Institute of Integrative and Interdisciplinary Research

Whether gravity has a bottom, what becomes of the energy that makes up most of your mass, and whether the universe turns one way — three open questions with instruments already pointed at them

Most of the famous unanswered questions in physics have no test attached. What happened before the Big Bang, why there is something rather than nothing, whether the constants could have been otherwise — these are wonderful to argue about and there is no instrument being built that will settle any of them.

A smaller set of questions is different. They are just as fundamental, they are genuinely open, and each one has a particular observation attached to it that will come out one way or the other. Three of them are below. None requires new physics to be invented before the test can be run. All three could be closed within a decade or two, and one of them could be closed in an afternoon by whichever star in our galaxy happens to explode next.

1. What happens to ninety-nine per cent of your mass if you squeeze it hard enough?

Start with a fact that is measured rather than theorized, and that still surprises people who have not met it.

Almost none of your mass comes from the Higgs field. A proton weighs about 938 MeV. Add up the masses of its three valence quarks and you get roughly nine. The missing 99 per cent is gluon field energy and chiral condensate — energy that exists because the strong force is holding it inside a volume about a femtometre across. Mass, for ordinary matter, is mostly confinement.

Now compress that matter. At sufficiently high density the strong force can no longer maintain confinement, and the quarks and gluons are no longer organized into separate nucleons. This happens somewhere between three and ten times the density of an atomic nucleus, and the reason nobody can say where is worth knowing: the calculation is blocked. Lattice QCD, the method that works beautifully for the hot, low-density corner of the phase diagram, breaks down at high baryon density because the mathematics develops a sign problem that no one has solved. Heavy-ion colliders reach deconfinement by heating rather than squeezing, which is the wrong corner. Cold dense quark matter is one of the few regimes in fundamental physics that is neither calculable nor reproducible in a laboratory.

So: what becomes of that 99 per cent?

There are exactly three possibilities, and the trichotomy is not a matter of taste. Energy that stays in the volume and is carried by particles with definite momenta contributes ordinary pressure. Energy that leaves is radiation. Energy that stays and is not carried by anything with a definite momentum has no rest frame with which to define a momentum flux, and a contribution to the stress tensor with no preferred rest frame is forced to look like vacuum energy: negative pressure, equal in magnitude to its energy density.

The third possibility is the interesting one, because negative pressure gravitates backwards. In general relativity the quantity that sources gravitational attraction is not the density alone but the density plus three times the pressure. Make the pressure sufficiently negative and the sign flips. The material repels.

How negative is sufficient? A recent analysis works it out in two lines. Split the deconfined medium into a kinetic part and a non-localized part, and call the non-localized fraction f. The effective equation of state is (1 − 4f)/3, and gravity turns repulsive when that falls below −1/3, which happens when f exceeds one half.

More than half the energy density has to be the non-localized kind. Standard accounting does not come close: in the bag model at five times nuclear density, f lands between 0.07 and 0.21, and the effective pressure is still positive. Worse, the shortfall grows under compression, because the kinetic term scales as the fourth power of the chemical potential while the vacuum term does not. Squeezing harder dilutes exactly the fraction you need.

That is the state of the question: a specific number, a specific gap, and a specific thing that would have to be true to close it — that the released confinement energy never becomes kinetic at all, because what carried it was never a set of localized objects in the first place. Nobody has shown that. Nobody has shown it is false either.

2. Is there anything between the heaviest neutron star and the lightest black hole?

This is the question the first one feeds into, and unlike the first it has a deadline.

Neutron stars have a maximum mass, somewhere around 2.2 to 2.3 solar masses, set by the point at which no known equation of state can hold the star up. Above it, according to the standard account, collapse runs to completion and a black hole forms. Whether anything can exist between that limit and the lightest black holes formed by stellar collapse is not currently known, and it is no longer an idle question, because objects keep turning up in exactly that range: the 2.6-solar-mass companion in the GW190814 merger, the compact object in GW230529, the heaviest pulsars creeping toward 2.3.

Are these light black holes, or are they something else?

If collapse does halt at a deconfinement density, the arithmetic that follows is unusually clean. An object whose interior sits at a fixed density has a radius that grows as the cube root of its mass. Its Schwarzschild radius grows linearly. So the ratio of the two is exactly (M/M_tr)^{2/3}, where M_tr is the mass at which they meet and a horizon finally closes. For a halt at five times nuclear density, M_tr comes out near 3.6 solar masses.

That leaves an interval roughly 1.4 solar masses wide in which the two accounts of the same object disagree about whether it has an event horizon. Not about a rate, or a spectrum, or an efficiency — about whether there is a horizon. This is the least negotiable disagreement two theories of the same thing can have, which is what makes it decidable by a crude observation.

Here is the crude observation. When a collapsing stellar core forms a black hole, its neutrino signal does not fade out. It stops, within tens of microseconds, on the light-crossing time of the emitting region, and it stops while the luminosity is still rising. That abruptness is the signature of a horizon closing over the emitting surface. If collapse halts instead, there is nothing to truncate the signal, and it decays over seconds like any hot object with a surface.

The next galactic supernova therefore performs a three-way test on a single observable. A brief second feature followed by a cutoff indicates a temporary stiffening — a quark-matter bounce that delays collapse without preventing it. A second feature followed by a cooling tail indicates that collapse has a permanent floor. No second feature at all, followed by a cutoff, leaves the textbook picture standing.

Detectors are ready for this now. A galactic event at ten kiloparsecs would give IceCube about a million counts with sub-millisecond timing, Super-Kamiokande around ten thousand events with energies, and DUNE the electron-neutrino channel that nobody else has. The measurement that actually discriminates plays out over hundreds of milliseconds, which is the easy part.

The inconvenience is that galactic supernovae happen a few times per century, and the last one was in 1604.

Meanwhile there is a second, weaker test that does not wait. If such objects exist below M_tr, then a horizonless remnant has a surface, and a surface can oscillate. A trapped-mode analysis shows that the frequency of any such oscillation inherits the geometry: it falls as the object gets heavier, because the gravitational redshift at the surface collapses faster than the radius grows, and it switches off entirely at M_tr. The absolute frequency depends on a quantity nobody can compute, but that quantity cancels out of ratios. Between 2.2 and 3.5 solar masses the frequency must drop by a factor of 4.03, with nothing adjustable in it. Neutron star modes do the opposite, rising with compactness. Two objects in that mass window with measured masses and a measured oscillation frequency would settle it, and the analysis can be run on archival gravitational-wave posteriors by anyone who wants to.

3. Does the universe have a preferred handedness?

The third question is unrelated to the first two in physics and closely related in structure: a symmetry that is assumed rather than measured, with a measurement now arriving.

Cosmology assumes parity symmetry. The statistical distribution of cosmic structure is taken to be unchanged under mirror reflection, so that left-handed and right-handed configurations occur equally often. Ordinary gravitational evolution cannot violate this, because the equations treat a configuration and its mirror image identically. A detected handedness would therefore point at something primordial, or at an unusual global shape for space, or at a systematic error — and each of those is worth knowing about.

Testing it has been frustrating. Searches using the winding direction of spiral galaxies have returned conflicting results for fifteen years, because spiral handedness is contaminated by how the galaxy is tilted, how it was imaged, and who classified it. Searches using the parity-odd part of the four-point galaxy correlation function initially looked promising and then turned out to depend sensitively on covariance modelling.

A new approach from Pedro da Silveira Ferreira and Renyue Cen sidesteps both problems. The cosmic web has nodes where several filaments meet, and a node provides something a galaxy does not: an unambiguous centre, from which each filament arm has a well-defined outward direction, and therefore a well-defined sense of winding. These structures span about sixty megaparsecs, which puts them far above the scales where baryonic physics muddies the signal. The authors measured both the amount of winding and its handedness across two independent reconstructions of the sky, and looked not only for a global preference but for a preferred axis — a direction about which the handedness organizes itself, changing sign between the two hemispheres of the sky.

They found nothing. The universe is even-handed to the precision of the measurement, as the standard model expects.

The interesting question is what that precision is. An analysis of the sensitivity against a specific competing prediction — a handedness dipole growing with distance, sourced by two cancelling rotations of a projective spatial geometry — finds the null is between eight and sixty-one times too coarse to see it, depending on how you translate between galaxy spin and filament winding. The channel is right; the sample is small.

What would not be too coarse is the next generation of surveys. DESI already provides an order of magnitude more spectroscopic tracers than SDSS, and Euclid, Rubin, and CSST will add more still. The sensitivity analysis puts the requirement at roughly 1.2 million filament arms if the estimator is evaluated along a direction specified in advance, and about 6 million if the analysis has to scan the whole sky for the best axis. That factor of five is the entire practical value of a theory naming a direction before the data arrive, and it is a decent argument for pre-registration in cosmology generally.

What these three have in common

Each of them is a question about something fundamental — the origin of mass, the end state of collapse, the symmetry of space. None of them is settled. And each has a specific instrument attached to it, with a number that could come out on either side.

They also share a structural feature worth pointing out. In all three cases the decisive quantity is not an absolute value but a ratio or a shape: the fraction of energy that is non-localized, the way an oscillation frequency changes with mass, the amplitude relative to a null. Absolute values in this subject usually depend on whatever is least known. Ratios and shapes survive the ignorance, and they are what you should look for when someone tells you a prediction is testable.

The last galactic supernova was observed by Kepler. Statistically another one is overdue, and when it arrives the neutrino detectors will have about ten seconds of data that answer a question people have been asking since Oppenheimer. It is an odd thing to be waiting for. But it is a good deal better than waiting for an answer with no instrument attached to it at all.

References:

Kriger, B. (2026, September). How far below the first cosmic-web handedness null does the counter-rotation chirality dipole sit? IIIR Cosmology and Theoretical Physics. https://doi.org/10.13140/RG.2.2.27645.12009

Kriger, B. (2026, September). Losing the jets, keeping the zones: A withdrawal and a replacement on the K-limit line, occasioned by Mameda and Sogabe (arXiv:2609.07835). IIIR Cosmology and Theoretical Physics. https://doi.org/10.13140/RG.2.2.17159.36004

Kriger, B. (2026, September). The vacuum fraction and the cutoff: What the confinement energy must do to arrest collapse, and how a galactic supernova would tell us whether it does. IIIR Cosmology and Theoretical Physics. https://doi.org/10.13140/RG.2.2.22821.67044

Boris Kriger is Lead Investigator at the Institute of Integrative and Interdisciplinary Research, Toronto, and a Research Fellow at the Information Physics Institute, Gosport. ORCID: 0009–0001–0034–2903