Einstein's Other Theory of Gravity Is Finally Getting Attention

Einstein's Other Theory of Gravity Is Finally Getting Attention

In 1915, Albert Einstein unveiled general relativity, a theory so elegant that it bent light, warped time, and turned gravity into geometry. It was his masterpiece. But here is the secret that most physicists don't tell you: Einstein also had a second theory of gravity. He published it in 1928, and for nearly a century, it was treated like an eccentric uncle at a family reunion. Interesting, sure. But not the main event.

That is changing. In a comprehensive 2022 review published in Reports on Progress in Physics, Sebastián Bahamonde, Konstantinos F. Dialektopoulos, Celia Escamilla Rivera, and Gabriel Farrugia catalogued the quiet revolution happening in what is called teleparallel gravity. The paper has already accumulated nearly 500 citations, a sign that something is shifting in the physics community. The authors argue that teleparallel gravity is not just a mathematical curiosity. It may be the key to solving problems that general relativity cannot touch, from the nature of dark energy to the behavior of gravity at quantum scales.

Here is the strange thing: teleparallel gravity makes the same predictions as general relativity for almost everything we have measured. But it does so by telling a completely different story about what gravity actually is.

The Geometry That General Relativity Left Behind

General relativity works because of a simple, beautiful idea. Mass tells spacetime how to curve, and curved spacetime tells mass how to move. Gravity is not a force. It is the shape of the universe itself.

Teleparallel gravity rejects this entirely. It says gravity is a force. A real force, like electromagnetism. And spacetime itself stays flat.

How is that possible? The answer lies in a mathematical object called torsion. In general relativity, the curvature of spacetime is described by the Riemann tensor, which measures how much spacetime bends. In teleparallel gravity, the curvature is zero. Spacetime is perfectly flat. But there is torsion, a twisting of the geometry that general relativity ignores.

Think of it like this. Imagine you are holding a piece of paper. General relativity says you can bend the paper to create gravity. Teleparallel gravity says you keep the paper flat, but you twist it. The twisting produces the same physical effects as bending. The mathematics is different. The predictions are identical for most experiments.

Einstein himself developed teleparallel gravity in the late 1920s, hoping to unify gravity with electromagnetism. He abandoned the idea when it did not work. For decades, it was a footnote. But in the last twenty years, physicists have realized that the footnote might contain the answer to questions Einstein never got to ask.

Why General Relativity Is Running Out of Road

General relativity has passed every test we have thrown at it. It predicted gravitational waves. It predicted black holes. It predicted the bending of starlight. It works.

But it also has problems. Three big ones.

The first is dark energy. In 1998, astronomers discovered that the universe is expanding faster and faster. Something is pushing it apart. General relativity cannot explain this without adding a fudge factor called the cosmological constant, a number that is 120 orders of magnitude smaller than what quantum mechanics predicts. That is not a small discrepancy. It is the worst prediction in the history of physics.

The second problem is dark matter. Galaxies rotate faster than they should, based on the visible matter they contain. General relativity says there must be invisible matter holding them together. But decades of searching have not found it.

The third problem is quantum gravity. General relativity breaks down at the Planck scale, the tiny realm where quantum mechanics rules. The equations produce infinities. They stop making sense.

Teleparallel gravity offers a way forward on all three fronts. Because it treats gravity as a force, it can be modified more naturally than general relativity. You can add terms to the equations without breaking the theory. You can build models of dark energy that do not require a cosmological constant. You can construct theories of gravity that might be compatible with quantum mechanics.

Bahamonde and his colleagues surveyed the landscape of these modified teleparallel theories. They found dozens of them, each with different assumptions and different predictions. Some are elegant. Some are messy. All of them are testable.

The Gauge Theory That Already Exists

Here is the part that made me sit up straight. Teleparallel gravity is a gauge theory. That sounds technical, but it is actually the most exciting thing about it.

Gauge theories are the language of modern physics. The Standard Model of particle physics is built on gauge theories. Electromagnetism is a gauge theory. The strong nuclear force is a gauge theory. The weak nuclear force is a gauge theory. These theories describe forces as arising from symmetries in nature.

General relativity is not a gauge theory. It is a geometric theory. It stands apart from everything else in physics. This has made it notoriously difficult to unify with quantum mechanics. You are trying to combine two fundamentally different kinds of mathematics.

Teleparallel gravity changes that. Because it treats gravity as a force, it can be formulated as a gauge theory of translations. The authors explain this in detail: "We give a comprehensive introduction to how teleparallel geometry is developed as a gauge theory of translations together with all the other properties of gauge field theory" (Bahamonde et al., 2022). This means gravity can be described using the same mathematical framework as the other fundamental forces.

This is not just neat. It is a potential roadmap to quantum gravity. If gravity is a gauge theory, then the techniques that successfully quantized the other forces might work for gravity too. The authors do not claim this is solved. But they show that teleparallel gravity opens a door that general relativity keeps locked.

The Horndeski Revival

One of the most intriguing sections of the review deals with something called Horndeski gravity. Horndeski theories are the most general class of scalar tensor theories of gravity that produce second order equations of motion. In plain language, they are the most flexible way to modify gravity without breaking it.

For years, Horndeski theories were a playground for cosmologists. They allowed you to add new fields and new interactions while keeping the equations manageable. But in 2017, a paper by physicists at the University of Geneva showed that most Horndeski theories were ruled out by the detection of gravitational waves. The speed of gravitational waves, they argued, had to match the speed of light. Most Horndeski models predicted a difference.

That seemed to kill the idea. But teleparallel gravity offers a resurrection.

The authors describe a "teleparallel analogue of Horndeski gravity which offers the possibility of reviving all of the regular Horndeski contributions" (Bahamonde et al., 2022). Because teleparallel gravity uses torsion instead of curvature, the constraints from gravitational wave observations do not apply in the same way. The old Horndeski theories, thought dead, are suddenly alive again.

This is a big deal. It means that entire classes of modified gravity models that were discarded can be reconsidered. The space of possible theories is larger than physicists thought. And some of those theories might explain dark energy or dark matter.

What the Cosmological Data Actually Shows

The review does not just catalog theories. It also examines how teleparallel gravity performs against real observations. The authors looked at data from the Planck satellite, from supernova surveys, from baryon acoustic oscillations. They wanted to know whether teleparallel models could fit the data as well as general relativity.

The answer is complicated. Some models fit beautifully. Others fail. The authors found that certain teleparallel theories can actually alleviate cosmological tensions, the persistent disagreements between different measurements of the expansion rate of the universe. The Hubble tension, as it is called, is one of the hottest problems in cosmology. The rate at which the universe is expanding today, measured directly, does not match the rate predicted from the early universe. Something is wrong. Teleparallel gravity offers a possible fix.

The authors write that they "examine works in observational and precision cosmology across the plethora of proposal theories. This is done using some of the latest observations and is used to tackle cosmological tensions which may be alleviated in teleparallel cosmology" (Bahamonde et al., 2022).

But they are careful. They do not claim victory. The data is still limited. The models are still being tested. What they show is that teleparallel gravity is not just a theoretical exercise. It makes predictions that can be checked against the real universe.

Machine Learning Meets Gravity

Here is a detail that surprised me. The review includes a section on machine learning. The authors describe how deep learning and Gaussian processes are being used to analyze gravitational theories. This is not common in reviews of gravity. It suggests that the field is changing how it does science.

The authors write that they "introduce a number of recent works in the application of machine learning to gravity, we do this through deep learning and Gaussian processes, together with discussions about other approaches in the literature" (Bahamonde et al., 2022).

What does this mean in practice? Machine learning can help identify which teleparallel models are consistent with data. It can find patterns in complex equations that humans miss. It can accelerate the process of testing theories against observations. The combination of teleparallel gravity and machine learning is still young, but it points toward a future where physicists use AI to explore the landscape of possible universes.

What This Does Not Prove

I need to be honest with you. Teleparallel gravity is not a solved theory. It is not even close.

The review makes clear that many open questions remain. The biggest one is local Lorentz invariance. In general relativity, the laws of physics are the same for all observers, regardless of their orientation or velocity. This is a fundamental principle. Some teleparallel theories violate it. That is a problem. The authors describe how to construct theories that preserve local Lorentz invariance, but not all proposed models succeed.

Another open question is the strong coupling problem. In some teleparallel theories, certain degrees of freedom become strongly coupled at short distances. This can lead to instabilities or inconsistencies. The authors acknowledge this but do not claim to have solved it.

There is also the question of experimental tests. Teleparallel gravity predicts the same results as general relativity for most current experiments. That is good for consistency, but bad for distinguishing them. To tell them apart, you need experiments that probe regimes where the theories diverge. Those experiments are just beginning.

Finally, the quantum gravity problem is not solved. Teleparallel gravity offers a more promising framework, but nobody has constructed a full quantum theory of teleparallel gravity. The authors do not pretend otherwise.

The Quiet Revolution in Physics

Teleparallel gravity is not going to replace general relativity. General relativity works too well. But it might supplement it. It might provide the missing pieces that general relativity cannot supply.

The review by Bahamonde and his colleagues is a signal that the field has matured. It is no longer a fringe idea. It is a serious research program with hundreds of active scientists, dozens of viable models, and a growing body of observational tests. The nearly 500 citations the paper has already received suggest that the community agrees.

What is happening here is a shift in how physicists think about gravity. For a century, we assumed that curvature was the only way to describe it. Teleparallel gravity says no. Torsion works just as well. And torsion might open doors that curvature cannot.

Einstein tried both paths. He chose curvature because it was beautiful and it worked. But he never forgot about torsion. He returned to it multiple times over his career. Maybe he sensed something that the rest of the field took a century to rediscover.

What This Actually Means

• ▸Teleparallel gravity is not a replacement for general relativity but a parallel framework that makes identical predictions for most experiments. The difference matters only at the frontiers: dark energy, dark matter, and quantum gravity.

• ▸The Hubble tension, the mismatch between measurements of the universe's expansion rate, might be resolved by teleparallel models. This is not proven, but the data is promising enough that cosmologists should pay attention.

• ▸Horndeski theories of gravity, which were thought to be ruled out by gravitational wave observations, are viable again in teleparallel form. This revives a large class of models for dark energy and modified gravity.

• ▸Teleparallel gravity can be formulated as a gauge theory, putting it on the same mathematical footing as the other fundamental forces. This makes it a more natural candidate for quantum gravity than general relativity.

• ▸Machine learning is being used to test teleparallel models against cosmological data. This is a new methodological development that could accelerate the search for the correct theory of gravity.