A simple tool for exposing the hidden premises that make weak conclusions sound convincing
This one is a sidebar to Essay 13 – call it Essay 13.5.
In Essay 13, I mentioned that Stephen Toulmin was pushing back against formal logic – the kind with syllogisms and proofs. Here’s the thing though – the more I’ve thought about it, the more I think formal logic is actually really useful. Not for building arguments from scratch, but for tearing bad ones apart.
A syllogism is a specific argument structure philosophers have been using since Aristotle.
It has exactly three parts. A major premise, which is a general claim. A minor premise, which is a specific case. And a conclusion.
Here’s the famous example – all men are mortal, Socrates is a man, therefore Socrates is mortal. If both premises are true, the conclusion HAS to be true. No exceptions. The structure guarantees it.
Now, when somebody makes an argument that sounds convincing on the surface but feels wrong underneath, you can often expose the problem by rebuilding it as a syllogism. This forces the hidden assumptions out into the open – and that’s usually where the argument falls apart completely.
Let me show you with a real example based on a premise I push back on a lot.
The argument – “AI uses too much energy, therefore we should stop using AI.”
Sounds reasonable enough on the surface. AI does use significant energy. Concerns about energy consumption are legitimate. But something about it is off. So let’s rebuild it as a syllogism and see what’s actually being claimed.
Major premise – we should stop using anything that uses too much energy. Minor premise – AI uses too much energy. Conclusion – therefore, we should stop using AI.
Now look at that major premise.
Do you actually believe it?
Because if you accept that premise, you also have to accept that you should stop using air conditioning, refrigerators, the internet, hospitals, airplanes, and basically the entire modern world. Every one of those things uses significant energy. None of them would survive the rule.
Nobody actually believes the major premise when it’s stated explicitly.
The argument only works as long as the major premise stays hidden. That’s the whole trick. Bad arguments almost always depend on unstated assumptions that nobody would agree to if they were forced to say them out loud.
And there’s a second problem with the AI argument.
Even if we granted that high energy use is a serious concern, “stop using AI” isn’t the only possible conclusion. It isn’t even the most logical one. You could make AI more energy-efficient. You could power it with renewable energy. You could use it specifically to optimize energy grids and reduce overall consumption. You could prioritize high-value uses and limit the frivolous ones. You could invest in fusion and other breakthrough energy technologies.
The jump from “X has a problem” to “eliminate X entirely” skips over a dozen more reasonable responses. It’s like saying cars cause accidents, therefore we should ban all cars – ignoring seatbelts, airbags, traffic laws, better road design, driver training, and roughly a hundred years of automotive safety evolution.
So when you run into an argument that leaps to an extreme conclusion, there are two questions to ask.
What would the major premise have to be for this argument to work? And are there other conclusions that follow more reasonably from the actual evidence on the table?
Most bad arguments fail at least one of those tests. Either the hidden premise is something nobody would actually endorse if you spelled it out, or the conclusion is skipping over several more sensible options to land on something dramatic.
The syllogism structure isn’t how people naturally argue in everyday life. But it’s a fantastic tool for reverse-engineering arguments to find where exactly they break down.
Force the premises into the open.
Check if they’re actually true.
Check whether the conclusion is the ONLY one that follows.
Nine times out of ten, it isn’t.