I’ve recently finished reading this book by Nassim Nicholas Taleb. It left a favourable impression with me and so I’m glad to see that it has become a best seller. The book exposes the conceits and false confidences in our ability to understand and/or predict that may accompany what are effectively toy town theoretical renderings of complex realities. One commentator suggested that Taleb’s book is repetitive, but given how easily human beings fall into the trap of what Taleb calls ‘epistemic arrogance’, the lesson needs to repeated over and again.
There is far more right with this book than there is wrong with it, but if I were to engage it critically I would want to delve a bit deeper into the following matters:
The Gaussian Bell curve: This curve is a product of the random sequence, the latter being a limiting form of complexity and therefore of fundamental interest. However, I think that Taleb is nevertheless right here: perfect randomness (and therefore the gaussian bell curve) has a limited application in our world. Our world exists somewhere in that vast trackless domain between high order and maximum disorder.
Fractal mathematics and the scalable power law treatment of ‘negative black swans’, like disasters: Fractal scalability only seems to exist in relatively small logarithmic ranges. Taleb is of course is aware of this, but suggests that the limits of scalability are academic if we live inside the scalable zone.
Retrospective explanatory narratives: Taleb pours scorn on those who rationalize things after the event. My own feeling is that the ‘NP’ like structure of some domains of study allows outcomes, like riddles, to be correctly rationalized only after the event. But on the whole I suspect Taleb is right again: retrospective analysis, like the Gaussian bell curve, sometimes works; often, however, it may simply amount to constellation spotting.
Theoretical Narrative: There is nothing wrong in theorizing per se: it’s just that this activity must proceed with caution and humility, always self-aware on an acutely self-critical meta-science level: how and why do we know we know?
The Cosmos is not sufficiently ordered for it to be universally amenable to elementary mathematical models and yet it is not disordered enough for comprehensive statistical treatment. Mathematical probability might seem to be a way of dealing with uncertainty, but it only works if the possible cases can be enumerated. Unlike games of chance our world is open ended with a space of possible cases of staggering dimension and unknown limit. Therefore formal probability cannot be used to quantify the unknown or provide a handle on those surprising ‘one-off’ events that hit us again and again. To attempt to do so is what Taleb calls the ‘Ludic fallacy’
Taleb is brash and abrasive about ‘empty suit’ forecasters and the ‘nerd knowledge’ of academia. This will undoubtedly upset those people who know they are being targeted. It was easy for me to laugh heartily at Taleb's amusing and quite extreme jibes and anecdotes, because I'm not the target. But those whose ivory tower positions have given them such a breath taking complacency that they think it is in order to bully and insult us into their beliefs have brought Taleb the Tornado on themselves. On balance I would rather we had people around like Taleb to challenge an endemic epistemic attitude problem than live in comfortable denial. If we take Taleb’s lesson to heart we will be better emotionally adjusted for the next big shock.
Right, after several months on atheist Larry Moran’s blog getting up to speed on evolution I’ve at last decided to invade the “opposition’s” airspace! See this link, and especially this link, for results.