In a cooperative argument, you should never* explicitly refer to logical fallacies.
*additional qualifications below
This is a point I’ve made several times before, especially in my How To Argue post.
In that post, I make a distinction between cooperative and competitive arguments. In a cooperative argument, you are trying to find the truth, which involves coming up with the best arguments you can, but does not necessarily involve “winning” the argument. In a competitive argument, you are trying to win the argument, which might involve coming up with truthful arguments, but not necessarily. I don’t mean to say there is anything inherently wrong with having a competitive argument, I’m just not talking about them here.
In a cooperative argument, you don’t want to antagonize the person you are arguing with (“interlocutor” is the term I would use). After all, your goals are aligned. You’re both trying to figure out the truth. And if you tell your interlocutor that they’ve just used a logical fallacy, I think there’s something inherently antagonizing about that. It’s like telling them not just that they’re wrong, but that they’re wrong in a particularly predictable and trite way.
It’s like, suppose you made one of those bingo cards, where each square contains an argument that you expect your opponent to make. While your opponent is making their argument, you visibly take out your bingo card, and then make a show of crossing out one of the boxes. Does this sound like it belongs in a cooperative argument? The bingo card is condescending, because you’re fitting your opponent into a small number of boxes, and you’re giving a stock response based on which box they used this time, rather than a respectful response based on actually listening to what they’re saying.
When you explicitly refer to logical fallacies, you’re pretty much doing the same thing. This goes not just for proper logical fallacies, but also other related terminology, like “strawman” or “burden of proof”.
Instead of referring to logical fallacies, I suggest just describing the problem that you are perceiving. For example, instead of “strawman”, you might say that they’ve misinterpreted what you said, or that you’ve never said it, or you said it but now you realize you didn’t mean it, or somebody else said it but you don’t agree with them, or hardly anybody else is saying it. Instead of “ad hominem”, you might say that their views don’t follow from their personal attacks, or that their personal attacks don’t advance the argument, or serve as a distraction, or that you just don’t like them.
Now, I will note some exceptions. Obviously if the argument is competitive, then maybe logical fallacies will score you rhetorical points. And maybe some of you have cooperative arguments where you openly antagonize your interlocutor. If it works for you, fine, I guess. I also have no issue with identifying logical fallacies in your own arguments, or in other people’s arguments, as long as you’re not identifying them in your interlocutor’s arguments. If you have no interlocutor (e.g. you’re writing a blog post for the world to read), then it depends on whether your intended audience identifies with the argument you want to say is fallacious.
I also note that some logical fallacies seem more antagonizing than others. For instance, I don’t have much of an issue with an issue with “non sequitur” because it’s a very broad term, that doesn’t necessarily imply that your interlocutor is being wrong in a specific predictable way. I also don’t have an issue with “sunk cost”, because “sunk cost” is a complicated concept that is pretty hard to convey without explicitly referring to it. Your mileage may vary on which logical fallacies are more or less appropriate, and it may vary by context as well.
Of course, besides antagonizing your interlocutor, there are other issues with explicitly referring to logical fallacies.
One problem is, it just looks amateurish. I get the sense that people who have just learned about a logical fallacy are the people who are most zealous about mentioning the fallacy whenever they have a chance. You don’t want to look like you’ve just learned about logical fallacies. The people who are long-familiar with these logical fallacies are not impressed by your apparent inexperience, and the people who are not familiar with the fallacies are not impressed by your apparent failure to communicate clearly. That means that maybe you want to avoid explicitly mentioning logical fallacies even in competitive arguments, where image is key.
The other problem is, I think logical fallacies encourage cookie-cutter thinking on your part. A few paragraphs ago, I described five distinct varieties of strawmen, and four distinct varieties of ad hominem. By labeling these many different situations as “strawmen” or “ad hominem”, you are not encouraging yourself to think about distinct situations that call for distinct responses. You are encouraging yourself to fit a large number of situations into a small number of boxes, offering stock responses to each box. This is useful in a pedagogical context, where you start out by learning about a small number of types of arguments, and the appropriate responses to each. But if you want to grow beyond critical thinking 101, then you should make increasingly fine-grained distinctions between different arguments, and learn to deliver the right response for each situation.