« A Victory Kiss | Main | Eisie in Print »

Wednesday, 17 August 2011


Fun column Ctein, congrats on 200 give or take! Your column seems to indicate that our concept of reality must be flawed assuming we believe the results of the outlined experiment(s). Is it possible our concept of a reality that is local is flawed? That reality is not local, whatever the appropriate term for that is?

Fascinating, glad you went off topic for #200.

I'm curious to know more - what was the experiment the scientists in the 70s did to violate Bell's inequality?

There are a few errors in this article, and they are somewhat important ... not just of the nitpicking variety.

1. "Local" means that we consider something to hold true everywhere, all the time, on all scales.

"Local" in physics generically means something that we can assert for very small scales (technically infinitesimal) in both space and time. For example, general relativity treats spacetime as a manifold -- a mathematical object that is local flat everywhere but global need not be. This is the essence of the Einstein equivalence principle: one may locally remove gravity by moving to an observer that is freely falling. This does not mean that everywhere at all scales is flat (i.e. gravity can still exist).

When a physicist describes a (quantum) theory as local, they can mean one of two closely related things. The most common is a statement that call observables commute outside the lightcone, so that you cannot get faster than light influences from one observation to another. Technically this is called microcausality, but many physicists are sloppy. The other (which ensures microcausality) is that any interactions only couple fields at the same point.

The EPRB paradox satisfies "local" dynamics when described by quantum theory. The only time something non-local occurs is when you try and describe to the collapse of the wave-function -- something outside quantum mechanics (and it is not even clear that one needs it, as decoherence seems to be able to give us a world that to us looks classical without invoking any collapse postulates.)

Oh dear.

2. "Induction"
The argument that A implies B and B implies C is known as syllogism, not induction. (It is used an infinite number of times in the principle of mathematical induction, and some philosophers are not convinced that an infinite chain of syllogisms are meaningful.) This is not under attack by physics in any way, shape or form, as the truth of syllogisms are at the heart of your formal logic system.

Physics at its heart is an experimental science. We don't have any statements that we know about the world (i.e. not just to do with our formal logic system) that we know to be true. For example, we don't know that "Energy is always conserved". We have seen it conserved in many experiments before, so we write down our theory in which this is one of the rules. Because we have not done all possible experiments, we may have to replace that rule eventually with something else. The more experiments we do, the more confident we are that the rule really is a law of nature, but we can never *know*. This process of making a set of observations that is true, and postulating that it is always true is sometimes called induction, but it is an assumption that is always up for revision when a counter-example rears its ugly head--it is not a logically bulletproof statement!

The conservation of energy is a good example: for all experiments on Earth we have found that energy is conserved to the accuracy of the apparatus. But when we started to learn more about cosmology and general relativity, we found that the expansion of the universe meant that the total ("global") energy was not conserved, and had to revise our rules. The "local" energy density is still conserved, and we have (yet) to find any counter-examples.

I would agree that we are entering interesting times, but so far my money (for what it is worth) is on quantum mechanics being the whole picture, and in large systems the wavefunction never collapses but decoheres into almost-classical parts with very little interference. That would leave all three postulates intact, and would require no new physics. Indeed we have something of a 'god-of-gaps' at the moment, as decoherence is used at all the levels we have been able to study systems so far (almost up to the scales of viruses!) but the claims are "maybe things are different a few scales up!".

I would appeal to scientific induction (the one that is not logically bulletproof, but allows us to progress with only a finite number of observations) and claim that after making so many observations that I need a reason to hold out on decoherence being the only solution.

I would disagree that now we are entering the age of experimental philosophy. Physics has been doing this for years. Aristotle postulated mechanics based on how he thought the world should work -- work that was toppled by the experimentally driven Newtonian mechanics. We thought that we were the center of the solar system, with the orbits related to the Plutonic solids. This idea was thrown out by observations by Kepler. Einstein (and others) held onto the idea that the universe as a whole was static and unvarying -- Einstein even introduced the cosmological constant into general relativity to achieve a static universe that was impossible in the first version of his theory with known matter -- until Hubble's observations showed that the universe was indeed expanding.

Physics has been digging deep into the nature of reality, and challenging philosophy for almost 300 years. This is another interesting example of it, but by no means the first or marking the beginning of an era.

Mayhaps those constructivists we've always pooh poohed are onto something after all?

Wherever you are, there you are.

Or we may find that Plato was right all along about his Forms. Certainly fits how I approach photography. The camera is seeing something, right? If that's questionable I'm giving up and going back to music.

Reality might be overrated, but it's the only place where you can get a good steak, no?

PS. I guess the rest of the comments will be similar. :)

We take it for granted that a dog cannot understand, say, how a camera works, or philosophy, or art, or what those little bright dots in the night sky actually are—even when we explained it to him. However we tend to believe we can understand anything if we only think hard enough. But while our brains are bigger than our dogs', compared to the size of the universe they still are pretty small objects. I guess reality includes concepts that simply are too big for us to get our heads wrapped around—it's not just a matter of doing more research but a fundamental limitation of what we can possibly think. Dogs' brains are limited ... so are ours.

You lost me at "Welcome to..."

Seriously, though. Great article.

Your WAG is the most comforting view I can think of. Accepting our basic philosophical view but rejecting some of those principles is deeply troubling.

My philosophical comfort may, of course, not be high priority for the real world.

This is very interesting, but I have to wonder: If there is sufficient reason to question objectivity, causality, and induction, doesn't that create problems for scientific method itself? And if so, how can scientific experiments be designed that could really get to the bottom of all this? If there is an explanation that a non-physicist, non-scientist such as myself could understand, I would be curious to know.

42 boxes of cereal.



Aha! Douglas Adams had it right (that and so much else) in "The Hitchiker's Guide to the Galaxy."

Perhaps he's the man behind the "Ctein"?


At some point, very close to where we are, irrelevance sets in.

What's the name of the experiment / experimenters you are referring to?

I think that counting the boxes multiple times is the problem. Ask any chicken farmer how well that's going to work.

I'd go with Lisa Randall ( I used to argue with a coworker whether she or Lene Hau* would be the physicist we would most like to go on a date with ) and say that what we think about time and or space has more to do with how our minds work than how or whether time or space exist.

Of course if time turns out to be a human construct, then causality goes out the window along with free will. I.E. if the teacup is going to fall off the table then something has to push it, and that something turns out to be you, but there needs to be a reason for you to push it and it turns out that it was thinking about causality.

*Lene Hau can stop and restart light but Lisa Randall calls into question whether it was going anywhere to begin with. Like I said I'd go with Lisa Randall but that's sort of a Ginger vs. Maryann kind of choice.

It has been a while since I looked at this stuff, but if Ctein is talking about the experiment I think it is with Bell's inequality I understood the experiment to show that wave-particle duality idea is false. This is roughly that fundamental particles such as electrons can sometimes behave like a particle and sometimes like a wave, but never at the same time. The Bell's Inequality experiment caught it doing both.

This was central to the Copenhagen interpretation of Quantum Mechanics, which had various other theoretical problems that were ignored until this nail in the coffin.

The result is that currently no one can "explain" what happens at a quantum level, but can calculate the result of whatever happens. Perhaps this is what Ctein is getting at, otherwise I'm really confused.

In any case, I rather thought everybody had been in agreement for a long time that quantum theory (in some form or other) is pretty fundamental and this "local reality" business was binned a long time ago!

if you have time, have a look at these two papers: http://dx.doi.org/10.4006/1.3231944
(also available from http://neuron2.net/QM/bell.pdf ) and http://bayes.wustl.edu/etj/articles/cmystery.pdf

The basis of the problems with the non-causality or non-locality lies in the foundations of probability theory, not in QM!

In a nutshell, probabilities are NOT physical. Rather, they attach to logical propositions in a given domain of discourse,
In this view of probability (due to Cox and Jaynes) there are no random quantities in the traditional sense, only "knowns" and "unknowns" (Example: Schrödinger's cat is either dead or alive, not a zombie. A coin only falls on head or tail, but nothing in between.) Over the "unknowns" we specify a probability distribution to characterize the *state of knowledge* about the unknown quantity. But this state of knowledge is not physical, it is logical.

On a side note, this contradicts both the frequentist view (who say the data are random while the parameters are constant ) and the standard Bayesian view (which assumes random parameters and constant data).

For the Bell theorem this implies that there is no physical non-locality, only a logical non-locality. Causation is tied with physical locality, so the mystery dissolves.

More info on the Cox theorem here:

Ed Jaynes has written a book on his view on the foundations of probability theory which unfortunately appeared only after his death:

Jaynes: Probability Theory, The Logic of Science ( http://www.cambridge.org/de/knowledge/isbn/item1155795/)

If you find the time, read this book, it is is certainly among the most influential science books of the last century (and this one too)

If you KNEW you wouldn't need wild ass guesses.


A really excellent optical illusion or hologram may appear to be completely real, but if you know what you are looking for you can observe that you are being tricked. Without knowing about the existence of a trick, or knowing what tell tale signs to look for, your observations of the illusion are not exactly wrong, but are wholly incomplete and misleading.

And you're saying that the entirety of the sciences are, almost certainly, being completely misled by some tricks of observation which are so pervasive we can barely find any tell anywhere in our universe. Except for those few really big tells which imply everything we know about everything is a mere fraction of the truth.

Life is fascinating.


...but, I prefer my physics to be Useful as well as thought-provoking!



Cheers! Jay

That dog may not understand man's explanation of the things which man considers worth trying to explain but it is entirely possible that the dog may understand, to its own satisfaction, another dog's explanation of the sort of things that dogs think about and then sleep well and die happy, which is a good goal for both man and dog.

Hi Ctein.
Congats on the 200th.

I wasn't going to post but the next site I went to was a blog about a book of the History of Space Art you might have an interest in, so here it is;

I like how CGI has made cosmology, dinosaurs and sci-fi cool as it was a lot different when I was growing up. Hasn't made it any easier to understand though!

Posts like yours remind me of Asimov's "The Final Question", or Poul Anderson's "Tau Zero", but cosmology is fascinating but incomprehensible to me; I still go wow! if I see the ISS or a meteor:-)

I reckon there is a strong affinity between telescopes and lenses as a site I go to quite often with anything from astrophotography to AFV telescopes is Pete Albrecht's;
which would repay a few hours scrolling through.

Just found this that made me laugh;

all the best phil

So this is what DSLR's dream about, until we turn them on and force them to take pictures for us. Explains why they cost so much.
Fortunately, I use a 2004 model (and my girlfriend is a retired exotic dancer)
Great article, though, Ctein, brings me back to Poul Anderson and multiple parallel universes, metaphysics..fun in life!

To quote Star Trek:

"It's worse than that Jim, it's physics!"

You know what's cool? Somewhere around "Houston", you can just click over to Domai real quick for a breather and then come back.

This isn't OT, it is a preamble to a very ON-T part 2 which discusses the implications for photographic philopsophy, i.e. is the emphasis on photos having to depict and not distort 'reality' overrated?

I just told my spousal equivalent: "Not tonight, I have a headache".

"...quantum mechanics might very well be wrong in some way, although this experiment didn't find it. What this experiment said, though, was that local reality was unquestionably wrong."

Now here's the thing: people have been trying to break QM for almost 100 years. Guess what? No one has done it. Of course they will keep trying to break QM... and this is a good thing. I'm just glad my career didn't hinge on attempting to find a flaw in QM.

Unfortunately for humans, intuition is essentially useless when it comes to QM.

Richard Feynman said:

"Do not keep saying to yourself, if you can possibly avoid it, "But how can it [QM] be like that?" because you will get "down the drain," into a blind alley from which nobody has yet escaped. Nobody knows how it can be like that."

Sorry to keep posting.

But Sebastian's post is relevant and insightful.

Of course parameters (what we estimate) such as the number of photon captured by a given digital sensor site during an exposure, can be random and the data (what we want to measure) is constant. Who knows what the real, but unknown value of a parameter is... ever. The best you can do is estimate a value for the parameter and some estimates are more certain than others. Just because the data is noisy, i.e. a measurement value has a high degree of uncertainty, doesn't mean what you need to measure (data) is random. Suppose 9 electrons are generated by photos interacting with a digital sensor site and the total number of electrons at the same site due to noise fluctuates between 6 and 1 photons for each exposure. The data – or the real, but unknown, number of photons – is constant (9). The uncertainty (which is not data) or noise is not.

Probability theory languished on the back burner for decades because computing probability density functions for meaningful problems was impractical. The advent of inexpensive parallel computing means Probability Theory will eventually replace Frequency of Occurrence statistics. Fields like astronomy (where the very best measurement have a high degree of uncertainty) are already heavy users of Probability Theory.

Ed Jaynes book is a master piece. Even if you skip over all the math, you will never think about statistics the same way after you read Jaynes book.

But don't read Jaynes book if you believe shooting in RAW is unnecessary. JPEG shooters will be much happier if they never read what Jaynes says about discarding, modifying or filtering original data. The same goes for people who believe noise reduction is literally possible. You really don't want to read Jaynes either.

Disclaimer: I worked closely with Larry Brethorst who edited and published Jaynes book.

Uhm... I just wanted to read something about high ISO ...or something. And go to bed.

Now I am wondering whether I really 'am', since I know that I, at some point this morning anyway, 'was', but now, uhm... where was I?

Not a dull day at TOP. Keep it coming - just reserve these articles for those days when I happen upon them in the morning.

Even if the physical theory aspects are discounted, there are still purely logical difficulties with notions of external reality vs our (joint and several) internal concepts and beliefs about reality. We can tie ourselves in nasty knots about whether or not we can safely trust evidence from our senses (many teenagers experience this worry at some point, if they are at all thoughtful, before sensibly setting it aside), and whether the actuality of evidence of something (assuming we don't doubt it altogether), resides in the evidence or in the something.

If at all interested in these considerations, which in daily life we can't afford to be [grin] - the most clear-minded and lay-accessible discussion that I have encountered is that by Gilbert Ryle, in for example "The Concept of Mind". Donnish, dated in style, yet elegant almost to a fault.

I make no claim to understanding contemporary physics other than to say that the world that we think we know and which forms the basis of our daily dialogue is at best a weak anthropomorphic representation of that which really is and at worst a complete misunderstanding of reality which itself may not actually exist. This quandry informs and is the basis of my photography which, if you're interested, you can view at http://www.ericperlberg.com.

Heck with the column, nice picture!

Yep. Sounds good to me. Can't say I disagree with anything in the post.

Hi Ctein,
great text - should i be worried if i find this much less 'messing with my head' than some of your past columns?

As a sci-fi fan you probably approciate today's wikipedia featured article which is about another 200th celebration (probably some kind of quantum entanglement going on there...)

Dear Hugh,

I am at Worldcon, so I can't pull the experiment dates. It is late 70's; a good article appeared in Scientific American in 1979 (I think).


Dear Scott,

So, yes this is old news... But not to most lay folk. And, I would agre with you that this is a long firmly-settled matter... But there are very good minds who disagree. The "philosophy" of QM, as opposed to the facts, is still hotly and properly debated, even though I think the consensus is clear. For example, read "kiwi"'s excellent set of posts. The only reason I am not commenting on them is that I think they should stand free, and it would raise the level of discussion to incomprehensibly stratospheric levels. Very much not what Mike wants the comments to be for. But do give his three posts a very close read.


Dear William,

...and continuing... Feynman is both right and wrong. We don't know why the formalisms of QM work, haven't for over fifty years. It is vexing to do physics and not have any really good idea of why it is working. So you just have to put that aside and move on with your life. OTOH, Feyman got his Nobel prize out of a failure, from his perspective. He did not solve the problem of the electron singularity and he did not come up with a methodology that made QM intelligible. He just came up with something that worked fabulously well.

Feynman could also be flat-out wrong. He poo-pooed experimental particle physics as something utterly incapable of giving us any real understanding of the universe. History has quite firmly disproven that assertion.

pax / Ctein

I read this yesterday and meant to post (and today I expected to see someone already say this): I read the headline 3 times, each time seeing it as "Ctein's 200th Column Really Is Overrated". Which I thought odd... why post if one knows its overrated? Then my brain started working again.


I recall watching a car moving along a road.

I had a sense that the wheels were not turning but only seeming to turn - dependent upon me appreciating a succession of moments.

So I will go with causality being the weak link.

I've always wanted to see an article in the popular press about Bell's theorem. Thanks

I believe scale has a lot to do with it Ctein, predictatbility exist in the same relative demensions as the observer. So in the world of real life physics (call it locality or call it scale) the universe seams to be a really big but sort of calm and predictable place (even a supernova can be more or less predicted). But strip away the venear and underneath you find the unpredictability on the small (quantum dynamics, dark matter, dark energy) and the very large scale, for instance branes and the mirror edges of time and space itself. Great post though, and if any of you would want to dive into this stuf, try a Brian Green book.

Greetings, Ed

Trying to figure this raeality stuff out myself. Closest I've come is The First and Last Freedom by Jiddu Krishnamurti. Constant awareness from moment to moment. There is an interval between two thoughts. The observer is the observed.

And for your next 'mind messing' column, how about TIME: direction? dimensionality? existance? the tachyon and time travel?

and then there's String Theory....

Maybe for Xmas (whenever that is, if it is

Once and for all, is there or is there not a rhinoceros in the room?

A few years back, I was reading about Bell's Theorem, and I was reading about the experiment. I read it, then read it again, not understanding. I read it a third time, stopping after each small explanation of the states of the tangled photons, and then all of a sudden I understood what it meant. And I was in awe. If locality is an illusion, then everything is all actually one big thing. Size, space, everything is a product of our understanding and perception; the universe is actually a singularity.

And then I was hungry, so I made a sandwich.

"how about TIME: direction? dimensionality? existance? the tachyon and time travel? and then there's String Theory...."

Or, he could consider a column called "Mike's Patience for Physics Columns on his Photography Website: Is It Infinite? Not." Just sayin'.




Paul Richardson

Dear Mike,

Careful- I am fully prepared to start the series on the Art of Tea.

pax / caffeinated Ctein

A neutron walks into a bar and orders a beer. The bartender says: "For you, there's no charge."

"Or, he could consider a column called "Mike's Patience for Physics Columns on his Photography Website: Is It Infinite? Not." Just sayin'."
But Mike, all of photography IS just physics - from the electrons which illuminate the subject to the electrical churning in the brain which with which we interpret and evaluate the image (real or virtual). No physics, no photography

You can see the physics in it, I'll see the poetry. It's that multivalence that makes it so interesting.


I've decided I am going to name my next cat "Schrödinger"

There is a science-fiction series I like wherein one of the alien races man encounters has a group of people (later there are human ones) called "sohon masters," who can manipulate matter at the quantum level with their minds. (Humans, being humans, refer to them as "mentats.")

At one point, one of them is trying to explain something to a reasonably bright human (who isn't a physicist but can hum a few bars.) She says, "You understand that the famous equation E=mc^2 means that matter and energy are forms of the same thing and they can be converted into each other. That understanding is completely and fundamentally incorrect. However, it's a useful metaphor to explain the observable effects of what's really going on."

That's all physics (and most other sciences) are: useful metaphors. Newton's equations of motion, even though we've known for decades that they're totally wrong and never, ever give the right answer to any problem, are such useful metaphors that we still use 'em, blissfully unconcerned about the relativistic/quantum mechanical errors that they allow to creep into our every calculation. :)

Our metaphors, they will get better. However, sometimes I wonder if it isn't kind of arrogant of us to think that we *can* understand, that someday we will know the "truth." I think we can certainly find a metaphor that will accurately describe the observable universe to the limits of our detection. But the problem is we'll never know if the universe is finer-grained than our detection ability. :)

Induction is logic. Logic cannot fail. Ask Spock.

The comments to this entry are closed.