What Are Cosmic Voids, and How Empty Are They Really?

Picture the universe as a sponge. The soapy walls and threads hold nearly all the galaxies, packed into filaments, walls, and clusters. The bubbles in between are the cosmic voids: enormous cavities spanning roughly 20 to 100 megaparsecs across, with an average density contrast of about −0.9. In plain words, they hold barely a tenth of the matter you’d expect from a uniform universe. Together, voids fill about 70% of the volume of the present-day cosmos.

These deserts weren’t always empty. Tiny density ripples right after the Big Bang set the pattern. Over billions of years, gravity drained matter from the under-dense patches and piled it onto the walls and filaments around them, inflating the voids as the cosmic web took shape. If the early universe fascinates you, our piece on Andromeda XXXVI as a window into the early universe makes a fine companion read, and the strange political history of Big Bang science comes alive in our article on Soviet science under Stalin.

Now, “empty” is doing heavy lifting in that description. Galaxies do live inside voids. They tend to be smaller, bluer, gas-rich, and busy forming stars from pristine material, precisely thanks to the calm around them. The question the new study asked is simpler and sharper: do these lonely galaxies ever gather into genuine, gravitationally bound groups?

How Do You Hunt for Galaxy Groups in Empty Space?

A team led by María Argudo-Fernández at the University of Granada, working within the Calar Alto Void Integral-field Treasury surveY (CAVITY), took on the challenge. Their study, accepted by Astronomy & Astrophysics in June 2026, examined 24,931 void galaxies living in 170 voids at redshift below 0.08, drawn from the Pan et al. (2012) void catalogue built on Sloan Digital Sky Survey data. Those voids have a median effective radius near 17.83 h⁻¹ Mpc.

Standard group-finding tools struggle here. Simulations show voids are emptier of galaxies than of mass, so density-based methods tuned for crowded regions can mislead. The team went with a friends-of-friends style approach, analysing each void on its own. Think of tracing constellations: you link galaxies that sit close together and move together, and chains of links become groups. Two filters did the work. First, candidate neighbours needed a line-of-sight velocity difference within ±500 km/s inside 1 Mpc of projected distance. Second, and this is the strict part, a physical bond required a separation of 450 kpc or less with velocities matching within ±160 km/s.

Once groups emerged, four classic dynamical quantities described their state. We’ve laid them out for you below, plain-language captions included.

One honest caveat, straight from the authors: with only a handful of galaxies per group, these estimates carry real uncertainty. The de-projection factor √3 assumes tidy, isotropic motions that small groups may not have. Science that admits its error bars is science you can trust.

What Did Astronomers Actually Find Inside 170 Voids?

The numbers tell a lovely story on their own.

Most void galaxies, 59% of them, roam without a single bound companion. Flip the environment and the picture inverts: in the control sample of galaxies living neither in voids nor in clusters (the “NCNV” sample), 60% belong to groups. Voids genuinely discourage company. Yet they never forbid it. Groups turned up in every single void, and, after correcting for observational bias, group richness showed no dependence on void density across the whole measured range of 0.001 to 0.017 galaxies per cubic h⁻¹ megaparsec.

There’s a fascinating wrinkle, too. Isolation is harder to earn than it looks. Of those 14,672 singlets, only about 7% (1,013 galaxies) qualify as potentially isolated, meaning nothing has appreciably disturbed them for around 5 billion years. And the densest groups avoid the innermost 60% of the void radius, hinting that as voids expand, their heftier groups drift toward the surrounding walls. Speaking of galaxies that keep surprising us, don’t miss our story asking whether galaxies can exist without dark matter.

Why Are Void Groups So Loose and Young?

Here’s where the dynamical parameters earn their keep. Compare the median values side by side.

Every arrow points the same way. Void groups are what astronomers call loose groups: velocity dispersions well under 200 km/s and members separated by distances many times their own size. Our own Local Group and the Leo Triplet belong to this family. The median crossing time of 0.66 dwarfs the 0.016 typical of compact groups, and even beats the 0.44 of the control sample. For reference, a system counts as relaxed, or virialised, when its crossing time sits far below the age of the universe, typically between 0.1 and 0.5. Void groups haven’t got there yet. They’re still finding their feet.

Are they babies, or permanent drifters?

The authors weigh two readings honestly. Perhaps void groups are simply young, destined to tighten up as members interact and merge. Or perhaps the void itself, with its global gravitational calm, keeps them in a permanent state of pseudo-equilibrium, sparse forever. One clue tips the scale: the merger rate inside voids today runs near 3%, comparable to filaments and walls. Interactions do happen, just later than in crowded places. That favours the first reading, groups caught in an early evolutionary stage, and it matches independent findings that mature, evolved groups concentrate in clusters and superclusters while poor, young ones scatter everywhere.

Do Void Groups Hide Extra Dark Matter?

Now the tantalising part. The median mass-to-light ratio of void groups lands near 105 h in solar units, roughly double the 50 h reported for compact groups. Taken at face value, that hints at a larger dark matter fraction wrapped around these sparse little systems, in line with what several recent simulations predict for void environments.

We’d love to hand you a headline here, and we won’t. The virial masses rest on velocity dispersions measured from two to six galaxies, so the uncertainties are wide, and the control sample shows a similar median at matched richness. The team says so plainly: no strong conclusion yet. A dedicated hunt for faint satellite galaxies around these groups would sharpen the answer. What the study does deliver is the halo occupation relation, N_gal ∝ M_vir^β, extended into the emptiest environments ever probed, with β = 0.71 ± 0.13, right where theory expects it to sit, below 1.

What Should We Take Away From All This?

Let’s gather the threads. Cosmic voids fill most of the universe’s volume yet hold almost none of its galaxies. Even so, a careful search through 170 of them found 1,367 bound galaxy groups, small families of two to six members, loose, light, slow-moving, and dynamically young. Their richness ignores how empty the surrounding void is. Their mass-to-light ratios whisper, without shouting, about generous helpings of dark matter. And their unhurried evolution offers us something precious: a laboratory where galaxies grow up with minimal interference, showing us what nature does when nobody crowds the room.

Next time you look at a map of the cosmic web, spare a thought for the bubbles between the threads. They aren’t failures of creation. They’re quiet neighbourhoods where gravity still works its patient magic, two or three galaxies at a time. What else might be waiting in the dark, just below our detection limits? Come back to FreeAstroScience.com soon, we’ll keep digging together, and your knowledge of the cosmos will keep growing with every visit.