Thursday, March 12, 2020

Scientific models: adjustment or validation?

Leonard Nimoy
as Mr. Spock
One of the ways in which science advances is by building models, which are often made up of more or less complex sets of mathematical equations, and trying to verify whether or not these models adapt to the functioning of the real world, as described by our senses and our instruments.
When building and using a model we must consider two distinct phases:
  • Model adjustment: it consists of assigning values ​​to the parameters of the model to ensure that it fits the data we already have about the real world. A model not adjusted to such prior knowledge would be totally useless.
  • Model validation: it consists of using the model to make surprising predictions that nobody could have foreseen without the help of the model. If these predictions are confirmed, they become surprising accurate predictions, validating the model. However, the validation is never final, for a new surprising inaccurate prediction could invalidate it in the future.
Let's look at a few examples:

  • Newton's theory of universal gravitation was adjusted to explain Kepler's three experimental laws, which had been known since the early 17th century. The value of G, for example, was adjusted so that these laws were correctly predicted by the model. But Newton's theory was not considered valid until it made several surprising accurate predictions, the most spectacular of which was the prediction of the existence of the planet Neptune by Adams and Le Verrier. The bad news: shortly after its great success, came its great failure: the surprising inaccurate prediction made by Le Verrier about the existence of an unknown planet (Vulcan) between Mercury and the sun, which (we know today) only exists in the Star Trek series, as Mr. Spock claims to have been born in that nonexistent planet. This prediction made it necessary to replace Newton's theory by a more approximate model.
  • Einstein's theory of general relativity was adjusted to explain everything that Newton's theory had predicted correctly, and to correct what it had predicted incorrectly. It was considered validated when it made two surprising accurate predictions: the deviation of light when passing near a large mass (such as the sun), confirmed by Eddington in 1919, and the gravitational redshift of sunlight, checked in 1959. There have been others, such as the existence of gravitational waves. So far, this theory has not resulted in any surprising inaccurate prediction, so it remains valid.
  • The Big Bang theory, devised by Georges Lemaître in 1931, was adjusted to explain the Hubble-Lemaître law, discovered by Lemaître four years before, and by Hubble two years after Lemaître. In 1948, George Gamow, Ralph Alpher and Robert Herman made two surprising predictions from this theory: the average composition of barionic matter in the cosmos, and the existence of a cosmic background radiation. Both were confirmed (and therefore became surprising accurate predictions) during the 1960s, which made the Big Bang theory the only cosmological theory valid at that time.
  • The standard cosmological model, formulated at the beginning of the 21st century. Its six independent parameters have been adjusted so that the theory correctly explains the acoustic waves detected in the cosmic background radiation, which were known before the model was formulated. In fact, five of the parameters have been adjusted to make the sixth (the curvature of the universe) as close as possible to zero (flat universe). However, this theory has not been validated, and is currently in the situation in which general relativity was between 1915 and 1919 (i.e. it’s a well-adjusted theory, but not a validated theory).
The cosmic background microwave radiation
Can the standard cosmological theory be validated? Has it made a surprising prediction that we can confirm? Yes, it has. It predicts that the universe contains something (we don't know what it is, thus call it dark matter) with a mass five times greater than that of the baryonic matter we know. If in the future dark matter were indeed discovered, and if its mass were exactly as expected, the standard cosmological theory would be validated... for the time being, until another surprising prediction (perhaps the existence and value of dark energy) would prove to be false. Then we should find another model to replace it. In fact, there is already one out there, called MOND (Modified Newtonian Dynamics) that claims not to need the dark matter hypothesis to explain the movements of galaxies.
This is how science advances.

The same post in Spanish
Thematic Thread on Standard Cosmology: Previous Next
Manuel Alfonseca

No comments:

Post a Comment