Tuesday, August 23, 2005

BS-95 Hunt

... When one person speaks and is encouraged to develop his or her ideas, then it is we, the audience, who provide the challenge. We provide the democracy. In each of our hearts and minds, we absorb, judge and come to our own conclusions. The dialectic is, thankfully, not between a group of equally ignorant people thrashing out a series of arbitrary subjects about which they know little and care less. It is between an informed individual who, we hope, has thought long and hard about their own area of specialisation, and an audience which is ready honestly to assess what the speaker has to say. Democracy, like everything else, thrives on preparation. (David Hare, in a Saturday Age syndication of an article originally published in The Grauniad (link via Nicholas Gruen at Troppo Armadillo).

I didn't sleep Sunday night. As a result, I took an inordinate amount of interest in this article in Monday's Age. It's an extract from Peter Costello's speech to the Australian-American Leadership Dialogue Forum (whatever that is) on Saturday. Looking it over, I thought it might be worth running a more or less complete BS-95 check over it, to see how the Treasurer rates as a bullshit artist.

This is what happens if you're the sort of person who wanders the streets picking out house numbers at random and then testing whether they're prime numbers. That's a lot easier than it sounds, incidentally - all you do is take an approximate square root and then look for a prime factor below that. As an example, my street number has no prime factors below 17 so it passes the test. If you're looking to narrow it down a little more, you might get a result with the help of this additional bit of information: if you knock off 100 the result isn't prime, having two factors below 10. Actually, that might just nail it. There's one extra hint I could give you, but that would hand you the result on a plate.

So I spent all of Monday in a crazy, insomniac headspace where it was absolutely the most important thing ever to check Costello's speech in detail and see how it rated on my home-grown bullshit quantifier. When I got down to it, the start wasn't all that promising:

Last Monday was the 60th anniversary of the victory in the Pacific. Australia and the United States went into World War II separately but came out of it together: as allies, as friends.

By February 1942 the continental mainland of Australia was under direct air attack. Australia faced its gravest security threat. But beginning with the naval battles in the Coral Sea, with the Australian land defence of Port Moresby and the island campaign under Douglas MacArthur, the war began to turn. It ended in circumstances that are well known. [my emphasis]

All up, there are seven assertions in those two paragraphs: dedicated nitpicking only produced one bullshit point. That was for the sentence in bold. According to this source, which I'm inclined to consider reliable, February 1942 is when Japanese air attacks on the Australian mainland actually started (with the bombing of Darwin on the 19th). In the light of Costello's later remarks on the subject of historical education, I considered it fair that the Treasurer should score a bullshit point for this pointless bit of lily-gilding. That little gloss over the exact circumstances which ended the campaign in the Pacific almost scored a bullshit point too, but on balance, I decided against it. It's a notable omission, but that's all.

The score really started to climb in this section of the speech:

It is common in this country, like so many others, to come across anti-American sentiment.

It is always there, but it rises at times of Australia's military engagement in coalition with the US. Most recently Australia's engagement in Iraq has raised these sentiments.

Critics commonly allege that Australia is only engaged in these theatres at the urging of, or in some supine gesture towards, the US. "After all," one senior school student aggressively asked me at a local school, "what have the Americans ever done for us?"

... I began my answer with the events of 1941. There was no flicker of recognition. It was clear to me that whatever the educational achievements of this school, the teaching of history was not among them. [emphasis added, one rhetorical question deleted]

By the end of this passage, the figures were up to 17 assertions with 9 of them bullshit of one kind or another. The anecdote about the senior school student at the local school scored as follows: accepting that the event described occurred, we have two facts, stated in the first two sentences. Next we have an opinion, passed off as fact and, finally, a daring leap to a conclusion about the school's educational performance - four assertions, two of them bullshit.

When I reached the end of the article I had the following figures. The Treasurer had made a total of 51 assertions of one kind or another and I was satisfied that 31 of those were bullshit. I'd also reached an assessment of the speech that's completely at odds with Tony Parkinson's in today's Age, but at the moment I'm nowhere near tempted to run a BS-95 check on Parkinson. With him, I prefer to use the less rigorous approach of reading the last paragraph first; if the conclusion strikes me as egregious, I immediately lose interest in how Parkinson got there.

Given the figures, how does the Treasurer stack up as a bullshit artist? On checking my revised BS-95 scoring table, I found that in an article with 50 substantive assertions, 32 of them had to be shown to be bullshit before the article could be written off (with 95% confidence) as total bullshit. The Treasurer's speech doesn't meet this criterion so in this case we'll have to retain the null hypothesis, a the statisticians say, and conclude that there's no more bullshit in the speech than you'd find if it were put together completely at random.

Acknowledgements: I am grateful to Comrade Zeppo Bakunin for his very precise formulation of the null hyothesis.

No comments: