Is The Universe Finite?

A new analysis of cosmic microwave background data is reviving a long-running question in cosmology: is the universe finite and “closed,” or infinite and flat? The claim hinges on gravitational lensing imprinted on the cosmic microwave background (CMB)—the leftover glow from the early universe. By finding more lensing than expected under an open, low-density geometry, the researchers argue that the amount of matter and energy inferred from the CMB is high enough to curve space inward, turning a previously favored flat picture into a finite hypersphere.

For years, measurements of the CMB’s speckled temperature pattern—especially the distribution of feature sizes summarized by the CMB power spectrum—have pointed toward a geometrically flat universe. Earlier work using the largest CMB “blobs” found curvature consistent with zero, though small positive or negative curvature could still hide within measurement uncertainty. The new study goes further by analyzing essentially all of the CMB structure rather than only the biggest features, and by modeling how gravitational lensing smooths the power spectrum. Light from the CMB does not travel through empty space; it passes through galaxies and galaxy clusters whose gravity deflects the paths of photons. That “lumpy glass” effect blurs the CMB features, and the amount of blurring becomes a proxy for how much mass—dark matter and ordinary matter alike—lies along the line of sight.

The researchers report that the lensing signal in the Planck satellite data is larger than what an open universe would predict. In their framework, that extra lensing implies a higher energy density, which in turn tends to produce positive curvature—space that eventually wraps back on itself. They fit a range of cosmological models that include the expansion rate, inflation-related parameters, and the behavior of different forms of mass and energy, and they find models with positive curvature that match the observed CMB power spectrum. The analysis is described as having greater than 99% statistical confidence for positive curvature, meaning that—assuming the modeling choices are correct—flat geometry would be unlikely to mimic the observed positively curved signature just from random noise.

But the story is not a clean win for a closed universe. A separate lensing diagnostic using the four-point correlation function points back toward the older, lower-energy-density, flat-universe result. That mismatch raises three possibilities: the universe really is closed; key assumptions in the model are wrong; or subtle issues in the Planck data processing—such as how microwave foregrounds were subtracted—introduced a bias. The discussion also connects geometry to the “crisis” in cosmology: Planck-derived expansion rates do not align with modern expansion rates inferred from supernova measurements, and a closed geometry would intensify that discrepancy. Even if space is finite, accelerated expansion and the presence of horizons mean recollapse is not guaranteed under current physics.

The upshot is a growing tension between CMB-based inferences and other cosmological measurements. Future observations are expected to determine whether the curvature signal is real, whether hidden systematics or modeling assumptions are at fault, or whether new physics—possibly involving evolving dark energy—must be added to reconcile the numbers.