🧪 This platform is in early beta. Features may change and you might encounter bugs. We appreciate your patience!
weak_man_fallacy
The weak man fallacy occurs when an arguer selects the weakest, least competent, or most extreme proponent of an opposing position and refutes their version of the argument, then presents this as a refutation of the position as a whole. Unlike the straw man fallacy, no distortion of the argument occurs — the weak version is genuinely held by someone. The fallacy lies in the selection: by cherry-picking the weakest representative rather than engaging the strongest formulation, the arguer creates the illusion of having defeated a position they have not seriously confronted.
"Want to see what climate activism looks like? Here's a clip of an activist who can't name a single greenhouse gas. That tells you everything about the environmental movement."
A pundit shares a video of a poorly informed protester at a gun control rally who confuses basic legal terms, then concludes: 'This is who wants to rewrite the Second Amendment. These people don't even understand the laws they want to change.' — The entire policy debate is judged by its least informed participant.
During a company all-hands meeting, a senior executive responds to employee concerns about the return-to-office policy by reading out a single poorly worded complaint email and saying: 'This is the level of argument we're dealing with from those opposed to this change.' — The weakest expression of dissent is used to dismiss all objections.
Binary (yes/no) questions an LLM must answer to identify this aspect:
Does the arguer respond to a real (not fabricated) but weak version of the opposing position?
Type: binaryAre stronger, more sophisticated proponents or formulations of the opposing position being ignored?
Type: binaryDoes the arguer treat the refutation of the weak version as a refutation of the position in general?
Type: binaryThe weak man fallacy occurs when an arguer selects the weakest, least competent, or most extreme proponent of an opposing position and refutes their version of the argument, then presents this as a refutation of the position as a whole. Unlike the straw man fallacy, no distortion of the argument occurs — the weak version is genuinely held by someone. The fallacy lies in the selection: by cherry-picking the weakest representative rather than engaging the strongest formulation, the arguer creates the illusion of having defeated a position they have not seriously confronted.
The refuted argument is real, making it feel like fair engagement. Audiences rarely notice the selection bias because the refutation is genuinely successful — the weakness is in which version was chosen to refute, not in the refutation itself.
Ask whether the refuted position represents the strongest version of the opposing argument. Invoke the principle of charity: engage the best formulation of a position, not the worst. Name stronger proponents and ask the arguer to address their version.
Extremely common in media coverage that selects the most inarticulate protesters for interviews, in political debates that target fringe elements, and in social media where outlier posts are presented as representative.
Distorting or caricaturing an opponent's argument to attack it more easily.
Selectively presenting only evidence that supports one's position while ignoring or suppressing evidence that contradicts it.
The discourse tactic of selecting the most extreme, foolish, or ridiculous members of an opposing group and presenting them as representative of the group as a whole. The name is a portmanteau of 'nut' (slang for a crazy person) and 'cherry-picking.' This is a group-level version of the straw man.
The discourse practice of constructing the strongest possible version of an opponent's argument before responding to it. The opposite of straw-manning. When done genuinely, it improves discourse quality. It can be misused when the 'steel man' is constructed in a way that subtly misrepresents the original position or when the practice is performed performatively without genuine engagement.
Use these tools to detect, analyze, or train this aspect.