marcus said:
well I've kept gnawing at it
read some more (mostly in the long paper) last night
willeke, would you to try to paraphrase or explain a bit of this?
say some of the obvious things about the chief rival of DM?
not-completely-lousy new year everybody
(if i could understand Teves better and see some more development in Dynamulation gravity, it would brighten the gloom some)
The very basics:
1. MOND and all theories spawned from it, say that the odd galactic dynamics and other phenomena attributed to dark matter can be more parsimoniously explained by tweaking the law of gravity so that in the spherically symmetric, weak field limit, it acts as a 1/r force instead of a 1/r^2 force, once the field reaches a point a sub zero at which the accelleration due to gravity is 1.2*10^-8 cm/s^2.
Comparisons of this theory to the data show that it explains the dynamics of galactic sized systems very well (the predictions are almost always well within the error bars of the data) without assuming any dark matter. In particular, it makes predictions about galaxies such as low surface brightness galaxies, whose existence was not known. It has made bona fide predictions that have born out, rather than post-dictions. It does this with basically one parametere (a sub zero) as the other potential variable in the theory (the mass to light ratio) is nearly 1 for almost all galactic objects.
Even if MOND is dead wrong and a space probe bonks its sensors on a big chunk of dark matter tomorrow and definitely proves that the stuff is completely made of WIMPZILLAs or whatever, MOND still creates a puzzle for DM theorists. DM theory doesn't say much about how dark matter must be distributed. It looks at a system looks at what Newtonian theory would predict and explains where the DM has to be for the Newtonian theory to match the result using about three parameters for any given galaxy. This calculation by the way has been show to requires a distribution that is impossible to fit with baryonic matter, i.e. matter made up primarily of protons and neutrons like the matter we know and love. But, MONDs successes show that even if DM is right that DM must have a very specific distribution in relation to the luminous matter in a galaxy. This is particularly problematic in ellipitcal galaxies which are formed from the merger of spiral galaxies because ellipitical galaxies do not show dark matter expectations consistent with a merger of two spiral galaxy DM halos, but do behave consistent with MOND. Galaxy formation theories do a poor job of explaining how such a uniform distribution of DM relative to luminous matter could arise.
MOND has failed primarily in three areas.
2. One, it has not predicted any dark matter in galactic clusters (which are within the area where MOND effects are not predicted in the basic theory). But, experimental data seem to show a 50%-80% DM content in those systems. This could be due to massive neutrinos (at about 2eV each) or it could be due to differences between basic MOND theory and advanced MOND theory. Still, considering that many kinds of galaxies are off by factors of 100+ and that MOND does not create a mass deficit and would not need non-bayronic matter to make up the deficit in galactic clusters, this isn't a great failing.
3. MOND has never made any specific prediction about lensing. It hasn't had a relativistic formulation and there has been doubt about whether it was even theoretically possible to great a formulation of MOND that was both well behaved and consistent with GR in non-MOND settings. MOND has not made any strong predictions cosmologically for the same reason The lack of a GR generalization has been cited as a major reason for rejecting MOND over DM.
I made a first shot at paraphrasing the technical details, and found my paraphrase even worse than the original (which is damn ugly in its notation and less than elegant in its exposition). So, let's start with the bottom line conclusions:
1. TeVes is designed so that it reduces, mathematically, to General Relativity in situations where gravitational fields are more than negligable. Only at weak gravitational fields do its unique characteristics come out. As a result, on a solar system scale (where graviational fields are sufficiently strong) and in essentially all the other places we have tested GR, TeVeS makes the same predictions.
2. TeVeS in the weak field is basically a Lagrangian formulation of MOND. The two are identical for point sources. The Lagrangian vector field skews naiive MOND predictions for mass distributions that are not spherically symmetric. The Lagrangian also tweaks the way the MOND acts when, e.g., two fields, one in the MOND limit and one in the ordinary limit are superimposed upon each other. The differences from naiive MOND are subtle (e.g. 10-15% adjustments in the MOND effect in some the areas of spiral galaxies), but have the effect of allowing center of gravity calculations, conservation of energy, etc.
3. TeVeS tweaks GR to a great extent through tensor big U. Because there is a tensor formulation, the theory does not have a preferred coordinate system.
4. TeVeS achieves a major result no prior theory of relativistic MOND did, its impact on light (twice the deflection created by gravitational effects acting on matter) is the same as GR. In other words, if the mass dynamics of a galaxy or other system are accurately described by both MOND in general and by DM as formulated in a particular system, then the lensing predicted by both theories will be the same.
5. TeVeS, like all prior formulations of MOND, uses an arbitrary function to transition between the GR/Newtonian regime and the MOND regime. In TeVeS this function is called "F" and it is a horrible ugly beast that Bekenstein himself describes as a toy model. It fits the data, but is really just a placeholder.
6. TeVeS is ugly in part because it is formulated not as a from scratch theory, but as a modification of GR. GR's actions can be basically broken into geometric and non-geometric components. TeVeS uses a four way split, modifying the two GR components modestly, and then adding two new components described as the vector and scalar field components. Heuristically one can think of the tensor part (the first two of the four components of the action) basically being old fashioned GR tweaks a little by the U tensor, the vector part connected with the adjustment in the Lagrangian necessary to make MOND well behaved, and the scalar part necessary to reproduce naiive MOND.