VOOZH about

URL: https://en.wikipedia.org/wiki/Definitional_extension

โ‡ฑ Extension by definition - Wikipedia


Jump to content
From Wikipedia, the free encyclopedia
(Redirected from Definitional extension)
In logic, defining a new symbol

In mathematical logic, more specifically in the proof theory of first-order theories, an extension by definition formalizes the introduction of a new symbol by means of a definition. For example, it is common in naive set theory to introduce a symbol ๐Ÿ‘ {\displaystyle \emptyset }
for the set that has no member. In the formal setting of first-order theories, this can be done by adding to the theory a new constant ๐Ÿ‘ {\displaystyle \emptyset }
and the new axiom ๐Ÿ‘ {\displaystyle \forall x(x\notin \emptyset )}
, meaning "for all x, x is not a member of ๐Ÿ‘ {\displaystyle \emptyset }
". It can then be proved that doing so adds essentially nothing to the old theory, as should be expected from a definition. More precisely, the new theory is a conservative extension of the old one.

Definition of relation symbols

[edit]

Let ๐Ÿ‘ {\displaystyle T}
be a first-order theory and ๐Ÿ‘ {\displaystyle \phi (x_{1},\dots ,x_{n})}
a formula of ๐Ÿ‘ {\displaystyle T}
such that ๐Ÿ‘ {\displaystyle x_{1}}
, ..., ๐Ÿ‘ {\displaystyle x_{n}}
are distinct and include the variables free in ๐Ÿ‘ {\displaystyle \phi (x_{1},\dots ,x_{n})}
. Form a new first-order theory ๐Ÿ‘ {\displaystyle T'}
from ๐Ÿ‘ {\displaystyle T}
by adding a new ๐Ÿ‘ {\displaystyle n}
-ary relation symbol ๐Ÿ‘ {\displaystyle R}
, the logical axioms featuring the symbol ๐Ÿ‘ {\displaystyle R}
and the new axiom

๐Ÿ‘ {\displaystyle \forall x_{1}\dots \forall x_{n}(R(x_{1},\dots ,x_{n})\leftrightarrow \phi (x_{1},\dots ,x_{n}))}
,

called the defining axiom of ๐Ÿ‘ {\displaystyle R}
.

If ๐Ÿ‘ {\displaystyle \psi }
is a formula of ๐Ÿ‘ {\displaystyle T'}
, let ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
be the formula of ๐Ÿ‘ {\displaystyle T}
obtained from ๐Ÿ‘ {\displaystyle \psi }
by replacing any occurrence of ๐Ÿ‘ {\displaystyle R(t_{1},\dots ,t_{n})}
by ๐Ÿ‘ {\displaystyle \phi (t_{1},\dots ,t_{n})}
(changing the bound variables in ๐Ÿ‘ {\displaystyle \phi }
if necessary so that the variables occurring in the ๐Ÿ‘ {\displaystyle t_{i}}
are not bound in ๐Ÿ‘ {\displaystyle \phi (t_{1},\dots ,t_{n})}
). Then the following hold:

  1. ๐Ÿ‘ {\displaystyle \psi \leftrightarrow \psi ^{\ast }}
    is provable in ๐Ÿ‘ {\displaystyle T'}
    , and
  2. ๐Ÿ‘ {\displaystyle T'}
    is a conservative extension of ๐Ÿ‘ {\displaystyle T}
    .

The fact that ๐Ÿ‘ {\displaystyle T'}
is a conservative extension of ๐Ÿ‘ {\displaystyle T}
shows that the defining axiom of ๐Ÿ‘ {\displaystyle R}
cannot be used to prove new theorems. The formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
is called a translation of ๐Ÿ‘ {\displaystyle \psi }
into ๐Ÿ‘ {\displaystyle T}
. Semantically, the formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
has the same meaning as ๐Ÿ‘ {\displaystyle \psi }
, but the defined symbol ๐Ÿ‘ {\displaystyle R}
has been eliminated.

Definition of function symbols

[edit]

Let ๐Ÿ‘ {\displaystyle T}
be a first-order theory (with equality) and ๐Ÿ‘ {\displaystyle \phi (y,x_{1},\dots ,x_{n})}
a formula of ๐Ÿ‘ {\displaystyle T}
such that ๐Ÿ‘ {\displaystyle y}
, ๐Ÿ‘ {\displaystyle x_{1}}
, ..., ๐Ÿ‘ {\displaystyle x_{n}}
are distinct and include the variables free in ๐Ÿ‘ {\displaystyle \phi (y,x_{1},\dots ,x_{n})}
. Assume that we can prove

๐Ÿ‘ {\displaystyle \forall x_{1}\dots \forall x_{n}\exists !y\phi (y,x_{1},\dots ,x_{n})}

in ๐Ÿ‘ {\displaystyle T}
, i.e. for all ๐Ÿ‘ {\displaystyle x_{1}}
, ..., ๐Ÿ‘ {\displaystyle x_{n}}
, there exists a unique y such that ๐Ÿ‘ {\displaystyle \phi (y,x_{1},\dots ,x_{n})}
. Form a new first-order theory ๐Ÿ‘ {\displaystyle T'}
from ๐Ÿ‘ {\displaystyle T}
by adding a new ๐Ÿ‘ {\displaystyle n}
-ary function symbol ๐Ÿ‘ {\displaystyle f}
, the logical axioms featuring the symbol ๐Ÿ‘ {\displaystyle f}
and the new axiom

๐Ÿ‘ {\displaystyle \forall x_{1}\dots \forall x_{n}\phi (f(x_{1},\dots ,x_{n}),x_{1},\dots ,x_{n})}
,

called the defining axiom of ๐Ÿ‘ {\displaystyle f}
.

Let ๐Ÿ‘ {\displaystyle \psi }
be any atomic formula of ๐Ÿ‘ {\displaystyle T'}
. We define formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
of ๐Ÿ‘ {\displaystyle T}
recursively as follows. If the new symbol ๐Ÿ‘ {\displaystyle f}
does not occur in ๐Ÿ‘ {\displaystyle \psi }
, let ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
be ๐Ÿ‘ {\displaystyle \psi }
. Otherwise, choose an occurrence of ๐Ÿ‘ {\displaystyle f(t_{1},\dots ,t_{n})}
in ๐Ÿ‘ {\displaystyle \psi }
such that ๐Ÿ‘ {\displaystyle f}
does not occur in the terms ๐Ÿ‘ {\displaystyle t_{i}}
, and let ๐Ÿ‘ {\displaystyle \chi }
be obtained from ๐Ÿ‘ {\displaystyle \psi }
by replacing that occurrence by a new variable ๐Ÿ‘ {\displaystyle z}
. Then since ๐Ÿ‘ {\displaystyle f}
occurs in ๐Ÿ‘ {\displaystyle \chi }
one less time than in ๐Ÿ‘ {\displaystyle \psi }
, the formula ๐Ÿ‘ {\displaystyle \chi ^{\ast }}
has already been defined, and we let ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
be

๐Ÿ‘ {\displaystyle \forall z(\phi (z,t_{1},\dots ,t_{n})\rightarrow \chi ^{\ast })}

(changing the bound variables in ๐Ÿ‘ {\displaystyle \phi }
if necessary so that the variables occurring in the ๐Ÿ‘ {\displaystyle t_{i}}
are not bound in ๐Ÿ‘ {\displaystyle \phi (z,t_{1},\dots ,t_{n})}
). For a general formula ๐Ÿ‘ {\displaystyle \psi }
, the formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
is formed by replacing every occurrence of an atomic subformula ๐Ÿ‘ {\displaystyle \chi }
by ๐Ÿ‘ {\displaystyle \chi ^{\ast }}
. Then the following hold:

  1. ๐Ÿ‘ {\displaystyle \psi \leftrightarrow \psi ^{\ast }}
    is provable in ๐Ÿ‘ {\displaystyle T'}
    , and
  2. ๐Ÿ‘ {\displaystyle T'}
    is a conservative extension of ๐Ÿ‘ {\displaystyle T}
    .

The formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
is called a translation of ๐Ÿ‘ {\displaystyle \psi }
into ๐Ÿ‘ {\displaystyle T}
. As in the case of relation symbols, the formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
has the same meaning as ๐Ÿ‘ {\displaystyle \psi }
, but the new symbol ๐Ÿ‘ {\displaystyle f}
has been eliminated.

The construction of this paragraph also works for constants, which can be viewed as 0-ary function symbols.

Extensions by definitions

[edit]

A first-order theory ๐Ÿ‘ {\displaystyle T'}
obtained from ๐Ÿ‘ {\displaystyle T}
by successive introductions of relation symbols and function symbols as above is called an extension by definitions of ๐Ÿ‘ {\displaystyle T}
. Then ๐Ÿ‘ {\displaystyle T'}
is a conservative extension of ๐Ÿ‘ {\displaystyle T}
, and for any formula ๐Ÿ‘ {\displaystyle \psi }
of ๐Ÿ‘ {\displaystyle T'}
we can form a formula ๐Ÿ‘ {\displaystyle \psi ^{\ast }}
of ๐Ÿ‘ {\displaystyle T}
, called a translation of ๐Ÿ‘ {\displaystyle \psi }
into ๐Ÿ‘ {\displaystyle T}
, such that ๐Ÿ‘ {\displaystyle \psi \leftrightarrow \psi ^{\ast }}
is provable in ๐Ÿ‘ {\displaystyle T'}
. Such a formula is not unique, but any two of them can be proved to be equivalent in T.

In practice, an extension by definitions ๐Ÿ‘ {\displaystyle T'}
of T is not distinguished from the original theory T. In fact, the formulas of ๐Ÿ‘ {\displaystyle T'}
can be thought of as abbreviating their translations into T. The manipulation of these abbreviations as actual formulas is then justified by the fact that extensions by definitions are conservative.

Examples

[edit]
๐Ÿ‘ {\displaystyle \forall x(x\times e=x\land e\times x=x)}
,
and what we obtain is an extension by definitions ๐Ÿ‘ {\displaystyle T'}
of ๐Ÿ‘ {\displaystyle T}
. Then in ๐Ÿ‘ {\displaystyle T'}
we can prove that for every x, there exists a unique y such that x ร— y = y ร— x = e. Consequently, the first-order theory ๐Ÿ‘ {\displaystyle T''}
obtained from ๐Ÿ‘ {\displaystyle T'}
by adding a unary function symbol ๐Ÿ‘ {\displaystyle f}
and the axiom
๐Ÿ‘ {\displaystyle \forall x(x\times f(x)=e\land f(x)\times x=e)}
is an extension by definitions of ๐Ÿ‘ {\displaystyle T}
. Usually, ๐Ÿ‘ {\displaystyle f(x)}
is denoted ๐Ÿ‘ {\displaystyle x^{-1}}
.

See also

[edit]

Bibliography

[edit]
  • S. C. Kleene (1952), Introduction to Metamathematics, D. Van Nostrand
  • E. Mendelson (1997). Introduction to Mathematical Logic (4th ed.), Chapman & Hall.
  • J. R. Shoenfield (1967). Mathematical Logic, Addison-Wesley Publishing Company (reprinted in 2001 by AK Peters)