The label “dark matter” encapsulates our ignorance regarding the nature of most of the matter in the universe. It contributes five times more than ordinary matter to the cosmic mass budget. But we cannot see it. We infer its existence only indirectly through its gravitational influence on visible matter.

The standard model of cosmology successfully explains the gravitational growth of present-day galaxies and their clustering as driven by primordial fluctuations in an ocean of invisible particles with initially small random motions. But this “cold dark matter” might actually be a mixture of different particles. It could be made of weakly interacting massive particles; hypothetical particles like axions; or even dark atoms that do not interact with ordinary matter or light. We have not detected any of these invisible particles yet, but we have measured the imprint of the fluctuations in their primordial spatial distribution as slight variations across the sky in the brightness of the cosmic microwave background, the relic radiation left over from the hot big bang.

Many experiments are searching for the signatures of various types of dark matter, both on the sky and in laboratory experiments, including the Large Hadron Collider. This search has so far been unsuccessful. In addition to specific types of elementary particles, primordial black holes have been mostly ruled out as a dominant component of dark matter, with a limited open window in the range of asteroid masses waiting to be eliminated.

In a 2005 paper, I showed, with Matias Zaldarriaga, that cold dark matter particles could cluster gravitationally on scales down to an Earth mass. Evidence for such tiny clumps of dark matter has not been found yet; observers have only studied much bigger systems, namely galaxies like our own Milky Way, containing gas and stars as their inner core, which is surrounded by a halo of dark matter.

As revealed by the pathbreaking work of Vera Rubin, the dynamics of gas and stars in galaxies indeed imply the existence of invisible mass in a halo that extends well outside the inner region where ordinary matter concentrates. Surprisingly, the need for dark matter in galaxies like the Milky Way appears only in the outer region where the acceleration drops below a universal value, which equals roughly the speed of light divided by the age of the universe. This is an unexpected fact within the standard dark matter interpretation. The fundamental flavor of a universal acceleration threshold raises the possibility that perhaps we are not missing invisible matter but rather witnessing a change in the effect of gravity on the dynamics of visible matter at low accelerations.

This was the idea pioneered by Moti Milgrom, who in 1983 proposed a phenomenological theory of “modified Newtonian dynamics” (MOND) to explain away the dark matter problem. Remarkably, his simple prescription for modified dynamics at low accelerations accounts for the nearly flat rotation curves in many galaxy halos extremely well, even after four decades of scrutiny. As expected in MOND, all existing data on Milky Way–size galaxies shows a tight correlation between the circular speed in the outskirts of galaxies and the total amount of ordinary matter (also labeled, baryonic matter), manifesting the so-called “baryonic Tully-Fisher relation.” In a 1995 paper, I showed with my first graduate student, Daniel Eisenstein, that the tightness of this relation is not trivially explained in the standard dark matter interpretation. Even if dark matter exists, MOND raises the fundamental question: why do the dark matter particles introduce a fundamental acceleration scale to the dynamics of galaxies? Is this an important hint about their nature?

MOND faces challenges on scales larger than galaxies. More massive systems such as galaxy clusters— where Fritz Zwicky first posited dark matter’s existence and coined its name—show evidence for missing mass even though their acceleration tends to be above the threshold scale in MOND. Moreover, the acoustic oscillations detected to exquisite precision in the brightness fluctuations of the cosmic microwave background, imply the presence of a dominant component of matter that streams freely, in addition to the ordinary matter and radiation fluids that are tightly coupled by electromagnetic interactions.

But what about the smallest scales? Together with my postdoc Mohammad Safarzadeh, I studied recently the latest data available from the Gaia survey of ultrafaint dwarf galaxies that are satellites of the Milky Way. We showed that their behavior deviates from MOND’s expectations. Just like clusters of galaxies, dwarf galaxies appear to argue against the universality of MOND on all scales.

Does the success of MOND on Milky Way scales and its failures on both smaller and larger scales offer new insights about the nature of dark matter? One possibility is that dark matter is strongly self-interacting and avoids galactic cores. With Neal Weiner, I showed in a 2011 paper that a dark sector interaction resembling the electric force between charged particles could facilitate the avoidance of galactic cores by dark matter, with a diminishing effect at the high collision speeds characteristic of galaxy clusters.

Another possibility that I suggested with Julian Muñoz in a 2018 paper, was inspired by the EDGES experiment, which reported unexpected excess cooling of hydrogen atoms during the cosmic dawn. We showed that if some dark matter particles possess a small electric charge, they could scatter off ordinary matter and cool hydrogen atoms below expectations, as reported.

Explaining one anomaly by the conjecture that a fraction of the dark matter particles are slightly electrically charged is far more speculative than explaining six anomalies by the conjecture that the interstellar object ‘Oumuamua is a thin film pushed by sunlight. Nevertheless, speculations on the nature of dark matter receive far more federal funding and mainstream legitimacy than the search for technosignatures of alien civilizations.

More definitive clues are needed to figure out the nature of dark matter. Here’s hoping that the coming decades will bring a resolution to this cosmic mystery, with all pieces of the jigsaw puzzle falling into place. Alternatively, we might seek a smarter kid on the cosmic block who would whisper the answer in our direction. Although it might feel like cheating in an exam, we should keep in mind that there is no teacher in sight looking over our shoulders.

This is an opinion and analysis article.