## Barendregt’s variable convention in rule inductions (2007)

### Cached

### Download Links

Venue: | In Proc. of the 21th International Conference on Automated Deduction (CADE), volume 4603 of LNAI |

Citations: | 20 - 8 self |

### BibTeX

@INPROCEEDINGS{Urban07barendregt’svariable,

author = {Christian Urban and Stefan Berghofer and Michael Norrish},

title = {Barendregt’s variable convention in rule inductions},

booktitle = {In Proc. of the 21th International Conference on Automated Deduction (CADE), volume 4603 of LNAI},

year = {2007},

pages = {35--50},

publisher = {Springer}

}

### OpenURL

### Abstract

Abstract. Inductive definitions and rule inductions are two fundamental reasoning tools in logic and computer science. When inductive definitions involve binders, then Barendregt's variable convention is nearly always employed (explicitly or implicitly) in order to obtain simple proofs. Using this convention, one does not consider truly arbitrary bound names, as required by the rule induction principle, but rather bound names about which various freshness assumptions are made. Unfortunately, neither Barendregt nor others give a formal justification for the variable convention, which makes it hard to formalise such proofs. In this paper we identify conditions an inductive definition has to satisfy so that a form of the variable convention can be built into the rule induction principle. In practice this means we come quite close to the informal reasoning of "pencil-and-paper " proofs, while remaining completely formal. Our conditions also reveal circumstances in which Barendregt's variable convention is not applicable, and can even lead to faulty reasoning. 1 Introduction In informal proofs about languages that feature bound variables, one often assumes (explicitly or implicitly) a rather convenient convention about those bound variables. Barendregt's statement of the convention is: Variable Convention: If M1; : : : ; Mn occur in a certain mathematical context (e.g. definition, proof), then in these terms all bound variables are chosen to be different from the free variables. [2, Page 26]

### Citations

704 |
Types and Programming Languages
- Pierce
- 2002
(Show Context)
Citation Context ...other variables of t are either in as or xs. That means we have for t[as; xs] that varsbp(t) ⊆ as and vars(t) ⊆ as ∪ xs hold. We extend this notation also to schematic rules: by writing ϱ[as; xs] for =-=(8)-=- we mean R ts1[as; xs] . . . R tsn[as; xs] S1 ss1[as; xs] . . . Sm ssm[as; xs] R ts[as; xs] However, unlike in the notation for schematic terms, we mean in ϱ[as; xs] that the as stand exactly for the ... |

286 |
The Lambda-Calculus, its Syntax and Semantics, volume 103
- Barendregt
- 1984
(Show Context)
Citation Context ...→ 1 M ′ [y := N ′ ] One4 (1) The substitutivity property states:s2 Christian Urban, Stefan Berghofer, and Michael Norrish Lemma. If M −→ 1 M ′ and N −→ 1 N ′ , then M[x := N] −→ 1 M ′ [x := N ′ ]. In =-=[2]-=-, the proof of this lemma proceeds by an induction over the definition of M −→ 1 M ′ . Though Barendregt does not acknowledge the fact explicitly, there are two places in his proof where the variable ... |

162 | Nominal Logic, A first order theory of names and binding
- Pitts
(Show Context)
Citation Context ...hapter on logical relation [4], and in various formalised proofs on structural operational semantics. 2 Nominal Logic Before proceeding, we briefly introduce some important notions from nominal logic =-=[9,12]-=-. In particular, we will build on the three central notions of permutations, support and equivariance. Permutations are finite bijective mappings from atoms to atoms, where atoms are drawn from a coun... |

151 |
Logic for Computer Science. Foundation of Automatic Theorem Proving
- Gallier
- 1986
(Show Context)
Citation Context ... In the POPLmark Challenge [1], the proof of the weakening lemma is described as a “straightforward induction”. In fact, mechanising this informal proof is not straightforward at all (see for example =-=[6,5,9]-=-). We have given a novel rule induction principle for the typing relation that makes proving the weakening lemma mechanically as simple as performing the informal proof. Importantly, this new principl... |

137 | Mechanized metatheory for the masses: The PoplMark challenge
- Aydemir, Bohannon, et al.
- 2005
(Show Context)
Citation Context ...dy deals with this situation by explicitly giving the information that x should appear in the induction context. However, we have not yet worked out the theory. 8 Conclusion In the POPLmark Challenge =-=[1]-=-, the proof of the weakening lemma is described as a “straightforward induction”. In fact, mechanising this informal proof is not straightforward at all (see for example [6,5,9]). We have given a nove... |

82 | Nominal techniques in Isabelle/HOL
- Urban
(Show Context)
Citation Context ...uction principles that come with the inductive definitions. Moreover, it will be shown how these new (“vccompatible”) induction principles can be automatically derived in the nominal datatype package =-=[12,10]-=-. The presented results have already been extensively used in formalisations: for example in our formalisations of the CR and SN properties in the λ-calculus, in a formalisation by Bengtson and Parrow... |

81 | Isar — a generic interpretative approach to readable formal proof documents
- Wenzel
- 1999
(Show Context)
Citation Context ...s. Now applying rule Type3 we can obtain Γ ′ ⊢ lam(x.M) : T1 → T2, again using the variable convention’s x # Γ ′ . This completes the proof. Its readable version expressed in Isabelle’s Isar-language =-=[13]-=- and using the nominal datatype package [10] is shown in Fig. 1. By way of contrast, recall that a proof without the stronger induction principle would not be able to assume anything about the relatio... |

18 | A recursion combinator for nominal datatypes implemented in Isabelle/HOL
- Urban, Berghofer
(Show Context)
Citation Context ...uction principles that come with the inductive definitions. Moreover, it will be shown how these new (“vccompatible”) induction principles can be automatically derived in the nominal datatype package =-=[12,10]-=-. The presented results have already been extensively used in formalisations: for example in our formalisations of the CR and SN properties in the λ-calculus, in a formalisation by Bengtson and Parrow... |

14 |
Formalising the pi-calculus using nominal logic
- Bengtson, Parrow
- 2008
(Show Context)
Citation Context ...en extensively used in formalisations: for example in our formalisations of the CR and SN properties in the λ-calculus, in a formalisation by Bengtson and Parrow for several proofs in the pi-calculus =-=[3]-=-, in a formalisation of Crary’s chapter on logical relation [4], and in various formalised proofs on structural operational semantics. 2 Nominal Logic Before proceeding, we briefly introduce some impo... |

12 |
Some type theory and lambda calculus formalised
- McKinna, Pollack
- 1999
(Show Context)
Citation Context ...ion principles once and for all. Proofs using the vc-compatible principles then do not need to perform any explicit renaming steps. Somewhat similar to our approach is the work of Pollack and McKinna =-=[6]-=-. Starting from the standard induction principle that is associated with an inductive definition, we derived an induction principle that allows emulation of Barendregt’s variable convention. Pollack a... |

12 | A formal treatment of the Barendregt Variable Convention in rule inductions
- Urban, Norrish
- 2005
(Show Context)
Citation Context ... induction relying on the substitution lemma, and the lemma fresh-atm, which states that x # y is the same as x �= y when y is an atom. 6 Related Work Apart from our own preliminary work in this area =-=[11]-=-, we believe the prettiest formal proof of the weakening lemma to be that in Pitts [9]. This proof uses the equivariance property of the typing relation, and includes a renaming step using permutation... |

6 |
Advanced Topics in Types and Programming Languages, chapter Logical Relations and a Case Study in Equivalence Checking
- Crary
- 2005
(Show Context)
Citation Context ...lisations of the CR and SN properties in the λ-calculus, in a formalisation by Bengtson and Parrow for several proofs in the pi-calculus [3], in a formalisation of Crary’s chapter on logical relation =-=[4]-=-, and in various formalised proofs on structural operational semantics. 2 Nominal Logic Before proceeding, we briefly introduce some important notions from nominal logic [9,12]. In particular, we will... |

3 |
Mechanising λ-calculus using a classical first order theory of terms with permutations
- Norrish
- 2006
(Show Context)
Citation Context ...s performing the informal proof. Importantly, this new principle can be derived from the original inductive definition of the typing relation in a mechanical way. This method extends our earlier work =-=[11,7]-=-, where we constructed our new induction principles by hand. By formally deriving principles that avoid the need to rename bound variables, we advance the state-of-the-art in mechanical theorem-provin... |