Against Godwin’s Law

Godwin’s law: true but abused

Originally formulated in 1989 by Mike Godwin, his law states: ‘As an online discussion grows longer, the probability of a comparison involving Nazis or Hitler approaches 1.’  Now I have no trouble with this statement: evidence suggests that it is almost certainly correct.

However there is another ‘Godwin’s Law’ (I shall use scare-quotes from now on to distinguish it from the original law).  In the words of Wikipedia:  ‘For example, there is a tradition in many newsgroups and other Internet discussion forums that once such a comparison is made, the thread is finished and whoever mentioned the Nazis has automatically “lost” whatever debate was in progress.‘  This statement, which is related to Godwin’s Law only in that it includes the word ‘Nazis’, is what I wish to deplore; I intend to show that it is a dangerous means of stifling debate that allows proponents of untenable propositions to maintain the appearance of being in the right by rubbishing those who differ from them.  

My attack will take the following form.  First I will demonstrate that ‘Godwin’s Law’ is useless as a tool for telling whether a statement is true or not.  It can give the wrong result and it can give the right result.  Second, I will examine some forms of argument where an appeal to Hitler or the Nazis may be justified (in context): specifically slippery-slope arguments and counter-examples.

Now the point of my argument is not that ‘The Nazis did it’ is always the best, or even a good argument: in fact it can become rather tiresome.  Rather, we should treat comparisons to Hitler and the Nazis on a case-by-case basis, rather than simply assuming that anyone who invokes them has lost the argument.  We should be wary of those who wish to reduce argument to a discussion of extreme cases, but a lazy application of ‘Godwin’s Law’ is not the answer.  As usual, we need better, subtler tools.


‘Godwin’s Law’ is useless

Let me show that ‘Godwin’s Law’ has no power, positive or negative, for determining the truth of an argument.  In other words it is useless as a tool for deciding who ‘won’ an argument.  

So here are two arguments.  Consider:

A: All vegetarians are peaceable people

B: Hitler was a vegetarian

A: ‘Godwin’s Law’.  You lose!

Here A claims to have won the argument, in spite of the fact that B has successfully negated their premise, so in this case ‘Godwin’s Law’ gives the wrong result.  Now consider:

A: Performing experiments on unwilling human subjects does not mean a society is not open

B: The Nazis did it

A: ‘Godwin’s Law’.  You lose!

Here A claims to have won the argument, and happens to be in the right, but only by accident, as it is not ‘Godwin’s Law’ that makes the Nazis an invalid counter-example, but rather the fact that a number of open societies (e.g. UK, US) have acted in ways that run counter to their founding ideals, but this does not negate their open status.  

Valid uses of Hitler and the Nazis in arguments

The slippery-slope argument

What I mean by a slippery-slope argument is a situation where I assert that some proposition is generally true.  If:

  1. I can point to some starting-point scenario in which it is provably true
  2. I can also show / assert that if the proposition is true for some one scenario, then it is true for all other scenarios sufficiently close to the specified one

Then we conclude that the hypothesis is true in any scenario that can be ‘connected’ to the starting-point, in the sense that I can get from one to the other by applying small modifications.  This is called the slippery-slope for obvious reasons: if you start at the top of the slope and can roll a little way down, there is nothing to stop you getting to the bottom.

Now if I can find a scenario where the proposition is not true, then there are two possibilities: either there is a class of scenarios that cannot be connected to the starting-point, where the proposition is not true; or assumption 2 is not (always) true, in which case the conclusion is not true.  So in either case the proposition is not universally applicable.  

As an example, consider the proposition ‘all disputes can be resolved peacefully’ which leads to the slippery slope proposition:

P: This dispute can be resolved peacefully

It is indisputably the case that there exist disputes that can be resolved peacefully, so P is true somewhere.  Moreover, it seems reasonable that if a given dispute can be resolved peacefully, a slightly more or less complex dispute can also be resolved peacefully.  So conditions 1 and 2 for the slippery-slope hold, and all disputes connected to my starting-point can be resolved peacefully.  But my original proposition says ‘all disputes’.  Are all disputes connected to a clearly peacefully-resolvable starting-point?

And this is where Hitler and the Nazis comes in.  They were so extreme, so unreasonable, so unpeaceable, so inhuman that they can provide a counter-example to more-or-less any attempt to assert a universal positive about human nature or sociology.  In this case, applying P’ to ‘the crisis precipitated by the invasion of Poland’ cannot (despite some rather inventive special pleading by agenda-ridden historians) realistically be said to give a positive result.  Thus P is false.  What is interesting, of course, is to determine:

  • Whether the Nazis are an isolated special case, or whether there is a more-or-less large class of disputes insusceptible to peaceful resolution
  • Whether there is some external criterion that we can apply to disputes to determine whether they are susceptible to peaceful resolution (think how much that would simplify the work of the UN!)

Now, you could argue ‘all you’ve done is find a counter-example, why bother with the slippery-slope stuff?’   Well, my intention here is not to prove the original proposition untrue, but to prove that it has limited applicability, to make some effort at determining the limits of said applicability, and to hint at the fact that there may be something deeper going on that repays further study.  If I merely wish to prove a proposition untrue, a simple counter-example will do, and to them we now turn.


The slippery-slope argument eventually came down to finding a counter-example to a proposition.  However, more generally, the sheer excessive negativeness of Hitler and the Nazis, the fact that they negate so many of the values we hold dear, makes them brilliant counterexamples for unwisely thought-through generalisations.  E.g.

  • P: Dog-lovers are good people
  • Hitler was a dog-lover


  • P: A pure, untainted Englishwoman could never be drawn to anything but good
  • Unity Mitford fell in love with Hitler

or even

  • P: Bad people must be as evil in their manner as they are in their hearts
  • Hitler was generally considered rather charming

In a sense what I am drawing attention to is a combination of the banality of evil, mixed in with the regrettable tendency of normally good people to be drawn to it.  Yes, Hitler was a monster, responsible for the deaths of tens of millions of people and architect of a systematic attempt to remove specific cultural / racial groups from the Earth, but he still loved his dog very much.  Regrettably, evil is not as simple as some would like us to believe, with its proponents being pantomime villains, and any tool that allows us to reveal the complexity beneath that simple label should be used.  So I say a resounding ‘no’ to ‘Godwin’s Law’. 


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s