Black holes do not exist

What is a black hole?

To anchor the argument, let us first pin down what a black hole is.

A black hole is an astronomical object that collapsed under its own gravity. It is often identified with a region of space in which the gravitational field is so strong that even light cannot escape it. The boundary of this region is called an “event horizon,” and inside this region resides a “singularity” – an entity of no volume that contains all the accreted mass. Everything that crosses the event horizon must end up at the singularity.

Black holes, their horizons, and singularities are usually modelled using Schwarzschild or Kerr metrics.

What mistakes have been made?

The concept of a black hole developed from the analysis of the Schwarzschild metric. The Schwarzschild metric is the simplest, commonly used model of a black hole. It is thought to describe non-rotating, non-accreting black holes adequately. Most importantly, the Schwarzschild metric is a stationary solution to the Einstein’s equations in a vacuum (the Kerr metric too; a stationary solution means that the metric and the distribution of matter are constant over time). Therefore, using it to describe massive compact bodies carries an assumption that these bodies are in some sort of equilibrium. This assumption is incorrect.

Let us consider the following analogy. Imagine a theory predicting that a physical system is described by a dynamic equation $a_n=0.5\times a_{n-1}$. If we assume the system is stationary, then $a_n=a_{n-1}$, and the only solution is $a_n=0$ for all $n$. However, this may turn out to be a non-physical solution – one that never describes an actual physical system. Such a system may be described only by a non-stationary solution $a_n=k\times 2^{-n}$ (for some $k$ that depends on the initial conditions).

Analogously, because Einstein’s equations are dynamical, assuming stationarity can yield a non-physical solution. Numerous paradoxes suggest that black holes indeed may not exist and that the Schwarzschild (and Kerr) metric, as a vacuum solution, although mathematically correct, does not correspond to physical reality in the region of space demarcated by the supposed event horizon.

Physicists have been aware that both stationary and non-stationary solutions exist. While the stationary solutions are described by the Schwarzschild and Kerr metrics, a non-stationary solution refers to a body in a state of permanent and asymptotic collapse, whose never-attainable limit is the stationary solution (in a similar way as $a_n=k\times 2^{-n}$ keeps approaching $a_n=0$ in the limit; see Oppenheimer and Snyder, 1939, for an example of a non-stationary solution).

Nevertheless, physicists have been biased toward stationary solutions as the preferred description of reality. The source of this bias seems to come from another assumption: that time must continue. A pure non-stationary solution has a “fatal” feature that the proper time of a co-collapsing observer is finite. However, if time must continue, such solutions cannot show the whole picture. Therefore, contemporary models of gravitational collapse stitch together the two types of solutions in an attempt to explain what happens after the proper time allotted to collapsing matter elapses.

The assumption that time must continue is not based on any physical principle. Eliminating it helps create a much clearer and more coherent picture of the celestial bodies we observe.

What exists instead?

If black holes do not exist, what do we observe as those compact invisible objects whose mass exceeds the maximum for a neutron star? To answer this question, let us imagine a neutron star (a pulsar) slowly accreting matter. As the mass of the star increases, the time on its surface slows down relative to the clock of a distant observer. That is, if an observer on the surface of the pulsar perceives the period to be 10 milliseconds and that value does not change as mass is added, a distant observer will see the period gradually increase, and the pulsar’s angular velocity will asymptotically approach zero but never reach it.

This is a physical effect, not an “illusion.” To see that, an observer in a rocket hovering over the neutron star can drop a rope ladder towards its surface. Then, he can quickly descend, spend some time at the star’s surface, and climb back up. A comparison between the readings on his watch and the clock left on the rocket reveals that time flows differently at the two locations.

The same applies to larger bodies. A supermassive object at the center of a galaxy does not have an event horizon. Nor does a singularity exist there. Instead, its matter is in a state of never-ending, ever-slowing collapse. The closer it gets to forming an event horizon, the more slowly this limit is being approached.

An event horizon and a singularity are defining features of a black hole. Because objects in the state of permanent collapse are not black holes, a new name is needed. For the purposes of this article, I will refer to them as brown holes. Holes because it is still very hard to retrieve something dropped into them, and brown because they can emit radiation, albeit extremely redshifted.

What does an observer feel as he stands on the surface of the above-mentioned accreting pulsar or flies his rocket into the brown hole at Sagittarius A*? In principle, he can calculate when the final pulse of the pulsar is supposed to happen, or more generally, when he is supposed to meet the event horizon. This should happen in a finite proper time, which I will call the remaining time from now on. So, let the observer pull out his iPhone and set the timer to his remaining time.

The observer waits for the countdown. He sees things moving with great speed, and feels the full forces of nature to the very last second. And then, when the timer is supposed to ring, time itself stops. There is no “after.” All physical processes cease, including the observer’s ability to perceive the flow of time. You can imagine that this is a movie whose final frame is still on the screen, but the motions initiated in previous frames will never continue because the movie is over.

Perhaps it is hard for some people to imagine that time will not continue forever. However, recall that modern theories hold that time began with the Big Bang. And there is little difference between the two: if time could start without a “before,” why can’t it end without an “after?” Or perhaps what is hard to swallow is that different observers may have different remaining times, especially if some have finite while others have infinite ones. But this is an obvious and natural consequence of a curved spacetime with time flowing differently for different observers. On the contrary, I argue that it should be even harder to imagine and accept the consequences of insisting that time must continue.

Black holes create disjoint regions of spacetime

Consider a Schwarzschild black hole created by symmetrically infalling matter and radiation. Let us examine the structure of a spacetime manifold in which such a black hole forms and exists for some time. We are going to examine the integrity of the manifold by connecting various points on it (events) using curves of finite length, i.e., curves combining timelike, spacelike, and lightlike intervals.

Let us place an observer at some distance from the star before it starts to collapse: this observer is present at event A, the first point on the spacetime manifold we note. In principle, the observer can fly his rocket into the collapsing star and observe its evolution from within. At a specific time and place – event B – the observer has enough information about the star to know that he, along with the surrounding matter, will cross the event horizon (within finite proper time) at event C.

The newly created event horizon is (supposedly) a three-dimensional submanifold (or a hypersurface in the language of Penrose, 2002) of the four-dimensional spacetime manifold. Thus, from event C, we should be able to draw a (mostly lightlike) curve to event D, which is again located at the event horizon but in the region of spacetime where the star has already collapsed, accreting most of the incoming matter and radiation, and the horizon has stabilized.

As the final step, we can draw a curve of positive proper distance from event D to some event E located far outside the fully-formed black hole. Thus, events A and E can be connected by a combination of curves with a finite proper distance or time, at least as long as some of these curves touch the event horizon.

The second observer starts from the same event A, but instead of going to B, goes to the expected location of event E and waits for it to happen. However, no matter how long he waits, E will never happen. The second observer can never fulfill the conditions placed on event E as we defined it above: he cannot be connected by a curve of positive proper distance with D, or any point on the event horizon, as the formation of a black hole is forever in his future (he can be connected with the event horizon only by a curve of positive proper time).

Where and when is E if it cannot be reached from A, even though both events are outside the event horizon? Or more generally, where (or what) is the neighborhood of a black hole, and how does it fit with the larger picture of the entire spacetime manifold? Neither matter nor energy, nor even information, can travel from A to E. It seems that the neighborhood of a black hole constitutes a reality disconnected from our own, one that the black hole creates or joins upon its formation (see Figure 1).

Alternatively, black holes do not exist, and the spacetime is not partitioned into semi-disconnected pieces. Brown holes solve this and many other paradoxes.

Figure 1. Spacetime diagram for a simplified version of the situation described in the text. A black hole is supposed to form from a shell of incoming particles. The path of a particle from this shell is denoted by S. The radial coordinate R, in typical fashion, is asymptotically identical with the proper distance measured by a distant stationary observer and is marked on the horizontal axis. R=0 is at the center of the shell. R=r_s is the coordinate at which the event horizon is supposed to stabilize (i.e., the Schwarzschild radius). On the vertical axis, temporal coordinate T, identical with the proper time of a distant observer, is projected hyperbolically. S crosses R=r_s in T=∞, at C. The coordinate time above the T=∞ reflects the proper time as measured by any clock capable of measuring the time after the stabilization of the event horizon. The first observer starts at event A and moves towards B. There, he waits for a finite amount of time for the shell to catch up and for the event horizon to form at C. Next, we draw a lightlike curve CD on the event horizon and connect D and E with a curve of a finite proper length. If the second observer starting at A wants to get to E without touching the event horizon, he finds this task impossible, because the curve between A and the question mark requires infinite proper time. In fact, events A and E are on separate patches of the spacetime manifold, which are not connected outside the black hole. Therefore, curve AE does not exist. Note that the description in the text does not refer to any coordinate system, so this conclusion is not an artifact of improperly chosen coordinates.

Larger context

The first well-known model of gravitational collapse was proposed by Oppenheimer and Snyder (1939). This model is compatible with the concepts described herein and can be treated as a model of a brown hole. The authors do not analyze what happens after the remaining time runs out, except to note that once it does, the observer loses the ability to send signals from the star. Although this suggests the “time must continue” way of thinking, it has no bearing on their model. The event horizon and the singularity do not form.

In subsequent models, the assumption that time must continue is clearly visible. In models like that of Penrose (1965), the observer crosses the horizon and falls into a singularity. Eddington–Finkelstein coordinates are used because the coordinates of a distant observer cannot meaningfully describe the spacetime manifold containing a black hole.

Noteworthily, Penrose (2002) writes: “It would be unreasonable to suppose that the observer’s experience could simply cease after some finite time, without his encountering some form of violent agency.” This is an explicit form of the “time must continue” assumption. By making this assumption, Penrose eliminates the possibility of describing spacetime using global inertial coordinates. Coordinate systems such as Eddington–Finkelstein or Kruskal–Szekeres coordinates are not needed to describe brown holes – the coordinates of a distant observer are sufficient. The possibility arbitrarily judged as “unreasonable” is in fact conceptually simpler (fewer unintuitive concepts), clearer and more coherent (no paradoxes), and mathematically more beautiful (no singularities or complicated coordinate systems).

As wrong assumptions had led the physicists to faulty models, several problematic ideas arose. The following problems can be easily solved by abandoning black holes in favor of brown holes: 1) bursts of Hawking radiation, 2) violation of the baryon number, 3) white holes and worm holes, 4) loss of information, 5) breakdown of physical laws at the singularity, etc. On the other hand, abandoning black holes will render much of the theory developed over the past decades vacuous. The amount of wasted effort is heartbreaking.

What about the presence of supposedly mounting observational evidence for black holes? Guidry (2019) notes that such evidence tends to fall into three categories: 1) binary stars in which an unseen companion is too massive to be a neutron star, 2) movement of stars indicating the presence of supermassive compact objects in centers of galaxies, and 3) mergers of massive compact objects detected by gravitational wave detectors. None of those observations indicates the presence of an event horizon. Instead, they only indicate the presence of compact objects with strong gravitational fields. They are thus compatible with the concept of a brown hole.

Recently, the idea that black holes might not exist and should be treated as if they were forever approaching the yet-to-be-formed event horizon in an ever-slowing collapse was raised by Janzen (2025). However, he is agnostic about which of the two solutions – stationary or non-stationary – better describes the physical reality. Instead, he focuses on the non-stationary solution as more empirically relevant. In contrast, this article goes a step further by taking a clear stance on the physical reality (hence the title) and describing what the infalling observer would experience.

Two horizons

In an actual brown hole, the remaining time for different observers can be different. Oppenheimer and Snyder (1939) suggest that an observer closer to the center of a brown hole has a shorter remaining time than the observer further away. For simplicity, let us assume that the brown hole in question has been collapsing long enough that the remaining time for 90% of its constituent matter is shorter than one second, and the remaining time for 99% of its matter is shorter than five seconds.

Light source X is located at a proper distance larger than five light-seconds from the “surface” of the brown hole. The light from X can never reach the particles of the collapsing star (except perhaps the 1% of the stragglers accreted later). The area from which light beams can reach the brown hole and the area from which light beams cannot reach the brown hole can be separated by a (perhaps somewhat fuzzy) boundary, which we can call the external horizon.

Light source Y is located near the center of the brown hole (for convenience, let us assume the brown hole is locally transparent). If within one light-second of Y the metric implies remaining time shorter than one second (and decreasing), then light emitted from Y cannot escape the brown hole. In principle, we should be able to find a region from which light cannot escape the brown hole and call its boundary the internal horizon.

This raises some interesting questions:

1. Does the internal horizon always reside inside the external horizon? Or alternatively: can the two horizons cross?
2. Do the horizons asymptotically approach the place where the event horizon is supposed to form at T=∞?

These, together with many other questions about brown holes, were neglected for too long and wait to finally be answered.

References

Guidry, M. (2019). Modern general relativity: black holes, gravitational waves, and cosmology. Cambridge University Press.

Jenzen, D. (2025). What if black holes never actually form? CosmiCave. https://cosmicave.org/2025/11/06/what-if-black-holes-never-actually-form/, date accessed: 2025-12-27.

Oppenheimer, J. R., & Snyder, H. (1939). On continued gravitational contraction. Physical Review, 56(5), 455-459.

Penrose, R. (1965). Gravitational collapse and space-time singularities. Physical Review Letters, 14(3), 57-59.

Penrose, R. (2002). “Golden Oldie”: Gravitational collapse: The role of general relativity. General Relativity and Gravitation, 34(7), 1141-1165.

How to cite

To cite this article in print, you can use the following template:

To cite this article on a website, please link it using the name of the article, for example:

Contact us

You can contact us by mail: