Jonah Thomas wrote:
>
>Guy Macon <http://www.guymacon.com/>
wrote:
>
>> 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 is a proper
>> subset of the null hypothesis alternative to the failings in
>> the language hypothesis. This is why the argument that there
>> is no other hypothesis and therefor the failings in the language
>> hypothesis must be correct is an invalid argument.
>
>But this is not the way science is actually done.
>
>Consider the following widespread scientific argument:
>
>Quantum mechanics always returns the correct answer for every scientific
>experiment.
>
>With no fudge factors or error bars, QM sometimes provides the correct
>answer to 20 decimal points. And it never ever fails.
>
>Therefore QM is correct and any alternative must be wrong.
>
>This argument is often made, minor variations on it are (in my
>experience) made whenever and wherever someone suggests that there might
>be an alternative to QM. This is how science is actually done, as
>opposed to standard scientific method.
I have worked with many scientists, and none of them argue as you
have done above. Engineers, maybe. Scientists, no. The closest
to the above that you will find among scientists is something like
"I have looked at all the proposed alternative theories as well as
the null hypothesis and found them all to be weak compared to QM.
Someone may come up with a better theory tomorrow, but right now
QM is the best theory we have."
Including the null hypothesis is key here. If QM makes a prediction
that is correct to 20 decimal places, the null hypothesis is that
this is a coincidence, which isn't likely. In the case of JD's
"failings in the language" hypothesis, he hasn't predicted a result,
but rather attempted to explain an already known result, and he isn't
talking about being correct to 20 decimal places, but rather about
a binary value [ becomes popular | doesn't become popular ]. That
makes the null hypothesis (that some languages become popular by
sheer luck) a far stronger contender. And, of course, several other
good alternatives to the "failings in the language" hypothesis have
been posted here.
--
Guy Macon
<http://www.guymacon.com/>
|
