I am sure I have made a gross misunderstanding of Gödel's completeness theorems, as to me, it seems to follow that all sets of formulas are consistent.

Let $\Gamma$ be a set of formulas. If $\Gamma\vdash\psi$, then by Gödel's completeness theorem, $\Gamma\models\psi$.

Then $\Gamma\not\models(\neg\psi)$.

By Gödel again, $\Gamma\not\vdash(\neg\psi)$.

Hence, $\Gamma$ is consistent.

I am pretty sure there is a flaw in the above argument, but I can't quite pinpoint it.

Any help is sincerely appreciated!

  • 6
    $\begingroup$ I think that if $\Gamma$ is inconsistent, then both $\Gamma\models\psi$ and $\Gamma\models\lnot\psi$ hold trivially. $\endgroup$ – Marc van Leeuwen Apr 26 '13 at 9:38

Two things.

First of all, $\Gamma \vdash \psi$ implying $\Gamma \models \psi$ is known as the Soundness Theorem for the proof system $\vdash$ (i.e., "true premises do not prove false conclusions").

Now on your purported proof, the flaw occurs at "Then $\Gamma \not\models (\neg \psi)$". Namely, $\Gamma \not \models (\neg\psi)$ can only be valid if there is a model of $\Gamma$ in which $\neg \psi$ is false. In particular, we have tacitly assumed existence of a model of $\Gamma$ in the first place.

Thus your proof is circular.


If $\Gamma$ is inconsistent to begin with then it has no models. Therefore $\Gamma\models\psi$ holds vacuously, for every $\psi$.

In particular it follows that if $\Gamma$ is inconsistent then $\Gamma\models\lnot\psi$, contrary to your "proof".

That is, the step $(\Gamma\models\psi)\rightarrow(\Gamma\not\models\lnot\psi)$ requires the assumption that $\Gamma$ is consistent, and therefore resulting in circularity.


Check the second step. $\Gamma \not ⊨ \phi$ means there is a model of $\Gamma$, which cannot happen if $\Gamma$ is inconsistent.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.