What do we know?

I wrote this after midnight, whilst quite tired. Apologies for any errors. If any are caught, please comment them and I'll update the post if I agree (or discuss it in comments if I don't!)----------------- I often hear people saying things like 'Science proves that....', or 'Prove to me that this is safe'.

For example, 'Scientists can't prove to me that MMR is safe'.

This is perfectly correct. Science cannot do this. Science can never do this. The thing is that many people have the wrong idea about how science works.

Science is about making 'models' of how the universe behaves, the models then allow us to make predictions about how the universe behaves in the future. When the models are right, we trust the model more, but we can never trust it 100% as we may not have tested it in the situation where the model fails. However, if it fails, just once, then we know that our model is incorrect and needs rethinking.

This was the fundamental idea of Karl Popper.

Of course, real life isn't like that. When physicists began to apply their understanding of Newton's law of gravitation, they found slight discrepancies in the outer planets. Rather than simply throw out Newton, they looked for other explanations, could there be something causing the discrepancies within the existing model that they hadn't taken account of. They did the calculations and suggested that there must be a planet in a larger orbit than Saturn. They were able to predict its position and mass.

The planet Uranus was just where it was expected to be.

After some time, Uranus' orbit was not as expected, and this in turn gave rise to the discovery of Neptune, and later, Pluto. Instead of throwing out an idea wholesale, where there are discrepancies the existing model is 'stretched' to encompass it.

There were similar problems with the orbit of Mercury, so physicists tried to repeat the idea which was so successful with the outer planets. When they did the sums they predicted a planet closer to the sun than mercury, but the planet was not there.

This problem could not be solved within the confines of Newton's laws, yet the 'perihelon of mercury' problem still did not cause Newton's laws to be disregarded, by this time, they were found to be too powerful, and too useful to bin. The problem was shelved.

There were several issues like this, Michelson and Morely failing to find the ether, for example. The ether was a medium which light was thought to travel through. Eventually the number of 'shelved problems' became so large that the shelf 'collapsed', and at the start of the 20th Century there was a 'paradigm shift'. The worldview changed. Quantum Mechanics came along, as well as Special and General Relativity. When the equations of General Relativity were applied to the perihelon of mercury problem, they predicted mercury's orbit correctly. In addition, they agreed with Newton for the outer planets. The new theory explained everything the old one did and something it didn't.

Similarly, Quantum Mechanics did a similar job in the realm of 'Black Body Radiation' (why hot things glow), why we have atoms, and so on.

Our new models are incredibly successful, though they may sometimes appear weird to us. The reason they appear weird is because their effects only become noticeable for the very fast (special relativity), near the very massive (general relativity) and the very small (quantum mechanics). The rest of the time, 'classical' or Newtonian physics is good enough.

Due to the weirdness of the new models of the universe, these models have been tested to phenomenal degrees of accuracy, and have not been found wanting. However, can physicists ever really say they have 'proved' relativity or quantum mechanics?

No. The most that can be said is that they have failed to disprove it.

Indeed, we know that at least one of the models (and probably both) will need revision, as there is conflict in the realm of the very fast small things!)

Going back to MMR and autism. When we hear a scientist saying 'we have failed to find a link', this should not be taken as 'it is safe', nor should it be taken as 'we didn't look hard enough'.

It should be taken as 'we looked hard for a link, but couldn't find one - that doesn't mean there is no link'. Similarly, it doesn't mean that there is!

It can be interpreted as 'If there is a link, it's quite hard to seperate out from the sort of fluctuations we'd get by chance alone'.

When one parent has a child which develops autism shortly after an MMR jab, their mind looks for causes, for links. These links may exist, or it may have just been coincidence - but it isn't easy to find out!

For example, science has shown that the average smoker dies younger than the average non smoker. This discrepancy is much larger than would be expected by chance alone. This does not mean that there might not be some other explanation, but any other explanation must explain the discrepancy at least as well. Despite this, I still know of people who justify smoking by citing an uncle who lived to a ripe old age on 40 a day... one exceptionally long lived smoker doesn't change the overall patterns!

To cite another example, there was a case where Israeli fighter pilots had an abnormal number of children of a particular sex. An investigation was done to try to see if it was something to do with the tight suits or the high accelerations. Over time, the distribution of the sexes reverted to the mean. Essentially, it was a freak anomoly. If one tosses a million coins in the air, it would be freakish if the distribution of heads and tails were evenly mixed, one would expect heads to come up in some areas in little clusters.

This is what happened with the Israelis, there was a freakish cluster of babies of one sex, produced by chance. The scientists can never say that there was no cause, but by making prolonged observations they can produce a 'confidence level' about how likely it is that the effect is other than chance - and after a while the 'blip' disappeared.

Science can't give certainties. Scientists don't know everything, they leave that for the politicians. What science can do is make postulates and attempt to show that they're flawed. The more these attempts fail, the more that the postulates are trusted, but we can never say they're 'correct'. Even when an idea is known to be wrong (e.g. Newton's theory of Gravitation), it is often 'good enough'. The apollo programme did not need Einstein's theory of Gravitation to land on the moon, it would have been cumberson and the corrections negligable!