hate this syntax. How about that?
> Besides, Moller called me a troll in response.
> Goldman: still, it's just syntax. any program in SRFI 49 is completely
> isomorphic to a regular Scheme program.
> Dillinger: OK, I shouldn't have called the author of this extremely
> stupid proposal an idiot
> Straaten: Dillinger was speaking for an overwhelming majority of Scheme
> users
> Goldman: most Scheme users don't call their colleagues idiots when they
> propose a new syntax
> Straaten: no single Schemer uses SRFI 49
> Chachie (puppet): no single Schemer uses SRFI 49
> Felix: you could not possibly know that
> Straaten: true, but what if I say that it's my SCIENTIFIC THEORY that
> no single Schemer uses SRFI 49; now it's a fact, a scientific fact!
> Goldman: not exactly
> Dillinger: let's laugh at him; the more we talk about whether Straaten
> is omniscient, the less we remember what started this discussion.
> Straaten: I enjoy being publicly humiliated by other men. Therefore,
> Moller should like being called an idiot too.
> Clinger: I enjoy horses and romantic walks in the park.
> Marshall: My beard is too long for a guy without a tenure or $1,000,000
> in the bank.
>

Sponsored Links