Looking on as a philosopher, I see the value of imaginative musings about
the might be, the techniques that haven't worked or might work (and
haven't worked for this but give a hint that might be appropriate for
that), etc., for the knowledgeable, those with the background to evaluate
the material. But I also consider what a treasure trove such material
would be for the nitwits with an axe to grind. It's tough enough to deal
with the quacks and true believers now. What would it be with the input
of all the half-baked and unchecked musings produced by even the best of
thinkers and experimenters?
I understand that there is a sharing of preprints and guesses among the
small number of workers in some of the specialized areas in physics. I
don't know to what extent this may be true in other disciplines.
Unfortunately, this seems to work only for an inner circle, not the
broad-based need for information that Wayne implies.
Is there any way for a scholar to get all the information that is
relevant to the task at hand and its total context? I recall reading many
years ago that, if a person spent all his working hours simply reading
the life science journals, at the end of a year he would be some three
months behind. The number of journals and the information put out has
increased amazingly since then. There was no comment on how to remember
all that information. I also recall that a nursing program specified
that, if the student had taken the required biology course more than two
years earlier, it had to be repeated--the material had changed that
radically. I note further the need for bioinformatics, people able to
design programs that will extract the relevant and present it in an
understandable form so that it can be used. Looks like information
overload is here to stay.
Dave
On Sat, 20 Oct 2001 10:38:28 EDT Dawsonzhu@aol.com writes:
Tim Ikeda wrote:
Most likely, >85% of this unpublished 90% is crap, but with search
engines, is there now enough storage and filtering capability to locate
gems and justify the effort? And how? Online notebooks? An E-journal of
Negative
and Partial Results? Is there some form of editorial control and
publication
quality assurance that would work in a high-data, low information
environment?
I second your observation. For years I thought splicing
occurs for all intron dropped into splicing extract. From
the way the biology literature on the spliceosome is written,
you sure would think so. Long behold, it turns out that
long sequences don't always work. No surprise per se, but
it would be nice to know that rather than get a false
"just so" impression that biology and biochemistry textbooks
often seem to claim.
As to Howard's remark on peer review, reviewers are
able to evaluate good experimental technique. So for
example, It would be useful to know that procedures A,
B, C, etc., fail to crystallize a protein. Crystallization
procedures are known and can be assessed by a reviewer.
The information is also quite useful because it would
tell me not to use those techniques and may even be
indicative of some important function or family of
proteins I might be investigating.
I see three advantages to reports of null, incomplete,
or unsuccessful results.
(1) Knowing that a certain set of well known procedures
fail to materialize results would saves others from making
the same mistakes and would encourage them to focus their
attention on a different strategy. Hence, even if a solution
is eventually discovered, it would spare a lot of waste making
square wheels when someone already found out they don't roll.
(2) It also helps identify problems of technical interest
which can attract the attention of ambitious thinkers
who are willing to be challenged by difficult and possibly
unsolvable problems. Likewise since these problems may be
unsolvable, there are no clear "answers" and a different
kind of ingenuity is required.
(3) Finally, no result is in fact sometimes an important
result. Michelson and Morley is perhaps a good historic
example.
by Grace we proceed,
Wayne
This archive was generated by hypermail 2b29 : Sat Oct 20 2001 - 13:53:35 EDT