[FRIAM] IS: Rhetoric in scientific arguments WAS: FW: Fractal discussion Landscape-bird songs

[Eric Charles]

Glen,
To "Peirce-up" the discussion of doubt a touch. To doubt something is to be
unable to act as-if-it-were-true without reservation. So, for example, you
do not doubt Newtonian mechanics under a wide range of conditions (you are
willing to act as if it is true under many circumstances), however, there
are conditions under which you would be nervous acting as if they were
true. Your doubt is not absolute, and the "caveats" you refer to, could be
expressed as a description of the circumstances under which you start to
get nervous.

Your description of replication is good, but non-typical (particularly
among the chemists, which are Peirce's favorite scientist). We now think of
replication as part of the falsification process, but that is actually a
weird way to think about (a symptom of the degenerate state of many current
fields). The most natural reasons to replicate a research report is a)
because the outcome is itself useful or b) because you intend to build upon
it. For example, if someone publishes a novel synthesis for artificial
rubber, I would probably try to replicate it because I need artificial
rubber, or because I intended to start with that artificial rubber and try
to synthesize something new. Under those conditions, anyone trying to
replicate would be very frustrated by a failure (given some tolerance for
first attempts).

Nick,
I think Glen is prodding, in his second part at an extremely important
point, and one that I have been wrestling with quite a bit lately. It is
quite unclear why "trusting the work of other scientists" - as currently
practiced - is not simply another deference to authority. The current
resurgence of assertions that people shouldn't argue with scientists (e.g.,
that we should care what an astrophysicists thinks about vaccines, or what
a geneticist thinks about psychology) is bad, and "science-skeptics" are
not wrong when they attach a negative valence to such examples. Let us
ignore those flagrant examples, however. How do I determine how much weight
to give to a report in a "top journal" in psychology? Does it matter that I
know full well most articles in top journals turn out to have problems (as
flashy reports of unexpected results are prone to do)? And if I am
suspicious of that, what do I make of the opinion of a Harvard full
professor of psychology vs. a bartender who has been helping people with
their problem for 5 decades? Etc. etc. etc. I think there is a very deep
issue here, which I'm not sure I've ever seen explained well. I think, in
part, it is a challenge that was not as prevalent even 50 years ago, but
that may just be imagined nostalgia.







-----------
Eric P. Charles, Ph.D.
Supervisory Survey Statistician
U.S. Marine Corps
<echarles at american.edu>

On Wed, Mar 1, 2017 at 4:13 PM, glen ☣ <gepropella at gmail.com> wrote:

>
> Heh, your lack of social salve has left me unclear on whether I should
> respond or which parts to respond to. >8^D  So, I'll just respond to what I
> think is the most important point.
>
> >  That implies that what you say below supports arguments from authority.
> [NST==>I don’t think we can EVER escape arguments from authority.  Science
> is locked in a matrix of trust.  Doubt in science is really important, but
> it has to be relatively rare, or we would never know which of a million
> doubts to take seriously.  <==nst]
>
> I think you use "doubt" differently than I do.  Even if we replace "doubt"
> with "falsified", it's not a binary thing.  When I doubt something an
> authority says, I'm not refuting, denying, or rejecting it.  I'm simply
> expressing that the saying probably has caveats, some of which I might know
> about, some of which I might not.  The same is true of (critical
> rationalist) falsification.  Even though we know Newtonian physics isn't
> end-all, be-all True with a capital T.  It's satisficing in most
> circumstances.  To (When I) say it's been falsified simply means it has
> caveats.
>
> And, in this sense of the two terms, doubt and falsification are _rampant_
> in science.  When you try to replicate some other lab's experiment, you
> must doubt, say, the methods section in their paper, usually because you
> don't have the exact same equipment and the exact same people ... doubt is
> what promotes reproducibility to replicability. ... at least in this
> non-scientist's opinion.
>
> >  I.e. we can't treat a lack of salve as an assertion of objectivity
> without implicitly asserting that every statement without such salve is
> fallacious. [NST==>Yep.  All statements are more or less fallacious.  So
> does that render all statements the same?  If I flip the coin once and it
> comes up heads, what evidence do I have that the coin is biased.   None.
> If I flip it twice, a little.  If I get a hundred heads, the probability
> that the coin flips represent a population of fair coin-flips is finite,
> but vanishingly small.  I’ld bet on it, wouldn’t you?  All statements of
> certainly of that character.  <==nst]
>
> No, not all fallacies are the same.  Different statements are fallacious
> in different ways.  And the argument from authority fallacy, in my opinion,
> is the worst one because it's opaque.  You can't learn from it.  At least
> with, say, assuming the conclusion, it encourages us to understand the
> relationship between premises and conclusion... it helps us grok deduction
> as well as the host of concepts surrounding languages, formal systems,
> algebras, etc.
>
> Your inductive argument, by the way, isn't obviously an argument from
> authority (obvious to me, anyway -- see how annoying it is), particularly
> with (as someone recently phrased it) interpersonally assessable things
> like coin flipping.  Anyone with the usual complement of sensorimotor
> manifolds (!) can put in place the kernel and carry out the iteration.  The
> only authorities involved are whatever physical structures are required for
> coin flipping and counting.
>
> But, more importantly, self-consistency (local coherence) is the governor
> of induction.  Can you imagine if "successor" were redefined at each
> iteration?  So, it helps make my larger point that it's irrelevant what you
> or I believe or state with authority.  What matters is whether the
> method(s) hang(s) together.
>
> --
> ☣ glen
>
> ============================================================
> FRIAM Applied Complexity Group listserv
> Meets Fridays 9a-11:30 at cafe at St. John's College
> to unsubscribe http://redfish.com/mailman/listinfo/friam_redfish.com
> FRIAM-COMIC http://friam-comic.blogspot.com/ by Dr. Strangelove
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://redfish.com/pipermail/friam_redfish.com/attachments/20170302/d9617fdd/attachment-0002.html>