Affirming Taleb-Style Skepticism in Scientific American


(Michael Shermer's , author of "Skeptic" column in Scientific American)

Although Nassim Taleb would be the first to point out that he did not invent the principles of skeptical empiricism described in his Black Swan book, it is interesting to see some (but not all) of those ideas affirmed in the Scientific American (July 2009 issue). It makes me think that those "ah hah!" moments I had when reading the book were legitimate.

One affirmation occurs in Michael Shermer's "Skeptic" column. Arguing that science is the best avenue to the truth, he describes the "null hypothesis, which assumes that the claim under investigation is not true until proven otherwise." Basically this means that in science (unlike law) a statement is presumed to be wrong (guilty) until it proven right - and even then you can never really be sure.

This is akin to Taleb's notion of negative empiricism (as taught by Karl Popper). Quoting my Black Swan Book Report... although you can never be absolutely sure that your theory or proposition is true, just one negative result can prove it wrong. That's where truth comes from. Evidence is asymmetrical. One piece of negative evidence can offset a lot of positive evidence. We should regard all theories as provisional. Don't look for what will prove you right but what will prove you wrong; it's a faster more certain process and you'll learn more.

The other affirmation is in "The Science of Bubbles and Busts" article by Gary Stix. He describes built-in biases which lead humans to make financial mistakes that in aggregate result in economic bubbles and busts. He says that we have...
  • A confirmation bias which prompts us look for evidence that confirms ideas we already have.
  • A herding bias which leads us to agree with everybody else.
  • An availability bias which leads us evaluate information based on context rather than merit.
  • A bias toward overrating our own abilities.
  • A bias to engage in heuristic-based intuition when we should be trying to think rationally.
These are the same biases and tendencies cited by Taleb as reasons for our inability to properly understand black swans. Taleb also notes our tendency to prefer a good narrative to the facts.

I think Taleb would disagree with the premise of the article that once these biases are understood that tools can be devised to predict economic bubbles. He would probably say that insofar as these events are true black swans (not gray swans) they are inherently unpredictable.

Ain't That Some Shit, Nassim Taleb? (Video)

Another attempt at this video...

Phase Space in Extremistan?

A series of accidents ...
  • A recent Black Swan presentation where I tried to make a connection between Nassim Taleb's Mediocristan/Extremistan idea and several other attempts to categorize the known/unknowable - Kant's phenomena/noumena thing, Castaneda's tonal/nagual concept, etc.
  • A slide in that presentation noting that "gray swans" are products of non-linear chaos.
  • A dusty clipping from the June 1994 issue of The Economist offering the best explanation I've ever read of the term "phase space".
...leading me to wonder if phase space is part of Extremistan.

Aside 1 - Is synergy a black swan? Maybe.

Aside 2 - What is phase space? It is a mathematical view of action in the "real" world. Consider the example of the pendulum described in the Economist article. The pendulum swings back and forth; its position and velocity constantly changing. In the real world the action of the pendulum is an arc. In phase space, the action is represented by a two-dimensional chart showing changes in position and velocity. An ideal pendulum with no friction would appear as a circle in phase space.

Non-linear chaos refers to fluctuating phenomena that appear disorderly in the real world but display order in phase space. As the article notes, "populations of animals, bubbling pots on a stove or measles epidemics" are all examples of non-linear chaos.

Patterns traversed in phase space are sometimes called "strange attractors" (one assumes because the behavior of systems is strangely attracted to these patterns). The picture below is a two dimensional view of the three-dimensional Lorenz attractor.

It describes the chaotic behavior of weather patterns - leading to the so-called butterfly effect meaning (more or less) that a butterfly flapping its wings in Africa can cause a hurricane in Florida. There is order here - constrained by the strange attractor. But within that order is infinite disorder.

Strange.

So, returning to my original question - Is phase space in Extremistan?

Extremistan is supposed to be the real reality underlying the man-made view of Mediocristan.

So, where does phase space belong?

Which is more real - the confused real world, or phase space which reveals the underlying order in the real world?

Is phase space real?

Does one peer through/past the behavior of the world of appearance to the underlying truth of phase space?

I suppose Taleb might say it doesn't matter. In the real world, the swans are definitely black. In phase space, they are gray. In the real world, you don't know the swans are coming. In phase space, you might know they are coming but you can't know when or where.