Showing posts with label physicists. Show all posts
Showing posts with label physicists. Show all posts

Friday, 26 February 2010

The Magic of Richard Feynman

Feynman DiagramsRichard Feynman (1918-1988) was one of the more colourful and charismatic characters in US physics.
He's remembered as one of the greatest physics minds of the Twentieth Century, which sometimes leaves non-physicists wondering exactly what it was (apart from Feynman diagrams) that he actually did to get that reputation. How did he end up being regarded as some sort of god amongst physicists, he never actually discovered anything that most people will have heard of?

One of Feynman's hobbies was stage magic. He was a keen practical joker, and was fascinated by the way that people are led to believe certain things, or why they end up acting in certain predictable ways. He was fascinated by fallibility, and predictable fallibility, which is one of the reasons why he was such a a great choice when they were picking people for the Rogers Commission, to investigate the reasons for the 1986 "Challenger" space shuttle disaster. Feynman understood the concept of system failure, both at the organisational and personal level, and he liked to play with people, including other physicists.

Stage magic often works through a process of misdirection. The practitioner demands with every element of their voice, facial gestures and body language that the audience may like to look over //here//, to the extent that we find it almost impossible not to look at their selected spot – perhaps an inch or so away from their extended, waggling fingertips – while with their other hand over //there//, they perform the mechanics of the actual trick.

A magician might announce before performing their stunt: "Look at this table. It's a perfectly ordinary table. It really, really is." And they bang on the table with their fist, and walk around it, and hit it with a stick, and mark an X on it with white chalk ... and you're concentrating so hard on the table to try to find why it's NOT an ordinary table, that you fail to notice the large black velvety cloth hanging above it, or the trap door behind it. The table is, in fact, completely ordinary. It's a double-bluff.

That's misdirection. You don't necessarily tell the audience something that's untrue or misleading, you give them a series of false clues, and let them work out the wrong story for themselves.

Another factor that makes stage magic effective is the way that people apply Occam's Razor. Technical stage tricks often require ludicrous amounts of preparation, absurd amounts of technical expertise or physical dexterity, and improbable investments in custom hardware. The assistant just happens to be double-jointed, or has an identical twin sister, or a false leg. At some subconscious level, the watcher's mind runs through a set of absurdly complicated and tortuous conspiracy theories that might explain what they're seeing and gives up, deciding that it's simpler to assume that the magician really can fly or make tigers disappear. The audience reasons that this isn't true (it's "only a trick"), but at a gut level they've already suspended disbelief enough to enjoy the show.



And so, to Feynman's magic trick.
One of the recurring stories about Richard Feynman goes something like this:
A physicist is working on a difficult problem. The physicist contacts Feynman. Feynman's secretary replies that Feynman is very busy, but could maybe schedule a meeting at some nebulous future date.
Several months pass. The physicist is contacted unexpectedly by the secretary to say that the secretary has just spotted that Mr Feynman now has a gap in his schedule, at quite short notice, and would the physicist still like to make use of it? The physicist eagerly agrees.

The physicist walks into Feynman's office.
"So,", says Feynman, "My secretary's just told me that you're working on some sort of interesting problem, but you'll have to forgive me, I've been really busy for the last couple of days, and haven't had the chance to look into it. Could you explain it to me? Oh, and could you start from scratch and make it simple, because, you know, this really isn't my field, and I'm not really up to speed with this subject. Start from the beginning."

The visitor is flattered and walks up to the board and starts explaining the nature of the problem. He pauses.

"So", says Feynman, "Let me see if I've got this right ..."

Feynman stares at the board and frantically marks symbols up while talking through what he's doing, until he has an equation.

"So your starting point would be something like that, yes? Okay, now tell me what you did next."

The visiting physicist is dumbfounded. What Feynman has just written on the board is the solution. And it's not just the solution, it's the solution to a more general version of the problem than the one that the visitor has been struggling with for months, or years. And Feynman's just done it in about three minutes flat.

The physicist leaves, ego totally destroyed, knowing that RF is in a totally different league to lesser mortals like himself.

Now, the "reveal".

If you were a suspicious stage-magician type, what you might suspect happened would be something like this: Physicist contacts RF's secretary, mentioning something about the problem. Secretary tells Feynman. RF researches the problem and all the relevant papers on the subject, and finds out how far the physicist has gotten. The secretary sends a stalling letter. Feynman adds the problem to his stack of other outstanding problems, playing them off each other, trying to cross-fertilise the different issues and bounce ideas between them, considering it a break from the problems he's actually trying to work on for himself. Finally, he works out the solution, and at this point, his secretary sends out the letter saying that RF now has an unexpected gap in his schedule.

Physicist arrives, RF plays dumb and asks them to outline the problem, RF "solves" it in three minutes flat, apparently using only the tools that the visitor has just provided.

Of course, this scenario still required RF to have been a damned good theoretical physicist. It also required RF to have had a wicked sense of humour, and to have done an awful lot of tough background work each time he pulled his stunt, just to create a few brief minutes of surprise for his "audience".

But that's exactly what stage magicians do.

Sunday, 12 July 2009

Remembering Emil Rupp

In the "impossible diamond" post, when I was talking about the impression given by C20th physicists had that fraud didn't happen in their profession, I forgot about Emil Rupp. Then again, almost everyone tends to forget about Emil Rupp.

Emil Rupp (1898-1979) studied under Nobel-prizewinning experimenter Philipp Lenard, and was considered by some to be one of the most exciting experimenters of his time. He did a series of experiments related to effects like electron diffraction that caught the imaginations of a number of key theoretical physicists, and his work was sometimes credited with being one of the most important influences on the development of quantum mechanics.

Rupp's work was central to some key questions in quantum mechanics. What is reality? is light really a wave or a particle? Is it emitted continuously or instantaneously? Can a state that is said not to exist still influence the outcome of an experiment?

Ironically, it then turned out that Rupp's own experiments, which had been so influential, didn't seem to have existed either. The thing supposedly came to light when some of his colleagues visited the lab where Rupp was working and confronted Rupp – he'd been describing experiments with 500kV electrons, but wasn't in possession of an accelerator that went up to 500kV. He'd been making up his experimental results.

Why did Rupp do it? Well, like Bernie Madhoff, for a while he was getting away with it, and was having a very, very good time. He was identifying problems that the physics community wanted solving, and solving them (albeit with fake experimental writeups). He was an enabler, and people (other than the fellow experimenters that he kept leapfrogging) liked him for it. Great names in theoretical physics would seek him out and cite him. Einstein spent quality time corresponding with Rupp in 1926, working through issues with wave-particle duality, and trying to work out what should happen in certain experiments ... and trying to come up with explanations for how it was that some of Rupp's experiments had come out so well, given some of the difficulties that he should have come up against. The collaboration was reasonably well-known, and people started referring to the "Einstein-Rupp experiments".

When the game was up, Rupp found that he'd now given the physics community a new headache. He'd shown that peer review didn't work as an efficient way of identifying "friendly" fraud within the system. If you had the right background, and you worked out which results people wanted and published those results, your paper tended to pass peer review unless the referees were so convinced that you couldn't possibly have gotten those results that they called you on it. And if an experiment produced the expected result, it was difficult for a referee to insist that an experiment was too successful. Results that don't agree with current thinking can be summarily rejected by peer review on the grounds that getting a "wrong" answer amounts to apparent evidence of error, but rejecting results that give the "right" answer is more awkward.
The lesson seemed to be that if you wanted a career as a scientific fraudster, the way to succeed was to agree with whichever theories were currently in vogue. So the physics community was now facing a potential upheaval – how would they assess how many other key papers by respected researchers might have been unreliable, or even outright fakes?

Rupp solved that problem for them with another piece of documentation. He sent a retraction of his five key papers, along with a letter from his doctor stating that Rupp had been in a "dreamlike" mental state when he'd written them.
It was a tidy conclusion – Rupp exited physics without there having to be a nasty inquiry, the community got to draw a line under the affair, quickly, and thanks to Rupp's explanation, they got to write off the matter not as an extended period of fraud lasting nine or ten years, but as the unfortunate actions of a guy who was having some mental health issues. That let the community off the hook – if Rupp hadn't been completely sane at the time, then we could still tell ourselves that physics was a special "fraud-free" field of science, and that no sane physicist would ever commit fraud. So everything was okay again.

Was Rupp's doctor's letter genuine? We didn't really care. We had the result that we wanted.


Refs: