A mathematician in London is using entropy to explain why the universe has any structure at all

At Queen Mary University of London, the mathematician Ginestra Bianconi has spent the past several years building a case that gravity is not a fundamental force in the conventional sense but a statistical phenomenon, the way that temperature is. Her argument, which appeared in coverage on 17 July 2026, is that the pull of mass is a manifestation of entropy, the same tendency toward disorder that drives heat to flow from hot bodies to cold and ice to melt in a warm room. If she is right, the smooth, galaxy-strewn cosmos is not so much a designed object as the most probable outcome of a system that cannot help but clump.

The proposal lands inside one of the longest-running puzzles in theoretical physics. The standard account treats gravity as a curvature of spacetime, formalised by Einstein's general relativity, and treats entropy as a bookkeeping device for counting microscopic configurations. The two ideas share no obvious mathematical scaffolding. Bianconi's move is to graft one onto the other: she treats spacetime geometry itself as an emergent description of how information is distributed, the way pressure is an emergent description of molecular collisions. Gravity, on this view, is what entropy looks like when it acts on the cosmic scale.

The argument in one diagram

The intuition begins with Jacob Bekenstein and Stephen Hawking's work in the 1970s, which established that black holes carry entropy proportional to the area of their event horizons. That result has haunted physics ever since, because it suggests that information about a region can be encoded on its boundary rather than in its volume. Bianconi extends the logic to the whole universe. If you treat the universe as a network of nodes exchanging information, the most likely configurations of that network are not uniform. They are clumpy. Mass collects where mass already is. Galaxies form where galaxies already are. The structure we see in deep-field images is, on this reading, the statistical signature of a system settling into its most probable state.

This is not a metaphysical claim. Bianconi is a network theorist by trade, and her tool of choice is the framework that physicists use to study disordered systems, from spin glasses to the wiring of the human brain. In her formulation, gravity emerges as an entropic force acting on a network whose connections obey quantum statistics. The mathematics is elaborate, but the underlying idea is the one the nineteenth-century physicist Ludwig Boltzmann committed to his tombstone: entropy is the thing the universe cannot avoid.

Why the second law matters here

The second law of thermodynamics holds that entropy in a closed system never decreases. The arrow of time points from order to disorder. Cosmologists have long struggled to reconcile this with the fact that the early universe was extraordinarily smooth, and that it has since clumped into stars, galaxies, and clusters. Something had to break the symmetry, and the standard account leans on quantum fluctuations during inflation, the brief period of exponential expansion shortly after the Big Bang, amplified over billions of years by gravity itself.

Bianconi's proposal offers a different emphasis. If gravity is itself an entropic phenomenon, then the clumping of matter is not an exception to the second law but an expression of it. The universe does not need a special initial condition to produce structure; it needs only to behave like every other statistical system and seek the most probable arrangement. Galaxies are what entropy looks like when it has a long time to work.

This is a subtle reframing rather than a contradiction. Inflationary cosmology and entropic gravity are not mutually exclusive. The former describes a mechanism for seeding perturbations; the latter describes the force that grew those perturbations into the cosmic web. If the entropic account holds, it would push gravity out of the fundamental category and into the emergent one, alongside temperature, pressure, and other properties that feel real but are ultimately statistical.

A contested corner of theoretical physics

The idea that gravity is emergent has a troubled history. The physicist Erik Verlinde proposed a version of entropic gravity in 2009, and the proposal has been debated, refined, and partially absorbed into the literature ever since, without settling into consensus. Critics have pointed out that the most discussed versions struggle to reproduce the detailed predictions of general relativity in the regimes where those predictions are most precisely tested, including the perihelion precession of Mercury and the bending of light by the Sun. Bianconi's network-based formulation is newer and more abstract, and the question of whether it survives contact with observational cosmology remains open.

The Phys.org coverage that surfaced the work does not claim vindication. It reports the proposal as a contribution to a long-running debate, which is the appropriate register for any single theoretical paper. What is unusual is the framing. Most discussions of entropic gravity proceed from black-hole thermodynamics and work outward. Bianconi proceeds from network theory and information geometry and works inward. The convergence of two such different starting points on similar conclusions is itself a fact about the state of fundamental physics: the discipline is short on data that distinguishes between competing accounts of gravity, and rich on mathematical frameworks that can be made to fit the existing observations.

What would have to be true for it to be right

For the entropic-gravity programme to displace general relativity, it would have to do more than reproduce the classical tests. It would have to predict something different, and then be confirmed. Candidate observations include the behaviour of gravity at very small scales, where the classical theory is silent, and the detailed statistics of the cosmic microwave background, the faint afterglow of the Big Bang that surveys the universe at roughly 380,000 years old. Existing data from the Planck satellite and, more recently, the Atacama Cosmology Telescope are precise enough to constrain some versions of the entropic picture, but the constraints are not yet tight enough to rule the programme out. New instruments, including the Simons Observatory and the planned CMB-S4, are expected to sharpen the picture over the next several years.

The deeper stake is conceptual. If gravity is emergent, then the search for a unified theory of physics looks different. The holy grail of the field, a quantum theory of gravity that reconciles general relativity with quantum mechanics, may not take the form physicists have assumed. It may instead resemble condensed-matter physics, the branch of the discipline that studies emergent phenomena in solids, liquids, and gases. That would be a quieter kind of revolution than a single dramatic equation. It would amount to the recognition that the force holding galaxies together is the same kind of phenomenon as the force pushing heat out of a cup of coffee.

What remains uncertain

The sources do not specify which observational data Bianconi considers the most decisive test of her framework, and they do not name collaborators beyond the institutional affiliation. The paper itself, which the Phys.org coverage references, would need to be examined directly to assess the technical content. The wider point stands on its own: a working theoretical physicist at a major London research university has now put forward a version of the entropic-gravity programme, and the proposal is in circulation. Whether it joins the small set of ideas that shape how physicists think about the early universe, or whether it is absorbed and forgotten, will be decided by work that has not yet been done.

This piece focused on the theoretical structure of the proposal rather than on the author's biography, because the available sources centre on the argument. Monexus will revisit the empirical status of entropic-gravity programmes as new cosmic-microwave-background data is released.