Friday, September 9, 2011

Kragh on scientific revolutions

The Danish historian Helge Kragh has a new book on failed physics theories, as discussed below. He is mainly known for writing Quantum Generations, a 1999 overview of 20th century physics. His account of relativity recites some common myths:
The famous 1887 Michelson Morley experiment was an attempt to measure the motion of the earth relative to the ether by means of an advanced interferometer technique. The experiment was performed at the Case School for Applied Science in Cleveland, Ohio, where Michelson was a professor of physics. ...

Lorentz’s first explanation of Michelson’s result was clearly ad hoc and not even based on his electrodynamic theory. During the following decade he greatly developed the theory, and in 1899 the Dutch theorist was able to derive the length contraction from the more general transformation formulas between the coordinates of a body moving through the ether and those of one at rest with regard to the ether. Lorentz wrote these transformations in a more complete form in 1904, the same form that we know today. He was not, however, the first to put the full "Lorentz transformations" in print. As a purely mathematical transformation, they can be found in a work on the Doppler effect published by Woldemar Voigt as early as 1887. More to the point, in 1900 Larmor derived the equations from his own version of electron theory. By means of the Lorentz Larmor transformations, the null result of the Michelson Morley experiment could be explained easily. Indeed, it followed from Lorentz’s theory that there could be no detectable effects of uniform motion through the ether, not just to the second order in v/c, but also to all orders.
That's right, Lorentz interpreted Michelson-Morley as showing that the speed of light is constant for all observers, and deduced his transformations from that. He later integrated electromagnetism into his relativity theory, and showed that the relativity principle held to all orders. By 1904, he had a relativity theory that could explain all of the known experiments. Others had similar formulas.

Kragh also credits Poincare:
No sketch of the prehistory of relativity, however brief, can avoid mentioning Henri Poincaré alongside Lorentz. Based on his conventionalist conception of science, around 1900 the French mathematician questioned whether the simultaneity of two events could be given any objective meaning. As early as 1898 he wrote, “Light has a constant speed.... This postulate can¬not be verified by experience, ... it furnishes a new rule for the definition of simultaneity” (Cao 1997, 64). Two years later, at the Paris world congress of physics, Poincaré discussed whether the ether really existed. Although he did not answer the question negatively, he was of the opinion that the ether was at most an abstract frame of reference that could not be given physical properties. In his Science and Hypothesis of 1902, Poincaré declared the question of the ether to be metaphysical, just a convenient hypothesis that some day would be discarded as useless. In his address to the St. Louis congress in 1904, he examined critically the idea of absolute motion, argued that Lorentz’s local time (t') was no more unreal than his general time (t), and formulated what he called the relativity principle, namely, the impossibility of detecting absolute, uniform motion. His formulation of 1904 is worth quoting: “According to the Principle of Relativity the laws of physical phenomena must be the same for a ‘fixed’ observer as for an observer who has a uniform motion of translation relative to him ... there must arise an entirely new kind of dynamics, which will be characterized above all by the rule, that no velocity can exceed the velocity of light” (Sopka and Moyer 1986, 293). Up to this point, Poincaré’s intervention in the discussion had been mainly programmatic and semiphilosophical. In the summer of 1905, without knowing about Einstein’s forthcoming paper, he developed an electrodynamic theory that in some respects went beyond Lorentz’s. For example, he proved the relativistic law of addition of velocities, which Lorentz had not done, and also gave the correct transformation formula for the charge density. Apart from restating the principle of relativity as “a general law of nature,” Poincaré modified Lorentz’s analysis and proved that the Lorentz transformations form a group with the important property that x2 + y2 + x2 - c2 t2 is invariant, that is, remains the same in any frame of reference. He even noticed that the invariant could be written in the symmetric way x2 + y2 + z2 + T2 if the imaginary time coordinate T = ict was introduced. Poincaré’s theory was an important improvement, a relativity theory indeed, but not the theory of relativity. Strangely, the French mathematician did not follow up on his important insights, nor did he show any interest in Einstein’s simultaneously developed theory of relativity.
Yes, this is correct up to the point were he says that Poincare's relativity theory was not the theory of relativity. It is Poincare's theory that was accepted by physicists, not Einstein's.

Kragh explains Einstein's version:

When twenty six year old Albert Einstein constructed the special theory of relativity in June 1905, he was unknown to the physics community. The paper he submitted to the Annalen der Physik was remarkable in several ways, quite apart from its later status as a work that revolutionized physics. For example, it did not include a single reference and thus obscured the sources of the theory, a question that has been scrutinized by later historians of science. Einstein was not well acquainted with the literature and came to his theory wholly independently. He knew about some of Poincaré’s nontechnical works and Lorentz’s work of 1895, but not about Lorentz’s (or Larmor’s) derivation of the transformation equations. Another puzzling fact about Einstein’s paper is that it did not mention the Michelson Morley experiment or, for that matter, other optical experiments that failed to detect an ether wind and that were routinely discussed in the literature concerning the electrodynamics of moving bodies. There is, however, convincing evidence not only that Einstein was aware of the Michelson Morley experiment at the time he wrote his paper, but also that the experiment was of no particular importance to him. He did not develop his theory in order to account for an experimental puzzle, but worked from much more general considerations of simplicity and symmetry. These were primarily related to his deep interest in Maxwell’s theory and his belief that there could be no difference in principle between the laws of mechanics and those governing electromagnetic phenomena. In Einstein’s route to relativity, thought experiments were more important than real experiments.
Yes, Einstein did not cite his sources or explain how his theory was any better than Lorentz's. Lorentz was one of the most famous theoretical physicists in Europe, and some say that it is very likely that Einstein had access to Lorentz's papers and lied about it. Even if Einstein did not have all the papers in 1905, he certainly did later, and there can be no excuse for his failure to acknowledge previous work throughout his life.

Einstein was able to ignore the experimental evidence because he was just giving an exposition of Lorentz's theory. He trusted Lorentz to have interpreted the experiments correctly. No one thought that Einstein's paper was revolutionary because it was seen as the same as Lorentz's theory.
Most unusually at the time, the first and crucial part of Einstein’s paper was kinematic, not dynamic.
No, that was not unusual. That is the same thing that Lorentz did 13 years earlier, as explained above. The first part was based on the speed of light being constant for all observers, and the second part reconciles the transformations with the equations for electrodynamics.
Contrary to those of Lorentz and Poincaré, Einstein’s formulas related to real, physically measurable space and time coordinates. One system was as real as the other.
Other historians say this, but it is nonsense. Lorentz and Poincare were directly and explicitly concerned with explaining the Michelson-Morley and other experiments. Those explanations do not make any sense unless they concern real physical measurements. Of course Lorentz's and Poincare's formulas used real space and time coordinates.

Those who credit Einstein need to say that he was original and that he revolutionized physics with pure thought. They have the problem that others had all the formulas years beforehand. So they have to somehow discount the meaning of those formulas, and deny that Einstein relied on experiments. The result is a story that does not make any sense.
Einstein’s theory was taken up and discussed fairly quickly, especially in Germany. Its true nature was not recognized immediately, however, and it was often assumed to be an improved version of Lorentz’s electron theory. The name “Lorentz Einstein theory” was commonly used and can be found in the literature as late as the 1920s. The most important of the early relativity advocates was Max Planck, ... Another important advocate was the Göttingen mathematician Hermann Minkowski who, in a 1907 lecture, presented relativity theory in a four dimensional geometrical framework with a strong metaphysical appeal. Minkowski introduced the notion of a particle’s world-line and explained enthusiastically how radical a break with the past the theory of relativity was: “Henceforth space by itself, and time by itself, are doomed to fade away into mere shadows, and only a kind of union of the two will preserve an independent reality” (Galison 1979, 97). However, Minkowski considered Einstein’s theory to be a completion of Lorentz’s and interpreted it, wrongly, to be within the framework of the electromagnetic worldview.
Yes, Minkowski and everyone else considered Einstein's theory to be an improved version of Lorentz’s electron theory. They were correct.

That four dimensional geometrical framework for relativity was what made the theory take off. Minkowski's paper was very widely circulated in 1908, and it was based on the work of Lorentz and Poincare. Einstein's paper was irrelevant to Minkowski and nearly everyone else. It only influenced Planck, as far as I can tell.

As you can see on this chart from Kragh's book, relativity papers took off in 1908, not 1905.
The picture of Einstein as an all knowing and somewhat arrogant rationalist who did not care about experiments is undoubtedly widespread and part of the Einstein myth. But it is basically wrong, at least as far as the younger Einstein is concerned.
Kragh has helped promote this myth by denying that relativity theory is the result of experiments.
Einstein’s theory of relativity shared with Darwin’s evolutionary biology, Röntgen’s invisible rays, and Freud’s psychoanalysis the fact that it was met with an enormous interest outside as well as within the scientific community. It became one of the symbols of the modernism of the interwar period and, as such its importance extended far beyond physics. Einstein’s theory was labeled “revolutionary,” a term commonly associated with the passage from Newtonian to Einsteinian physics. The theory of relativity was indeed a kind of conceptual revolution and in the early 1920s the revolution metaphor, freely associating to political revolutions, became a trademark of Einstein’s theory. It was a trademark that Einstein did not want to sanction. Einstein did not consider himself a revolutionary; in papers and addresses, he repeatedly stressed the evolutionary nature of the development of science. The theory of relativity, he often said, was the natural outcome of the foundations of physics laid by Newton and Maxwell. Thus, in a 1921 paper, Einstein noted, “There is someting [sic] attractive in presenting the evolution of a sequence of ideas in as brief a form as possible, and yet with a completeness sufficient to preserve throughout the continuity of development. We shall endeavour to do this for the theory of relativity and to show that the whole ascent is composed of small, almost self evident steps of thought” (Hentschel 1990, 107).
No, Einstein lied about his sources throughout his life, and always describe his relativity work as a revolutionary flash of genius that has no need for the work of Lorentz and Poincare.
. In his best selling The Structure of Scientific Revolutions, Kuhn argued against the positivistic view of science, suggested that there is no such thing as scientific progress across periods of revolutionary change, and hinted that science develops in a nonrational manner. The message, it seemed to many of his young readers, was that physics was no more scientific than psychology, art history, or literary criticism. Nor was modern astronomy to be believed any more than astrology. ...

The works of Kuhn and Feyerabend formed the background of other historical, philosophical, and sociological studies of science, the latest fashion being known as the program of sociology of scientific knowledge or social constructivism. Contructivist [sic] sociologists of the 1980s and 1990s denied that the scientific world view is grounded in nature and should therefore be given higher priority than any other worldview. Science, they said, is basically a social and cultural construction fabricated by negotiations, political decisions, rhetorical tricks, and social power. Since truth and falsehood are always relative to a given local framework, scientists’ beliefs about nature are not inherently superior to those of any other group.
Kragh ends up concluding:
The great changes that have occurred in twentieth century physics have built on existing knowledge and a healthy respect for traditions. There have been several attempts to base physics on an entirely new worldview (such as those proposed by Eddington and Milne in the 1930s), but they have all failed. It may seem strange that respect for traditions can produce revolutionary changes, but this is just what Thomas Kuhn described in 1962 under the label “normal science.” On the other hand, the changes that sometimes follow paradigm ruled or “normal” science are not revolutions in the strong sense that Kuhn suggested in 1962, namely, new paradigms incompatible with and totally different from the old ones. No such revolution has occurred in twentieth century physics. After all, a theoretical physicist of the 1990s will have no trouble in understanding the spirit and details of Planck's work of 1900 in which the quantum discontinuity was introduced, nor will a modern experimentalist fail to appreciate J. J. Thomson's classical paper of 1897 in which the electron was announced. There is no insurmountable gap of communication, no deep incommensurability, between the physics of the 1990s and that of a century earlier.

The lesson to be extracted from the latest century of physics is that physical knowledge has greatly expanded and resulted in new and much improved theories, but that these have been produced largely cumulatively and without a complete break with the past. It has always been important to be able to reproduce the successes of the old theories, and this sensible requirement guarantees a certain continuity in theoretical progress. The great discoveries and theories of our century have not, of course, left earlier knowledge intact, but neither have they turned it wholesale into non knowledge. Most experimental facts continue to be facts even in the light of the new theories. ...
I agree with this. Kuhn's ideas are widely accepted, but wrong.

My quarrel with Kragh is that he seems to concede that Kuhn's thesis was correct insofar as Einstein's relativity was revolutionary break from previous theories. If Einstein's 1905 paper were really so original, so radical, so independent of experiment, and so influential, then there might be some merit to Kuhn's view of revolutionary science. But all of that is completely wrong, as you can see by parsing Kragh's own facts.

I think that there is an Einstein reality distortion field. People idolize him even when it goes against their own facts and theories.

All of this is explained in my book, altho I did not know of Kragh's work when I wrote it. Kragh confirms everything I say, and I would have mentioned him if I had known about him.

No comments:

Post a Comment