
The First Selection Among Universes is Geometric
November 19, 2026Free Audiovisual Version Here

By C. Rich
The last time we visited the Cosmic Garden, we discovered something that would probably annoy anyone hoping the universe could be explained with a perfect sphere and a sufficiently impressive collection of Greek letters. We found that the gravitational destiny of a primordial universe could depend on how its geometry was moving. Two universes could possess the same amount of shape motion, yet the boundary separating expansion from collapse would occupy different positions depending on whether their shapes were rotating or oscillating back and forth. Apparently, even the universe has to worry about its dance moves, although the consequences of getting them wrong are considerably more serious than stepping on somebody’s toes.
That discovery gave me the foundation for Pillar 3, but it also left an obvious problem. I had been conducting my investigation in a universe containing pressureless matter, which cosmologists rather charmingly call dust. Dust is a perfectly legitimate ingredient in theoretical cosmology, but the early universe was also filled with radiation. If the Cosmic Garden was ever going to develop into a serious investigation of how our universe might have emerged from a primordial geometric state, I needed to find out whether my discovery survived when radiation entered the picture. There is little point in designing an elegant mathematical doorway into the universe if the entire contraption falls apart the moment somebody turns on the lights.
Heat is particularly interesting because of the way it influences physical structure. We see this throughout nature. Ice crystals lose their intricate architecture when they melt, magnetic materials can lose their ordered alignment when heated, and substances that display elaborate structures at low temperatures can behave very differently when thermal energy becomes important. The details depend on the particular physics, of course, but the broader question was irresistible. If the shape of a primordial universe helps determine its gravitational future, could radiation alter that relationship or eliminate it altogether?
I decided to gather my team of AIs called The Mash and find out. The first task was to construct a family of mathematically consistent primordial universes containing different combinations of dust and radiation. I began with the pressureless model from Cosmic Garden’s Pillar 3 and gradually increased the radiation contribution until I reached a configuration containing radiation but no dust at all. Each member of this family had to satisfy Einstein’s gravitational equations while remaining in the balanced, static condition I was investigating. I could not simply pour radiation into the equations like gasoline into a lawn mower and assume the machine would continue running.
This produced a continuous family of Einstein-static configurations extending from pure dust to pure radiation. At the radiation endpoint, the radiation energy density and vacuum energy density were equal. That balance would become important later, especially when I started pondering the relationship between the Cosmic Garden and the Hot Big Bang. With the family established, we could examine what happened to the geometric effect discovered in Pillar 3. I already knew that the distinction between rotating and linear shape motion influenced the critical gravitational boundary. The question was whether that distinction would survive as radiation became increasingly important.
The mathematics produced an intriguing answer. The polarization-dependent contribution remained present and retained its sign at every one of the eleven configurations we sampled along the dust-to-radiation family. I then derived exact results at the pure-radiation endpoint and at a mixed dust-and-radiation configuration. The effect became smaller in absolute magnitude as radiation increased, but something interesting happened when I compared it with the contribution from the overall amount of shape motion. Radiation reduced that contribution even faster. This was cool stuff.
In other words, the polarization effect became smaller in one sense while becoming relatively more important in another. That was not quite the outcome someone might expect if they imagined radiation simply washing away the geometric distinctions I had found. At the pure-radiation endpoint, the magnitude of the polarization coefficient had fallen by approximately 46 percent compared with the dust model. Yet its relative importance had increased. The relationship was subtle, and it was precisely the sort of thing that justified doing the mathematics instead of relying on a good analogy and a dangerous amount of confidence.
I had another reason to be encouraged. The analytical calculations were conducted with independent checks. In one particularly useful exercise, Astra derived the pure-radiation result without seeing the values generated by an earlier numerical exploration. When the results were compared, the independently derived quantities agreed with the withheld values to every printed digit. I also had the midpoint calculations independently verified. That gave me a mathematical prediction I could put in front of a computer and challenge. The prediction concerned the location of the gravitational dividing line between two different futures. I would construct universes with carefully matched amounts of geometric motion, change the polarization of that motion, and measure how the critical radiation content shifted. I would repeat the procedure at different amplitudes, using increasingly small disturbances, and see whether the results approached the values predicted by the mathematics.
I froze the predictions before running the experiment. This is one of the disciplines I have come to appreciate most about doing theoretical research with AI. It is remarkably easy to produce a beautiful explanation after seeing what a computer has done. The much harder task is to write down what you expect before the computer has had an opportunity to embarrass you. As it turned out, the computer was perfectly willing to oblige.
Our first major numerical experiment was designed to examine thirty-two combinations of background conditions, amplitudes, and polarization states. The program began working through them, and the initial results looked promising. But after ten scans, it stopped. Both of our background configurations had encountered the same problem, and the experiment ended with the official verdict INCONCLUSIVE.
There are few words in scientific research quite as irritating as inconclusive, particularly after you have spent a considerable amount of time constructing what you thought was a perfectly reasonable test. That shit can piss you off. It is the mathematical equivalent of calling a contractor to find out whether your house is finished and being informed that the building inspector has discovered an interesting philosophical question about doors. But the reason for the failure turned out to be one of the most interesting discoveries in the entire investigation.
My testing procedure contained a safeguard intended to ensure that a certain mathematical quantity changed consistently as we searched for the boundary between expansion and collapse. That sounds reasonable enough. Unfortunately, I discovered that the rotating geometry possessed a symmetry that placed its critical boundary precisely at a maximum along the path I was measuring. Imagine walking up a hill to locate a particular elevation. Your measuring procedure assumes that every step carries you either consistently upward or consistently downward. That assumption works beautifully on a hillside. It becomes a problem when the very thing you are trying to locate is the summit. My rotating universe had placed its gravitational boundary on the hilltop.
This was not a random numerical accident. The geometry possessed a particular symmetry involving time reversal and reflection. In mathematical terms, that symmetry tied together two different aspects of the gravitational boundary and placed the rotating family’s critical value at the maximum of the release path. The safeguard I had written to protect the experiment was incompatible with that structure. The first, smaller-amplitude case had slipped through the safeguard almost by chance. At the next amplitude, the program correctly enforced its instructions and stopped.
There is something rather satisfying about discovering that a computer failed because the mathematics was more interesting than the assumptions I had made about it. I would still prefer to discover such things before launching an experiment, but theoretical cosmology, as I have grown to learn, has apparently declined to organize itself around my convenience. The nerve, right? I did not change the verdict of the first experiment. It remains INCONCLUSIVE. The stopped run was preserved, the reason was investigated, and the correction was documented. The analytical predictions were also left untouched.
I then designed a follow-up experiment, openly acknowledging that it was constructed after examining the first run’s partial results. The rotating configurations would be examined using the hilltop property implied by their symmetry, while the linear configurations would retain the original safeguards. I also wanted to investigate another issue the first experiment had revealed. At the smallest original amplitude, the numerical measurements differed from the theoretical predictions by approximately two to three percent.
The direction of the shift matched what I had predicted, but the magnitude was not as close as our original tolerance required. This presented a straightforward question. Was I seeing a genuine disagreement with the mathematical prediction, or was I observing additional effects that became less important as the geometric disturbances grew smaller? The next experiment was designed to answer that question.
I had the Mash investigate four amplitudes, progressively reducing the size of the geometric motion, and watched what happened to the discrepancy. Before the new run, we also recorded a specific expectation for how the numerical results might approach the analytical values. That expectation came from the earlier partial data, so it was a follow-up hypothesis rather than an independent prediction from first principles. This time, the experiment completed all thirty-two scans.
My office computer spent approximately one hour and fifty-six minutes carrying out the work. Across the full investigation, the numerical audit recorded 5,696 integrations. Two different numerical solvers were employed, additional calculations were performed at tighter accuracy settings, and the results were checked for violations of the gravitational constraint. The independent review of the completed records found no solver disagreements, no unresolved trajectory classifications, and no constraint failures. The largest recorded constraint errors were vastly smaller than the permitted limits. The controls also behaved as expected, including the agreement between opposite senses of rotation, which the mathematical symmetry required.
Most importantly, the discrepancy between the numerical measurements and the analytical predictions decreased as the amplitude became smaller. At the pure-radiation endpoint, the extrapolated result differed from my frozen mathematical prediction by approximately 0.06 percent. At the mixed dust-and-radiation configuration, the difference was approximately 0.04 percent. Both were comfortably inside the tolerance I had established before running the follow-up experiment.
The numerical pattern also suggested that the leading remaining discrepancy was associated with a higher-order effect whose influence diminished with amplitude. I could measure that behavior and estimate its size, although its coefficient had not been independently derived from the underlying mathematics. Under the frozen criteria of the follow-up experiment, both backgrounds passed.
That was an important result for the Cosmic Garden. I had extended the geometric selection effect from the original dust model into a family containing radiation, carried the calculation all the way to the pure-radiation endpoint, derived exact analytical predictions, and obtained numerical convergence toward those predictions in two independently examined backgrounds.
The result also clarified something about the nature of the geometric motion itself. What mattered was the polarization, meaning the distinction between rotating and back-and-forth oscillations. I had to learn that phrase too. That word always had a political connotation to it when I thought of it. Anyway, it was not a preference for clockwise motion over counterclockwise motion. Those opposite rotating senses remained equivalent in our calculation. The gravitational boundary responded to the form of the motion, not to an arbitrary preference for which way the universe happened to turn.
That distinction may sound minor to someone unfamiliar with the mathematics, but it is fundamental to what I actually discovered. Nature has enough mysteries without me inventing additional ones through careless descriptions. There is a larger scientific tradition behind this investigation, and it deserves recognition. Einstein gave us the gravitational framework. The mathematics of Bianchi cosmologies provided a way to investigate universes whose geometry need not behave identically in every direction. Researchers such as John Barrow and Kei Yamamoto investigated the instabilities of Einstein-static configurations, while George Ellis and Roy Maartens developed the Emergent Universe approach, exploring how an initially static cosmos might transition into an expanding one. These men are giants.
Andrew Zardecki had already investigated anisotropy, inflation, and reheating in Bianchi IX cosmology back in 1985, when I was considerably more interested in the music of the hard rock band Triumph than the gravitational behavior of primordial universes. Pierre-Philippe Dechant, Anthony Lasenby, and Michael Hobson subsequently examined an anisotropic early universe evolving toward inflation. These researchers were exploring important parts of the landscape long before I arrived, playing in a scientist’s sandbox uninvited with my particular questions.
My contribution is narrower and more specific. I have been investigating how matched amounts of primordial geometric motion, arranged into different polarization states, shift the critical boundary between gravitational outcomes. Cosmic Garden’s Pillar 4 establishes that this distinction survives the introduction of radiation within the mathematical regime we examined. And that brings us to an observation that has been occupying my thoughts throughout this work. I have been talking about radiation, temperature, primordial geometry, and the universe emerging from the Garden. Naturally, I began wondering whether we might be approaching the moment traditionally associated with the Big Bang itself.
I have never been particularly satisfied with the assumption that the Hot Big Bang must represent the absolute beginning of everything. The observable universe unquestionably passed through an extraordinarily hot, dense early phase. That is one of the great achievements of modern cosmology. But describing that phase does not automatically tell us whether anything preceded it, or whether the physical conditions responsible for it had an earlier history. I think it does. I think it has a geometric memory, but what I think and what I can prove is a painful slog.
The Cosmic Garden that comes from my theory, Cosmological Pangaea, began with the possibility that a primordial geometric configuration could exist before our universe entered its familiar expanding history. I am investigating whether that geometry possesses lawful gravitational exits. Now that radiation has entered my mathematics, the temptation is to imagine an extraordinary transition in which the object leaves its primordial state and erupts into a hot, expanding universe. It is a compelling picture. Unfortunately, Einstein’s equations have an irritating habit of requiring an actual mechanism before they agree to perform the scene.
The radiation in my present model is already there. I have introduced it as a conserved component of the universe, not generated it by converting some other form of energy into heat. I have also retained a genuine cosmological constant, meaning the vacuum energy does not spontaneously disappear when it becomes inconvenient to my cosmological ambitions. As the universe expands, the radiation becomes more dilute while the constant vacuum energy remains unchanged. This means my present model does not yet contain the transition that would end a vacuum-dominated phase and produce the hot radiation-dominated universe associated with the conventional Hot Big Bang history.
That is what I am hunting next. That is a separate physical problem, and an important one. Ellis and his collaborators explored ways in which scalar fields could connect an Einstein-static state to inflation and a subsequent hot cosmological era. Their work provides one of the foundations for thinking about such a transition. Other researchers have examined how anisotropic geometry might influence inflation, scalar fields, and the particles produced during the early universe.
What interests me is whether the particular geometric distinctions I have now established could influence such a transition. Suppose two primordial configurations possessed matched amounts of geometric motion but different polarizations. I already know that their critical gravitational boundaries differ. Could those differences also influence the later evolution of a field responsible for ending vacuum domination? Could the primordial geometry affect when energy becomes radiation, or whether some trace of the original shape survives into the universe we observe?
That is big game hunting in the world of theoretical physics. Those questions remain open, but they are no longer disconnected from the mathematical work I have completed. The Cosmic Garden has now reached an interesting position. I began by asking whether primordial geometry could influence gravitational fate. I found a distinction between rotating and linear shape motion, and I have now shown that radiation does not erase that distinction near the Einstein-static state. I have also encountered a genuine symmetry that explained why my first numerical procedure failed, and the corrected follow-up has strengthened the agreement between analytical predictions and numerical evolution.
This does not yet establish that our universe emerged from such a Garden. What it establishes is that the mathematical doorway I have been investigating retains its polarization-dependent structure even when radiation is present. I am knocking on that door. For me, that is a meaningful advance because it brings the investigation closer to the conditions any proposed primordial history must eventually confront. Our universe was not merely a collection of abstract shapes following elegant equations. It became a hot, expanding physical environment in which matter, radiation, gravity, and eventually galaxies participated in an extraordinary sequence of transformations. I am still examining the doorway, but I am beginning to understand more about what can pass through it.
And after spending a fair amount of time worrying that adding heat might destroy the geometric distinction The Mash and I discovered, I find it rather satisfying that the mathematics delivered a different answer. The Cosmic Garden survived our attempt to turn up the heat. We even discovered that our supposedly troublesome hilltop was exactly where the rotating universe had placed its gravitational boundary.
The next great question is whether the geometry that helps determine how a universe exits its primordial state can also influence the hot cosmic history that follows. Somewhere between that doorway and the distant galaxies measured by Hubble and his successors lies the physical story I am trying to reconstruct.
I have not reached those galaxies yet, but I have learned something worth carrying along: even in the presence of radiation, the way a primordial universe moves through its geometry matters to the gravitational threshold it encounters. Apparently, the universe remembers enough of its choreography to keep the mathematics interesting, even after I turn up the heat. What is that saying? Oh yeah, if you can’t stand the heat, get out of the kitchen. Well, I grew up on the streets and beaches of South Florida; heat just flows off my shoulders.
C. Rich




