Reading about history of science is usually about a history of succes, often presenting it as a very straight "highway of progress" (this is at least what one person remember from, admittedly, usually "popular" science texts).
Is there any "history of scientific failures" as some sort of counterweight?
(Of course You may deny that such a history could have anything interesting to say, if you deny any such failures, but that may be questionable).
Of course, examples as "geocentrism", "Lysenko", and other are well known, but perhaps there are other, even more relevant to todays "mainstream"?
(they can be both seen as "special" or even "exotic" examples, and some may not se much relevance to contemporary science).
I think such a history, if written by the right author, should have a chance to be read by "ordinary people" (it should have a chance to be an interesting story, either written as a "history" or as a novel).

You're right, of course- mankind tends to tell history as a mostly unblemished record of success. Early failures or repeated trial-and-error are generally rolled into the eventual tale of success, because human history likes to view itself as a logical progression forward.

Check out the History Channel's "Great Blunders in History" series. It has showcased topics as "Great Bridge Disasters" and the Chernobyl disaster. Yes, since it is the Hitler Channel, there's a large portion of it that is WWII -related.

Indeed it does. History is littered with scientific mistakes and lies. Mistakes? The Challenger disaster I'm sure could be counted in this category. Lies? Well I'm on rocky ground with the moderators, but certain people taking until 1992 to officially recognise that Gallileo was right and they were wrong could fall in to this category.

__________________
Of all the things I've ever lost, I miss my mind the most!

As soon as one becomes immersed in the details of specific scientific areas, you find lots of stuff in the literature. In one of my own areas of expertise, defect mechanisms in crystals, the literature is full of old ideas that newer evidence shows are wrong. But to 99.999999% of the planet, it is of absolutely no interest.

__________________
At night the stars put on a show for free (Carole King)

One Earth, One Sky - IYA 2009
All moderation in purple

All scientific enquiry will have more failures than successes - you've got to kiss a lot of frogs before you find a prince

But why should these blind alleys be of any interest?

__________________
Spike
:)

Amazon lists this book.

__________________
"The truth may be out there, but lies are inside your head" Terry Pratchett

To me, a history of failure would be interesting if it focused on failures where in hindsight, it would seem that the players had all the tools in place to succeed, but failed by some kind of mistake. This allows us to learn how to avoid mistakes by understanding their nature. Failures that occured because the data available was incomplete or faulty don't count-- that's not very interesting, it is a natural part of science to adjust to new data.

The book The Sleepwalkers, by Koestler, was about the development of the modern theory of the solar system, by a series of blunders by the famous that made it seem that they just "sleepwalked" and happened to arrive at the right place.
Where does the "lie" come in, in 1991? Nobody alive then was disputing it.

Why should such a history of failure interest anyone?
1erhaps there are some good stories (i guess there is).
2:As a "test" or reminder that there is such a risk of failure - perhaps even getting some indication of how big it is. Of course it may be that some disciplines have the "large picture", and a right one too, while for other fields there still may be room for doubt about important points. (examples of the first may be found in geography, where questions about landmasses are solved).

Also, while Lewis, Clarke, and Vasco da Gama (as an example) are the ones remembered today, both Rocky Mountains and African west coast are littered with bones of their less lucky predecessors and/or competitors.

__________________
Fiction has to be plausible. Reality is under no such constraint.

There is at least one journal today that published basically failures: The Journal of Negative Results in Biomedicine http://www.jnrbm.com/

It appears that their main goal is to create a record of what has been done so that the next researchers come along and don't make the same mistake.

That's amazing-- I wonder how they convince people to publish with them? I can just see people looking over a c.v.-- "well, he's got ten publications... wait, 5 of them are in the journal of failures!" And what about citation counts for such a publication-- "see how many people I've saved from making the same mistakes I did...?" But seriously, a null result is more than just a "trail closed" sign, it can be used in actual scientific arguments. So it's generally not a failure at all, just a disappointment.

Still, there is a well-known principle called "positive outcome bias", where scientific researchers are much more likely to publish positive results than negative ones. I would probably publish in that journal, but not until after I get tenure. I am not sure how the tenure comittee would react to publications in that sort of journal.

__________________
I met this wonderful girl at Macy's. She was buying clothes and I was putting Slinkies on the escalator.
-Steven Wright



My Website: The Black Cat's Web Page

amen to that! This is where the amazing power of statistics comes into its own. Transform your data, track down an obscure test to run... poof! you can prove almost anything!

Better yet-- set up your experiment to produce an absolutely common sense result, then make people forget it's common sense, publish it, and call it your own!

That is the difficult part. How do you do this?

__________________
I met this wonderful girl at Macy's. She was buying clothes and I was putting Slinkies on the escalator.
-Steven Wright



My Website: The Black Cat's Web Page

There is a large push to make it mandatory that all clinical trials be registered, and the results reported, published or not. I think the hard part is actual enforcement.

Well the geocentristic idea isn't a failure, it's wrong but not a failure. Neigher was it bad science, not begin with. It became bad sience when it was defended after a better idea (heliocentrism) had been showed to be more likely. The same principle are true for a lots of old now to proven to be wrong scientific teories. Our ancestors where no idiots (well some where, just like people today).

__________________
"-UFO detected, interceptor launched"

A "single" example that comes instantly to mind of historical "bad science" is Blondlot and his N-Rays.

The Skeptic's Dictionary has a page about him. His was basically a case of a man who wanted to believe so much, that he ended up fooling himself.

__________________
"The facts gentlemen, and nothing but the facts, for careful eyes are narrowly watching." Isaac Asimov

Use your imagination. Jargon is the most popular way. You translate all the common-sense results into ethereal-sounding jargon, and people quickly forget the common sense, they're just trying to remember what the words mean. Then you tie the jargony-sounding words together into a nice package using your experimental results, thereby proving the value of the precise use of your jargon. If they haven't forgotten the common sense behind it by now, you aren't trying hard enough! Of course, the best test for this flavor of research is to always translate all jargon terms back into simple English before addressing the results of the experiments. The red flag is generally an exaggerated tendency to choose jargon even when normal words are perfectly appropriate (like the "sapience" vs. "self-awareness" issue).

But I'm not saying most research is like this, I'm just saying this is a good way to make sure you get a positive experimental result. There is probably a law that says, the higher the chance of getting a null result even though you've used a well-designed experiment, the more important the research, because the more unknown is the phenomenon under investigation. That must be a significant paradox for those seeking tenure...

This may work on laypeople, but not on your colleagues who know the jargon as well as they know common English (or whatever language they happen to speak). The jargon may seem mysterious and unintelligible to people not in the field, but to people in the field it makes perfect sense. There is no need to translate the jargon, to people in the field it can be read just as easily as any other text. If people had to translate jargon it would make no sense, the very reason it is used at all is because it is quicker, easier, and more specific to use jargon terms rather than having to have a lengthy explanation every time they have to mention something. If the jargon had to be translated by people who use it regularly, it would be of absolutely no use. Jargon is used to simplify things, not make them more difficult.

For instance, a single sentence from my Principles of neuroscience textbook:
This is a very simple sentence to me, I know exactly what all the terms mean and their significance when used in this context. A research paper full of these sort of terms would be no more difficult to me than reading the newspaper, and that is after only studying the material for one semester. If someone tried to bamboozle me with a lot of these terms in a research paper I would know it. However, this would seem like a lot of useless jargon to someone not already familiar with the terms. Now allow me to translate this exact same sentence into common English:
See the problem? What is more, this is an extremely simplified version. The terms, to someone familiar with them, have a much greater significance than can be covered here, to such an extent that I cannot do them proper justice without having several paragraphs (or even a page or more in some cases) dedicated to each one.

__________________
I met this wonderful girl at Macy's. She was buying clothes and I was putting Slinkies on the escalator.
-Steven Wright



My Website: The Black Cat's Web Page

Actually, the field I had most in mind is psychology, or one of the "soft" sciences, not biomedicine or biophysics. Those latter fields have spectacular amounts of jargon, but it passes the test: there is no common-vernacular equivalent. Thus the jargon is not translatable. In terms of your example, the real test would be to ask ten different practitioners to do what you just did. If they all came out nearly the same, then the test is passed. If they come out quite different, it is failed. I have seen many cases in soft science where I would expect the latter, though in "hard" science it would tend to rather be the former, and this may be one of the fundamental distinctions. I agree that jargon is usually good, in the right audience, since good use of jargon reduces ambiguity. But bad use intentionally generates it, as a kind of smokescreen. That of course happens all the time in pseudoscience, which may enter into the tenure process in some cases but hopefully that's not really what we are talking about. Nevertheless, I have seen many soft-science offerings of the latter stripe, and once it was really sifted through, the result was indeed common sense. But no offense to soft scientists, most are not this way, and the fields are simply a whole lot more difficult to obtain a "clean" result that was not already obvious to a population whose whole lives are one giant psychology experiment.

I have a great book called "The Measure of All Things" that details the development of the metric system, spending a lot of time on the failures that went along with that. It's a really good political history lesson, too, in addition to being about the basic units of modern science.

In artificial intelligence journals, especially in a lot of the earlier papers, you run across results that are not so good. I think that they publish those because it's nice to see very precisely the boundary at which a system is complex enough to solve some problem.

According to this link
http://www.acfnewsource.org/science/...e_results.html
the negative results journal is focussed on reporting results that challenge existing beliefs, not necessarily experiments that failed. I don't like that idea, because you should be able to publish a positive and a negative result about an idea in the same journal, so long as the experiment is valid. The only journals that don't publish (potentially) contradictory things like that tend to be the fringe journals that support a particular cause -- things like "Journal of the American Society of Homeopathy." (I don't know if that's a real journal, or if they really don't publish negative results if it is, but that's the kind of publication that would ignore or refuse to publish contradictory results.)