BewaConsensus Science

>
> > --- In ssi_list@... "Arthur P. Smith"
> >> If you're expecting certainty from science, you're barking up
> >> the wrong tree.
> >
> > No, I'm expecting more substance to be attached to what is
> > published and then dissemenated. Sloppiness like this feeds the
> > very deniers that more solid arguments and presentations would
> > suppress.
>
> Scientific publication should be understood as more like an
> interoffice memo than like a final press release. This is what I
> mean when I refer to public lack of understanding of how science
> works - lots of stuff gets published that is later shown to be
> wrong, not just based on new results, but merely on more careful
> analysis. You publish when you reach a point where you feel that
> you have something new to contribute to the ongoing scientific
> conversation, and peer review exists to make sure that your
> contribution is original and not trivially falsified. The next
> stage is more careful critical examination, which takes place
> *after* publication.

This is a very concise summation and this is the creative
destruction I referred to in my earlier post. I agree. The point as
I see it isn't to publish a magnum opus but to advance the art,
defend your position and contribute to the effort of falsifying
other's work. The goal is to weed out the weak arguments and I
suppose scientists shouldn't have thin skins if they're going to be
opening themselves to such a process of criticism.

My point has more to do with a subjective call about standards. The
study I referred to in "Nature," IMHO, doesn't meet the conditions
(as you note above) of "something new to contribute to the ongoing
scientific conversation, and peer review exists to make sure that
your contribution is original and not trivially falsified."

I don't see the utility in presenting results from running a
computer model and then predicting species extinction if the
conditions in the model come to pass. Why even bother running the
model and instead just come out and say if these conditions come to
pass species will become extinct. The error laden computer model
doesn't add any authority to the statement. So why does the
inclusion of the computer model make the predictions *scientific*?
I'd much rather see results on how closely the model represents the
climatic cycle.

Perhaps this has to do with the urgent need to publish in order to
advance one's career. A market always expands to fill a demand.
There certainly are a heirarchy of journals for every science
discipline and I'd be surprised if most scientists can stay abreast
of ALL the published work pertinent to their field (marginal
journals included.)

As the signal/noise ratio decreases, standards fall and even though
there is peer review at each journal, what passes muster does a
disservice to the very standards you describe above.

For those scientists who are employed on a university campus, they
must in their heart of hearts wonder about some of the disciplines
they see being practiced on campus. Most social sciences aspire to
the rigor of the physical sciences, and they do follow the same
publishing paradigm, so doesn't much of that activity seem like
useless spinning of the wheels?

I contend that across the spectrum of human knowledge there is no bi-
modal distribution of rigor with all of the physical sciences
clustered together at one end on a scale of rigor and standards and
all the other fields of study clustered together at a lesser level.
Rather, I think there is some type of linear distribution to this
application of rigor and standards. My impression is that
environmental sciences, due in large part to the difficulty in
understanding the complexity of the systems that are the focus of
their study, fall far short of the rigor found in physics and
chemistry and are down at the level of, let's say, clincial
neuropsychology and macroeconomics, but above the level of sociology
and cultural anthropology. If this is indeed the case, then the
predictive value of their work should be held to the same standards
as comparative fields. I don't think that this is currently the case
and that they are afforded more respect than they've earned from
demonstrating the predicitive value of their science.

Simply, ISTM that the environmental sciences are punching above
their weight.

> That's why consensus is important - the consensus *is* science,
> more so than any individual paper. Scientific truth emerges out
> of the interaction of many individual scientists, and papers are
> the medium of that interaction. It's quite possible that the
> consensus is wrong, but more often than not (much more often) it's
> right.

What backs the consensus? That's the important distinction. Maybe
you've got your back up about this consensus issue because you have
to deal with cranks (I hope I'm not one :-) who use Crichton's
consensus argument as justification for their outlandish theories.
If there is evidence that withstands attempts at falsification and
the comunity reaches consensus on its utility, then I contend that
the need for the community to stress its consensus is minimal
because the facts and data can speak for themselves and the
consensus is that scientists are saying "we all agree on the
thoroughness of our understanding". Conversely, I'm leery when
consensus is trotted out front and center to substitute for the lack
of rigor in the field. This form of consensus is a political
statement.

Here is an excellent essay on consensus that examines the role of
cranks and their attacks on consensus and puts them in their place.
http://www.huppi.com/kangaroo/L-consensus.htm

Perhaps that is where you think I'm coming from. I'm not. You'll
have to parse my argument in more detail to understand my concern. I
think consensus is too widely bandied about and should be reserved
for matters that are solidly supported. There is compelling evidence
of anthropogenic causation of human causation in global climate
change thus a position of consensus is supportable. The extent of
the causation and its long range effects are less supportable, thus
when consensus is trotted out it is a political statement.

There was wide spread consensus that homosexuality was a mental
disorder and then "In 1973, the Board of Trustees of the American
Psychiatric Association voted to delete homosexuality as a mental
disorder from the seventh printing of the second edition of the
Diagnostic and Statistical Manual of Mental Disorders, or DSM-II."
http://www.priory.com/psych/disparat.htm

The case above is illustrative of a field lacking the rigor of
physics and thus more reliant on subjective standards and therefore
the field is more prone to political basis for reaching consensus.
The same is the case for environmental sciences. Measuring levels of
pollution and observing a relationship between such levels and
demonstrable effects on the environment is one thing. Determining
the carrying capacity of the environment and the ability of numerous
and disparate environmental cycles and their various interactions to
ameliorate our negative influence is wholly another. In the former
case a position of consensus is supportable, in the latter case, we
simply don't know enough and to posture as though we do and to
advocate the adoption of the precautionary principle becomes a
consensus based on ideology and politics and is absent specific
evidence that can support the consensus.

Now, having said the above that doesn't mean that there should be no
public policy to ameleorate our impact on the environment. We should
be taking such measures and we should acknowledge many different
approaches that each seek in their own way to address the issue.
Lending a scientific imprimatur to a political compromise is what
I'm uncomfortable with. Science should clearly acknowledge what it
knows, what is doesn't know, and what it speculates on. The
environmental sciences have abandoned such clarity.

I wonder if the religous aspects of environmentalism, which I think
we on this list have all acknowledged do exist, has a spillover
effect on environmental sciences. Perhaps an affintiy to such a
world view serves to guide some people into the field. I'm not aware
of any other scientific discipline that has this type of related
religious movement as a close parallel that sits across the
epistomological divide. Perhaps in days of yore, people would study
astronomy so that they could search for the face of god, but do
teenagers today hold plasma as being a wonderous state of matter and
then strive to study plasma physics so that they can achieve a more
spiritual bonding with this different state of matter. I'm curious
about the polluting *sub-conscious* influence of mysticism on the
scientific method and tolerance for dissenting opinion that
challenges the orthodoxy. Yes, there are cranks and industry hacks
that launch spurious attacks, but surely shining a bright light on
such idiocy should solve that problem. More problematic for me is
when there are people of good conscience who point out that
consensus is advancing faster than warranted by the evidence and
they are attacked rather than debated. That rings more true of
religious behavior and lends support to the hypothesis that, when
push comes to shove, the primary allegiance may be to philosophy
rather than method.

> Politicization of science is a real problem, but it's a problem in
> large part because people don't want to wait until conclusions are
> firmed up. Even with moneyed interests actively working to
> subvert the system, the truth will eventually come out. The
> problem is that the more active and aggressive the interests are
> in subversion, the longer the timescale before the truth
> eventually emerges, and in some cases that's just too long.

You know, from my reading of this debate I don't think we're that
far apart in opinion. I think I'm just doing a poor job at conveying
the point I'm trying to make but I'm unsure of how else I can get my
point across. Oh well.

I'm sure you've come across the Science Court idea before. I'd be
interested in your, and others', opinions on its usefulness for
science that is to be used to guide public policy?

TangoMan