Thursday, July 17, 2008

Godwin's law is wildly overated

Godwin's Law is a statistical formulation of people's tendency to compare things they think are bad or unfair one the one hand with Hitler on the other. Technically, it's theorized that the probability of any heated or controversial argument referencing Hitler approaches 100% as the conversation gets longer.

Recently, the idea of Godwin's law has been reflected on the one of my favorite (and hilarious) blogs "Stuff White People Like"

Yes I admit, sometimes people compare things that shouldn't be compared. Sometimes people exaggerate. And yes, in particular, people like to use Hitler as a point of exaggeration.

But, there are a couple of things that bug me about the (now pretty old) sensation of Godwin's law.

First, the idea that people sometimes exaggerate and over compare things to Hitler is not new, nor is it a great discovery made by some guy named "Godwin." Any intelligent person who has ever had a conversation in which somebody made an inappropriate rhetorical reference to Hitler has probably figured this out.

Second, as wikipedia pointed out when I read the article on Godwin's Law, (note: wikipedia changes-so it may no longer be on the site) the exact formulation of Godwin's law (that probability of the mention of Hitler goes up as a conversation continues) is pretty much simply a law of big numbers. Indeed the probability of ANY idea being mentioned is more probable as the conversation gets longer. Thus, Godwin's pseudo-intellectual formulation of the old concept is also pretty worthless.

Finally, sometimes references to Hitler are indeed informative. For example, suppose someone argues that a policy would be good for the economy. The other person may argue that
thats nice, but an economics's policy isn't everything. Suppose the other person says that economics should drive every decision. Wouldn't a sensible response be "even Hitler (at first) improved the economy with his polices?"

Indeed, comparisons to Hitler are what make the history of Hitler instructive. If we are too careful to ban comparisons to Hitler, we make the entire story of Hitler irrelevant and unhelpful toward preventing things like that from recurring (which-by the way-we definitely have not done a very good job of-if things like the Rwandan genocide, the current situation in Sudan, Cambodia under Pol Pot, Bosnia in the 90's, China under Mao, Russia under Stalin, or Saddam Hussein's Iraq are any indication).

True, sometimes these comparisons are so exaggerated and fanciful that they are irrelevant and unhelpful-but the mentality of Godwin's law simplistic formulation encourages us to ignore even those comparisons which would be helpful.


Martin Weiss said...

I think this is a case where you have to consider the difference between the detonation and the connotation.

Yes, strictly speaking (the denotation), Goodwin's Law is trivially true.

However, the connotation is that people use the Hitler comparison far too much (the reductio ad Hitleriam) and that, in a heated conversation, this is fairly likely to happen. When I first heard about Goodwin's law I thought it was something like, "In a heated debate the mean time before someone accuses their ideological opponent of being either Hitler or Chamberlain, is 15 comments."

George Weiss said...

yes. but there is a problems with the connotation too. basically-that people who like Godwin's law tend to label every Hitler analogy as bad.

now that there is Godwin's law-people who were upset with the Hitler analogues can more easily label even good Hitler comparisons as bad by saying "oh Godwin's law."