Translations of intuitionistic logic
(Definitions of three translations) |
m (→Call-by-name Girard's translation A\imp B \mapsto \oc{A}\limp B: use of \rulename) |
||
Line 25: | Line 25: | ||
<math> |
<math> |
||
− | \LabelRule{\textit{ax}} |
+ | \LabelRule{\rulename{ax}} |
\NulRule{A\vdash A} |
\NulRule{A\vdash A} |
||
\DisplayProof |
\DisplayProof |
||
\qquad\mapsto\qquad |
\qquad\mapsto\qquad |
||
− | \LabelRule{\textit{ax}} |
+ | \LabelRule{\rulename{ax}} |
\NulRule{A^n\vdash A^n} |
\NulRule{A^n\vdash A^n} |
||
\LabelRule{\oc L} |
\LabelRule{\oc L} |
||
Line 41: | Line 41: | ||
\AxRule{\Gamma\vdash A} |
\AxRule{\Gamma\vdash A} |
||
\AxRule{\Delta,A\vdash B} |
\AxRule{\Delta,A\vdash B} |
||
− | \LabelRule{\textit{cut}} |
+ | \LabelRule{\rulename{cut}} |
\BinRule{\Gamma,\Delta\vdash B} |
\BinRule{\Gamma,\Delta\vdash B} |
||
\DisplayProof |
\DisplayProof |
||
Line 49: | Line 49: | ||
\UnaRule{\oc{\Gamma^n}\vdash \oc{A^n}} |
\UnaRule{\oc{\Gamma^n}\vdash \oc{A^n}} |
||
\AxRule{\oc{\Delta^n},\oc{A^n}\vdash B^n} |
\AxRule{\oc{\Delta^n},\oc{A^n}\vdash B^n} |
||
− | \LabelRule{\textit{cut}} |
+ | \LabelRule{\rulename{cut}} |
\BinRule{\oc{\Gamma^n},\oc{\Delta^n}\vdash B^n} |
\BinRule{\oc{\Gamma^n},\oc{\Delta^n}\vdash B^n} |
||
\DisplayProof |
\DisplayProof |
||
Line 108: | Line 108: | ||
\LabelRule{\oc R} |
\LabelRule{\oc R} |
||
\UnaRule{\oc{\Gamma^n}\vdash \oc{A^n}} |
\UnaRule{\oc{\Gamma^n}\vdash \oc{A^n}} |
||
− | \LabelRule{\textit{ax}} |
+ | \LabelRule{\rulename{ax}} |
\NulRule{B^n\vdash B^n} |
\NulRule{B^n\vdash B^n} |
||
\LabelRule{\limp L} |
\LabelRule{\limp L} |
||
Line 117: | Line 117: | ||
\UnaRule{\oc{\Gamma^n},\oc{(\oc{A^n}\limp B^n)}\vdash \oc{B^n}} |
\UnaRule{\oc{\Gamma^n},\oc{(\oc{A^n}\limp B^n)}\vdash \oc{B^n}} |
||
\AxRule{\oc{\Delta^n},\oc{B^n}\vdash C^n} |
\AxRule{\oc{\Delta^n},\oc{B^n}\vdash C^n} |
||
− | \LabelRule{\textit{cut}} |
+ | \LabelRule{\rulename{cut}} |
\BinRule{\oc{\Gamma^n},\oc{\Delta^n},\oc{(\oc{A^n}\limp B^n)}\vdash C^n} |
\BinRule{\oc{\Gamma^n},\oc{\Delta^n},\oc{(\oc{A^n}\limp B^n)}\vdash C^n} |
||
\DisplayProof |
\DisplayProof |
||
Line 146: | Line 146: | ||
\DisplayProof |
\DisplayProof |
||
\qquad\mapsto\qquad |
\qquad\mapsto\qquad |
||
− | \LabelRule{\textit{ax}} |
+ | \LabelRule{\rulename{ax}} |
\NulRule{A^n\vdash A^n} |
\NulRule{A^n\vdash A^n} |
||
\LabelRule{\with_1 L} |
\LabelRule{\with_1 L} |
||
Line 155: | Line 155: | ||
\UnaRule{\oc{(A^n\with B^n)}\vdash \oc{A^n}} |
\UnaRule{\oc{(A^n\with B^n)}\vdash \oc{A^n}} |
||
\AxRule{\oc{\Gamma^n},\oc{A^n}\vdash C^n} |
\AxRule{\oc{\Gamma^n},\oc{A^n}\vdash C^n} |
||
− | \LabelRule{\textit{cut}} |
+ | \LabelRule{\rulename{cut}} |
\BinRule{\oc{\Gamma^n},\oc{(A^n\with B^n)}\vdash C^n} |
\BinRule{\oc{\Gamma^n},\oc{(A^n\with B^n)}\vdash C^n} |
||
\DisplayProof |
\DisplayProof |
||
Line 242: | Line 242: | ||
\DisplayProof |
\DisplayProof |
||
\qquad\mapsto\qquad |
\qquad\mapsto\qquad |
||
− | \LabelRule{\textit{ax}} |
+ | \LabelRule{\rulename{ax}} |
\NulRule{A^n[\tau^n/\xi]\vdash A^n[\tau^n/\xi]} |
\NulRule{A^n[\tau^n/\xi]\vdash A^n[\tau^n/\xi]} |
||
\LabelRule{\forall L} |
\LabelRule{\forall L} |
||
Line 251: | Line 251: | ||
\UnaRule{\oc{\forall\xi A^n}\vdash \oc{(A^n[\tau^n/\xi])}} |
\UnaRule{\oc{\forall\xi A^n}\vdash \oc{(A^n[\tau^n/\xi])}} |
||
\AxRule{\oc{\Gamma^n},\oc{(A^n[\tau^n/\xi])}\vdash C^n} |
\AxRule{\oc{\Gamma^n},\oc{(A^n[\tau^n/\xi])}\vdash C^n} |
||
− | \LabelRule{\textit{cut}} |
+ | \LabelRule{\rulename{cut}} |
\BinRule{\oc{\Gamma^n},\oc{\forall\xi A^n}\vdash C^n} |
\BinRule{\oc{\Gamma^n},\oc{\forall\xi A^n}\vdash C^n} |
||
\DisplayProof |
\DisplayProof |
||
Line 287: | Line 287: | ||
\DisplayProof |
\DisplayProof |
||
</math> |
</math> |
||
− | |||
== Call-by-value translation <math>A\imp B \mapsto \oc{(A\limp B)}</math> == |
== Call-by-value translation <math>A\imp B \mapsto \oc{(A\limp B)}</math> == |
Revision as of 23:15, 12 February 2009
The genesis of linear logic comes with a decomposition of the intuitionistic implication. Once linear logic properly defined, it corresponds to a translation of intuitionistic logic into linear logic, often called Girard's translation. In fact Jean-Yves Girard has defined two translations in his linear logic paper[1]. We call them the call-by-name translation and the call-by-value translation.
These translations can be extended to translations of classical logic into linear logic.
Contents |
Call-by-name Girard's translation
Formulas are translated as:
This is extended to sequents by .
This allows one to translate the rules of intuitionistic logic into linear logic:
Call-by-value translation
Formulas are translated as:
The translation of any formula starts with , we define such that .
The translation of sequents is .
This allows one to translate the rules of intuitionistic logic into linear logic:
We use .
Alternative presentation
It is also possible to define as the primitive construction.
If we define , we have and thus we obtain the same translation of proofs.
Call-by-value Girard's translation
The original version of the call-by-value translation given by Jean-Yves Girard[1] is an optimisation of the previous one using properties of positive formulas.
Formulas are translated as:
The translation of any formula is a positive formula.
The translation of sequents is .
This allows one to translate the rules of intuitionistic logic into linear logic:
We use .
References
- ↑ 1.0 1.1 Girard, Jean-Yves. Linear logic. Theoretical Computer Science. Volume 50, Issue 1, pp. 1-101, doi:10.1016/0304-3975(87)90045-4, 1987.