Why Some Truths Cannot Be Proved

What came before

Earlier parts of this book looked at problems in modern physics. They found that our best theories about reality, matter, cause, and method may not fully capture the world. This chapter moves to a different but related question. Are there limits not just to certain theories, but to our very tools of knowledge, logic, and formal reasoning? Can any complex system fully describe itself or prove its own base rules? The chapter looks at formal systems, self-reference, and limits found by mathematical logic.

The dream of perfect systems

In the early 1900s, many thinkers wanted to put all of mathematics on rock-solid ground. They hoped to build formal systems: precise structures with no ambiguity and no need for human guesswork. A formal system has four parts. First, a formal language: a fixed set of symbols and clear rules for building meaningful statements. Second, axioms: starting statements that are not proved. Third, rules of inference: mechanical rules for making new statements from old ones. Fourth, proofs: step-by-step chains where each step follows by rule from what came before. A theorem is the last statement in such a chain. The rules must be mechanical, so a machine could check any proof.

David Hilbert sought one single system that would cover all of mathematics. He wanted it to be consistent—never yielding a contradiction—and complete—able to settle every question. He also hoped to prove the system's consistency using only the simplest, most trusted methods. If this worked, mathematics would rest on unshakable ground. Gödel's work destroyed this hope.

Gödel's two blows

In 1931, Kurt Gödel showed that any formal system strong enough for basic arithmetic must have built-in limits.

The first blow: such a system, if consistent, is incomplete. This means there is at least one statement that the system can neither prove nor disprove. Gödel achieved this by making the system talk about itself. He gave numbers to symbols, statements, and proofs. Then he built a special sentence, the Gödel Sentence, which says of itself: "This sentence cannot be proved in this system." If the system could prove this sentence, it would be inconsistent. If it could disprove it, that too would cause trouble. So the sentence is undecidable: neither proved nor disproved within the system. Yet from outside, we can see the sentence is true. This shows that truth and provability are not the same. No single formal system can catch all arithmetic truths. And the problem never ends: add the new sentence as an axiom, and the bigger system will have its own undecidable sentence.

The second blow: such a system cannot prove its own consistency. The statement "this system is consistent" can be written as an arithmetic sentence inside the system. Gödel showed that the system proves: "If I am consistent, then my Gödel Sentence is true." Since the Gödel Sentence cannot be proved, consistency cannot be proved either. To show a system is consistent, you must step outside it and use stronger tools.

These results apply only to systems that are consistent, have mechanical rules, and are strong enough for arithmetic. Weaker or non-mechanical systems escape the theorems.

Old paradoxes, same pattern

Gödel's method of self-reference echoes older puzzles. The Liar Paradox says: "This statement is false." If true, it is false; if false, it is true. This shows that naive talk of truth inside a language leads to contradiction. Tarski later showed that a formal language cannot define its own truth without risk; we need levels of language.

Russell's Paradox asked: consider the set of all sets that do not contain themselves. Does this set contain itself? Either answer leads to its opposite. This broke naive set theory. The fix was axiomatic set theory, which blocks such strange sets by stricter rules.

These puzzles, like Gödel's theorems, arise when systems try to close in on themselves—defining truth in their own language, forming sets by any rule, proving their own consistency. Complete self-containment leads to contradiction or incompleteness.

The same limits in computing

Alan Turing found a matching limit in computation. He imagined an abstract machine, a simple model of any computer. He then asked: could a program tell, for any other program and input, whether that program will eventually stop or run forever? He proved that no such general program can exist. His proof used the same diagonal trick as Gödel: imagine a stopping-tester, then build a program that does the opposite of what the tester predicts, fed to itself. The result is a contradiction, so the tester cannot exist.

This undecidable halting problem is deeply tied to Gödel's first theorem. If a complete formal system existed, it could solve the halting problem by proving "stops" or "does not stop." Since no program can solve the halting problem, no complete formal system can exist. The theorems are fundamentally about the limits of rule-based, mechanical processes.

What this does and does not mean

The theorems have real philosophical weight. They show that no single formal system captures all mathematical truth. Mathematics is inexhaustible: there is always more outside any system. Truth and provability inside a system are different things. This challenges views that see mathematics as nothing but formal game-playing.

But many uses of the theorems outside mathematics are wrong or overstated. Some argue that human minds must be non-mechanical because humans can "see" the truth of Gödel sentences. This argument is widely considered flawed. It assumes humans are perfectly consistent and confuses informal insight with formal proof.

Some claim the theorems block a final theory of physics. This is speculative. Physical theories are not fully formalized in Gödel's sense. They may not need full arithmetic power. Undecidable statements may not matter for physics. The theorems might hint at limits on predicting complex systems, but they do not rigorously forbid a workable theory of everything.

Worst are uses in theology, ethics, law, or social theory. These fields lack the precise language, mechanical rules, and arithmetic content that the theorems require. Calling any gap or paradox "Gödelian" in these areas is loose analogy, not valid logic.

The bottom line

Gödel's theorems, with related paradoxes and the halting problem, show hard limits in complex formal systems. Any consistent system strong enough for basic arithmetic cannot prove all its own truths and cannot prove its own consistency. This dashes the hope of absolute certainty through purely formal methods.

These formal limits echo the problems in physics found earlier in this book. Just as quantum mechanics and general relativity seem unable to give a full, coherent picture of physical reality, our most rigorous thought-systems cannot fully capture truth or validate themselves. Perhaps physical limits are not just missing data, but reflect deeper limits like Gödel's, built into any complex system trying to represent itself.

This matters for how we think about science. It warns against expecting final answers, total certainty, or complete formalization. There may always be a gap between our formal descriptions and the world they try to capture. Other tools—experiment, philosophical critique, creative insight—remain vital. Knowing what formal methods cannot do helps us see where other ways of knowing must take over.

Words used here

Axiom

a starting statement in a logical system, treated as given without proof.

Consistent

a system that never lets you prove both a statement and its opposite.

Formal system

a set of symbols, rules, and starting points built so that a machine could check every step.

Self-reference

when a statement talks about itself, like "this sentence has five words."

Learn more in plain language

• Set theorySimple English Wikipedia
• Turing machineWikipedia (English, more technical)
• Philosophy of mathematicsSimple English Wikipedia
• EpistemologySimple English Wikipedia
• David HilbertSimple English Wikipedia
• AxiomSimple English Wikipedia