Re: On Godwin's Law

Mike Godwin (
Mon, 16 Mar 1998 18:21:10 -0500

Certainly I never said anything originally about whether the mention of
Nazis means the discussion is over -- there are countless instances to the

The primary mutation of Godwin's Law (not discussed, infra, for obvious
reasons) states that when such a reference or comparison occurs,
*meaningful* discussion is over.

Speaking empirically, I think there is a lot of truth to this mutation.

It is a common mistake to interpret the original version of Godwin's Law as
primarily a probability statement. If that were so, it would never have had
any currency. It is intended, of course, as a statement about online
rhetorical excesses, and as an implicit criticism of one of those excesses.


At 4:41 PM -0500 3/16/98, Lloyd Wood wrote:
>On Mon, 16 Mar 1998, Joseph M. Reagle Jr. wrote:
>> On Fri, 13 Mar 1998, Gregory Alan Bolcer wrote:
>> >BTW, this is fuel for my theory that any thread longer than
>> >3 replies will converge on the topic of sex or death, 97% of the time.
>> Isn't this just a variant of Godwin's Law?
>> Godwin's Law
>> /prov./ [Usenet] "As a Usenet discussion grows longer, the
>> probability of a comparison involving Nazis or Hitler approaches
>> one." There is a tradition in many groups that, once this occurs,
>> that thread is over, and whoever mentioned the Nazis has
>> automatically lost whatever argument was in progress. Godwin's Law
>> thus practically guarantees the existence of an upper bound on
>> thread length in those groups.
>You did a search on altavista for Godwin's law, right? Did you think
>to verify the accuracy of the source you quote? No, I thought not -
>but we'll get to that in a moment.
>First, I see that baisley added:
>| Thus Godwin's Law is the end of its own losing, one-post argument.
>A nice, quick, but not completely thought out, riposte.
>Logic would suggest that since Joe first mentions Hitler in the
>thread when he invokes Godwin's law:
>1. Joseph loses the argument, by Godwin's law as stated above (and
> Greg's assertation of convergence is not just a variant of Godwin's
> law).
>2. Since losing the argument means Joseph must be wrong, Godwin's law
> must be wrong.
>3. By continuing the thread as we are doing, we are providing a
> practical demonstration that Godwin's law is, indeed, wrong.
>So far, so good - at first glance. But #1 assumes Godwin's law is
>right, and #2 and #3 demonstrate that it is wrong. How do we resolve
>this evident paradox?
>The first obvious fallacy here lies in step 2; in fact, on usenet and
>in any forum of self-selecting public discussion, arguments are
>*always* lost by the people who are right and who know what they are
>talking about (generally because no-one else understands the
>ramifications of the points they make, and because the people who
>actually know what they are talking about and who are willing to waste
>their time contributing to usenet are in an ever-decreasing minority.
>No wonder those people get frustrated and are said to invoke Godwin's
>law before eventually leaving in disgust at the lowest common
>denominator that mass consensus brings, you say.)
>That much would be consistent with baisley's conclusion. Godwin knew
>what he was talking about, therefore he is right, but would appear to
>lose his own public argument on usenet in one post simply _because_ he
>is in actuality right.
>But continuing the thread - as we are doing now - disproves Godwin's
>law as stated (even though he's publically wrong yet actually right).
>How can we account for this?
>Quite simply, anyone on usenet who remembers Godwin's law and quotes
>it as received wisdom must be wrong, as all received usenet wisdom is
>obviously wrong, since the people with the time and inclination to
>post to usenet and spread such information are clueless morons who
>can't get anything right and who always shout down the people who know
>what they are talking about. Those few knowledgeable people then
>leave, embittered, to write books to put their points of view across
>(other than those usenet-like computer books with shelflives of less
>than three months, of course).
>This is a corollary of arguments on usenet being lost by the people
>who are right; people who are always wrong will thrive on usenet.
>We can find support for this view in Godwin's own description of his
>law, 'Meme, Counter-meme':
>'I developed Godwin's Law of Nazi Analogies: As an online discussion
> grows longer, the probability of a comparison involving Nazis or
> Hitler approaches one.
> I seeded Godwin's Law in any newsgroup or topic where I saw a
> gratuitous Nazi reference. Soon, to my surprise, other people were
> citing it - the counter-meme was reproducing on its own! And it
> mutated like a meme, generating corollaries like the following:
> Gordon's Restatement of Newman's Corollary to Godwin's Law:
> Libertarianism (pro, con, and internal faction fights) is the
> primordial discussion topic. Any time the debate shifts
> somewhere else, it must eventually return to this fuel source.
> Morgan's Corollary to Godwin's Law: As soon as such a comparison
> occurs, someone will start a Nazi-discussion thread on
> alt.censorship.
> Sircar's Corollary: If the Usenet discussion touches on homosexuality
> or Heinlein, Nazis or Hitler are mentioned within three days.
> Van der Leun's Corollary: As global connectivity improves, the
> probability of actual Nazis being on the Net approaches one.
> Miller's Paradox: As a network evolves, the number of Nazi comparisons
> not forestalled by citation to Godwin's Law converges to zero.'
>That article clearly outlines the corruption of the original statement
>of the law, but at no point does it mention the end of a thread.
>It would appear that the popular misstatement of Godwin's Law that
>Joseph cites has actually evolved to serve the ends of the clueless
>morons who propagate it, as adding the perceived relevance to usenet
>helps to ensure its continued propagation in the environment of
>usenet. Memes can, and will, adapt to their hosts.
>Thus, this misquoted Godwin's law, in the form given by Joseph above,
>is a fallacy spread by clueless morons on usenet and now on the web
>(and some of us even knew that already), because that's how it best
>survives. The earlier problem I raised of the thread being continued
>appearing to thwart Godwin's law, as stated by usenet, is resolved by
>usenet and accepted wisdom being, as usual, completely wrong. There is
>no paradox.
>Dave Crook adds:
>< I think there is also a corrolary stating that as a usenet
>< discussion grows longer, the probability of someone mentioning
>< Godwin's Law also approaches one.
>In general, as a thread grows, the probability of someone mentioning
>_anything_ approaches one. This is Pretty Damn Obvious - and it would
>even include sex and death, too. Where I come from, we call this
>As a thread grows, the probability of Hitler being mentioned
>approaches one. That is all. Godwin is shown to be right in his
>statement, since, after all, it's just simple probability. Received
>usenet wisdom and Godwin's law disagree on the details of Godwin's
>law, because, as previously stated, usenet is filled with clueless
>morons who will misquote you (and who incidentally can't do stats,
>Now, as a corollary to the treatment we've seen of Godwin's (original)
>law, I am able to state that anyone who mentions Hitler in a thread is
>very probably right and deserves to win their argument, despite
>popular opinion as asserted by clueless morons who will appear to win
>in public and who will then go on to misquote them for the rest of
>eternity. Godwin stating his own law in public is obviously a very
>clear example of this. (Thanks for the germ of that one, baisley.)
>Mentioning Hitler and being completely dead-on absolutely right has
>certainly been the case for me in the many times I've been forced
>through desperation to bring up Hitler in the past while arguing with
>the accepted wisdom of clueless morons on usenet and elsewhere; do
>note that I'm mentioning Hitler right now, and I hope that you all
>appreciate the ramifications of the point that I'm making here.
>Feel free to disagree with and misquote me, since it's pretty obvious
>from the corollary I've just inferred that you'll be completely wrong,
>and since you're going to do that anyway I may as well give you
>permission to go ahead and make fools of yourselves. You're just
>victims of an adaptive meme; you're paralysed by peer pressure, you're
>mere victims of what you think you know; you're mesmerised by a
>mutated meme.
>See if I care. I'm off to write a book and I'll have the last laugh.
>History will remember my words; it won't remember usenet's.
>Overall, this seems pretty straightforward to me. Any questions?
>Release new memes at your peril, for the mutations will come back to
>haunt you.
><>PGP<>+44-1483-300800x3641 '

We shot a law in _Reno_, just to watch it die.

Mike Godwin, EFF Staff Counsel, is currently on leave from EFF,
participating as a Research Fellow at the Freedom Forum Media Studies
Center in New York City. He can be contacted at 212-317-6552.