There is a claim that forms an integral part of the Teleological Argument for the existence of God, an argument that claims the universe has clear signs of purpose. This claim is that a life-permitting universe is so unbelievably unlikely that no thinking person should be able to say it happened due to chance. The argument implies the assertion, then, that life was the purpose of the universe.
Trying to explain the argument ends up taking you in circles: the fact that the odds of life are so astronomically small, means it is apparent that life if the goal; then it is only because life is the goal that we can apply significance to the odds against it. And around and around it goes.
But, ignoring the circularity for a moment, what are the actual claims?
The idea is that there are fundamental values in the physical laws that, if they were anything different, would not permit life. For example, if gravity were stronger the early universe would have collapsed back in on itself, or if the strong nuclear force were any weaker, then nucleii would never form and there would be no element to compose life out of.
It seems to depend where you look to see exactly what the numerical odds are against life. William Lane Craig says Roger Penrose calculated this value to be 1 in 10123 (Craig, no date). Penrose’ work isn’t actually cited, just his name, so I can’t find out what assumptions Penrose made or if he even made this calculation. (I also can’t find when Craig published the essay in which he says this; none of his references are post 1989.)
The argument that goes that:
The actual argument
- This low probability event actually happening has 3 possible explanations: physical necessity, chance or design;
- It isn’t chance or necessity;
- Therefore it is design.
It is worth saying that Craig doesn’t refer to this as an “event”, but as “Fine tuning”, and other apologists tend to follow his lead. But, that term is slightly confused and so I have avoided it.
Quick Fire rebuttals
- You can go about striking off the options in a different order and reaching a different conclusion: it isn’t design or chance, therefore it is physical necessity; it isn’t physical necessity or design, therefore it is chance. We have good reasons to not accept any of the three options and remain in that dreaded cloud of ‘I don’t know’.
- We can rule out chance because the odds presented are so very low.
- We can rule out design because there are no known entities not contingent on the universe capable of such a design.
- We can rule out physical necessity as there are no known mechanisms that would lead to such a necessity.
- In principle, we can avoid the probability question by invoking a multiverse. We don’t have to believe a multiverse is real ― just to point out that the possibility of a multiverse creates an option in the argument that combines physical necessity and chance: it’s a probabilistic outcome, but nature played the numbers game.
- In a game of poker, every hand has the same odds against it. It is only when a hand that is already defined as significant comes up that we notice the odds against our hand. A royal flush isn’t less likely than any other hand, it’s just more significant. So, I take you back to the circularity mentioned earlier: the fact that the odds of life are so astronomically small, means it is apparent that life if the goal; then it is only because life is the goal that we can apply significance to the odds against it.
- God doesn’t need these fundamental values to be ‘just right’ for life; God can do what God wants, and if God wants to create life in a universe where life appears physically impossible, then God can. The fact that life emerges in a universe where life can emerge appears like evidence for naturalism, not supernaturalism or theism.
- These might not be fundamental laws, and the fine tuning problem might go away with a better understanding. This has already happened once. The expansion rate of the universe used to be one of these ‘finely tuned’ parameters, with a probability against it being what it is of 1 in 1060. But, if you derive the expansion rate of the universe from General Relativity instead, it becomes a physical necessity (i.e. probability of 1 in 1).
The rebuttal I mean to focus on today
Sean Carroll describes the kind of work you would have to do to actually decide on the probabilities against the universe fostering intelligent life. Step one is to create some mathematical space that counts all the possible universes. Step two is to create some mathematical space of all universes that could have life arise in it. Step three is to do an integral of two against one to derive a probability. His point is that nothing like this has been done or is being done.
Here are the problems with the steps:
Starting, methodically, with step one, we will take a brief look into what a space for all possible universes might look like. Limited by my own lack of intelligence and a 2D screen to display my work on, I will start by illustrating the issue with one arbitrarily chosen physical parameter. Lets just call it ‘Parameter 1’.
We have a lot of important questions to ask at this stage: what are the upper and lower bounds for Parameter 1; what are the highest possible and lowest numbers? Give a continuum from the (unknown) lowest possible value (“-P?”) to the (unknown) highest possible value (“+P?”), what is the probability distribution: are all outcomes equally likely, or are some outcomes more likely than others? How we could discern the probability, given that we can only sample the one universe we are in, seems like a hugely speculative project.
This one parameter, Parameter 1, does not constitute a space that counts all possible universes yet. Wikipedia claims there are 25 such parameters (Wikipedia contributors, 2018) for us to sort out like this. And I say “sort out” rather loosely, as we haven’t actually sorted anything out ― we’ve just raised some questions fundamental to the methodology.
But let’s move on to step two before we start confusing step one with many parameters. (We’ll come back to that.) The first question is where the current value for Parameter 1 lies between -P? And +P?. If the probabilities are equal, then this doesn’t matter, however, if different outcomes have different probabilities, then whether we are sat in a relatively high or low probability value does matter; it describes the probability (along that particular parameter) of the Life that is us (LU). More interestingly we can ask: is there some other completely different value that ― holding all other constants equal ― would help give rise to a other types of life (LO)?
We don’t actually have a good definition of life or a good understanding of how life came about in this universe. We don’t really know if the universe is teeming with life or if life arose of Earth against the odds. Therefore, we really don’t know what the upper and lower bounds for the values of Parameter 1 could be that would still give rise to life as we know it; that is why they are denoted with question marks, again: -LU? and +LU?. And we really don’t know if other types of life would emerge under very different values.
There’s far too many unknowns here to discuss clearly, so let’s simplify the model down to considering life only as we would recognise it and then go back to step one to introduce another of these 25 parameters: Parameter 2. Consider the whole graph space the mathematical model of all possible universes, and only where I have coloured Green to be a ‘Life as we know it’-permitting universe. You can already see that if you throw a dart at the graph from a distance, you’re not likely to hit it; that is quite a small area where life permitting conditions for both Parameter 1 and Parameter 2 line up.
But there is a paler green area and a diagonal dotted line. What is all that about? Well, it’s a visual representation of yet another question: if we change the value of both Parameter 1 and Parameter 2 together, is there some other area on the graph where life as we know it could emerge; the pale green area with a question mark in it? Even more intriguing, is there an entire relationship along the dotted line where turning one value up and the other down preserves life as we know it?
I originally created the graph with other life included in different colours, but it became too difficult to read. That said, don’t take that as a cue to discard the question of other types of life.
The pattern continues: where two parameters create a large 2 dimensional ‘space’ to depict all possible universes, and a smaller ‘space’ in green depicting all ‘life as we know it’-permitting universes, considering 3 parameters creates a 3D conceptual space with a green cube inside it. After that, a 4 dimensional space… etc.
Step three is to do an integral equation on the 25 dimensional space and the 25 dimensional green area within it. If, in step one, you figured out that the probability of each value in each dimension is equal, then you have an easy division question: what percentage of the total space is occupied by green space? However, if the probabilities are not equal, then you have a much more complicated picture in front of you: imagine the space varies in density (‘probability’) across its entire reach; the question now is what percentage of the total mass of the area is overlapping with the green space?
Every step of the method to actually calculate the probability against a life permitting universe is wrought with currently unanswered questions. And this makes the whole project meaningless.
Firstly, are the multiple variables you are considering really the variables you need to consider? As we saw briefly with the expansion rate of the universe, the variable can evaporated quite easily.
Secondly, once you have established with variables, what are the upper and lower bounds for what that variable could take?
Third, what is the probability distribution across each variable?
Then, life: what is it? This question defines the areas across the distribution that are considered significant. I simplified the models I presented by reducing it to life as we would recognise it: LU. But there is the consideration of other life (LO) out there in conceptual space that would ask the same questions if it came to be; that life is equally significant to the question.
Lastly, how does the permissibility of life in a universe change as more than 1 variable is altered at once? If one parameter is reduced life becomes impossible, but could you compensate by turning another parameter up?
All of these questions have to be answered before we can get to some mathematical count of all possible universes and then the count of life permitting universes inside it.
Craig, W.L. (no date) The Teleological Argument and the Anthropic Principle | Reasonable Faith. Available from: https://www.reasonablefaith.org/writings/scholarly-writings/the-existence-of-god/the-teleological-argument-and-the-anthropic-principle/ [Accessed 6 September 2018].