#77: Falsifiability and Historical Counter-factuals

August 26, 2010

All the pundits agree that high unemployment is the single most serious obstacle to the Democrat Party in the upcoming November elections. The standard scripted spin on this from the Democrat side is the argument that “things would have been immeasurably worse” had the Obama regime not spent the astronomically large amounts of money that it did. This claim is what philosophers call a “historical counter-factual” and the regime’s making it provides us with what Obama would call a “teachable moment.”

Students of history are regularly warned against these “historical counter-factuals” which are a species of “contrary-to-fact conditionals.” Karl Popper is at least as famous for his “Falsifiability Principle” as he is for taking on Plato and Marx. Popper’s principle states that a proposition is scientifically acceptable if and only if it is susceptible to empirical falsification. In intuitively simpler terms, it holds that we must be able to imagine real-world circumstances which would make a proposition false, if and if only that proposition is to be admitted to scientific inquiry. Please note: the principle only requires that we know what circumstances would make a proposition false, not that it be false. Since the problem with historical counter-factuals is that they are in principle unfalsifiable, they fall afoul of Popper’s Principle.

All of this must be looked at in the context of the early 20th c. philosophy called “Logical Positivism” and its “criterion of empirical verifiability.” These philosophers, using science as their foundation, were aggressively bent on erasing traditional philosophies as meaningless bafflegab. In order to do this, they needed a test for the meaningfulness of a proposition or sentence. Their candidate for such a test was empirical verifiability. That is, they argued that a given proposition has meaning if and only if it is capable of being empirically verified. But, unfortunately, this was a test that no proposition could pass. “Verification” means being shown to be true. But no amount of confirming observation can show a proposition to be true, the best that it can do is increase the probability that the proposition is true. The reason for this is that as long as the number of observations falls short of the infinite, there always remains the possibility that observations not yet made will go against the proposition. But even if we change the test to confirmation as opposed to verification, there will be questionable propositions that will pass the test. And, in addition, the claim that “unverifiable” propositions are “meaningless” is a hard sell. On the one hand, consider the following example:

Steel surfaces are not hard at all, they only become hard in direct proportion to how hard we press on them.

Now it should be clear that every experiment we conduct will turn out to confirm this proposition. But, at the same time, it should be equally clear that there is something wrong with it that is not caught by the requirement of empirical confirmability.

On the other hand, both verifiability and confirmability equally would toss out such statements as “There are no supernatural entities” as meaningless, not to mention all statements attributing values to acts, men, or objects. But, in addition, note that it also makes meaningless such sentences as:

Had Hitler not chosen to open the eastern front, he would have won the war.

This is an example of a historical counter-factual. What sort of empirical evidence could be adduced to verify or confirm it? Is it really true, though, that this sentence is meaningless? Do we not understand it? This is not a desirable outcome.

Popper avoids such problems by dropping the challenge to meaningfulness (shifting to scientific acceptability) and by making the test Falsifiability rather than either verifiability or confirmability.

The huge benefit of the latter move is that it only takes a single contrary observation to falsify a proposition, while it takes an infinite number to verify one and a very large number to confirm one. Further, any proposition that is falsifiable is also by that token confirmable.

Now historical counter-factuals are easy to identify structurally. They are always conditionals (if…then statements) and they are always in the subjunctive mood (would, had). These sentences make claims about outcomes or results in other worlds than the real one; they change something about the real world, and then propose to tell you what the consequence of that change is in their fictitious world. Thus, change the real history in which Hitler did attack Russia to a fictitious history in which he did not; the narrative immediately stops being a history at all and becomes instead a historical novel. But in a novel anything is possible and what happens is determined not by facts but rather by the virtually unlimited imagination of the author.

Once we give up our commitment to working only with facts by even a single exception, we have left the domain of Falsifiability. Since we are no longer submitting ourselves to the control of reality, how can there be anything at all like evidence any more? Evidence comes from observations of the real world and applies only to the real world. Once we have left the real world, observations no longer have authority.

You might wish to argue that, after all, you are only interested in making a single little change in the gigantic set of facts that makes up world history, but is it really possible to make only a single change? When we assume, contrary to fact, that Hitler did not attack Russia, we are no longer assuming the real Hitler. We are assuming a different Hitler entirely. But who, exactly, is this new Hitler? What is altered in his mind, in his personality, to make him avoid an Eastern Front? Were there perhaps contrary-to-fact events you want to assume that changed his mind? In which case you have gone beyond a single little fact. But even if you are going to put it all down to just a different Hitler, surely there still had to be different facts in your history that led to the existence of this different Hitler. And, further, once your Hitler decided to not attack Russia, how do you know that he would have won the war?

Once you change a fact in history, you put yourself into the position of any person today who is trying to predict the future. How successful are we today in predicting the outcomes of wars, of economic changes, of climatic changes, of cultural changes, of technological changes? We are very unsuccessful, and we are unsuccessful because of the innumerable interacting causes, impossible to even specify, much less to monitor. The counter-factual Hitler claim is as good or as bad as any claim today about the outcome of the Afghanistan war, neither claim is susceptible of either falsification or confirmation.

“Things would have been immeasurably worse had we not spent the money” is a historical counter-factual. It cannot be confirmed, it cannot be falsified, it should never enter the field of political discourse. It is, further, a strategically bad talking point for the Democrats, since it opens the door for historical counter-factuals on the other side.

I just heard Steve Forbes being interviewed on the television. His criticism of the Obama regime was interesting. He said, “had Obama not implemented his various policies and spent all that money, the economy would now be booming.”

What’s sauce for the goose is sauce for the gander. Once we open the Pandora’s box of counter-factuals, they flood out into the world, and while they are none of them capable of falsification or confirmation, right now the ones coming from the free-market spinners are far more likely to be believed and to be influential.