Chapter 13bThe Cosmic Web — Voids, Sheets, and Coordination

“We shape clay into a pot, but it is the emptiness inside that holds whatever we want. We hammer wood for a house, but it is the inner space that makes it livable. We work with being, but non-being is what we use.” — Lao Tzu, Tao Te Ching, chapter 11 (author’s rendering)

I. The Largest Structures Are Empty

Strip away every galaxy, every star, every atom of visible matter, and you still have most of the universe left over. The largest structures in the cosmos are the absences between the things we can see. Voids between galaxies are where dark energy, initial conditions, and the thermodynamics of structure formation stand most exposed.

Typical cosmic voids span 100 to 400 million light-years, though rare supervoids extend far larger. Our Milky Way measures roughly 100,000 light-years; a single void could swallow a thousand Milky Ways laid end to end.

First identified in the 1970s by Gregory and Thompson,⁷ these regions of exceptionally low matter density are essential components of the cosmic web. The Boötes Void, one of the largest, spans 330 million light-years⁸ and is informally called “the Great Nothing.”

How can “nothing” be a structure? How can emptiness be the largest thing in the universe?

Western metaphysics offers little help here. From Parmenides (“Being is; non-being is not”) through substance metaphysics (the tradition treating concrete things as fundamental), existence is primary and absence derivative. The word “void” itself suggests privation.

Cosmic voids are structure of a different kind: negative space where dark energy (whatever is accelerating cosmic expansion) is most visible. The universe is made of emptiness and fullness together; each gives shape to the other.

II. The Tao of the Cosmic Web

Lao Tzu understood that emptiness is functional space: the enabling condition for everything within it. The pot’s usefulness lies in its hollowness. Non-being is essential to being.

Buddhist śūnyatā (emptiness) goes further: emptiness is the nature of all phenomena, the lack of inherent, independent existence. Form is emptiness; emptiness is form.

Cosmic voids embody this at the largest scale: an arrangement that enables the structures at their boundaries to exist. In the cosmic web, the density contrast between filaments and surrounding voids helps channel matter toward galaxy-forming regions. Matter flows “downhill” from void to filament, like water draining from a plateau into a valley. The voids are part of the web’s topology, shaping the contrast through which matter flows.

III. Entropy’s Spatial Expression

Emptiness is not the same as disorder: whether a void holds more or less entropy than a cluster depends on how hot its gas is and what it has been through, not on how empty it looks. Cosmic voids express a spatial density contrast. Classifying their thermodynamic entropy requires temperature, mass, and thermal history; dilution alone supplies no ordering.

As matter flows toward denser filaments and clusters, global entropy increases. Radiative cooling can lower the entropy of gas while exporting entropy into its surroundings. As matter falls together it heats and shines; the balance depends on temperature, density, and thermal history, rather than structure alone. Think of a refrigerator: the cold interior comes at the cost of hot exhaust coils on the back. For a star, the coils are its own light, poured into the dark.

Voids, filaments, and clusters divide the cosmic web by density and morphology. Their thermodynamic entropies cannot be ranked from those labels alone; hot intracluster gas is a particularly important part of the accounting.

Two kinds of entropy are at work here, and the distinction matters for everything that follows. For gravitating systems, the statistics flip: clumping is the high-entropy direction. This sounds contradictory.

A rough analogy: thermodynamic entropy is a drop of ink dispersing through a glass of water, spreading until every region is the same pale shade. Gravitational entropy runs the opposite way, like a snowball rolling downhill, gathering mass as it goes, growing precisely because it is already large. One process scatters; the other concentrates. The cosmic web is shaped by both at once.

Roger Penrose, the Nobel laureate in physics, identified the paradox. For an ideal gas in a box, the highest-entropy state is uniform distribution: particles spread evenly, gradients erased. Gravity reverses the picture.

When matter attracts other matter, a uniform spread can be unstable on scales where gravity overcomes pressure or velocity dispersion. Perturbations on those scales can grow into clumps that pull in more matter. The early universe’s near-uniformity represented extraordinary gravitational order: a coiled spring waiting to release.

The result: clumping raises gravitational entropy globally while forming visible structure; the local thermodynamic entropy change requires separate accounting.

The cosmic web combines expanding underdense regions with condensing filaments. Volume or density alone does not determine thermodynamic entropy: a reversible adiabatic gas can expand with no entropy increase. Gas temperature, mass normalization, thermal history, and the chosen gravitational coarse-graining must enter any entropy account of this structure.

The cosmic web is bilateral at the thermodynamic level: two complementary processes running in opposite directions, producing structure together.

The constructal reading treats the cosmic web as a candidate geometry for easing flow. A maximum-entropy-production claim (a formulation of the MEPP met in Chapter 4) would additionally need an admissible set of morphologies, fixed constraints, an entropy-production functional, and an optimization result. The Second Law supplies nonnegative production, without selecting its maximum. Stars and other dissipative structures need gradients to sustain their activity; that fact alone does not prove an optimal cosmic topology.

IV. The Constructal Cosmic Web

The emptiness of voids is active, shaped by the Constructal Law introduced in Chapter 3.

Adrian Bejan’s constructal theory holds that natural systems evolve to ease flow, the way a river delta branches into ever-finer channels. On Bejan’s reading, the cosmic web is flow finding its easiest routes at the largest scale.

The Constructal Law applies equally to channels and the spaces between them. In a river network, optimizing flow defines the interfluvial spaces (regions between rivers where water does not flow). These spaces emerge from the same optimization as the rivers themselves.

At cosmic scale, matter flows along filaments toward nodes (galaxy clusters). The voids are the interfluvial spaces: regions evacuated by gravity, where matter has drained toward denser structures. Simulations by Mark Neyrinck and others show voids and filaments emerging naturally from nearly uniform matter distributions. Guided by gravity and cosmic expansion, the process closely parallels river network formation under the Constructal Law.

The ZOBOV algorithm (ZOnes Bordering On Voidness) identifies void boundaries in galaxy surveys using topological methods that work equally well for watersheds.⁹ Scale invariance extends to void ancestry: small voids merge into larger ones, like tributaries feeding rivers. The Boötes Void is consistent with ΛCDM (Lambda-CDM, the standard cosmological model combining dark energy and cold dark matter) via hierarchical merging of smaller progenitor voids.¹⁷

Filaments do more than scaffold; they imprint. Of 93 quasars (intensely luminous galactic cores) surveyed, 19 showed significant polarization (a measurable directional bias in their light). Their rotation axes aligned with their host filaments, either parallel or perpendicular, across billions of light-years. The odds of such alignment arising by chance are roughly one percent.⁶ Perpendicular counts as alignment here. A random universe would scatter those axes across every angle; finding them settled on two particular angles, both fixed by the filament’s own direction, is the signature that the filament set them.

Read constructally, the explanation lies in the channel: gas within a filament moves and rotates coherently. Galaxies condensing within the same filament can inherit correlated angular momentum. An isolated system’s total is conserved, while individual galaxies and filaments exchange it through torques and matter flows. The channel carries matter to the nodes and tells that matter which way to spin.

The web shapes more than spin. A 2025 study of warped disk galaxies found that the warp itself tracks the cosmic web. The fraction of galaxies with warped disks climbs as they approach a filament, within a few megaparsecs (a megaparsec is about three million light-years). The more common warps are S-shaped: one edge of the disk lifts while the opposite edge dips, tracing the shallow S of an integral sign. Satellites orbiting those galaxies line up with the nearest filament rather than scattering at random.⁶⁰⁶ The same large-scale tide that tells a galaxy which way to spin also bends the plane it spins in.

In 2024, the flow became directly visible. The eROSITA X-ray space telescope, cataloging nearly 900,000 X-ray sources in its first six months, imaged cosmic-web filaments in X-rays across the whole sky.² The haul exceeded the combined 25-year discoveries of earlier X-ray telescopes Chandra and XMM-Newton.

The largest single detection was a bridge of hot gas between two galaxy clusters, Abell 3667 and Abell 3651, spanning 42 million light-years in projection. Projection is all the sky can show, a length flattened onto two dimensions, which foreshortens any structure tilted toward us. The true three-dimensional length, once redshifts along the bridge supply the missing depth, is closer to 80 to 100 million light-years. The bridge contains roughly 270 trillion solar masses of gas, heated to 10.6 million degrees. The mass estimate carries broad uncertainty, spanning perhaps 190 to 410 trillion solar masses. It was the first individual filament detected extending beyond three times the virial radius (the gravitational boundary) of both clusters, at 11-sigma significance. At that confidence the detection itself is not marginal, whatever the systematics do to the mass figure.

The filament is anomalously enriched: denser and more X-ray-luminous than models predict. Flow channels concentrate matter beyond what gravitational evacuation alone would produce. The channel deepens the river.

The cosmic web is no longer an abstraction. You can see the current.

A complementary 2024 result arrived from the Subaru Telescope’s Hyper Suprime-Cam, which used weak gravitational lensing (detecting the subtle bending of light by invisible mass) to find dark matter filaments attached to the Coma Cluster. This was the first confirmation through lensing alone, without gas emission or absorption.³

Where eROSITA saw hot baryonic gas (ordinary matter), Subaru’s lensing revealed the dark matter that defines the channels: the invisible banks of the cosmic river. Two methods, different substances, different physics, yet the same constructal architecture.

A third confirmation came from statistical stacking, which combines many faint signals to reveal a pattern too weak for any single observation. Merging eROSITA data from 7,817 filaments across four all-sky surveys yielded a 5.4-sigma detection of X-ray emission from the warm-hot intergalactic medium (diffuse gas between galaxies, heated to 100,000 to 10 million degrees).⁴

The finding addresses a major puzzle. The early universe’s census of ordinary matter (baryons) predicts more than astronomers had ever found in stars, galaxies, and clusters. A substantial fraction of these missing baryons appear to reside in filaments, too hot for optical telescopes and too diffuse for previous X-ray instruments.

The constructal picture remained incomplete. Filaments are one-dimensional channels; voids are three-dimensional reservoirs. Counting dimensions here means counting the directions in which a structure is large: a filament runs tens of times longer than it is wide, so it behaves like a line, while a void is vast in all three directions at once. The intermediate geometry, sheets serving as drainage basins on which cosmic rivers run, required a different discovery.

V. The Sheet Nobody Expected

For decades, the standard view held that the Local Group’s mass sits in a roughly spherical distribution several million light-years across, containing our galaxy, Andromeda, their satellites, gas, and the gravitational scaffolding that holds the group together.

Clean, intuitive, and incorrect.

In January 2026, Wempe, Helmi, and collaborators published in Nature Astronomy a study that upended this model.²⁶ Their tool was BORG (Bayesian Origin Reconstruction from Galaxies), a technique that infers where dark matter sits by working backward from observed galaxy positions. Running 169 resimulations of the local cosmic environment, they found that a spherical halo fit nothing. Galaxy motions diverged, galaxies receded at the wrong speeds, satellite distributions made no sense.

When they changed the shape, everything fell into place. Modeled as a flattened sheet stretching 30 million light-years, the data converged: peculiar velocities (galaxy motions relative to overall cosmic expansion), satellite arrangements, and Local Group dynamics all became consistent.

The sheet’s central plane is roughly twice cosmic average density. The regions above and below it are near-empty voids. Combined Local Group halo mass: approximately 3.3 ± 0.6 trillion solar masses.

The sheet geometry also resolved a long-standing discrepancy: nearby galaxies recede faster and more smoothly than a Local Group this massive should allow, since that much mass ought to be dragging them back. Mass distributed farther out in the plane pulls outward and partially offsets the inward force, so recession is braked less than a sphere would predict. The analogy is being held in a wide net rather than pulled by a single rope: the net’s threads tug from many angles at once, and neighboring threads partially cancel each other’s sideways pull, so the total inward force is gentler than a single concentrated tug.

The researchers did not impose a sheet; they asked the data what shape it required. The data answered: flat. Spherical symmetry had been assumed without evidence.

The sheet lies within a far larger flattened arrangement, the supergalactic plane, home to major galaxy clusters including Virgo, Centaurus, the Great Attractor, Hydra, and Perseus-Pisces. At both scales the galaxies trace a skeleton built by dark matter underneath.

VI. Ancient Architecture, Full Hierarchy

The sheet geometry is ancient. In 2017, Marrone and collaborators used the ALMA radio telescope to resolve SPT0311-58, a distant galaxy system, at redshift 6.9.²⁷ We see this system as it appeared when the universe was only 780 million years old. The surrounding gas already showed sheet-like distribution: the same flattened geometry present in the first 6% of cosmic history.

The pattern imprints at smaller scales too. Koch & Grebel (2006) found Andromeda’s satellite galaxies arranged in a highly flattened plane, with early-type dwarfs aligned within 5 to 7 degrees of Andromeda’s pole at 99.7% significance.²⁸

Ibata et al. (2013) extended this: roughly half of Andromeda’s satellites form a vast, thin, co-rotating plane, at least 400 kiloparsecs (about 1.3 million light-years) across yet less than 14.1 kiloparsecs thick.²⁹ The Milky Way’s satellites show a similar planar arrangement.

These satellite planes are the constructal pattern fracturing downward: cosmic sheet, filaments, galaxy planes, satellite planes. The same flow optimization, nested.

The smallest rung carries a caveat. Satellite planes may instead be merger debris: dwarfs condensed from a single tidal stream would co-rotate in a plane because they inherited one orbit, with no flow optimization at work.^29b The hierarchy above them, sheet to filament to node, stands on independent evidence either way.

The cosmic web has four structural elements: voids, walls (sheets, where voids meet), filaments (where walls intersect), and nodes (where filaments intersect). Yakov Zel’dovich’s pancake picture (1970) predicts this full geometric hierarchy; the constructal reading offers an account of why the flow takes it. Matter drains from three-dimensional voids onto two-dimensional sheets, into one-dimensional filaments, into zero-dimensional nodes. Each stage concentrates the flow further.

A spherical arrangement is the low gravitational-entropy default in the sense of Section III: matter still spread the same way in every direction, no symmetry broken yet, no history written into the geometry. Sheets are what gravity produces given time to organize, one step along the clumping direction that raises gravitational entropy. The transition from spherical assumption to observed sheet is itself a constructal finding: flow systems develop directional structure (anisotropy, variation depending on direction) rather than remaining isotropic (uniform in all directions).

The sheet completes the Taoist picture. Lao Tzu’s pot: “We shape clay into a pot, but it is the emptiness inside that holds whatever we want.” The sheet is the pot itself, the surface between fullness and emptiness.

VII. Dark Energy’s Laboratory

Whatever drives accelerating expansion expresses itself most clearly in voids, where matter’s gravitational pull is weakest.¹⁰ Four competing programmes illustrate what voids might teach us.

First, sound waves from the universe’s first few hundred thousand years remain frozen into the arrangement of matter: the spacing between galaxies today still carries their imprint, the way tree rings record past seasons. Astronomers call these fossil ripples baryon acoustic oscillations, and DESI’s measurements of them hint at evolving dark energy (a 3.1-sigma preference from DR2 combined with the cosmic microwave background alone, and 2.8 to 4.2 sigma once supernova datasets are added).¹⁴

Second, the Keenan-Barger-Cowie (KBC) void hypothesis proposes that we sit inside a local underdensity roughly 300 Mpc in radius (about a billion light-years, so a diameter near two billion). If real, this void could account for the Hubble tension (the persistent disagreement between different methods of measuring the universe’s expansion rate) without new physics. Measured against the baryon acoustic oscillation baseline the source itself uses (not the headline supernova-versus-Planck figure, usually quoted near five sigma), the KBC void would reduce the tension from 3.3 sigma to 1.1 to 1.4 sigma across twenty years of data.^{15 16} The size is contested. A 2025 analysis, fitting void models directly to measured distances of nearby galaxies (Tully-Fisher distances from the CosmicFlows-4 catalog), prefers a void radius below 70 Mpc.⁶⁰⁷ That is under a tenth of the radius in the model Haslbauer and colleagues fitted to counts of galaxy brightness.

Third, Wiltshire’s Timescape cosmology dispenses with dark energy entirely. It derives the Hubble tension from differential clock rates between voids and walls. Clocks in emptier regions tick faster than clocks in denser ones, and the discrepancy mimics acceleration.^{23 24}

Buchert’s averaging formalism supplies the mathematics behind Timescape and its relatives. Together they make the backreaction case: the idea that the universe’s lumps change how the average universe expands. General relativity is nonlinear, so you cannot average the lumpy, uneven universe and expect the average to behave like a smooth one. The average temperature of a room with one corner on fire and one corner frozen is “comfortable,” yet no one in the room is comfortable. Voids contribute the dominant share of the correction term.^25a

Fourth, whatever produces the dark energy signal may adjust its local expression depending on the density of surrounding matter: a phenomenon called screening. In 2025, physicist Slava Turyshev of NASA’s Jet Propulsion Laboratory set out how solar-system experiments could hunt for it.⁶⁰⁸ The analogy is a radio signal in a dense city: buildings scatter and absorb the broadcast until a receiver cannot distinguish it from noise. Move to open countryside, and the same signal comes through clearly.

Two models capture the physics. In the “chameleon” model, a hypothetical scalar field (a quantity defined at each point in space, like temperature or pressure) acquires effective mass in dense environments; for such a field, more mass means shorter reach, so its range shrinks until it becomes undetectable. In voids, the field extends freely. Near the Sun, it retreats into a thin outer shell at the stellar surface, suppressed to levels five orders of magnitude below current instrument sensitivity. In the Vainshtein model, the field’s own nonlinear self-interaction creates a suppression zone around massive objects. For the Sun, this zone extends roughly 400 light-years, encompassing the local stellar neighborhood.

The prediction is structural: a force that is real and present everywhere, yet invisible where observers live. Dense regions screen it. Sparse regions reveal it. Turyshev argues that detecting such a force locally requires two steps: first, using large-scale cosmological surveys (DESI, Euclid) to derive precise local predictions; second, designing targeted solar-system experiments around those predictions. The instruments that confirmed Einstein were designed to test Einstein. Finding what screening hides requires instruments designed to test screening.

What the four programmes share is where they place the blame. Each looks to the late or local universe, to the structure that has assembled since the cosmos cooled enough for atoms to form, rather than to the physics of the first few hundred thousand years.

A 2026 stellar census bears on that division. Banik and colleagues dated 155,600 subgiant stars within 5 kiloparsecs (about 16,000 light-years) of the Sun. They kept only stars low in iron yet rich in oxygen, magnesium, and silicon. That ratio works as a birth certificate. Exploding massive stars forge oxygen, magnesium, and silicon quickly, within a few million years of a burst of star formation. Most iron arrives later, produced mainly by a slower class of explosion (Type Ia supernovae, in which a white dwarf detonates after accumulating matter from a companion). A star carrying the first trio of elements without much iron must have formed before the iron arrived. The oldest of them dated to 13.73 billion years.⁶⁰⁹

Fixes that resolve the Hubble tension by changing physics before atoms formed generally require a universe of 12.9 billion years to match the low-redshift measurements, which would put the oldest stars near 12.7 billion. By this preprint’s dating, the stars in our own neighborhood are older than that. If the result holds, it picks no winner among the four; it narrows the field to explanations of their kind.

The sheet geometry may bear on the Hubble tension as well. If the dark matter environment around the Milky Way is a dense plane rather than a sphere, the gravitational influence on nearby galaxy motions differs from standard assumptions. The KBC void hypothesis may need reconsideration: we sit inside a particular mass distribution, and its shape affects measurements taken from within it.

Cosmologists assumed homogeneity: that our patch of the universe is typical of the whole. The KBC void challenges that assumption. We may have been measuring the universe from inside a particular emptiness, mistaking the local for the global.

If the speculation in Chapter 16 has merit, that dissipative systems might be causally significant to cosmic structure, voids are where such effects become measurable. These regions are relatively lifeless; if life affects expansion, they expand differently from dense ones. Speculative, yet testable in principle, though life’s energy budget is gravitationally negligible by many orders of magnitude, and the test would most likely come back null.

VIII. The Informative Absence

Despite their emptiness, voids encode detailed information about the universe’s initial conditions. Regions with less matter contain more information about fundamental physics. This can sound like a paradox, since there is almost nothing inside a void to describe. Both statements hold, but they measure different things. What a void preserves instead is a nearly untouched record of the conditions it started from, because so little has happened inside it to overwrite them.

Sherlock Holmes understood this:

“Is there any other point to which you would wish to draw my attention?” “To the curious incident of the dog in the night-time.” “The dog did nothing in the night-time.” “That was the curious incident.”

The dog’s silence told Holmes that the intruder was familiar. What didn’t happen was maximally informative.

Voids carry information because they are depleted:

• Their sizes encode the spectrum of primordial density fluctuations from the early universe.
• Their shapes reveal the balance between dark energy and gravity.
• Their expansion rates probe dark energy more cleanly than dense regions.
• The galaxies within them show what is possible at the edge of formation thresholds.

Sutter and others have shown that void distributions carry imprints of early-universe physics, including the nature of dark matter and cosmic inflation.¹¹

Voids are the universe’s control group: they show what happens when most of the usual factors are removed.

Voids are also dynamically active. The Integrated Sachs-Wolfe Effect shows this directly.

Emptiness is uphill, which takes a moment to accept. The matter heaped along a void’s walls pulls backward on anything traveling inward, so the middle of a void sits at the top of a broad, shallow rise. Photons from the cosmic microwave background lose energy as they climb into a void’s elevated gravitational potential. They regain less energy descending out the far side, because accelerating expansion has flattened the hill during the crossing. In a flat universe containing only matter, with structure still growing gently, the hill would hold its height and the books would balance. A hill that changes height during the crossing therefore tests how the universe expands and grows, dark energy included, though curvature and other dynamics can also reshape it. The analogy: a ball rolling over a hill that sinks beneath it. It gains speed descending, yet recovers less than it spent climbing because the far slope has shrunk. It exits slower than it entered.

CMB photons traversing voids arrive slightly cooler than expected.¹⁹ The signal from supervoids is anomalously strong, stronger than ΛCDM predicts, and remains unresolved: a dog that barked louder than it should have.

Every void-based constraint expected from Euclid’s void catalogs, Rubin’s galaxy surveys, and DESI’s spectroscopic mapping rests on this control-group principle. Voids are the cleanest laboratory in the universe because of their emptiness. Precision void cosmology has arrived.²⁰

IX. Void Galaxies and the Environmental Gradient

Void galaxies, such as the isolated MCG+01-02-015, raise a central question: how do galaxies form in low-density environments?

The Void Galaxy Survey and CAVITY project have mapped these properties in detail:¹²

• Higher specific star-formation rates: more active per unit mass than counterparts in denser environments. Gas remains available as fuel, so the apparent enhancement is generally interpreted as delayed consumption.
• Lower stellar metallicities: fewer heavy elements at a given mass (roughly 20% less than filament galaxies and 60% less than cluster galaxies), reflecting less stellar recycling.
• Different morphologies: predominantly late-type (spiral rather than elliptical), bluer, more irregular. The gravitational harassment that transforms spirals into ellipticals never arrived.
• Less interaction history: isolated from the mergers, tidal stripping, and ram pressure (the headwind a galaxy feels plowing through cluster gas) that shape cluster galaxies.

The void environment preserves galaxies in an earlier evolutionary state. They retain gas and morphology that denser environments strip away.

JWST weak-lensing mapping of the COSMOS field (Scognamiglio et al., 2026) achieved twice Hubble’s resolution.¹ Galaxies form at “thick knots” along dark-matter filaments; where the scaffold concentrates, luminous matter follows. The map also reveals mass peaks with little visible counterpart, indicating regions of predominantly dark matter.

The scaffold reliably produces complexity given sufficient density and time, though some concentrations remain dark. Structure and dissipation are coupled yet distinct.

The supergalactic plane extends this gradient. Inside the dense plane, frequent mergers destroy spiral structure, producing smooth ellipticals. M87, hosting the black hole first imaged by the Event Horizon Telescope in 2019, is a product of this environment. On the periphery, galaxies evolve more freely, retaining gas, morphological diversity, and star-forming capacity.

The gradient runs from dense-plane interior to plane edge to void. Density builds galaxies; past a point, it grinds them toward one shape. On the book’s reading, diversity and optionality (the range of futures still open to a system) rise as density falls: environments that force convergence use up the diversity their history produced, and sparser ones preserve it. Section XX treats the parallel as an illustration, not a test.

X. We Are Children of the Void

We are inside a void, though of a much smaller kind.

The Local Bubble is a region roughly 1,000 light-years across, filled with hot, low-density gas. Supernova explosions beginning about 14 million years ago carved it out, sweeping away the interstellar medium (the gas and dust between stars) and creating the emptiness through which our solar system drifts.

Zucker and collaborators (2022) mapped the Local Bubble’s history using Gaia data.¹³ Nearly all the young stars near the Sun, those formed within the last 14 million years, were born on the surface of this bubble.

The emptiness gives us clearer sight lines, a particular radiation environment, and the stellar nurseries that birthed our youngest neighbors. Our neighborhood is the work of an absence.

The absence is salted with ash. Iron-60, a radioactive form of iron forged in massive stars and flung out when they explode, is still settling onto Antarctic ice, atom by atom, in recently fallen snow.^13a The supernovae that carved our void did not simply leave; their fallout is still arriving.

That fallout rose and fell between 40,000 and 81,000 years ago. Recovered from a Dronning Maud Land ice core, it records the changing density of the interstellar cloud we are now drifting through.^13b The clearing remembers its makers, and we live inside the record of them.

The makers had a birthplace. At its nearest, about a thousand light-years from the Sun, runs the Radcliffe Wave: a chain of dense gas clouds some 9,000 light-years long that rises as much as 700 light-years above the plane of the galaxy and dips as far below it, with the stellar nurseries of Orion and Cygnus X strung along its length. In 2024, measurements of its gas and its young star clusters showed that the Wave moves. The disk’s gravity pulls each cloud back toward the midplane, so the clouds overshoot and swing back like pendulums. Neighboring clouds swing slightly out of step, which sends the crest traveling along the chain while each cloud only rises and falls. The whole structure is also drifting outward from the galactic center at about 5 kilometers per second. Run that drift backward, and the cluster whose supernovae blew out the Local Bubble may have been born inside the Wave.^13c

Our solar system has passed through it. Tracing 56 young clusters that ride with the Wave back through time, astronomers placed the crossing in the Orion region, with the closest approach between about 15 and 12 million years ago.^13d That window overlaps the Middle Miocene climate transition, when Antarctica settled into the lasting ice cover it still carries. The same team asked whether extra interstellar dust from the crossing could have driven that cooling, and estimated that the dust fell one to five orders of magnitude short; on present numbers the crossing played at most a minor part. Nor can the ash settle the question yet: iron-60 in deep-sea crusts and sediments, the longer archive, reaches back only about 10 million years.

Bubbles like ours dot the galactic disk, and one of them spent forty years impersonating something far grander. A radio arc looming over the Milky Way’s center, cataloged since 1984 as the Galactic Center Lobe, was long interpreted as the plume of an ancient eruption from the core. Optical mapping in 2026 revealed instead a closed shell of ionized hydrogen roughly 115 light-years across and only 6,500 light-years away, a quarter of the distance to the center it appeared to crown.^13e Its gas drifts within 5 kilometers per second of rest, showing none of the churn expected near the core, and young stars light it from within, much as the stars of Orion light the nearby Barnard’s Loop. The astronomers who resolved the confusion propose keeping the acronym and rereading it: the Greatly Confused Loop.

The nesting extends to grander scales. The KBC void, identified through near-infrared galaxy number counts, spans approximately 2 billion light-years on those counts, which would make it one of the largest structures ever mapped.¹⁵ A 2025 analysis based on measured galaxy distances prefers a much smaller void (Section VII). The CMB Cold Spot, a 10-degree anomaly in the cosmic microwave background, overlaps low-density structure on the sky, though later analysis found the known foreground voids unable to account for most of its temperature decrement under the standard model.²⁵

If confirmed, we sit inside a cosmic underdensity vast enough to bias our measurements of the universe’s expansion rate.

Children of the void at nested scales: the Local Bubble shaped our stellar neighborhood; the KBC void may shape our cosmology.

The Milky Way’s disk is oriented almost perpendicular to the supergalactic plane. A galaxy aligned with the plane would be embedded in its densest flows, destined for merger-driven simplification. Perpendicular orientation offers geometric shelter: disruptive flows run along the plane rather than through the disk. The Milky Way sits on the plane’s outskirts, which may be why it retains spiral arms, active star formation, and the conditions that produced us. The perpendicular orientation is observed; the protective consequence is inferred from the geometry and has not yet been modeled.

Many creation narratives begin with void: the earth of Genesis, “without form, and void”; Hesiod’s Chaos; the Norse Ginnungagap. Ours, it turns out, did too.

XI. The Galactic Thermostat

Galaxies self-regulate. Active Galactic Nuclei (intensely bright galactic cores powered by matter falling into supermassive black holes) operate as thermostats:

1. Gas cools and flows toward the galactic center.
2. Accretion (the process of matter spiraling inward) feeds the black hole.
3. The black hole launches jets of superheated matter.
4. Jets inflate bubbles of hot gas.
5. Heated gas stops cooling.
6. Accretion slows. Jets weaken.
7. Gas cools again. The cycle repeats.

This cycle operates over tens to hundreds of millions of years.^{30 31}

This negative feedback loop shares the formal structure of a household thermostat: sensor (accretion rate), effector (jets), controlled variable (circumgalactic gas temperature). Nobody regulates the galaxy; it regulates itself.

XII. The Baryon Cycle: Galaxies as Metabolic Systems

Galaxies also metabolize, processing gas through cycles that mirror biological metabolism.

The circumgalactic medium (the vast gas reservoir surrounding every galaxy, extending well beyond the visible disk) is dynamic. Gas flows in from the intergalactic medium, passes through star formation and feedback, and returns enriched, heated, and restructured.³²

Anglés-Alcázar et al. (2017), using FIRE simulations, showed that galaxies grow substantially through re-accretion of previously ejected gas.³³ Late-time fuel supply is dominated by recycled material, overshadowing fresh accretion: consume, process, return, re-consume. Metabolism.

Wright et al. (2024) compared the baryon cycle across three independent simulation suites (EAGLE, IllustrisTNG, and SIMBA) and found similar final stellar masses through markedly different baryon-cycling pathways.³⁴ The endpoint is robust; the route is flexible. (Each suite was tuned to reproduce the stellar masses of real galaxies, so the shared endpoint is partly built in; the divergence of the routes is the finding.)

This is equifinality: multiple paths leading to the same destination. A river reaches the sea whether it takes the northern or southern fork; a dropped ball reaches the bottom of a bowl regardless of where on the rim you release it. Galaxies converge on similar outcomes through different histories. Equifinality is the signature an attractor leaves.

XIII. Galaxy Conformity: Coordination Without Contact

In 2006, Weinmann and collaborators studied galaxy group catalogs from the Sloan Digital Sky Survey.³⁵ They discovered that satellite galaxies’ star-formation properties correlate with their central galaxy’s properties at fixed halo mass (when comparing galaxies living in equally massive dark-matter halos). Quiet centrals are surrounded by quiet satellites; actively star-forming centrals by active satellites.

Kauffmann et al. (2013) extended this: the conformity signal persists to approximately 4 Mpc (about 13 million light-years), roughly ten times the gravitational boundary of a galaxy’s dark-matter halo.³⁶ Galaxies with no direct gravitational contact appear to remain coordinated in their properties. The large-radius signal is debated: later work argues that part of it may be a selection or halo-mass systematic rather than genuine long-range conformity, so the effect’s true reach remains unsettled.

Sheets may explain this. If a sheet channels gas and sets the tidal field for everything within it, conformity at 4 Mpc is what you would expect from galaxies embedded in the same one.

Coordination by shared environment: the same logic by which biofilm cells coordinate through signal molecules pooling in their common medium, or forest trees synchronize mast years (seasons of heavy seed production) by responding to the same weather. No signal passes directly between galaxies; coordination emerges from the shared substrate.

XIV. Bilateral Influence: Baryons Reshape Dark Matter

Pontzen and Governato (2012) addressed the cusp-core problem: simulations predict steep central density peaks in dark-matter halos, yet observers measure gentler profiles. They showed that repeated supernova-driven outflows irreversibly transform dark-matter halo structure.³⁷ The steeply peaked profiles flatten as gas is repeatedly blown out and re-accreted. Ordinary matter reshapes the dark-matter scaffolding it inhabits, the way a tenant’s renovations alter the building’s structure.

Di Cintio et al. (2014) generalized this: dark-matter profile shape depends systematically on stellar-to-halo mass ratio.³⁸ The more baryonic processing, the more dark-matter structure is modified. The scaffold and what it scaffolds co-evolve.

The same bilateral dynamic operates at every other scale: the environment shapes the organism; the organism reshapes the environment. Niche construction.

XV. The Gas Regulator: A Literal Attractor

Lilly et al. (2013) proposed the “gas regulator” model: galaxies maintain quasi-equilibrium star formation through balanced accretion, consumption, and outflow.³⁹ Disturb the equilibrium; the system recovers.

Peng & Maiolino (2014) formalized the dynamics, showing the gas regulator behaves as a damped oscillator:⁴⁰ perturbations are damped back toward equilibrium, each swing smaller than the last, like a playground swing with friction settling to rest.

The gas regulator shares mathematical structure with the Trust Attractor (formalized in Chapter 17): both are basins of attraction, states toward which a system naturally returns when disturbed, as every stable equilibrium is. The shared structure is generic; the interest lies in what each basin holds. Star-formation rate is drawn toward equilibrium because the feedback architecture makes that state stable. The galaxy returns to its equilibrium rate the same way a thermostat returns to its set temperature.

XVI. Magnetic Coherence: Non-Gravitational Order

Gravity is not the only organizing force at cosmic scales.

Vernstrom et al. (2021) detected coherent magnetic fields of 30 to 60 nanogauss (billionths of a gauss, roughly ten million times fainter than Earth’s magnetic field, yet organized) in cosmic-web filaments at scales exceeding 3 Mpc. This was the first detection, by stacking, of ordered magnetic fields along cosmic-web filaments.⁴¹ Carretti et al. (2023/2025) extended this to supercluster scales with LOFAR, finding 10 to 145 nanogauss.⁴²

These fields do real work. Magnetic pressure opposes gravitational collapse along certain axes, channels gas flows, and shapes structure formation. Filaments organize fields; fields influence matter flow.

XVII. Mergers as Embrace

Galaxy mergers can dismantle old structures and kindle new stars in the same encounter.

Direct collisions between individual stars are extremely rare during a galaxy merger.⁶¹⁰ Stars are so small relative to their separations that two galaxies can pass through each other like two swarms of fireflies crossing a field, their members mostly missing one another. The Antennae (NGC 4038/4039), one of the nearest major mergers and one of Hubble’s most studied objects, illustrates the outcome. Two spiral galaxies have been interacting for hundreds of millions of years. Hubble images reveal a rich population of young star clusters formed during the encounter.⁶¹¹

Gas responds differently. Clouds collide, shocks heat them, and turbulent motion dissipates. In the Antennae’s overlap region, infrared observations of molecular hydrogen reveal emission powered by shocks, tracing this loss of turbulent kinetic energy.⁶¹² The same encounter that mostly lets stars pass one another brings their gas into direct, energetic contact.

Compression can help that gas form stars. The Jeans mass is the smallest mass a cloud must carry, at a given temperature and density, before its own gravity overcomes thermal pressure. At fixed temperature, squeezing a cloud lowers this threshold. If the gas can radiate away heat and lose enough turbulent support, it can collapse under its own gravity. Heating and turbulence can also delay collapse; the outcome depends on how these processes balance.

François Schweizer’s 2005 review of merger-driven galaxy evolution describes how mergers trigger galaxy-wide starbursts and chemical enrichment.⁴³ The encounter changes the conditions available to the gas, producing dense star-forming regions and eventually a deeper gravitational potential well. The first offers an analogy with the Trust Attractor: interaction can open possibilities unavailable in isolation. The parallel concerns access to new states; galaxies have no demonstrated capacity to choose cooperation or withhold consent. The second, as Section XX argues, is settling rather than persistence.

A density gradient governs how often individual stars make contact. In the diffuse outer regions, their wide separation makes direct impacts vanishingly unlikely; gravity rearranges their orbits across the intervening space. In dense globular clusters and nuclear star clusters, stellar collisions become more important.⁶¹³ Gas occupies much more of the available volume, so rare stellar impacts tell us little about how violently the gas interacts.

The consequences cross those scales. Tidal forces pull stars and gas into long tails while compressed gas forms new stars. Dense environments with frequent mergers still simplify morphology (Section IX): spiral structure is disrupted, gas is consumed or stripped, and the remnant may settle into a deeper, quieter well. Creation and disruption belong to the same encounter.

The Milky Way may undergo the same process if it meets Andromeda. A 2025 reanalysis of the two galaxies’ masses and motions found only about even odds of a merger within the next ten billion years, with a median merger time near 7.6 billion years.⁶¹⁴ The earlier, once-confident 4.5-billion-year appointment is no longer secure. At that revised median the Sun has already left the main sequence (the long, stable hydrogen-burning phase it is in now), so the Earth-bound vantage point the older accounts assumed no longer exists. If the merger comes, our solar system’s orbit will be rearranged; stellar collisions will remain negligible. The sky will transform.

Chapter 7 (Entropic Evolution) describes eukaryotic origins as embrace rather than capture. Galaxy mergers lend that image a cosmic counterpart: an encounter can create the conditions for something neither participant was making alone. Here the work is done by gravity, gas compression, cooling, and dissipation. The new stars are real; the embrace is our metaphor.

XVIII. Fertile Emptiness

Complex systems need void space. Tissues need the spaces between cells, which the extracellular matrix (the structural scaffolding between cells) holds open. Brains depend on synaptic gaps (the narrow spaces between neurons where chemical signals pass). Cities need parks and plazas; Jane Jacobs showed that a park dies without busy streets around it. Creativity requires slack, because a mind fully allocated leaves no room for surprise.

The ethical framework names this optionality, the preserved possibility space met in Section IX. Voids are possibility space at cosmic scale, where the future is least constrained by the past.

Over-coordination is as pathological as under-coordination. Optimal topologies maintain specific densities: sparse enough to permit novelty, dense enough to sustain interaction.

XIX. The Pattern at Every Scale

The feedback architectures share mathematical structure; the attractor states share stability properties. The same control-loop architecture describes a galaxy regulating its gas supply and a body regulating its blood sugar. Markus Aschwanden’s 2018 review catalogs seventeen self-organization processes across planetary, solar, stellar, galactic, and cosmological scales, all operating without central control.⁴⁵

Structure formation, working through gravity, feedback loops, magnetic fields, radiation pressure, tidal torques, and baryon cycling, is no more “just gravity” than a mind, working through neural networks, hormonal feedback, immune signaling, and synaptic plasticity, is “just electrochemistry.”

XX. The Trust Attractor at Cosmic Scale

The Trust Attractor (the book’s central ethical claim, developed formally in Chapter 17) names a proposed basin toward which coordinating systems tend, one in which coordination by invitation outlasts coordination by coercion. The cosmic web shows the same structural signature. Gravity would carve voids, sheets, and mergers regardless of any invitation-versus-coercion claim, so this is illustration rather than a falsifiable test of the framework. With that caveat, the inferred parallel at cosmic scale:

Sheet-edge galaxies interact gently through shared environment, preserving spiral structure, star formation, gas reserves, and morphological diversity; their futures remain open.

Sheet-interior galaxies undergo forced convergence through repeated mergers, simplifying into ellipticals with spiral arms destroyed, gas consumed or stripped, and optionality lost.

Two senses of “stable” must be kept apart here, and they pull in opposite directions. By the gravitational-entropy measure of Section III, the coerced interior wins: the merger-formed elliptical is more massive, more gravitationally bound, and occupies a deeper potential well, which is the thermodynamically favored direction for gravitating matter. That is settling, not persistence.

The Trust Attractor means stability in the other sense: a persistent, self-maintaining dissipative steady state, the kind a gas regulator can approach under sustained inflow (Section XV). In the comparison drawn here, sheet-edge spirals exemplify that persistence: they last for billions of years, maintaining quasi-equilibrium star formation; the interior reaches a deeper but quenched equilibrium, thermodynamically simplified, with reduced capacity for further complexity. The edge does not occupy the deeper gravitational well; it sustains ongoing dissipation and preserves optionality, and that is what “more stable” denotes in the framework.

Permissive environments preserve the diversity that enables adaptation. Coercive environments simplify it away. A pattern hundreds of millions of light-years across, consistent with the framework and, as noted, not a test of it.

“Coordinative” does not mean intentional. The term refers to systems whose components mutually influence each other’s states through feedback, producing emergent stability that no component imposes and none could achieve alone. By that definition, which applies equally from autocatalytic chemistry to bilateral alignment, the cosmic web is coordinative. The sheet works as a facilitation architecture: flow runs along it.

XXI. The Void-Dominated Future

The standard cosmological projection is stark.²¹ As dark energy accelerates expansion, voids grow faster than the universe as a whole. Filaments stretch thin. Connections between galaxy clusters grow longer than light can cross. Gravitationally bound structures become islands in a void-sea, each unable to communicate with the others or detect them.

The isolation reaches into knowledge itself. When the last external evidence drifts beyond reach, the capacity to know the universe recedes with it, a severed feedback loop that Chapter 17c (Entropic Epistemology) follows to its end.

Complexity requires gradients: differences in density, temperature, or energy between one region and another, the way a waterfall requires both a high point and a low point. Growing voids isolate bound structures without immediately erasing their internal gradients. Stars will exhaust their fuel. The conditions that produced galaxies, neurons, love, and this book will not persist.

The trajectory ending in void domination is the same trajectory that produced everything worth valuing: the same river flowing in the same direction that carved channels where complexity briefly flourished. We are expressed by it, and that expression is finite.

Dark energy might evolve. The voids might harbor physics we have not yet discovered. The universe, for a time, built something extraordinary out of the same emptiness that will eventually reclaim it.

XXII. Looking Forward

The instruments are already at work. eROSITA has revealed the cosmic web’s hot baryonic rivers (Section IV). Euclid, launched in July 2023 and in routine survey operations since February 2024, has already issued two quick data releases; its first major release begins with a foundation tranche in November 2026, and the higher-level products that include void catalogs follow in mid-2027. The Vera C. Rubin Observatory began full LSST survey operations in early 2026.

DESI has already produced its first void catalogs from DR1: 1,489 voids mapped via VoidFinder at redshifts below 0.24.²² Techniques developed for void mapping, including ZOBOV, have found uses from network analysis to materials science.

We are children of the void. The cosmic web teaches what every other scale corroborates: emptiness enables structure, structure enables coordination, coordination enables persistence. The pattern holds. All the way up.

Notes for this chapter