Pi in the sky

April 1, 2009 6:39 PM

New physics research: Time variation of a fundamental dimensionless constant
posted by shothotbot (26 comments total)

That's some fine LaTeX.
posted by swift at 6:50 PM on April 1, 2009 [1 favorite]

I found figure 2 quite illuminating. On the whole, a well reasoned treatment of the subject.
posted by demiurge at 6:55 PM on April 1, 2009

It's only a slightly funny idea, but the execution is just top notch. Like the fact that he proposes a theory that gives results in direct opposition to the data. And the section on the Oklo reactor. It's the details that do it, really.
posted by mr_roboto at 6:57 PM on April 1, 2009

Is April Fool's Day over yet? It's been a long one.
posted by Optimus Chyme at 6:58 PM on April 1, 2009 [1 favorite]

The section on the Oklo reactor is inspired.
posted by unSane at 7:07 PM on April 1, 2009 [1 favorite]

I was very impressed that this paper included a section on the Oklo reactor.
posted by TheOnlyCoolTim at 7:16 PM on April 1, 2009

The idea actually isn't that funny at all. If any fundamental dimensionless constants change, why not π? I'm talking about the "observed value" of π in the overall shape of the fabric of space time here. It's less clear that "absolute π" in some platonic 2D realm is up for variation.

Also, the spatial variation of π gives me an idea. What is the shape of spacetime if π is exactly 3, as it nearly was in Indiana? Did they narrowly avoid creating a wormhole?
posted by DU at 7:20 PM on April 1, 2009 [2 favorites]

(For a randomly-selectedcollectionof such papers, seeRefs. [2, 3, 4, 5, 6, 7, 8, 9, 10]).


Chortle.
posted by lalochezia at 7:29 PM on April 1, 2009 [1 favorite]

I guess this means I should stop telling people how neat the Oklo reactor is when I'm at dinner parties.
posted by fantabulous timewaster at 7:35 PM on April 1, 2009

Friggin hell... I actually thought up this idea months ago.

Granted, I'm not a physicist with the background to actually put it into a paper..
posted by leviathan3k at 8:15 PM on April 1, 2009

This theory is supported by the well-known mathematical result that 2 + 2 = 5, for large values of 2.
posted by ubiquity at 8:33 PM on April 1, 2009 [2 favorites]

Heh.
posted by turgid dahlia at 8:49 PM on April 1, 2009

No discussion of anything would be complete without a mention of the Oklo natural fission reactor.
posted by Salmonberry at 9:39 PM on April 1, 2009

lol Fyi, you only score a 2 page paper if the journal publishes it as 2 pages.
posted by jeffburdges at 9:43 PM on April 1, 2009

Yeah but how does the Oklo reactor in 2009 not know that Sayid shot it in 1977?
posted by WolfDaddy at 9:45 PM on April 1, 2009 [2 favorites]

The quantity plotted on the vertical axis has been chosen to make the time variation appear larger than it really is.
posted by alby at 3:36 AM on April 2, 2009

[13] Is it Friedman or Friedmann? Are you sure?
posted by beerbajay at 3:59 AM on April 2, 2009

An unusual new class of galaxy cluster
posted by Johnny Assay at 4:51 AM on April 2, 2009 [2 favorites]

This paper is clearly incomplete. It fails to include the datum that the constant in question was briefly 3.2 in part of a building in Indiana in the year 1897. It's time for a new theory that can account for localized saltation of fundamental constants.
posted by metaquarry at 4:56 AM on April 2, 2009

More speculatively, one might consider the possibility that the values of the integers could vary with time, a result suggested by several early Fortran simulations.

Heh.
posted by Anything at 5:48 AM on April 2, 2009

Vaguely related: why π is wrong.
posted by dmd at 6:06 AM on April 2, 2009

What is the shape of spacetime if π is exactly 3, as it nearly was in Indiana?

You know that this was in itself an April Fools joke, right?
posted by goodnewsfortheinsane at 9:46 AM on April 2, 2009

What is the shape of spacetime if π is exactly 3

Circles become hexagons, since the perimeter of a regular hexagon is exactly 3 times the distance between opposite corners.
posted by DevilsAdvocate at 10:13 AM on April 2, 2009

q. What is the shape of spacetime if π is exactly 3?
a. Circles become hexagons

Well, a hexagon doesn't have the same symmetries as a circle: you have to decide which direction leads to the first corner.

If you have a locally flat geometry with constant curvature, like the surface of the earth (closed) or a lettuce leaf (open), "π" depends on the scale you're looking at. If you draw a circle on your sidewalk, you get the flat π. If your circle's radius is from the north pole to the equator, the circumference (equator) is four "radii," or π = 2. So there's some length scale with π = 3. (Maybe historical time variation in π is correlated with changes in the tilt of the earth's axis! Too bad I have to wait a year to write that paper.)

Of course a circle where π depends on the radius doesn't have scale invariance, so that doesn't have all the same symmetries as a Euclidean circle, either. A manifold with π = 3 everywhere wouldn't be locally flat and would probably have lots of pathologies --- "spacetime" might not be a sensible word for such a shape.

The possibility that α has changed over time is a different creature, since nobody knows where its value comes from: it's just some number, with lots of decimal places, that tells you how strong electromagnetism is, which looks to be the same everywhere in the universe. Sure π is just some number too, but algorithms for calculating π correspond (at least in principle) to some sequence of operations on a circle. For the electromagnetic constant the sequence of measurements are different physical processes --- one photon exchange, pair production, etc. But α tells you how much weight to give each sub-process when you have a composite, like an electron transition in an atom, and really seems to come from nowhere.
posted by fantabulous timewaster at 11:33 AM on April 2, 2009 [1 favorite]

[13] Is it Friedman or Friedmann? Are you sure?

I think it's Friedmann only for very large values of n.
posted by Avelwood at 12:49 PM on April 2, 2009 [5 favorites]

Surely it's Friedmann iff n=2.

friedman( x ) = friedman( x-1 ) + "n"
friedman( 1 ) = "Friedman"

now if I could only figure out friedman( -1 ).
posted by unSane at 9:29 PM on April 4, 2009