Wednesday, 23 November 2016

I Had No Need Of That Hypothesis

There's a common argument that science must exclude God and the supernatural from its enquiries. This argument fundamentally misunderstands what science is, what it does, and what its remit is.

I've been plugging away at an offering on the inanity that the Earth is expanding, but my eyes keep glazing over, so I thought I'd set it aside for a short while and delve into the more light-hearted arena of scientific epistemology and address this and some related arguments having to do with the burden of proof and absence of evidence. These arguments are somewhat ubiquitous crutches of the religious apologist, and I intend to shred them, but let's first some housekeeping.

The place for us to begin is to talk about what it is to be a sceptic. Scepticism is often presented by the apologist as the position that everything should be doubted. This is a massive simplification of what scepticism really is. Scepticism is a process by which we assess truth-claims. It involves only ascertaining whether the available evidence is sufficient to warrant tentative acceptance of any given assertion, and whether there's anything standing in opposition to it. It's a heuristic tool, formulated to improve our understanding of the world and our confidence in the results of our enquiries. Properly applied, scepticism places limits on what can be deemed knowledge, and teaches us to take great care in our thinking. This self-same heuristic, when applied to phenomena, goes by another name: Science.

Science is a branch of philosophy. Many treat philosophy and science as having different remits, and even go so far as to say that philosophy deals with the questions that science can't answer. In some respects, I've already addressed these claims in earlier posts, particularly in Who, What, When, Where, How, Why? and Deduction, Induction, Abduction and Fallacy, which dealt with what philosophy really is and how logic is employed in the sciences respectively, but here I want to be a little more explicit, particularly to address some popular misconceptions.

In the first of those posts, I talked at length about the disconnect between what people perceive philosophy to be and what it actually is. Specifically, I argued that philosophy is about ensuring that we're asking the right kinds of question. Different branches of philosophy deal with different kinds of question, largely differentiated by the means by which we can test proposed answers to said questions. 

In mathematics, we test our proposals by constructing proofs which, in practice, entails starting with statements that are taken to be axiomatic, and progressing from them to consequences that must also be axiomatic, as long as we haven't made any logical mis-steps. In science, we test our ideas by means of the predictions they generate and their correlation with observation. In other areas, there's not a huge amount we can do beyond testing our ideas for logical consistency. Ultimately, unless we can make contact with an observation at some point, or otherwise demonstrate that our starting assumptions are true, then our ideas have little epistemological value (there's an important caveat here; thought experiments have had great value in science in helping us to formulate better questions, but they must still make testable predictions).

There are some important ideas that we need to take into account when we're evaluating hypotheses, and among the most important is the idea of parsimony. Parsimony is simply economy, and it enters scientific thought in the form of a principle popularly known as Occam's Razor. This principle is often horribly misunderstood, not just by the laity, but even academic sources such as the Stanford Encyclopaedia of Philosophy and the Internet Encyclopaedia of Philosophy. In those and other sources, it's formulated as something along the lines of  'the simplest hypothesis is the best'. This is not a million miles away but, as we discussed in Irreducible Complexity and Evolution, simplicity and economy don't reside on the same spectrum.

Properly, the opposite of simplicity is complicatedness, while the opposite of economy is complexity. Complexity is what Occam's Razor actually deals with. Where two competing hypotheses are compared, we should select the one with the fewest entities or assumptions. That's not to say that the more economical or parsimonious hypothesis is necessarily correct, only that we should rule out the more parsimonious hypothesis before moving on to more complex hypotheses. The motivation for doing this should be reasonably clear. Obviously, the more economical ideas are going to be the easiest to rule out, in a process that underpins all scientific progress. This is where we'll go next.

Another critical idea in science is the notion of falsifiability. This was famously formalised by Karl Popper as a solution to the demarcation problem, the problem of how to separate scientific ideas from non-scientific. It tells us in a nutshell that, if there is no way in principle to show that an idea is incorrect, it isn't possible to test it, thus it isn't a scientific idea, and ultimately has no epistemological utility. Any such idea can and should be discarded on that basis alone. 

A really important feature of any scientific idea is the null hypothesis. In rigorous terms, this is a prediction about what, if observed, will falsify a hypothesis. This is what solves the demarcation problem, as it provides a means by which we can robustly test our ideas. 

Is God a scientific idea? This is a thorny problem, and it's kept philosophers up at night for centuries. Part of the problem is that there's very little information in that word for us to be able to reasonably assess it for scientific merit. There's actually an entire school of atheists whose position can be summed up as 'no deity has ever been sufficiently coherently defined'. These are the theological non-cognitivists, and they have a point. Many apologists will direct you to the 'God of classical theism', as if that answers the definition question but, insofar as this entity has been defined at all, the definition is highly problematic. I address some of the issues in All Kinds of Everything, in which I treat the famous 'omnis' from a logical standpoint, but the problems run far deeper than that because, aside from these logically absurd attributes, the entity still hasn't been defined in any robust sense. That said, this particular entity can be considered a scientific idea of sorts, because it's at least possible to falsify it, and this has been done.

Those uninitiated in public debate will define it as 'the God of the bible' but this is even more vague, and runs into issues as soon as you begin to address the claims, not least because the apologist will often insist that the counters offered don't apply to their conception for one reason or another. It's a kind of bait-and-switch, in which the switch is never completely made, only denial of the specific attribute at the root of the objection.

A corollary problem with this is that God seems to be a different entity for each believer, and it's difficult to reach agreement between one believer and another. This is the reason that there are so many denominations of the large religions. The problem goes even deeper than that, and there's an old joke that deals with it. It concerns the First Baptist Church, and how the Second Baptist Church can tell you everything that's wrong with the first. Matt Dillahunty quipped that it's even worse, and that those in the second row of pews in the First Baptist Church can tell you everything that's wrong with those in the first row.

The point is that there seem to be as many conceptions of deity as there are believers, and it's changeable. It simply isn't possible to debunk several billion conceptions of deity at once, and that makes the idea of god functionally, if not in principle, unfalsifiable, and therefore unscientific. 

Individual conceptions are very much falsifiable, though. Any entity that is proposed to manifest in any way in the world should be falsifiable, dependent on the details. Indeed, it isn't stretching the point to far to suggest that any entity that can circumvent the laws of the universe on a whim is comprehensively falsified by the existence of science, as any universe containing such arbitrary processes would be one in which science would almost certainly not be possible.

There's a marvellous book by physicist Victor Stenger, God: The Failed Hypothesis, which fairly concretely eviscerates the idea of a deity as a scientific hypothesis, and I highly recommend it.

Anyhoo, there's a famous anecdote about Pierre-Simon Laplace presenting his masterwork Mécanique Céleste to Napoleon. Napoleon, apparently fond of asking awkward questions, remarked:

"M. Laplace, they tell me you have written this large book on the system of the universe, and have never even mentioned its Creator."
To which Laplace replied:
"Je n’avais pas besoin de cette hypothèse-là."
No need of that hypothesis indeed.

In short, science doesn't remotely need to exclude god, it just needs not to include it until there's need of the hypothesis, and until said entity is sufficiently well-defined to bring it into the realm of testability. If you can present evidence for such an entity, bring it on so we can test it. The same goes for any other asserted entities and phenomena.