Since Nassim Nicholas Taleb published his bestseller The Black Swan: the impact of the highly improbable back in 2007, “Black Swans” and “Black Swans events” have become part of everyday language. They are used as a catchphrase to mean two different things. First, as was the case recently in the Brookings interesting interactive “briefing book” Big Bets and Black Swans: Foreign Policy Challenges for President Obama’s Second Term, “black swans” represent high impact, low probability events, what is also known as wild cards.[i] Second, “black swans” refer to events that could absolutely not be predicted, as, for example for the Economist in ”The prediction games: Our winners and losers from last year’s edition”. Unfortunately, in this case, the label “black swans” excuses foresight errors. It tends to stop explanations and evaluation. Similarly, some will make statements along the line of “oh, but there is no point to do any foresight (or futures work or forecast), did you not read Taleb’s Black Swan? One cannot predict or foresee anything.”
This is a rather crucial assertion for us and it needs to be investigated.
What are exactly those Black Swans about which Taleb writes: they cannot be absolutely unpredictable and low probability events at the same time? Thus, what did Taleb really describe? After having read this book, should we just resign, abandon any work related to anticipation, and, instead, do something else? Or is there more to Taleb’s argument than that? Can we use this book to improve our foresight methodologies, consider deep uncertainty, yet without giving up?
This first post will review the whole book and see if it really points to the absurdity of foresight. The next one will outline some of Taleb’s points that could be more than useful to foresight, forecast, warning, and more generally all anticipation methodologies.
Black and Grey Swans
Taleb is interested in understanding better uncertainty and randomness, notably those “Black Swan events” defined as unpredictable (outliers), with an extreme impact and which are, after the fact, revised as explainable and predictable.
The existence of black swans is logical and even may be seen as obvious, but it does not imply that foresight is doomed, only that it cannot be 100% certain (for the anecdote, the ancient Chinese divination method, the Yi Ching, which used the tossing of yarrow stalks, always removed one stalk before the cast, to account for this unpredictability).
In a nutshell, The Black Swan denounces the problem and risks of induction, building upon David Hume and Karl Popper. Extremely briefly, an inductive reasoning runs as follows: all the swans observed are white, thus all swans are white… which is proven wrong when one black swan is spotted. Hence the danger of this reasoning, if it is not done cautiously. Incidentally, this is quite crucial for us as so many of our analyses come from collected evidences, and we should always keep the danger of induction in mind, but more on that with the next post.
The Black Swan attacks quantitative predictions made with a Gaussian distribution (or normal distribution or bell curve) when applied to a world that is increasingly complex, notably because of the evolution of social interactions, and includes unexpected events. Such a world should rather be understood through fractals, that would then allow us to anticipate those events that Taleb names “Mandelbrotian Gray Swans.” Here I trust the author because of the consistency and logic of his argument, while I don’t have the mathematical knowledge necessary to go more in-depth into fractals.
The Economist usage is thus correct, whilst Brookings foreign policy experts’ actually refer to “Gray Swans.”
All doomed: forecasting, prediction, foresight, and also meaning
Taleb’s attack is directed at statistics and quantitative methodologies (correlations, trends, statistical forecasts, etc.), NOT qualitative ones, thus concerns only one part of “foresight.” Nonetheless, we are not safe yet, as Taleb also denounces a “narrative fallacy” and underlines various cognitive biases, which make us believe we can understand history or the present or try to anticipate the future, when, according to him, such endeavours are near impossible as everything is “ruled” by randomness or luck. For Taleb, most historical events with large impact are Black Swans, thus could not be predicted, not even with a low probability of success.
This brief summary would let us believe that, indeed, prediction, forecast, foresight, and even worse understanding and finally meaning are impossible human endeavours. Those stem from our human needs and cognitive make up. To believe in their possibility emerges from our lack of introspection and reflection.
This book is thus not only about anticipating the future, and more broadly the philosophy of science and epistemology but also about meaning. The specific narrative Taleb uses – intertwining personal experience and stories as examples, with logic, (sceptical) empiricism and references to philosophers, novelists and scientists – is part and parcel of the demonstration. Don’t even hope to skim through The Black Swan to understand it. It is here in the literary part of the book, in its very construction, that we find the key to Taleb’s work and to answering our question “is foresight doomed?”
The choice of hope over despair
The demonstration made by the author is extremely well done, very consistent, save for one single paragraph. Our very smart author could not be unaware that someone, somewhere, would tell him: “Hey, wait a minute, if you say that nothing can be predicted and understood, that something random and unexpected (the famous black swan) may always happen, then it may also happen that your argument will be proven wrong by something unexpected or something you do not know.”
Taleb had thought about that, but used something quite akin to a sophist argument (p.192-193, hardback edition): He started by explaining the “Black Swan asymmetry,” which allows you only to assert, for example that “all swans are not white” (nothing more but nothing less), stops his reasoning and then quotes Popper, who, asked about “a possible falsification of falsification,” answered that his questioner was an idiot. Conclusion, the reader does not dare anymore to question Taleb (or Popper or anyone for that matter) for fear of being an idiot. Well done, but unfair, and somehow a shame to use such a device because some new insight could have emerged. Actually, I don’t really care about being named an idiot, but as a reader, I would very much prefer, if I don’t understand, to be explained why, and if I do, to see the author pushing his argument further.
This does not challenges most of the points made in the book, but allows questioning its overall conclusion. Thus, foresight might still be possible, should the impossibility to understand the world, the “narrative fallacy” be questionable too.
If human cognitive biases are numerous, witness the amount of research and findings of cognitive sciences, and as wonderfully synthesised by Heuer, if indeed our knowledge and understanding of the world through historical, political and more broadly social science is most of the time imperfect and still minimal, Taleb here, despite his huge erudition, seems himself to be prey – as all human beings – to the problem of induction, to generalization and to seeking confirmation as validation. Only one properly foreseen event should be sufficient to make foresight and understanding not impossible.
I’ll give here an example (because it is simple, but we could quote many others, as Jay Ulfelder correct 2012 forecast of coups in Mali and Guinea-Bissau or all the correct Economist 2012 predictions etc.). In 1919, Max Weber explained and underlined the main characteristics of the State, including the crucial importance of the legitimate monopoly of violence.[ii] Armed with this knowledge and common sense, and referring to the second Iraq War, it comes that if one destroys the State, it is very highly likely that civil war will occur. It could thus have been (very easily) anticipated that if one destroyed the Iraqi State (through the Ba’ath party), then civil war would more than probably occur, something that was carefully avoided in Germany after WWII despite the “denazification goal.”[iii] This is NOT a reinterpretation of facts after the events, but could have been made beforehand. Why this was not done or rather not heard is another story, let’s not confuse issues. We have here at least one instance of a potentially accurate anticipation (actually many with the other examples). Thus foresight is not impossible.
Of course, something completely unexpected may always happen, forbidding certainty. Of course, our imperfect knowledge and understanding will only at best help us outlining possible futures. Yet, in the meantime, do we have the right to disregard an understanding that could save lives and protect our security (individual, national, and global)? Using our previous example, the knowledge on State and its processes could definitely be helpful in assessing needs and potential impacts before to decide about austerity policies and IMF backed structural adjustment policies as was done worldwide with disastrous consequences, before to enter post-conflict countries and determine strategies. It would be crucial in evaluating what is happening in countries such as those that made the Arab (winter)-spring, and the evolution of situations where the international community intervenes, such as Afghanistan.[iv] Do we have the right not to work towards improving our understanding? We have to live with uncertainty but isn’t it our job? We have to live with incomplete mastery and imperfection, but isn’t it what it is to be just human?
What seems to me to lie at the core of the book and to lead to Taleb’s overall conclusion, is a very deep despair and a revolt when confronted to uncertainty, to the unexpected, to unfairness, to death and war and suffering (hence the importance of the autobiographical and literary part of the book). For the author, the world has become meaningless, and he finds hope only in the miracle of being alive and of building upon the opportunities of “positive Black Swans.”
Someone else could very well use similar valid points (the problem of induction, the risk of anachronistic projections, etc.), but also see and emphasize connections (even hidden ones, or different ones, for example the probably not yet completely understood fractals, where one can also marvel at the self-similarity property) rather than disconnections, meaning rather than meaninglessness and reach a different conclusion. From a physicist such as Omnes, you would get a completely different outlook on life. This is often the same with astrophysicists, who truly wonder (in both meanings of the word) about the world.
If we choose not to despair but, on the contrary, to wonder and hope and work hard, we may and even must continue our foresight and warning endeavour, assuming that a few rules are respected, those very rules that Taleb underlines (common sense, humility and using the right causality or the right tool for the right problem, etc.), as we shall see next.
[i]“A wild card is a future development or event with a relatively low probability of occurrence but a likely high impact on the conduct of business,” BIPE Conseil / Copenhagen Institute for Futures Studies / Institute for the Future: Wild Cards: A Multinational Perspective, (Institute for the Future, 1992); then popularised with John L. Petersen, Out of the Blue, Wild Cards and Other Big Surprises, (The Arlington Institute, 1997, 2nd ed. Lanham: Madison Books, 1999).
Ertman, T., (1997), Birth of the Leviathan: Building States and Regimes in Medieval and Early Modern Europe, (Cambridge, Cambridge University Press).
Mann, M., (1986), The sources of social power. 1, A history of power from the beginning to A.D. 1760, (Cambridge: Cambridge University Press).
Tilly, C., (1990), Coercion, capital, and European states, AD 990-1990, (Oxford: Blackwell).
Weber, M. (1919) Le savant et le politique, (Paris : 10/18, 1963) Paru originalement en allemand «Wissenschaft als Beruf » & « Politik als Beruf » 1919.
[iii] Nina Serafino, Curt Tarnoff, and Dick K. Nanto, (Foreign Affairs, Defense, and Trade Division), U.S. Occupation Assistance: Iraq, Germany and Japan Compared, CRS Report for Congress, March 23, 2006,http://www.fas.org/sgp/crs/natsec/RL33331.pdf
[iv] For example, one may wonder, considering that the modern state is essentially a territorial state, if the necessary resources to build or rebuild a state should not be proportionate to territory. If this is correct and if the amount spent on Germany is a good indication, although approximate, then the amount needed for Afghanistan might have been closer to 76 billion USD rather than to the 14 billion pledged. The way to implement a reconstruction, as well as the time necessary to succeed, would most probably have to be revisited. Hélène Lavoix « La construction de la paix et l’estimation des besoins et de l’impact, » Policy paper, Centre d’Etudes et de Recherches Internationales (CERI – Sciences Po) programme on « Peace and Human Security » (CERI-CPHS), April 2008.
Featured image: Comet Bradfield from Cactus Flats in NE Colorado. Just before sunrise. 5/20/2004 10:40:20 AM. By TheStarmon (Own work) [CC-BY-3.0 (http://creativecommons.org/licenses/by/3.0)], via Wikimedia Commons