The Many Worlds of Hugh Everett III: Multiple Universes, Mutual Assured Destruction, and the Meltdown of a Nuclear Family

by Peter Byrne

  John Bell, 1981.1

  The rise of mentalism

  A century has passed since Max Planck and Albert Einstein discovered the quantum world. The basic quantum paradoxes—uncertainty, non-locality, and the measurement problem—are either unsolved or remain highly contentious. The most influential proposals for adjudicating the measurement paradox are the Copenhagen Interpretation (which declares the problem not a problem), Bohm’s non-collapsing hidden variables theory (which supplements the Schrödinger equation), the “GRW” spontaneous collapse theory2 (which modifies the Schrödinger equation), and the Everett interpretation, which makes do with the Schrödinger equation alone.

  Apart from the Copenhagen interpretation, these theories are “realist,” i.e. they assume that the wave function is physically real, that the human mind is a natural object, and that consciousness does not play any role in shaping a reality that exists independently of human agency.3 It is an interesting phenomenon that, faced with the measurement problem, resort is sometimes made to theories of consciousness, as if the mind is an all-purpose repository of the inexplicable. For instance, in 1988, even as Everett’s fundamentally realist theory was gathering steam in scientific and philosophical circles, an attempt was made to transform it into a theory called “many minds” by philosophers David Albert and Barry Loewer, (Albert later disavowed this model).4

  The Albert-Loewer interpretation set out to keep the Schrödinger equation intact and to solve the preferred basis and probability problems in Everett’s formulation. It postulated that a continuous infinity of non-physical minds exist and operate—not according to a physical dynamics, not according to quantum mechanics, but according to mental dynamics that exist apart from our physical reality (there are no superpositions in this model). In this scheme, subjective beliefs about nature are considered as physically determinate in nature, i.e. every physical outcome in a probabilistic universe is correlated to a single mental state drawn from the vast array of independently evolving minds that collectively define the identity of the observer.5 Everett, on the other hand, deliberately replaced human minds with record-keeping machines in his theory, obviating the need for a special theory of consciousness or a mentalist approach. There are, unremarkably, many minds or rather many brains in Everett’s theory because each splitting observer-brain embarks on its own future. But after branching, these purely physical minds are no longer linked, nor part of an over-arching Mind.

  It is worth recalling what Everett wrote in 1980 about Wheeler:

  [He was] wondering if somehow human consciousness was a distinguished process and played some sort of critical role in the laws of physics.

  I, of course, do not believe any such special processes are necessary, and that my formulation is satisfactory in all respects.6

  Inserting non-physical processes into Everett’s theory is antithetical to his intention and what it says.

  Bohm again

  In 1957, Everett tried to find an address for the ex-patrioted physicist, David Bohm.7 Perhaps he wished to commiserate with him about Bohr’s resistance to new ideas. There is no evidence that the two mavericks ever communicated, but Bohm was fascinated by Everett’s theory. He died in London shortly before his book on quantum ontology (with Basil Hiley), The Undivided Universe, was published in 1993. In it, Bohm argued that the wave function does not collapse, and that human consciousness is not a causal force in physics. He considered the many worlds theory as vital and worthy of

  special attention because it has come to be fairly widely accepted, especially by those who work in general relativity and cosmology and who therefore feel a need to regard the universe as existing in itself whether observed or not.8

  But Bohm did not understand Everett. He misinterpreted Everett’s use of memory records to register splitting events as a statement that only consciousness branches. Bohm demarcated what he called Everett’s “speculative theory of mind” from DeWitt’s account of many worlds as a physical theory. Specifically, he stated that Everett’s theory did not support DeWitt’s statement that observers and universes physically split, or branch.9

  Regarding Dewitt’s “version” of Everett, Bohm criticized it for assuming, not deriving the Born rule, and requiring, he thought, additions to the quantum formalism. More to the point:

  It defies the imagination to grasp intuitively how the universe could split, and even more as to how it could be doing so in a stupendous number of ways.10

  But Everett and DeWitt were in accord on the existence of splitting (as we have seen in previous chapters). It is a red herring (and an all-too-common mistake) to divide Everett’s view from DeWitt’s view in regard to splitting: They both asserted the “reality” of branching universes. DeWitt and Everett differed regarding the method of deriving a probability measure, but they agreed that the branches were “equally real.” This agreement is substantiated by the published and unpublished writings of both men, including Everett’s “split” footnote, rebelliously inserted into the published thesis of 1957. Nonetheless, Bohm and Hiley insisted that

  Everett did not contemplate the splitting of the universe…. Indeed Everett’s view should not even be called the many-worlds interpretation but rather, as Albert and Loewer have suggested, the many minds interpretation.11

  This error was amplified by philosopher David Chalmers in The Conscious Mind (1996). Chalmers was overtly promoting mind-body dualism, i.e. the hypothesis that human consciousness is not subject to physical laws. And he looked to Everett in support of his non-Everettian notion. Attracted by Everett’s argument that the Schrödinger equation does not break down in the face of measurement, Chalmers echoed Bohm-Hiley by insisting that Everett had constructed an Albert-Loewer type of many minds theory. For Chalmers, Everett’s splitting is “only in the minds of the observers.”12

  Other philosophers of physics have construed Everett as a many-minder, but those arguments consistently fail in the light of his long thesis, and the basement papers that speak to his intention. Philosophers and physicists may create Everett-type many minds theories, but it is not correct to attribute them to Everett.

  It was up to an iconoclastic physicist allied with a small group of philosophers at the University of Oxford to protect Everett’s materialism.

  The Oxford connection

  David Deutsch talked to Everett only that one time at lunch in Austin in 1977, but he has devoted much of his career to making a case that quantum mechanics itself is the best proof of the existence of multiple universes, or the multiverse, as he calls the set of all physically possible worlds. Born in 1953, Deutsch is slightly built, pale of pallor, long-haired, and somewhat reclusive. A night owl, he seldom ventures out of the dreamy confines of Oxford town. In the early 1980s he introduced the idea of a universal quantum computer.13 And in the early 1990s, he co-invented the Deutsch–Jozsa algorithm, demonstrating the potentially amazing speed and parallel processing capacity of a quantum computer. The insight that is possible to use superposed microscopic systems as “qubits” running computations exponentially faster than a classical processor was the shot that set off a billion-dollar research race.14 He is a Visiting Professor of Physics at Oxford. And in 2008, he was elected a fellow of the prestigious Royal Society, which has 60 Nobel laureates on its roster.

  Deutsch recounts that DeWitt convinced him of the truth of many worlds. But, DeWitt used to tell him there was a terrible problem with probability in the otherwise great theory: “And I would understand the problem only for as long as it took me to walk out of DeWitt’s office. After several years, I finally got it and realized it was important and started working on it.”15 In 1984, Deutsch proposed a method of considering quantum mechanics as universal and the Everett worlds as “continuously infinite.”16 But this particular method added to the equations of quantum mechanics, defeating the point of the Everett interpretation, and he later turned to another method of introducing a measure of probability in the multiverse.17 Nonetheless, Oxford philosopher Michael Lockwo
od used Deutsch’s method to construct a many minds interpretation of Everett:

  Instead of postulating a continuous infinity of worlds, we could credit every sentient being with a continuous infinity of simultaneous minds or conscious points of view, which differentiate over time.18

  An intense debate over Lockwood’s proposal ensued in the pages of the British Journal of Philosophy in 1996. Deutsch wrote an essay congratulating the philosopher on taking physics seriously, but he noted,

  Lockwood’s preference for the term ‘many minds’ over ‘parallel universes’ risks giving the impression that it is only minds that are multiple, and not the rest of reality…. Admittedly, [we can] detect other universes only indirectly. But then, we can detect pterodactyls and quarks only indirectly too. The evidence that other universes exist is at least as strong as the evidence for pterodactyls or quarks.19

  Deutsch expressed amazement that physicists and philosophers are reluctant to consider the Schrödinger equation as universally valid:

  Despite the unrivaled empirical success of quantum theory, the very suggestion that it may be literally true as a description of nature is still greeted with cynicism, incomprehension and even anger.20

  In the same journal issue, Oxford philosophy professor Simon Sanders said it was a mistake to focus physics debates on “mentality and the nature of consciousness.” Saunders, a realist, is a pivotal player in many worlds research. Intellectually, he is a hybrid, typical of philosophers of science: he is learned in both the higher mathematics of physics and the syllogisms of philosophy. And he is a bit of a rabble-rouser inside his rarified circle of colleagues.

  After a year as an undergraduate in the early 1970s in the newly founded Physics and Philosophy degree school at Oxford, Saunders eschewed the socially elite, all-male “champagne glass twirling” atmosphere of his ancient, walled-in college to live communally on a farm while puzzling out the quantum paradoxes and earning his degree, and later his doctorate. By 1990, he was working as a professor of philosopher at Harvard University in Cambridge, Massachusetts, where he met his no-nonsense wife, Kalypso Nicolaidis, an expert in international relations. Kalypso recalls that Saunders talked about multiple universes on their first date. So, she married him.

  Saunders returned to Oxford in the mid 1990s as a philosophy professor. Soft-spoken and unassuming, the professor is the polar opposite of Everett’s persona. But for two decades, he has been grappling with the problem of defining probability in many worlds theory, picking up where the original theorist abruptly left off. Saunders was one of the first philosophers of science to recognize the importance of decoherence to the many worlds interpretation. He has also developed distinct parallels between Everett’s relative state idea and Einstein’s theory of relativity according to which the difference between the past, present and future is purely relational. Just as it seems that there is no “flow of time” in Einstein’s relativistic view of the whole of space and time, so there is no “transition from the possible to the actual” in Everett’s wave function of the universe. Put simply, “relational fact” theory speaks to the problem of identity by asserting that one can locate oneself as an object inside the multiverse of all actualities by identifying the fact or sets of facts to which the self-object is related, e.g. I am that Simon that exists in a universe in which London has a population of 7,556,900 at noon sharp on Tuesday, not the other Simon where the population is 7,556,901!

  In the late 1990s, Deutsch dropped a theoretical bombshell that astonished Saunders: he proposed that the Born rule can be derived in quantum mechanics from considerations of pure rationality.21 In a move that would have delighted Everett, Deutsch turned to decision theory—the modern incarnation of game theory—to define probability in the multiverse. Deutsch recalls

  I eventually got around to publishing a decision-theoretic version of probability, and there was a welter of criticism, almost entirely missing the point, from people objecting to the Everett interpretation itself, rather than to this particular derivation of probability. They got rather hot under the collar about it.22

  His technical argument concluded that rational decisions are expected to satisfy personal preferences with a certain probability. Taunting one-worlders, Deutsch argued that people are rationally compelled to make quantum mechanical decisions in a branching universe as if the Born rule applied. This addressed one of the main objections to the many worlds interpretation: that if everything physically possible happens, then assessing outcomes probabilistically is not necessarily a reasonable criterion for deciding what to do. He concluded:

  Thus we see that quantum theory permits what philosophy would hitherto have regarded as a formal impossibility, akin to “deriving an ought from an is,” namely deriving a probability statement from a factual statement. This could be called deriving a “tends to” from a “does.”23

  Savage axioms

  As the millennium turned, a graduate student in physics at Oxford, David Wallace, improved the rigor of Deutsch’s argument, knocking down some of the objections to it.24 Saunders quickly endorsed the newly christened “Deutsch-Wallace” theorem’s consideration of the “practical” problem in a theory of branching universes: How should we act as to realize our goals when every outcome actually occurs? According to Everett’s basic theory, the quantum world appears to be indeterminist—even though the multiverse is objectively determinist because everything occurs in some measure.

  Wallace observes,

  The co-existence of determinism with the in-principle-unknowability of the future is from a philosophical point of view perhaps the Everett interpretation’s most intriguing feature.25

  The name of the game is to derive probabilistic judgments from constraints of rationality without postulating a probability rule or changing the essential formalism of quantum mechanics in which the Schrödinger equation evolves deterministically. This research is guided by the statement: “Whatever (objective) probability is, our empirical access to it is via considerations of rationality and behavior: it must be the case that it is rational to use probability as a guide to action.”26

  In defining rationality, Wallace referred to a set of logical axioms developed in 1954 by Leonard J. Savage.27 Savage noted:

  It is one of my fundamental tenets that any satisfactory account of probability must deal with the problem of action in the face of uncertainty.28 … From the personalistic point of view, statistics proper can perhaps be defined as the art of dealing with vagueness and with interpersonal difference in decision situations.29

  Interpreting quantum mechanics with decision theory is not as far-fetched as it may sound. Wallace constructed (and continues to refine) a technical proof that it is rational to use the Born rule (squaring the wave function of a quantum event to determine its classical probability) to make decisions in a multiverse—without postulating that rule.

  One weakness in the original Deutsch-Wallace approach (acknowledged as such by its proponents) was that it required making a different kind of assumption: that the Everett interpretation is true. If one does not believe in branching universes, then the decision-theoretic argument loses meaning because it is an argument that one should accept the Born rule if one believes in branching universes. This limitation was notably addressed by Hilary Greaves, a fellow in philosophy at Oxford’s Summerville College. Despite the fact that everything happens, the multiverse theory is still accountable to empirical test on the basis of measuring relative frequencies, says Greaves. Rigorously applying Savage’s axioms, which are independent of quantum mechanics, she argued that it is rational to use the Born rule to make personal decisions in both single universe and multiverse models. And her argument did not require believing that the Everett worlds exist.30

  In sum, the Oxford philosophers want to “operationalise” probability in quantum mechanics by reducing it—not to relative frequencies—but to quantifiable “preferences of rational agents.”31 For Wallace, “Rationality is the only way in which the concept of ‘pro
bability’ makes contact with the physical world.”32 And a rational being (or machine-driven algorithm) ranks preferences among decision options according to principles of utility, i.e. usefulness to the ranker.33 Utilities are here treated as values—prices—placed upon optimizing certain goals. Rational people compare the value they have assigned to the achievement of a goal with its perceived probability of occurrence before they choose to act. But in the Everettian multiverse, of course, this method is questionable: how can rational decisions be based upon relative frequencies when there appears to be no such thing as probability?34

  Yet, probabilistic quantum mechanics works.

  Savage’s basic tenet is that probability is personal. So, say the Oxford Everettians, in a multiverse, it is functionally logical to make decisions based on personal preferences and beliefs about possible outcomes in one branch even though “you” do not know which post-decision branch “you” will end up on.35 The Born rule serves to quantify this uncertainty, and becomes a rational guide for decision-making even when all consequences appear to be possible. The argument gives Everettians permission to keep using probability in quantum mechanics, making a case that it is rational to accept the Everett interpretation.36