Did Einstein predict Bose-Einstein condensation?
Introduction
In the mid-90s, physicists at JILA1 and MIT created, for the first time, a fascinating phenomenon: Bose-Einstein condensation (BEC)2 in ultracold atomic gases. Since then, many laboratories have been built around the world in which BECs are created and further investigated. In these experiments, an atomic gas is produced within a vacuum chamber, laser light slows down the atoms (strange but true), they are trapped and cooled further, until they have a temperature of only a few hundred nanokelvin. If the atoms are bosons, and the clouds are very cold, they can undergo a phase transition and become a Bose-Einstein condensate. A BEC is a state of matter, like “liquid” or “solid”. But in contrast to those states, it can only be understood within quantum theory. So, what is a Bose-Einstein condensate? Here are a few of its features. In the Bose-Einstein state, all atoms of the condensate are in the same quantum state, the lowest state of the magnetic trap that confines them.3 The wavefunctions of the single atoms are so large that they overlap. The phases of the waves are aligned, so that the whole cloud behaves like one big wave. And “big” really means big: even though the condensate populates the quantum realm, its size is a few micrometers in diameter. You can even “see” the macroscopic wavefunction if you let two BECs overlap: the atomic clouds create interference fringes, just like light in a double-slit experiment!
Because of this deeply quantum mechanical behaviour of a macroscopic system, many physicists are enthralled by Bose-Einstein condensation. The first successful realisations of Bose-Einstein condensation in atomic gases were rewarded with Nobel Prizes in 2001, for Eric A. Cornell, Wolfgang Ketterle, and Carl E. Wieman. And these Nobel Prize winners are not the only famous people that appear in the story of Bose-Einstein condensation: it is often claimed that this phenomenon was predicted a long time ago, already in 1925. And that this prediction was made by none other than Albert Einstein himself. Here is the quote from the 2001 Nobel Prize committee:
Einstein predicted that if a gas of such atoms [a “certain type of atoms”] were cooled to a very low temperature all the atoms would suddenly gather in the lowest possible energy state. The process is similar to when drops of liquid form from a gas, hence the term condensation. Seventy years were to pass before this year's Nobel Laureates, in 1995, succeeded in achieving this extreme state of matter.4
Similar claims – Einstein predicted something that was confirmed decades later – can be found in other sources as well, for example in scientific articles (like Anderson 1995), encyclopedias (like the Britannica5), and introductory books on BECs (like Pitaevskii & Stringari, 2016) This seems incredible: Einstein made the prediction in 1925, and it took 70 years till it was confirmed! This narrative suggests that quantum theory in 1925 already contained all the basic information about Bose-Einstein condensation, and the experiments in the 1990s merely showed that information is true.
This narrative about a specific prediction is linked to a more overarching view: predictions are of central importance to science. This view has been around for a long time, but in the 20th century, it was famously expressed in Karl Popper's influential work on falsificationism. Popper set out to explain the difference between pseudo-science (for example, Marxism and Freud's psychoanalysis, in his view) and real science. He claimed that it is the method that distinguishes the two: In science, theories are used to make predictions, and these predictions are then tested. If a prediction fails the test, the theory is rejected. That is not the case for pseudo-science:
Pseudoscientific theories attempt to accommodate all known observations by ad hoc manoeuvers, to avoid falsification at any cost. Scientific theories ‘stick their necks out’ by making bold predictions, which are then honestly checked against experience (Popper, 1972).
In this view, a business that does not make and test predictions is not science! An example of a scientific episode that Popper was particularly impressed by was the 1919 eclipse test of Einstein's theory of relativity. According to the theory, heavy objects influence the path of light. During a solar eclipse, it should be possible to observe that the light coming from a star behind the sun would be deflected by the sun. During the 1919 solar eclipse, measurements were done by Arthur Eddington and his team, and they confirmed that the position of the star was the one predicted by general relativity. To Popper, this was a beautiful example of how science works. Within philosophy of science, Popper's claims have been discussed extensively, and the modern consensus is that the situation is a lot more complicated than falsificationism suggests. However, there is still a widespread belief that predictions play a central role within science.6 And Popper's view – or a popularized version of it – is still well-known, also among scientists. The story of Bose-Einstein condensation seems to speak for it. Look, we might say, Einstein “stuck his neck out” and made a bold prediction, which could have been falsified. But it wasn't: it took a long time, but eventually, his prediction was confirmed.
In this article, I will assess whether this really is a story of a prediction that was confirmed decades later. To that end, I will analyse the history of Bose-Einstein condensation. This history contains several surprising turns. It begins 1924 with theoretical work on light, switches to an analysis of atomic gases, then turns to superfluid helium, and back to atoms. In the 1970s, experimentalists enter the story, first with work on hydrogen, and then on alkali gases. In 1995, finally, BECs in atomic gases are created. After this historical part, I will discuss whether this is the story of a confirmed prediction. I will argue that Einstein did not say what you need to do to create a BEC with an atomic gas, he did not say what evidence for a BEC would look like, and his concept of “condensation” was significantly different from our modern concept of Bose-Einstein condensation. I will judge that at least, Einstein's claims were not a prediction in the sense that goes back to Popper. Since this is the kind of scientific prediction that researchers in both philosophy and physics tend to have in mind, I find the narrative of Einstein having predicted BEC misleading. However, I will point out how the prediction narrative can be defended in a weaker sense. I will suggest that it would be worthwhile for philosophers to investigate further into the notion of prediction and distinguish different kinds and aspects of predictions, and the roles they play in scientific research.
Section snippets
The history of Bose-Einstein condensation
The story begins with Satyendra Nath Bose, a physicist at the University of Dacca in India (now Bangladesh). He wrote a 4-page long article in 1924 (Bose, 1924) which was so influential that today, we still call one kind of particle after its author: bosons. Even though here is where the story of Bose-Einstein condensation starts, the article is not about atoms at all, but about a statistical treatment of light.7
The prediction narrative
Having presented the key stages of the history of Bose-Einstein condensation, I will now discuss the question of whether this is the story of a confirmed prediction. Here is the claim I will assess:
Einstein made the novel prediction: “if you do X, atoms will form a condensate”. 70 years later, X was done, and atoms formed a condensate.
I will now analyse different parts of this claim: the conditional “if you do X”, the question of how we can tell whether there is a condensate, and the question
Lessons from the case study
I argued above that Einstein did not envision what you need to do to create a BEC, he did not say how to recognize a BEC, and what he means by the term “condensate” is quite different from the modern notion of a BEC. Therefore, I find the prediction narrative unconvincing. Rather than making a successful prediction, Einstein made a claim about an idealized system, which was certainly very insightful and extremely influential. He introduced the concept of “condensation”, which was scrutinized,
Conclusion
I investigated the often-repeated claim: Einstein predicted Bose-Einstein condensation, and his prediction was confirmed in 1995. I discussed historical work on the phenomenon and its realization, starting with the work of Bose in 1924 and ending with the 1995 experiments at JILA and MIT. I argued that there are three reasons to think that Einstein did not predict the condensation phenomenon produced in these experiments: he did not say what you need to do to create a BEC with an atomic gas, he
Funding
I acknowledge funding from the Cambridge Trust and the Cambridge AHRC DTP.
Acknowledgments
For very helpful feedback and discussions, I thank Hasok Chang, Claus Zimmermann, and the participants of Hasok Chang's Wednesday reading group. I am indebted to two very insightful anonymous reviewers. Furthermore, I am particularly grateful to David Pritchard and Wolfgang Ketterle for their helpful feedback and discussions. I acknowledge funding from the Cambridge Trust and the Cambridge AHRC DTP.
References (46)
- et al.
Observation of Bose-Einstein condensation in a dilute atomic vapour
Science
(1995) - et al.
Observation of interference between two Bose condensates
Science
(1997) Representing electrons
(2006)The making of modern physics in colonial India
(2020)Who discovered the Bose-Einstein statistics?
From the necessary to the possible: The genesis of the spin-statistics theorem
The European Physical Journal H
(2014)- et al.
Gravitational waves and the long relativity revolution
Nature Astronomy
(2018) Plancks Gesetz und Lichtquantenhypothese.“ Zeitschrift für Physik 26, 178-181. Repr
Dynamics of theory change: The role of prediction
PSA
(1994)- et al.
Experiments in dilute atomic Bose-Einstein condensation