15 June, 2007Issue 6.3PhilosophyScience

Email This Article Print This Article

All in a Flap: Beware Unknown Unknowns

Will Davies

Nassim Nicholas Taleb
The Black Swan: The Impact of the Highly Improbable
Penguin Books, 2007
400 pages
ISBN 978-0713999952

“To know that we know what we know, and that we do not know what we do not know, that is true knowledge.” Like many examples of literary refinement, Thoreau’s remark strikes us with the familiarity of a platitude and the certainty of commonsense. In reality, however, we seem consistently to underrate the extent and significance of our ignorance: true knowledge, in Thoreau’s sense, is truly elusive knowledge. This shortcoming would be inconsequential, if only it were true that what we don’t know won’t hurt us. However, arguably some of the most significant events of our times have fallen beyond our epistemic ken. In 1929, the Weimar Republic had enjoyed several years of relative political stability and economic growth, aided in no small part by loans from the United States under the Dawes and Young Plans. This semblance of security, however, was short-lived: the Wall Street Crash, an unknown unknown, and the resultant recall of American loans sparked a German Depression that allowed extremists-in-waiting to capitalise on the unrest of a nation. Unfortunately, we all know what happened next. The model – an improbable, unpredictable, unknowable event with colossal consequences – retains a perfectly contemporary significance. The recent shock collapse of the American sub-prime mortgage market has caused sizeable global jitters, which could ultimately cost trillions of dollars in banking revenues and thousands of jobs. What we don’t know can and will hurt us; the question is, how badly?

Nassim Taleb is fairly seen as a present-day guardian of Thoreau’s epistemic nonpareil, bitterly conscious that society is inherently and precariously unaware of its own ignorance. Taleb’s book, The Black Swan, is in essence a manifesto on the mysterious and obscure category of momentous but unpredictable events, and the need to heed their importance. A Black Swan, as Taleb employs the concept, is a highly improbable, highly consequential event; unlike winning the lottery, however, Black Swans are events of which we lack even the faintest conception before their occurrence. Black Swans pervade all aspects of life, and range in form from market crashes to the portentous rise of the National Socialist party. The spread of the internet, 9/11, the Virginia Tech massacre, the invention of the printing press, the Indian Ocean tsunami of 2004, Microsoft, the rise of Islamic extremism and the ensconcing of a free market system in Europe during the Medieval period are all Black Swans. All have, to some extent, shaped the society in which we live today, but none seems to have occurred in accord with a prior discernible pattern. For Taleb, this suggests only one conclusion: history abhors a narrative, at least as far as the major plays are concerned.

Why a Black Swan? Until 1697, Europeans had strong inductive grounds for the claim that all swans are white: every swan that had been seen by them had been white. However, Willem de Vlamingh set that theory on its head while exploring Australia, eponymically naming the Swan River in Western Australia in honour of his piceous discovery. Karl Popper made the black swan famous as an example of the problem of induction, namely, that one event, no matter how improbable or unforeseeable, can upset any scientific generalisation. On Popper’s model, confirming instances are well and good, but perhaps meaningless in the face of one good counterexample. Taleb first developed an appreciation of the Black Swan in his former career as a trader. A firsthand witness of the shattering US market crash of 1987, Taleb became convinced that his colleagues were lulled into overplaying their knowledge, and downgrading their ignorance; “state of the art” predictive techniques had convinced most traders that the market was robust, yet Taleb discerned an entrenched and institutional blindness to the looming crisis. While others despaired in the aftermath of the crash, Taleb oozed triumphalism; he smugly recalls the experience as of a “deafening trumpet signalling to me that I was right“. For Taleb, it seems, the crash was nothing less than Popper’s revenge, a thunderous comeuppance for sins of over-generalisation. Thus convinced, Taleb has been thinking about Black Swans, in one form or another, ever since.

Taleb came to see that the Black Swan phenomenon had a direct bearing on the predictive techniques prevalent in what he calls “mainstream economics”, as employed in major financial institutions. Mainstream models, Taleb claims, are typically based on “Gaussian”, or normal distributions, where expectations are geared towards the centre, or fattest point of the familiar bell-shaped curve. Taleb argues that the presence of Black Swans in a population wholly nullifies the Gaussian model of reality. While forecasters in 1987 were giving financial advice and trading based on average expectations at the centre of the curve, for instance, the crunch event ultimately occurred at the extreme end of the scale. Taleb’s quarry here is not new, of course. The inadequacy of the Gaussian model was noted by Taleb’s idol, Benoit Mandelbrot, in a series of journal articles in the 1960s. However, for Taleb, the importance of this moral extends far beyond finance and economics. In any Black Swan-susceptible domain, predictions based on norms are liable to come unstuck. Consider projections of home computer usage, prior to the development of the Black Swan single-chip microprocessor. In the mid-seventies, predicted usage was so low that even experts at the Digital Equipment Corporation and Hewlett Packard saw little point in pursuing the idea of a personal computer. By 1980, however, sales had reached a million units and by 2000, around 140 million and rising. The PC now dominates many aspects of work, entertainment and information access, but even the best industry models were blind to its potential impact. At the socio-political level, could anyone have predicted the multifarious consequences of 9/11, both in respect of the immediate physical effects, such as the collapse of the Twin Towers, and the subsequent shifts in foreign policy? Perhaps defence specialists, or technology consultants, aided by increasingly sophisticated predictive models, will get better at forecasting the next 9/11 or microprocessor. However, Taleb remains fundamentally sceptical: the deep complexity of social phenomena, he claims, should “make us moresuspicious of scientific claims of precise models of reality”. The most extreme Black Swans will always, Taleb believes, slip through the theoretical net, eluding even our most sophisticated attempts to predict them.

It would be quite wrong, however, to think that our systematic ignorance of the Black Swan is a wholly theoretical deficiency. It is pertinent here to reflect on the words of philosopher Henri Bergson, who proclaimed, “the eye sees only what the mind is prepared to comprehend”. Taleb marshals a host of highly interesting findings from psychology and cognitive science that suggest the mind is congenitally primed to shroud the Black Swan. More or less all inhabitants and visitors to Indonesia in December 2004 were taken completely by surprise by the disastrous inundation following an offshore earthquake. Students in Blacksburg, Virginia, assumed that 17 April 2007 would be just like any other normal day. None of this will seem particularly untoward: nobody expects to predict the unpredictable; cognitive economy, for one thing, dictates that we simply cannot entertain every open possibility pertinent to our lives and actions. Rather, like genetically-primed gamblers, our brains induce us to play the odds, ignoring the small probability of accident or catastrophe, and to derive patterns or regularities from the jumbled chaos of experience on which to base our decisions. In this respect, our default lack of Black Swan reasoning is not apt for blame or reproach. Nonetheless, Taleb seems to consider it an intellectual duty to recognise this tendency in ourselves. Once we shun what Taleb calls our steady state of “epistemic arrogance”, we will at least start to extricate ourselves from the illusion that events are under our control and within our intellectual grasp.

All well and good, but what more can we do? If the prospects for precise models of social reality are as bleak as Taleb claims, what hope do we have at predicting and averting the next World War or 9/11? If our blindness to Black Swans is truly inherent, what option is there, other than to acknowledge our cognitive limitations and pray for the best? Taleb’s answer is augured in Thoreau’s conception of true knowledge; simply, that we ought to know that we don’t know that we don’t know. In this somewhat clumsy iteration of the Socratic adage, Taleb descries a necessary precondition for epistemic enlightenment and, resultantly, the taming of the Black Swan. Only in fully acknowledging the possibility of extreme outliers, unforeseeable catastrophes or fortunes, can we begin to insulate against and prepare for the occurrence of the unknown unknown. The possibility of a tsunami is thus met with increased investment in disaster planning, the terrorist attack with general defence-readiness and strategic response teams; improved insurances are introduced to protect citizens, particularly in volatile markets, against crises and shocks. The importance of such schemes is clearly starting to be appreciated, but the lessons are far from being fully assimilated. Absolute vigilance and overall preparedness, Taleb feels, is the only viable defensive strategy when faced with Black Swans of such immense destructive force.

Taleb’s skill in conveying the significance of the Black Swan is unfortunately coupled throughout the book with a tiresome penchant for irreverent hyperbole and bumptious conclusions. One would indeed be forgiven at times for mistaking The Black Swan for the lesser gaudy parrot: flashy, repetitive, and exceedingly irritating. One may grant the stylistic liberty, but at times Taleb’s haughty style borders on the intellectually irresponsible. He is generally far too quick to dismiss research agendas such as “mainstream economics” for notable predictive failures – the 1987 market crash being a pertinent example – without exploring the possibility that such theories are subject to more complex standards of falsification. While a simple generalisation such as “‘all swans are white” is clearly falsified by an instance of a black swan, “laws” in the social sciences, or for that matter in everyday practical reasoning, are generally subject to numerous ceteris paribus, or “all things being equal”clauses. This fact should at least give us pause for thought in considering the impact of a seeming counterexample. The Virginia Tech tragedy, for instance, might be taken as a clear counterexample to the folk psychological generalisation, “students will not take to arms against their peers”. However, like a market crash or terrorist strike, murderous rampages generally occur when some rogue element enters the equation; in this case, it would seem, a terrible pathology or delusion. One can agree that a Black Swan is contrary to the operative generalisation in a given domain, without abandoning that law or principle: when all is not equal, all bets are off. Predictive success with respect to human behaviour, whether individual or aggregate, has to be measured holistically, wherein “successful on the whole” is perhaps a more apposite criterion than “free of all counterexamples”. For Taleb, however, his obsession with Popper’s criterion of falsification is sufficient to damn entire explanatory-predictive enterprises. While Taleb’s counsel about preparedness for Black Swans is timely, his dismissal of great swathes of our intellectual and rational landscape seems an expression more of frustration and petulance than scholarly contemplation.

Taleb’s response would undoubtedly be that, in erecting theory on the foundations of ceteris paribus laws, we are irresponsibly “Platonising”, or idealising in such a way as to neglect the manifold anomalies and fractal elements of reality. Such theories, Taleb believes, are destined for failure. At first pass, the rejoinder certainly has some teeth: if Taleb has demonstrated anything, perhaps it is that there is latent danger in ignoring society’s rough edges and Black Swans, and that unwavering faith in the predictive power of our current best theories is premature, if not misguided. However, it seems too soon, and our sciences too young, for Taleb’s resoundingly defeatist mentality. If we see any value whatsoever in trying to predict the next dot.com bubble, or to obviate the rise of our generation’s Hitler or the next schoolyard slayer, surely the wisest prescription is to continue to hone, improve and test our theories and social predictive practices – not to rail against them. Perhaps Jefferson was right that ignorance is preferable to error; but in the engine room of progress sits an unsated desire to understand the seemingly unfathomable. Black Swans certainly pose an acute challenge both to our theories and our society, but if the notion of true knowledge requires one to accept their eternal intangibility, then perhaps that is an ideal worth forsaking.

Will Davies is a BPhil student in Philosophy at Balliol College, Oxford. His research lies at the intersection of Philosophy of Mind and Language.