Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

Original proof of
Gdel's completeness
theorem
From Wikipedia, the free encyclopedia

The proof of Gdel's completeness theorem given

by Kurt Gdel in his doctoral dissertation of 1929
(and a rewritten version of the dissertation,
published as an article in 1930) is not easy to read
today; it uses concepts and formalism that are no
longer used and terminology that is often obscure.
The version given below attempts to represent all
the steps in the proof and all the important ideas
faithfully, while restating the proof in the modern
language of mathematical logic. This outline
should not be considered a rigorous proof of the
theorem.

Contents
1 Denitions and assumptions
2 Statement of the theorem and its proof
2.1 Theorem 1. Every valid formula
(true in all structures) is provable.
2.2 Theorem 2. Every formula is
either refutable or satisable in
some structure.

1 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

2.3 Equivalence of both theorems

2.4 Proof of theorem 2: rst step
2.5 Reducing the theorem to formulas
of degree 1
2.6 Proving the theorem for formulas
of degree 1
2.6.1 Intuitive explanation
3 Extensions
3.1 Extension to rst-order predicate
calculus with equality
3.2 Extension to countable sets of
formulas
3.3 Extension to arbitrary sets of
formulas
4 References
5 External links

Denitions and assumptions

We work with rst-order predicate calculus. Our
languages allow constant, function and relation
symbols. Structures consist of (non-empty)
domains and interpretations of the relevant
symbols as constant members, functions or
relations over that domain.
We x some axiomatization of the predicate
calculus: logical axioms and rules of inference.
Any of the several well-known axiomatisations will
do; we assume without proof all the basic

2 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

well-known results about our formalism (such as

the normal form theorem or the soundness
theorem) that we need.
We axiomatize predicate calculus without equality,
i.e. there are no special axioms expressing the
properties of equality as a special relation symbol.
After the basic form of the theorem is proved, it
will be easy to extend it to the case of predicate
calculus with equality.

Statement of the theorem and

its proof
In the following, we state two equivalent forms of
the theorem, and show their equivalence.
Later, we prove the theorem. This is done in the
following steps:
1. Reducing the theorem to sentences (formulas
with no free variables) in prenex form, i.e.
with all quantiers ( and ) at the
beginning. Furthermore, we reduce it to
formulas whose rst quantier is . This is
possible because for every sentence, there is
an equivalent one in prenex form whose rst
quantier is .
2. Reducing the theorem to sentences of the
form
x1x2...xk y1y2...ym (x1...xk, y1...ym).

3 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

While we cannot do this by simply

rearranging the quantiers, we show that it
is yet enough to prove the theorem for
sentences of that form.
3. Finally we prove the theorem for sentences of
that form.
This is done by rst noting that a
sentence such as

B = x1x2...xk y1y2...ym (x1...xk, y1...y

is either refutable or has some model in
which it holds; this model is simply
assigning truth values to the
subpropositions from which B is built.
The reason for that is the completeness
of propositional logic, with the
existential quantiers playing no role.
We extend this result to more and more
complex and lengthy sentences, Dn
(n=1,2...), built out from B, so that
either any of them is refutable and
therefore so is , or all of them are not
refutable and therefore each holds in
some model.
We nally use the models in which the
Dn hold (in case all are not refutable) in
order to build a model in which holds.

Theorem 1. Every valid formula (true

in all structures) is provable.
This is the most basic form of the completeness
theorem. We immediately restate it in a form more
4 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

convenient for our purposes:

Theorem 2. Every formula is either

refutable or satisable in some
structure.
" is refutable" means by denition " is
provable".

Equivalence of both theorems

To see the equivalence, note rst that if Theorem
1 holds, and is not satisable in any structure,
then is valid in all structures and therefore
provable, thus is refutable and Theorem 2
holds. If on the other hand Theorem 2 holds and
is valid in all structures, then is not
satisable in any structure and therefore
refutable; then is provable and then so is ,
thus Theorem 1 holds.

Proof of theorem 2: rst step

We approach the proof of Theorem 2 by
successively restricting the class of all formulas
for which we need to prove " is either refutable
or satisable". At the beginning we need to prove
this for all possible formulas in our language.
However, suppose that for every formula there
is some formula taken from a more restricted
class of formulas C, such that " is either

5 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

refutable or satisable" " is either refutable or

satisable". Then, once this claim (expressed in
the previous sentence) is proved, it will suice to
prove " is either refutable or satisable" only for
's belonging to the class C. Note also that if is
provably equivalent to (i.e., () is provable),
then it is indeed the case that " is either
refutable or satisable" " is either refutable or
satisable" (the soundness theorem is needed to
show this).
There are standard techniques for rewriting an
arbitrary formula into one that does not use
function or constant symbols, at the cost of
introducing additional quantiers; we will
therefore assume that all formulas are free of such
symbols. Gdel's paper uses a version of
rst-order predicate calculus that has no function
or constant symbols to begin with.
Next we consider a generic formula (which no
longer uses function or constant symbols) and
apply the prenex form theorem to nd a formula
in normal form such that ( being in normal
form means that all the quantiers in , if there
are any, are found at the very beginning of ). It
follows now that we need only prove Theorem 2
for formulas in normal form.
Next, we eliminate all free variables from by
quantifying them existentially: if, say, x1...xn are
free in , we form
. If is

6 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

satisable in a structure M, then certainly so is

and if is refutable, then
is
provable, and then so is , thus is refutable.
We see that we can restrict to be a sentence,
that is, a formula with no free variables.
Finally, we would like, for reasons of technical
convenience, that the prex of (that is, the
string of quantiers at the beginning of , which is
in normal form) begin with a universal quantier
and end with an existential quantier. To achieve
this for a generic (subject to restrictions we
have already proved), we take some one-place
relation symbol F unused in , and two new
variables y and z.. If = (P), where (P) stands
for the prex of and for the matrix (the
remaining, quantier-free part of ) we form
. Since
is clearly provable, it is easy
to see that
is provable.

Reducing the theorem to formulas of

degree 1
Our generic formula now is a sentence, in
normal form, and its prex starts with a universal
quantier and ends with an existential quantier.
Let us call the class of all such formulas R. We are
faced with proving that every formula in R is
either refutable or satisable. Given our formula
, we group strings of quantiers of one kind

7 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

together in blocks:

We dene the degree of to be the number of

universal quantier blocks, separated by
existential quantier blocks as shown above, in
the prex of . The following lemma, which Gdel
adapted from Skolem's proof of the LwenheimSkolem theorem, lets us sharply reduce the
complexity of the generic formula we need to
prove the theorem for:
Lemma. Let k>=1. If every formula in R of
degree k is either refutable or satisable, then so
is every formula in R of degree k+1.
Comment: Take a formula of degree k+1
of the form
,
where
is the remainder of (it is thus
of degree k-1). states that for every x there
is a y such that... (something). It would have
been nice to have a predicate Q' so that for
every x, Q'(x,y) would be true if and only if y
is the required one to make (something) true.
Then we could have written a formula of
degree k, which is equivalent to , namely
. This formula is indeed equivalent to
because it states that for every x, if there is a
y thatsatises Q'(x,y), then (something) holds,

8 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

and furthermore, we know that there is such

a y, because for every x', there is a y' that
satises Q'(x',y'). Therefore follows from
this formula. It is also easy to show that if the
formula is false, then so is . Unfortunately,
in general there is no such predicate Q'.
However, this idea can be understood as a
basis for the following proof of the Lemma.
Proof. Let be a formula of degree k+1; then we
can write it as

where (P) is the remainder of the prex of (it is

thus of degree k-1) and is the quantier-free
matrix of . x, y, u and v denote here tuples of
variables rather than single variables; e.g.
really stands for
where
are some distinct variables.
Let now x' and y' be tuples of previously unused
variables of the same length as x and y
respectively, and let Q be a previously unused
relation symbol that takes as many arguments as
the sum of lengths of x and y; we consider the
formula

Clearly,

9 of 19

is provable.

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

Now since the string of quantiers

does not contain variables from x or y, the
following equivalence is easily provable with the
help of whatever formalism we're using:

And since these two formulas are equivalent, if we

replace the rst with the second inside , we
obtain the formula ' such that ':

Now ' has the form

, where (S) and
(S') are some quantier strings, and ' are
quantier-free, and, furthermore, no variable of
(S) occurs in ' and no variable of (S') occurs in .
Under such conditions every formula of the form
, where (T) is a string of quantiers
containing all quantiers in (S) and (S')
interleaved among themselves in any fashion, but
maintaining the relative order inside (S) and (S'),
will be equivalent to the original formula '(this is
yet another basic result in rst-order predicate
calculus that we rely on). To wit, we form as
follows:

and we have

10 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

Now
is a formula of degree k and therefore by
assumption either refutable or satisable. If
is
satisable in a structure M, then, considering
, we see that is satisable
as well. If
is refutable, then so is , which is
equivalent to it; thus
is provable. Now we can
replace all occurrences of Q inside the provable
formula
by some other formula dependent on
the same variables, and we will still get a provable
formula. (This is yet another basic result of
rst-order predicate calculus. Depending on the
particular formalism adopted for the calculus, it
may be seen as a simple application of a
"functional substitution" rule of inference, as in
Gdel's paper, or it may be proved by considering
the formal proof of
, replacing in it all
occurrences of Q by some other formula with the
same free variables, and noting that all logical
axioms in the formal proof remain logical axioms
after the substitution, and all rules of inference
still apply in the same way.)
In this particular case, we replace Q(x',y') in
with the formula
. Here
(x,y|x',y') means that instead of we are writing a
dierent formula, in which x and y are replaced
with x' and y'. Note that Q(x,y) is simply replaced
by
.
then becomes

11 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

and this formula is provable; since the part under

negation and after the sign is obviously
provable, and the part under negation and before
the sign is obviously , just with x and y
replaced by x' and y', we see that
is provable,
and is refutable. We have proved that is either
satisable or refutable, and this concludes the
proof of the Lemma.
Notice that we could not have used
instead of Q(x',y') from
the beginning, because
would not have been a
well-formed formula in that case. This is why we
cannot naively use the argument appearing at the
comment that precedes the proof.

Proving the theorem for formulas of

degree 1
As shown by the Lemma above, we only need to
prove our theorem for formulas in R of degree 1.
cannot be of degree 0, since formulas in R have
no free variables and don't use constant symbols.
So the formula has the general form:

Now we dene an ordering of the k-tuples of

natural numbers as follows:

12 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

should hold if either

, or
, and
precedes
in lexicographic order. [Here
denotes the sum of the terms of the
tuple.] Denote the nth tuple in this order by
.
Set the formula

as
. Then

put

as

Lemma: For every n,

Proof: By induction on n; we have

, where the latter implication holds by variable
substitution, since the ordering of the tuples is
such that
. But the
last formula is equivalent to
.
For the base case,
is obviously a corollary of as well. So the
Lemma is proven.
Now if

13 of 19

is refutable for some n, it follows that

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

is refutable. On the other hand, suppose that

is not refutable for any n. Then for each n there is
some way of assigning truth values to the distinct
subpropositions
(ordered by their rst
appearance in
; "distinct" here means either
distinct predicates, or distinct bound variables) in
, such that
will be true when each
proposition is evaluated in this fashion. This
follows from the completeness of the underlying
propositional logic.
We will now show that there is such an
assignment of truth values to
, so that all
will be true: The
appear in the same order in
every
; we will inductively dene a general
assignment to them by a sort of "majority vote":
Since there are innitely many assignments (one
for each
) aecting
, either innitely many
make
true, or innitely many make it false and
only nitely many make it true. In the former case,
we choose
to be true in general; in the latter
we take it to be false in general. Then from the
innitely many n for which
through
are
assigned the same truth value as in the general
assignment, we pick a general assignment to
in the same fashion.
This general assignment must lead to every one of
the
and
being true, since if one of the
were false under the general assignment,
would also be false for every n > k. But this

14 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

contradicts the fact that for the nite collection of

general
assignments appearing in
, there
are innitely many n where the assignment
making
true matches the general assignment.
From this general assignment, which makes all of
the
true, we construct an interpretation of the
language's predicates that makes true. The
universe of the model will be the natural numbers.
Each i-ary predicate
should be true of the
naturals
precisely when the proposition
is either true in the general
assignment, or not assigned by it (because it never
appears in any of the
).
In this model, each of the formulas
is true by
construction. But this implies that itself is true
in the model, since the
range over all possible
k-tuples of natural numbers. So is satisable,
and we are done.
Intuitive explanation
We may write each Bi as (x1...xk,y1...ym ) for some
x-s, which we may call "rst arguments" and y-s
that we may call "last arguments".
Take B1 for example. Its "last arguments" are
z2,z3...zm+1, and for every possible combination of

15 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

k of these variables there is some j so that they

appear as "rst arguments" in Bj. Thus for large
enough n1, Dn1 has the property that the "last
arguments" of B1 appear, in every possible
combinations of k of them, as "rst arguments" in
other Bj-s within Dn. For every Bi there is a Dni
with the corresponding property.
Therefore in a model that satises all the Dn-s,
there are objects corresponding to z1, z2... and
each combination of k of these appear as "rst
arguments" in some Bj, meaning that for every k
of these objects zp1...zpk there are zq1...zqm, which
makes (zp1...zpk,zq1...zqm) satised. By taking a
submodel with only these z1, z2... objects, we have
a model satisfying .

Extensions
Extension to rst-order predicate
calculus with equality
Gdel reduced a formula containing instances of
the equality predicate to ones without it in an
extended language. His method involves replacing
a formula containing some instances of equality
with the formula

16 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

Here
denote the predicates appearing in
(with
their respective arities), and ' is the
formula with all occurrences of equality
replaced with the new predicate Eq. If this new
formula is refutable, the original was as well;
the same is true of satisability, since we may take
a quotient of satisfying model of the new formula
by the equivalence relation representing Eq. This
quotient is well-dened with respect to the other
predicates, and therefore will satisfy the original
formula .

Extension to countable sets of

formulas
Gdel also considered the case where there are a
countably innite collection of formulas. Using the
same reductions as above, he was able to consider
only those cases where each formula is of degree
1 and contains no uses of equality. For a countable
collection of formulas
of degree 1, we may
dene
as above; then dene
to be the
closure of
. The remainder
of the proof then went through as before.

17 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

Extension to arbitrary sets of formulas

When there is an uncountably innite collection of
formulas, the Axiom of Choice (or at least some
weak form of it) is needed. Using the full AC, one
can well-order the formulas, and prove the
uncountable case with the same argument as the
countable one, except with transnite induction.
Other approaches can be used to prove that the
completeness theorem in this case is equivalent to
the Boolean prime ideal theorem, a weak form of
AC.

References
Gdel, K (1929). "ber die Vollstndigkeit
des Logikkalkls". Doctoral dissertation.
University Of Vienna. The rst proof of the
completeness theorem.
Gdel, K (1930). "Die Vollstndigkeit der
Axiome des logischen Funktionenkalkls".
Monatshefte fr Mathematik (in German) 37
(1): 349360. doi:10.1007/BF01696781.
JFM 56.0046.04. The same material as the
dissertation, except with briefer proofs, more
succinct explanations, and omitting the
lengthy introduction.

External links
Stanford Encyclopedia of Philosophy: "Kurt
18 of 19

07/21/2016 06:41 PM

Original proof of Gdel's comp...

https://en.wikipedia.org/w/inde...

Gdel (http://plato.stanford.edu/entries
/goedel/)"by Juliette Kennedy.
MacTutor biography: Kurt Gdel.
(http://www-groups.dcs.st-and.ac.uk/~history
/Mathematicians/Godel.html)

Retrieved from "https://en.wikipedia.org

/w/index.php?title=Original_proof_of_Gdel%27s_completen
oldid=679095433"
Categories: Logic Model theory
Mathematical proofs Mathematical logic
Proof theory Works by Kurt Gdel
This page was last modied on 2 September
2015, at 12:13.
Text is available under the Creative Commons
Attribution-ShareAlike License; additional
terms may apply. By using this site, you agree
to the Terms of Use and Privacy Policy.
Wikipedia is a registered trademark of the
Wikimedia Foundation, Inc., a non-prot
organization.

19 of 19

07/21/2016 06:41 PM