Proof Theory: MultiContext Systems

The starting point in the definition of a MultiContext Logic is a family of logical languages {Li}i in I defined over a set of indexes I (in the following we drop the index "i in I"). Intuitively, each Li is the (formal) language used to describe what is true in a context. For the purpose of our work we suppose that I is (at most) countable. We restrict also ourselves to to (classes of) first order languages as this is the case we studied most, but the same ideas apply to different logical languages as well.

Notationally, let us write i:A to mean A and that A is a formula of Li. We say that A is an Li-formula, and that i:A is a formula or, also, a labelled Li-formula. This notation and terminology allows us to keep track of the context we are talking about.

Given our family of languages {Li} over I, a MultiContext system (MC system) MS is a pair
\begin{displaymath}
	      MS = \langle \{T_i\}, \Delta_{br}\rangle
	      \end{displaymath}

where:
  • for each i in I, is an axiomatic formal system where Li is the language, is the set of axioms, and is the set of inference rules;
  • $\Delta_{br}$ is a set of inference rules with premises and conclusions in different languages.

An MC system is essentially a set of logical theories, plus a set of inference rules which allow for the propagation of consequences among theories. MC systems are a generalization of Natural Deduction (ND) systems  [prawitz1]. The generalization amounts to using formulae tagged with the language they belong to. This allows for the effective use of the multiple languages. The deduction machinery of an MC system is composed of two kinds of inference rules: the inference rules in each $\Delta_i$, called internal rules, and the inference rules in $\Delta_{br}$, called bridge rules. Internal rules are inference rules with premises and conclusions in the same language, while bridge rules are inference rules with premises and conclusions belonging to different languages. Notationally, inference rules are written as follows:

\begin{displaymath}
	      \begin{array}{c@{\hspace{3 cm}}c}
	      \infer[ir]
	      {i\hspace{-0.1c...
	      ...s \quad i_n\hspace{-0.1cm}:\hspace{-0.05cm}\phi_n }
	      \end{array}\end{displaymath}

where ir is an internal rule, while br is a bridge rule. Internal rules allow us to draw consequences inside a theory, while bridge rules allow us to export results from one theory to another. Indeed ir allows us to derive the formula C from the formulae A1, ..., An in the theory tagged with i, while br allows us to export the formula C to the theory tagged with j because of the fact that all the A1, ..., An are derivable in the theories tagged with i1, ..., in, respectively. $\Delta = \bigcup_{i
\in I}\Delta_i \cup \Delta_{br}$. is the deduction machinery of MS. Using ND and following [prawitz1] in the notation and terminology, the deduction machinery of MS contains also inference rules which discharge assumptions, written as:

\begin{displaymath}
	      \infer[dr]
	      {j\hspace{-0.1cm}:\hspace{-0.05cm}\psi }
	      {i_1\hsp...
	      ...i_{m+n}}
	      {[k_{m}\hspace{-0.1cm}:\hspace{-0.05cm}\gamma_{m}]}
	      }
	      \end{displaymath}

dr represents an inference rule which allow to infer j:C from A1, ..., An discharging the assumptions D1, ..., Dm

Figure 2 illustrates the contruction of a MC system with three languages and four bridge rules. We start from different languages, e.g., L1, L2, and L3. Then, we associate each of them with a logical theory Ti. Finally, we connect different logical theories with bridge rules, e.g., br1, br2, br3, and br4. The final result is an MC system.

Deductions in MC systems are trees of formulae built starting from a finite number of assumptions and axioms, possibly belonging to distinct languages, and by applying a finite number of inference rules. A formula i:A is derivable from a set of formulae B in a MC system MS, in symbols |-MS, if there is a deduction with bottom formula i:A whose un-discharged assumptions are in B. A formula i:A is a theorem in MS, in symbols |-MS if it is derivable from the empty set. Notice that a deduction in an MC system can be seen as composed of sub-deductions in distinct languages, obtained by repeated applications of internal rules, any two or more sub-deductions being concatenated by one or more applications of bridge rules.

Relevant Publications

L. Serafini and F. Giunchiglia
ML Systems: A Proof Theory for Contexts
Journal of Logic, Language and Information Spring 2002, Volume 11, Issue 2 pp. 471-518. (pdf)
F. Giunchiglia and L. Serafini
Multilanguage Hierarchical Logics (or: How we can do without modal logics)
Artificial Intelligence, 65:29-70, 1994. (ps)

 

Web mastering: Alessandro Tomasi
and Chiara Ghidini
Last Update: June 2003