Why Some Truths Are So Certain They Can't Be False
Bulletproof Truths: When a Sentence Can’t Go Wrong
It’s a rainy afternoon, and your friend says, “If it’s raining, then it’s raining.” You roll your eyes. Of course that’s true. Nobody needs to look out the window. This kind of truth feels unshakable. Philosophers call them logical truths.
A logical truth isn’t just any true sentence. It’s a statement that stays true no matter what words you plug in for the non‑logical bits — as long as you keep the logical words exactly as they are. If you take a sentence like “Either it’s raining or it’s not raining,” you can swap “raining” for “snowing,” “dancing,” or “spaghetti,” and the sentence remains true. The magic is in those small logical words: if, then, and, not, all, some. They create a skeleton that supports truth all on its own.
But why do logical truths feel so bulletproof? For centuries, thinkers have argued that a logical truth must be true; it couldn’t possibly be false. This “must” is called necessity. Figuring out exactly what that necessity means has been one of philosophy’s longest-running puzzles.
Aristotle (384–322 BCE) gave one of the earliest answers. Look at this his classic pattern, or syllogism:
No desire is voluntary.
Some beliefs are desires.
Therefore, some beliefs are not voluntary.
If the first two sentences are true, the conclusion must follow. Why? Aristotle said the necessity comes purely from the logical form. Strip away the specific words (desire, belief, voluntary), and you get a shape like this:
No Q is R. Some P are Q. So some P are not R.
No matter what real-world things you put into P, Q, and R, if the premises turn out true, the conclusion will be true too. So the “must” is just a universal fact about the actual world: this pattern never fails. The necessity doesn’t reach into other realities; it only says the form works here and now.
The Form Trap: Why “A Widow Runs” Isn’t a Logical Truth
Not every bulletproof truth is a logical truth. Imagine the sentence:
If a widow runs, then a female runs.
This seems undeniably true. A widow is a female whose husband has died, so whenever a widow does something, a female does it. The sentence feels necessary and you might never need to check any facts to accept it. Yet its logical form — the shape you get when you drop the specific words — is something like this:
If a P Qs, then an R Qs.
Plug in new words: “If a log runs, then a dog runs.” Suddenly the sentence can be false. A log doesn’t run, but a dog might. Because the form doesn’t guarantee truth for every replacement, the original isn’t a logical truth. Logical truths must stay true no matter what non‑logical words you fill in. That’s what formality demands.
So what makes a word logical? The standard idea is that logical words are topic‑neutral. They can appear in any subject without tying you to a specific thing. “And” works the same whether you’re talking about stars, elephants, or video games; “under” doesn’t. Over time, philosophers have tried to pin down the exact boundary.
One modern proposal uses the idea of permutation invariance. Imagine you swapped every single object in the universe — your cat becomes the Moon, the Moon becomes your pencil, your pencil becomes the cat, and so on. Words like “philosopher” would scramble (your cat probably isn’t a philosopher), but “and” or “identical to” would still do exactly the same job. Logical expressions don’t care which object is which. Still, tricky cases pop up, like “male widow,” which never describes anything but might technically pass the swap test. The boundary stays fuzzy, but the core insight remains: logical truth depends only on the skeleton of logical words and their arrangement.
Knowing Without Looking: The A Priori Puzzle
Logical truths also seem knowable in a special way. You don’t need to count every raindrop to be sure “If it’s raining, then it’s raining” is correct. Knowledge that doesn’t depend on sense experience is called a priori. But where does that knowledge come from?
Gottfried Wilhelm Leibniz (1646–1716) answered like a rationalist. He believed the mind has a built-in capacity to perceive pure ideas and their connections, much like a mathematician sees that 2+2=4 just by thinking about numbers. On this view, logical truths are etched into the structure of thought itself.
Radical empiricists saw things differently. John Stuart Mill (1806–1873) argued that what feels like pure reason is just an extremely early generalization from experience. We’ve never seen a case where it rains and doesn’t rain at the same time in the same spot, so the pattern becomes a deeply ingrained habit that seems unshakeable. Later, W.V.O. Quine (1908–2000) suggested that logical truths are part of our whole web of belief. In principle, if future observations make it convenient, we might revise even basic logic — but he admitted that would be wildly disruptive.
Yet many philosophers find that hard to swallow. Could you really wake up one morning and decide that “and” no longer means what it always did? Probably not. So the debate stays open: is our sense of logical certainty a window into the mind’s deepest structure, or just a very useful habit carved by a lifetime of noticing patterns?
Cooking Logic: The Recipe for Finding All Logical Truths
By the late 1800s, logicians wanted a precise recipe for logical truth. Gottlob Frege (1848–1925) invented a formal language — a stripped‑down symbolic system where every sign has exactly one meaning. This let him build a deductive calculus: a few axioms (starting logical truths) and simple rules that generate new truths. Any sentence you can cook up from those ingredients is derivable. It’s like having a machine that prints guaranteed logical truths.
Around the same time, another idea took shape: model‑theoretic validity. Take a formal sentence, keep its logical words fixed, but let its non‑logical words change their meanings wildly — even in made‑up worlds. If the sentence comes out true under every possible reinterpretation you can dream up, then it’s model‑theoretically valid. Alfred Tarski (1901–1983) gave this idea a rock‑solid mathematical foundation.
For a large and very useful chunk of logic (called first‑order logic), these two approaches match perfectly. A great result by Kurt Gödel (1906–1978) proved that every model‑theoretically valid sentence is also derivable in a well‑built calculus, and vice‑versa. So for those logics, we have a crystal‑clear characterization.
Things get trickier when we allow ourselves to talk about “all properties” or “all sets” (higher‑order logic). In that case, some valid sentences resist derivation. Are they still genuine logical truths? Philosophers argue. The mathematical recipe may not capture every flavor of logical truth, but it has sharpened our questions enormously.
Why It Matters: The Key That Opens Itself
You might wonder: does all this nitpicking about logical truth really matter? It does, because logic is the backbone of every argument you’ll ever make. If you started doubting the simplest logical truths, you couldn’t even state your doubt without using them.
There’s a famous puzzle from the writer Lewis Carroll that captures the problem. A tortoise accepts “If A then B” and also accepts “A.” But the tortoise stubbornly refuses to accept “B” until you add a new rule: “If A and (if A then B), then B.” Once you add that, the tortoise demands yet another rule, and another, forever. You can never force the tortoise to take the last step by logic alone — you must just see that the pattern works. The rule modus ponens eventually has to be accepted without external proof. That’s a kind of logical truth you simply trust because reasoning can’t get going without it.
So next time you find yourself thinking “If it’s raining, then it’s raining,” you’re grazing a deep mystery. Logical truths sit at the very bottom of thought. Whether they reflect the shape of the world, the structure of your mind, or an unprovable starting point, they are the quiet engine behind every “therefore” you’ll ever utter.
Think about it
- Imagine a world where the word “and” doesn’t work the way you learned — maybe “p and q” means only one of the two is true. Could you still describe that world using sentences with “and”? What would happen to a truth like “It’s raining and it’s raining”?
- Suppose your friend refuses to believe that “or” means what you think it means. Can you prove the meaning of “or” without ever using the word “or” itself? Try it.
- Do you think a mind that had never encountered the world — perhaps a pure thinking machine — could figure out that “All cats are cats” is true? Or does that truth somehow depend on having met real cats?
