bigideas

So, as you will no doubt know, telly writer and journalist Ariane Sherine has used her Comment Is Free blog to raise donations for an advertising campaign for atheism. One word in the ads has caused far more comment so far than the others.



The ad in full reads: “There’s probably no God. Now stop worrying and enjoy your life”. The word in question is, of course, “probably”. Sherine herself explains that the word was put there to get around a requirement from the bus companies that ads “must not offend religious people”. To me it smacks of more of a cop-out than that.

Evidence

After all, I take it Sherine and fellow-travellers Polly Toynbee and Richard Dawkins — who have both endorsed the campaign — think there actually is no God, in the same sort of way that some people think there actually is a God. But one of their key arguments against religions is that they can’t offer empirical evidence for the claims they make, and nobody can offer any evidence for the claim “There’s no God”, as the organisers know:

even though there’s no scientific evidence at all for God’s existence, it’s also impossible to prove that God doesn’t exist (or that anything doesn’t). As Richard Dawkins states in The God Delusion, saying “there’s no God” is taking a “faith” position. He writes: “Atheists do not have faith; and reason alone could not propel one to total conviction that anything definitely does not exist”.

Better not print it in big letters on the side of a bus then: good catch, that.

Nobody can, though, offer any evidence that there’s “probably” no God either. To do this we should have to show that the probability that there’s no God is some specific number greater than 0.5. Probability is about numbers. I can’t imagine how you would start doing such a calculation, or what set of data would be either useful or sufficient for it. It may be a deficit in my imagination but it just seems a very silly thing to spend your time trying to do.

The contemporary version of this, which the Atheist Bus website refers to, is Dawkins’s argument that God is “the Ultimate Boeing 747” (The God Delusion, ch 4), but this is an attempted refutation of a very specific philosophical attempt to prove God’s existence from the complexity of the universe. More rigorous attacks have been made but Dawkins’s populist one isn’t bad as far as it goes. Yet this isn’t — and can’t be — a positive claim that the probability of God’s existence could, in theory, be calculated.

So the statement in the ad, taken as a statement about probability, is just the kind of unfounded assertion that people like Sherine are complaining about.

Evidentiality

I can, though, say something like “I probably won’t go to the pub tonight” without committing a maths-crime. By “probably” here I mean, roughly, “this statement is something I believe somewhat”. This is what linguists call evidentiality: “probably” modifies a statement in a way that indicates how much the speaker believes it to be true.

This is a different “probably” and it is — excuse me — probably what the authors of the ad were thinking of. They would — ahem — perhaps argue that they don’t mean “probably” in a scientific, mathematical sense, any more than I have to do a lot of calculations before I say “I probably won’t go to the pub”.

Yet there’s no obvious speaker in the case of a bus advert: it simply can’t mean “That God doesn’t exist is something I believe fairly strongly. Now stop worrying and enjoy your life”. The strength of belief of the imagined speaker of the bus advert has no bearing on my own ability to “stop worrying”.

This very widespread usage of “probably”, if I’m reading the ad right, serves to conceal its nonsensical nature when read as a statement about mathematical probability. That’s dishonest. Well, some people will say, quelle surprise: it’s advertising.

[Aside: Prescriptivists sometimes say that sentence adverbs (as in "Hopefully I'll go to the pub", "Probably God doesn't exist" etc) are "incorrect", about which they're mistaken. Interestingly Sherine is a bit of a prescriptivist about language as well as belief.]

[General Note: This is our first Big I blog post in a long, long while. As some of you know we've been awfully busy doing various other things and we decided to put the blog on the back burner. We reserve the right to post occasionally when something jumps out at us.]

[Update, 20090115: Inevitably, this whole business is stimulating much verbiage.

Christian campaigner Stephen Green has made a complaint to the ASA, which regulates adverts in the UK. His complaint, bizarrely, is that there is "plenty of evidence" for the existence of God, so he's decided to argue over probabilities that nobody knows how to calculate. If Christian Voice were smarter and less interested in driving their own agenda a complaint might stick, although if you ask me freedom of speech should win out here even if the ad is a bit silly.

The BHA has responded, but their response is badly flawed too. Saying that "There's probably no God" isn't warranted doesn't entail that "There's probably a God" is. The middle ain't excluded.

Philosopher Stephen Law thinks an ASA ruling against the ads would mean that statements like "God loves you" in religious ads would automatically be disbarred too. But that doesn't sound right: specific, unfounded claims about probability are different in kind from expressions of faith. Perhaps you don't like the latter, but this isn't a suitable stick to beat them with.

Philosopher A C Grayling has weighed in against the word "probably" but for the opposite reason I give: he thinks it's been established that there is definitely no God. Possibly I missed a memo, it wouldn't be the first time.

]