Chapter 8: The Invisible Majority

THE INVISIBLE MAJORITY

Indra's Net, Dark Matter, and the Physics of the Scaffold

> The things which are seen are temporal. > — 2 Corinthians 4:18

Introduction: The Curve That Should Have Fallen

Take a spiral galaxy and measure how fast things orbit at different distances from the centre.

You know what the answer should look like, because you know what it looks like in the solar system. Mercury moves at forty-seven kilometres per second. Neptune moves at five. The further out you go, the slower things travel, because almost all the mass is concentrated in the middle and the gravitational pull weakens with distance. Orbital speed falls off as the square root of the radius. Kepler worked this out in 1619 and it has not needed revision.

A spiral galaxy is also a system with most of its light concentrated in the middle. So the orbital speeds should fall off in the outer regions the same way.

They don't.

They go flat. You measure outward through the bright inner disk, and out past where the stars thin to almost nothing, and out into the faint outer gas where there is no visible material at all — and the rotation speed just stays there, roughly constant, kilometre for kilometre, as far as anyone can measure. `[VERIFIED]`

There are only two things that can mean. Either there is a great deal of mass out there that we cannot see, distributed very differently from the light. Or gravity does not work the way we think it does at those accelerations.

Ninety years after the discrepancy was first noticed, we have an extremely detailed model of the first option, no detection of the substance it requires, and a minority tradition insisting on the second option that keeps making predictions that come true.

This chapter is about that situation, and about what it means to be held up by something you have never touched.

Section I: The Net and the Jewels

1.1 Indra's Net

In the Huayan school of Chinese Buddhism there is an image, drawn from the Avatamsaka Sutra, of a net stretched infinitely in every direction across the palace of the god Indra.

At every knot of the net there hangs a jewel. Each jewel is polished, and each reflects every other jewel in the net. And because each reflection contains all the other reflections, every jewel contains an infinite regress of the entire net, and any change to one is registered — instantly, completely — in all the rest.

The seventh-century patriarch Fazang is said to have demonstrated this to Empress Wu by installing mirrors on the walls, floor and ceiling of a room, placing a Buddha statue in the centre with a lamp beside it, and letting her look.

The philosophical point is not decorative. In the Huayan reading, no jewel has independent existence. What a jewel is consists in its reflections of everything else. The relations are not something the jewels have; the relations are what constitute them.

And an honest note on the parallel, because this book has broken two already. `[ILLUSTRATIVE]` Indra's net is about mutual constitution and interdependence. It is not about invisible mass, and I am not going to claim the Huayan patriarchs anticipated a halo. The part that transfers is narrower and worth having: the net is not itself a jewel. It is the connective structure, it is load-bearing, it does not shine, and every account of the image describes it in one sentence before moving on to the beautiful part.

1.2 The Web of Wyrd

The Norse have a structurally similar object with a different emotional temperature.

The Norns sit at the well of Urd beneath the world tree, and they weave — or in some tellings, carve into wood, or pour water on the roots to keep the tree from rotting. What they maintain is wyrd, from the same root as the verb to become.

Wyrd is often flattened into "fate" in translation, and the flattening loses the thing that makes it interesting. It is closer to the accumulated weight of everything that has already happened, constraining what can happen next. Not a script written in advance, but a state that narrows the space of futures. Pull one strand and the whole web registers it.

Again: what is real and consequential in the image is the web. What gets illustrated on the cover of the book is the Norns.

1.3 Nobody Writes About the Net

In Chapter Four I pointed out that Hestia has almost no mythology, because her function is the absence of events, and narrative requires events.

There is a second and related gap, and this chapter is standing in it. Cultures narrate the nodes and not the connections. We have thousands of pages about the jewels and one line about the net. We have named Norns and an unnamed web. We have gods of the harvest and no god of the soil chemistry.

This is not a failure of imagination. It is a property of the observing apparatus. Stories are about things that do things, and connective tissue does not do things in the way a story can register — it makes it possible for other things to do things, which is a different grammatical position and one that language handles badly.

Keep that in mind through the next four sections, because we are about to watch physics make exactly the same error and take ninety years to stop.

Section II: The Discrepancy

2.1 Zwicky, and Forty Years of Nothing

In 1933, Fritz Zwicky measured the velocities of galaxies in the Coma Cluster and applied the virial theorem — a straightforward relation between the kinetic energy of a bound system and the gravitational potential holding it together.

The galaxies were moving far too fast. By his estimate the cluster needed hundreds of times more mass than its luminous content to remain bound. Without it, Coma should have flown apart long ago.

He called the missing component dunkle Materie. `[VERIFIED]`

And essentially nothing happened for forty years. Zwicky was abrasive, his distance scale was off by a large factor, and the result sat in the literature as a curiosity. Horace Babcock found anomalous rotation in Andromeda in 1939 and attributed it to something else. Jan Oort had found related discrepancies in the local stellar disk in 1932.

The anomaly was on the table for four decades and nobody built on it. That is worth noticing alongside the perytons and the solar neutrino problem: this book keeps collecting cases where the data were fine and the field's attention was the failure point.

2.2 Rubin and Ford

In the late 1960s and through the 1970s, Vera Rubin and Kent Ford began systematically measuring rotation curves using an image-tube spectrograph Ford had built, which was sensitive enough to get spectra from the faint outer regions of spiral galaxies.

They started with Andromeda. Then they did dozens more.

Every one of them was flat. `[VERIFIED]`

The strength of the result was not any individual galaxy. It was that it was universal — a systematic property of spirals, not an oddity of one system, showing up regardless of size or type or environment. By around 1980 the case was difficult to argue with, and the field's attitude shifted from dismissal to acceptance over roughly a decade.

Rubin died in 2016 without a Nobel Prize. I mention it, briefly, because Chapter Six had a similar note about Jocelyn Bell Burnell, and two data points in one book is a pattern worth registering rather than a coincidence worth remarking on.

2.3 Past Where the Light Stops

The decisive extension came from radio. Neutral hydrogen emits at twenty-one centimetres, and hydrogen gas extends far beyond the optical edge of a galaxy's disk — often to twice the radius or more.

Which means you can measure orbital velocities out where there is essentially no starlight at all, and this is where the argument became unavoidable. If the flat curve continued into a region containing almost no visible material, the mass responsible for it could not be some unaccounted population of faint stars. It had to be distributed differently from the light entirely, extending far past it, in a roughly spherical halo. `[VERIFIED]`

And a related precision, because it corrects a common assumption. "Visible matter" and "ordinary matter" are not the same thing. Stars are only a small fraction of the baryonic mass in the universe — most ordinary matter is diffuse ionised gas between and around galaxies, and for decades roughly half of the expected baryons could not be located at all. This "missing baryon problem" was substantially resolved around 2020, when the dispersion of fast radio bursts — the objects from Chapter Six — was used to weigh the intergalactic medium along their sightlines and the baryons turned up where the models said they should be. `[VERIFIED]`

So there are two distinct invisibilities here, and they are routinely conflated. Most ordinary matter is invisible but findable by other means. Dark matter is something else.

2.4 Why One Anomaly Would Not Be Enough

If galaxy rotation were the only evidence, the honest position would be that we have a persistent unexplained dynamical result and no idea what it means.

It is not the only evidence, and this is the part that popular treatments consistently underweight. There are at least six substantially independent lines, drawing on different physics at different scales and different epochs:

1. Galactic rotation curves. 2. Cluster dynamics — Zwicky's original velocity dispersions, now vastly better measured. 3. Gravitational lensing — the bending of light past clusters maps their mass directly, without any dynamical assumptions, and the maps require far more mass than the light accounts for. 4. The cosmic microwave background. The relative heights of the acoustic peaks from Chapter One depend on how much matter was present that did not couple to photons. A universe with only baryons produces a visibly different peak structure. This is a measurement from 380,000 years after the Big Bang, entirely independent of anything happening in galaxies. `[VERIFIED]` 5. Big Bang nucleosynthesis. The deuterium abundance pins the baryon density to about five per cent of the critical density — and it agrees with the CMB's independent baryon measurement, while the total matter density from both is around thirty per cent. The gap is not baryons. 6. Structure formation. Baryons cannot begin clumping until after recombination, because before that they are locked to the radiation. A universe with only baryons has not had enough time since then to build the structure we observe. Something that could start collapsing earlier is required.

These are not six versions of one observation. They involve different instruments, different centuries of cosmic history, and different physics. And they agree on the same number: roughly five times as much dark matter as ordinary matter, or about eighty-five per cent of all matter. `[VERIFIED]`

One precision on that figure. The 85% is the universal matter budget. Individual systems vary enormously — within the bright inner disk of a large spiral, baryons can locally dominate, while dwarf galaxies can be almost entirely dark-matter-dominated. And of the total energy density of the universe, matter of all kinds is only about thirty-two per cent; the rest is dark energy, which is Chapter Eleven's problem and is not the same thing at all.

Section III: The Case, and the Gap

3.1 The Bullet Cluster

The single most quoted piece of evidence is a collision.

1E 0657-558, imaged in detail by Douglas Clowe and collaborators in 2006, is two galaxy clusters that have passed through each other. Three components can be tracked separately.

The galaxies themselves are effectively collisionless — the space between them is so vast that they pass through without interacting — so they continued on their trajectories and are now on either side.

The hot X-ray gas, which is the majority of the baryonic mass, is collisional. It rammed into itself, shocked, slowed, and is sitting in the middle, dragged out of position.

The gravitational mass, mapped independently by weak lensing, does not follow the gas. It follows the galaxies. `[VERIFIED]`

So the bulk of the gravitating mass has separated from the bulk of the ordinary matter. That is very hard to explain by modifying gravity, because in a modified-gravity picture the gravity ought to be where the matter is, and it isn't.

A fair caveat. The Bullet Cluster is often deployed as though it single-handedly ends the argument, and it does not quite. Its collision velocity has been argued to be uncomfortably high for standard cosmology, and modified-gravity proponents have constructed responses. It is a very strong result, not a knockout.

3.2 The Null Results, Honestly

Now the part that the confident popular account tends to skip.

We have been looking for the particle for forty years, with increasingly enormous detectors, and we have not found it.

The favoured candidate has long been a WIMP — a weakly interacting massive particle, attractive partly because a particle at roughly the weak scale would naturally have the right relic abundance, a coincidence known as the WIMP miracle. Experiments have looked for the tiny nuclear recoil from a WIMP striking a detector nucleus: XENON1T and XENONnT, LUX and LUX-ZEPLIN, PandaX, and a long line of predecessors, each improving sensitivity by orders of magnitude.

Nothing. `[VERIFIED]`

The most natural parts of WIMP parameter space have been excluded. And the search is now running into a hard limit: at current sensitivities, coherent scattering of solar and atmospheric neutrinos produces a signal nearly indistinguishable from the one being sought. Both LZ and XENONnT reported observing this neutrino background in 2024. The "neutrino fog" is no longer theoretical; it has arrived, and it means further gains get much harder. `[VERIFIED]` `[BOUNDARY]`

Axion searches have likewise found nothing so far. Primordial black holes remain viable in narrow mass windows and excluded in most. The one persistent positive claim — the DAMA/LIBRA collaboration's annual modulation signal, reported for over two decades — has not been replicated by experiments using the same target material, and results from COSINE-100 and ANAIS disfavour the dark matter interpretation. `[BOUNDARY]`

We have a component that shows up in six independent gravitational measurements and has never once shown up in a detector.

3.3 MOND, Taken Seriously

In 1983 Mordehai Milgrom proposed that instead of adding invisible mass, one might modify the dynamics: that below a characteristic acceleration of about 1.2 × 10⁻¹⁰ metres per second squared, the relation between force and acceleration departs from Newton's.

The standard treatment of MOND in popular science is a dismissive paragraph. I think that is intellectually dishonest, and the honest version is more interesting.

What MOND gets right is genuinely remarkable. It predicted, from a single free parameter, the baryonic Tully–Fisher relation — the tight empirical link between a galaxy's rotation speed and its total baryonic mass — before that relation was well established. And the radial acceleration relation, documented in detail by Stacy McGaugh and collaborators in 2016 across a large sample, shows that the observed acceleration in a galaxy at any radius is a tight function of the acceleration predicted from the visible matter alone, with remarkably little scatter. `[VERIFIED]`

Think about what that means in the dark matter picture. The dark halo, which is supposed to dominate the mass and to have formed through a stochastic merger history largely independent of the baryons, nonetheless arranges itself so precisely in relation to the visible matter that you can predict the rotation curve from the light. That is not impossible in ΛCDM — feedback processes can couple the two — but it is an unnatural coincidence, and it is the strongest thing MOND has going for it.

What MOND gets wrong is decisive at larger scales. It does not account for cluster dynamics without adding dark matter anyway. It does not explain the Bullet Cluster. It does not reproduce the CMB acoustic peaks. It does not build large-scale structure. And its leading relativistic extension, TeVeS, was severely damaged by GW170817 — the neutron star merger from Chapter Five — because the gravitational waves and the gamma rays arrived within about 1.7 seconds after travelling 130 million years, constraining the speed of gravitational waves to match light to about one part in ten to the fifteenth, which ruled out a large class of modified-gravity theories at a stroke. `[VERIFIED]`

The defensible position: dark matter is overwhelmingly favoured as the description of the universe at large scales, and MOND has identified a real, tight, unexplained regularity in galactic dynamics that dark matter models have to work hard to reproduce. Both of those sentences are true. A book that told you only the first would be misleading you about the state of the field.

3.4 What We Actually Have

Let me state the position without the varnish.

We have named a discrepancy. We have built a model of it — cold dark matter — that is quantitatively successful across an extraordinary range of scales and epochs, from the CMB to the cosmic web to the internal structure of galaxies. The model has known difficulties at small scales — the core-cusp problem, the detailed satellite population — most of which improve when baryonic physics is included properly, and not all of which are settled. `[BOUNDARY]`

And we have no idea what the stuff is.

This is a genuinely unusual epistemic state, and I want to name it accurately rather than resolve it prematurely in either direction. We have measured something we have not identified, repeatedly, from six directions, and every attempt to touch it has come back empty.

Section IV: Why It Stays a Halo

There is one more piece of physics, and it is the one that ties this chapter to the argument of the book.

Dark matter cannot radiate.

That is nearly the whole definition of it. It does not couple to the electromagnetic field, which is why it is dark, and the consequence goes far beyond invisibility.

Recall Chapter Three. A gas cloud collapses only because it can get rid of its heat. Compression heats gas; heated gas pushes back; and the collapse can continue only insofar as the heat can be radiated away into the cold universe. Radiation is how the cloud pays for its own contraction. That is why gravitational clumping increases total entropy, and it is why the whole process works.

Dark matter has no such channel. Compress a cloud of dark matter and it heats up — its particles move faster — and it has no way to shed that energy. So it cannot cool. And because it cannot cool, it cannot contract further. It virialises into a diffuse, roughly spherical, pressure-supported halo and it stays there.

This is why baryons form disks and stars and planets and people, and dark matter does not. The baryons can pay. They radiate, they cool, they sink to the centre of the halo, they settle into a rotating disk, they fragment, they ignite. Every structure in Chapters Three through Seven exists because ordinary matter has a way to lose energy.

Dark matter provides the potential well that made all of it possible — the halos formed first, before recombination freed the baryons, and the baryons fell into scaffolding that was already there — and dark matter can never become anything at all.

The invisible majority holds everything up and is structurally incapable of being any of it. Not by choice, not by neglect, not by anyone's decision. By the absence of a coupling.

Section V: A Worked Example of Getting It Wrong

I need to do something here that is not usually done in books like this, which is to correct a claim I would have used without checking.

The standard neuroscientific analogy for dark matter is glia. The version I have read many times, and would have written, runs like this: neurons are the stars, and glia are the dark matter of the brain, making up something like ninety per cent of its cells while producing no electrical signal — ignored for a century because they were quiet.

The ninety per cent is wrong. So is the frequently quoted ten-to-one ratio.

Direct counts using the isotropic fractionator method, published by Suzana Herculano-Houzel, Frederico Azevedo and colleagues in 2009, put the adult human brain at roughly eighty-six billion neurons and roughly eighty-five billion non-neuronal cells. A ratio close to one to one. `[VERIFIED]`

In 2016 Christopher von der Bartheld, Jami Bartlett and Roberto Lent traced where the myth came from and found what you would expect: an early, tentative, regionally specific estimate that was cited, then cited again, then cited as established, with each repetition dropping the qualifications, until a figure with no primary support was in every textbook. `[SOURCED]`

And the necessary nuance, since the corrected number is itself a compression. The one-to-one ratio is a whole-brain average and it conceals enormous regional variation. The cerebral cortex is glia-rich, with ratios of three or four to one. The cerebellum contains roughly eighty per cent of all the neurons in the brain in about ten per cent of its mass, and is overwhelmingly neuron-dominated. Whole-brain averages of two very differently structured regions are exactly the kind of summary this book keeps warning about. `[VERIFIED]`

The underlying argument survives intact. Glia are not packing material. Astrocytes regulate cerebral blood flow, buffer extracellular potassium, recycle neurotransmitters, and participate directly in synaptic signalling. Oligodendrocytes myelinate axons and set conduction velocity. Microglia are the resident immune population and actively prune synapses during development. A brain without them does not work at all. All of that is well established and none of it depended on the ninety per cent.

I have kept this section in rather than quietly fixing the number, because the myth's propagation is itself an instance of the thesis. A claim was compressed, the provenance was dropped, and the summary went on circulating for decades in a form nobody could decompress back to a source — until someone went and looked. That is the failure mode of every archive in this book, occurring inside the literature of the organ that does the archiving.

Section VI: The Low-Bitrate Image

`[ILLUSTRATIVE]` — application, not evidence.

6.1 The Operator Is the Instrument

Every previous chapter in this book had a physical operator. Expansion cooled the plasma. Gravity collapsed the cloud. Fusion consumed the hydrogen. The universe did the compressing.

This chapter is different, and the difference is the point. Dark matter is not compressed by anything. It is fully present, at full detail, exerting its full influence at all times.

We are the lossy operator. The compression happens in the measurement — in the fact that our primary channel is electromagnetic, and that anything which does not couple to that channel is absent from the image regardless of how much of the mass it constitutes.

A photograph of a galaxy is not a picture of a galaxy. It is a picture of the fifteen per cent of the matter that interacts with light, rendered as though it were the object. Every galaxy image you have ever seen is a low-bitrate summary in which the majority of the subject is not represented at all — and the summary is so persuasive that it took ninety years and six independent lines of evidence to convince the field that the picture was mostly missing.

6.2 What Transfers

The transferable claim is not that some people are dark matter. It is narrower and, I think, more useful.

Invisibility in a measurement channel is a property of the coupling, not of the importance. Dark matter is invisible because it does not interact electromagnetically. That is all. It is not invisible because it is unimportant, or peripheral, or because anyone decided to overlook it. It is invisible because the instrument measures one kind of interaction and this thing does not do that kind.

Most institutional measurement has the same structure. Organisations instrument what generates legible events — output, decisions, shipped work — and the connective functions produce no events of that type. Maintenance is visible only as the absence of failure. Institutional knowledge is visible only when the person holding it leaves. Care work produces no artefact. None of this is hidden. It is unmeasured by the specific channel in use, and the resulting picture is confidently wrong about the mass distribution.

The correction is not to look harder in the same channel. That is what the field did for forty years after Zwicky. The correction is to find a second channel — lensing rather than starlight, the CMB peaks rather than galaxy dynamics — and to notice that the discrepancy between channels is itself the measurement.

6.3 And the Uncomfortable Part

Section IV had an edge to it that I do not want to file down.

Dark matter cannot become anything. It provides the well, and the things that form inside the well are made of something else entirely, and no amount of time changes this, because the constraint is the absence of a coupling rather than a shortage of opportunity.

I am not going to convert that into a moral about invisible labour, in either direction. It would be too easy to make it a consolation — the scaffold is the real hero — and equally easy to make it a warning. Both would be smuggling.

What I will say is that the physics distinguishes cleanly between two things our language does not: being unseen, and being unable to participate through the channel that is being observed. Those get treated as the same problem and they are not, and the remedies are entirely different. One is a measurement failure and is fixed by instrumenting differently. The other is structural and is not fixed by attention at all.

The Four Slots

| Slot | The dark matter halo | |---|---| | Input | The primordial density perturbation field — the same fluctuations recorded in the microwave background of Chapter One, imprinted on a collisionless, non-radiating component | | Operator | Gravitational collapse without dissipation. The halo virialises and stops, because there is no channel through which it can shed the energy that further contraction would require | | Invariant | A diffuse, roughly spherical, pressure-supported potential well: a mass, a density profile, a total angular momentum — and no internal structure whatsoever | | Cost | The capacity to form anything. No disks, no stars, no chemistry, no planets. In exchange, permanence: a halo does not burn out, disperse, or evolve, and will outlast every structure it made possible |

Set this against Chapter Three's table. The molecular cloud paid a colossal price — dispersed, destroyed, ninety-five per cent lost — and bought stars and planets. The halo pays nothing and buys nothing. It cannot pay, so it cannot buy, so it persists unchanged.

In this book, the ability to lose is the ability to become.

The Protocol: Instrumenting the Net

`[ILLUSTRATIVE]` — application, not evidence.

Assume your primary channel is missing most of the mass. Not some of it — most. Any measurement system captures a particular kind of interaction, and things that do not interact that way are absent from the record entirely rather than under-represented in it. The default assumption should be that the picture is a minority of the subject.

Find the discrepancy, not the object. Nobody found dark matter. They found that two independently derived numbers — mass from dynamics, mass from light — did not match, and the mismatch was the discovery. If you want to detect what your instruments are missing, compute the same quantity two different ways and look at the gap. The gap is the signal.

Take the minority position seriously enough to state its best case. MOND is wrong about the universe and right about a real regularity that the standard model has to strain to explain. Being able to say precisely what an opposing view gets right is the difference between holding a position and having been told one.

Check the numbers you would have used without checking. The ninety per cent glia figure survived for decades in textbooks because it was plausible, useful, and nobody went back to the source. Every field has several of these. The ones most likely to be wrong are the ones that are most quotable and that you have never seen a citation for.

And distinguish unseen from uncoupled. If something is invisible because you are not looking, look. If it is invisible because it does not interact through the channel you are measuring, looking harder will produce nothing, and the answer is a different instrument.

Where This Leaves Us

• Spiral galaxy rotation curves are flat well beyond the optical disk, contrary to the Keplerian falloff expected from visible mass alone. `[VERIFIED]`
• Fritz Zwicky inferred missing mass in the Coma Cluster in 1933 from velocity dispersions; the result was largely neglected for four decades. `[VERIFIED]`
• Vera Rubin and Kent Ford established flat rotation curves as a universal property of spirals through the 1970s; 21 cm radio measurements extended the curves past the starlight. `[VERIFIED]`
• Most baryonic matter is not in stars. The "missing baryon" deficit was substantially resolved around 2020 using fast radio burst dispersion measures. `[VERIFIED]`
• Dark matter is supported by at least six substantially independent lines: rotation curves, cluster dynamics, gravitational lensing, CMB acoustic peak structure, Big Bang nucleosynthesis combined with total matter density, and structure formation timing. `[VERIFIED]`
• Dark matter constitutes roughly 85% of matter and ~27% of total cosmic energy density; individual systems vary widely. `[VERIFIED]`
• In the Bullet Cluster (Clowe et al., 2006), lensing-derived mass tracks the collisionless galaxies rather than the X-ray gas that dominates the baryonic mass. Its collision velocity has been argued to be in tension with standard cosmology. `[VERIFIED]` `[BOUNDARY]`
• Decades of direct detection experiments have produced no confirmed signal; much of the natural WIMP parameter space is excluded, and coherent neutrino scattering backgrounds were observed by LZ and XENONnT in 2024. `[VERIFIED]` `[BOUNDARY]`
• The DAMA/LIBRA annual modulation claim remains unreplicated and is disfavoured by COSINE-100 and ANAIS. `[BOUNDARY]`
• MOND predicts the baryonic Tully–Fisher relation and the tight radial acceleration relation from baryonic distribution alone, which ΛCDM must reproduce through feedback. MOND fails on cluster dynamics, the Bullet Cluster, CMB peaks, and structure formation, and TeVeS was heavily constrained by GW170817. `[VERIFIED]` `[BOUNDARY]`
• Small-scale challenges to cold dark matter (core-cusp, satellite populations) are substantially but not entirely addressed by baryonic physics. `[BOUNDARY]`
• Dark matter cannot radiate and therefore cannot dissipate energy, cannot cool, and cannot collapse beyond a virialised halo. Baryonic structure formation depends entirely on the ability to radiate energy away. `[VERIFIED]`
• The human brain contains roughly 86 billion neurons and roughly 85 billion non-neuronal cells, a ratio near 1:1. The widely repeated 10:1 or "90% glia" figures are unsupported and were traced to uncritical citation propagation. Regional ratios vary greatly. `[VERIFIED]` `[SOURCED]`
• Glial function — metabolic support, potassium buffering, neurotransmitter recycling, myelination, immune response, synaptic pruning — is well established independently of any cell-count ratio. `[VERIFIED]`
• The mapping from measurement-channel invisibility to institutional invisibility is analogical, and asserts only that absence from a record may reflect the instrument's coupling rather than the subject's significance. `[ILLUSTRATIVE]`