How good is your theory?

Ethan Siegel in How Good is Your Theory? Open Thread I : Starts With A Bang : Starts With A Bang, describes these levels of support, which I have numbered:

  • 3: Scientific Law. Very well-supported theories that have successfully passed numerous tests.
  • 2: Validated. Well-supported but with some problems, like unconfirmed or untested parts.
  • 1: Speculative. May or may not be testable. Often gets a lot of publicity, however.
  • 0: Ruled Out. The opposite of the first category.

Ethan Siegel’s spectrum is, I think, a good way of comparing the level of support for various theories, and it is a good antidote for “scientists always change their theories!” arguments. Many theories come and go, but they are usually on the speculative side, around 1. Those that get promoted to 2 or 3 usually last. Some of them may get their domains of validity restricted, as with Newtonian mechanics after the rise of relativity and quantum mechanics, but they still remain very successful theories within those domains.


I’ll now consider some speculative hypotheses that are now very well-supported.

Continental drift. Alfred Wegener proposed it in 1912, but it languished a long time between 0 and 1 until the 1950’s and 1960’s. In those decades, it moved up to 3, and it has stayed there ever since. When I was old enough to understand the issue, it had already gotten there. I remember Natural History magazine having an article about it that had a cartoon of an anthropomorphized North America walking and carrying a suitcase.

The late-Pleistocene giant floods, like the Missoula Floods, were glacial-dam-break floods for continental glaciers in the Columbia River Basin and the Altai Mountains and some other places. They were essentially large versions of similar floods that nowadays happen in glaciated mountainous areas. J Harlen Bretz proposed them for the Columbia River Basin in 1923, and it stayed between 0 and 1 until the 1970’s, when it got elevated to 3. In 1980, geologist R.B. Wiatt proposed that there were at least 40 of them, something which indicates the hypothesis’s high status by then.

Impact cratering was more-at-less at 1 until the 1960’s, when the discovery of minerals with “shock metamorphism” firmly established impacts, bumping the hypothesis into 3. Some impact-related hypotheses still remain at around 1 or 2, however. Did the K-T impact (or impacts) cause the K-T mass extinctions? Were other mass extinctions caused by big impacts? Did an impact start the Younger Dryas cold snap and mass extinctions at the end of the last Ice Age? Was the impact that created Burckle Crater in the Indian Ocean remembered in numerous flood legends? The K-T one I’d rate as around 2; the others around 1.

Quarks. In 1964, Murray Gell-Mann, George Zweig, and Yuval Ne’eman proposed that hadrons are composed of them. From then to the 1970’s and 1980’s, quarks moved from 1 to 3, with gluons and Quantum Chromodynamics following along.

The particle-physics Standard Model in general has been pretty close to 3 since the 1980’s, and if the LHC Higgs detection holds up, that will move it even closer.

Endosymbiosis of eukaryotic organelles. It was first proposed in 1905 by Konstantin Mereschkowski, but it languished at close to 0 until Lynn Margulis revived it in 1967. Here is what she proposed it for:

  • Mitochondria and chloroplasts: endosymbiosis gradually went from 1 to 3 over the 1970’s and 1980’s, and it’s still at 3.
  • Eukaryote flagella and cilia: while the origin of the eukaryotic flagellum is still obscure, endosymbiosis is gradually drifting from 1 to 0.

I’ll now consider some theories that seem rather speculative to me but that seem to have gotten a lot of acceptance in recent decades.

Very early evolution of life: the RNA world. It states that there were some early organisms which used RNA for both information storage and enzyme action. These organisms then devised proteins and DNA, leading up to the most recent common ancestor of all known present-day cellular organisms.

I’d rate it at least a 1, and over the last few decades, a lot of discussions of it I’ve seen imply that it’s nearly 2 or even a little past 2. I don’t know how much that can be justified, however. Ribozymes (enzymes made of RNA) suggest that a RNA world is feasible, but that’s not really evidence for a RNA world. I think that vestigial RNA would qualify, however, and there is lots of RNA that’s reasonably interpreted as vestigial.

The main criticism I’ve seen of the RNA-world hypothesis is the difficulty of a prebiotic origin of RNA. But there’s a solution: RNA was preceded by something with a similar base-pairing mechanism, but that could more easily form prebiotically. RNA would later take over from it. There are several possibilities for that hypothesis, and I’d rate it as 1.

Mars’s former ocean. The Mars Ocean Hypothesis. The Vastitas Borealis is a large northern lowland region that has features around it that strongly suggest that some liquid had poured into it, likely water. It’s still at 1 and maybe a bit toward 2.

The giant-impact origin of the Moon. The Giant impact hypothesis or the “Big Whack”. It was proposed in 1946 by Reginald Aldworth Daly, but it was ignored until it was revived in 1984 by William K. Hartmann and Donald R. Davis. This rather Velikovskian hypothesis is the most successful hypothesis so far for the origin of the Moon. A Mars-sized object collided with the early Earth, producing splatters that went into orbit and condensed into the Moon. It is a plausible consequence of how the inner planets had formed, and it explains why the Moon is (1) mostly rocky and (2) depleted in volatiles as if it had been baked. I’d rate it at least a 1 and possibly somewhat close to 2.

Inflation: early-Universe exponential expansion. Proposed by Alan Guth in 1980, it states that the Universe had an early phase of exponential expansion. It successfully accounts for primordial fluctuations; these were quantum fluctuations that got frozen in place when they got stretched beyond the horizon size. It also accounts for the great size of our Universe; it needs only about 60 e-foldings to produce that size. It’s much like the steady-state cosmology, though that is now at 0. There are still problems with it, like the nature of the inflation-making particle or “inflaton” (no i), and how the inflation phase ended.

I’d say that it’s a little less than 2; Ethan Siegel rates it as 1.7.

Supersymmetry. It states that every elementary particle has a relative of it with the same interactions but with a spin differing by 1/2. It successfully eliminates certain awkward infinities, and the Minimal Supersymmetric Standard Model has the most success at achieving Grand Unified Theory gauge-coupling unification. According to GUT’s, the QCD and electroweak interactions should be parts of the same interaction, meaning that they ought to have the same “charge” value, and the MSSM is best at achieving that at GUT energy scales.

The main downside is that no particle is known which has another known particle as its supersymmetry partner. In fact, if the LHC detection of the Standard-Model Higgs particle holds up, it may be hard to observe supersymmetry partners of known particles.

So I’d rate supersymmetry halfway between 1 and 2. However, various other people prefer to rate it as close to 1, from its lack of experimental support.

Grand Unified Theories. Somewhat greater than 1.

String theory. An attempt at a Theory of Everything. It includes quantum gravity and GUT’s, and one can get much of the Standard Model out of it, but it does not do so anywhere close to uniquely. I’d put it at 1, maybe a tiny bit more.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: