 | ||
This drawing is from Alan Guth, MIT Dept. of Physics. It is what physicists call an 'embedding diagram'. It represents three-dimensional space as a two-dimensional surface that is 'bent' or curved by gravity. This diagram depicts one of the basic ways that an existing universe (e.g., the one we live in) can generate, via quantum tunneling, an entirely new and separate child universe. Currently known and accepted laws of physics appear to allow this — a quantum 'phase change' of a small 'seed' of matter (just a few grams) into a peculiar state called a 'false vacuum' or something like it, which has negative pressure, and can therefore expand into a whole universe at an astounding rate. The diagram depicts the moment that the 'umbilical cord' (wormhole or Ellis Drainhole) connecting the parent and child universe snaps, and the child universe becomes a completely separate new universe. See Farhi, Guth, and Guven (Nuclear Physics B, Volume 339, Issue 2, 30 July 1990, Pages 417-490) for all the detail. |
* * *
Universe Self-Replication Cosmology: A Metaparadigm worth discussing?
Peter J. Wetzel, NASA Goddard Space Flight Center, retired
(last updated 26 August 2025)
SUMMARY
Universe Self-Replication (USeR) Cosmology is the hypothesis that the physical laws of our universe permit it to self-replicate. Several different mechanisms by which this may occur are identified, including mechanisms that restart the universe chronology from an early epoch and mechanisms that only act upon a current state and then advance the multiple 'copies' forward in time. The robustness of the hypothesis stems from this abundance of different non-prohibited mechanisms rather than relying on the (still debated) details. Some of these mechanisms allow for mutation during the replication process. The important consequence of the hypothesis is that in a USeR Cosmology, any given universe is necessarily the offspring of a previous universe. There is an inheritance of physical laws and constants during replication, which seems analogous to the process of reproduction in living things, and thus points to a correspondence with Darwinian Evolution. Even evaluated in a narrow sense, this analogy points to the possibility that the physics governing the earliest epochs of our universe need not have been a predictable (falsifiable) or even a unique sequence of events. To paraphrase Einstein: "God had a choice." Further consequences of this analogy are discussed.
* * *
Introduction and Personal Note: This manuscript addresses in broad-brush conceptual terms the simple, hard, physical argument of whether and how a universe with the physical laws that ours has can self-replicate. Six largely independent theoretical ways that this can occur are addressed. The details of each mechanism are in various states of uncertainty/debate, but the over-arching argument gains its power from the fact that it stands on at least six different independent pillars. The far-reaching consequences that seem to inevitably result are discussed. Many of the self-replication mechanisms have been under discussion in published technical literature for decades (see the headline figure for perhaps the earliest example), and yet discussion of the over-arching big-picture implications seems largely absent. It is argued that the consequences are so significant that they are worthy of 'a seat at the table' in the Cosmology discourse regardless of, or independent of the particulars of any one mechanism.
The author is not a Cosmologist, but a PhD Atmospheric Physicist who worked at NASA GSFC (Greenbelt, MD) as a Civil Service research Atmospheric Scientist for 25 years. As of this writing, he is approaching the age of 77. This manuscript is written at a non-technical but scientifically literate level. It is intended to be a catalyst for scholarly discussion among a new generation of physicists who have an interest in expanding the horizons of Cosmology and Cosmogony and Foundational Theoretical Physics. As such, the author is seeking review, comment, guidance, and inspiration from the physics community. Most importantly, it is a call for rigorous consideration (mathematically as well as philosophically) of the potentially intrinsic non-falsifiability of the paradigm, and, in parallel, the search for experimental evidence supporting any of the six self-replication mechanisms.
Question 1: Does a universe with the physical laws that our observed universe currently exhibits permit self-replication?
Answer: We do not understand the full set of physical laws describing all aspects of our universe's behavior. In particular, the earliest period (the 'birth') of our universe seems to be best described by extrapolating toward high energy and small dimensions, which is the realm where the two highly successful models we have of our universe, Quantum Mechanics and General Relativity, would need to merge (actually emerge [going forward in time] perhaps via some form of symmetry-breaking in a more fundamental, more general theory). To date, despite at least a century of intensive theoretical and experimental study, no satisfactory (generally accepted, experimentally confirmed) form of such a theory has been identified.
One way of attacking this question is to assume that the two models of our universe are both correct, and that no fundamental merger is required because self-replication can occur within the framework (domain of scales) of either model. The physical systems that both Quantum Mechanics and General Relativity describe include mechanisms that do appear to allow self-replication (see next question).
We have no experimental evidence that self-replication has happened, though there are scenarios where the event could be undetectable. Importantly, there is also no evidence that self-replication is prohibited.
Question 2: What are the six ways that our currently accepted physical laws permit self-replication to happen?
First way: Within Quantum Mechanics: a natural, spontaneous quantum tunneling. It seems to require very high density and high energy, states that have very low probability, but not zero probability. A number of papers, over the years, have been published describing how a universe could spontaneously form from vacuum fluctuations. In 2012, Lawrence M. Krauss wrote a popular-level book on the subject, 'A Universe from Nothing,' which hit the NYT best-seller list. The popular interpretation of these discussions is usually along the lines that our universe could appear from some kind of pure vacuum, i.e. from no prior substrate and no prior laws; but, as critics have noted, that argument is not well posed. 'Nothing' is an absolute. You can't chip a piece off of it and get 'something.' The best we can do in our contingent, non-abstract, non-absolute world, is argue that what is at the root of it all is a core irresolvable, inexplicable, paradox—a virtual superposition of nothing and something, as represented by the relationship 0 ~ 1/∞. What we observe emerging from this paradox when we interrogate this 'nothing' (the observable vacuum) is a quantum froth. And that's just fine for purposes of this discussion. Since vacuum fluctuations are an essential part of the quantum theory of our universe, the arguments work if some pre-existing state of our universe was the starting point of the process of emerging (or tunneling). Sean Carroll and Jennifer Chen, in 2005, proposed such a mechanism that is favored in the late universe, assuming the persistence of a cosmological constant. As the universe continues to expand, and all available hydrogen fuel is spent, and all black holes have evaporated, the resulting empty space would be prone to form low-entropy fluctuations that could initiate inflation. Perhaps we do not have to wait for the entire universe to achieve 'emptiness' for this mechanism to be actuated. Two possibilities are in an advanced laboratory (see the Third Way below) and in the heart of the vast Cosmic Voids between galaxies, galaxy clusters, walls, and filaments, as these voids continue to expand.
Second way: From General Relativity: Lee Smolin has discussed the reproduction of universes on the inside of black holes. His 'Cosmological Natural Selection' theory, and the related 'Black Hole Cosmology' are not commonly accepted General Relativity solutions. Standard General Relativity solutions lead to intractable singularities at the heart of Black Holes, but various transient high-energy processes that lead to them, such as Super-Nova explosions, Neutron Star mergers, etc., concentrate matter and energy into high densities, within which conditions found at the earliest times of our universe might be replicated. This is a generalization rather than a theory, but if Quantum Mechanics is then invoked to describe what is possible within such high-density matter, the probability of quantum tunneling producing a 'false vacuum'-like (inflation) state should steadily increase as the energy density increases.
Third Way: Technology. 'Creating a baby universe in a test tube.' In general, there is no theoretical prohibition to generating the kind of high energy density that black holes produce within 'normal' space, i.e., not beyond the event horizon of a black hole. This is the thought experiment discussed in the 1990 Nuclear Physics B article cited in the caption to the introductory image at the top of this post—Creation of a child universe in the laboratory or in a particle accelerator. We do not possess the technology to achieve these conditions, but there seems no theoretical barrier. It boils down to an engineering problem, and the successful creation of a child universe should produce a detectable signature. Perhaps the application of metamaterials can bring the needed conditions closer to achievability in the less-remote future.
Fourth way: 'Mitosis'. This covers any number of related concepts, most of which would benefit from rejecting the Standard Model of Cosmology's hypotheses that the universe would appear the same as we see it no matter where we happened to be in it, that is, perfectly flat, isotropic and homogeneous. The unexplained phenomenon known as 'Dark Energy' is effectively self-replicating. It creates new space and that new space contains the same amount of dark energy as pre-existing space. As of about 4 billion years ago, so much dark energy has been created, that it has begun causing the universe to expand at an accelerating rate. If Dark Energy is constant, as the Standard Model assumes, the result is that parts of the universe that are currently visible to us will eventually be receding from us at greater than the speed of light and will thus causally separate from us. These separated parts then evolve completely independently. If there is any mechanism for 'mutation', by which the laws of physics in these separated regions could diverge from each other, then 'interesting things' such as selection processes could happen. Further, if dark energy is variable in time, which the latest results from the Dark Energy Survey Instrument studies suggest, it could also be variable in space. Various forms of Quintessence Theory, in which Dark Energy is hypothesized to be a separate force/particle, give it this ability. This is mere speculation, but areas where Quintessence clumps (if it does) would yield more rapid expansion, and would, through time, create voids that would separate individual 'cells' and cell-connecting filaments organized in a manner similar to the structure of galaxies and galaxy clusters in our observable universe. Note that a sufficient concentration of Quintessence could behave like inflation, perhaps initiating a new universe on its own after a manner similar to the 'universe in a test tube' mechanism just discussed. Unification of Dark Energy and Cosmic Inflation is a topic of active research, potentially via Quintessence or other aspects of the String Theory Landscape of diverse vacuum states and a set of conjectures known as the Swampland.
Fifth way: Many Worlds. The so-called Everett interpretation of Quantum Mechanics taken to be a truly physical process. The Many Worlds interpretation seems to be gaining favor among quantum physicists largely because it stays truest or least invasive to the formalism of the theory. What annoys me about it is the compounded, seemingly inexhaustible fecundity of a new universe being required to form each time any particle 'encounters' another—the "ontological extravagance" as David Wallace recently put it. The basic concept of this interpretation is that each physical interaction that 'forces' a quantum superposition to actualize into a specific state, causes a split in the timeline (e.g. into two separate Worldlines), such that the original superposition is preserved across the combined Worldlines. For the purposes of Universe Self-Replication, it seems useful to hypothesize one or more thresholds or 'gatekeepers' that contains/restricts this overwhelming fecundity. Foremost among those is a distinction between Strong and Electroweak interactions. A further form of threshold that might be considered is the bound beyond which perturbation theory can no longer be applied (i.e., where the excitation being studied mathematically develops too great of an amplitude).
In the simplest form of this mechanism, Strong Interactions are ignored because all Worldlines that are created are on the scale of femtometers (the scale of protons and neutrons—the effective limit of the range of the Strong Nuclear Force). In the case of Electromagnetic and Weak interactions, each virtual particle (or more generally, each excitation of the quantum field) that is generated during an interaction between real particles is automatically endowed with an 'inner life' that is the complete state of the universe that spawned it (a Worldline is created). Each such actualization is thus imbued with the current laws of physics and values of constants, etc., in force at the time of the interaction or excitation but also subject to mutations associated with the 'dressing' of vacuum excitations that envelop the interaction.
Inherent in this mechanism is the assumption that our universe itself amounts to nothing more than one more quantum excitation—just another virtual particle, albeit with a stunningly elaborate 'inner life'—within the greater 'frothing sea' of the vacuum. How it achieved that astounding complexity is at the crux of this discussion. As said, the 'dressing' of vacuum fluctuations (interference from fluctuations of other fields during the particle interaction) provide the opportunity for the newly emergent world to be 'born' with effective mutations. If so, then we have all the processes needed for Darwinian evolution, lacking only sexual reproduction.
An adjunct or extension to this view includes two more difficult questions/considerations: first, the vacuum fluctuations that 'dress' real particles at all times, even when no obvious interactions are taking place, and second, the virtual entities that make up the so-called quantum foam, which pervades space even in the complete absence of real particles. Do these excitations also represent new worlds when they appear? Given that no 'observation' has taken place (with 'observation' defined in the minimalist sense of an interaction—any event where one real particle is affected by another—or, alternatively, defined by some further unknown threshold criterion that governs interactions, if any), I suspect that these virtual entities are representative of the general vacuum, which I formally define in a manner consistent with the simple mechanism described above, as the complete 'frothing sea' within which our universe is merely one virtual entity. This extension might speculatively suggest that the some or all vacuum excitations in general population of this 'greater sea' may also be 'other worlds', not necessarily closely 'related' to our own. The vacuum, then, would be the virtual venue where inter-universe interactions—an instantaneous exchange of influence between universes—may take place. This adjunct mechanism—a physical process that is ubiquitous—is here taken to its logical end (or extreme), where it points to the possibility of a form of 'sex' between universes.
Sixth way: Simulation. Current computer power allows scientists to simulate selected physical systems in great detail, but such calculations fall far short of being capable of simulating processes in sufficient detail to address some of the fundamental questions that we do not yet have answers to, such as how life emerged and how self-aware 'intelligent' beings emerged. But there does not seem to be any obstacle that would prevent us from someday being capable of answering such deep questions via simulation. If future technology makes such robust simulations possible, and because our known universe contains only a finite amount of information, there is no known physical or philosophical reason to expect that we could not someday create a simulated subset of our universe (say, of the human brain and all its sensory inputs) that could pass any pre-established arbitrarily test of 'adequacy'. The first such simulation is likely to focus on an individual 'AI being' that will surpass human intelligence and will profess to be conscious, and this may not be more than a few years or decades away.
Simulations can be analog models as well as digitized mathematical ones. The creation of physical analog models doesn't seem to get the attention or volume of research work that digital simulations do. But there is an overlap in the area of life's genetic code. Experimental efforts to produce synthetic life starting with inorganic chemicals is a primitive example. There seems no barrier to extrapolating such work toward far more complex systems, perhaps creating an Artificial Biological Intelligence that could rival human intelligence and perhaps exceed the self-replication capability of any digital form of Artificial General Intelligence. This is not a projected area of thought that seems to be given much attention, but perhaps it should.
Much existing work discusses the possibility that our universe, as a whole, is a simulation. There is certainly no law that limits the amount of information that could be contained in a universe. A simple thought experiment frequently posed is the idea that our universe is merely a video game or school science project of a kid (who might have 6 x 10^80 bits of information—the total amount in our observable universe—in a single eyelash), living in some advanced civilization in a universe with far greater information content/density and complexity than ours. The simulated 'toy universe' could represent a physically simplified version of the progenitor universe. Perhaps only in that higher-order realm could the physical conundrums that we face (the myriad unsolved problems in physics) be correctly resolved.
Simulated universes have some interesting consequences for the world's religious communities. For example, simulations that are intentionally created clearly have a creator, which could be thought of as a true, real 'God.' You and I could also have an effectively real, truly immortal afterlife as an archived dataset containing all data describing our mind and body at any point in our lives. At the will of the 'creators', these simulations could be re-activated at any time.
Question 3: What are some consequences if universes like ours can self-replicate?
Answer: Analysis of the potential consequences of this cosmology have been sparse. Three of these consequences are of particular note:
First: Our universe came from a 'preceding' parent universe. If self-replication is occurring, then it is nearly certain that it is the explanation of the origin of the particular universe that we observe. Our universe, with all its complexity, was not original - not cut from whole cloth, so to speak. If self-replication is accompanied by mutations, then theoretical models that have been criticized for being unfalsifiable because they can predict nearly any outcome, such as Quintessence and String Theory, can play a role in a manner analogous to the way models of cell biology can lead toward the 'prediction' of the particular evolutionary outcome that is the human species. In Cosmology, we do not have the benefit of studying the examples of other 'species'. We can observe only our one example. But perhaps we have a more grounded theoretical basis than models of cell biology provide life scientists. We can, at least pursue this hope/expectation.
Second: The 'Fine Tuning' question: In the cases where self-replication occurs in the 'open' universe, i.e., not inside the event horizon of a black hole, and particularly in any of the set of scenarios where intelligent beings (or even simple forms of life) are present to interact with or direct the replication process, there is a potential for the laws of physics in the offspring universes to be skewed favorably toward the living forms involved. This could provide a natural explanation of what has been called the 'Fine Tuned Universe' conjecture or hypothesis or question, which asks why or how our universe seems to be so well-suited for the development of life, when such an outcome is estimated (by the majority of physicists) to have a prohibitively low probability within the range of possible sets of governing laws and constants.
An intriguing consequence of the 'Universe in a Test Tube' replication mechanism (and also of the whole-universe-as-a-simulation concept) is its implication for the Fermi Paradox, which asks "Where are the intelligent aliens?" Advanced civilizations, which are capable of creating child universes (say, the way we create Higgs Bosons in our accelerators, or by creating exceedingly cold, empty vacuum states) do not have to exist in our present universe, only in an 'ancestor' universe. Perhaps there is some 'Great Filter' that makes appearance of stable advanced intelligence extremely rare, but given a vast genealogy of precursor universes, the odds of one successful emergence rise dramatically.
Third: Evolutionary processes (mutation and some form of selection), analogous to those that pertain to the development of complex life forms from simple ones, if operative during universe self-replication, would push back the difficult question of the ultimate origin of reality (e.g., the mechanism or process that produced the "Big Bang") into a veil of obscurity that might be even more difficult to unravel than the question of abiogenesis—the means by which life emerged from the approximately 92 natural elements of the Periodic Table, as generated by cosmic and/or stellar nucleosynthesis.
The example of the development of photosynthesis in early cyanobacteria, and its consequent Great Oxidation Event is good stark example. New chemical pathways have emerged that did not occur in the evolutionary process that led to them, and the earlier forms of anaerobic life, to which oxygen was poisonous, died out, leaving little trace of the chemical pathways from which these life forms originated. If there is any analog between the evolution of life and a not-prohibited process of universe self-replication with mutations, and Stephen Hawking and Thomas Hertog have proposed that there is, then the questions of emergent physics and lost physics become relevant. Quantum Mechanics and General Relativity in the form that we observe them could, conceivably, not have directly participated in the physical processes that were extant in simpler ancestor/precursor universes.
 |
| Stephen W. Hawking (8 January 1942 – 14 March 2018) NASA photo from 21 April 2008. |
Hawking and Hartog's 'Top-Down Cosmology' (see also Phys. Rev. D73 (2006) 123527) echoes this perspective. It confronts the certainty that we do not understand the initial conditions that lead to our universe by positing a 'no initial boundary condition' theory that incorporates the string theory landscape and allows us to decipher a series of selection events and quantum accidents that lead to the late time (present-day) boundary conditions that we observe. These selection events and accidents can be viewed as a series of symmetry-breaking processes that "describe the beginning of a new, independent universe with a completely self-contained ‘no boundary’ description" with the previous state becoming irrelevant. They are silent about whether each of these 'beginnings' could represent 'genealogy with mutation' self-replication events in which a parent universe gives birth to a child with new traits; and they would probably argue that it does not matter, since we can only observe our one universe. It does matter, however, if that (unobservable) process carries explanatory power. Nearly all of this evolution appears to have taken place in a tiny fraction of a second, as we observe the resultant effects. An ancestral line of precursor universes provides a more natural time frame for all these improbable events to have occurred. If our universe is not the simplest self-sustaining and self-replicating model that its physical laws allow, and it certainly appears to be far from it, then the biological analog seems favored. This does not rise to the level of a testable hypothesis unless or until some observable consequence is identified. On the other hand, the biological analog suggests an entirely different perspective, not often discussed in the physics community, in which it might be rigorously provable (philosophically if not mathematically—e.g., advancements in the theory of complex adaptive systems), that the observed outcome (our universe) is not a predictable result of precursor conditions, any more than the human species is a predictable consequence of cell biology.
Pursuing that analogy to life processes further, we conclude with a thought experiment: Imagine that our universe's Hubble volume (meaning everything that the speed of light allows us to observe) lies within one single neuron cell in a grand, diverse organism such as the human being—just one of thirty-seven trillion cells in this 'multiverse'. In the development of a human being from its original stem cell to mature adult, there were multiple cell-differentiation events that produced the observed attributes of the neuron, and the multitude of steps along that pathway represent symmetry-breaking events where specific genetic instructions are activated and others de-activated, radically changing the final cell's function and morphology. Note well that the analogy is to cell differentiation only (or to the process evolutionary development). A living cell is not a self-contained system as the universe seems to be. It requires input of nutrients and oxygen and export of waste products. It is rank speculation (and not parsimonious based on our current best effective theories) to suppose that our universe could be part of some greater structure, let along some sort of functioning organism.
Research Questions and some Foundational Questions (this section is under construction):
1. What can be said about entities that lie outside of the observable universe? This discussion equates 'the universe' with that region we can now observe, which leaves open the term 'multiverse' as a valid concept ...
2. Is there a protocol for quantifying the probability of the existence of other universes (the various manifestations of the concept of the multiverse)?
3. If the multiverse exists, does the Copernican Principle apply? In other words, what is the validity of the assertion that our universe is nothing particularly special? What are the implications of that application of the Principle?
4. What is the simplest theoretical non-trivial self-sustaining and self-replicating universe? Is our universe that simplest entity, or, if not, what can be said about the relationship between our universe and those simpler constructions?
5. Using Complex Adaptive Systems theory or Replicator Theory (or other protocol), can it be rigorously proven whether or not the outcomes are predictable based on initial boundary conditions? Conversely, is it possible to retrieve initial boundary conditions from observed current state of Complex Adaptive Systems such as the human being?
6. Can the answers to 5 be applied to our observable universe? Is Complex Adaptive Systems Theory applicable to the Hawking-Hertog top-down cosmology proposal?
7. Extending the parable of the 'Blind men and the elephant' to cell biology, what could be learned about cell biology using nothing but the internal behavior of a single cell. Starting with several different kinds of living cells, from an archaeon to a human neuron, from a cell in a banana leaf to the eye of a fruit fly, etc., how would the information that is gleaned about general cell biology be different? How much of what is concluded would be incorrect? (Conducting this 'blind' study would require great care that the results are not contaminated by the researcher's bias based on knowledge about patterns of similarity between different cell types. The purpose is to apply Hawking-Hartle 'top-down' methods to cell biology.)
