It’s time to abandon dark matter in favor of a new theory of gravity.

The barred spiral galaxy UGC 12158. (Image credit: Wikimedia, CC BY-SA)

This article was originally published on The conversation. (opens in a new tab) The publication contributed the article to’s Expert Voices: Editorials and Perspectives.

Indranil Banik (opens in a new tab)postdoctoral researcher in astrophysics, University of St Andrews

We can model the motions of the planets in the solar system quite accurately using Newton’s laws of physics. But in the early 1970s, scientists noticed that this didn’t work for disk galaxies – the stars at their outer edges, away from the gravitational pull of all the matter at their center – moved much faster than it did. predicted Newton’s theory.

This led physicists to propose that an invisible substance called “dark matter” provided additional gravitational pull, causing the stars to speed up (opens in a new tab) – a theory that has become extremely popular.

However, in a recent review (opens in a new tab) my colleagues and I suggest that observations over a wide range of scales are much better explained in an alternative theory of gravity proposed by Israeli physicist Mordehai Milgrom in 1982 called Milgromian dynamics or Monday (opens in a new tab) — requiring no invisible material.

Related: What is dark matter?

Mond’s main postulate is that when gravity becomes very weak, as it happens at the edge of galaxies, it begins to behave differently from Newtonian physics. In this way it is possible to Explain (opens in a new tab) why the stars, planets and gas on the outskirts of over 150 galaxies are spinning faster than expected based on their visible mass alone. But Mond is not content with Explain such rotation curves, in many cases it predicted their.

philosophers of science argued (opens in a new tab) that this predictive power makes Mond superior to the standard cosmological model, which proposes that there is more dark matter in the universe than visible matter. Indeed, according to this model, galaxies have a very uncertain amount of dark matter that depends on the details of galaxy formation – something we don’t always know. This makes it impossible to predict how fast galaxies should spin. But such predictions are regularly made with Mond, and so far they have been confirmed.

Imagine that we know the distribution of visible mass in a galaxy but do not yet know its rotational speed. In the standard cosmological model, it would only be possible to say with some confidence that the rotation speed will come out between 100km/s and 300km/s at the periphery. Mond makes a more accurate prediction that the rotational speed should be between 180 and 190 km/s.

The galaxy cluster Abell 520, with presumed dark matter highlighted in blue. (Image credit: NASA)

If observations later reveal a rotational speed of 188 km/s, this is consistent with both theories – but clearly Mond is preferred. This is a modern version of Occam’s razor (opens in a new tab) — that the simplest solution is preferable to the most complex ones, in this case that the observations must be explained with as few “free parameters” as possible. Free parameters are constants – certain numbers that we need to put into equations to make them work. But they are not given by the theory itself – there is no reason for them to have any particular value – so we must measure them by observation. An example is the gravitational constant, G, in Newton’s theory of gravity or the amount of dark matter in galaxies in the standard cosmological model.

We introduced a concept known as “theoretical flexibility” to capture the underlying idea of ​​Occam’s razor that a theory with more free parameters is consistent with a wider range of data, making it more complex. In our review, we used this concept to test the Standard and Mond Cosmological Model against various astronomical observations, such as the rotation of galaxies and movements within galaxy clusters.

Each time, we assigned a theoretical flexibility score between -2 and +2. A score of -2 indicates that a model makes a clear and accurate prediction without peeking at the data. Conversely, +2 implies “everything is fine” – theorists could have fitted almost any plausible observational result (because there are so many free parameters). We also assessed how well each model matches the observations, with +2 indicating excellent agreement and -2 reserved for observations that clearly show the theory to be false. We then subtract the theoretical flexibility score from the agreement with observations score, because matching the data well is good – but being able to adjust anything is bad.

A good theory would make clear predictions that would later be confirmed, ideally achieving a combined score of +4 in many different tests (+2 -(-2) = +4). A bad theory would get a score between 0 and -4 (-2 -(+2)= -4). Accurate predictions would fail in this case – they are unlikely to work with the wrong physics.

We found an average score for the Standard Cosmological Model of -0.25 across 32 tests, while Mond averaged +1.69 across 29 tests. The scores of each theory in many different tests are shown in Figures 1 and 2 below for the Standard and Mond Cosmological Model, respectively.

Comparison of the standard cosmological model with observations based on the correspondence of the data with the theory (improvement from bottom to top) and on the flexibility of adjustment (increase from left to right). The hollow circle is not counted in our evaluation, because this data was used to define free parameters. Reproduced from Table 3 of our review. (Image credit: Arxiv)

Similar to Figure 1, but for Mond with hypothetical particles that interact only by gravity called sterile neutrinos. Notice the absence of clear falsifications. Reproduced from Table 4 of our review. (Image credit: Arxiv)

It is immediately apparent that no major issues have been identified for Mond, which is at least plausibly consistent with all data (note that the bottom two rows indicating falsifications are blank in Figure 2).

dark matter problems

One of the most striking failures of the Standard Cosmological Model are the “galaxy bars” – bright rod-shaped regions made up of stars – that spiral galaxies often have in their central regions (see main image). The bars rotate in time. If galaxies were embedded in massive halos of dark matter, their bars would slow down. However, most, if not all, of the galaxy bars observed are fast-moving. This falsifies (opens in a new tab) the standard cosmological model with very high confidence.

Another problem is that the original models (opens in a new tab) who suggested that galaxies had dark matter halos made a big mistake – they assumed that dark matter particles provided gravity to the matter around them, but were unaffected by the gravitational pull of the normal matter. This simplified the calculations, but it does not reflect reality. When taken into account in later simulations (opens in a new tab) it was clear that dark matter halos around galaxies did not reliably explain their properties.

There are many other failures of the standard cosmological model that we investigated in our review, with Mond often able to explain naturally (opens in a new tab) observations. The reason the Standard Cosmological Model is nevertheless so popular could be due to miscalculations or limited knowledge of its failures, some of which have been discovered quite recently. It could also be due to people’s reluctance to modify a theory of gravity that has been so successful in many other areas of physics.

Mond’s huge lead over the standard cosmological model in our study led us to conclude that Mond is strongly favored by the available observations. While we don’t claim Mond is perfect, we still think he gets the big picture right – galaxies are seriously lacking in dark matter.

This article is republished from The conversation (opens in a new tab) under Creative Commons license. Read it original article (opens in a new tab).

Follow all Expert Voices issues and debates — and join the discussion — on Facebook and Twitter. The opinions expressed are those of the author and do not necessarily reflect the views of the publisher.

Sharon D. Cole