The quote in the question is a paraphrase of the physicist Max Planck's statement:
Eine neue wissenschaftliche Wahrheit pflegt sich nicht in der Weise durchzusetzen, daß ihre Gegner überzeugt werden und sich als belehrt erklären, sondern vielmehr dadurch, daß ihre Gegner allmählich aussterben und daß die heranwachsende Generation von vornherein mit der Wahrheit vertraut gemacht ist.
A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.
Was he right, with regards to major paradigm shifts in the mid 19th to late 20th centuries? I'm specifically thinking about Planck's own field of theoretical physics, as well evolutionary biology (Darwinian evolution vs. competing models, the Recent African Origin model of human evolution, the abandonment of hereditary telegony, etc.), anthropology, and linguistics (UG vs competing models). Did modern accepted theories "win over" scientists from competing models, or did the old models' advocates just die?
I'm also concerned solely with the transmission and acceptance of scientific models among scientists themselves, not popular culture responses to scientific theories.
Thank you!
Did modern accepted theories "win over" scientists from competing models, or did the old models' advocates just die?
They won converts. Not always quickly, and this is where Planck's comment comes from. Usually, there is an existing older theory modelling/explaining much of the same things as the new theory. Generally, for wide acceptance, the new theory needs to be better than the old theory. "Better" is not always obvious. Things that can affect better include:
The theory explaining a wide range of phenomena. Often, the old theory has an advantage here, since more time and effort has gone into using it to model/predict/explain things. It can take time for the new theory to accumulate a sufficient body of work to demonstrate broad explanatory power.
Explaining some poorly-understood phenomena well, with the same mechanism as well-understood phenomena. E.g., special relativity explains (a) mechanics at high speeds and (b) electromagnetic phenomena involving relative motion in a unified way with low-speed mechanics and no-motion electromagnetics.
Being easily understood.
The second of these can result in a new theory being considered more or less acceptable, but in need of better experimental evidence. That "explaining some poorly-understood phenomena well" can require experimental verification of a prediction of the new theory. A good example is the rapid acceptance of relativity (both the special and general theories) in the English-speaking physics world in the early 1920s, when better experimental evidence was available (notably, Eddington's measurements of the gravitational deflection of light during the 1919 solar eclipse).
The acceptance of atoms took even longer - atoms were too small to see, and direct measurement wasn't possible. Indirect evidence was needed, but indirect evidence can be less convincing. Key steps were quantitative theory of measurable phenomena distinct from the indirect evidence from chemistry, and experimental evidence supporting that theory. These were provided by Einstein, in his 1905 work on Brownian motion, and Perrin's 1911 experimental verification (which led to a Nobel prize for Perrin).
For some theories, it is less a question of quantitative predictions than the plausibility of qualitative parts of the theory. Wegener's theory of continental drift was rejected for decades before plausible mechanisms for continental drift were developed. When these mechanisms were proposed, and it appeared that they were actually operating, continental drift was rapidly accepted. A slow-to-be-accepted theory making it in the end, due to better explanations and better evidence rapidly winning converts, rather than Planck's model of death-and-replacement.
Being understandable helps a lot. This isn't just a matter of whether the theory is, e.g., very difficult to work with mathematically. If a theory is littered with ad hoc assumptions, it's less likely to be widely accepted. One example of the acceptance of a theory being relatively slow was Maxwell's electrodynamics. His full theory was published in 1865, with his widely read book of 1873 spreading it further. However, acceptance of his theory was hung up on two main points: the complexity of his theory, and his presentation of it as a theory of the electromagnetic aether. Hertz's work on Maxwell's theory was essential to its wide acceptance in German physics - he simplified Maxwell's equations down to the modern version (the 4 field equations in D,E,B, and H) from Maxwell's version involving both the fields and the potentials, and presented an alternative derivation, not dependent on a particular model of an electromagnetic aether.
Quantum mechanics had a similarly slow, and quite controversial, road to acceptance. First, the early theories had multiple ad hoc assumptions. Second, the connections between quantum mechanics and the older classical theories hadn't yet been made (e.g., proper connection with classical mechanics took until de Broglie's work in 1923). Third, there is much about the theory that still isn't understood (e.g., what is the wavefunction?, is collapse of the wavefunction real?, etc.) - there are multiple competing interpretations of quantum mechanics. Despite Planck's statement, quantum mechanics did win converts - of the pioneers, some grew up with the early theory (e.g., Bohr, Born, Schrödinger), but others were converts (e.g., Sommerfeld).
Where new theories contradict accepted conceptions/mis-conceptions, they can have a harder time. This was yet another obstacle that quantum mechanics faced. An example of a much quicker-to-resolve controversy is that over the Hanbury Brown-Twiss effect. A simple "billiard ball" picture of a photon is contradicted by the effect, but, judging by the controversy itself, this was a common mental model. It was not just a case of Hanbury Brown and Twiss against the world, but of proponents of both sides of the controversy arguing against the other - Hanbury Brown and Twiss had converts. For more on this controversy, see
Darwinian evolution vs competing theories is a complex case. The idea of evolution - and much evidence supporting it - was around long before Darwin. If Darwin's work is seen as a step in the ongoing process of acceptance of evolution, it provided a coherent explanation of the process in place of earlier ad hoc suggestions. The coherence of Darwin's theory, and the wealth of evidence he presented, led to a rapid acceptance of evolutionary theory - rapid, that is, in comparison to the slow progress of the century before. Mayr discusses Darwinism along these lines - Darwin's Origin of species as "the midpoint of the so-called Darwinian revolution rather than its beginning":
Mayr notes the disparity in acceptance of Darwin's theory, with some (e.g., Agassiz) never converting, and many partial converts (e.g., accepting evolution, but not Darwin's mechanism of natural selection), and full converts.
Planck's general point, that scientific acceptance is frequently generational, and that people who have massive "sunk costs" in a disproved theory rarely convert or re-tool, is accurate. There have been many qualitative and even quantitative studies to this effect, looking at the ages of scientists and how it correlates to their willingness to adopt new theories, and looking at how, over time, you see a "displacement" (again, correlated with age) of theories in scientific journals.
The place where it can get more complicated is when you have other factors playing a role as well. So anthropology is an interesting example: the shift from the racist anthropology of the Madison Grant sort to the ethnographic pluralism of the Boasians was both a generational displacement as well as a deliberate "take over" of the main institutional bodies by said Boasians. Boas and his school won not only because Boas (literally) out-produced new PhDs who took his approach, but because he helped them seize control of the American Anthropological Association and from there had immense power to dictate the future direction of the field. That's a little more "active" than Planck's model would have you think.
One of the points that Kuhn makes in The Structure of Scientific Revolutions, that I think is a good one, is that the "paradigms" that win out are not necessarily the ones that give the best or most complete answers (certainly the early quantum revolution did not do that), but the ones that allow practitioners, especially new ones, to ask new and more interesting questions. Quantum mechanics appealed to the young scientists the 1920s (to the point that it was called Knabenphysik — young man's physics) not because it gave them some kind of new, great understanding of everything in the universe, but because it promised a new "frontier" that they could conquer in order to achieve professional success. That's very exciting; certainly much more exciting than what classical physics was offering ("come up with slight improvements on aether theory"). But to take part in that enterprise, you needed to be very specifically "tooled up" with certain types of mathematics, physical ideas, etc., that were quite different than the educational requirements of the previous generation, and "re-tooling" for quantum physics, after a career as a classical physicist, was a non-trivial endeavor. So one sees, a bit, of the practical reasons (beyond any psychological ones) that "conversion" can be tricky, at least in this sort of field.