An excerpt from Maarten Boudry (Persuasion):

My doubts [about the value of focusing on logical fallacies] began when I was still in academia, teaching critical thinking to philosophy students and science majors alike. Fallacies are a favorite chapter in such courses. In some ways, they are ideal teaching material: they come in tidy lists and seem easy to apply. Many trace back to Aristotle and still parade under their Latin names—ad hominem, ad populum, ad ignorantiam, ad verecundiam (better known as the argument from authority), the slippery slope, affirming the consequent, and so on.

So I dutifully taught my students the standard laundry list and then challenged them to put theory into practice. Read a newspaper article or watch a political debate—and spot the fallacies!

After a few years, I abandoned the assignment. The problem? My students turned paranoid. They began to see fallacies everywhere. Instead of engaging with the substance of an argument, they hurled labels and considered the job done. Worse, most of the "fallacies" they identified did not survive closer scrutiny.

It would be too easy to blame my students. When I tried the exercise myself, I had to admit that I mostly came away empty-handed. Clear-cut fallacies are surprisingly hard to find in real life. So what do you do if your professor tells you to hunt for fallacies and you can't find any? You lower the bar. To satisfy the assignment, you expand your definition….

In 2015, I published a paper in the journal Argumentation with two colleagues arguing that fallacy theory should be abandoned. Here is its crux: every so-called fallacy closely resembles forms of reasoning that are perfectly legitimate, depending on the context. In formal terms, good and bad arguments are often indistinguishable. Worse, there is almost always a continuum between strong and weak arguments. You cannot capture that gradient in a rigid formal scheme. As my friends Hugo Mercier and Dan Sperber succinctly put it in The Enigma of Reason: "most if not all fallacies on the list are fallacious except when they are not." …

Read the whole thing here.