Arxiv:0905.1283V1 [Physics.Pop-Ph] 8 May 2009 Rae Diversity

Home , Max Tegmark

arXiv:0905.1283v1 [physics.pop-ph] 8 May 2009 rae diversity. progressivel allowing greater 1) (Figure in- four-leve multiverses natural theories of a hierarchy physics form survey which I universes, parallel article, volving this In universe. our also universe, observable our our called defines radius this of sphere h i-agepninbgn h otdsatvisi- distant 4 most about since now The years are billion objects 14 began. ble the expansion during big-bang travel the to able been has have that distinguish universes also to parallel are proposed. forget of They been many types since different speculation. confusion, purview between silly of the is source merely within it a are or and they popular, science, whether are of ask they However, one.” to as controversial this important as not are but in — they exam your universes passed “You parallel many jokes: even and movies books, n hte hsi cec rpiooh.Frnw the now, For philosophy. or science is this whether and levels many how has. rather it but cosmo- uncontroversial model), mul- rather concordance a the logical is is there I whether Level not (since tiverse therefore is question key The 1 n r o bu bu 0blinlgtyasaway. years light billion expansio 40 cosmic about the of about dis now because most receded are the have and us, see reaching can now we is things that light the emitting After h atetyucnosrei h itneta light that distance the is observe can you farthest The in up cropping rage, the all now are universes Parallel eo ewl ics tlnt h su fevidence of issue the length at discuss will we Below • • • • Dtd eray2 05 ulse in published 2005, 20 February (Dated: in naTsit ieetrgoso pc can space of regions different different T-shirt, equa- exhibit with a capture on to hope tions day one physicists that ee II: Level away. nldn nietclcp fyuaot10 about you of — copy conditions identical initial an contains all including which realizing space, volumes “ergodic” infinite Hubble an is tion I: Level a osat,dmninlt,pril content, particle dimensionality, constants, cal eetfnaetleutoso hsc o htT- that for shirt. physics of equations fundamental ferent has IV: controversial. level Level this most that the given been ironic historically is qualita- which nothing new, add tively wavefunction the of branches III: land- Level a in possibilities. minima of scape local different to corresponding et fPyis&MTKviIsiue ascuet Insti Massachusetts Institute, Kavli MIT & Physics of Dept. ubevolume Hubble eei rdcino omlgclinfla- cosmological of prediction generic A ie the Given te ahmtclsrcue iedif- give structures mathematical Other nuiayqatmmcais other mechanics, quantum unitary In effective our , fundamental × 10 oio volume horizon aso hsc (physi- physics of laws 26 eesaway meters h utvreHierarchy Multiverse The aso physics of laws Uies rMultiverse?” or “Universe rsimply or 1 n a and , 10 a Tegmark Max etc. 29 tant m n, y ) l l t rdcin,mrl tlatoeo hm Consider them. test of and one observe analogy: least following to at the able merely be predictions, not its need all we theories falsifiable, certain be of to prediction a but that theory, is remember to point key esul rdce aytig htwe that things suc- many has predicted Relativity General cessfully of theory Einstein’s Because ie ucsflpeitoso h hoiso cosmologi- of unitary theories and Like- the inflation cal of horizon. misconcep- predictions the early at successful right to wise, happens (contrary funny nothing that tions) can- and we horizons things event for predictions observe, its not seriously take also we oa odtosfrteuiesllw htsol oon go should that laws observable T-shirt. universal entire that the our ou for misinterpreting fills conditions into local in it us space that tricking of potentially de- so universe, patch the phase small has particular a could a stretching inflation but also of fish, cosmological is property than ceptive smarter there fooled: be that similarly may We realizing be not steam. and universal, ice properties are the water that conclude of mistakenly (parti- might physics ocean of the laws effective o symmetries, liquid phases, cles, solid, different be In can water gas. like just can possibilities land- a of space in scape that phases suggest different but between also transitioning theory uni- “freeze”, void, unified expanding static a (gravity) (the for curve boring Searches and stretch waves) a can (gravitational vibrate merely that verse), entity not dynamic taught is Einstein a space us. that tricking in- in be have us head may insights its Nature theoretical that with recent dicated van- ostrich Moreover, the our sand. of from the reminiscent them is observe point cannot things tage dismissing we and because imagina- world, physical merely of the underestimate lack of to vastness humans Hübris and us the caused repeatedly details. have tion the into delving predic- universes. other parallel of their types seriously various including more tions, take scientists some 2 ntr unu mechanics quantum Unitary Inflation Relativity General u ehnc s“ntr” akn h otoesa proc controversial the collapse. lacking wavefunction “unitary”, as known is of mechanics version simplest tum mathematically the below, described As e scnld ihtocuinr eak before remarks cautionary two with conclude us Let .Cr d,CmrdeUiest rs,2007)) Press, University Cambridge ed., Carr B. , ueo ehooy abig,M 02139 MA Cambridge, Technology, of tute e.g. htsaecniusisd lc hole black inside continues space that , etc. )cuddffr s ee leaving never fish A differ. could .) 2 unu ehnc aemade have mechanics quantum ee I aalluniverses parallel III Level universes parallel I Level interiors hole Black aalluiessaenta not are universes parallel o theory a For . can observe, quan- ess r r . 2 3

I. LEVEL I: REGIONS BEYOND OUR COSMIC HORIZON

Let us return to your distant twin. If space is infinite and the distribution of matter is sufficiently uniform on large scales, then even the most unlikely events must take place somewhere. In particular, there are infinitely many other inhabited planets, including not just one but infinitely many with people with the same appearance, name and memories as you. Indeed, there are infinitely many other regions the size of our observable universe, where every possible cosmic history is played out. This is the Level I multiverse. FIG. 2: Although an infinite universe has always been a possibility, the lower limit on the size of our universe has kept growing. A. Evidence for Level I parallel universes

Although the implications may seem crazy and them. Infinite models fit the data, and strong limits have counter-intuitive, this spatially infinite cosmological been placed on the alternatives (de Oliveira-Costa et al. model is in fact the simplest and most popular one on 2003; Cornish et al. 2003). In addition, a spatially infi- the market today. It is part of the cosmological concor- nite universe is a generic prediction of the cosmological dance model, which agrees with all current observational theory of inflation (Garriga & Vilenkin 2001b), so the evidence and is used as the basis for most calculations striking successes of inflation listed below therefore lend and simulations presented at cosmology conferences. In further support to the idea that space is after all simple contrast, alternatives such as a fractal universe, a closed and infinite just as we learned in school. universe and a multiply connected universe have been se- Another loophole is that space is infinite but matter is riously challenged by observations. Yet the Level I mul- confined to a finite region around us–the historically pop- tiverse idea has been controversial (indeed, an assertion ular ”island universe” model. In a variant on this model, along these lines was one of the heresies for which the Vat- matter thins out on large scales in a fractal pattern. In ican had Giordano Bruno burned at the stake in 16003), both cases, almost all universes in the Level I multiverse so let us review the status of the two assumptions (infi- would be empty and dead. But recent observations of the nite space and “sufficiently uniform” distribution). three-dimensional galaxy distribution and the microwave How large is space? Observationally, the lower bound background have shown that the arrangement of matter has grown dramatically (Figure 2) with no indication of gives way to dull uniformity on large scales, with no co- an upper bound. We all accept the existence of things herent structures larger than about 1024 meters. Assum- that we cannot see but could see if we moved or waited, ing that this pattern continues, space beyond our observ- like ships beyond the horizon. Objects beyond cosmic able universe teems with galaxies, stars and planets. horizon have similar status, since the observable universe grows by a light-year every year as light from further away has time to reach us4. If anything, the Level I mul- B. What are Level I parallel universes like? tiverse sounds trivially obvious. How could space not be infinite? Is there a sign somewhere saying ”Space Ends The physics description of the world is traditionally Here–Mind the Gap”? If so, what lies beyond it? In split into two parts: initial conditions and laws of physics fact, Einstein’s theory of gravity calls this intuition into specifying how the initial conditions evolve. Observers question. Space could be finite if it has a convex curva- living in parallel universes at Level I observe the exact ture or an unusual topology (that is, interconnectedness). same laws of physics as we do, but with different initial A spherical, doughnut-shaped or pretzel-shaped universe conditions than those in our Hubble volume. The cur- would have a limited volume and no edges. The cosmic rently favored theory is that the initial conditions (the microwave background radiation allows sensitive tests of densities and motions of different types of matter early such scenarios. So far, however, the evidence is against on) were created by quantum fluctuations during the inflation epoch (see section 3). This quantum mechanism generates initial conditions that are for all practical pur- poses random, producing density fluctuations described 3 Bruno’s ideas have since been elaborated by, e.g., Brundrit by what mathematicians call an ergodic random field. (1979), Garriga & Vilenkin (2001b) and Ellis (2002), all of whom Ergodic means that if you imagine generating an ensem- have thus far avoided the stake. 4 If the cosmic expansion continues to accelerate (currently an ble of universes, each with its own random initial con- open question), the observable universe will eventually stop grow- ditions, then the probability distribution of outcomes in ing. a given volume is identical to the distribution that you 4 get by sampling different volumes in a single universe. In the testable prediction that we should observe no oxygen other words, it means that everything that could in prin- here, and is therefore ruled out by observation. ciple have happened here did in fact happen somewhere As a more serious example, the Level I multiverse else. framework is routinely used to rule out theories in mod- Inflation in fact generates all possible initial conditions ern cosmology, although this is rarely spelled out explic- with non-zero probability, the most likely ones being al- itly. For instance, cosmic microwave background (CMB) − most uniform with fluctuations at the 10 5 level that observations have recently shown that space has almost are amplified by gravitational clustering to form galaxies, no curvature. Hot and cold spots in CMB maps have a stars, planets and other structures. This means both that characteristic size that depends on the curvature of space, pretty much all imaginable matter configurations occur and the observed spots appear too large to be consis- in some Hubble volume far away, and also that we should tent with the previously popular “open universe” model. expect our own Hubble volume to be a fairly typical one However, the average spot size randomly varies slightly — at least typical among those that contain observers. from one Hubble volume to another, so it is important to A crude estimate suggests that the closest identical copy be statistically rigorous. When cosmologists say that the 29 91 of you is about ∼ 1010 m away. About ∼ 1010 m away, open universe model is ruled out at 99.9% confidence, there should be a sphere of radius 100 light-years identical they really mean that if the open universe model were to the one centered here, so all perceptions that we have true, then fewer than one out of every thousand Hubble during the next century will be identical to those of our volumes would show CMB spots as large as those we ob- 115 counterparts over there. About ∼ 1010 m away, there serve — therefore the entire model with all its infinitely should be an entire Hubble volume identical to ours.5 many Hubble volumes is ruled out, even though we have This raises an interesting philosophical point that will of course only mapped the CMB in our own particular come back and haunt us in Section V B: if there are Hubble volume. indeed many copies of “you” with identical past lives and The lesson to learn from this example is that multi- memories, you would not be able to compute your own verse theories can be tested and falsified, but only if they future even if you had complete knowledge of the entire predict what the ensemble of parallel universes is and state of the cosmos! The reason is that there is no way specify a probability distribution (or more generally what for you to determine which of these copies is “you” (they mathematicians call a measure) over it. As we will see in all feel that they are). Yet their lives will typically begin Section V B, this measure problem can be quite serious to differ eventually, so the best you can do is predict and is still unsolved for some multiverse theories. probabilities for what you will experience from now on. This kills the traditional notion of determinism. II. LEVEL II: OTHER POST-INFLATION BUBBLES C. How a multiverse theory can be tested and falsified If you felt that the Level I multiverse was large and hard to stomach, try imagining an infinite set of distinct Is a multiverse theory one of metaphysics rather than ones (each symbolized by a bubble in Figure 1), some per- physics? As emphasized by Karl Popper, the distinc- haps with different dimensionality and different physical tion between the two is whether the theory is empirically constants. This is what is predicted by most currently testable and falsifiable. Containing unobservable enti- popular models of inflation, and we will refer to it as ties does clearly not per se make a theory non-testable. the Level II multiverse. These other domains are more For instance, a theory stating that there are 666 paral- than infinitely far away in the sense that you would never lel universes, all of which are devoid of oxygen makes get there even if you traveled at the speed of light forever. The reason is that the space between our Level I multiverse and its neighbors is still undergoing inflation, which keeps stretching it out and creating more volume 5 This is an extremely conservative estimate, simply counting all faster than you can travel through it. In contrast, you possible quantum states that a Hubble volume can have that are 8 115 could travel to an arbitrarily distant Level I universe if no hotter than 10 K. 10 is roughly the number of protons 6 that the Pauli exclusion principle would allow you to pack into you were patient and the cosmic expansion decelerates. a Hubble volume at this temperature (our own Hubble volume contains only about 1080 protons). Each of these 10115 slots can 115 115 be either occupied or unoccupied, giving N = 210 ∼ 1010 possibilities, so the expected distance to the nearest identical 6 115 115 Astronomical evidence suggests that the cosmic expansion is N 1/3 ∼ 10 ∼ 10 Hubble volume is 10 Hubble radii 10 29 meters. currently accelerating. If this acceleration continues, then even Your nearest copy is likely to be much closer than 1010 meters, the level I parallel universes will remain forever separate, with the since the planet formation and evolutionary processes that have intervening space stretching faster than light can travel through tipped the odds in your favor are at work everywhere. There it. The jury is still out, however, with popular models predicting are probably at least 1020 habitable planets in our own Hubble that the universe will eventually stop accelerating and perhaps volume alone. even recollapse. 5

A. Evidence for Level II parallel universes tio mp/me ≈ 1836 and the cosmological constant, which appears to be about 10−123 in so-called Planck units. Inflation is an extension of the big bang theory and There are models where also such non-integer parameters can vary from one one post-inflationary bubble to ties up many of the loose ends of that theory, such as 8 why the universe is so big, so uniform and so flat. A another. In summary, the Level II multiverse is likely rapid stretching of space long ago can explain all these to be more diverse than the Level I multiverse, contain- and other attributes in one fell swoop (see reviews by ing domains where not only the initial conditions differ, Linde 1994 and Guth & Kaiser 2005). Such stretching is but perhaps the dimensionality, the elementary particles predicted by a wide class of theories of elementary par- and the physical constants differ as well. ticles, and all available evidence bears it out. Much of This is currently a very active research area. The pos- space is stretching and will continue doing so forever, but sibility of a string theory “landscape” (Bousso & Polchin- ski 2000; Susskind 2003), where the above-mentioned some regions of space stop stretching and form distinct 500 bubbles, like gas pockets in a loaf of rising bread. In- potential has perhaps 10 different minima, may of- finitely many such bubbles emerge (Figure 1, lower left, fer a specific realization of the Level II multiverse which with time increasing upwards). Each is an embryonic would in turn have four sub-levels of increasing diver- Level I multiverse: infinite in size7 and filled with matter sity: IId: different ways in which space can be compact- deposited by the energy field that drove inflation. Recent ified, which can allow both different effective dimension- cosmological measurements have confirmed two key pre- ality and different symmetries/elementary articles (corre- dictions of inflation: that space has negligible curvature sponding to different topology of the curled up extra di- and that the clumpiness in the cosmic matter distribution mensions). IIc: different “fluxes” (generalized magnetic use to be approximately scale invariant. fields) that stabilize the extra dimensions (this sublevel is where the largest number of choices enter, perhaps 10500). IIb: once these two choices have been made, there may be a handful of different minima in the effec- B. What are Level II parallel universes like? tive supergravity potential. IIa: the same minimum and effective laws of physics can be realized in a many differ- The prevailing view is that the physics we observe to- ent post-inflationary bubbles, each constituting a Level I day is merely a low-energy limit of a much more general multiverse. theory that manifests itself at extremely high tempera- Before moving on, let us briefly comment on a few tures. For example, this underlying fundamental theory closely related multiverse notions. First of all, if one may be 10-dimensional, supersymmetric and involving a Level II multiverse can exist, eternally self-reproducing in grand unification of the four fundamental forces of na- a fractal pattern, then there may well be infinitely many ture. A common feature in such theories is that the other Level II multiverses that are completely discon- potential energy of the field(s) relevant to inflation has nected. However, this variant appears to be untestable, many different minima (sometimes called “metastable since it would neither add any qualitatively different vacuum states”), and ending up in different minima cor- worlds nor alter the probability distribution for their responds to different effective laws of physics for our properties. All possible initial initial conditions and sym- low-energy world. For instance, all but three spatial di- metry breakings are already realized within each one. mensions could be curled up (“compactified”) on a tiny An idea proposed by Tolman and Wheeler and recently scale, resulting in an effectively three-dimensional space elaborated by Steinhardt & Turok (2002) is that the like ours, or fewer could curl up leaving a 5-dimensional (Level I) multiverse is cyclic, going through an infinite space. Quantum fluctuations during inflation can there- series of Big Bangs. If it exists, the ensemble of such in- fore cause different post-inflation bubbles in the Level carnations would also form a multiverse, arguably with II multiverse to end up with different effective laws of a diversity similar to that of Level II. physics in different bubbles — say different dimension- An idea proposed by Smolin (1997) involves an ensem- ality or different types of elementary particles, like two rather than three generations of quarks. In addition to such discrete properties as dimensionality and particle content, our universe is characterized by 8 Although the fundamental equations of physics are the same a set of dimensionless numbers known as physical con- throughout the Level II multiverse, the approximate effective stants. Examples include the electron/proton mass ra- equations governing the low-energy world that we observe will differ. For instance, moving from a three-dimensional to a four- dimensional (non-compactified) space changes the observed gravitational force equation from an inverse square law to an inverse cube law. Likewise, breaking the underlying symmetries of parti- 7 Surprisingly, it has been shown that inflation can produce an in- cle physics differently will change the lineup of elementary parti- finite Level I multiverse even in a bubble of finite spatial volume, cles and the effective equations that describe them. However, we thanks to an effect whereby the spatial directions of spacetime will reserve the terms “different equations” and “different laws of curve towards the (infinite) time direction (Bucher & Spergel physics” for the Level IV multiverse, where it is the fundamental 1999). rather than effective equations that change. 6

hotels, you could infer the existence of other hotel rooms to explain the coincidence.

As a more pertinent example, consider the mass of the sun. The mass of a star determines its luminosity, and using basic physics, one can compute that life as we know it on Earth is possible only if the sun’s mass falls into the narrow range between 1.6 × 1030kg and 2.4 × 1030kg. Otherwise Earth’s climate would be colder than that of present-day Mars or hotter than that of present-day Venus. The measured solar mass is M ∼ 2.0 × 1030kg. At first glance, this apparent coincidence of the habitable and observed mass values appears to be a wild stroke of luck. Stellar masses run from 1029 to 1032kg, so if the sun acquired its mass at random, it had only a small chance of falling into the habitable range. But just as in the hotel example, one can explain this apparent coincidence by postulating an ensemble (in this case, a number of planetary systems) and a selection effect (the fact that we must find ourselves living on a habitable planet). Such observer-related selection effects are referred to as “anthropic” (Carted 1973), and although the “A-word” is FIG. 3: Why we should not be surprised to find ourselves notorious for triggering controversy, physicists broadly living in 3+1-dimensional spacetime. When the partial dif- agree that these selection effects cannot be neglected ferential equations of nature are elliptic or ultrahyperbolic, when testing fundamental theories. In this weak sense, physics has no predictive power for an observer. In the re- the anthropic principle is not optional. maining (hyperbolic) cases, n> 3 admits no stable atoms and n < 3 may lack sufficient complexity for observers (no gravitational attraction, topological problems). From Tegmark What applies to hotel rooms and planetary systems (1997). applies to parallel universes. Most, if not all, of the attributes set by symmetry breaking appear to be fine- tuned. Changing their values by modest amounts would ble similar in diversity to that of Level II, but mutating have resulted in a qualitatively different universe–one in and sprouting new universes through black holes rather which we probably would not exist. If protons were than during inflation. This predicts a form of a natu- 0.2% heavier, they could decay into neutrons, destabiliz- ral selection favoring universes with maximal black hole ing atoms. If the electromagnetic force were 4% weaker, production. there would be no hydrogen and no normal stars. If the In braneworld scenarios, another 3-dimensional world weak interaction were much weaker, hydrogen would not could be quite literally parallel to ours, merely offset in a exist; if it were much stronger, supernovae would fail to higher dimension. However, it is unclear whether such a seed interstellar space with heavy elements. If the cos- world (“brane”) deserves be be called a parallel universe mological constant were much larger, the universe would separate from our own, since we may be able to interact have blown itself apart before galaxies could form. In- with it gravitationally much as we do with dark matter. deed, most if not all the parameters affecting low-energy physics appear fine-tuned at some level, in the sense that changing them by modest amounts results in a qualita- C. Fine-tuning and selection effects tively different universe.

Although we cannot interact with other Level II par- Although the degree of fine-tuning is still debated (as allel universes, cosmologists can infer their presence in- exemplified in the rest of this book; see Barrow & Tipler directly, because their existence can account for unex- 1986, Tegmark 1998 & Hogan (2000) for more technical plained coincidences in our universe. To give an analogy, reviews), these examples suggest the existence of parallel suppose you check into a hotel, are assigned room 1967 universes with other values of some physical constants. and note that this is the year you were born. What a The existence of a Level II multiverse implies that physi- coincidence, you say. After a moment of reflection, how- cists will never be able to determine the values of all ever, you conclude that this is not so surprising after all. physical constants from first principles. Rather, they The hotel has hundreds of rooms, and you would not have will merely compute probability distributions for what been having these thoughts in the first place if you had they should expect to find, taking selection effects into been assigned one with a number that meant nothing to account. The result should be as generic as is consistent you. The lesson is that even if you knew nothing about with our existence. 7

FIG. 4: Hints of fine-tuning for the parameters α and αs which determine the strengths of the electromagnetic force and the strong nuclear force, respectively (from Tegmark 1997). The observed values (α, αs) ≈ (1/137, 0.1) are indi- cated with a filled square. Grand unified theories rule out everything except the narrow strip between the two vertical lines, and deuterium becomes unstable below the horizontal FIG. 5: Difference between Level I and Level III. Whereas line. In the narrow shaded region to the very left, electro- Level I parallel universes are far away in space, those of Level magnetism is weaker than gravity and therefore irrelevant. III are even right here, with quantum events causing classical reality to split and diverge into parallel storylines. Yet Level III adds no new storylines beyond levels 1 or 2. III. LEVEL III: THE MANY WORLDS OF QUANTUM PHYSICS function. According to the Schrödinger equation, this There may be a third type of parallel worlds that are state evolves over time in a fashion that mathematicians not far away but in a sense right here. If the fundamental term “unitary”, meaning that the wave function rotates equations of physics are what mathematicians call uni- in an abstract infinite-dimensional space called Hilbert tary, as they so far appear to be, then the universe keeps space. Although quantum mechanics is often described branching into parallel universes as in the cartoon (Fig- as inherently random and uncertain, the wave function ure 5, bottom): whenever a quantum event appears to evolves in a deterministic way. There is nothing random have a random outcome, all outcomes in fact occur, one or uncertain about it. in each branch. This is the Level III multiverse. Al- The sticky part is how to connect this wave func- though more debated and controversial than Level I and tion with what we observe. Many legitimate wave func- Level II, we will see that, surprisingly, this level adds no tions correspond to counterintuitive situations, such as new types of universes. a cat being dead and alive at the same time in a so- called superposition. In the 1920s physicists explained away this weirdness by postulating that the wave func- A. The quantum conundrum tion “collapse” into some definite classical outcome whenever someone made an observation. This add-on had In the early 20th century the theory of quantum me- the virtue of explaining observations, but it turned an chanics revolutionized physics by explaining the atomic elegant, unitary theory into a kludgy, nonunitary one, realm, which does not abide by the classical rules of New- since there was no equation specifying when or how this tonian mechanics. Despite the obvious successes of the collapse occurred. The intrinsic randomness commonly theory, a heated debate rages about what it really means. ascribed to quantum mechanics is the result of this pos- The theory specifies the state of the universe not in classi- tulate, triggering Einstein’s objection that “God doesn’t cal terms, such as the positions and velocities of all parti- play dice”. cles, but in terms of a mathematical object called a wave Over the years many physicists have abandoned this 8 view in favor of one developed in 1957 by Princeton collapse while preserving unitarity–prevents them from graduate student Hugh Everett III. He showed that the seeing Level III parallel copies of themselves. collapse postulate is unnecessary. Unadulterated quan- Whenever observers are asked a question, make a snap tum theory does not, in fact, pose any contradictions. decision and give an answer, quantum effects in their Although it predicts that one classical reality gradu- brains lead to a superposition of outcomes, such as “Con- ally splits into superpositions of many such realities, ob- tinue reading the article” and “Put down the article”. servers subjectively experience this splitting merely as a From the bird perspective, the act of making a decision slight randomness (Figure 5), with probabilities in exact causes a person to split into multiple copies: one who agreement with those from the old collapse postulate (de keeps on reading and one who doesn’t. From their frog Witt 2003). This superposition of classical worlds is the perspective, however, each of these alter egos is unaware Level III multiverse. of the others and notices the branching merely as a slight randomness: a certain probability of continuing to read or not. B. What are Level III parallel universes like? As strange as this may sound, the exact same situation occurs even in the Level I multiverse. You have evidently Everett’s many-worlds interpretation has been bog- decided to keep on reading the article, but one of your gling minds inside and outside physics for more than alter egos in a distant galaxy put down the magazine after four decades. But the theory becomes easier to grasp the first paragraph. The only difference between Level I when one distinguishes between two ways of viewing a and Level III is where your doppelgaängers reside. In physical theory: the outside view of a physicist study- Level I they live elsewhere in good old three-dimensional ing its mathematical equations, like a bird surveying a space. In Level III they live on another quantum branch landscape from high above it, and the inside view of an in infinite-dimensional Hilbert space (Figure 5). observer living in the world described by the equations, like a frog living in the landscape surveyed by the bird.9. From the bird perspective, the Level III multiverse C. Level III parallel universes: evidence & is simple. There is only one wave function. It evolves implications smoothly and deterministically over time without any kind of splitting or parallelism. The abstract quantum The existence of Level III depends on one crucial as- world described by this evolving wave function contains sumption: that the time evolution of the wave function within it a vast number of parallel classical story lines, is unitary. So far experimenters have encountered no de- continuously splitting and merging, as well as a number partures from unitarity. In the past few decades they of quantum phenomena that lack a classical description. have confirmed unitarity for ever larger systems, includ- From their frog perspective, observers perceive only a ing carbon 60 buckyball molecules and kilometer-long op- tiny fraction of this full reality. They can view their own tical fibers. On the theoretical side, the case for unitarity Level I universe, but a process called decoherence (Zeh has been bolstered by the discovery of decoherence (see 1970; Giulini et al. 1996) — which mimics wave function Tegmark & Wheeler 2001 for a popular review). Some theorists who work on quantum gravity have questioned unitarity; one concern is that evaporating black holes might destroy information, which would be a nonuni- 9 Indeed, the standard mental picture of what the physical world tary process. But a recent breakthrough in string theory is corresponds to a third intermediate viewpoint that could be known as AdS/CFT correspondence suggests that even termed the consensus view. From your subjectively perceived quantum gravity is unitary. If so, black holes do not frog perspective, the world turns upside down when you stand on your head and disappears when you close your eyes, yet you destroy information but merely transmit it elsewhere. subconsciously interpret your sensory inputs as though there is If physics is unitary, then the standard picture of how an external reality that is independent of your orientation, your quantum fluctuations operated early in the big bang must location and your state of mind. It is striking that although change. These fluctuations did not generate initial con- this third view involves both censorship (like rejecting dreams), interpolation (as between eye-blinks) and extrapolation (say at- ditions at random. Rather they generated a quantum tributing existence to unseen cities) of your inside view, inde- superposition of all possible initial conditions, which co- pendent observers nonetheless appear to share this consensus existed simultaneously. Decoherence then caused these view. Although the inside view looks black-and-white to a cat, initial conditions to behave classically in separate quan- iridescent to a bird seeing four primary colors, and still more dif- tum branches. Here is the crucial point: the distribution ferent to bee a seeing polarized light, a bat using sonar, a blind person with keener touch and hearing, or the latest overpriced of outcomes on different quantum branches in a given robotic vacuum cleaner, all agree on whether the door is open. Hubble volume (Level III) is identical to the distribution The key current challenge in physics is deriving this semiclas- of outcomes in different Hubble volumes within a single sical consensus view from the fundamental equations specifying quantum branch (Level I). This property of the quantum the bird perspective. In my opinion, this means that although fluctuations is known in statistical mechanics as ergodic- understanding the detailed nature of human consciousness is an important challenge in its own right, it is not necessary for a ity. fundamental theory of physics. The same reasoning applies to Level II. The process 9 of symmetry breaking did not produce a unique outcome

G but rather a superposition of all outcomes, which rapidly 0 ? g g 0 went their separate ways. So if physical constants, space- h Quantum

General time dimensionality and so on can vary among parallel Field

0 Relativity

g 8 quantum branches at Level III, then they will also vary h Theory g c among parallel universes at Level II. Gg0

Nonrelativistic

Particle In other words, the Level III multiverse adds nothing Special Nuclear Quantum

phenomenology Relativity Physics new beyond Level I and Level II, just more indistinguish- Mechanics able copies of the same universes–the same old story lines hg0 playing out again and again in other quantum branches. Atomic Electro- Classical

Physics magnetism The passionate debate about Everett’s theory therefore Mechanics seems to be ending in a grand anticlimax, with the dis-

Solid State covery of less controversial multiverses (Levels I and II) Statistical

Chemistry

Physics that are equally large. Mechanics Needless to say, the implications are profound, and

Electric

Mechanics Hydro- physicists are only beginning to explore them. For in- Thermo-

properties

of solids dynamics dynamics stance, consider the ramiﬁcations of the answer to a long- of solids standing question: Does the number of universes expo- Astrophysics nentially increase over time? The surprising answer is

Electrical Mechanical no. From the bird perspective, there is of course only Earth Evolution

Engineering Engineering one quantum universe. Science From the frog perspective, what matters is the number

Computer of universes that are distinguishable at a given instant– Biology Science that is, the number of noticeably diﬀerent Hubble vol-

Psychology umes. Imagine moving planets to random new locations, Medicine imagine having married someone else, and so on. At the 118 quantum level, there are 10 to the 10 universes with Sociology temperatures below 108 kelvins. That is a vast number, but a finite one. FIG. 6: Theories can be crudely organized into a family tree From the frog perspective, the evolution of the wave where each might, at least in principle, be derivable from more function corresponds to a never-ending sliding from one fundamental ones above it. For example, classical mechanics of these 10 to the 10118 states to another. Now you are can be obtained from special relativity in the approximation in universe A, the one in which you are reading this sen- that the speed of light c is infinite. Most of the arrows are tence. Now you are in universe B, the one in which you less well understood. All these theories have two components: are reading this other sentence. Put differently, universe mathematical equations and words that explain how they are B has an observer identical to one in universe A, except connected to what we observe. At each level in the hierarchy of theories, new words (e.g., protons, atoms, cells, organisms, with an extra instant of memories. All possible states cultures) are introduced because they are convenient, cap- exist at every instant, so the passage of time may be turing the essence of what is going on without recourse to in the eye of the beholder — an idea explored in Greg the more fundamental theory above it. It is important to re- Egan’s 1994 science-fiction novel Permutation City and member, however, that it is we humans who introduce these developed by physicist David Deutsch of the University of concepts and the words for them: in principle, everything Oxford, independent physicist Julian Barbour, and oth- could have been derived from the fundamental theory at the ers. The multiverse framework may thus prove essential top of the tree, although such an extreme reductionist ap- to understanding the nature of time. proach would of course be useless in practice. Crudely speaking, the ratio of equations to words decreases as we move down the tree, dropping near zero for highly applied fields such as medicine and sociology. In contrast, theories near the top D. Two world views are highly mathematical, and physicists are still struggling to understand the concepts, if any, in terms of which we can The debate over how classical mechanics emerges from understand them. The Holy Grail of physics is to find what quantum mechanics continues, and the decoherence dis- is jocularly referred to as a “Theory of Everything”, or TOE, covery has shown that there is a lot more to it than just from which all else can be derived. If such a theory exists letting Planck’s constanth ¯ shrink to zero. Yet as Fig- at all, it should replace the big question mark at the top of ure 6 illustrates, this is just a small piece of a larger puz- the theory tree. Everybody knows that something is missing here, since we lack a consistent theory unifying gravity with zle. Indeed, the endless debate over the interpretation of quantum mechanics. quantum mechanics — and even the broader issue of parallel universes — is in a sense the tip of an iceberg. In the Sci-Fi spoof “Hitchhiker’s Guide to the Galaxy”, the answer is discovered to be “42”, and the hard part is finding 10 the real question. Questions about parallel universes may equations are equally real. This is the Level IV multi- seem to be just about as deep as queries about reality can verse. First we need to digest two other ideas, however: get. Yet there is a still deeper underlying question: there the concept of a mathematical structure, and the notion are two tenable but diametrically opposed paradigms re- that the physical world may be one. garding physical reality and the status of mathematics, a dichotomy that arguably goes as far back as Plato and Aristotle, and the question is which one is correct. A. What is a mathematical structure? • ARISTOTELIAN PARADIGM: The subjectively perceived frog perspective is physically real, Many of us think of mathematics as a bag of tricks that and the bird perspective and all its mathematical we learned in school for manipulating numbers. Yet most language is merely a useful approximation. mathematicians have a very different view of their field. They study more abstract objects such as functions, sets, • PLATONIC PARADIGM: The bird perspec- spaces and operators and try to prove theorems about tive (the mathematical structure) is physically real, the relations between them. Indeed, some modern math- and the frog perspective and all the human lan- ematics papers are so abstract that the only numbers guage we use to describe it is merely a useful ap- you will find in them are the page numbers! What does proximation for describing our subjective percep- a dodecahedron have in common with a set of complex tions. numbers? Despite the plethora of mathematical struc- What is more basic — the frog perspective or the bird tures with intimidating names like orbifolds and Killing perspective? What is more basic — human language fields, a striking underlying unity has emerged in the or mathematical language? Your answer will determine last century: all mathematical structures are just special how you feel about parallel universes. If you prefer the cases of one and the same thing: so-called formal sys- Platonic paradigm, you should find multiverses natu- tems. A formal system consists of abstract symbols and ral, since our feeling that say the Level III multiverse rules for manipulating them, specifying how new strings is “weird” merely reflects that the frog and bird perspec- of symbols referred to as theorems can be derived from tives are extremely different. We break the symmetry by given ones referred to as axioms. This historical devel- calling the latter weird because we were all indoctrinated opment represented a form of deconstructionism, since it with the Aristotelian paradigm as children, long before stripped away all meaning and interpretation that had we even heard of mathematics - the Platonic view is an traditionally been given to mathematical structures and acquired taste! distilled out only the abstract relations capturing their In the second (Platonic) case, all of physics is ulti- very essence. As a result, computers can now prove the- mately a mathematics problem, since an infinitely intel- orems about geometry without having any physical intu- ligent mathematician given the fundamental equations of ition whatsoever about what space is like. the cosmos could in principle compute the frog perspec- Figure 7 shows some of the most basic mathematical tive, i.e., compute what self-aware observers the universe structures and their interrelations. Although this fam- would contain, what they would perceive, and what lan- ily tree probably extends indefinitely, it illustrates that guage they would invent to describe their perceptions there is nothing fuzzy about mathematical structures. to one another. In other words, there is a “Theory of They are “out there” in the sense that mathematicians Everything” (TOE) at the top of the tree in Figure 6 discover them rather than create them, and that contem- whose axioms are purely mathematical, since postulates plative alien civilizations would find the same structures in English regarding interpretation would be derivable (a theorem is true regardless of whether it is proven by and thus redundant. In the Aristotelian paradigm, on a human, a computer or an alien). the other hand, there can never be a TOE, since one is ultimately just explaining certain verbal statements by other verbal statements — this is known as the infinite B. The possibility that the physical world is a regress problem (Nozick 1981). mathematical structure

Let us now digest the idea that physical world (specif- IV. LEVEL IV: OTHER MATHEMATICAL ically, the Level III multiverse) is a mathematical struc- STRUCTURES ture. Although traditionally taken for granted by many theoretical physicists, this is a deep and far-reaching no- Suppose you buy the Platonist paradigm and believe tion. It means that mathematical equations describe not that there really is a TOE at the top of Figure 6 — and merely some limited aspects of the physical world, but all that we simply have not found the correct equations yet. aspects of it. It means that there is some mathematical Then an embarrassing question remains, as emphasized structure that is what mathematicians call isomorphic by John Archibald Wheeler: Why these particular equa- (and hence equivalent) to our physical world, with each tions, not others? Let us now explore the idea of mathe- physical entity having a unique counterpart in the math- matical democracy, whereby universes governed by other ematical structure and vice versa. Let us consider some 11

GENERAL RELATIVITY QUANTUM FIELD THEORY (3+1-dimensional pseudo-Riemannian (Operator-valued fields on R 4 obeying certain Lorentz- LORENTZ SU(2) manifolds with tensor fields obeying PDE's of, say, invariant PDE's and commutation relationships, structure would not correspond to a single frame of it, GROUP Einstein-Maxwell theory with perfect fluid component) acting on an abstract Hilbert space) Specialize but to the entire videotape.

LIE LIE DIFFERENTIAL MANIFOLDS WITH METRIC Given a mathematical structure, we will say that it has GROUPS ALGEBRAS OPERATORS TENSOR FIELDS MANIFOLDS physical existence if any self-aware substructure (SAS) LINEAR TENSOR REAL COMPLEX HILBERT DISTRI- ALGEBRAS within it subjectively, from its frog perspective, perceives OPERATORS SPACES MANIFOLDS MANIFOLDS SPACES BUTIONS Define linear Define charts and Add inner Define linear vector atlases product functionals itself as living in a physically real world. What would, TRIPLE mappings Define linear FIELDS functions of BANACH REAL n vectors Rn Cn Add 4th SPACES FUNCTIONS mathematically, such an SAS be like? In the classical binary op. VECTOR etc. SPACES Define . physics example above, an SAS such as you would be DOUBLE Add new ckass REAL =, + & COMPLEX COMPLEX Add ordering, for pairs FIELDS (vectors), define least upper bound NUMBERS NUMBERS FUNCTIONS vector sum & property scalar product Add a tube through spacetime, a thick version of what Ein- Add 3rd norm & binary operation, etc. Define =, + & .for complete- MEASURABLE ABELIAN Dedekind cuts ness SPACES stein referred to as a world-line. The location of the tube FIELDS RATIONAL METRIC Add NUMBERS measure would specify your position in space at different times. Add commutativity SPACES ABELIAN Define =, + & . for pairs GROUPS FIELDS Add INTEGERS metric Within the tube, the fields would exhibit certain complex Add commutativity Add multiplicative inverse Define =, + & . for TOPOLOGICAL GROUPS RINGS negative numbers SPACES behavior, corresponding to storing and processing infor- Add inverse Add commutativity and 2nd NATURAL Add assoc. & distrib. binary operation NUMBERS open Specialize set ABSTRACT mation about the field-values in the surroundings, and SEMI- Add number theory axioms GEOMETRIES GROUPS symbols & axioms at each position along the tube, these processes would DODECA- Add symbols Add = & associative LOWER HEDRON and axioms for binary operation with identity PREDICATE SETS RELATIONS give rise to the familiar but mysterious sensation of self- GROUP =, in, union Add CALCULUS and intersection cartesian products Add quantifiers awareness.From its frog perspective, the SAS would per- BOOLEAN ceive this one-dimensional string of perceptions along the ALGEBRA Add symbols & axioms for brackets, not, and, etc. tube as passage of time. MODELS FORMAL Although our example illustrates the idea of how our SYSTEMS physical world can be a mathematical structure, this particular mathematical structure (fields in four-dimensional FIG. 7: Relationships between various basic mathematical space) is now known to be the wrong one. After real- structures (Tegmark 1998). The arrows generally indicate izing that spacetime could be curved, Einstein doggedly addition of new symbols and/or axioms. Arrows that meet searched for a so-called unified field theory where the uni- indicate the combination of structures — for instance, an al- verse was what mathematicians call a 3+1-dimensional gebra is a vector space that is also a ring, and a Lie group pseudo-Riemannian manifold with tensor fields (top cen- is a group that is also a manifold. The full tree is probably ter in Figure 7), but this failed to account for the ob- infinite in extent — the figure shows merely a small sample served behavior of atoms. According to quantum field near the bottom. theory, the modern synthesis of special relativity theory and quantum theory, the universe (in this case the Level III multiverse) is a mathematical structure known as an examples. algebra of operator-valued fields (top right in Figure 7). Here the question of what constitutes an SAS is more A century ago, when classical physics still reigned subtle (Tegmark 2000). However, this fails to describe supreme, many scientists believed that physical space black hole evaporation, the first instance of the Big Bang was isomorphic to the mathematical structure known as 3 and other quantum gravity phenomena, so the true math- R : three-dimensional Euclidean space. Moreover, some ematical structure isomorphic to our universe, if it exists, thought that all forms of matter in the universe cor- has not yet been found. responded to various classical fields: the electric field, the magnetic field and perhaps a few undiscovered ones, mathematically corresponding to functions on R3 (a handful of numbers at each point in space). In this C. Mathematical democracy view (later proven incorrect), dense clumps of matter like atoms were simply regions in space where some fields Now suppose that our physical world really is a math- were strong (where some numbers were large). These ematical structure, and that you are an SAS within it. fields evolved deterministically over time according to This means that in the Mathematics tree of Figure 7, one some partial differential equations, and observers per- of the boxes is our universe. (The full tree is probably ceived this as things moving around and events taking infinite in extent, so our particular box is not one of the place. Could, then, fields in three-dimensional space few boxes from the bottom of the tree that are shown.) be the mathematical structure corresponding to the uni- In other words, this particular mathematical structure verse? No, since a mathematical structure cannot change enjoys not only mathematical existence, but physical ex- — it is an abstract, immutable entity existing outside of istence as well. What about all the other boxes in the space and time. Our familiar frog perspective of a three- tree? Do they too enjoy physical existence? If not, there dimensional space where events unfold is equivalent, from would be a fundamental, unexplained ontological asym- the bird perspective, to a four-dimensional spacetime metry built into the very heart of reality, splitting mathe- where all of history is contained, so the mathematical matical structures into two classes: those with and with- structure would be fields in four-dimensional space. In out physical existence. As a way out of this philosophical other words, if history were a movie, the mathematical conundrum, I have suggested (Tegmark 1998) that com- 12 plete mathematical democracy holds: that mathemat- This argument also makes assumption 2 more appeal- ical existence and physical existence are equivalent, so ing, since it implies that any conceivable parallel uni- that all mathematical structures exist physically as well. verse theory can be described at Level IV. The Level IV This is the Level IV multiverse. It can be viewed as a multiverse, termed the “ultimate Ensemble theory” in form of radical Platonism, asserting that the mathemat- Tegmark (1997) since it subsumes all other ensembles, ical structures in Plato’s realm of ideas, the Mindscape therefore brings closure to the hierarchy of multiverses, of Rucker (1982), exist “out there” in a physical sense and there cannot be say a Level V. Considering an en- (Davies 1993), casting the so-called modal realism the- semble of mathematical structures does not add anything ory of David Lewis (1986) in mathematical terms akin to new, since this is still just another mathematical struc- what Barrow (1991; 1992) refers to as “π in the sky”. If ture. What about the frequently discussed notion that this theory is correct, then since it has no free parame- the universe is a computer simulation? This idea occurs ters, all properties of all parallel universes (including the frequently in science fiction and has been substantially subjective perceptions of SASs in them) could in princi- elaborated (e.g., Schmidthuber 1997; Wolfram 2002). ple be derived by an infinitely intelligent mathematician. The information content (memory state) of a digital computer is a string of bits, say “1001011100111001...” of great but finite length, equivalent to some large but fi- D. Evidence for a Level IV multiverse nite integer n written in binary. The information processing of a computer is a deterministic rule for chang- We have described the four levels of parallel universes ing each memory state into another (applied over and in order of increasing speculativeness, so why should we over again), so mathematically, it is simply a function f believe in Level IV? Logically, it rests on two separate mapping the integers onto themselves that gets iterated: 7→ 7→ 7→ assumptions: n f(n) f(f(n)) .... In other words, even the most sophisticated computer simulation is just yet an- • Assumption 1: That the physical world (specif- other special case of a mathematical structure, and is al- ically our level III multiverse) is a mathematical ready included in the Level IV multiverse. (Incidentally, structure iterating continuous functions rather than integer-valued ones can give rise to fractals.) • Assumption 2: Mathematical democracy: that A second argument for assumption 2 is that if two en- all mathematical structures exist “out there” in the tities are isomorphic, then there is no meaningful sense same sense in which they are not one and the same (Cohen 2003). In a famous essay, Wigner (1967) argued that “the This implies assumption 2 when the entities in question enormous usefulness of mathematics in the natural sci- are a physical universe and a mathematical structure de- ences is something bordering on the mysterious”, and scribing it, respectively. To avoid this conclusion that that “there is no rational explanation for it”. This argu- mathematical and physical existence are equivalent, one ment can be taken as support for assumption 1: here the would need to argue that our universe is somehow made utility of mathematics for describing the physical world of stuff perfectly described by a mathematical structure, is a natural consequence of the fact that the latter is but which also has other properties that are not described a mathematical structure, and we are simply uncover- by it. However, this violates assumption 1 and implies ing this bit by bit. The various approximations that either that it is isomorphic to a more complicated math- constitute our current physics theories are successful be- ematical structure or that it is not mathematical at all. cause simple mathematical structures can provide good The latter would be make Karl Popper turn in his grave, approximations of how a SAS will perceive more complex since those additional bells and whistles that make the mathematical structures. In other words, our successful universe non-mathematical by definition have no observ- theories are not mathematics approximating physics, but able effects whatsoever. mathematics approximating mathematics. Wigner’s ob- Another appealing feature of assumption 2 is that it servation is unlikely to be based on fluke coincidences, provides the only answer so far to Wheeler’s question: since far more mathematical regularity in nature has been Why these particular equations, not others? Having uni- discovered in the decades since he made it, including the verses dance to the tune of all possible equations also standard model of particle physics. resolves the fine-tuning problem of Section II C once and A second argument supporting assumption 1 is that for all, even at the fundamental equation level: although abstract mathematics is so general that any TOE that is many if not most mathematical structures are likely to definable in purely formal terms (independent of vague be dead and devoid of SASs, failing to provide the com- human terminology) is also a mathematical structure. plexity, stability and predictability that SASs require, we For instance, a TOE involving a set of different types of course expect to find with 100% probability that we of entities (denoted by words, say) and relations between inhabit a mathematical structure capable of supporting them (denoted by additional words) is nothing but what life. Because of this selection effect, the answer to the mathematicians call a set-theoretical model, and one can question “what is it that breathes fire into the equations generally find a formal system that it is a model of. and makes a universe for them to describe?” (Hawking 13

1993) would then be “you, the SAS”. A. Future prospects

There are ample future prospects for testing and per- E. What are Level IV parallel universes like? haps ruling out these multiverse theories. In the com- ing decade, dramatically improved cosmological measure- The way we use, test and potentially rule out any the- ments of the microwave background radiation, the large- ory is to compute probability distributions for our future scale matter distribution, etc., will test Level I by fur- perceptions given our past perceptions and to compare ther constraining the curvature and topology of space and these predictions with our observed outcome. In a mul- will test level II by providing stringent tests of inflation. tiverse theory, there is typically more than one SAS that Progress in both astrophysics and high-energy physics has experienced a past life identical to yours, so there is should also clarify the extent to which various physical no way to determine which one is you. To make predic- constants are fine-tuned, thereby weakening or strength- tions, you therefore have to compute what fractions of ening the case for Level II. If the current world-wide effort them will perceive what in the future, which leads to the to build quantum computers succeeds, it will provide fur- following predictions: ther evidence for Level III, since they would, in essence, be exploiting the parallelism of the Level III multiverse • Prediction 1: The mathematical structure de- for parallel computation (Deutsch 1997). Conversely, ex- scribing our world is the most generic one that is perimental evidence of unitarity violation would rule out consistent with our observations. Level III. Finally, success or failure in the grand challenge of modern physics, unifying general relativity and quan- • Prediction 2: Our future observations are the tum field theory, will shed more light on Level IV. Either most generic ones that are consistent with our past we will eventually find a mathematical structure match- observations. ing our universe, or we will bump up against a limit to the unreasonable effectiveness of mathematics and have • Prediction 3: Our past observations are the most to abandon Level IV. generic ones that are consistent with our existence.

We will return to the problem of what “generic” means B. The measure problem in Section V B (the measure problem). However, one striking feature of mathematical structures, discussed in There are also interesting theoretical issues to resolve detail in Tegmark (1997), is that the sort of symmetry within the multiverse theories, first and foremost the and invariance properties that are responsible for the sim- measure problem. plicity and orderliness of our universe tend to be generic, As multiverse theories gain credence, the sticky issue of more the rule than the exception — mathematical struc- how to compute probabilities in physics is growing from a tures tend to have them by default, and complicated ad- minor nuisance into a major embarrassment. If there are ditional axioms etc. must be added to make them go indeed many identical copies of you, the traditional no- away. In other words, because of both this and selection of determinism evaporates. You could not compute tion effects, we should not necessarily expect life in the your own future even if you had complete knowledge of Level IV multiverse to be a disordered mess. the entire state of the multiverse, because there is no way for you to determine which of these copies is you (they all feel they are). All you can predict, therefore, are prob- V. DISCUSSION abilities for what you would observe. If an outcome has a probability of, say, 50 percent, it means that half the We have seen that scientific theories of parallel uni- observers observe that outcome. verses form a four-level hierarchy, in which universes be- Unfortunately, it is not an easy task to compute what come progressively more different from ours. They might fraction of the infinitely many observers perceive what. have different initial conditions (Level I), different effec- The answer depends on the order in which you count tive physical laws, constants and particles (Level II), or them. By analogy, the fraction of the integers that are different fundamental physical laws (Level IV). It is ironic even is 50 percent if you order them numerically (1, 2, that Level III is the one that has drawn the most fire in 3, 4, ...) but approaches 100 percent if you sort them the past decades, because it is the only one that adds no digit by digit, the way your word processor would (1, 10, qualitatively new types of universes. 100, 1,000, ...). When observers reside in disconnected Whereas the Level I universes join seemlessly, there are universes, there is no obviously natural way in which to clear demarcations between those within levels II and III order them. Instead one must sample from the differ- caused by inflating space and decoherence, respectively. ent universes with some statistical weights referred to by The level IV universes are completely disconnected and mathematicians as a “measure”. need to be considered together only for predicting your This problem crops up in a mild and treatable manner future, since “you” may exist in more than one of them. at Level I, becomes severe at Level II (see Tegmark 2004 14 for a detailed review), has caused much debate at Level entire set can be generated by quite a trivial computer III (de Witt 2003, Mukhanov 2005), and is horrendous at program, whereas a single number can be hugely long. Level IV. At Level II, for instance, Linde, Vilenkin and Therefore, the whole set is actually simpler. others have published predictions for the probability dis- Similarly, the set of all solutions to Einstein’s field tributions of various cosmological parameters. They have equations is simpler than a specific solution. The for- argued that different parallel universes that have inflated mer is described by a few equations, whereas the latter by different amounts should be given statistical weights requires the specification of vast amounts of initial data proportional to their volume (e.g., Garriga & Vilenkin on some hypersurface. The lesson is that complexity in- 2001a). On the other hand, any mathematician will tell creases when we restrict our attention to one particular you that 2 × ∞ = ∞, so there is no objective sense in element in an ensemble, thereby losing the symmetry and which an infinite universe that has expanded by a factor simplicity that were inherent in the totality of all the el- of two has gotten larger. Moreover, a finite universe with ements taken together. the topology of a torus is equivalent to a perfectly peri- In this sense, the higher-level multiverses are simpler. odic universe with infinite volume, both from the mathe- Going from our universe to the Level I multiverse elimi- matical bird perspective and from the frog perspective of nates the need to specify initial conditions, upgrading to an observer within it. So why should its infinitely smaller Level II eliminates the need to specify physical constants, volume give it zero statistical weight? After all, even in and the Level IV multiverse eliminates the need to spec- the Level I multiverse, Hubble volumes start repeating ify anything at all. The opulence of complexity is all in (albeit in a random order, not periodically) after about the subjective perceptions of observers (Tegmark 1996) 118 10 to the 10 meters. — the frog perspective. From the bird perspective, the If you think that is bad, consider the problem of assign- multiverse could hardly be any simpler. ing statistical weights to different mathematical struc- The complaint about weirdness is aesthetic rather than tures at Level IV. The fact that our universe seems rel- scientific, and it really makes sense only in the Aris- atively simple has led many people to suggest that the totelian worldview. Yet what did we expect? When correct measure somehow involves complexity. we ask a profound question about the nature of reality, do we not expect an answer that sounds strange? Evolution provided us with intuition for the everyday C. The pros and cons of parallel universes physics that had survival value for our distant ancestors, so whenever we venture beyond the everyday world, we So should you believe in parallel universes? We have should expect it to seem bizarre. Thanks to clever in- seen that this is not a yes/no question — rather, the most ventions, we have glimpsed slightly beyond the frog per- interesting issue is whether there are 0, 1, 2, 3 or 4 levels spective of our normal inside view, and sure enough, we of multiverses. Figure 1 summarizes the evidence that have encountered bizarre phenomena whenever depart- we have discussed for the different levels. The principal ing from human scales in any way: at high speeds (time arguments against them are that they are wasteful and slows down), on small scales (quantum particles can be that they are weird. at several places at once), on large scales (black holes), The wastefulness argument is that multiverse theories at low temperatures (liquid Helium can flow upward), at are vulnerable to Occam’s razor because they postulate high temperatures (colliding particles can change iden- the existence of other worlds that we can never observe. tity), etc.. Why should nature be so wasteful and indulge in such A common feature of all four multiverse levels is that opulence as an infinity of different worlds? Yet this ar- the simplest and arguably most elegant theory involves gument can be turned around to argue for a multiverse. parallel universes by default. To deny the existence of What precisely would nature be wasting? Certainly not those universes, one needs to complicate the theory by space, mass or atoms–the uncontroversial Level I multi- adding experimentally unsupported processes and ad hoc verse already contains an infinite amount of all three, so postulates: finite space, wave function collapse, ontologi- who cares if nature wastes some more? The real issue cal asymmetry, etc. Our judgment therefore comes down here is the apparent reduction in simplicity. A skeptic to which we find more wasteful and inelegant: many worries about all the information necessary to specify all worlds or many words. Perhaps we will gradually get those unseen worlds. more used to the weird ways of our cosmos, and even But an entire ensemble is often much simpler than one find its strangeness to be part of its charm. of its members. This principle can be stated more for- mally using the notion of algorithmic information con- Acknowledgements: The author wishes to thank tent. The algorithmic information content in a number Anthony Aguirre, Aaron Classens, Marius Cohen, Angel- is, roughly speaking, the length of the shortest computer ica de Oliveira-Costa, Alan Guth, Shamit Kachru, An- program that will produce that number as output. For drei Linde, George Musser, David Raub, Martin Rees, example, consider the set of all integers. Which is sim- Harold Shapiro, Alex Vilenkin and Frank Wilczek for pler, the whole set or just one number? Naively, you stimulating discussions and George Musser for exten- might think that a single number is simpler, but the sive editing and improvement of the text of a shorter 15 predecessor of this article that was published in Scien- Giulini, D., Joos, E., Kiefer, C., Kupsch, J., Stamatescu, tific American. This work was supported by NSF grants I. O., & Zeh, H. D. 1996, Decoherence and the Appear- AST-0071213 & AST-0134999, NASA grants NAG5-9194 ance of a Classical World in Quantum Theory (Berlin: & NAG5-11099, a fellowship from the David and Lucile Springer) Packard Foundation and a Cottrell Scholarship from Re- Guth, A., & Steinhardt, P. J. 1984, Sci. Am., 250, 116 search Corporation. Hawking, S. 1993, A Brief History of Time (Touchstone: New York) Lewis, D. 1986, On the Plurality of Worlds (Blackwell: VI. REFERENCES Oxford) Linde, A. 1994, Sci. Am., 271, 32 Anglin, J. R., & Zurek, W. H. 1996, Phys. Rev. D, 53, Maldacena, J. 2003, in Science and Ultimate Reality: From 7327 Quantum to Cosmos, ed. Barrow, J. D., Davies, P. Barbour J B, 2001 (The End of Time: Oxford) C. W., & Harper, C. L. (Cambridge Univ. Press: Cam- Univ. Press;Oxford bridge) Barrow, J. D. 1991, Theories of Everything (Ballantine: Nozick, R. 1981, Philosophical Explanations (Harvard New York) Univ. Press: Cambridge) Barrow, J. D. 1992, Pi in the Sky (Clarendon: Oxford) Rees, M. J. 2002, Our Cosmic Habitat (Princeton Univ. Barrow, J. D., & Tipler, F. J. 1986, The Anthropic Cos- Press: Princeton) mological Principle (Clarendon: Oxford) Rucker, R. 1982, Infinity and the Mind (Birkhauser: Bousso, R., & Polchinski, J. 2000, JHEP, 6, 6 Boston) Brundrit, G. B. 1979, Q. J. Royal Astr. Soc., 20, 37 Schmidthuber, J. 1997, A Computer Scientist’s Bucher, M. A., & Spergel, D. N. 1999, Sci. Am. 1/1999 View of Life, the Universe, and Everything, in Carter, B. 1974, in IAU Symposium 63, ed. Longair, S. Foundations of Computer Science: Potential- (Reidel: Dordrecht) Theory-Cognition, Lecture Notes in Computer Chaitin, G. J. 1987, Algorithmic Information Theory Science, C. Freksa, ed., (Springer: Berlin), (Cambridge U. P: Cambridge) http://www.idsia.ch∼juergen/everything/html.html Cohen, M. 2003, Master’s thesis, Dept. of Philosophy, Smolin, L. 1997, The Life of the Cosmos (Oxford Univ. Ben Gurion University of the Negev, Israel Press: Oxford) Davies, P. C. W. 1982, The Accidental Universe (Cam- Steinhardt, P. J., & Turok, N. 2002, Science, 296, 1436 bridge U. P: Cambridge) Susskind, L. 2003, hep-th/0302219 Davies, P. 1993, The Mind of God (Touchstone: New York) Tegmark, M. 1996, Found. Phys. Lett., 9, 25 Davies, P. 1996, Are We Alone? (Basic Books: New York) Tegmark, M. 1997, Class. Quant. Grav., 14, L69 Deutsch, D. 1997, The Fabric of Reality (Allen Lane: New Tegmark, M. 1998, Ann. Phys., 270, 1 York) Tegmark, M. 2000, Phys. Rev. E, 61, 4194 Deutsch, D. 2003, in Science and Ultimate Reality: From Tegmark, M. 2002, Science, 296, 1427 Quantum to Cosmos, ed. Barrow, J. D., Davies, P. Tegmark, M., & Wheeler, J. A. 2001, Sci.Am., 2/2001, C. W., & Harper, C. L. (Cambridge Univ. Press: Cam- 68-75 bridge) Tegmark, M. 2004, astro-ph/0410281 de Witt, B. 2003, in Science and Ultimate Reality: From Wigner, E. P. 1967, Symmetries and Reflections (MIT Quantum to Cosmos, ed. Barrow, J. D., Davies, P. Press: Cambridge) C. W., & Harper, C. L. (Cambridge Univ. Press: Cam- Wolfram, S. 2.002, A New Kind of Science (Wolfram Me- bridge) dia: New York) Egan, G. 1995, Permutation City (Harper: New York) Zeh, H. D. 1970, Found. Phys., 1, 69 Garriga, J., & Vilenkin, A. 2001a, Phys. Rev. D, 64, Zurek, W. 2003, in Science and Ultimate Reality: From 023507 Quantum to Cosmos, ed. Barrow, J. D., Davies, P. Garriga, J., & Vilenkin, A. 2001b, Phys. Rev. D, 64, C. W., & Harper, C. L. (Cambridge Univ. Press: Cam- 043511. bridge)