Scott Aaronson was asked five questions about the "progress in philosophy". In this text, I will replace his talkative nonsensical answers by the concise and correct ones and clarify some of his minuderstandings of quantum mechanics, too.
OK, the first question is:
1. Why are you so interested in philosophy? And what is the social value of philosophy, from your perspective?The social value of philosophy was hiding in its role of a subject that used to attract – and, to a lesser extent, still attracts – high-IQ people and makes them think about important questions. Historically, philosophy was therefore the ultimate "protoscience" and became the seed of science as we know it today, too. And that was good for the mankind.
However, its modus operandi is a flawed approach to learning the truth. The old philosophy was studied before the scientific method was understood; and the modern philosophers – by the very definition of philosophers – are still failing to use the scientific method. They don't understand that Nature is smarter than us which is why they still hope to "guess the important truths" without any accurate empirical input; and, more importantly, they fail to formulate their musings sharply enough and eliminate the falsified ones.
Therefore, we may say that philosophy as a human enterprise has a "social value" but philosophy as a body of knowledge, methods, and results has no "epistemic value".
The second question was related:
2. What are some of your favorite examples of illuminating Q-primes [i.e., scientifically-addressable pieces of big philosophical questions] that were solved within your own field, theoretical computer science?In general, there aren't any big questions posed by philosophers that were solved within science simply because philosophy's modus operandi is not only a flawed method to find the right answer; it is a flawed method to choose the right questions, too. For this reason, virtually all important enough questions first posed by philosophers were scientifically shown to be meaningless or building on invalid assumptions (and all "specific enough" theories invented by philosophers – whether they have called them "questions" or, which was more typical, "teaching" – were shown scientifically false). The philosophy's unscientific method not only fails to eliminate the blunders and misconceptions from the answers; it fails to eliminate them from the questions, too.
As Sidney Coleman wisely said, and he was not joking, "not even thousands of philosophers meditating for thousands of years would manage to invent something as strange as quantum mechanics". At least so far, listening to Nature's voice has been far more important and fruitful in the process of learning the truth than a purely man-made invention of principles that should be followed. Men may guess the right two or three more steps after they have learned something about Nature while paying no attention to Nature. However, if they try to get too far with their predetermined "philosophy", they are pretty much guaranteed to be stuck in a dead end.
Not every man is refined enough to be hired as a cop in Prague. ;-) Check also the Prague advent interactive street Flash.
Scientists – and more generally, researchers following some rational, quantitative, science-like protocols and principles – sometimes use the term "philosophy" for the "seemingly big questions" and the kind of thinking that resembles the philosophers' thinking. However, if they think about it carefully enough, they must acknowledge that the similarity is mostly coincidental and it is never exact.
Despite all superficial similarities, science is ultimately solving at least somewhat different questions than those that were originally posed by philosophers and it uses very different methods to find the answers. Quite generally, the new and groundbreaking scientific results are unacceptable for most philosophers who are stuck with their centuries-old dogmas.
The third question wants reforms:
3. Do you wish philosophy-the-field would be reformed in certain ways? Would you like to see more crosstalk between disciplines about philosophical issues? Do you think that, as Clark Glymour suggested, philosophy departments should be defunded unless they produce work that is directly useful to other fields … ?Philosophy has already been reformed by Galileo and his followers about 400 years ago; the product of the reformation is known as "science". Today, by definition, the word "philosophy" means the "unreformed old philosophy" or "whatever is using the same basic methods and building on the same basic principles". And as long as we will avoid new changes of the terminology, this "unreformed philosophy" will remain "unreformed" i.e. as flawed as it has always been and as it is today. Philosophy is unreformable.
It is up to any sponsor to decide what he or she or it or they is or are paying the money for. The main problem with the philosophical method is not that it produces no results for other fields; the main problem is that it doesn't produce the true answers in its own field.
4. Suppose a mathematically and analytically skilled student wanted to make progress, in roughly the way you describe, on the Big Questions of philosophy. What would you recommend they study? What should they read to be inspired? What skills should they develop? Where should they go to study?String theory, with the usual prerequisites – some maths, mechanics, field theory, general relativity, quantum mechanics, quantum field theory. In the future, string theory may become just another stair in the hierarchy of prerequisites but today it's the only state-of-the-art framework to reliably answer the deepest questions about the being.
5. Which object-level thinking tactics... do you use in your own theoretical (especially philosophical) research? Are there tactics you suspect might be helpful, which you haven’t yet used much yourself?The examples that the author of the question presents are random parts of the thinking process (and most of them are concerned with some technicalities of the algorithmic complexity rather than general, philosophical features of reasoning). There are just way too many of them to enumerate them all. However, a viable way to use any of them has to be rational and non-dogmatic – has to eliminate assumptions that have been shown invalid. In other words, the right approach is not the philosophical one.
Scott Aaronson vs Werner Heisenberg
I promised you to focus on one particular topic penetrating Aaronson's blog post: Heisenberg's reasoning. Scott isn't among those who completely misunderstand quantum mechanics but one may still see that he mostly misunderstands its meaning.
Bram Cohen began the exchange by this, ehm, somewhat loaded question:
Hey Scott, what do you make of Heisenberg’s more, ahem, obscure writings?Well, they're mostly obscure for readers who are not sufficiently talented in quantitative, rational, scientific reasoning. Heisenberg not only wrote down the right words about the inner workings of quantum mechanics (and the rest of modern physics); he was the first man who actually figured many of these key things out, too.
Aaronson writes lots of things in his reply, for example:
Personally, I’d say that neither Bohr nor Einstein really understood entanglement or decoherence in a modern way.Well, already in the 1920s, Bohr understood all the elementary rules of physics that governed entanglement – and he understood everything right about entanglement when the term "entanglement" was introduced in the 1930s. He wouldn't understand all the modern technical results about entanglement that are studied as "quantum information theory" – but these are not fundamental questions in physics, they are just particular engineering applications of fundamental physics.
He couldn't have understood "decoherence" with everything we associate with it today – after all, the term was introduced in the 1980s, long after Bohr's death. But he understood that the chaotic evolution of macroscopic etc. systems suppresses all the intrinsically quantum mechanical behavior such as interference by which a quantum system may be operationally distinguished from a classical one. And this is really what decoherence does – everything completely different that people try to attach to the term "decoherence" is a myth.
You might say that Bohr and Heisenberg got closer to what we now know to be the truth about QM (i.e., that local hidden-variable theories can’t work, and the probabilities in QM can’t have an ordinary ignorance interpretation like in QM).Scott is "better" than some other zealots because he at least realizes that local hidden variables are wrong.
The last statement is subtly wrong but it is wrong, however. The probabilities in quantum mechanics do have exactly the same physical interpretation as the probabilities resulting from ignorance in classical statistical physics – i.e. as the probabilities calculable from the distribution functions on the classical phase space. The actual difference between classical statistical physics and quantum mechanics is the uncertainty principle that governs the latter. The uncertainty principle says that some degree of the ignorance is fundamentally unavoidable, independently of the chosen observer, her measuring apparatus, or methodology. Mathematically speaking, the new feature of quantum mechanics is the nonzero commutator between generic enough observables. It is not just a specific technical feature of some particles; it is a key conceptual rule that holds everywhere in this quantum world: The truths themselves refuse to commute in this Universe.
It seems very obvious to me that Scott Aaronson doesn't understand these basic conceptual findings about the character of probabilities in statistical physics and quantum mechanics. He's not the only one; almost everyone else who loves to write "popular" texts about quantum mechanics these days is similarly deluded. These people maintain some insane anti-Bohr, anti-Heisenberg sentiments that prevent them from seeing that by these assaults against the deepest findings done by these two men (and their school), they are exactly as canonical crackpots as "biologists" who love to constantly assault Darwin's "mistakes".
It's too bad that such a majority of the folks misunderstands the fundamental role of the uncertainty principle as the source of all the novelties of quantum mechanics. If they were using better textbooks, the situation would perhaps be less hopeless. For example, Volume III of Feynman's lectures contains a philosophical Section 2.6 where Feynman exactly explains that
For already in classical mechanics there was indeterminability from a practical point of view.That's the last sentence that follows his explanation that the true novelty of quantum mechanics is the fundamental impossibility to know the values of all observables at the same moment – it is the uncertainty principle, stupid. (The closely related operational interpretation says that a measurement must always affect the system; the more accurate measurements have a greater unavoidable impact.) The word "uncertainty" appears 26 times in the Chapter 2 itself. Ludwig Boltzmann figured out that some important questions in physics – those about thermodynamics – require one to introduce the probability calculus to physics in order to be understood from the first principles and the probabilities wouldn't go away. For this and related reasons, Boltzmann was the forefather of quantum mechanics.
Aaronson writes lots of other idiocies about "who was right about quantum mechanics":
But on the other hand, Einstein and Schrödinger were clearer in realizing that you couldn’t restrict QM to “microscopic phenomena” only using nothing but mountains of verbiage about complementarity—that once you adopted QM consistently, there would be no inherent limit to the size or spatial range of superpositions. Both sides were sort of groping toward points that Everett and Bell would make a lot sharper in the 50s and 60s.Bohr, Heisenberg, and their school would never restrict the applicability of the quantum postulates to microscopic phenomena. Instead, it was their very school that laid the foundations for the research of macroscopic objects using quantum mechanics – Bloch waves as the basic insights about crystals and solid state physics and similar findings about quantum statistical physics, semiconductors, and many other types of macroscopic matter.
The stunning lie that the founders of quantum mechanics have believed that quantum mechanics didn't apply to macroscopic objects is one of the key "Earth is flat" dogmas that the anti-quantum bigots love to brainwash each other with. The reality is that they realized very well that quantum mechanics applies to all phenomena in the world. The macroscopic phenomena don't differ from the microscopic ones because quantum mechanics fails for them, they knew. Quantum mechanics never fails. The macroscopic phenomena differ because the classical physics or at least its logic is becoming also valid, albeit just approximately, for the macroscopic phenomena. And the classical logic is ultimately needed for us to interpret any sharp statements about observed facts and for this reason, the existence of some approximately classically behaving quantities is necessary for our statements about observed facts to be meaningful. But the exact relationship between all these facts is always governed by the laws of quantum mechanics, whether we study microscopic or macroscopic systems.
Einstein and Schrödinger were wrong about all foundational issues of quantum mechanics in which they disagreed with the founders of quantum mechanics. Everett and Bell were semi-wrong, trying to pay lip service to quantum mechanics but being confused about it at the level of "their research" pretty much in the same way as Einstein and Schrödinger. In particular, they have never found – and nobody has ever found – any "third way", a framework that would interpolate between Bohr's and Einstein's views. There can't be any viable "third way"; Bohr was 100% right and Einstein was 100% wrong, everything else is just pure spin and demagogy.
Scott escalates his anti-quantum crackpottery in another comment:
Bram #4: I haven’t read much of Heisenberg, but from the little I have, I think most of what I said about Bohr in #6 would also apply. Bohr and Heisenberg both had the properties ofI have already discussed the point (1). Aaronson et al. only put insufficient focus on the uncertainty principle because they don't understand what the whole novelty of quantum mechanics is all about. It is about the uncertainty principle or, equivalently, about the unavoidable degree of ignorance about observables or, mathematically speaking, about the nonzero commutators between pairs of generic observables. If the commutators were zero, the predictions of quantum mechanics would be reduced back to predictions of classical physics.
(1) putting way more stress on “wave/particle complementarity” and the uncertainty principle than we’d put today,
(2) bizarrely, saying almost nothing about the aspects of QM we do see as central today, like entanglement, the enormous size of Hilbert space, or amplitudes being complex-valued analogues of probabilities,
(3) repeatedly, seeming to walk right up to the cusp of saying that consciousness is implicated in wavefunction collapse, reality is created by our perceptions, and various other “insane” things, but then never saying those things, and
(4) generally, being a lot more ponderous and obscure than not only their successors, but even contemporaries like Schrödinger and Dirac.
Bohr's complementarity conveys the same general "philosophical" principle as Heisenberg's uncertainty principle but the role of "momentum" in Heisenberg's principle is replaced by the "wavelength" (or any related property that waves may possess) which was known to be related to the momentum at least since the 1924 findings by de Broglie. So Heisenberg says that a particle can't have an exact particular position and an exact particular momentum at the same moment; Bohr says that this object can't simultaneously have a well-defined position (be a particle) and preserve a nice interference pattern with some wavelength (be a wave). These two verbal declarations are not quite equivalent but they are close and each of them captures a big part of the "lesson" that may be extracted from the full mathematical formalism of quantum mechanics – something that Bohr, Heisenberg, and others have fully agreed about.
Wolfgang gave an appropriate answer to Scott's first bullet about the "less stress" on Bohr's and Heisenberg's principles:
I am not sure what you mean by that.Exactly – the trouble with Scott is that he is not doing physics. He is indeed attending a superficial debate club where the "debaters" find it OK to remove the "stress" from some key findings in science before they deny them completely because they are inconvenient for their narrative and philosophy.
We understand the propagation e.g. of electrons in a crystal or e.g. absorption / emission of photons thanks to the work of Heisenberg et al. and nothing has changed about the uncertainty principle in recent decades as far as I know.
Physics is not a debate club where we put more or less stress on some facts...
The point (2) by Aaronson is completely incorrect, too. The tensor-product structure of the Hilbert space for a composite system has been understood since the mid 1920s, since the very birth of quantum mechanics, and it has been included among the basic postulates of quantum mechanics (check the second bullet) ever since.
One may also view the tensor-product structure of the composite systems' Hilbert spaces as a derived technical result from the representation theory of the operators; the truly fundamental starting point isn't the tensor-product structure itself but rather the vanishing commutators between observables describing two different, decoupled subsystems. It's the existence of these two (or many) mutually commuting operator subalgebras that forces the tensor-product structure on the minimal nontrivial representations. All these points were implicitly clear to Heisenberg since his 1925 papers.
The "exponentially growing" dimension of the Hilbert space that's been known for nearly 90 years also implies – and was understood by Bohr et al. to imply – that virtually all states in the Hilbert space are entangled i.e. refusing to be written as a (tensor) product of states in the factor spaces. This "entanglement" is really a "negative result" – impossibility to factorize a general multi-particle state – and a trivial one. It only makes sense to mention this trivial fact in front of students who still want to think classically (all pure states happen to be "unentangled" in classical physics) but it isn't enough to uncover the novelties of quantum mechanics (most classical probability distributions on the phase space also refuse to be factorized!). It's nothing else than the uncertainty principle that captures all the key novelties of quantum mechanics. Nothing whatsoever has changed about these fundamental issues since the 1920s.
"Amplitudes are complex-valued analogues of probabilities" is an insight that Max Born of the Copenhagen school formulated this clearly in 1926 and he has won the 1954 Nobel prize in physics for that. The claim that this was unknown to the Copenhagen school is a 100% lie, too. They not only knew it; they must be 100% credited with this groundbreaking discovery.
Now, let me copy the third point by Aaronson so that you don't have to return.
(3) repeatedly, seeming to walk right up to the cusp of saying that consciousness is implicated in wavefunction collapse, reality is created by our perceptions, and various other “insane” things, but then never saying those things, andThe reason why these "insane" statements were "almost made" by Heisenberg but not quite is that they are "almost true" but not quite and Heisenberg wrote the exact truth concerning all these conceptual questions. Quantum mechanics isn't a classical model of reality; it is a framework to formulate right statements, explanations, and (probabilistic) predictions of the real phenomena.
Because it switches from a "classical model" to "what correct statements we can make about Nature", it requires "someone" to care about the truth, about the validity of statements, and it has been a tradition to "personify" each of the frameworks that care about the validity of propositions about observations and call them an "observer". The "consciousness" is needed to the extent to which the existence of "someone or something" that can perceive the results of measurements is necessary to "animate" the predictions of quantum mechanics.
No connections with any special features of animals, human beings, their DNA, their location, anthropomorphic themes, or any particular religious or spiritual movements or religions has ever been made by Heisenberg because quantum mechanics requires and implies no such connections. But the in-principle subjective nature of the knowledge about the observed facts that are related by the laws of quantum mechanics – the in-principle multiplicity of possible frameworks in which propositions may be chosen before they are assigned truth values – is a genuine and irreversible result of modern science. It may "remotely resemble" some of the "insane" spiritual proclamations but it is true, anyway. Hardcore Marxists like Scott Aaronson may dislike this breakthrough because it's just another result showing that their "objective materialist" hero Marx was completely wrong – but the self-evident fact that Marx was an idiot can't make quantum mechanics any less valid. Heisenberg has not only understood these key principles of modern science since the mid 1920s; he was largely the first man in the world who realized them.
That's why Heisenberg was one of the greatest 20th century scientists while Scott Aarronson who not only failed to discover these insights himself but who still misunderstands them even though they have been around for nearly 90 years is just another obnoxious, permanently dissatisfied anti-quantum kibitzer (yes, I know some Yiddish, too).
(4) generally, being a lot more ponderous and obscure than not only their successors, but even contemporaries like Schrödinger and Dirac.In general, anti-quantum cranks find texts about quantum mechanics "ponderous" because anti-quantum cranks have a limited IQ and an excessive emotional relationship to outdated philosophical beliefs which is why they prefer to read porn novels and superficial philosophical (anti-quantum) books – among other light genres – than to read high-quality texts that reflect how Nature actually works according to modern science.
And that's the memo.