Saturday, February 9, 2013

The Odd Ends of Odds and Ends

Yes, it's that time again.  Clearing out bunch of ends and odds.
  • The Little Ironies of Life
  • Life Used to be So Much Simpler
  • The Music of a Generation
  • Hot Enough For Ya? 
  • Return of the Quanta

The Little Ironies of Life

A fellow in the UK who loves animals, but not so much as he is willing to live with them in his house, trapped a troublesome mouse.  Rather than exterminate the dirty little vermin -- perhaps he was influenced by too many Disney cartoons in his youth -- he took it to a nearby public park and released it.

A friend who had accompanied captured the liberation on his cellphone camera.  Within moments of Mickey's release, a hawk swooped out down, seized it, and flew off with the dinner

Life Used to be So Much Simpler

A Topeka man answered a CraigsList ad for a sperm donor placed by two lesbians who wanted to pretend that they had had a baby together.  He donated the sperm for free and the two women made him sign a contract in which he was barred from ever seeking any parental rights.  Well, you guessed it.  The two lesbians got "divorced" and the birth mother, falling on hard times, sought public relief, and the State of Kansas tracked down the sperm donor and is dunning him for child support payments.
Some of us recollect that, civilly, marriage was a contract between a copulating couple and the State that the couple (and not the State) would pay for the upbringing of any child that might result from their activities.  It reminds me of "dalliance songs" in folk music, only without the momentary pleasures of the dalliance. 
Speaking of the Product of Dalliance
The Guardian contemplates the consequences of sex with children and recollects that
In 1976 the National Council for Civil Liberties, the respectable (and responsible) pressure group now known as Liberty, made a submission to parliament's criminal law revision committee. It caused barely a ripple. "Childhood sexual experiences, willingly engaged in with an adult," it read, "result in no identifiable damage … The real need is a change in the attitude which assumes that all cases of paedophilia result in lasting damage."
Used to be, too, that publicizing such incidents were thought to do more harm to the child than keeping quiet about it, and that counseling could cure the offender.  All the counselors said so.   Charges were not brought unless the parents insisted on it.  It is hard to imagine in the present era of resurgent Calvinism in which people do not so much commit sins as they are irredeemably sinful, but folks who relied on the Best Professional Advice in one decade might be pilloried for having done so in another. 

The Music of a Generation

James Chastek makes an interesting point about a generation's music.  What they grow up listening to may be quite different from the music they go on to create:
The Baby Boomers (those born 1946-1965) are associated with the music of Elvis (b. 1935) the Beatles (Paul ’42; John ’40) Grace Slick (’39) Dylan (’41) Chuck Berry (’26)  Hendrix (’42) The Mamas and The Papas (John Philips ’35) The Beach Boys (Brian Wilson ’42) etc. But when we stop looking at the music they listened to and consider the music they made, we get Black Sabbath (Ozzy and Tony Iommi ’48)  Ice-T and Chuck D (’58 and ’60) The Sex Pistols (Sid Vicious ’57 Johnny Rotten ’56) Led Zeppelin (Plant and Bonham  ’48 , Jones ’46 – though not Page ’44) etc. So what the Baby Boomers listened to was one thing – when they had to express themselves, we got  Heavy Metal, Punk, and Gangsta Rap.
All that peace-and-love music of the Sixties was created and performed by members of the Silent Generation.  (Chuck Berry, man, he was only a little younger than my Dad!  So he was older than the Silents.)  The Boomers listened, grooved to it, and then went on to lay waste to it.  

Hot Enough For Ya?  

If not, we can make it hotter.
One gets used to the temperatures growing warmer, but it is harder to contemplate past temperatures getting warmer or cooler, ex post facto as it were.  The chart shows the amount of adjustment made to previous temperatures between the 2008 GISS report and the 2013 report.  Curiously, nearly all the recorded anomalies prior to 1965 have been adjusted to be lower than they were reported in 2008 while those after about 1975 have been adjusted to be higher than they were reported to be in 2008.  That is, 1991, say, was warmer in 2013 than it was in 2008, which cannot but arouse a certain degree of wonderment.

BTW, data adjustment can be perfectly legitimate.  You will notice the notice at the top of each data sheet: "using elimination of outliers and homogeneity adjustment."  What changed between 2008 and 2013 was the nature of the homogeneity adjustment.  They got a new tweak on.  It has added about a 0.1° to the slope of the trend line since 1900. 

Deming gave the example of the measurements of the angles of a triangular plot of land whose sum totaled less than 180°.  But since such a sum must equal 180°, it is perfectly legitimate to add the deficit into the raw data.  There are (at least) two different ways of doing this: a) add ⅓ of the deficit to each angle; b) add a proportionate share of the deficit to each angle based on the size of the angle.  There are circumstances when a) is appropriate and when b) is appropriate.  What is never appropriate (from the statistician's POV) is to blindly apply one sort of adjustment in all circumstances by means of an automated algorithm.

Return of the Quanta

Speaking of Chastek, he has a nice discussion of why they mythical causeless quantum events might refute Einstein or Bohm -- or for that matter, materialists -- but don't lay a glove on Aristotle or Aquinas.  Both fellows were quite aware of coincidence and chance events

Apparently, there are these people out there who think that if you cannot predict when and where a virtual "particle" pops into and out of existence that the virtual particle therefore has no efficient cause.  But the efficacy of a cause has nothing to do with its predictability.  Gravitational theory cannot predict which apple will fall on Newton's head and when.  Time (t) does not appear anywhere in the equations, except insofar as the apple is already in motion.  Quantum mechanics likewise eschews the term t.  In neither case does it mean that there is no cause to the event of an apple detaching from a branch or a virtual particle virtually existing for a virtual moment.  These "particles" are a superposition of quantum states, so it is evident that the particle is caused by the quantum states that are logically prior. 

It is not even clear in what manner they exist.  Wikipedia says
they appear in calculations of interactions, but never as asymptotic states or indices to the scattering matrix. As such the accuracy and use of virtual particles in calculations is firmly established, but their "reality" or existence is a question of philosophy rather than science.
Which to this aged ear sounds an awful lot like deferents, equants, and epicycles, the accuracy and use of which in calculations was firmly established: purely mathematical devices used to "save the appearances," but which might or might not have physical existence.  It's not entirely clear whether something that does not physically exist needs to have a cause. 
Random events are the result of the intersection of two or more causal lines.  We have used the example before, but recall the hammer sliding off a roof that brained the unlucky office worker walking toward his favorite luncheonette.  Everything in the worldline of the hammer is caused: the geometrical position of the hammer, the nudging of the roofer's foot, the coefficient of friction of the roofing tiles, the slope of the roof, the acceleration of gravity, and the kinetic energy of the fall.  Everything in the office clerk's worldline is caused: he is hungry, it is lunchtime, he chooses to eat at his favorite luncheonette, he walks at a certain pace along a certain street, and....  It is only chance that he is under the hammer when it falls.  The two worldlines cross -- and there is no reason why they ought to have crossed any more than there is a reason why you should be reading this while sipping a cup of your favorite beverage.  Correlation is not causation.  

Our unfortunate's death is caused indeed by the hammer, but there is no ratio, no reason, no law of nature that requires hammers to brain pedestrians, let alone to predictably brain this pedestrian at this particular time and place. 

Then, too, genuine randomness, should it ever be encountered, is a fair to middling argument for design with tongue not entirely all the way into the cheek. 


27 comments:

  1. I think you're conflating two separate concepts in your example of the hammer and the office worker. When we say something's accidental/unpredictable/random, we can mean either that it is theoretically impossible to predict, or merely that in practice nobody actually has the ability to predict it. A good example of the distinction is a roulette table; theoretically, a sufficiently intelligent being or sufficiently powerful computer, given enough data about the starting state, could predict where the ball would land. But since we don't have that much information or intelligence, the result is practically random even though it is theoretically deterministic.

    The example of the man walking under the hammer seems similar. It is a practical accident but a theoretical certainty and therefore can't really be called chance in the philosophical sense.

    ReplyDelete
    Replies
    1. theoretically, a sufficiently intelligent being or sufficiently powerful computer, given enough data about the starting state, could predict where the ball would land.

      Depends on the theory. If you assume there is no random effect, then it's true; if you assume there can be a random effect, then it's false.

      Not very helpful, since the "answer" is just what assumption you start with.

      Delete
    2. Yes, that was... kind of my point. I was drawing a distinction between cases where there is and is not a random effect.

      Delete
    3. I guess I didn't put it right.... I recognize the distinction; the problem is that your example is of a situation where the difference depends entirely on the theory.

      Contrast with, say, a computer "random number generator"-- same input, same result, but the variables are far enough beyond what humans can predict to be "good enough" for random. (Also have enough things like multiple algorithms that make for wildly different results that another computer won't be reasonably able to figure it out, either; reasonable being "without more examples than it is likely to get.")

      Delete
    4. Yeah, I actually thought of using an RNG as the example, but didn't want to get bogged down in a description of how they work. (I also thought of using dice, but there are some people who can control the fall of dice.)

      I don't see any plausible mechanism by which roulette could be truly random. A roulette ball is way too big to be influenced by quantum effects. Even frictive effects are on a scale orders of magnitude greater than anything truly random.

      Delete
    5. Addendum: now that I come to think of it, computer RNGs could be affected by truly random effects, though it's highly unlikely. There are trace amounts of thorium in chip casings, so it's theoretically possible that an alpha particle could flip a bit in the memory used by the RNG at exactly the wrong time. Extremely improbably, but not quite impossible.

      Delete
    6. How on earth could something actually be established as theoretically predictable vs theoretically unpredictable?

      Either you believe that if you only had enough information you'd be able to predict it, or no.

      It's like the "you only think you have free will" argument applied to physical effects-- if you think you've got enough data to predict something, and the prediction is wrong, then you just didn't have enough data rather than it not being predictable....

      Delete
    7. Well, John Stewart Bell actually did prove quantum effects to be theoretically unpredictable (there is a loophole; Bell's theorem assumes there's no way to exceed the speed of light. If that assumption proves to be untrue, then the argument falls apart). In the case of the roulette wheel, I suppose it's possible that there's some other effect we don't know about that makes it truly random. However, this would mean positing some previously unknown effect. The only known physical phenomenon that is truly random can have no effect here.

      Delete
  2. Don't forget that the hippies and peaceniks of the 60s were lead in public prayer in school. So much for the positive influence of public prayer. :)

    ReplyDelete
  3. "The two worldlines cross -- and there is no reason why they ought to have crossed any more than there is a reason why you should be reading this while sipping a cup of your favorite beverage."

    The frightening thing is that I actually *was* drinking a cup of tea while reading this.

    ReplyDelete
    Replies
    1. No one seems to wonder if there might not be something caused that is not predictable. "Caused" need not mean "determined."

      The most difficult thing in the world to arrange is a random system. Consider all the effort that must go into ensuring the randomness of a casino. Roulette balls and wheels must be designed, fabricated to tolerance, weighed, precisely balanced, etc. No one says that if A causes B that B cannot be a probability distribution.

      Delete
    2. Again, casino games aren't actually random. They're just beyond the ability of human beings to predict.

      A quick aside here. I should preface this by saying that, although my degree is in physics, it's only a BA, so don't take this as authoritative or anything. However, my understanding is that, due to energy-time uncertainty, particles can temporarily pop into existence literally anywhere, without being dependent on anything else. There is no conceivable situation under which they would not pop into existence. Therefore, they have no physical cause. I suppose you could say they have a non-physical cause (the laws of physics themselves) but I think that's changing the definition of "cause" a bit.

      Delete
    3. These so-called "virtual" particles are not particles as we understand them. They are caused by the superpostion of quantum states. That they are unpredictable does not make them causeless. You could even argue that since "they appear in calculations of interactions, but never as asymptotic states or indices to the scattering matrix," they are no more real than epicycles were for an earlier cosmological model. Like the epicycles (and their associated hardware), the virtual particles are computationally useful. They "save the appearances," in the medieval phrase. But even the Ptolemaics did not assert that epicycles were physically real and did not require that they have a "cause."

      Delete
    4. 1] Virtual particles, according to theory, arise into "reality" from the quantum vacuum foam in Hawking radiation at the boundary of singularities.

      2] A Florida judge recently ordered a birth certificate to indicate three parents:

      http://usnews.nbcnews.com/_news/2013/02/07/16889720-florida-judge-approves-birth-certificate-listing-three-parents?lite


      JJB

      Delete
    5. I vaguely recall that a ball oscillating in a sloped cup follows a chaotic path, much like droplet formation.

      However:

      http://suite101.com/article/chaos-theory-and-water-droplets-a86485


      JJB

      Delete
    6. They are caused by the superpostion of quantum states.

      I think you're being misled by the common study of virtual particles. We're mostly interested in them for their role in interactions between non-virtual particles, but energy-time uncertainty is written into the laws of physics. Even without any non-virtual particles around, they can and do "fizz" -- a vacuum devoid of all non-virtual particles would still have this fizz.


      That they are unpredictable does not make them causeless.

      I'm not saying it does. It's certainly possible to imagine an event that is caused but unpredictable. I'm just saying it's not clear that virtual particles are like that. For A to be a cause of B, then it must be the case that if A did not exist, B would not happen/exist. There's nothing like that for virtual particles. The only preconditions for them are that: (1) quantum theory must be valid, and (2): there must be space for them to inhabit. Neither of these things are really "causes" in the sense usually meant, I think (though I'm uncertain of my judgment in this area).

      I vaguely recall that a ball oscillating in a sloped cup follows a chaotic path, much like droplet formation.

      Different kind of chaos. Chaos theory deals with systems that are "chaotic" in the sense that very small experimental uncertainties in the starting conditions lead to wildly different results, not necessarily with true randomness.

      Delete
    7. Addendum: the fact that virtual particles don't appear as asymptotic states is just a consequence of the fact that they're temporary. An asymptotic state is one to which the actual state of a particle approaches as time goes to negative or positive infinity. Since virtual particles have finite lifespans, they don't approach any state outside of those lifespans.

      I think the difference between a mathematical fiction and "reality" isn't so clear-cut when it comes to QFT. If our understanding of physics is just flat-out wrong, of course, then that's one thing, but since our description of quantum states is purely mathematical to begin with, any attempt to classify some features as real and others as mathematical fiction is necessarily going to be a bit arbitrary. (earlier physical models also described things in mathematical terms, but they were models in which there was a clear built-in assumption of the definition of a physical object. That's not the case here).

      There is of course the difference between the quantum fizz in general and whether any one particle comes into existence, and since this is a product of uncertainty, there's a point of view in which it's not even meaningful to talk about individual virtual particles except when... well, you get the idea. So that could be the answer to individual particles' need for a cause, but what about the fizz's need for a cause?

      Delete
    8. what about the fizz's need for a cause?

      Carbonation, obviously.
      + + +

      since our description of quantum states is purely mathematical to begin with, any attempt to classify some features as real and others as mathematical fiction is necessarily going to be a bit arbitrary.

      Easier, say. It's all mathematical fiction. Neopythagorean number woo-woo has been in the catbird seat ever since the revival of idealism over empiricism that took place during the Renaissance. This was the notion that "If the math worked, the reality followed." The old empiricists would have said that since the reality worked, the math describing it would also obviously work, but the arrow of causality ran from reality to model: R→M. You cannot say "M, therefore R."
      Now, the math-working-out is a good clue. If it doesn't work, your model has not captured reality. Denying the consequent is valid. But never forget that the math of the Tychonic model worked marvelously well and was replaced by the Keplerian model only because the math there was easier. It held its own for half a century against the Copernican model. Even the Ptolemaic model held up for more than two millennia, until the phases of Venus put paid to it.

      Delete
    9. *wry* It would be nice if folks would keep a little mental voice that says, over and over, "in theory" or "the model says that" and such.

      It'd make my life a lot easier, and I'm not even in an obviously scientific field!

      Delete
    10. Again, I think you're conflating two separate things here. Something can be a "fiction" in the sense that it's wrong, or it can be a fiction in the sense that it's a useful mathematical shorthand without direct real-world equivalent -- and, crucially, that the people who use it know this. An example of the first sort would be the luminiferous aether. An example of the second sort would be the Lagrangian in classical mechanics (arguably). What I'm saying is that, if our understanding of quantum physics is correct (which of course it may not be), then it's hard to come up with a consistent division by which the aspects of a quantum system can be separated into classes that correspond with our notions of reality in any robust way. The presence of non-virtual particles is an aspect of the underlying state of the system, and there's no immediately obvious reason why it should be classified differently from any other aspect of the system. I suppose we could just define "real" to mean "corresponding to non-virtual particles", but then we've entered into circular reasoning.

      Delete
    11. It's all a "fiction" in the sense that the mathematical model is entirely a non-physical, axiomatic system, whereas physical reality is, er, physical at least (I'll let you argue about "axiomatic"). There is no "class" of "aspects" which is more "physical" than any other-- and this is true of the Newtonian model as well as the quantum model.

      A homey example: As a paper map is to the terrain, so is physics theory to physical reality. The map allows you to make predictions about future results, and to the extent that it is easy to use and makes accurate predictions, it is a "good" map. But 1) your experiences/experiments serve to validate or invalidate the map, not the terrain, 2) the terrain is much more than a map (e.g. you can't mine ore out of a map), and 3) requirements of the map don't necessarily translate to requirements of geography (e.g. a stereographic map projection requires a fixed reference point; the real terrain does not).

      The Newtonian model is fully deterministic-- but that is a property of the model. It doesn't follow that the physical reality is also fully deterministic. Quantum theory requires energy-time uncertainty to make it "work"-- but again, it doesn't follow that reality itself is necessarily uncertain. The theory doesn't prove or disprove reality; instead, reality disproves (or fails to disprove) the theory.

      Delete
    12. But again, you're talking about something else entirely. You're discussing whether the theory is correct; I'm saying that if the theory is correct, there's no obvious ground for classifying non-virtual particles as "real" and virtual particles as "unreal". This is different from the Newtonian model, which has a clearly defined distinction between objects and other, for lack of a better term, "things". Quantum physics has a much fuzzier distinction.

      Delete
  4. I did the same analysis as Chastek with some of my favorite recent bands: Thrice, Muse, an the Black Keys are late Gen-X (b. '75-'80) while Imagine Dragons and Mumford & Sons are Millennials (both headmen were born in 1987) Just that overview seems to indicate a difference in tone between our generation and the previous one, but this obviously requires more data points, especially regarding the Millennials. As an aside, Nirvana was early Gen-X (b. late 60's). Interestingly, it can be argued that the Millennials grow up on three generations of music, Silent, Boomer, and Gen-X, with all the variations of tone and emphasis between the three. There seems to be a mini folk revival among the Millennials, which will be interesting to see how much it catches on. [Full disclosure: I am firmly a Millennial (b. 1990) and am typing this on an iPhone]

    ReplyDelete
  5. Going off of the number of birds that have figured out "haying=food," I'd guess that it's not that uncommon for birds of prey to snatch up catch-and-release housemice.

    It's daytime, they're outside, they don't know where cover is, and birds are pretty good hunters.

    ReplyDelete
  6. I'm starting to suspect that the tendency for the "Best Professional Advice" to swing from one extreme to another is a matter of changing passwords to keep infiltrators out.

    ReplyDelete
  7. Someone in the comments of Chastek's post linked to a more detailed post on virtual particles that made me go back and check my math, and it looks like I got it pretty seriously wrong. For energy-time uncertainty to come into play, there must be a changing observable -- a stationary state won't do it, so there would have to be at least two (overlapping) stationary states there ahead of time. I really should have gone through and worked it out mathematically, rather than going from what seemed the obvious interpretation.

    ReplyDelete

In The Belly of the Whale - Now Available

    Dear Readers, Dad's final (? maybe?) work is now available at Amazon, B&N, and many other fine retailers. I compiled a list a fe...