Saturday, October 15, 2011

THE PROBLEM OF INDUCTION: KARL POPPER AND HIS ENEMIES

The “Black Swan” refers to the fact that as far as Europeans knew, for thousands of years, hundreds of generations, all swans were white.  Then, after discovering Australia, they found black swans.  In the stock market, the "black swan" is the event no could have foreseen.  This is the problem of induction.  We can collect all the instances we want, but at some level, our knowledge must be admittedly contingent.  Objectivism asserts a positive solution to the problem of induction.

In the lecture series, “The Basic Principles of Objectivism” Nathaniel Branden said that if you disagree with someone on the relevant morality of some current political news topic, if you stay engaged, pretty soon the discussion gets down to philosophical principles at an ever more basic level until you must finally get to metaphysics.  For Ayn Rand, most disagreements centered on epistemology.  
In The Logical Leap, David Harriman provides one solution to the problem of induction:  a single example is enough if your generalization is based on reality.  (See my review on this blog.)
In the Open Society and Its Enemies, Karl Popper makes a plea for tolerance based on an admission of ignorance.  Ultimate knowledge eludes us: all claims are contingent.  Therefore, it is wrong to force your views on another person because you cannot be sure that your claims are not to be falsified at some later date.  This clearly rests on Popper’s statements in The Logic of Scientific Discovery that inductive reasoning is insecure and that falsifiability is the test of truth.  Moreover, says Popper, deductive reasoning is not much help.  The claim that all ravens are black leads to the equivalent claim that all non-ravens are non-black.  Thus, everything you see that is not black validates the claim that all ravens are black; and this is clearly ridiculous.
The solution via Ayn Rand and via David Harriman is that a proper induction is based on the essential characteristics.  In the cases of ravens and swans what was lacking was the knowledge of genetics, of the biochemistry of inheritance.  Indeed, when black swans were discovered, they were not called vorks or gabongs.  They were called swans because whiteness was not essential to swanness after all.  Objective knowledge is not an accumulation of instances.  Objective knowledge integrates reason and experience, theory and data.  It is true that lacking any understanding of physics, we could only say that the sun would come up tomorrow again in the east, but a little bit moved over, only because it always had.  The claim was inductive  only.  After Newton, this was no longer merely an inductive claim, but an objective one.  And, to the point, the problem with the precession of the perihelion of Mercury that was solved by Einstein, did not invalidate the entire body of knowledge.  Inductive claims are contingent.  Objective truths are not.
Admirable as Popper’s passion for an open society remains, the weakness of his argument is revealed by the absolutism of his opponents.  They are not so laissez faire.  They are absolutely sure of their claims.  They unequivocally assert the right to dictate for others. 


I posted this observation at the OrgTheory blog in the Comments for Occupy Everything:
Popper placed our problems in Plato’s lap. These Occupy people are acting out a desire for a coup d’etat by the Philosopher Kings. They know what is best. The idle class is not really the bourgeoisie that shows up to the office every day to trade equities. Those are only workers of a different trade. The true leisure class are the philosophers and writers who proscribe our fuels, our foods, our trades and deals, our monetary media and margins. Those hawks ... always win over the doves because those conflict managers have an agenda while the merchants of peace beg for reasoned debate. So, the Occupiers will push and shove for their goals, while the advocates of reason fail to mount a coordinated defense. In fact, one is not possible.
Perhaps the best thing is just to ignore them. No headlines. No sound bites. No tweets. A thousand years ago, Durk Pearson and Sandy Shaw penned an essay titled, “What if they gave a millennium and nobody came?” It was about securing your own personal freedom here and now.
In the news now, the Occupiers of Wall Street demand the power to decide what is best for everyone else.  Admitting that you are not certain will not carry the day.  Is your mind so open that a loose array of slogans from a flash mob can occupy it?  If not, then why not?  If your individual sovereignty is your inviolable right, what is the social extension of that? 
Despite the fact that we survive and thrive through trade and commerce based on productivity, by the nature of self-interest, the bourgeoisie is difficult to arouse.  The guardians march to their battle hymns and anthems, but we have no tuneful lyrics about open exchange or scientific discovery.  Perhaps this is as it must be.  In Systems of Survival, Jane Jacobs delineated the guardian and the trader as mutually exclusive modes.  Perhaps the bottom line only needs to be laissez nous faire. 

ADDENDUM
For more on Black Swan Theory see
This blurb from NUMB3RS on YouTube
Wikipedia on the Black Swan Theory here
The orignal work of Nassim Nicholas Taleb on Google Documents
Taleb's website here.

ALSO ON NECESSARY FACTS
Is Physics a Science?
The Man Who Loved Only Numbers
LOVERS OFFLINE
Hypatia of Alexandria

2 comments:

  1. Michael: I think the "problem of induction" is limited to the fact that there is no logical necessity that conclusions drawn from induction will be true in the way that the conclusion of a valid deductive argument follows by necessity. It's a distinction that can't be gotten around, and the Peikoff/Harriman attempt fails. It's the very inability to know when we've identified the "essential characteristics" of a subject, that imposes the limit. That's why science makes claims that are contingent, subject to the appearance of contradictory evidence; deductive argument faces no such problem. The necessity of accepting this distinction can be seen in the reversal of the fundamental, widely accepted theories of the cosmos as a result of Riess's work. We thought we understood the basics. We found we didn't. This doesn't invalidate observation or fact-gathering, it just undermines the claim that conclusions based on such observations, even deeply integrated observations tied to credible fundamental theories, don't remain vulnerable to the counter example. I think the Peikoff/Harriman effort raises a question of how much of current Objectivism, and how much of Rand's thought, owe to Aristotle's absorption of some of Plato's essentialism.

    ReplyDelete
  2. Mike and John: The "problem of induction" is the problem of justifying induction in the absence of necessary factual truths. However, if there are necessary factual truths then there is no problem, and I have argued that there are necesary factual truths in my book of that name. Hume thought that necessary truths could not be factual because they were mere relations of ideas; Logical Positivists said that they could not be factual because they were mere products of language. However, I have argued that this view were simply mistake.
    Also, John, remember that Harriman is using "induction" in the original sense, in which it does not mean probabilistic reasoning, but reasoning from individuals to generalizations.

    ReplyDelete