Monday, May 28, 2007

Birds of a different feather

"ALL swans are white. This bird is black. Therefore..." But it IS a swan. Cygnus atratus,
if you'll pardon the language.

Until some years ago, the only places I had seen black swans were on the postage stamps once issued by Western Australia, and the labels of a brand of lager (whatever happened to it?), also hailing from that part of the world,

Then I was pleased to see they had introduced some in St.James' Park in London. Apparently Regents Park has some too. But what was a really nice surprise was to come across this one in the pictures, who had evidently come up from the park along the Grand Union Canal, introduced itself to the locals in Alperton, and was hanging out with the gang (or should we say swanning about?) on the stretch between the Plough and Pleasure Boat pubs and Sainsbury's. Not unlike our own perambulations at the time.
Apparently the black swans from Australia are a different species from the white European varieties, though I've noticed that cygnets are sort of grey, so could both have evolved from a common ancestor, going in different directions in northern and southern hemispheres, like the way water is supposed to go down the plughole in different ways north and south of the Equator? (Before you explode in derision, let me be first to admit my ignorance of natural history. Blame our education system.) But the difference was not apparently enough to prevent the antipodean guest from knowing it was a swan, nor the other swans (whom I'd previously thought stand-offish) from apparently accepting it.
However, the reason I've brought you on safari in south Wembley is not to discourse on the social life of our feathered friends but to muse on some philosophical points. There's been a good deal of talk recently about the ideas in a book called The Black Swan: The Impact of the Highly Improbable by Nassim Nicholas Taleb. The author is not an ornithologist either; his interest is in Cygnus atratus as an example of the unexpected, turning up to confound the generalisers. His previous work was Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets.
I've not read either book yet, but I have seen some impressive reviews.
According to one admirer The Black Swan " is the most important book in social science since Adam Smith's The Wealth of Nations. Nassim Taleb's book also happens to be the most significant contribution to the science and philosophy of uncertainty since Andrey Kolmogorov axiomitized probability theory (which along with Bayes, gave us the solid foundation necessary to think clearly about chance) and made progress (with contributions by Chaitin and Solomonoff) towards a more mathematically precise definition of randomness. In terms of epistemology, reading The Black Swan gave me a sense of intellectual kinship that I have not felt since reading Isaiah Berlin's "The Hedgehog and the Fox."

Now, I'm no mathematician, but I have a smattering of logic. The statement we began with was a categorical syllogism, it begins with a generalisation, derived from experience - that all swans are white - which was disproved when people from the north discovered a black bird in Australia, and decided it was a swan. But was this unpredictable, even improbable? Not to native Australians, who might have been surprised to encounter the white version. Saying "All swans are white" was not like saying 2+2=4. The whiteness of swans was not a priori deductible from the nature of swans (or what was known about it) but a generalisation from empirical observation. This does not devalue it as knowledge but limits its value for prediction.
If you were mapping a coastline it would be important to do it accurately, for navigation, but would not tell you reliably what the parts as yet undiscovered would be like.
'Nassim Taleb mentions the analogy to maps in relation to his disdain for what he calls "Platonicity." Taleb defines Platonicity (named after the philosopher Plato) as "our tendency to mistake the map for the territory, to focus on pure and well-defined 'forms,' whether objects, like triangles, or social notions .... we privilege them over other less elegant objects, those with messier and less tractable structures ..." The cardinal sin of Platonicity is that it "makes us think that we understand more than we actually do."'
Fair warning. But consider other cases. If tossing a coin it had landed heads ninety-nine times, you might still think there was a 50:50 chance of it landing tails next time. Or you might consider the possibility that it was a biased coin. In backing horses you look at their form, even though as we know it is impossible as yet even when laying odds to predict the outcome of a race with any certainty. If we knew enough to do that it would be the end of horse racing, or at least of betting on it. However, Nassim Taleb and his readers are concerned with the predictability or otherwise of still more serious issues.
Nassim Taleb, a scientist-philosopher-businessman, makes the case that 9/11 was a black swan. A black swan is a unpredictable event that defies prediction. An outlier. I agree. He expands:
A vicious black swan has an additional elusive property: its very unexpectedness helps create the conditions for it to occur. Had a terrorist attack been a conceivable risk on Sept. 10, 2001, it would likely not have happened.
In their analysis of black swans (which by definition will likely never be repeated), human beings engage in what is called hindsight bias. This is the tendency to believe that the event was predictable based on knowledge gained after the event occurred. In effect, people unknowingly substitute current knowledge of outcomes into the gaps of knowledge that were present when building earlier expectations of potential events'.
Notice how easily the writer has slipped into discussing "black swans" (and even "a vicious black swan"!) as meaning unpredictable events, even though the book has not been out that long. I wonder if anyone can predict how widespread this usage will become, and whether it will be taken up by media hacks without even a clue what it means before, like other trendy expressions it fades into obscurity? I'm pleased to see one commentator is skeptical:
I'm all for an attack on Platonism and the complacency of scholastic generalisations (often based on far less certain knowledge than people thought they had about swans) assuming we know everything about what is, and what's more, what is going to be. But I'm also suspicious of political leaders and other so-called "experts" who always claim something was "unpredictable" when they mean, they failed to predict it. I also remember how eagerly some of our gurus on the left grasped at "uncertainty" and "chaos theory" when they wanted to acquire mystic status in place of any pretence at political strategy.
When we say that black swans were "unpredictable", that was not a matter of some unknowable quirk in nature, but of our inadequate knowledge. It is perhaps a telling sign of our times that whereas primitive human beings devised religion to cope with the mysteries of nature and natural catastrophes, modern human beings fear what appears the unpredictable results of human social, economic and political processes. Even when such events as major floods devastating Asian villages or an American city are predicted, it is society and government that hamstring efforts to cope with them.

Was 9/11 predicted? So far as I am aware, not. Was it unpredictable? Well, there were warning signs. On February 26, 1993 at 12:17 PM, a man called Ramzi Yousef took a truck with 1,500 pounds of explosives into the underground garage of the World Trade Centre's North Tower. The explosion tore a 100 foot hole through 5 sub levels of concrete leaving six people dead and 50,000 other workers and visitors gasping for air in the shafts of the 110 story towers. Many people inside the North Tower were forced to walk down darkened stairwells which contained no emergency lighting, some taking two hours or more to reach safety. As the Port Authority was a bi-state agency, the towers were exempt from New York City building codes. Subsequent to the bombing the Port Authority installed emergency lighting in the stairwells. In 1997-8 six Islamicists were convicted and sentenced to life in prison for their part in the bombing. According to a judge, the conspirators' aim had been to cause the north tower to crashing into the south tower, toppling both landmarks.
So it might have been reasonable to guess that the World Trade Centre was going to be a target again, and what political direction the attack might come from. Had the US intelligence services known enough about what was being planned ... well, we cannot say. But the attack was surely "unpredictable" only in the sense that those who might have prevented it did not know enough (for all their expenditure on surveillance around the globe, which makes one wonder whether they were too busy looking at the rest of us) .
Nassim Taleb is Lebanese, and looking at the tangle of factions and outside interests preying on his native country I can understand why, quite apart from what experiences he might have had with stock markets, he is inclined to stress the unknown, even irrational. All the same... Does anyone know, let alone understand, what was behind the armed clash that suddenly brought fresh deaths and misery to Nahr el-Bared camp, making its residents once again refugees? Maybe even the gunmen of Fatah al-Islam did not know.
But that does not mean nobody did.
Not knowing, we should not fall for glib and fabricated conspiracy theories. But admitting we don't know should not make us think that nobody does, or accept that some things are without reasons and unknowable. The swan was there all the time before the scholars knew about it.

Labels: , ,


At 7:34 AM, Anonymous Black Swan Rupert said...

I realize this is a philosophy site as opposed to a bird site, but I couldn't resist -- here in California, not only did we have a black swan suddenly appear, but the freshwater bird appeared in & adapted to saltwater! After 15 years of roaming about, he was adored like a community pet. When he was run over by a boat, there was a large funeral for the bird, reported in the LA Times. (How's that for posing a philosophical question?!). We have photos of Rupert the black swan, and made a movie about him, if you want to see.

At 6:50 AM, Blogger Charlie Pottins said...

Thanks. I am really glad to get out of the stuffy world of politics and theory now and then for some outdoor life. It is refreshing, and you have added to our knowledge.
Did Rupert the black swan really make it to California on his own, and not from some introduced stock?
I did read that some black swans had flown from Australia to settle in New Zealand, but a cross-Pacific Ocean migration would be really something!
I shall take a look at your site.

At 8:55 PM, Anonymous Phoebe said...

Local lore says Rupert the swan was a pet that got loose or was let loose.

But you can imagine the shock of people enjoying the beach, kayaking, or swimming, only to find themselves eye-to-eye with a large black bird, unlike any they'd ever seen in the harbor!

He kept showing up in various spots within a 5-mile radius, demanding drinks from hoses. Sometimes he would chase people.

At 4:23 PM, Blogger Chris Martin said...

Charlie, not a bad posting, but I really think you should read the book before you comment on it. When you do, you'll see that much of what you criticise, he does too.
However, [in my view]the claim that Taleb is as significant an author as Adam Smith is far fetched. His reasoning is interesting and at times incisive, but in the end, I felt that he was just as biased as those he aims at. Worth a read never-the-less.

Chris Martin

At 9:07 AM, Anonymous Rob Edwards said...

Chris makes a good point concerning NNT and bias except to say that he does challenge his bias throughout and finds no other conclusion; exactly as a skeptical empiricist ought. And let's face it anyone in the business of strategy or prediction needs to have the models re-defined. I am reminded of the endless conflict between the formal and informal schools which seemed to occupy most of the arguments in which I found myself some time ago. I think when considering the predictability of the Sub-Prime catastrophe one could say it was "predictable" in hindsight but ultimately did not register as significant in the risk models used. Platonicity at work.


Post a Comment

<< Home