Morning quote.

I thought this bit was well stated and highlights a frequent error you see in science journalism.

"Warning klaxons should sound whenever alleged scientists" [or science journalists -- ed.] use the word 'consensus.' What we’re looking for is 'evidence' or 'experimental results,' not 'consensus.' I believe you could easily get a consensus among scientists that 'my field of study should be better funded' or 'hot members of the targeted gender should find scientists more attractive' or maybe 'huge, throbbing foreheads should be the key marking of sexual desirability.' Whatever. But nothing in a consensus indicates anything about that consensus being backed by science – only backed by *scientists*, a subtle but critical distinction."

(from a post at Yard Sale of the Mind -- here: http://yardsaleofthemind.wordpress.com/2012/04/27/science-distinguishes-hard-thinking-verus-faith-using-the-ever-popular-smart-versus-stupid-cleaver/ )

+ + +

The preponderance of the evidence pointing a certain way generally produces a consensus. But a preponderance of evidence is not the only phenomenon that can produce one -- no, not even among scientists.

As a lifelong geek who loves a good story, I count "history of science and technology" as my favorite nonfiction genre. Probably the biggest literary influence on my career choice was the collection of historical essays gathered together under the title _Asimov on Chemistry_, which I found in my public library when I was fifteen or so. It is a good thing I did not discover _The Making of the Atomic Bomb_ by Richard Rhodes until it was too late for me to major in physics. I am beyond pleased that James Burke's _Connections_ series remains among my children's favorite DVDs.

To be deep in history is to cease to be impressed by consensus qua consensus. It isn't the same as that disdain the public sometimes has for rapidly-changing headlines, such as you see as they cycle through the latest health-danger-warning ("so x is bad for you now too, huh? I remember when not getting enough x was supposed to be bad for you. Just goes to show they don't know what they're talking about"). Nor is it the same as the raised eyebrows that go with the often-too-swiftly connected dots between "Results" and "Discussion." (Although spotting the exact point along that path when a paper author steps across the boundary of his expertise is always a fun pastime.) It is not even just the awareness of how many times a consensus has been overturned, nor a philosophical rejection of the idea that a "science" can ever be "settled" and still remain science.

I think for me it is an interest in the paths to consensus. Every consensus has been arrived at by a certain sequence of events. When I was fifteen or so, I was captivated by Asimov's essay "Slow Burn," which could have been subtitled "How The Consensus Against Phlogiston Got Started," particularly the ending (Priestley and Cavendish did not live long enough to be convinced that phlogiston was bunk). I think it rather programmed me to wonder about the backstory-- all the time. Exactly how did we get where we are? Which are the key papers? Which personalities have championed which theories, and why? When some future, technically-adept storyteller sits down to pull it all together in a canonical account, one that (along with the technical insights, tests and conclusions) assesses the human strengths and weaknesses, quirks and mediocrities too, in all the key players, what tale will emerge?

Comments

[Barb]

Very interesting.

In medicine, the push for using only evidence based diagnostic tests and treatments keeps getting stronger and stronger.
But the "consensus" opinions are necessary for practical reasons.

If there is no good science to guide a particular medical decision, all you have is anecdotal evidence.

A consensus from the experts who have studied and reviewed the problem in detail can be very helpful until better evidence is available.

And by the way, I have wanted to tell you for some time how much I enjoy your blog. We are a homeschooling medical scientific family :-)

[bearing]

Barb,

Consensus is not necessarily an indication of *truth,* but it is *very often* an indication of *usefulness.*

Useful consensus-es (consensi?) are often the basis of an answer to the question, "What do we DO about this problem?" Which is why they are valuable in medicine.

But let's remember that clinical work in medicine is "applied" science -- it is not really akin to science per se, and is more akin to engineering. Which, like medicine, is a field in which consensus is often useful. The best practitioners of both, I hope, know the difference between "true" and "useful."

[Joseph M]

No one would argue that consensus is not often useful, or is not often the best thing we can get. BUT - consensus is not *science*. There is a scientific consensus that water boils at 212F under normal conditions at sea level - but the consensus flows from repeatable, well-controlled experiment.

So, the warning is: if scientists lead their argument with claims of consensus, rather than with compelling evidence or argument (that would lead to a consensus), then we should be wary.

BTW: Nice blog. Thanks for the link.

[sara]

so I presume you've read and enjoyed Oliver Sack's memoir "Uncle Tungsten." It was our book club selection last month--before that, being an English major, I'd never heard of Pholigiston

[bearing]

Sara: Indeed!

[Joseph M]

Always wanted to see the headline:

"Phlogiston Blamed in Antique Shop Fire"