Dark Matter Is Not What Scientists Thought — And the Real Answer Is More Disturbing

[00:00:00]For most of the 20th century, the dominant assumption in particle physics was that dark matter would eventually be found in the form of a single, well-defined particle, massive, electrically neutral, leftover from the early universe in exactly the quantity needed to account for the gravitational effects observed in galaxies and galaxy clusters.

[00:00:24]This assumption drove the construction of the largest, most sensitive particle detectors ever built. It shaped the experimental program at the Large Hadron Collider. It defined the career trajectories of hundreds of physicists across multiple decades. And so far, after more than 40 years of progressively more sensitive searches, it has produced nothing.

[00:00:49]Not a null result that narrowly missed, a comprehensive, systematic, multi-decade failure to find the particle that the best theoretical models said should have been detectable by now. What is emerging in the wake of that failure is a picture of dark matter that is considerably stranger, considerably less certain, and in several specific ways considerably more disturbing than the tidy particle physics story most people have been told.

[00:01:21]The history of dark matter begins not in particle physics, but in astronomy. In 1933, the Swiss-American astronomer Fritz Zwicky measured the velocities of individual galaxies within the Coma galaxy cluster and found that they were moving far too fast to be held together by the gravitational pull of the visible matter in the cluster.

[00:01:45]The cluster's galaxies were flying around at velocities that, given the amount of luminous mass present, should have sent them scattering off into the cosmos long ago.

[00:01:56]Zwicky proposed that some form of Dunkel Materie, dark matter, must be providing additional gravitational mass that his instruments could not see.

[00:02:07]His result was largely ignored for decades, treated as a curiosity without obvious physical explanation or theoretical backing.

[00:02:16]The confirmation that galactic dynamics genuinely require something unseen came four decades later through the meticulous work of Vera Rubin and Kent Ford at the Carnegie Institution, who measured the rotation curves of spiral galaxies, the way stars at different distances from a galaxy's center orbit around it, and found galaxy after galaxy that stars far from the center rotate at roughly the same velocity as those close in, rather than slowing down as Kepler's laws would predict for a system where most of the mass is concentrated near the center.

[00:02:58]This flat rotation curve is the direct observational signature of a dark matter halo surrounding each galaxy, an invisible component of matter distributed in a roughly spherical cloud extending well beyond the visible disc of stars, providing the gravitational mass needed to maintain the observed rotation velocities.

[00:03:22]By the 1980s, the theoretical framework supporting dark matter had crystallized around the specific class of candidates called weakly interacting massive particles, or WIMPs.

[00:03:36]The WIMP hypothesis was attractive for a specific reason. Beyond simply fitting the astronomical observations, it emerged naturally from independent calculations in supersymmetric extensions of the standard model of particle physics.

[00:03:53]Supersymmetry, a theoretical framework that postulates a fundamental symmetry between fermions and bosons, predicts the existence of a lightest supersymmetric particle, typically called a neutralino, that would be stable, massive, electrically neutral, and produced in the early universe in approximately the right quantity to account for the observed dark matter abundance.

[00:04:22]This apparent coincidence, that a particle predicted by supersymmetry for completely different theoretical reasons, would also naturally explain the cosmological dark matter abundance, was known as the WIMP miracle. And it gave the WIMP hypothesis a compelling theoretical basis that went beyond simply fitting observations after the fact.

[00:04:47]The experimental program to detect WIMPs has been conducted through three complementary approaches.

[00:04:53]Direct detection experiments attempt to detect the recoil of atomic nuclei struck by passing dark matter particles.

[00:05:02]Underground detectors like LUX in the Homestake Mine in South Dakota, XENON at Gran Sasso in Italy, PandaX in China, and the current XENON100 and LUX experiments have progressively improved sensitivity by orders of magnitude, with LUX representing sensitivity to WIMP nucleon cross sections more than a thousand times smaller than the sensitivity available in the 1990s.

[00:05:33]Indirect detection experiments search for the products of dark matter annihilation in regions of high dark matter density, like the galactic center or dwarf spheroidal galaxies. The Fermi Gamma-ray Space Telescope and ground-based Cherenkov telescopes like Hess and Magic have provided the most sensitive searches of this type.

[00:05:56]Collider experiments at the LHC search for WIMP pair production in proton-proton collisions where the missing energy signature of neutral weakly interacting particles escaping the detector would provide a distinctive signal.

[00:06:13]None of these three complementary approaches has found the WIMP signal.

[00:06:17]The LHC found no supersymmetric particles across more than a decade of operation and hundreds of inverse femtobarns of data at collision energies up to 13.6 teraelectronvolts.

[00:06:32]The direct detection experiments have excluded the most theoretically motivated regions of WIMP parameter space to the point where simple neutralino dark matter is effectively ruled out for the mass range where it was most strongly predicted.

[00:06:49]The indirect detection experiments have constrained dark matter annihilation cross-sections to levels below what simple thermal relic WIMPs require. This comprehensive experimental failure does not by itself rule out WIMPs. There remain viable WIMP candidates in less theoretically motivated parameter space and the direct detection experiments will continue improving sensitivity, but it has substantially eroded the theoretical motivation for WIMPs as the natural, well-motivated dark matter candidate and has driven a significant fraction of the particle physics and cosmology communities to consider alternatives seriously that were previously regarded as fringe or exotic. The most thoroughly studied alternative to WIMPs is the axion, a particle originally proposed not as a dark matter candidate, but as a solution to an entirely different problem, the strong CP problem in quantum chromodynamics, by Roberto Peccei and Helen Quinn in 1977, with the observational implications for a new light particle worked out independently by Frank Wilczek and Steven Weinberg. The axion is extraordinarily light, with a mass potentially as small as 1 millionth of an electron volt, about a trillion times lighter than a WIMP, but it can be produced in the early universe in large enough quantities to account for the observed dark matter if the Peccei-Quinn symmetry was broken during inflation.

[00:08:31]Axion dark matter would form not a gas of discrete particles drifting through galaxies, but a coherently oscillating quantum field, more analogous to a classical wave than a collection of individual particles, occupying all of space with a slight, rapidly oscillating density variation.

[00:08:52]The experimental program to search for axions, while less well funded and less mature than the WIMP program, has made significant progress in recent years. The ADMX experiment at the University of Washington uses a high-Q microwave cavity and a strong magnetic field to convert axions into detectable photons through the Primakoff effect, and has now excluded axion masses in a specific frequency range consistent with theoretical predictions. Sterile neutrinos represent a third candidate class, heavier cousins of the known neutrinos, which do not interact through the weak force, but can mix with ordinary neutrinos through quantum mechanical oscillation, producing a distinctive x-ray spectral line when they decay.

[00:09:46]In 2014, observations with the XMM-Newton and Chandra X-ray Observatories reported a tentative detection of an anomalous spectral line at 3.5 keV in the x-ray spectra of galaxy clusters and the Andromeda galaxy, consistent with the decay signature expected from a sterile neutrino with a mass of roughly 7 keV.

[00:10:11]This result generated considerable excitement and extensive follow-up investigation, which has produced inconsistent results, some observations confirming the line, others failing to detect it at comparable significance, leaving the situation unresolved as of the current date, and awaiting data from future x-ray telescopes with higher spectral resolution.

[00:10:38]The most disturbing possibility in dark matter research, and I want to be precise about why it is disturbing in a way that the alternatives are not, is not that dark matter consists of an exotic particle we have not yet found. It is that dark matter might not be a particle at all.

[00:10:58]Modified gravity theories, which propose that the gravitational force law itself deviates from Newtonian and Einsteinian gravity on galactic scales in a way that mimics the effects attributed to dark matter, have never been fully killed by observation, despite several decades of theoretical development and observational testing.

[00:11:20]Modified Newtonian dynamics, or MOND, proposed by Mordehai Milgrom in 1983, postulates that Newton's second law is modified at very low accelerations, below a critical value of about 1.2 * 10 to the minus 10 m/s squared, causing gravity to fall off more slowly than expected at large distances. The simple modification, with a single free parameter, reproduces the observed rotation curves of hundreds of galaxies with a precision that dark matter models struggle to match on a galaxy-by-galaxy basis. The primary observational argument against modified gravity as a complete replacement for dark matter is the Bullet Cluster, a system of two galaxy clusters caught in the act of passing through each other.

[00:12:16]Observed in X-ray wavelengths by Chandra and in gravitational lensing maps reconstructed from distortions in background galaxy shapes, the hot intracluster gas of the two clusters, which makes up most of the visible baryonic mass, has been slowed by ram pressure during the collision and lags behind the galaxies. The gravitational lensing map, however, shows that the mass peaks are centered on the galaxies rather than on the gas, which would be the expected signature if most of the cluster mass is in a collisionless dark matter component that passed through the collision unimpeded as particles do.

[00:12:59]This lensing gas offset is extremely difficult to explain within purely modified gravity frameworks, and it remains the strongest single observational argument for the existence of some form of dark matter as a genuine additional mass component, rather than a modification to gravity.

[00:13:21]But the Bullet Cluster does not identify what form of matter produces the additional mass. Only that something gravitationally significant was present and collisionless during the cluster merger. The discovery of the galaxy NGC 1052-DF2 in 2018 by Pieter van Dokkum and colleagues at Yale University added a genuinely unexpected complication to the dark matter picture from an entirely different direction.

[00:13:53]This galaxy appeared to contain essentially no dark matter at all. Its velocity dispersion was consistent with the gravitational mass of its visible stellar component alone, with no significant dark matter contribution required. A galaxy that seemingly exists without its expected dark matter halo in a universe where all other comparable galaxies appear embedded in massive dark matter halos, does not fit comfortably into either a simple particle cold dark matter framework or a simple modified gravity framework. If dark matter is a universal particle background, why would this particular galaxy lack it?

[00:14:37]If modified gravity is the explanation, why does the rotation curve modification not apply here?

[00:14:45]Subsequent research has complicated the picture further. Some measurements have revised DF2's distance, which affects all the derived mass estimates.

[00:14:56]But the basic discovery of galaxies with apparently anomalous dark matter content has now been extended to a small population of objects with similarly puzzling kinematics, suggesting that whatever dark matter is, its distribution relative to visible matter is more complex and variable than simple models predict. A more recent theoretical development that has attracted serious attention is the possibility that dark matter is not a single particle species but a complex dark sector, an entire set of dark particles with their own forces and interactions that are invisible to ordinary matter except through gravity.

[00:15:39]In the same way that ordinary matter has quarks, electrons, photons, and a rich set of nuclear and electromagnetic interactions, dark matter might have its own internal physics. Dark photons, dark nuclei, dark chemistry, accessible only through its gravitational effects on visible matter.

[00:16:02]This idea, sometimes called mirror dark matter or dark atom models, is much harder to test than single particle dark matter.

[00:16:13]Precisely because the dark sector interactions are not directly accessible to any instrument sensitive only to ordinary matter.

[00:16:23]It is not ruled out by any current observation. In fact, some models within this class can explain features of galaxy structure that simple cold dark matter struggles to reproduce, such as the apparent scarcity of satellite galaxies around large galaxies, and the observed structure of dark matter density profiles at galactic centers.

[00:16:47]But it offers no obvious experimental prediction that would distinguish it from alternatives, which is itself a scientifically uncomfortable position.

[00:16:58]What I find genuinely disturbing about the current state of dark matter research is not the absence of detection. Negative results are scientifically valuable and have certainly narrowed the parameter space significantly.

[00:17:14]It It the possibility that the thing making up roughly 27% of the total mass energy of the universe, the dominant mass component of every galaxy and galaxy cluster we can observe, may be so fundamentally different from anything in the standard model of particle physics, so much more complex or so much more weakly interacting, or so much more entangled with modifications to gravity itself, that no experiment currently conceivable could detect it directly, leaving us with astronomical observations as our only window into its properties.

[00:17:58]We infer dark matter's existence from how it bends light, how it sculpts the large-scale structure of the cosmos, how it determines the rotation of galaxies that formed billions of years ago.

[00:18:13]But after four decades of progressively more sensitive searches, we still do not know what it is, where it came from, or whether the concept of a particle at all applies to it. That uncertainty about something that outweighs all the ordinary matter in the universe combined is, I think, the most consequential unsolved problem in modern physics.

[00:18:38]And the honest answer to what scientists currently think dark matter is is that the most they can say with confidence is that it exists.

[00:18:48]And that everything they expected it to be appears to be wrong.

[00:18:54]I want to revisit several threads with additional depth before closing, because there are dimensions of this problem that a single pass through the evidence cannot adequately convey.

[00:19:06]Return to the WIMP miracle and consider what it means that the miracle has apparently failed to materialize experimentally.

[00:19:15]The theoretical elegance of the WIMP hypothesis, the coincidence that a particle postulated by supersymmetry for reasons unrelated to cosmology would also produce the right dark matter abundance, was not merely aesthetically pleasing.

[00:19:32]It was taken as genuine evidence that WIMPs were the right answer in the same way that physicists take mathematical consistency and theoretical naturalness as genuine evidence about the physical world.

[00:19:46]The failure of WIMP detection experiments to find anything is therefore not simply an experimental negative result. It is evidence that the theoretical reasoning that made WIMPs seem natural was itself wrong or missing something important in a way that undermines a broader theoretical framework, supersymmetry, that many physicists spent decades assuming was correct. This is a significant blow not just to dark matter particle physics, but to the theoretical foundations that supported a generation of theoretical work in particle physics and cosmology.

[00:20:25]Return to the axion and consider the specific physical picture it implies for what dark matter actually is.

[00:20:34]An axion dark matter background filling the galaxy would not behave like a gas of particles.

[00:20:41]It would behave like a coherent quantum field oscillating in time at a frequency determined by the axion mass with a de Broglie wavelength potentially large enough to be macroscopically observable.

[00:20:56]Very light axions with masses around 10 to the minus 22 electron volts, sometimes called fuzzy dark matter, would have de Broglie wavelengths of kiloparsec scale, comparable to the sizes of dwarf galaxy cores, and would produce a characteristic quantum pressure at galactic centers that smooths out the density cusps predicted by cold particle dark matter simulations.

[00:21:24]This quantum pressure signature, the solitonic core that fuzzy dark matter models predict at the centers of small galaxies, is currently a subject of active observational investigation using high-resolution measurements of dwarf galaxy stellar kinematics.

[00:21:43]Return to the modified gravity alternatives and consider the specific theoretical challenge that tensor vector scalar gravity, or TeVeS, developed by Jacob Bekenstein in 2004 as a relativistic version of MOND, faces from observations of the cosmic microwave background.

[00:22:05]The detailed pattern of temperature fluctuations in the CMB encodes information about the acoustic oscillations of the early universe, and the relative heights of successive peaks in this pattern depends sensitively on the amount and distribution of dark matter present in the early universe.

[00:22:24]The CMB acoustic peak heights are beautifully explained by standard cold dark matter models, while modified gravity theories, including TeVeS, struggle to reproduce the same detailed peak structure without invoking some form of additional mass that at least functionally resembles dark matter in the early universe, even if its nature is different. This constraint makes pure modified gravity a difficult sell as a complete replacement for dark matter in cosmological contexts, even while MOND continues to provide the best available description of galactic rotation curves on galaxy by galaxy basis. I want to close with a thought about what the current situation in dark matter research reveals about the relationship between observational evidence and theoretical understanding in modern physics.

[00:23:22]We have overwhelming multi-source independently confirmed evidence that dark matter exists from galaxy rotation curves, from gravitational lensing, from the CMB, from the large-scale structure of the cosmos.

[00:23:38]The evidence for dark matter as a gravitationally significant component of the universe is arguably as strong as the evidence for any well-established fact in modern astrophysics.

[00:23:52]And yet our understanding of what dark matter actually is, what particle or field it consists of, what its mass and interactions are, how it was produced in the early universe remains essentially zero after four decades of dedicated experimental effort. We know that something is there, we know roughly how much of it is there, we know how it is distributed, at least statistically, across the cosmos, and we know essentially nothing else about its fundamental nature. This is a genuinely unusual situation in physics [snorts] where observational certainty about a phenomenon and theoretical understanding of it have so completely diverged.

[00:24:37]Whatever dark matter turns out to be, the story of how we failed to find it for so long, and eventually found it, or found whatever actually replaces it in our understanding, will be one of the central narratives of 21st century physics.