The Transcendental Argument for God (TAG) seems to be becoming more popular. It's one of the preferred arguments of Oxbridge academics and not one that you're likely to face from your typical, rolling eyes, creationist.

It picks some abstract attribute, such as logic, scientific laws or morality and then asserts that these must be contingent on a rational creator, therefore that rational creator exists. In the case of Christian apologists, they then go off on some hair-brained argument to "prove" that only a triune God as worshipped by Christianity can possibly fit the bill. I'm going to ignore this latter part.

When I first heard this argument, I was so thrown by it that I couldn't think of a response. I wasn't thrown by its insightfulness, but rather that it was such a blatant non-sequitur that I thought I must be missing something so had better shut up. It threw me in exactly the same way as the Ontological Argument did (by definition God is perfect, existence is more perfect than non-existence, therefore God exists). It seemed to be sheer sophistry.

Before addressing the three most typical versions of TAG, I'd like to look at the argument itself. The first thing to note is that it is not an argument at all, it is a simple assertion that a rational God is the only possible explanation for something. There are no clearly defines premises. There is no chain of deductive reasoning. There is no examination of alternative conclusions.

The second thing to realise is that this is just a fancy dress version of the millennia old God Of the Gaps Argument (GOGA). GOGA used to be relatively simple, something like "Who makes earthquakes? Must be the Gods." Well, no, it's caused by tectonic plates shifting. "Who made the sun? Must be God." Well, no, actually the Sun is a condensed ball of hydrogen, where gravitational collapse is exactly balanced by the outward pressure of nuclear fusion in its core. "Who designed all the animals and plants? Must be God." Well, no actually, life evolved through natural selection.

You'd think that theists would start to see a pattern here, but no, they replay the same old mistakes every time. As the gaps have shrunk and God becomes ever more squeezed and diminished, they now resort to the God of Abstractions.

I'll dispense with the specific argument from morality first. By "morals" here, I don't mean morals in the sense that the Catholic Church almost invariably does: a euphemism for sex. I mean morals in the sense of that collection of actions and traits that we regard as "good", specifically good to humans, but we could throw in humans being good to animals as well.

Given that empathy, sympathy, charity and a sense of justice have been observed in other primates and even in dogs, it's clear that morality is not something unique to humans. We see behaviour analogous to human morality in even the lowliest of creatures. The self sacrifice of an ant defending its colony from attack is just as moral from an ant's perspective as any moral action we might take through instinct. We see love and care in even the most stereotypical creatures that are "red in tooth and claw". A lioness, for all her ferocity as a hunter, will do almost anything to protect her cubs, if she didn't, her species would not have survived.

We are a social species and like all social species must cooperate and help one another, just like the ant. We must look after our children, just like the lioness. All of this is essentially instinctive. They are instincts which have evolved and survived because they are necessary for our own survival. This does not make human morals any less important. Indeed, in some ways it enhances our perspective on ethics. Knowing that our tribal nature is evolved allows us to study it in other species and perhaps learn something about ourselves.

There is also a sense in which human morality is unique. Humans posses an unparalleled degree of foresight and planning. Someone who dives into a river to save a child from drowning knows full well the danger to themselves in a way that no other animal would. They can foresee the possibility of their own death yet may still choose to follow their instincts despite this. Understanding this conflict between instinct and rational thought provides new perspectives on ethics that no amount of theology ever could.

In short, morality is both better explained and better understood through evolution than through scripture.

The next specific assertion of TAG is that the laws of logic are divinely inspired. A specific example comes from Aristotelian logic: a statement cannot be both true and false. My first objection to this is simple: who says it's a law? It's not a law by divine diktat, it's not one of the ten commandments. In fact, logic would not seem to be one of Hebrew scripture's strong points. Logic is a purely human construct arrived at through inductive reasoning.

It's childishly simple to construct statements that have no obvious truth value. "This sentence is a lie," is the classic example. So this "law" is definitely not universal. Nor is it universally applicable in the physical world: an electron can be both spin up and spin down simultaneously, it doesn't have to be in one state or the other. Other statements do have a definite truth value. "Two is less than three," is a statement that is true because it is axiomatic of Number Theory. In this case the law of Aristotle is really the tautology of Aristotle.

So what distinguishes the sets of statements with and without a definite truth value? Aristotle's law only applies to those statements that we believe have two distinct and mutually exclusive outcomes. In other words we select the statements for which we believe Aristotle's law to be true and then apply it accordingly. This is not a sufficient answer however. We need to explain how it is that we arrive at this belief that there are a set of statements that can only have one of two truth values, and the fact is, we arrive at it inductively.

Our mental model of our environment recognises it as "true" that cups fall to the ground, and "false" that they float up into the sky. We deduce from millions of empirical observations, from the moment of our birth, that there are large classes of actions which preclude other alternatives. We're not the only ones who do so. Our friend the lioness recognises that if her prey is ahead of her it is not behind her. The way to obtain lunch is to move forward. We have the ability to abstract this property of the world and call it Aristotle's law.

"You cannot base the laws of deduction on purely inductive reasoning," counter the philosophers and they may be right, but as I've suggested above, Aristotle's law is not universal and as the example of the electron shows, we should be wary of applying it outside our normal experience.

The final example of TAG that I'd like to look at is the argument that the universe is orderly and that someone must have imposed that order upon it. Once again this can be dismissed on the basis that it's an unfounded assertion whose conclusion simply doesn't follow, but I'll go a little further.

There's no reason that I know of why every electron has to be identical. Yet this appears to be the case, so much so that it has quite profound implications - the periodic table would be impossible if it weren't true. There's no obvious reason why the universal gravitational constant, or Planck's constant should actually be "constant". Facts like these strongly suggest, so the argument goes, the imposition of a design.

Let's consider the case of a universe that is completely random. Forces come and go, they vary in strength over time. Such a universe would be chaotic and unpredictable. Or would it? The first thing to point out is that many of our most successful scientific theories are, at root, based on total randomness. Quantum Mechanics, Thermodynamics and Evolution are all based on randomness. Yet out of these apparently random behaviours, emerge statistical patterns that we label as "laws". The second law of thermodynamics, that the amount of energy available to do work always decreases, is one of the few laws of physics that can be deduced from first principles and it is entirely statistical in nature. The point here is that "randomness" does not necessarily lead to chaos.

Many of the "laws" of physics are thought to arise through symmetries in nature (in a way that I've explained here). This combination of geometry and statistics explains really rather a lot, but it doesn't explain everything. It doesn't explain identical electrons or physical constants. Yet we have no way of knowing whether these are inevitable because of the underlying geometry of nature. If the worst comes to the worst, we can resort to anthropic reasoning and posit the multiverse where all things are possible and we just happen to exist in one where our particular geometry causes the laws of nature to take the form that they do. The point is, it doesn't follow from the laws of nature that a supernatural intelligence must exist. You cannot conclude that there isn't a natural explanation.

In summary, TAG is simply an assertion. Its conclusion does not follow from its premise. In specific instances, morality is better explained by evolution, the laws of logic are really empirical conclusions and the order of the universe has alternative explanations that do not require a god of any kind.

[Edit - there's a truly superb critique of TAG available on the Freethought & Rationalism Archive]