Chapter XVIII: In-Principle And In-Practice | The Philosophy Of Science by Steven Gussman [1st Edition]

        “In theory, theory and practice are the same. In practice, they are not.”

        – Attributed to Albert Einstein^I

        A very important dichotomy that even working scientists frequently mix up is the difference between in-principle and in-practice.  In-principle means something is actually, ideally, ontologically the case.  For example, in-principle, one can deterministically predict which face a flipped coin or rolled die will land on, every single time.  This is because the system is understood: in-principle, if you had perfectly accurate and precise initial condition measurements (such as the object's orientation and the force exerted by the thumb) and applied Newton's kinematic equations, you would get the same (correct) answer every time you performed the calculation.  In-practice, however, means that in the real world where you do not (and perhaps cannot realistically) have perfect measurements, your predictions might fall prey to chaotic outcomes as a result^II (in fact, even without chaos theory, the fact is that you likely do not have a good way to measure the initial conditions, nor time to calculate the prediction, during a natural coin flip).

        When you say something cannot be done even in-principle, you are saying it is strictly impossible (read: physically impossible).  When you say that something cannot be done in-practice, you are merely saying that (as far as you understand it) it is an intractable engineering project for people to undertake under real-world conditions (at least given present understanding and technology).^III  A result of determinism is that, in-principle, counter-factuals (wondering about the courses of events in “what if” scenarios) require the changing of initial conditions at the branching point (this is typically the point of such "what if" exercises).  But sometimes, scientists become mixed up due to their misunderstanding of statistical mechanics, quantum physics, or even evolutionary theory.  In my view, thermodynamics is graded on a curve (Eddington famously said that it is the only law he's absolutely sure of)^IV because it is an in-practice law: technically, entropy could decrease, in-principle!^V  Boltzmann showed that the second law of thermodynamics (that entropy, or disorder, increases over time) is merely a statistical law: entropy tends to increase because of the astronomical probability in favor of this happening in-practice, but not because it is an actual, in-principle, deterministic law.^VI  The reason the second law of thermodynamics tends to be where much confusion occurs is that it truly pushes towards the limit of where in-practice becomes closer to in-principle (saying that it is “unlikely” the universe's entropy will decrease tomorrow is an understatement—there are few predictions one can hang their hat on more safely than that it will increase); but one must remember that these are ultimately differences of kind and not degree.  The fact is that decreasing entropy is merely astronomically improbable, whereas exceeding the speed of light is (as far as we currently understand) literally impossible.  Some think that if you could rewind the clock and hit play on the world again, that you would get a different outcome as a result, but this is untrue.  The "statistical" in “statistical mechanics” is not meant to imply in-principle, literal randomness in the motions of particles.  Each particle is in random motion with respect to the microstate of the whole system (that is, with respect to the other particles)—but each is where it is, and moving at the speed it is moving at, because of in-principle deterministic collisions in its recent past.  As a result, one would not just probabilistically get the same macrostate upon rewinding and playing such a system again; one would get the exact same microstate as well.  The reason different trials of such many-particle systems have different (and basically random) microstates each time the system is set up, is because the system is not being set up (and in-practice it is at least currently intractable to do so) in the same microstate—where the number of particles, and each particle's initial position and velocity are set to the same starting point in the different trials.  It is our ignorance and inability in-practice to set up such systems which causes us to use statistical mechanics rather than directly track the individual motions of particles in this way.  But when we do thought experiments such as rewinding time and playing it back, we are doing so in-principle, and must specify any counter-factual change in conditions we want to make; we would be wrong to assume that statistical mechanics (and I argue, even quantum physics—though this is far more controversial) is such that we may not specify that the conditions are exactly the same upon rewinding and still claim a different outcome.  One can see why by recognizing that statistical mechanics is based on the idea that groups of individual particles' interactions (the microstate) gives rise to a macrostate, and that the statistical behavior of such many-particle systems (how microstates are related to macrostates) are constrained by statistical mechanics.^VII  The empirical success of this theory was taken as indirect empirical evidence for the existence of molecules at a time when the ontological status of these tiny objects was controversial (we have since directly observed even subatomic particles much smaller than these).^VIII  Given these facts, we must understand the rewinding process of our thought experiment to be the smooth rewinding of the motions of each individual particle (it is meaningless to “rewind” the statistical mechanics which is not an deterministic ontological mechanism but a statistical prediction).  When we do that, we get the very same initial conditions for each particle (in a way that we currently cannot achieve with in-practice experiments on gases), leading not just statistically to the same macrostate, but deterministically to the very same microstate and attendant macrostate.^IX

        Now take evolutionary biology.  There is a popular school of though (perhaps most associated with biologist Stephen Jay Gould)^X which takes the pseudo-random parts of evolution (both mutation and historical circumstance) and treats them as truly random.  Such thinkers mistake the complexity of their system as literal, in-principle chaos, imagining that if one turned the clock back and hit play, that different mutations would arise and different adaptive (and genetic drifting) evolutionary histories would unfold, in which wildly different species and environments would result when the clock caught up to the time in which it had been rewound.  But mutations are only random with respect to phenotype, not literally effects with no deterministic cause (again, this becomes complicated by the fact that some mutations may be caused by quantum processes such as cosmic rays colliding with gametes,^XI at which point I fall back on my argument that fundamental physics will be shown not to be in-principle random, after all).  The cause of the mutations, the locus mutations occur at, and the character of the mutations are not themselves in-principle random: a cosmic ray, for example, was going to hit that particular gene and change it in that particular way (we are just not currently in possession of the in-principle laws nor in-practice ability to apply them to make such predictions), and it would happen again, resulting in the same phenotypic change, if one rewound the clock and hit play without explicitly changing those initial conditions as part of a counter-factual exercise.^XII

        Finally, we come to the trickiest case of quantum physics—a case where I may perhaps be wrong, but in which I cannot possibly see how, in-principle.^XIII  Quantum physics, to the best of our understanding, has random components to it.  Due to the requirements of determinism and mechanical philosophy, this precludes it from being a fundamental theory, and yet it is currently the lowest level theory of energy that we have (on the ontological stack).  Most physicists treat quantum physics as-if it is fundamental, and then extrapolate extreme philosophical consequences such as reality being fundamentally random (or even more incoherent, that reality is somehow un-real, merely measured and made up at the moment that it occurs).  Most physicists would say that even if you found an election at a particular place in an atom, if you rewound nothing but the clock and left the initial conditions the same, you would find that electron in a different spot when you hit play again (in fact, they would tell you that while there are many electrons, we cannot even in-principle tell them apart and so it is meaningless to even try and find a particular electron again in the counter-factual future).  This is because such thinkers interpret the chance elements that arise from their own ignorance in-practice as in-principle chance!  Unfortunately, without time travel, we cannot easily check this prediction and must make due with the fact that our blunt tools only seem to tell us regions in which to expect electrons.  If we could rewind the clock and hit play, one could falsify (or confirm!) quantum anti-realism (or quantum indeterminacy) immediately.  It is clear to me that the philosophy of science is such that in-principle, mechanical philosophy and determinism must be met for any full theory of an ontological phenomenon (because the world is a deterministic machine).  Thus, whereas I can already convince one that statistical mechanics and evolutionary history do not contain random components (ignoring quantum effects) as-formulated, quantum physics currently does, as-formulated—I believe this must be wrong, that quantum physics is, like statistical mechanics, merely a probabilistic approximation but not the fundamental theory underpinning the cosmos.  If one performs a thought experiment where the clock is rewound on a quantum system, in-principle, one should expect the same exact outcome of the quantum measurement.

        It is scientifically uncontroversial that the first two of the above examples are examples of in-practice pseudo-randomness being misconstrued as in-principle randomness, because we lack the hard theoretical laws to predict every aspect of these systems and, in our hubris, we interpret our ignorance as uncertainty built into the cosmos.  I believe the same must be true of quantum physics (and that this will one day be shown to be the case), as an argument from philosophy of science.  Remember: in-principle speaks of possibility and impossibility; in-practice speaks of probability and improbability.

Footnotes:

0. The Philosophy Of Science table of contents can be found, here (footnotephysicist.blogspot.com/2022/04/table-of-contents-philosophy-of-science.html).

I. See Our Mathematical Universe by Tegmark (pp. 92).

II. Chaos theory is the discovery (discovered through computer simulations modeling weather) that even given deterministic laws, many systems are highly sensitive to differences in initial conditions such that the outcomes are highly (rather than subtly) divergent (these “differences” in initial conditions could easily be of the size of measurement error or otherwise occur beyond the precision of the measurement apparatus), see “A Profusion Of Place”by Gussman (https://footnotephysicist.blogspot.com/2020/03/a-profusion-of-place-part-i-of-unity.html#FN42B) which further cites The Great Unknown by du Sautoy (pp. 21-71).

III. Interestingly, many people end up being wrong when they speculate that they believe something is in-principle doable yet in-practice impossible: Einstein for example did not believe that gravitational waves nor black holes (predictions of his general theory of relativity; the former he nevertheless believed in, the latter he did not) would ever be observed, yet both have been in the century since his death, winning Nobel Prizes in physics, see Parallel Worlds by Kaku (pp. 116, 258-259); "The Nobel Prize in Physics 1993: Russell A. Hulse And Joseph H. Taylor Jr." (The Royal Swedish Academy Of Sciences / Nobel Foundation) (1993) (https://www.nobelprize.org/prizes/physics/1993/press-release/) (for indirect detection of gravitational waves); "The Nobel Prize in Physics 2017: Rainer Weiss, Barry C. Barish, And Kip S. Thorne" (The Royal Swedish Academy Of Sciences / Nobel Foundation) (2017) (https://www.nobelprize.org/prizes/physics/2017/press-release/) (for direct detection of gravitational waves, and indirect detection of black holes); and "The Nobel Prize in Physics 2020: Roger Penrose, Reinhard Genzel, Andrea Ghez" (The Royal Swedish Academy Of Sciences / Nobel Foundation) (2020) (https://www.nobelprize.org/prizes/physics/2020/summary/) (for indirect evidence for the existence of black holes). Though it has not (yet!) won a Nobel Prize, see “Astrophysicist Sheperd Doeleman Awarded National Academy Of Sciences Henry Draper Medal: Prestigious Award Recognizes Doeleman’s Work In Realizing The First Image Of A Supermassive Black Hole” by Nadia Whitehead (Smithsonian) (2021) (https://www.si.edu/newsdesk/releases/astrophysicist-sheperd-doeleman-awarded-national-academy-sciences-henry-draper) (for direct detection of a black hole).

IV. See Enlightenment Now by Pinker (pp. 16-17).

V. That is to say that a deterministic theory which attempts a complete model of a system with no ignorance, providing a deterministic mechanism (such as the general theory of relativity) is much easier to falsify because it makes a narrow, exact prediction; on the other hand, when statistical mechanics or quantum physics makes predictions, they are fuzzy, probabilistic predictions—it is like grading the former based on whether they can hit a bull's eye whereas the latter gets the same grade as long as it manages to hit anywhere on the target board. As impressive as the empirical predictions of both theories are, I believe (contrary to most), that the general theory of relativity is actually the most impressive empirically-verified theory we have, for this reason. For a contrarian essay which mentions the statistical nature of entropy, see "Entropy" by Bruce Parker (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25524) in This Idea Must Die edited by Brockman (pp. 40-43). For a courageous argument that the general theory of relativity will have the last laugh over quantum physics, see Fashion, Faith, And Fantasy by Penrose (at least pp. 198-215).

VI. See Modern Physics by Serway et al. (pp. 334-345) (I was assigned sections of this textbook in a “Modern Physics” course in college some six-or-seven years back, so I do not remember how thoroughly I read a given portion) and "Entropy" by Parker (https://www.edge.org/response-detail/25524) in This Idea Must Die edited by Brockman (pp. 40-43).

VII. See Modern Physics by Serway et al. (pp. 334-345).

VIII. See Modern Physics by Serway et al. (pp. 334-336).

IX. For more on these topics, look forward to the “Physics” chapter in the “Ontology” section.

X. See “Evolutionary Contingency” by Douglas H. Erwin (CellPress Current Biology) (1998) (https://www.cell.com/current-biology/pdf/S0960-9822(06)02143-9.pdf) (though I have only read the sentence containing Gould's name for confirmation of my claim, as well as the section headers) which further cites Wonderful Life: The Burgess Shale And The Nature of History by Stephen Jay Gould (W. W. Norton & Company) (1989 / 1990 / 2007) (though I have not yet read this work).

XI. See Cosmos by Sagan (pp. 34, 245).

XII. Interestingly, there is a related debate as to the predictability of which adaptations are likely to be selected—essentially an argument about just how different selective-environments really are, with one camp seeing infinite complexity, and the other suspecting regularity inherited from the shared laws of physics and chemistry. For a range of views on the tension between variety and regularity in evolutionary history, see Cosmos by Sagan (pp. 29, 38-40); "The Genetic Book Of The Dead" by Richard Dawkins (Harper Perennial) (2017 / 2018) (https://www.edge.org/response-detail/27237), “Variety” by Lee Smolin (Harper Perennial) (2017 / 2018) (https://www.edge.org/response-detail/27138), “Phylogeny” by Richard Prum (Harper Perennial) (2017 / 2018) (https://www.edge.org/response-detail/27179), and “Chronobiology” by Douglas Rushkoff (Harper Perennial) (2017 / 2018) (https://www.edge.org/response-detail/27083), all from This Idea Is Brilliant edited by Brockman (pp. 8-10, 124-126, 262-263, 452-454); "Fully Random Mutations" by Kevin Kelly (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25264) and "Human Evolutionary Exceptionalism" by Michael McCullough (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25434), both from This Idea Must Die edited by Brockman (pp. 174-176, 447-451); “Evolutionary Contingency” by Erwin (https://www.cell.com/current-biology/pdf/S0960-9822(06)02143-9.pdf); “Evolutionary Convergence” by Simon Conway Morris (CellPress Current Biology) (1998) (https://www.cell.com/current-biology/pdf/S0960-9822(06)02143-9.pdf) (though I have not yet read these latter-most two CellPress pieces); and "Sam Harris | Bret Weinstein's DarkHorse Podcast #8" by Bret Weinstein and Sam Harris (DarkHorse) (2019) (https://www.youtube.com/watch?v=id6AqKIxd94&t=) (though I have only so far heard the first 55 minutes and 23 seconds, I have already heard B. Weinstein agree with me that the “many worlds” interpretation of quantum mechanics is ridiculous and disagree with me about determinism, which he utterly denies). There have been a spate of recent books on convergent evolution, see The Equations Of Life: How Physics Shapes Evolution by Charles S. Cockell (Basic Books) (2018); The Zoologist's Guide to the Galaxy: What Animals on Earth Reveal About Aliens—and Ourselves by Arik Kershenbaum (Penguin Press) (2021); Furry Logic: The Physics of Animal Life by Matin Durrani and Liz Kalaugher (Bloomsbury Sigma) (2016 / 2017); Improbable Destinies: Fate, Chance, And The Future Of Evolution by Jonathan B. Losos (Riverhead Books) (2017); Evolutionary Worlds Without End by Henry Plotkin (Oxford University Press) (2010); and Imagined Life: A Speculative Scientific Journey among the Exoplanets in Search of Intelligent Aliens, Ice Creatures, and Supergravity Animals by James Trefil and Michael Summers (Smithsonian Books) (2019) (though I have yet to read any of these works I have been tracking).

XIII. As you read this paragraph, recall what I have already written on the topic (especially in the “Determinism” and “Mechanical Philosophy” chapters), wherein I further cite others across the spectrum on the quantum foundations debate: Quantum Physics by Raymer; What Is Real? by Becker; Our Mathematical Universe by Tegmark; Einstein's Unfinished Revolution by Smolin (though I have only yet read through the first half of pp. xxii in this work); Fashion, Faith, And Fantasy by Penrose; and “In Defense Of Philosophy (Of Science)” by Gussman (https://footnotephysicist.blogspot.com/2021/05/in-defense-of-philosophy-of-science.html). For more views on quantum foundations, see "The Collapse Of The Wave Function" by Freeman Dyson (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25350), "There Is No Reality In The Quantum World" by Anton Zeilinger (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25548), "Quantum Jumps" by David Deutsch (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25453), "The Uncertainty Principle" by Kai Krause (Edge / Harper Perennial) (2014 / 2015) (https://www.edge.org/response-detail/25531), all from This Idea Must Die edited by Brockman (pp. 73-76, 107-108, 253-255); "Maxwell's Demon" by Jimena Canales (Harper Perennial) (2017 / 2018) (https://www.edge.org/response-detail/27214) in This Idea Is Brilliant edited by Brockman (pp. 193-196); and "What Twentieth-Century Physics Says About The World Might Be True" by Carlo Rovelli (Edge / Harper Perennial) (2006 / 2007) (https://www.edge.org/response-detail/11431) from What Is Your Dangerous Idea? edited by Brockman (pp. 120-121).