Science writers on finding truth

SciAm writer John Horgan reviews the book of fellow science writer George Johnson, contrasting it with his own. They also discuss it in this Bloggingheads video.

I think that the XX century will always be considered the greatest for finding fundamental enduring truths, and Horgan has a similar view:

In 1965 Richard Feynman, prescient as always, prophesied that science would reach this impasse. "The age in which we live is the age in which we are discovering the fundamental laws of nature, and that day will never come again." After the great truths are revealed, Feynman continued, "there will be a degeneration of ideas, just like the degeneration that great explorers feel is occurring when tourists begin moving in on a new territory."

They argree on this:

On the issue of superstrings, Johnson and I are in complete agreement. Superstring theory seems to be a product not of empirical investigation of nature but of a kind of religious conviction about reality's symmetrical structure. Some particle physicists have the same view. Sheldon Glashow, one of the architects of the electroweak theory, has ridiculed superstring researchers as the "equivalents of medieval theologians."

Where they disagree is where Johnson seems to be influenced by foolish philosophers who deny truth and treat science as just another social activity driven by fads:

Johnson proposes that science, too, might be "a construction of towers that just might have been built another way." Although he never mentions The Structure of Scientific Revolutions, Johnson's view evokes the one set forth in 1962 by the philosopher Thomas Kuhn. In that book's coda, Kuhn compared the evolution of science to the evolution of life. Biologists, he noted, have banished the teleological notion that evolution advances toward anything--including the clever, featherless biped known as Homo sapiens. In the same way, Kuhn suggested, scientists should eschew the illusion that science is evolving toward a perfect, true description of nature. In fact, Kuhn asserted, neither life nor science evolve toward anything, but only away from something; science is thus as contingent, as dependent on circumstance, as life is. Johnson espouses a similar view, though more eloquently than Kuhn did. ...

My main disagreement with Johnson is that he plies his doubts too even-handedly. The inevitable result is that all theories, from the most empirically substantiated to those that are untestable even in principle, seem equally tentative. I also cannot accept Johnson's evolutionary, Kuhnian model of scientific progress. ...

Johnson further attempts to undermine the reader's faith in physics by reviewing ongoing efforts to make sense of quantum mechanics.

Johnson's view is probably (unfortunately) the majority among historians, philosophers, social scientists, science writers, and physics popularists.

Shimony's opinion of Bell's Theorem

Following up my EPR comments, I post this to show that my opinions are not much different from respected experts on the subject.

Abner Shimony spent his whole career writing papers on quantum spookiness, and here is his opinion of the consequence of Bell's Theorem:

There may indeed be "peaceful coexistence" between Quantum nonlocality and Relativistic locality, but it may have less to do with signaling than with the ontology of the quantum state. Heisenberg's view of the mode of reality of the quantum state was briefly mentioned in Section 2 - that it is potentiality as contrasted with actuality. This distinction is successful in making a number of features of quantum mechanics intuitively plausible - indefiniteness of properties, complementarity, indeterminacy of measurement outcomes, and objective probability. But now something can be added, at least as a conjecture: that the domain governed by Relativistic locality is the domain of actuality, while potentialities have careers in space-time (if that word is appropriate) which modify and even violate the restrictions that space-time structure imposes upon actual events. The peculiar kind of causality exhibited when measurements at stations with space-like separation are correlated is a symptom of the slipperiness of the space-time behavior of potentialities. This is the point of view tentatively espoused by the present writer, but admittedly without full understanding. What is crucially missing is a rational account of the relation between potentialities and actualities - just how the wave function probabilistically controls the occurrence of outcomes. In other words, a real understanding of the position tentatively espoused depends upon a solution to another great problem in the foundations of quantum mechanics - the problem of reduction of the wave packet.

What he is trying to say is that Heisenberg's quantum states represent potential measurements, not actual reality. There are apparent nonlocalities in the potentialities, but that is just in our minds. It does not violate relativistic causality, because that is about actual reality. There is no nonlocality in actual reality.

He still wants to believe in some sort of spookiness, but this shows that there is no proof of nonlocality in actual reality.

Explaining the EPR paradox

A reader asks me to explain my view of the EPR Bohm paradox, since I seem to reject what everyone else says.

Actually, my view is just textbook quantum mechanics. I do not subscribe to hidden variables, superdeterminism, action-at-a-distance, or anything exotic.

I am also a logical positivist, so I stick to what we know, and try not to jump to unsupported conclusions.

Experiments show that electrons and photons have both particle and wave properties. Quantum mechanics teaches that an electron is a mysterious quantum that is not exactly a particle or a wave, but something else. It obeys various physical laws, like conservation of energy and momentum, and Heisenberg uncertainty.

The EPR paper looks at a physical process that emits two equal and opposite electrons. Only I prefer to call them quanta, because they are not really particles with definite position and momentum at the same time.

Mathematically, the two quanta are represented as a single quantum state. A measurement of one collapses the state, according to the rules of quantum mechanics. Quantitative predictions are in excellent agreement with experiment.

In particular, you can measure the momentum of one quantum, and know that the other must be equal and opposite. Physically there is nothing strange about this, as it is a consequence of momentum being conserved.

But it is a little strange if you combine this with Heisenberg uncertainty, which normally prevents us from making a precise statement about momentum until it is measured. Measuring one quantum allows us to say something about a distant quantum.

Bohm pointed out that you get the same paradox when measuring spin, and tried to argue for hidden variables.

One way people have tried to explain this is with action-at-a-distance, with measuring one quantum having an instantaneous physical effect on the other. But that is so contrary to everything else we know about physics, then such an explanation would only be a last resort.

Another is to model the quantum with hidden variables. All such attempts have failed. In particular, if you think of the quantum as having a definite position, momentum, and spin before the measurement, you get contradictions with experiment.

So what is left? There is still the original Bohr logical positivist view. Things are clearer if you avoid trying to answer unanswerable questions. Physics is about observables. Concepts like position, momentum, and spin are not really properties of quanta. They are properties of observations.

We have no true mathematical representation of the quantum. We have a mechanics of observables. We predict observations, and avoid trying to say what is not observed.

When people try to visualize the quantum, they inevitably form some classical (pre-quantum) picture that we know is wrong. So they get paradoxes and say that quantum mechanics is incomprehensible.

Or they try some mathematical model that ends up being a hidden variable model, and it does not work.

So again, here is what happens. A physical process emits two equal and opposite quanta. They seem particle-like and wave-like, but they are physical objects that lack a true mathematical formulation. From the physics, we know that they are equal and opposite, and from quantum mechanics formulas, we can make certain predictions about observables. In particular, observations about the two quanta are correlated. Correlation is not necessarily causation.

Does some physical process like radioactive decay determine the state of an emitted quantum? I have an open mind on that, because I don't see how the question makes any sense. How can any experiment tell us one way or the other? You can believe it or not believe it; it is all the same to me.

Physicists make arguments that EPR-like experiments prove true randomness. I have posted denials of that, and I do not even believe that there is any such thing as true randomness. Randomness is a mathematical formalism that is actually deterministic, or a figure of speech for how certain theories fail to predict certain outcomes. That's all. When physicists talk of true randomness, they are usually talking nonsense.

What about the quanta being separable? Action-at-a-distance seems like hokum to me. It is as implausible as a perpetual motion machine. There is no evidence for it, and a lot of reasons for thinking it impossible.

I say that the physical quanta are separate and cannot influence each other. At the same time, our knowledge of the two quanta are linked, and info about one tells us something about the other.

In the terminology of Bell's theorem, I am rejecting counterfactual definiteness, just as all mainstream physicists have for decades.

Counterfactual definiteness says that the photons in the double-slit experiment must entirely go thru one slit or the other, just as if they were measured as particles at the slits. But mainstream quantum mechanics teaches that this is completely false, and such measurement would destroy the interference pattern. The light goes thru both slits at once.

You cannot do a completely passive observation of a quantum giving it a particular position, momentum, or spin. Any such measurement changes it, because quanta do not have such properties.

Rejection of counterfactual definiteness is essential to XX century physics, and is embodied by these slogans:

Another thing that people have emphasized since quantum mechanics was developed is the idea that we should not speak about those things which we cannot measure. (Actually relativity theory also said this.) [Feynman]

Unperformed experiments have no results. [Peres]

Somehow the 21st century has brought us more and more physicists who would rather believe in spookiness or parallel universes. A serious disease has infected Physics.

You will probably say I am cheating because I am not seeking a complete mathematical description of an electron, or that it is a cop-out to say that the wave function is just a representation of our knowledge.

My answer is that this issue goes to the heart of what science is all about. The job of physics is to use mathematics to predict outcomes for experiments. It is not to provide mathematical representations for things you cannot observe. All of the EPR paradoxes are based on naive expectations for counterfactuals, and not observations. Stick to observables, and the experiments are not so strange.

Philosophers confused over causality

Physicists usually have a dim view of philosophers. One reason is that philosophers reject causality.

Here is a new philosophy paper with more nonsense on the issue:

Issues surrounding the role of causation/causal reasoning in physics have recently been the subject of considerable philosophical discussion (e.g., Norton, 2009, Smith, 2013, Frisch, 2014). There is a spectrum (or perhaps, more accurately, a multi-dimensional space) of different possible positions on this issue: Some, echoing Russell, 1912 take the view that features of fundamental physical laws or the contexts in which these laws are applied imply that causal notions play little or no legitimate role in physics – or at least that they play no “fundamental” role. Others take the even stronger position that causal notions are fundamentally unclear in general and that they are simply a source of confusion when we attempt to apply them to physics contexts (and presumably elsewhere as well. ) A more moderate position is that while causal notions are sometimes legitimate in physics, they are unnecessary in the sense that whatever scientifically respectable content they have can be expressed without reference to causality. Still others (e. g., Frisch, 2014) defend the legitimacy and even centrality of causal notions in the interpretation of physical theories. Those advocating this last position observe that even a casual look at the physics literature turns up plenty of references to “causality” and “causality conditions”. Examples include a micro-causality condition in quantum field theory which says that operators at spacelike separation commute and which is commonly motivated by the claim that events at such separation do not interact causally, and the clustering decomposition assumption referred to in section 5 which is also often motivated as a causality condition. Another example is the preference for “retarded” over “advanced” solutions to the equations of classical electromagnetism (as well as the use of retarded rather than advanced Green’s functions in modeling dispersion relations, as discussed below) where this is motivated by the claim that the advanced solutions represent “non-causal” behavior in which effects temporally precede their causes (violation of another causality condition). Similarly, there is a hierarchy of causality conditions often imposed in models of General Relativity, with, for example, solutions involving closed timelike curves being rejected by some on the grounds they violate “causality”. Causal skeptics (e.g., Norton, 2009 and to some extent, Smith, 2013) respond, however, that these conditions are either unmotivated (because, e.g., vague or unreasonably aprioristic) or superfluous in the sense that what is defensible in them can be restated without reference to any notion of causation.

This is hopelessly confused. If a philosopher cannot see that causality is essential to modern physics, then he has no understanding of physics.

Denying that physics is about causality is like denying that physics is about energy.

Even the guy (Frisch), who defends causal notions in physics, is also confused. See his paper (pdf), where he only allows for causality in the narrowest subfields of physics. His arguments have no merit either.

There are several arguments in play, all nutty. One is that a theory making precise predictions is not causal, because the outcomes are constrained by the math, not the cause. So Newton's F = ma is not causal. Another is that action at a distance is not causal, because it cannot be broken into a chain of events. So Newtonian gravity is not causal. Another is that time-symmetric theories cannot be causal, because the past has to cause the future without the future causing the past. Most physics differential equations are time symmetric, so this eliminates a lot of theories. Another is that indeterministic theories cannot be causal, because outcomes should be determined by the causes, and not by chance. So quantum mechanics cannot be causal.

Of course stochastic theories can be causal. Without getting into mathematical definitions, consider the statement "smoking causes lung cancer". Everyone understands this. But smoking does not cause lung cancer in every single smoker, and is not the sole cause of lung cancer. Smoking raises the probability of lung cancer. It has always been understood that causality can work with probability in this way.

Statements like "the moon causes the tides" are also widely accepted under either Newtonian gravity or general relativity. Under Newtonian gravity, you might object that there was no local causality, but there was non-local causality.

Philosophers do not just attack physics. Some have attacked evolution biology, in the Causalist-Statisticalist Debate. Here is a recent review of this issue. The argument is that principles like "survival of the fittest" have no predictive power, and do not give causal explanation. They just give a framework for generating statistics about nature.

More on Musser's spooky book

SciAm editor George Musser complains that I trashed his book without reading it. Okay, fair point. His book is on Amazon as Spooky Action at a Distance: The Phenomenon That Reimagines Space and Time -- and What It Means for Black Holes, the Big Bang, and Theories of Everything.

The endorsement from Frank Wilczek says:

Locality has been a fruitful and reliable principle, guiding us to the triumphs of twentieth-century physics. Yet the consequences of local laws in quantum theory can seem 'spooky' and nonlocal-and some theorists are questioning locality itself. Spooky Action at a Distance is a lively introduction to these fascinating paradoxes and speculations.

Wilczek is a distinguished and level-headed physicist. I read this as saying that he firmly believes in locality as a great triumph of XX century physics. Some quantum experiments may seem spooky and nonlocal, so they are fun to talk about, but he is not endorsing any spooky or nonlocal interpretations or speculations that are in the book.

I can agree with that. Maybe Musser put "spooky" in the title to sell more books. If so, I do not fault him for that, as long as he describes the physics correctly.

I was once appalled by a 1979 book called The Dancing Wu Li Masters: An Overview of the New Physics. While the book was filled with goofy speculations, the actual description of the known physics was pretty accurate. So I ultimately decided that it was a decent book.

The current SciAm has a preview of his book, where he relates nonlocality to general relativity:

When I first learned about the quantum phenomenon known as nonlocality in the early 1990s, I was a graduate student. But I didn't hear about it from my quantum-mechanics professor: he didn't see fit to so much as mention it. Browsing in a local bookshop, I picked up a newly published work, The Conscious Universe, which startled me with its claim that “no previous discovery has posed more challenges to our sense of everyday reality” than nonlocality. The phenomenon had the taste of forbidden fruit. ...

Points in the gravitational field must be interlinked with one another so that they can flop around while collectively still producing the same internal arrangement of objects. These linkages violate the principle that individual locations in space have an autonomous existence. Marolf has put it this way: “Any theory of gravity is not a local field theory. Even classically there are important constraint equations. The field at this point in spacetime and the field at this point in spacetime are not independent.” ...

In short, Einstein's theory is nonlocal in a more subtle and insidious way than Newton's theory of gravity was. Newtonian gravity acted at a distance, but at least it operated within a framework of absolute space. Einsteinian gravity has no such element of wizardry; its effects ripple through the universe at the speed of light. Yet it demolishes the framework, violating locality in what was, for Einstein, its most basic sense: the stipulation that all things have a location. General relativity confounds our intuitive picture of space as a kind of container in which material objects reside and forces us to search for an entirely new conception of place.

I have no quarrel with this, except that I would not use the word "locality" this way. To me, locality means that the physics of a point can be understood from tevents, matter, and fields in its local neighborhood. General relativity satisfies locality in that sense. Musser says that relativity makes a global definition of location more difficult. Yes, that's right, but I would say that is a consequence of locality, not a contradiction to locality.

Update: Motl has just posted a good explanation of locality, and why it is a mistake to give up locality in order to get a more intuitive understanding of quantum mechanics. The spookiness of nonlocality is always less intuitive. I will respond to the first comment below tomorrow.

SciAm book promotes spooky action

Physicist Sabine Hossenfelder blogs this book review:

Spooky Action at a Distance: The Phenomenon That Reimagines Space and Time -- and What It Means for Black Holes, the Big Bang, and Theories of Everything
By George Musser
Scientific American, To be released November 3, 2015

“Spooky Action at a Distance” explores the question Why aren’t you here? And if you aren’t here, what is it that prevents you from being here? Trying to answer this simple-sounding question leads you down a rabbit hole where you have to discuss the nature of space and time with many-world proponents and philosophers. In his book, George reports back what he’s found down in the rabbit hole.

Locality and non-locality are topics as confusing as controversial, both in- and outside the community, and George’s book is a great introduction to an intriguing development in contemporary physics. It’s a courageous book. I can only imagine how much headache writing it must have been, after I once organized a workshop on nonlocality and realized that no two people could agree on what they even meant with the word. ...

In his book, George lays out how the attitude of scientists towards nonlocality has gone from acceptance to rejection and makes a case that now the pendulum is swinging back to acceptance again. I think he is right that this is the current trend (thus the workshop).

This sums up what is wrong with physics today. Clear-eyed XX century physicists purged the medieval spooky mysticism of nonlocality from science, and now it is back with no one even knowing what it is.

She goes on to say that the book is a confusing mish-mash of buzzwords and personalities, without ever explaining the physics or saying what is accepted:

I found the book somewhat challenging to read because I was constantly trying to translate George’s metaphors back into equations and I didn’t always succeed. But then that’s a general problem I have with popular science books and I can’t blame George for this. I have another complaint though, which his that George covers a lot of different research in rapid succession without adding qualifiers about these research programs’ shortcomings. There’s quantum graphity and string theory and black holes in AdS and causal sets and then there’s many worlds. The reader might be left with the mistaken impression that these topics are somehow all related with each other. ...

For my taste it’s a little too heavy on person-stories, but then that seems to be the style of science writing today.

This sums up what is wrong with popular science writing. SciAm used to be better than this.

A friend of hers wrote the book, and she recommends it. Sigh.

The Wikipedia article on action at a distance is pretty good. It explains how Maxwell developed his electromagnetic theory by seeking to get rid of action at a distance. This was one of the most important intellectual developments of all time.

To date, all experiments testing Bell-type inequalities in situations analogous to the EPR thought experiment have results consistent with the predictions of quantum mechanics, suggesting that local hidden variables theories can be ruled out. Whether or not this is interpreted as evidence for nonlocality depends on one's interpretation of quantum mechanics.

That's right. You can interpret quantum mechanics to give a scientific causal view of the world, or you can interpret it to allow for unverifiable spooky actions. Your choice. Apparently the current trend among physicists and popular science writers is for the latter.

Lawrence M. Krauss has a pretty good entanglement article in the latest New Yorker mag:

No area of physics causes more confusion, not just among the general public but also among physicists, than quantum mechanics. On the one hand, it’s the source of New Age mythology, and has enabled hucksters to peddle new self-help cures; on the other, for the philosophically inclined, it has provided some illusory hope of free will in an otherwise deterministic universe. Of the aspects of quantum mechanics that confuse and dismay observers, perhaps nothing approaches the property called “entanglement.” Einstein, who never really accepted entanglement’s existence, called it, derisively, “spooky action at a distance.”

Unfortunately for Einstein, entanglement, “spooky” or not, is apparently real, as researchers in the Netherlands demonstrated last week, just in time for Halloween. In doing so, the researchers affirmed once again that quantum mechanics, as strange as it may seem, works in every way we can test it.

Yes, physicists themselves cannot agree on whether entanglement is spooky.

He attacks the essay I cited last week:

Similarly, last week, the Pulitzer prize-winning writer Marilynne Robinson published an essay in which she challenges the nature and relevance of modern science. The essay argued that entanglement “raises fundamental questions about time and space, and therefore about causality.” She went on to say that this called into question the ability of science to explain reality as a whole. It’s easy to understand how Robinson arrived at this incorrect idea: when a measurement of one electron here can instantaneously affect the measurement of another electron on the opposite side of the universe, faster than the speed of light, it does seem as though causality has been thrown out the window.

Yes, fundamental questions about causality would be raised if the "measurement of one electron here can instantaneously affect the measurement of another electron on the opposite side of the universe". But it cannot.

Krauss's explanation is above average, but defective:

As long as the two electrons remain entangled, then this link endures — even if they are separated across the galaxy. If I measure one electron in my lab, the second electron is affected by the measurement of the first electron with no time delay — instantaneously — even though a signal travelling at the speed of light would take millenia to cross the distance between them.

That instantaneous link is the “spooky action at a distance” of which Einstein was so skeptical.

No, measuring an electron does not affect a distant electron.

Quantum mechanics gives a mathematical representation of the electron pair, and a way of making probabilistic predictions about spin. Measuring one electron does affect the prediction being made for the other. But the math is not the same as the physics. The actual electrons may be deterministic and separable.

Spooky action at a distance was debated by physicists in the time of Newton, of Maxwell, and of Bohr and Einstein. I thought that mainstream physics was solidly convinced that no such thing exists. Maybe that was true 50 years ago. What happened? Why have all these otherwise-hard-headed guys gone mushy about physics that perfected in 1930? Physics is in a sorry state when you have to read the Lubos Motl blog or my own to find something sensible on this subject.

Speculation about NSA and quantum computing

I posted a couple of months ago that NSA is cautious about quantum computers. The NSA is a secretive govt spy agency that does very little to explain itself, so there is a lot of speculation about why it would become cautious about quantum computing. Does it know something we don't?

One could also ask why Microsoft and Google are excited about quantum computing. Do they know something that we don't?

As I have often noted on this blog, quantum computing has been a colossal failure, and has no hope of any commercial applications in the foreseeable future. It is doubtful whether it is even physically possible.

Cryptography professor Matthew Green writes:

If you’re looking for a nice dose of crypto conspiracy theorizing and want to read a paper by some very knowledgeable cryptographers, I have just the paper for you. Titled “A Riddle Wrapped in an Enigma” by Neal Koblitz and Alfred J. Menezes, it tackles one of the great mysteries of the year 2015. Namely: why did the NSA just freak out and throw its Suite B program down the toilet?

These guys are leading experts in elliptic curve cryptography, and long-time NSA watchers. So their speculation is probably better than mine.

The popular press has somehow convinced everyone that Snowden proved the NSA has tricked people into using elliptic curves in order to use a pseudorandom number generator that has an NSA trapdoor, thereby allowing the NSA to spy on everyone.

This story is exaggerated. The so-called trapdoor was publicly known without Snowden, and no one had to use it. The basic elliptic curve technology remains sound.

It is curious that the NSA has deprecated the P-256 elliptic curve, as it has no publicly known weaknesses, and is used for all Bitcoin transactions. The Bitcoin network is hugely successful and out of control, and maybe the NSA is trying to cast fear, uncertainty, and doubt (FUD) on it.

My guess is that either the NSA has been suckered by quantum computing hype like Microsoft and Google, or it wants to discourage elliptic curve cryptography because it is too secure.