Matt Springer is a graduate student of physics at Texas A&M university. He is also an occasional writer and tinkerer, and he is probably too curious for his own good.
when you order from Amazon, and a fraction of the purchase price will be sent to me at zero cost to you. Much obliged, and thanks for your patronage.Has anyone checked out ScienceBlogs' new election site, A Vote for Science? It's nominally about science issues in the presidential and congressional elections, but of course in practice it's a pretty standard near-self-parody of the ultraviolet end of the political spectrum. Well, I've got an account there as well. Haven't used it yet, but I think I probably will at some point. My own nutty borderline minarchist politics will be decidedly in the minority, but it's no fun if everyone agrees with you anyway.
Moving on, a couple of links. Here is Dr. Frank Close of Oxford University discussing the much discussed (and deservedly so) LHC. It's a good essay, but he says this which I think is unintentionally funny:
As the 21st century begins, physics can explain almost all of the fundamental phenomena revealed in the search for our origins, yet there are niggling loose ends. We see hints of a unified theory vaguely in the shadows, but what it is and how the structures that led to the particles and forces that molded us are still perceived only vaguely.
Loose ends, eh? I believe this has been said once or twice in physics before. Usually it's right before some colossal revolution upends our understand and reveals that the well-charted island we lived on is merely the tip of a much vaster peninsula. Will the collider experiments being conducted allow us to wrap physics up with a Theory of Everything? Maybe. But I wouldn't bet on it. History is strongly against it.
Via Swans on Tea, we have Assume a Spherical Physicist from the excellent blog The First Excited State. There's an exploration of several approximations, including how N + 1 = N is for all purposes true in percentage terms for very large N. But how about this one, mentioned in my undergrad thermodynamics textbook (Schroeder
1023N = N
You don't believe me, but it's in some respects true for very large N. Look:
Yeah, the approximation is inaccurate by a factor of Avogadro's number. But in lots of thermodynamics calculations that's a trivial error.
Here's a really interesting post from Dot Physics analyzing some physics from a video game. Why so interesting? Not just for the physics but also for the metaphysics. In a way analyzing the physics of a game is like trying to discover the laws of nature in and entirely different universe with entirely different rules.
Have a great weekend!
Physical Science
Comments
Yeah, I like that one. Likewise, e^x = 10^x for sufficiently large x. Even better, the units (Planck times or Hubble times) don't matter.
Posted by: Sean Carroll | September 20, 2008 11:49 AM
the approximation is inaccurate by a factor of Avogadro's number. But in lots of thermodynamics calculations that's a trivial error
Greenhouse Effect or economics, "trivial error" (heteroskedasticity for cause rather than effect) means "wrong." In neither case does this affect application (Kyoto and Montreal Protocols for both). Being unconstrained by empirical validation is marvelously liberating (e.g., string theory, NINJA mortgages, candidate VP MILF).
Geoffrey De Havilland demanded square windows for the Comet 1 to differentiate it from US airliners' round portholes. The "trivial error" and seven crashes later... Comet 2 had oval windows. 13 more craters followed.
Posted by: Uhcle Al | September 20, 2008 11:58 AM
conservatives are bad people who should be sent to reeducation camps. since conservatives have more children than liberals, and many liberals are not reproducing, it is probably morally acceptable to transfer children from evil conservative families to good liberal ones so that they may be raised as moral upright people who care about their fellow person. or, perhaps we should reintroduce polling tests, except we would disenfranchise people who listen to rush limbaugh, who should be imprisoned for hate crimes.
Posted by: razib | September 20, 2008 10:46 PM
Well, good thing we stocked up on all those guns!
Posted by: Matt Springer | September 21, 2008 1:21 AM
"Well, good thing we stocked up on all those guns!"
...just wait until president obama and the democratic congress are on charge.
Posted by: razib | September 21, 2008 2:15 PM
Wow, I've never seen an approximation that approximate! That's a good one. Apparently Uncle Al can do physics without approximations, but they are clearly good for the rest of us to help understand the physical world.
Thanks for the link to my blog, although I would like to point out that the name of the blog is The First Excited State, and "Assume a Spherical Physicist" was the title of that particular post.
Posted by: Excited State | September 22, 2008 12:01 AM
Fixed!
Posted by: Matt Springer | September 22, 2008 1:04 AM
My favourite approximation (that I sometimes use on the back of the envelope) is the small-circle approximation:
pi = 1. pi^2=10.
Posted by: Chris | September 25, 2008 7:06 AM
pi^2 is so close to 10 that we could switch to base pi^2 and no one would notice. But I always thought of the small circle approximation as 2*pi=1. And more generally, the small hypersphere approx is S_d=1. Makes field theory calculations a lot easier!
Posted by: Ben V-L | September 25, 2008 10:13 AM
DeHaviland's square windows were not a matter of trivial error. The stress concentration on a square corner is practically infinite, while the stress concentration on a circular edge is more or less zero.
Posted by: Jeffrey W. Baker | September 25, 2008 11:40 AM
It is probably morally acceptable to transfer children from evil conservative families to good liberal ones so that they may be raised as moral upright people who care about their fellow person. or, perhaps we should reintroduce polling tests.
Posted by: cabbagepow | September 25, 2008 5:44 PM