Don’t Get Triggered by Analogies – They are a Powerful Tool for Logical Reasoning

I was recently reading a book by US General McChrystal, and his experiences fighting the Al Qaeda terrorist insurgency in Iraq. In the book, he brought up something that I found particularly interesting.

The world has become “flatter” and faster. People are more connected, more mobile, and move faster than ever before. These changes have ushered in a universe of new possibilities for players operating outside the conventional systems: Mark Zuckerberg, without family connections, starting capital, or an undergraduate degree, changed the world before his mid-twenties. Justin Bieber posted a self-made video in 2007, and has since sold 15 million albums. And Abu Musab al-Zarqawi, buoyed by online bomb-making instructions and the power to recruit and disseminate propaganda worldwide, incited a war.

Interconnectedness and the ability to transmit information instantly can endow small groups with unprecedented influence: the garage band … the viral blogger, and the terrorist cell

First and most important, the above was very insightful and interesting, especially coming from a general who had first had experiences fighting against modern terrorist organizations. Second, as an online blogger myself, I could immediately picture the way simpletons and character assassins can quote the above passage out of context to demonize the author. 

“General McChrystal compares Mark Zuckerberg to mass-murderer and terrorist al-Zarqawi.” 

Or if you want to take a different angle: 

“General McChrystal draws similarities between online bloggers and terrorists.” 

The clickbait and bad-faith arguments practically write themselves.

The above might seem like a stretch. Thankfully for the general, nobody would believe that he’s actually trying to attack Zuckerberg or online bloggers. Unfortunately for the rest of us, the above bad-faith arguments are par for the course in almost any political discussion.

Adam: People should never face a legal obligation to stay at a job they no longer want to do. Even if they agreed to it previously. Otherwise, it’s indentured servitude, and that should never be allowed.

Ben: Military personnel are legally obligated to serve for the period they committed to. Are you accusing the US government of hiring indentured servants?

Adam: Oh my God, did you just compare our soldiers to indentured servants? Those are completely different things, how can you even compare them? That is so offensive to our brave men and women in uniform.

Or:

Adam: Joe Biden is the leader of our republic. Therefore, we have a duty to support him

Ben: Hitler was once also the leader of a republic. Did the German people have a duty to support him as well?

Adam: Oh My God, did you just compare Biden to Hitler!?

Breaking News: Ben draws comparison between Biden and Hitler. Says we shouldn’t support either of them

(the above example works equally well if you replace Biden with Trump, if that better suits your political sympathies)

I’ve heard variants of the above countless times in my life. I’ve heard it from random internet commenters, respected newspaper journalists, and even PhD-equipped professors. There’s even an entire social rule around this topic – “when a Hitler comparison is made, the thread is finished and whoever made the comparison loses whatever debate is in progress.”

 All of them seem to have missed the point of the above exchange, and similar conversations. Some similarities can and do exist even among people or things that are extremely different from one another. Pointing out these similarities is a valid and useful way to either draw insights or examine the validity of broad statements.

In the last example above, Adam stated that Biden (or Trump) deserves our support based on his electoral success. And Ben pointed out that if we adopted this principle of always supporting electoral winners, then it would logically require the German people in the mid 1930s to support Hitler. Clearly, such a proposition is nonsensical, and hence, we must reject this principle. The above is a perfectly valid point raised by Ben, and one deserving of a thoughtful response from Adam.

  • Perhaps Adam could say that Hitler’s electoral victories were illegitimate because he used violence to suppress his political rivals
    • Though that would imply that Hitler would be deserving of support, despite his morally unconscionable behavior, if he had fairly won the popular vote
  • Perhaps Adam could say that we should support our leaders, but only if they aren’t responsible for the death of innocent people
    • Though if we’re making an exception for the deaths of innocents, why not also make an exception for other bad actions that hurt people in other ways?
  • Perhaps he could say that we should support our leaders as a general rule, though we should make exceptions for leaders who pursue very bad policies or exhibit poor conduct.
    • Though he would then need to defend Biden from accusations of bad policymaking or poor conduct
  • Or perhaps he could recant and admit that winning an election doesn’t make someone worthy of support at all. And that we should judge all leaders by their actions, not their popularity.

Either way, Adam’s initial statement was overly simplistic, and needs to be further refined. And by pointing out the most extreme counter-example possible, Ben clearly and concisely pointed out the error in Adam’s original statement. 

This doesn’t mean that Ben is accusing Biden (or Trump) of being very similar to Hitler. One of them is clearly far far more evil – only a fool doesn’t see that. But only a fool would also refuse to acknowledge that they have some things in common. Such as the fact that all three were popular, performed well in national elections, and became the leaders of their republic at some point. And these partial similarities can be used to test the validity of broadly stated principles – such as the principle of always supporting the leaders of our republics.


In fact, this is something that scientists, philosophers and mathematicians all do on a regular basis. When someone proposes a new principle, theory, or theorem, the fastest way to sanity check it is to apply the most extreme example you can think of, and see if it still holds up strong. And if it doesn’t, that single counterexample is all that is needed to disprove the theory and point out the need for further refinement. Most famously, Newton’s laws of motion, after standing tall for centuries, were disproven and superseded by Einstein’s more refined theory of general relativity – all precipitated by an experiment measuring minuscule changes in the position of starlight during a total solar eclipse.

Saying that the example provided is “too extreme” or “too different” or “applies only to special circumstances” and “therefore not relevant” would be the most idiotic defense imaginable. And so too with anyone who makes an overly broad statement – and then complains when it is disproven using a counter-example they prefer to sweep under the rug. And for goodness’ sakes, please don’t go around tarring and feathering people for “comparing online bloggers to terrorists.”


Addendum: If ABC is very bad, and similarities exist between ABC and XYZ, that does not mean that XYZ is also bad. Or that the thing they have in common is bad. Hitler was said to be a great public speaker. Obama is also said to be a great public speaker. Only a nitwit would then conclude that being a good public speaker is a bad thing. Or that Obama is a bad person because he has something in common with Hitler.

This sounds obvious when you explicitly point it out. But the number of political arguments based entirely on “guilt by association” or “guilt by similarity” is mind boggling. Don’t fall for it. Case in point.