In an interesting comment Daniel Fischer (22.07.2010, 17:00) points out the contribution by Ethan Siegel. In essence Ethan Siegel writes “saying that, ‘since the naive predictions we can make are inadequate, the entire idea of dark matter needs to be thrown out’ is absurd.”
“What is absurd is to presume that a theory that is adequate on one scale will inevitably succeed on another scale. That LCDM is consistent with large scale measurements is no guarantee that it will work on smaller scales. The widespread presumption among cosmologists seems to be that it must surely work out, though this is hardly a scientific attitude.
The failure of dark matter to explain galaxy scale phenomena is not just a failure of detail. A single effective force (modified gravity) suffices to explain (and indeed,
predicted a priori) many aspects of galaxy dynamics. To explain this with dark matter is like asserting that really the solar system operates on an inverse cube law; there just happens to be dark matter arranged just so as to always make it look like an inverse square law.
It is noteworthy that the community seems to stress aparent successes of dark matter but when failures occur these are often put down as occuring in a regime not testable. A good example is the current statement by Daniel Fisher in contrast to the statements found at the beginning of Section 2.4.
We see that Ethan Siegel’s argument is wrong logics. The dark matter hypothesis was introduced to specifically solve the “small-scale” problem of flat rotation curves of disk galaxies. To later argue that the dark matter hypothesis cannot be tested on these scales is throwing out logics and the very basics of how scientific inference works. Actually, in our research paper the problems found are not only on small scales but extend to scales of a million pc.
A Gedanken experiment helps to clarify the problem: Take one observed normal disk galaxy which you know to be in equilibrium (i.e. is not disturbed by another galaxy). Call this galaxy 1. It has a flat rotation curve. Put in the correct distribution of dark matter to explain the rotation curve. Now observe another disk galaxy which you know to be in equilibrium (galaxy 2). Measure the distribution of normal matter. Ask the astronomer to predict the shape of the rotation curve of galaxy 2 based on the observation of galaxy 1. Then measure the rotation curve of galaxy 2. The predicted and observed rotation curve will almost certainly be different.
Now, repeat the exercise in MOND (no dark matter, but a gravitational theory modified according to Milgrom). The astronomer will be able to exactly predict the rotation curve of galaxy 2 based purely on the observed distribution of normal matter in galaxy 2 and one universally valid number (a universal acceleration scale, a_0) obtained from the fit to galaxy 1. In fact, the astronomer can predict the shapes of all rotation curves using this one number a_0.
Therefore, Milgrom’s theory is far more superior in describing galactic dynamics than the dark matter plus Newtonian hypothesis. What is more, the dark-matter hypothesis does not allow us to understand rotation curves. And this does not depend on unknown small-scale effects, as we are talking about scales of 10,000 to 50,000 pc.
Some explicit references:
In McGaugh & de Blok (1998a) the authors state: “Interpreting the data in terms of dark matter leads to troublesome fine-tuning problems. Different observations require contradictory amounts of dark matter. Structure formation theories are as yet far from able to explain the observations.”
The companion paper, McGaugh & de Blok (1998b) finds: “One hypothesis, Modified Newtonian Dynamics (MOND), is consistent with the data. Indeed, it accurately predicts the observed behavior. We find no evidence on any scale that clearly contradicts MOND and much that supports it.“
Within the dark-matter theory galaxies would be embedded in dark-matter halos with well-known properties. That they do not match observed roation curves is documented convincingly by Kuzio de Naray et al. (2009): “The shape of the modeled NFW rotation curves does not reproduce the data”.
The observed similarity of galaxies being in contradiction to the expected large variation of galaxies if dark-matter theory were correct is documented by Disney et al. (2008).
In fact, galaxies tell us that the dark matter (if it were to exist) arranges itself according to the distribution of the normal matter, although the normal matter makes only a small fraction of the mass of a galaxy. To explain this insurmountable fine-tuning problem one would need to invoke a dark force coupling dark matter to baryons in a hitherto not understood way (Kroupa et al. 2010), or simply accept that galaxies are made up only of normal matter without dark matter and that gravity is non-Newtonian.
A lesson from history:
In this vain we might remember history: Once upon a time, not too long ago in fact, very clever people knew that heavenly bodies were either at rest or moving about other bodies on perfect circles (the planets and Sun moving around the Earth). So important and widely accepted was this idea that when the observations (of planetary motions) failed to fit the predicted motions, a circle on a circle was introduced to describe the motion of the planets. And if this was still not good enough, then yet another circular motion about a centre which moved on a circle about a centre which moved on a large circle about the Earth was added. But, in this theory of epicyclic motion an astronomer was not able to predict the motions of a new planet.
And here too, one could have argued that “just because there are minor deviations from the calculated motion of the planet and just because a Kepler and a Copernicus thought that the actual motion was on an ellipse about the Sun (!) it is surely absurd to throw out the whole large and divine picture of perfect motions of heavenly bodies, which was so successful overall.”
Note also that two major mental changes had to be accepted: The centre of motion is not the Earth but the Sun and the motions are elliptical.
In our case (Kroupa et al. 2010) two mental changes are also required: the satellite galaxies of the Milky Way are not dark-matter sub-structures but tidal-dwarf galaxies, and dynamics is non-Newtonian (e.g. Milgrom’s MOND or Moffat’s MOG).
And, although the Kepler laws and Copernican idea proved endlessly more successful in describing planetary motions, they were quite useless in describing the dynamics of star clusters because the deeper (Newtonian) gravitational theory had not been discovered yet. But it was absolutely evident that the epicyclic ansatz was out.
Today we would be saying that modified gravity is endlessly more successful in describing rotation curves of galaxies and galactic dynamics in general, and that cold or warm dark matter is out. But we do not yet have the full underlying gravitational theory which is likely to unite matter and space time as a single entity.
The concordance cosmological model arose by adding dark matter to the Theory of General Relativity, then adding inflation and then dark energy to get a model which has not been able to account correctly for the way galaxies work nor how they evolve and arrange themselves in space time. Furthermore, an extra dark force is required as a further addition.
by Anton Ippendorf, Pavel Kroupa and Marcel Pawlowski (28.07.2010): “Is it absurd to throw out the idea of Cold or Warm Dark Matter?” in “The Dark Matter Crisis – the rise and fall of a cosmological hypothesis” on SciLogs. See the overview of topics in The Dark Matter Crisis.