Honestly I’m somewhat critical of “paradigm shifts.” There is a bit of an obsession in academia when it comes to metaphysics, both in philosophy departments and physics departments, of academics trying to come up with “paradigm shifts.” Academics seem to have a natural tendency to find extraordinary explanations more interesting, and seem to have a strong bias against more banal explanations, always seeing them as dubious and questionable.
The result is that there is almost never an attempt to steel man the non-extraordinary interpretation before adopting the extraordinary one. “Paradigm shifts” are almost taken to be de facto correct, whereas someone who defends the old paradigm is often met with extreme hostility, painting them as an old outdated kook who just needs to die off so we can all accept the new mode of thinking.
The reality is that hardly any of the “paradigm shifts” in the last century and a half have actually been empirically necessary, and the old way of thinking has always been defensible. Physicists often create the false illusion that a paradigm shift is a necessity by falsely equating the physics with their metaphysics and then using the success of the physics to “prove” their metaphysics, equating anyone who doesn’t believe their metaphysics as equivalent to a science denier, but these things are indeed logically separable.
Take, for example special relativity, that supposedly “proves” space and time is relative because it empirically predicts that rods and clocks can deviate from one another. The physicist John Bell published a paper “How to Teach Special Relativity” pointing out that this only proves space and time is relative if we define space and time in terms of a metric corresponding to whatever any arbitrary rod and clock measures, because that metric is undeniably relative.
If we instead define space and time in terms of whatever a specific rod and clock measures, then we will have an absolute space and time metric. Deviations from that metric would then be interpreted as apparent effects, and thus there would be a distinction between real time and real space vs apparent time and apparent space. This would be wholly consistent with an absolute, Galilean space and time, while also being wholly consistent with the empirical predictions of special relativity.
Indeed, when Einstein introduced his special relativity in 1905, it actually made no new empirical predictions at all, because Lorentz had proposed a theory in 1904 which was mathematically equivalent and thus made all the same predictions, with the only difference being that Lorentz does not assume an invariant speed of light (which, despite common misconception, has never been measured, as it is not physically possible to measure the speed of light) with a preferred rest frame of the universe.
Einstein’s criticism of Lorentz’s theory was that he did not think such a preferred rest frame was sensible because there is no good reason to treat any particular frame as “preferred” over any others, and he developed special relativity, and late general relativity, on these premises. However, Einstein’s criticism turned out to be false, because to actually make GR consistent with our universe, you have to adopt specifically the Friedman solutions to Einstein’s field equations, and in the Friedman solutions, GR spontaneously gives rise to an empirically measurable rest frame of the universe known as cosmic time, and you can measure your motion relative to the rest of the universe by analyzing the cosmic dipole in the cosmic background radiation.
A similar thing is also true of quantum mechanics. Many physicists, like Niels Bohr, insisted that we must interpret it as a “paradigm shift” whereby particles have no real values until you look at them, then suddenly “collapse” down to a definite state. Later, you had Hugh Everett who went even further and claimed that we should stop believing that a single universe exists, but in fact there is an infinite multiverse with an infinite number of copies of you.
However, the physicist Dmitry Bloktinsiev had pointed out that you can just interpret quantum mechanics as a holistic form of statistical mechanics, whereby particles never spread out as waves, they have real values at all times, but that the waves are purely a description of their statistical dynamics mechanics and thus are only observable over an ensemble of systems. Hence, we see waves in experiments we run many many times over (like firing a single photon many many times over and accumulating statistics in the double-slit experiment forms a wave-like interference pattern), but we never see it in the individual case (the photon, in any individual case, is always found to exist in a single place, with a single trajectory).
Bloktinsiev made this argument in a 1950s paper and a 1960s book, although he never really rigorously formulated it. However, fairly recently in 2025, the physicist Jacob Barandes published a paper “The Stochastic-Quantum Correspondence” proving that quantum mechanics is mathematically equivalent to a holistic statistical theory, the holism coming from history-dependence, what he calls non-Markovianity. This history-dependence means that the behavior of a system is a function over its whole statistical history, which directly leads to a breakdown in classical divisibility.
In some ways, this breakdown in classical divisibility is a “paradigm shift,” but far less radical than something like Copenhagen or Many Worlds. There is no empirical necessity to actually believe that particles do not have real values when you are not looking at them, that there are invisible infinite-dimensional waves, that these waves “collapse” when you look at them, or that there is a multiverse, or any of these exotic interpretations.
It is perfectly logically consistent to interpret quantum theory as just a form of holistic statistical mechanics, and so particles really are just particles that evolve in a way with real values at all times in the real world, but you just cannot know those values because the theory is stochastic and not deterministic, and all the deviations between classical and quantum mechanics arise solely from the lack of divisibility in quantum systems.
The “paradigm shifts” are always overblown and overstated. There is actually nothing logically inconsistent with interpreting reality in terms of just particles with well-defined positions at all times that evolve through absolute spacetime. That is logically and mathematically consistent with unmodified orthodox quantum theory and special relativity. The “paradigm shifts” are ultimately not necessary, but an intentional metaphysical choice. John Bell tried to point this out in his paper “On the Impossible Pilot Wave,” that, whether or not pilot wave theory is correct, the very fact it is even possible to formulate such a theory, even if the theory turns out to be wrong, disproves the idea that it is empirically necessary to adopt exotic interpretations of QM, like Copenhagen or Many Worlds.
You see these “paradigm shifts” in philosophy as well, like David Chalmers supposedly proving the “hard problem of consciousness,” but they do not steelman the simpler position either in this case, which is that of direct realism in which the problem is not derivable.
I think the Kuhnian “normal science --> anomalies --> paradigm shift” model of scientific progress can be seen as a kind of dialectical process.
Honestly I’m somewhat critical of “paradigm shifts.” There is a bit of an obsession in academia when it comes to metaphysics, both in philosophy departments and physics departments, of academics trying to come up with “paradigm shifts.” Academics seem to have a natural tendency to find extraordinary explanations more interesting, and seem to have a strong bias against more banal explanations, always seeing them as dubious and questionable.
The result is that there is almost never an attempt to steel man the non-extraordinary interpretation before adopting the extraordinary one. “Paradigm shifts” are almost taken to be de facto correct, whereas someone who defends the old paradigm is often met with extreme hostility, painting them as an old outdated kook who just needs to die off so we can all accept the new mode of thinking.
The reality is that hardly any of the “paradigm shifts” in the last century and a half have actually been empirically necessary, and the old way of thinking has always been defensible. Physicists often create the false illusion that a paradigm shift is a necessity by falsely equating the physics with their metaphysics and then using the success of the physics to “prove” their metaphysics, equating anyone who doesn’t believe their metaphysics as equivalent to a science denier, but these things are indeed logically separable.
Take, for example special relativity, that supposedly “proves” space and time is relative because it empirically predicts that rods and clocks can deviate from one another. The physicist John Bell published a paper “How to Teach Special Relativity” pointing out that this only proves space and time is relative if we define space and time in terms of a metric corresponding to whatever any arbitrary rod and clock measures, because that metric is undeniably relative.
If we instead define space and time in terms of whatever a specific rod and clock measures, then we will have an absolute space and time metric. Deviations from that metric would then be interpreted as apparent effects, and thus there would be a distinction between real time and real space vs apparent time and apparent space. This would be wholly consistent with an absolute, Galilean space and time, while also being wholly consistent with the empirical predictions of special relativity.
Indeed, when Einstein introduced his special relativity in 1905, it actually made no new empirical predictions at all, because Lorentz had proposed a theory in 1904 which was mathematically equivalent and thus made all the same predictions, with the only difference being that Lorentz does not assume an invariant speed of light (which, despite common misconception, has never been measured, as it is not physically possible to measure the speed of light) with a preferred rest frame of the universe.
Einstein’s criticism of Lorentz’s theory was that he did not think such a preferred rest frame was sensible because there is no good reason to treat any particular frame as “preferred” over any others, and he developed special relativity, and late general relativity, on these premises. However, Einstein’s criticism turned out to be false, because to actually make GR consistent with our universe, you have to adopt specifically the Friedman solutions to Einstein’s field equations, and in the Friedman solutions, GR spontaneously gives rise to an empirically measurable rest frame of the universe known as cosmic time, and you can measure your motion relative to the rest of the universe by analyzing the cosmic dipole in the cosmic background radiation.
A similar thing is also true of quantum mechanics. Many physicists, like Niels Bohr, insisted that we must interpret it as a “paradigm shift” whereby particles have no real values until you look at them, then suddenly “collapse” down to a definite state. Later, you had Hugh Everett who went even further and claimed that we should stop believing that a single universe exists, but in fact there is an infinite multiverse with an infinite number of copies of you.
However, the physicist Dmitry Bloktinsiev had pointed out that you can just interpret quantum mechanics as a holistic form of statistical mechanics, whereby particles never spread out as waves, they have real values at all times, but that the waves are purely a description of their statistical dynamics mechanics and thus are only observable over an ensemble of systems. Hence, we see waves in experiments we run many many times over (like firing a single photon many many times over and accumulating statistics in the double-slit experiment forms a wave-like interference pattern), but we never see it in the individual case (the photon, in any individual case, is always found to exist in a single place, with a single trajectory).
Bloktinsiev made this argument in a 1950s paper and a 1960s book, although he never really rigorously formulated it. However, fairly recently in 2025, the physicist Jacob Barandes published a paper “The Stochastic-Quantum Correspondence” proving that quantum mechanics is mathematically equivalent to a holistic statistical theory, the holism coming from history-dependence, what he calls non-Markovianity. This history-dependence means that the behavior of a system is a function over its whole statistical history, which directly leads to a breakdown in classical divisibility.
In some ways, this breakdown in classical divisibility is a “paradigm shift,” but far less radical than something like Copenhagen or Many Worlds. There is no empirical necessity to actually believe that particles do not have real values when you are not looking at them, that there are invisible infinite-dimensional waves, that these waves “collapse” when you look at them, or that there is a multiverse, or any of these exotic interpretations.
It is perfectly logically consistent to interpret quantum theory as just a form of holistic statistical mechanics, and so particles really are just particles that evolve in a way with real values at all times in the real world, but you just cannot know those values because the theory is stochastic and not deterministic, and all the deviations between classical and quantum mechanics arise solely from the lack of divisibility in quantum systems.
The “paradigm shifts” are always overblown and overstated. There is actually nothing logically inconsistent with interpreting reality in terms of just particles with well-defined positions at all times that evolve through absolute spacetime. That is logically and mathematically consistent with unmodified orthodox quantum theory and special relativity. The “paradigm shifts” are ultimately not necessary, but an intentional metaphysical choice. John Bell tried to point this out in his paper “On the Impossible Pilot Wave,” that, whether or not pilot wave theory is correct, the very fact it is even possible to formulate such a theory, even if the theory turns out to be wrong, disproves the idea that it is empirically necessary to adopt exotic interpretations of QM, like Copenhagen or Many Worlds.
You see these “paradigm shifts” in philosophy as well, like David Chalmers supposedly proving the “hard problem of consciousness,” but they do not steelman the simpler position either in this case, which is that of direct realism in which the problem is not derivable.