Abstract:We provide a semi-grammatical description of the set of normal proofs of positive formulae in minimal predicate logic, i.e. a grammar that generates a set of schemes, from each of which we can produce a finite number of normal proofs. This method is complete in the sense that each normal proof-term of the formula is produced by some scheme generated by the grammar. As a corollary, we get a similar description of the set of normal proofs of positive formulae for a large class of theories including simple type theory and System F.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: in minimal predicate logic, to provide a semi - grammatical description of normal proofs of positive formulae. Specifically, the author aims to generate a series of schemes by defining a context - free grammar, from which a finite number of normal proofs can be produced. This method is complete, that is, the normal proof terms of each formula can be generated by some scheme generated by the grammar.
### Background and Motivation of the Paper
1. **Decidability Problem**
- In some logical systems, such as certain formulations of classical propositional sequent calculus, the provability of a formula can be verified by enumerating all possible proofs, and when the formula is provable, it has only a finite number of proofs.
- For some formulations of intuitionistic or minimal propositional sequent calculus, the search tree is infinite but regular, which means that although the set of proofs may be infinite, we can still describe it with a context - free grammar.
2. **Zaionc's Results**
- Zaionc has proved that the set of normal proof terms in minimal propositional logic (that is, the set of normal terms in simply - typed lambda calculus of a given type) is not a context - free language. This explains why previous grammatical descriptions of the set of normal terms of a given type require an infinite number of symbols.
3. **Improved Methods**
- Takahashi, Akama and Hirokawa, as well as Broda and Damas, have proposed that if a pattern language is used to represent proof terms, in which the same assumptions are referred to by the same names, the set of proof terms in minimal propositional logic can become a context - free language. Each pattern proof term corresponds to a finite number of real proof terms.
4. **Research Motivation**
- There are two main reasons for studying the positive fragment of minimal logic:
- In classical logic, undecidability mainly comes from negative quantifiers, and the positive fragment is decidable.
- In System F, data types are represented as positive types, such as unary natural number types and binary natural number types. Although some positive types are not data types, we still hope to describe the set of normal terms of these types because they are used in higher - order abstract syntax or as input types of algorithms extracted from constructive completeness theorem proofs.
### Main Contributions
- The author has extended previous work and given a semi - grammatical description of the set of β - normal η - long proof terms in the positive fragment of minimal predicate logic.
- Provided a method of regular search tree based on a parenthesis system, ensuring the decidability of the positive fragment.
- This method is applicable not only to minimal predicate logic, but also to other theories, such as simple type theory and System F.
### Conclusion
By this method, the author has successfully solved the enumeration problem of normal proof terms in the positive fragment of minimal predicate logic and provided a similar description for positive types in System F. This result is helpful for better understanding and handling the proof structures in these logical systems.