The Cases When Ad Hominem Arguments are Not a Fallacy

Eli Lyons
5 min readDec 31, 2022

--

Oh the falshoods they teach in school

I almost failed out of high school. And after being a PhD student at a top university, publishing scientific research, and working in industry, I can say with confidence that I was right about the pointlessness of most of high school ‘education’. However, not only is high school often a waste of time, you can actually learn the wrong things. Sometimes in college as well. I remember a class called Critical Thinking. I remember Logical Fallacies. Yet, I believe you can learn more of value from watching the show Succession than from a year of high school or a college Critical Thinking course.

(very minor spoiler alert)

There’s a scene in Succession where Kendal is in a tough position in a power struggle for his father’s company and gives Tom a pitch to team up. Tom rejects Kendal’s sneaky offer:

“Having been around a bit, my hunch is that you’re (Kendal) going to get fucked. Because I’ve seen you get fucked, a lot. and I’ve never seen Logan get fucked once.”

And this is why ad hominem arguments are not always fallacies.

This is also why Nassim Taleb uses phrases like ‘Intellectual Yet Idiot’.

What schools teach is that arguments need to follow something like boolean logic to be correct, and these are the kinds of arguments we should present publicly or in a debate. This is misleading. Arguments need to be useful. Ad hominem attacks are often useful for multiple reasons and I’ll cover two important ones here:

One

The case when the audience has difficulty assessing the logic used by a party in the debate, but can assess an ad hominem attack. This is kinda like what Trump does, sometimes passive-aggressively. This resonates with people because debates are just words, but character often matters more in many real-world situations. For example, say I tell some guy I am going to kick his ass. It doesn’t matter if I explain why I think that to be true, we’ll probably have to find out the truth through experimentation. That’s what’s going on in Ukraine now — playground logic. If I tell my friend I will meet him for lunch tomorrow but I flaked on him last time, it doesn’t matter how good my argument is, it matters if I have the kind of character that shows up. Most real-world (meaning a place that’s not a classroom) situations are not abstract discussions of logic, but rather discussions about what will happen in the future, or what we should actually do! Execution matters. I’m not promoting this ad hominem use case, but it has utility.

Two

Ad hominem arguments work when they represent a collection of events so complex, that all the data samples can’t be analyzed in detail clearly, or when a previous sample and a current sample are only vaguely similar. This is what most people learn from a young age, and why this logical fallacy has always seemed a bit off to me. Just like Tom, we know some people are often right and some people are often wrong. In the end, that’s much more information, like a meta-analysis on prediction accuracy, and more robust, than analyzing the inductive logic of an argument. This is perfectly useful and may lead to the correct result more than an attempt at analyzing a complex situation with many unknowns (isn’t a debate usually this case?), especially if you think probabilistically anyway. That’s how machine learning and such methods work as well. So Tom thinks that there’s a higher probability that Logan won’t get fucked and Kendal will, that’s what his training data predicts. No boolean logic. And it’s completely ad hominem.

The real world is extremely complex. You cannot even articulate all the processing going on in your brain in a debate. Note that Tom doesn’t say he thinks Logan will always win in the future, which is another common error, making incorrect predictions when the number of observations is low, he is just making a decision based on the observations he has so far…Tom is the modern day philosopher we didn’t know we needed. So the nuance is that ad hominim attacks are not always a fallacy, but even when they are not, they do not guarantee a correct result.

Of course, using this method of assessing an argument or predicting the future needs to be constrained by the context of the training data. For example, Tom is kinda considering Kendal’s argument in the context of business. People naturally trust experts in a field (sometimes too much), and what seems to be happening lately (or always has?) is that people start to apply that trust outside the context of the training data. For example, Elon Musk becomes a trusted source on ethics or in a somewhat new technology area, and creates a company that doesn’t deliver what it promises.

Usually, you have a friend that is a trusted source for the best restaurants, another that has good taste in movies, or another friend that always picks the wrong man to pursue. You don’t always need to pick apart the argument about why this one is not different from the other ones they’ve pursued, in fact, probably one of the reasons that won’t work is because one of the presented facts is just wrong and you have no way to prove it yet. “He has a great sense of humour!” cannot be argued against easily. “You have a poor sense of humour” is ad hominem, but also shows the reason they often picked the wrong man, they cannot consciously perceive the quality and nature of their judgement objectively.

Finally, ad hominim arguments include the ‘appeal to authority’ fallacy, because anytime you attack someone’s character, this is equivalent of saying the person is not an authority, and if the opinion of an authority is appealed to, this is ad hominim because it implies the opposing party opinion is not equally credible. I notice this is also often used when the person executing the fallacy does not have sufficient technical skills to analyze a situation or is lazy…etc. Of course sometimes appeals to authority can also yield the correct results, but often not…Remember when parents and universities told you going college would allow everyone to have a middle-class life, a life of luxuries like a house and maybe a german car? You MUST master the five paragraph essay! That will prepare you!

My next essay will be on why debates are a luxury.

--

--

Eli Lyons

A Hungarian man said to me, 'You don't talk much do you.' Co-founder/CEO of www.genomeminer.ai.