Structures

Mort Yao

Structure. A structure (or interpretation) \(\mathfrak{A}\) for a first-order language \(\mathcal{L}\) is a function whose domain is the set of parameters and such that

  1. \(\mathfrak{A}\) assigns to the quantifier symbol \(\forall\) a nonempty set \(|\mathfrak{A}|\) called the universe (or domain) of \(\mathfrak{A}\).
  2. \(\mathfrak{A}\) assigns to each \(n\)-place predicate symbol \(P\) an \(n\)-ary relation \(P^\mathfrak{A} \subseteq |\mathfrak{A}|^n\), i.e., \(P^\mathfrak{A}\) is a set of \(n\)-tuples of members of the universe.
  3. \(\mathfrak{A}\) assigns to each constant symbol \(c\) a member \(c^\mathfrak{A}\) of the universe \(|\mathfrak{A}|\).
  4. \(\mathfrak{A}\) assigns to each \(n\)-place function symbol \(f\) an \(n\)-ary operation \(f^\mathfrak{A}\) on \(|\mathfrak{A}|\), i.e., \(f^\mathfrak{A} : |\mathfrak{A}|^n \to |\mathfrak{A}|\).

Let \(\varphi\) be a wff of the language \(\mathcal{L}\), \(\mathfrak{A}\) be a structure for the language \(\mathcal{L}\), \(s : V \to |\mathfrak{A}|\) be a function from the set \(V\) of all variables into the universe \(|\mathfrak{A}|\). We say that \(\mathfrak{A}\) satisfies \(\varphi\) with \(s\), if and only if the translation of \(\varphi\) determined by \(\mathfrak{A}\), where the variable \(x\) is translated as \(s(x)\) wherever it occurs free, is true, written as \[\models_\mathfrak{A} \varphi[s]\] Formally, we define the extension function from the set \(T\) of all terms into the universe of \(\mathfrak{A}\): \[\bar{s} : T \to |\mathfrak{A}|\] such that

  1. For each variable \(x\), \(\bar{s}(x) = s(x)\).
  2. For each constant symbol \(c\), \(\bar{s}(c) = c^\mathfrak{A}\).
  3. If \(t_1, \dots, t_n\) are terms and \(f\) is a \(n\)-place function symbol, then \[\bar{s}(f t_1 \cdots t_n) = f^\mathfrak{A}(\bar{s}(t_1), \dots, \bar{s}(t_n))\]

For atomic formulas,

  1. \(\models_\mathfrak{A} = t_1 t_2 [s]\) iff \(\bar{s}(t_1) = \bar{s}(t_2)\).
  2. \(\models_\mathfrak{A} P t_1 \cdots t_n [s]\) iff \(\langle \bar{s}(t_1), \dots, \bar{s}(t_n) \rangle \in P^\mathfrak{A}\).

For other wffs,

  1. \(\models_\mathfrak{A} (\lnot \varphi) [s]\) iff \(\not\models_\mathfrak{A} \varphi[s]\).
  2. \(\models_\mathfrak{A} (\varphi \to \psi) [s]\) iff either \(\not\models_\mathfrak{A} \varphi[s]\) or \(\models_\mathfrak{A} \psi[s]\) or both.
  3. \(\models_\mathfrak{A} \forall x \varphi[s]\) iff for every \(d \in |\mathfrak{A}|\), we have \(\models_\mathfrak{A} \varphi[s(x|d)]\), where \(s(x|d)\) is defined as \[s(x|d)(y) = \begin{cases} s(y) & \text{if } y \neq x \\ d & \text{if } y = x \end{cases}\]

Example 1. \(\mathfrak{A} = (\mathbb{N}; \leq, S, 0)\) is a structure for a first-order language:

  • \(|\mathfrak{A}| = \mathbb{N}\), the set of all natural numbers,
  • \(P^\mathfrak{A} =\) the set of pairs \(\langle m, n \rangle\) such that \(m \leq n\),
  • \(f^\mathfrak{A} =\) the successor function \(S\): \(f^\mathfrak{A}(n) = n + 1\),
  • \(c^\mathfrak{A} = 0\).

Theorem 2. Assume that \(s_1\) and \(s_2\) are functions from \(V\) into \(|\mathfrak{A}|\) which agree at all variables that occur free in the wff \(\varphi\). Then \[\models_\mathfrak{A} \varphi[s_1] \iff \models_\mathfrak{A} \varphi[s_2]\]

Proof. (By induction on \(\varphi\).)

Theorem 3. If two structures \(\mathfrak{A}\) and \(\mathfrak{B}\) agree at all parameters that occur in \(\varphi\), then \[\models_\mathfrak{A} \varphi[s] \iff \models_\mathfrak{B} \varphi[s]\]

Corollary 4. For a sentence \(\sigma\), either

  1. \(\mathfrak{A}\) satisfies \(\sigma\) with every function \(s : V \to |\mathfrak{A}|\), or
  2. \(\mathfrak{A}\) does not satisfy \(\sigma\) with any such function.

Model and truth. If a structure \(\mathfrak{A}\) satisfies the sentence \(\sigma\) with every function \(s : V \to |\mathfrak{A}|\), we say that \(\sigma\) is true in \(\mathfrak{A}\) or that \(\mathfrak{A}\) is a model of \(\sigma\), written as \(\models_\mathfrak{A} \sigma\). Otherwise \(\sigma\) is false in \(\mathfrak{A}\). For a set \(\Sigma\) of sentences, \(\mathfrak{A}\) is a model of \(\Sigma\) iff it is a model of every \(\sigma \in \Sigma\).

Logical implication and validity. Let \(\Gamma\) be a set of wffs, \(\varphi\) be a wff. \(\Gamma\) logically implies \(\varphi\), written as \(\Gamma \models \varphi\), iff for every structure \(\mathfrak{A}\) for the language and every function \(s : V \to |\mathfrak{A}|\) such that \(\mathfrak{A}\) satisfies every member of \(\Gamma\) with \(s\), \(\mathfrak{A}\) also satisfies \(\varphi\) with \(s\).

We say that \(\varphi\) and \(\psi\) are logically equivalent iff \(\varphi \models \psi\) and \(\psi \models \varphi\).

A wff \(\varphi\) is valid iff \(\emptyset \models \varphi\), written simply as \(\models \varphi\). That is, \(\varphi\) is valid iff for every \(\mathfrak{A}\) and every \(s : V \to |\mathfrak{A}|\), \(\mathfrak{A}\) satisfies \(\varphi\) with \(s\).

The concept of validity is the first-order analogue of the concept of tautology in sentential logic.

Corollary 5. For a set \(\Sigma; \tau\) of sentences, \(\Sigma \models \tau\) iff every model of \(\Sigma\) is also a model of \(\tau\). A sentence \(\tau\) is valid iff it is true in every structure.