If you can't explain it simply, you don't understand it well enough. ~ Albert Einstein
"Using one unsolved mystery to solve another makes some people sceptical," says Marcus Chown in his article on the latest weird theory involving dark matter. To which I can only add, "Amen, brother." It seems that dark matter caused the re-ionization of matter when the universe was about a billion years old. Well, at least that's the what the gang over at Fermilab is postulating these days.
But that's not all. Somehow while dark matter was busily annihilating itself to ionize hydrogen, it was also responsible for the growth of the supermassive black holes at the centers of galaxies, in a process called, so help me, "dark gulping."
Oh lawsey, lawsey, lawsey, when will it stop?
It's not that I have anything against dark matter or its evil twin, dark energy. They've never done anything to me. No, my problem is with scientists who, in their rush to publish something, anything at all, will invoke those undetected and unmeasured phantoms at the drop of a tachyon. The dark things are the deus ex machina of modern physics.
I've been trying to figure out why there are so many half-baked, ill-formed, and generally unsatisfactory theories coming out of physics and cosmology these days. Certainly string theory can take its share of the blame. I've written a few thousand words on the subject (just search the blog on "string theory" and see), but string theory isn't to blame for dark matter/energy.
The need to "publish or perish" is, of course, always an issue. Scientists trying to get grants or tenured positions all always under pressure to publish theories as soon as possible, often without the requisite due diligence. And then there's just the issue of keeping your organization in the scientific limelight. Certainly, the advent of the Large Hadron Collider has had the folks at Fermilab trying to show how relevant they still are. But even those pressures can't explain the weird sorts of theories being published almost daily.
I've got a theory about theories.
First, there's the problem of data. It used to be that data about the workings of the universe at the scales of the very large or the very small were difficult, if not impossible, to come by. Nowadays, though, there is a veritable flood of information from satellites and sophisticated ground-based telescopes. Even without the LHC, there is a mountain of data created by existing particle colliders, and, thanks to our ability to actually "see" things happening at quantum levels, there is data on the behavior of atomic particles.
The trouble is that there is so much data, no one can stop long enough to analyize it. Consider that recently, planets were discovered to have been imaged by the Hubble Space Telescope. This is great news, except that they were imaged years ago, yet astronomers are just now reviewing those images closely enough to tell what they show. Hubble has taken such a huge volume of measurements that no one can get their arms around it. Add Spitzer, Chandra, and a host of other devices, and it becomes obvious that we simply have more data than we can analyze in a human lifetime.
Now, in itself, such a huge volume of information is not a bad thing. The problem comes when scientists cannot impose enough discipline on themselves to actually sift through the mountain of data with some degree of focus. What we tend to see is a researcher taking a single observation and making a theoretical mountain out of an single-data-point molehill.
The other cause of the increase in cockamamy theories is the ease of constructing computer models. Now, I'm a networking specialist by profession. I understand a lot of in's and out's of programming. Over the years, I've learned that the results that a computer program produces are completely dependent on the conditions set in the program.
Now consider that no one has ever actually found any dark matter. There is some putative evidence of gravitational effects that would be explained by some unknown source, but we have no idea what it is and what its properites are. If one is to construct a program to model the interaction between dark matter and black holes, one is going to make a lot of assumptions. Just because that model turns out something called "dark gulping" (oh lawsey, lawsey, lawsey), that doesn't mean the universe engages in such asto-gastronomical adventures.
In other words, as one scientist once put it on some show or another, if you want to prove that pink elephants exist, you can create a computer model that will generate them given certain starting conditions.
This is not to say that computer models are all phony or useless. On the contrary, a model based on solid initial conditions can be very useful. But, those initial conditions must be well understood, and the laws of interaction that work on those conditions also need to be well understood.
Consider weather prediction. Around here, hurricane season makes us all weather researchers. Many of us will monitor predicted hurricane paths with a passion borne of a cross between pure fear and morbid curiosity (if it looks like the hurricane is going somewhere else). One weather site used to publish a map showing the paths predicted by all the current computer weather models. I presume they do this to provide a little comic relief to lessen the tension, because some of the models will show a hurricane making it to Minnesota as a category 2 storm.
You might say I'm being unfair because the weather is a complex business, with many factors to consider, the interplay of which is still not fully understood. My answer is that you're right. The weather here on Earth, which we've been studying for centruies like our lives depend on it (because it does) is so difficult to model because we don't understand all the physics of what's going on in the atomosphere.
So how can we be confident that a model involving an unknown substance (dark matter) with unknown properties interacting with a known object (a black hole) with poorly understood properties is actually producing a meaningful result? We can't be.
Computer models can be useful (even the weather ones) as a guide, but we should be aware that they could also be leading us to Minnesota.
What physicists need to do is start focusing their efforts more on collating some of those mountains of data and interpreting what they tell us directly, rather than cherry picking an interesting observation here or there and creating a model that mimics the behavior based on initial conditions that may or may not have anything to do physical reality.
The various forms of science used to be called "disciplines." What's needed is to restore some discipline to the sciences.