The theory of evolution appears to have a subtle flaw, according to a thought-experiment described here. Evolution has a vast range of scientific facts supporting it, of course, but facts are only half of the truth-equation. The other half is logic. According to the thought-experiment, the theory of evolution does not pass the test of logic and would appear to be wrong on that account. The difficulties are the same as those causing the 'measurement problem' in atomic physics.
The videos on YouTube are:
http://www.youtube.com/watch?v=C_oEALmYr-g (Free Will Explained as Force)
http://www.youtube.com/watch?v=wDysumXVmsU (Free Will Explained in Terms of Logic)
http://www.youtube.com/watch?v=9QzAMbBL_e4 (Free Will Not Explained by Mechanism)
These links refer to the videos on YouTube. They are not hosted on this site. This is due to a current lack of space. If you would like to host the original videos (about 100 MB each), please send an email to the professor at this site. (thanks!) 100328
I was writing an article comparing two approaches to scientific truth - facts versus logic - and used Richard Dawkins and the theory of evolution to demonstrate the fact-based approach. Then I chose Einstein for the logic-based approach. In the case of Einstein, I set up a 'thought-experiment' to show how a new biological species might arrive on Earth according to logic. But instead of coming up with a logical alternative to evolution's facts, the experiment actually showed the theory of evolution to be unsound.
The problem is that evolution produces a new biological object by relying on randomness, at least in part. This makes the new object unpredictable in principle i.e., there are no logical pointers to it. The only way we can say that the object exists is to observe it, and until we do that the default position is that it does not exist. This poses an obvious problem for the evolution of man-the-observer, because the evolutionary pre-cursors that are going to evolve into man do not achieve reality until man observes them.
If the thought-experiment is valid, it is necessary to explain how the facts of evolution can be wrong. It turns out that the facts of evolution are not wrong (of course), but they have been selected. They point to 'selected reality' rather than 'reality'. Selected reality is different on account of it being selected.
Imagine that scientists have set up a computer experiment modelling biological evolution. The scientists have programmed a computer with the Darwinian algorithm which includes a random starting point, random variation and a suitable selection environment. The scientists hope to see evolutionary products - animals, for example - on the computer screen, starting from nothing. For the purpose of this thought-experiment it does not matter whether the randomness used by the algorithm is genuine or provided by a pseudo-random number generator. Neither does it matter if the selection environment is biased towards evolution-friendly conditions. The important thing is that the scientists have no idea what the Darwinian algorithm will output to the screen. [link not yet implemented - chapt 3 Blind Watchmaker]
At first the computer shows nothing of interest. The scientists leave it unattended for a few weeks and when they return, the screen is showing a strange new animal species with 3 eyes and 5.1 ears. The experimenter who views the screen tells the others. Soon they are crowded around the screen congratulating each other on how well their experiment has modelled real-life evolution.
(Present-day computer simulations of evolution do not have this level of sophistication. They produce only changed allele frequencies and other low-level effects. But this is a thought-experiment. Sophisticated evolutionary outcomes are not ruled out in principle.)
Now comes the important part. Let us imagine that the computer's output has not been viewed by the scientists but has been streamed to magnetic tape, which has not yet been played. Question: Is the creature with 3 eyes, etc, on the tape? The answer strongly polarises people. One group, the majority, say that of course the creature is on the tape. We can prove it by pressing play. They also say that the creature does not come into existence when we play the tape but was always present in the pattern of the tape's magnetic particles. The fact that we have yet to view the tape is irrelevant to the reality of what is on the tape... and so on. This seems such a common-sense position that most people adopt it.
The other group of people - and I imagined Einstein as one of them - would say that the creature does not exist on the tape. They would say that the word 'it' - as in 'it is on the tape' - does not refer to anything. The statement 'it is on the tape' only makes sense when the tape has been viewed and we know what corresponds to the 'it'. Most philosophers would agree with this position, but they would express a caveat which I deal with shortly.
Suppose that we get God to decide the issue. He will use his X-ray vision to see if the creature is on the tape. (God speaking): 'OK. I will check to see if it is there. Tell me what I am looking for.' The scientists say they do not know. (Remember, the scientists have not viewed the tape.) 'Well, what is it called?' They do not know. 'Describe it.' They cannot. 'Well, whereabouts on the tape is it?' Again they do not know. 'Are you sure it is on that tape? Is it the only thing on the tape? Is it next? Is it first?' The scientists tell him they know absolutely nothing about it. 'In that case,' says God, 'it is irrational for you to say it exists.' The scientists ask God to play the tape and report what he sees. God says that there are a near-infinite number of objects on the tape (combinations of the metal particles in various ways). Which of them do they want him to report on? If he reports on them all (and particularly if the tape is infinite in length and the creature exists only for an hour) then he would not be picking out a particular object and no particular object would exist. And if he did pick out an object, the reality of that particular object would be determined by his selection. It would not be reality determined by itself.
In this way God would agree with the philosophers that the creature is not on the tape. But the philosophers would express the following caveat. The creature could be on the tape, without any need to view it, if there are logical pointers directed at the creature. God was fishing for such logical pointers when he asked if the object was the 'only' thing on the tape, or the 'first' thing or 'next' thing. A number of such pointers exist in real life - new, only, first, next, previous, after, before, now, this, that, my, our, like, similar to, and so on - which can be relevant in various contexts. (There can even be non-verbal pointers, like someone grabbing you by the elbow and taking you to see 'this'.) But evolutionary theory has been designed to avoid logical pointers. Richard Dawkins expresses this by saying that the watchmaker is blind. He means that evolution is not determined to produce any particular object - instead, what you get is what you get. Einstein would attribute evolution's lack of logical pointers not to a blind watchmaker, but to the incorporation of randomness in the theory. If evolutionary outcomes are produced by randomness (or are dependent on randomness to an essential degree) then of course they will not be predictable.
The lack of logical pointers in evolutionary theory means that we are absolutely dependent on seeing evolution's products to effect their existence in our world. Until we view the tape, the creature with 3 eyes and 5.1 ears does not exist in our world. We might imagine that it exists in some other world - particularly its 'own' world - but there is a logical gulf between that world and our own. We can know nothing about worlds that are not 'ours'. It is not helpful to say that something exists, or does not exist, in some world that we have no knowledge of - particularly when we cannot even conceive the object to ask whether or not it might exist in that world.
The reasoning suggests that we cause the products of evolution to exist by seeing them. The products of evolution do not exist for us to stumble upon, but come into existence upon our observation of them. This bizarre-sounding idea is a well-known phenomenon in atomic physics, where the result of a quantum measurement only becomes real when the output dial is read by the experimenter. In atomic physics, if the experimenter avoids looking at the dial but takes a photo of it, the result of the measurement only becomes real when the photograph is viewed. This is the 'Copenhagen' viewpoint analysed by John Von Neumann in the middle of last century. The alternative to the Copenhagen viewpoint is the 'Many Worlds' theory which I will have more to say about later on. The Many Worlds theory suggests a possible resolution to the evolution problem.
Let us accept for the moment the result of the thought-experiment - that animals, plants and other evolutionary products only become real when we observe them. How does an evolutionary object, created by our seeing it, come to be the objective scientific truth that it is? The answer lies in what we do immediately after seeing the new object. It seems that once we have observed an unpredictable object into existence, we immediately give it a grounding in reality so that it no longer depends (and never did depend!) on our seeing it. We do this because we conceive our world to be an orderly one of objective causes and effects with ourselves as outside observers of those effects. We make a determined effort to distance ourselves from the reality we observe. So if we discovered a new species of animal in the jungles of Borneo - a 'meringue-outang', say - our first reaction would be surprise at seeing something new but then we would move on to working out a habitat to support the animal, to identifying a genetic chain linking it to others already on our books, and to working out the animal's niche in reality. The time sequence would be (a) observing something, then (b) justifying what we have observed. This is a sequence that our logic ought to respect. If observation is needed at the start, we must accept that this is how things are. It would be logically unsound to rely on observation then to pretend that the observation was not necessary and did not need to be incorporated in our theories. I will be suggesting later on that 'observation' is a matter of logic, i.e., there is a logical reason for the world to have the objects that it has.
The result of this thought-experiment is quite devastating and it is natural to look for holes in the argument. It is of course possible to finesse a way round any argument (or to ignore an argument entirely if that's what you want to do). But evolution, the procedure which brought you and me into existence, is extremely important to us. We want the theory to be robust and to command our respect. Also, it would not be honest to reject the argument here without rejecting the equivalent argument affecting atomic physics. The measurement problem in atomic physics is still a very pressing problem.
One finesse is to break the link between the thought-experiment and real life. We could say, for instance, that computer simulations do not accurately model evolution. But that in itself would be a Big Thing. Biological scientists tend to agree that there is nothing wrong in principle with computers modelling evolution. (See chapter 3 of Richard Dawkins' The Blind Watchmaker, for instance, including the appendix written 5 years later.) To make this objection stand up, somebody would have to show why computers are not logically capable of doing the job.
Another possible approach is to attack randomness. The ingredients of the argument that cause the trouble are the creature being 'new' and its origin being random. If we could adjust evolutionary theory so that its products were predictable, we could avoid reliance on observation. But it is difficult to see how evolution could remain a satisfactory theory without randomness.
Most of us react in horror to the idea that observation is required to make things real. Things just are real, we like to think - they don't need observation to make them so. That reaction, while natural, does not stand up to the logic. The logic clearly indicates that things new and unpredictable-in-principle are not real until they are seen. It is logically unsound to use the word 'it' to refer to something that doesn't exist.
But people often do that. They say to themselves 'There is no need for someone to observe an object to make it real - the object is real in itself. Effectively it sees its own existence and that's all that matters. What's all the fuss about?'
The fuss is over the words that the person is using - the words it, object, thing, and itself. What do those words refer to? In computer jargon, where do they point? The words cannot point to 'the thing itself' because that would be circular. If the only knowledge of a thing's existence is supposedly held by the thing itself, then we are excluded from that knowledge and are not able to refer to the thing.
Many people conjure up in their mind a vague object that has no characteristics at all except that it exists in itself ('really exists'). Then by observation of the world and making intelligent deductions they proceed to clothe the imaginary object in properties which eventually make it concrete enough to demonstrate to other people. At that point everybody agrees that the object is real. The person uses this endorsement to justify the first step in this process - their conjuring-up of the object to exist 'in itself'. The object must have really existed back then, they say to themselves, otherwise it wouldn't exist now and contribute to the reality that everybody agrees is in front of us.
This lacks scientific rigour. The first step, where we imagine the existence of a 'thing' with no characteristics except the possession of a 'self' that allows it to exist in itself, is irrational. The remaining steps are dependent on human action. The whole process is a drawn-out type of 'observation'. It is us tweaking the world's reality to get something coherent from our point of view. If reality is to exist 'in itself', it cannot depend on us tweaking things like this.
'Evolutionary theory might defy logic, but that's the beauty of it - it works regardless!' This is another would-be solution to evolution's logical problem. Of course it is irrational. But many people seem to think that irrationality - a defiance of logic - can be acceptable in tricky situations. Scientists think otherwise. They accept the pre-eminent position of logic. If something offends logic, it is wrong. When logic comes into conflict with the facts, there is always the possibility of reinterpreting the facts or looking at different facts - but defying logic is never an option.
Personally, I see no need to be precious about an absolute status for reality. I come from a physics background and physicists have long considered that reality might not exist independently of us. Most physicists agree that things exist independently of individual people seeing them, but we are open to the idea that humans might be involved in bringing that reality into existence. This happens with the Anthropic Principle, for instance. That principle suggests that the early conditions of our world were/are selected by the brute fact of our human existence. It is an example of us unwittingly causing the reality of the past. In other physics' experiments, the world's reality seems to depend on what we choose to measure, or a single object might appear from our point of view to have parts separated by a million kilometres when in reality the object does not have parts. There are several such reality quirks lurking in atomic physics, and they lead physicists to be rather liberal about reality's status.
But this is not the case with ordinary people. Most ordinary people are fiercely dogmatic about reality existing before all else. They do not want to listen to philosophical objections. 'Let's just assume that reality comes before all else and get on with the job of learning about it.'
Physicists have more respect for the truth and are particularly critical of people who think they that truth can be represented by assertion or dogma.
In the following pages I attempt a solution to the logical problem revealed by this thought-experiment. My approach is based on the similarities between evolution and quantum measurement. Physicists have proposed several solutions to the quantum measurement problem, but until now there has been little reason to take any of them seriously. That's because quantum effects do not impinge on the lives of ordinary people. That now needs to change. Evolution is important to people. It is intolerable to suppose that the mechanism that brought you and me into existence is wrong.
Physicists tend to agree [link not yet implemented - Nick Herbert] that the solution to the quantum measurement problem will probably require a change to our generally accepted notion of reality. Evolution's logical problem will require a rethink of reality too, given its similarities with the quantum measurement problem. The solution I put forward here - called 'logic first' - is based on the idea that reality is not the primary existent thing in the world. Instead, the thing that exists before all else is logic. Logic chooses reality, or, to express it another way, there is a logical reason for our world.
Before describing that approach, I will address the vexed question of how the facts of evolution can be correct but the theory wrong. Next page - Page 2