Prove, using only the field axioms of real numbers, that $0/0$ is undefined.
I have thought about it for a while and come up with an idea how to solve this. First, I would like to prove (using field axioms - again) that $x \cdot \frac{1}{y} = \frac{x}{y}$, (but I am still working on this - any hints would be most appreciated),therefore $\frac{0}{0} = 0 * \frac{1}{0}$ and then, say that - by virtue of one of the axioms there exists only one number $b$ satisfying the equation $ab = 1$ for all $a \ne 0$. Since in our case we want to take the reciprocal of 0 - the axiom clearly states that it does not exist.
Is this a good way to prove that the expression in question is undefined?

  • $\begingroup$ What does $\frac xy$ mean if it is not an abbreviation for $xy^{-1}$? $\endgroup$ – Mees de Vries Nov 3 '17 at 14:19
  • $\begingroup$ @MeesdeVries This is obvious, but as long as we are limited to the axioms, I like to think thoroughly through every step in my proof. $\endgroup$ – Aemilius Nov 3 '17 at 14:20
  • $\begingroup$ According to filed axioms $\dfrac 1 a$ is defined only for $a \ne 0$. $\endgroup$ – Mauro ALLEGRANZA Nov 3 '17 at 14:20
  • $\begingroup$ I am speaking on an "every step of the proof" level. Typically, the field operations are $+, \times, 0, 1$. Thus $\frac xy$ is not something you can formulate directly in the language of fields. Hence it must be defined. How is it defined? $\endgroup$ – Mees de Vries Nov 3 '17 at 14:21
  • $\begingroup$ Thus, the issue is that $\dfrac 0 0 = 0 \times \dfrac 1 0$ is undefined because $\dfrac 1 0$ is undefined. $\endgroup$ – Mauro ALLEGRANZA Nov 3 '17 at 14:23

Let $R$ be any ring with $1$. Suppose that $0^{-1}$ makes sense. Then that means $0 \cdot 0^{-1}=1 = 0^{-1}\cdot 0$. Let $a \in R$ be arbitrary. Then $$ a = 1a = 0^{-1}0a = 0^{-1}0=0. $$ Hence $R=\{0\}$.

Thus, if you want your ring to have more than one element, you better not let $0$ have an inverse. So, not only should $\frac{0}{0}$ not make sense, no "fraction" of the form $\frac{a}{0}$ should make sense.

In case you do not believe that $a0=0$, we have $$ a0 = a(0+0) = a0+a0. $$ Now cancel an $a0$ from each side. If you do not believe in subtraction/cancellation someone else will have to help you.

  • $\begingroup$ Your equations rely on $a \times 0 = 0$, but if you already know that then you're done (since $0 \neq 1$ in a field). $\endgroup$ – Mees de Vries Nov 3 '17 at 14:26
  • $\begingroup$ I was avoiding assuming $1 \neq 0$ until the end. Some people allow this, but either way, the argument above forces $R=\{0\}$ no matter what. $\endgroup$ – Randall Nov 3 '17 at 14:28
  • $\begingroup$ Fine, but your argument is still missing the proof that $a \times 0 = 0$. $\endgroup$ – Mees de Vries Nov 3 '17 at 14:29
  • $\begingroup$ I'm assuming that's not a dealbreaker/showstopper. The OP can insert it. $\endgroup$ – Randall Nov 3 '17 at 14:29
  • $\begingroup$ I disagree; I think it is the "difficult" part of the proof. But I will provide my own answer then. $\endgroup$ – Mees de Vries Nov 3 '17 at 14:30

The expression $\frac xy$ is not immediately in the language of fields, which only has the symbols $+,\times,-,0,1$. Thus we must define what $\frac xy$ means, and the fastest thing to do is as follows: if the equation $ya = x$ has a unique solution $a$, then $a = \frac xy$. (Typically you would define $y^{-1}$ first, and then set $\frac xy = xy^{-1}$, and we would have to rule out the existence of $0^{-1}$, but this route also works.)

Thus, let us show that $0 \times a = 0$ does not have a unique solution. In fact, for any $a$ we have $$ 0 \times a = (0 + 0) \times a = 0 \times a + 0 \times a. $$ Now subtract $0 \times a$ from each side to obtain $0 = 0 \times a$. This means that the equation defining $\frac 00$ does not have a unique solution -- it has at least the solutions $0,1$ -- and thus $\frac 00$ is not well-defined.

  • $\begingroup$ How do you infer that because of the fact that $0 = 0 a$, the expression $0/0$ has more than one value? $\endgroup$ – Aemilius Nov 3 '17 at 15:38
  • $\begingroup$ Like I mentioned before, $\frac xy$ does not a priori mean anything. We have to give it meaning. The standard way to do this (without going via $y^{-1}$) would be to say: $\frac xy$ is defined if and only if $ya = x$ has a unique solution in $a$, and then that solution is $a = \frac xy$ by definition. Since $0 = 0a$ for at least two distinct $a$, we do not have a unique value for $\frac 00$. $\endgroup$ – Mees de Vries Nov 3 '17 at 15:42

Suppose $\frac{0}{0}$ is defined.

Then, by the Existence of a Multiplicative Inverse, $\frac{0}{0}=1$

By the Existence of an Additive Inverse, we have $a+(-a)=0$. Using this with $a=1$, we can write:

$1-1=0 \Leftrightarrow \frac{0}{0} - \frac{0}{0} =0 \Leftrightarrow \frac{0-0}{0}=0 \Leftrightarrow \frac{0}{0}=0$

which leads to a contradiction by uniqueness of your multiplicative inverse.

  • 1
    $\begingroup$ Thanks, your answer is clear and hits the spot. However, I've got one question - you wrote that $0/0 - 0/0 = (0-0)/0$. But none of the axioms tells anything about fraction addition. Can I use it? $\endgroup$ – Aemilius Nov 3 '17 at 15:33
  • $\begingroup$ Consider the Distributivity of Multiplication over Addition that tells you $a\times(b+c)=a\times b+a\times c$ and let $a$ be the inverse of $0$. $\endgroup$ – Zachary Nov 3 '17 at 22:44

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.