The facts of evolution are things like the fossil record, the DNA similarities between species, and controlled experiments where evolution can be seen in action. No one seriously disputes the veracity of these facts. Indeed, it is illogical to suppose that facts can be wrong - they are correct by definition. Nevertheless, the theory of evolution that binds these facts together is not logically sound, so we need to look into the matter more closely.
As it happens, science has always recognised a problem with 'contingent' facts. Contingent facts are facts that are true by observation rather than by logic. They are not obliged to be true, but we can see that they are. The problem is that there are an infinite number of contingent facts about the world and it is not possible for us to have knowledge of them all. So any contingent knowledge we have will be selective. It will still be true, but what we see is not so much 'reality', but 'selected reality'. This does not seem to be significant difference, but any difference must be accounted for. In science, the selection involved in 'selected reality' will not be deliberate, because scientists try to be scrupulously honest in their methods. This is particularly so in the 'hard' disciplines such as physics. But the problem is with the selection itself, not whether or not we do it deliberately. If we make the selection accidentally, the reality revealed will still be selected reality.
Here is an example. Suppose we want to know whether there are purple heptagons on Mars. After much effort we could find out the answer, which, yes or no, would thereafter contribute to the world's logic as a fact of reality.
Instead of purple heptagons on Mars, we might want to know how many holes it takes to fill the Albert Hall. Again, we could apply a certain amount of effort and find out the answer - say 20,000. That answer would make a contribution to the world's logic as a fact of reality. But it is a different answer from the purple heptagon answer. The world's reality is now different on account of the different questions we asked.
So we end up with a different world - a world with different logic - because we decided to investigate different facts. This is not to say that one of the facts is wrong - both are correct. But they do not apply at the same time. Actually, 'time' is not the correct word. It is more correct to say that the two facts apply to different worlds. We can think of it as one fact 'displacing' the other one, rather than negating it. Both facts can coexist but for various reasons they do not. If we humans had infinite power and intellect, we might know all of the world's facts simultaneously, without some preceding others. But we have limited abilities. We investigate some facts after we have ascertained other facts. Our choice of facts - the facts we choose to investigate - will be coloured by the facts we already know about. This is not to say that the facts themselves will be coloured. No, they will be 100% correct. But the reality portayed by those facts will be coloured by the selection process. It is a simple matter of a world with fact A being different from a world with fact B, regardless of whether the two facts are related to each other.
We can scale this to the whole world we live in. Our world includes a huge bundle of contingent facts that are true because we see them to be so. This bundle of facts that we know about displaces a bundle of other facts that we do not know about. The alternative facts, if we had learned about them instead, would have resulted in a different world.
Science's answer, historically, is to accept contingent facts as interesting but not the final answer. Scientists want to know what the facts mean, their value in developing an hypothesis for predicting further facts. If the scientists are successful - i.e., their hypothesis predicts further facts subsequently confirmed to be true - the status of the hypothesis is elevated to law. The scientists can then use the law to retrodict the facts that went into the formulation of the original hypothesis, showing that, given the law, the original facts were not contingent but only appeared so at the time. Those original facts were based on law and could have been predicted back then if the law had been known at that time.
This procedure worked well until atomic physics was invented and it was found impossible to predict the result of quantum measurements. To make progress, the scientists modified the idea of physical law so that the predictions made by new physical laws could be statistical in nature rather than deterministic, as they used to be. This required a subtle change to the nature of randomness. The previous understanding of randomness, as it was employed in statistics, was as a tool for scientists to analyse a large number of objects or events. Randomness was a tool only - it did not 'cause' anything. The subtle change required by quantum physics and adopted by most scientists - with the notable exception of Einstein - was to make randomness a causal agent in its own right. This allowed physicists to say that the statistical outcome of a quantum measurement was caused by the action of randomness, or 'occurred at random'. The randomness was no longer an artefact of our wielding a convenient statistical tool, but existed as a reality in its own right. Although Einstein disagreed with this, the procedure seemed to get results and was accepted by most physicists. (To be more correct, the randomness influencing quantum reality is constrained - stochastic - rather than totally at random.)
So far no great problem. But then scientists began to consider what else might randomness cause. If randomness was real, where else besides atomic physics might it be influencing the world? Scientists were now tacitly accepting that randomness was an agent that implemented reality, at least some parts of it. It was not just a tool commenting on reality. One field in which randomness might play a part was cosmology. Cosmologists worked out theories of how randomness might have caused the entire world to be what it is. Another field - and the one concerning us here - was evolutionary biology. Almost every day now, a biologist (or journalist?) seems to bring out a theory of how some biological phenomenon can be explained by the action of randomness. Randomness 'causes' something, then the environment selects the thing for continued existence. (These explanations have become tiresome and I detect the beginnings of a backlash against them.) [link not yet implemented - New Scientist]
The main beneficiary of these randomness-based explanations has been the theory of evolution itself. If evolution can explain all these things, surely it must be correct.
Not necessarily. The problem is that evolutionary theory has been built upon contingent facts rather than upon deterministic law, and contingent facts lead to coloured reality. The reality is coloured because if we were to obtain different contingent facts, perhaps through asking different questions, we could obtain a different coloured reality. To get real reality - reality that is uncoloured by human selection activities - we need to move away from facts and embrace logic.
There are actually two problems. The first is that the theory of biological evolution rests uncomfortably on contingent facts. The second is that evolution requires randomness to 'cause' things. The latter is a dodgy aspect of science. Randomness being a causal agent is dodgy because the concept of cause seems to rule out such a notion. In daily life we treat randomness as the opposite of cause. This is evident when we do not know the cause of something and say 'it happened at random'. The elevation of randomness to causal status has remained controversial in physics, but in biology it is essential to the theory of evolution so biologists simply assume it.
So we arrive at a possible reason why evolution's facts can be correct but the theory wrong. The contingency of evolution's facts means that they are a selection from an infinite range of possible facts that could have been selected instead. If we had selected different facts we would have ended up with a different world. The theory that binds the facts together needs to take the selection process into account. Alternatively, the theory could bring logic into the equation and show that the products of evolution were inevitable rather than contingent. That would mean plants and animals having a deterministic origin rather than random. But evolutionary theory would not make sense if we deleted randomness.
It is a sobering thought to realise that there is nothing setting the reality of the biological world around us other than our viewing it. When we look at a fossil embedded in shale, for instance, logic obliges us to accept that it would not exist as an element of our world if we did not see it there. That's one of the consequences of contingent facts. They do not contribute to the logic of the world - which is to say, they do not exist - until they are observed to exist.
It seems that the reason we see fossils and other contingent things in the world is because those particular facts fit into the logic of our preexisting worldview - in a sense, we expect to see things of that sort. If we had a different preexisting worldview we would not expect to see those things and they would not register as existing things i.e., they would not exist.
So this is why the facts of evolution can be correct but the theory incorporating those facts can be wrong. The theory can be wrong if it does not take into account the contingency of the facts that go into its formulation. Certainly evolutionary theory takes facts into account - but does it take into account the contingency of those facts? In our haste, perhaps with some wishful thinking thrown in, we view the facts supporting evolution as absolute - as fixed-points of reality from which other absolute facts follow. We should not be so hasty. As the thought-experiment demonstrates, contingent facts derive their truth value from us seeing them. So any theory incorporating such facts must take the seeing process into account. If it does not, it is logically unsound.
One of the reasons I am happy to accept the truth of the thought-experiment exposing the deficiencies of evolution is that the situation is very similar to the quantum measurement problem in atomic physics. The similarity means that someone refusing to accept the conclusion of the evolution thought-experiment is also refusing to accept that a quantum measurement problem exists. That is a bold step. Physicists agree that the quantum measurement problem has not been solved. Wikipedia If either problem gets solved - the evolution one or the quantum measurement one - the other will automatically get a solution too.
I will describe a possible solution to the quantum measurement problem on the next page. It hinges on the 'Many Worlds' viewpoint, rather than on the 'Copenhagen' version I introduced earlier. Next page - Page 3
Back to beginning of Page 1 (Thought-experiment)