Soundness and Completeness
Suppose that \(\Gamma\) is a set of qf formulas of \(LR(\sigma)\) and \(\psi\) is a qf formula of \(LR(\sigma)\).
\(\Gamma\vdash_\sigma \psi\) means that there is a derivation of natural deduction of \(\psi\) where all undischarged assumptions are elements of \(\Gamma\).
\(\Gamma\models_\sigma\psi\) means that for all \(\sigma\)-structures \(A\), if for all \(\chi\in\Gamma\), \(\models_A\chi\), then \(\models_A\psi\).
Soundness of Natural Deduction with qf Sentences
Let \(\sigma\) be a signature, \(\Gamma\) a set of qf sentences of \(LR(\sigma)\) and \(\psi\) a qf sentence of \(LR(\sigma)\). Then:
Completeness of Natural Deduction with qf Sentences
Let \(\sigma\) be a signature, \(\Gamma\) a set of qf sentences of \(LR(\sigma)\) and \(\psi\) a qf sentence of \(LR(\sigma)\). Then:
Hintikka Set
Let \(\sigma\) be a signature and let \(\Gamma\) be a set of qf sentences of \(LR(\sigma)\) (with only \(\wedge\) and \(\neg\) in the language). We say that \(\Gamma\) is a Hintikka set if it meets the following conditions:
- If \((\varphi\wedge\psi)\in\Gamma\), then \(\varphi\in\Gamma\) and \(\psi\in\Gamma\),
- If \(\neg(\varphi\wedge\psi)\in\Gamma\), then either \(\neg\varphi\in\Gamma\) or \(\neg\psi\in\Gamma\),
- If \((\neg(\neg\varphi)) \in\Gamma\), then \(\varphi \in\Gamma\).
- \(\bot\not\in\Gamma\).
- There is no atomic sentence \(\varphi\) such that both \(\varphi\) and \((\neg \varphi)\) are in \(\Gamma\).
- For every closed term \(t\), \((t=t)\) is in \(\Gamma\).
- If \(\varphi\) is atomic and \(s,t\) are closed terms such that both \((s=t)\in\Gamma\) and \(\varphi[s/x]\in\Gamma\), then \(\varphi[t/x]\in\Gamma\).
Lemma
If \(\Gamma\) is a Hintikka set for \(LR(\sigma)\), then there is some \(\sigma\)-structure \(A\) that is a model of \(\Gamma\).
Lemma
If \(\Gamma\) is a consistent set of qf sentences, then there is a Hintikka set \(\Delta\) such that \(\Gamma\subseteq \Delta\).
Decidability Theorem for qf Sentences
There is an algorithm which, given a finite set of qf sentences, will decide whether or not the set has a model.