A blog about spam by Tommaso Dorigo ( The Spam Of Physicist Mailboxes ) got me thinking about this issue. How can one know if a theory which is less favored or "alternative" to the accepted "standard model(s)" is legitimate science? These points will apply to any area of science, but I know astronomy and astrophysics the best. So, I will use an example from that area of science.
1.) The alternative theory proposes a different underlying mechanism for a set of observations and matches most or all of those observations about as well as the standard models.
In this case the standard models are accepted over the alternatives due to their calculational simplicity, conceptual simplicity, or in the worst case....simply because senior researchers favor them.
A good example of this is in cosmology. The current accepted standard model of cosmology is called . This model depends on dark energy and cold dark matter. We don't understand either of those phenomena very well at all. There is real reason to think we may never be able to detect, let alone experiment with, and learn about the nature of dark matter. This model requires General Relativity + cold collisionless dark matter and dark energy in certain amounts and nothing more.
The alternatives to Lambda CDM require modifications to General Relativity. For example what is known as f(R) gravity. R is the scalar curvature of space time, lets just say it is a very complicated function of position and time. However using it, all the physics of Einstein's general theory of relativity can be written like so.
2.) The proposed alternative embraces the established physics as it proposes alternative new physics.
f(R) gravity says lets try this instead.
Where the function of R can be as simple as a R^2 or very complicated. This model fits the data about as well as Lambda CDM but it is much more complicated. Furthermore Lambda CDM is backed by more senior scientist. This model does not say "Einstein was wrong" it says "let us add to what Einstein did".
3.) The proposed alternative is either conceptually simple yet mathematically complex, OR it is mathematically complex and conceptually simple.
Legitimate theories are either conceptually simple and mathematically complex OR they are mathematically simple yet conceptually complex. The examples of Lambda CDM and F(R) are good examples of this. Lambda CDM is conceptually more complicated than f(R) gravity. The concept of a new type of sub atomic particle with all the quantum field theory it would take for it to be cold, massive, and collisionless is very challenging. However in so far as cosmologist and astronomers care it is mathematically simple. The LCDM universe simply needs a certain amount of dark matter.
f(R) gravity is conceptually simple, "lets modify General Relativity a tiny bit". This leads to all manner of mathematical complications.
The chief sign of a total crackpot theory is it will be both conceptually complicated and mathematically complicated to the point of incoherence.
Such a theory will have a list of "postulates" or basic assumptions half a page or a page long. Such a theory has no underlying principle guiding it. Such a theory will be a collection of equations one which cannot be derived from the other, even by way of mathematical errors. In other words it will not be a mathematically or logically coherent structure.
Sometimes physicist call a theory crack pot or "cranky" because it is simply unpopular or the person proposing it is unsavory for whatever reason. Sometimes they call a theory crackpot because it is counter to the preferences of a senior researcher. In those cases it is simply an ad hominem. (Let me make it clear that did not happen in the comments on Dorigo's blog. He was very patient and respectful of those people as much as can be expected. ) These few points allow you as a consumer of science to judge for yourself.
TLDR: Legitimate alternative theories are known by their conceptual or mathematical simplicity, internal consistency, and ability to match the data at least as well as the generally accepted model. Usually, alternatives are minor extensions to accepted models. Sometimes, they propose a different underlying physics when the accepted model is merely popular speculation.
Illegitimate alternatives are just a mish mash of mathematical incoherence (not just incorrectness) and lack even logical internal consistency. Such models never match the data.... if a model looks bizarre but matches some of the data it is not crackpot.
- Planck confirms WMAP results, and is largely silent on theoretical astrophysics.
- Direct detection of CMB B-Mode Polarization by the BICEP2 team. Yay! Now what does any of that mean?
- The Cyclic Model Vs Inflation, Steinhardt et. al. Vs Hawking et al Data Always Wins.
- Dark Energy Probably Exists, It’s Just Not That Simple
- A most elegant explanation for Dark Matter's unexpected elusiveness