## Propositional Functions and Families of Types (1989)

Venue: | In Workshop on Programming Logic |

Citations: | 1 - 0 self |

### BibTeX

@INPROCEEDINGS{Smith89propositionalfunctions,

author = {Jan M. Smith},

title = {Propositional Functions and Families of Types},

booktitle = {In Workshop on Programming Logic},

year = {1989},

pages = {140--159}

}

### OpenURL

### Abstract

Introduction In order to capture some of the programmers errors, several computer languages, like Pascal and ML, are equipped with a type system. Using the Curry-Howard interpretation of propositions as types [3, 8], or as we shall say here, propositions as sets, a type system can be made strong enough to be used to specify the task a program is supposed to do. This is one of the basis for Martin-Lof's suggestion in [11] to use his formulation of type theory for programming; his ideas are exploited in [14] and there are several computer implementations of type theory [4, 16]. Similar ideas are also behind Coquand and Huet's calculus of constructions [2]. The idea of propositions as sets is closely related to the intuitionistic explanations of the logical constants given by Heyting [7]. In Martin-Lof's type theory, the interpretation of propositions as sets is fundamental since the notions of proposition and set are identical. So a logical constant is definitionally equal to th

### Citations

696 | A Framework for Defining Logics
- Harper, Honsel, et al.
- 1993
(Show Context)
Citation Context ...declarations. This is important 5 when building a computer system since you then do not want to make major changes of the implementation when introducing a new set forming operation. The Edinburgh LF =-=[6]-=- is based on similar ideas. The type level introduced by Martin-Lof has judgements of the forms 6 ff is a type, ff and fi are equal types, a is an object of the type ff , and a and b are equal objects... |

471 |
The calculus of constructions
- Coquand, Huet
- 1988
(Show Context)
Citation Context ...eory for programming; his ideas are exploited in [14] and there are several computer implementations of type theory [4, 16]. Similar ideas are also behind Coquand and Huet's calculus of constructions =-=[2]-=-. The idea of propositions as sets is closely related to the intuitionistic explanations of the logical constants given by Heyting [7]. In Martin-Lof's type theory, the interpretation of propositions ... |

441 |
The formulae-as-types notion of construction
- Howard
- 1980
(Show Context)
Citation Context ...ion In order to capture some of the programmers errors, several computer languages, like Pascal and ML, are equipped with a type system. Using the Curry-Howard interpretation of propositions as types =-=[3, 8]-=-, or as we shall say here, propositions as sets, a type system can be made strong enough to be used to specify the task a program is supposed to do. This is one of the basis for Martin-Lof's suggestio... |

342 |
Intuitionistic type theory
- Martin-Löf
- 1984
(Show Context)
Citation Context ...framework, we obtain a monomorphic theory. We may define a stripping function on the expressions which takes away the set information and we would then obtain expressions of the polymorphic theory in =-=[11, 13]-=-. However, the polymorphic theory is fundamentally different from the monomorphic theory; in Salvesen [18] it is shown that there are derivable judgements in the polymorphic theory which cannot come f... |

265 |
Constructive mathematics and computer programming
- Martin-Löf
- 1982
(Show Context)
Citation Context ...as we shall say here, propositions as sets, a type system can be made strong enough to be used to specify the task a program is supposed to do. This is one of the basis for Martin-Lof's suggestion in =-=[11]-=- to use his formulation of type theory for programming; his ideas are exploited in [14] and there are several computer implementations of type theory [4, 16]. Similar ideas are also behind Coquand and... |

261 |
Programming in Martin-Löf’s Type Theory: An Introduction
- Nordstrom, Petersson, et al.
- 1990
(Show Context)
Citation Context ...be used to specify the task a program is supposed to do. This is one of the basis for Martin-Lof's suggestion in [11] to use his formulation of type theory for programming; his ideas are exploited in =-=[14]-=- and there are several computer implementations of type theory [4, 16]. Similar ideas are also behind Coquand and Huet's calculus of constructions [2]. The idea of propositions as sets is closely rela... |

157 |
An intuitionistic theory of types: Predicative part
- Martin-Löf
- 1975
(Show Context)
Citation Context ...lf. Such a universe would have been very practical to use but, by Girard's paradox, V 2 V implies that all sets are non-empty; hence, it is impossible to interpret propositions as sets. In Martin-Lof =-=[12]-=-, the universe V is replaced by a series of universes U 0 ; U 1 ; : : : where U 0 is the set of small sets and U n 2 U n+1 . Following the semantics in Martin-Lof [11], where a set is defined by presc... |

77 |
An introduction
- Intuitionism
- 1956
(Show Context)
Citation Context ...deas are also behind Coquand and Huet's calculus of constructions [2]. The idea of propositions as sets is closely related to the intuitionistic explanations of the logical constants given by Heyting =-=[7]-=-. In Martin-Lof's type theory, the interpretation of propositions as sets is fundamental since the notions of proposition and set are identical. So a logical constant is definitionally equal to the co... |

66 | An Intuitionstic Theory of Types
- Martin-Löf
- 1995
(Show Context)
Citation Context ...f lists of elements in a set A . However, many specifications can still not be expressed with these sets but require a universe. 3 The need of a universe Martin-Lof's first formulation of type theory =-=[10]-=- contained a universe V in which all sets were elements, including V itself. Such a universe would have been very practical to use but, by Girard's paradox, V 2 V implies that all sets are non-empty; ... |

46 | A natural extension of natural deduction
- Schroeder-Heister
- 1984
(Show Context)
Citation Context ...n particular, to define propositional functions by recursion without using a universe. It is then important that the elimination rules are formulated in the general way suggested by Schroeder-Heister =-=[19, 20]-=-. I will first briefly describe, following [15], how sets in Martin-Lof's type theory can be viewed as specifications and then why a universe sometimes must be used when expressing propositions. A pre... |

30 |
Zur Deutung der Intuitionistischen Logik
- Kolmogorov
- 1932
(Show Context)
Citation Context ...s as sets The idea of of viewing a specification of computer programs as a set in Martin-Lof's type theory has its origin both in understanding propositions as sets and in Kolmogorov's explanation in =-=[9]-=- of propositions as problems. Kolmogorov explains the sentential constants in the following way: AsB is the problem of solving both of the problems A and the problem B . AsB is the problem of solving ... |

17 | The independence of Peano’s fourth axiom from Martin-Lof’s type theory without universes
- Smith
- 1988
(Show Context)
Citation Context ...t of natural numbers. Assuming Id(N; 0; 1) it is easy, using tt 2 F (1) , to show that F (0) is nonempty. Since F (0) = ; we then obtain Id(N; 0; 1) ! ; , i.e., by definition, :Id(N; 0; 1) . In Smith =-=[22]-=- it is shown that in type theory without a universe, no negated equalities at all can be proved. 4 The logical framework The main reason to introduce a type level, more basic than the level of sets, i... |

15 |
A programming system for type theory
- Petersson
- 1982
(Show Context)
Citation Context ...e of the basis for Martin-Lof's suggestion in [11] to use his formulation of type theory for programming; his ideas are exploited in [14] and there are several computer implementations of type theory =-=[4, 16]-=-. Similar ideas are also behind Coquand and Huet's calculus of constructions [2]. The idea of propositions as sets is closely related to the intuitionistic explanations of the logical constants given ... |

13 |
An interpretation of Martin-Lof's type theory in a type-free theory of propositions
- Smith
- 1984
(Show Context)
Citation Context ...rtin-Lof [11], where a set is defined by prescribing how the canonical elements are formed, it is natural to view an element in a universe as a code for the corresponding set; this is the approach in =-=[14, 21]-=- and will be used here. 3 I will in this section give two examples how one is forced to use a universe in two basic applications; the first is when defining a simple proposition by induction and the s... |

9 |
The strength of Martin-Löf’s type theory with one universe
- Aczel
- 1977
(Show Context)
Citation Context ...s I will in this section sketch an interpretation of set theory with type valued recursion but without a universe into set theory with a universe but without type valued recursion. Aczel has shown in =-=[1]-=- (see also [5]) that the proof theoretic ordinal of Martin-Lof's type theory with a universe is, in Veblen's notation, OE ffl 0 (0) . So we will then get an upper limit on the strength of set theory e... |

3 | Polymorphism and monomorphism in Martin�L�of�s Type Theory
- Salvesen
- 1988
(Show Context)
Citation Context ...way the set information and we would then obtain expressions of the polymorphic theory in [11, 13]. However, the polymorphic theory is fundamentally different from the monomorphic theory; in Salvesen =-=[18]-=- it is shown that there are derivable judgements in the polymorphic theory which cannot come from any derivable judgement in the monomorphic theory by stripping. The cartesian product on a family of s... |

2 |
Deriving Rules for Inductive Sets in Martin-Lof's Type Theory
- Synek
- 1989
(Show Context)
Citation Context ...El(A))El(B(x)) ((y)y) : (x : El(A))El(B(x)) So apply \Pi(A;B) is defined by just applying Funsplit (x : El(A))El(B(x)) on the identity function of ((x : El(A))El(B(x))) (x : El(A))El(B(x)) . In Synek =-=[23]-=- type valued recursion is used when defining a set constructor for mutual recursive sets in terms of well-orderings. In this application, type valued recursion is crucial since otherwise, using a univ... |

1 |
Iterated Inductive, Fixed Point Theories
- Feferman
- 1982
(Show Context)
Citation Context ...is section sketch an interpretation of set theory with type valued recursion but without a universe into set theory with a universe but without type valued recursion. Aczel has shown in [1] (see also =-=[5]-=-) that the proof theoretic ordinal of Martin-Lof's type theory with a universe is, in Veblen's notation, OE ffl 0 (0) . So we will then get an upper limit on the strength of set theory extended with t... |

1 |
A chapter on Martin-Lof's type theory in vol
- Petersson, Smith
(Show Context)
Citation Context ...are formulated in a natural deduction style, but we will here not give the general rules concerned with substitution, equality and handling of contexts; the semantics and rules are given in detail in =-=[17]-=-. If we have a type ff and a family fi of types on ff , then we can form the dependent function type from ff to fi : Fun-formation ff : type fi : type [x : ff] (x : ff)fi : type Functions are introduc... |

1 |
Generalized Rules for Operators and the Completeness of the Intuitionistic Operators
- Schroeder-Heister
- 1984
(Show Context)
Citation Context ...n particular, to define propositional functions by recursion without using a universe. It is then important that the elimination rules are formulated in the general way suggested by Schroeder-Heister =-=[19, 20]-=-. I will first briefly describe, following [15], how sets in Martin-Lof's type theory can be viewed as specifications and then why a universe sometimes must be used when expressing propositions. A pre... |