Properties of Recurisve and Recursively Enumerable Languages

We already know that the best way to solving a number of problems, it must be reduced to
another. The reduction process involves a combination of many Turing machines to form a
combined machine. The combined Turing machine will be able to perform an action if the
algorithm accepts the action and another action if the algorithm doesn’t accept it. A normal
Turing machine wouldn’t be able to perform the above action since if the Turing machine didn’t
accept, it could run forever and the combined Turing machine would never begin the next
action.//Keeping this notion in mind, we can show following properties or theorems of
recursive language and subsequently we can also show the properties of recursively
enumerable languages.
Theorem 1: The complement of a recursive language is also recursive.
Proof:Let L be a recursive language and L¯ be its complement. Also let M be a Turing machine

m
er as
that halts on all inputs and accepts L. Here, the accepting states of M are made non-accepting
states of M′ with no transitions, i.e., here M′ will halt without accepting. If s is new accepting

co
eH w
state in M′, then there is no transition from this state. Also if L is recursive, then L = L(M) for
some Turing machine M, which always halts. Then M is transformed into M´ so that M´ accepts

o.
rs e
when M does not and vice-versa. So M´ always halts and accepts L¯. Hence L¯ is recursive. It
ou urc
shows that the complement of L is also recursive.
Theorem 2: A language is recursive if and only if both it and it’s complement are recursively
o

enumerable.
aC s

Proof:Here it is given that L and L¯(complement of L) are recursively enumerable. Let L = L(M1)
vi y re

and L¯ = L(M2). Now let us construct a Turing Machine ‘M’ that simulates M1 and M2 in
parallel, using two tapes and two heads. If input to M is in L, then M1 accepts it and halts,
hence M accepts it and halts. If input to M is not in L, then it is in L¯, and so M2 accepts and
ed d

halts, hence M halts without accepting. Hence M halts on every input and L(M) = L which shows
ar stu

that L is recursive. Hence, it shows that if both L and L¯ are recursively enumerable, then L is
recursive
is

Theorem 3: The union of two recursive languages is recursive.

Th

Proof:Let L1 and L2 be two recursive languages accepted by Turing machines M1 and M2,
respectively. First of all, we construct a Turing machine M which first acts on M1. If M1 accepts
then M accepts as well but if M1 rejects then M moves on to M2 and simulates it. Same as
sh

before, M accepts if and only if M2 accepts. Due to the fact that both M1 and M2 are
algorithms, M is sure to halt. Thus, L1 ∪ L2 is accepted by Turing machine M, where x = w1 ∪
w2, for w1 ∈ L1 and w2 ∈ L2. Hence, L1 ∪ L2 is also recursive since a Turing machine M exists
for it.
Theorem 4: The union of two recursively enumerable languages is recursively enumerable.

This study source was downloaded by 100000823000995 from CourseHero.com on 04-07-2021 06:44:11 GMT -05:00

https://www.coursehero.com/file/83439812/Properties-of-Recurisve-and-Recursively-Enumerable-Languagespdf/
 Proof:Let L1 and L2 be two recursively enumerable languages accepted by Turing machines M1
and M2, respectively. Then L1 ∪ L2 is accepted by Turing machine M, where x = w1 ∪ w2, for
w1 ∈ L1 and w2 ∈ L2. To determine if M1 or M2 accepts x, we run both M1 and M2 at once,
using a two-tape Turing machine M. M first simulates M1 on the first tape and then M2 on the
second tape. If either one enters the final state, the input is accepted by Turing machine M.
Hence it shows that L1 ∪ L2 is recursively enumerable.

Theorem 5 (Rice’s Theorem): Let C be a set of proper recursively enumerable languages and
LC be a language defined as LC = {M: L(M) ∈ C}. Then LC is undecidable.

Proof:Without the loss of generality, we may suppose that a Turing machine that identifies the
bare language does not have the property C. But if a Turing machine does, then we just take the
complement of C. The undecidability of the complement of C would instantly imply the
undecidability of C.

m
Here let us suppose towards a contradiction that for some class C, the language LC is not

er as
empty. It does not contain the descriptions of all Turing machines, and also it is decidable. Then

co
complement of LC is also not empty which also doesn’t contain descriptions all Turing

eH w
machines, and is also decidable. We can assume that Ï• ∉ C, otherwise we have to apply the

o.
argument below to the complement of LC instead of LC. Let Ma be a machine such that <Ma> is
rs e
ou urc
in LC. Here we have to show that the acceptance problem is decidable to reach a contradiction.
Given an input (, w) for the acceptance problem, we construct a new Turing machine Mb that
does the following: on input x, Mb first simulates the actions of M on input b and shows that
o

If M on input b loops, then so does Mb;

aC s
vi y re

If M on input b rejects, then so does Mb;

If M on input b accepts, then Mb continues with a simulation of Ma on input x.
ed d
ar stu
is
Th
sh

This study source was downloaded by 100000823000995 from CourseHero.com on 04-07-2021 06:44:11 GMT -05:00

https://www.coursehero.com/file/83439812/Properties-of-Recurisve-and-Recursively-Enumerable-Languagespdf/
Powered by TCPDF (www.tcpdf.org)