1
$\begingroup$

I was playing around with Euler's identity

$$\begin{align}e^{i\pi}&=-1\\\\\frac{\log{(-1)}}{i}&=\pi\\\\\frac{x \log(-1)}{i}&=\pi x \end{align}$$

And when I had $x=2$

$$\frac{\log(1)}{i}=2\pi$$Which is equivalent to $$\frac{0}{i}=2\pi$$

How can this be?

$\endgroup$
  • $\begingroup$ $2\ln(-1)\neq \ln((-1)^2)$ $\endgroup$ – kingW3 Mar 11 '17 at 18:04
  • $\begingroup$ @kingW3 a constant times a logarithm is equal to the object of the logarithm to the constant $\endgroup$ – user406613 Mar 11 '17 at 18:05
  • $\begingroup$ This works for the real-valued logarithm and not for the complex valued logarithm. $\endgroup$ – kingW3 Mar 11 '17 at 18:07
  • 1
    $\begingroup$ The logarithm of a complex numbers is not single valued. $\ln 1$ is not just 0, but also $2k\pi i$ as well. $\endgroup$ – fleablood Mar 11 '17 at 19:57
1
$\begingroup$

When considering complex numbers, the usual rules about logarithms and exponentials do not apply. For example, $e^{2\pi i} = 1 = e^0$, even though $2\pi i \neq 0$. Most notably in this case, the rule $a\ln{b} = \ln(b^a)$ is not true when $b$ is not a positive real or when $a$ is not real. So it's not correct to say that $2\ln(-1) = \ln{1}$.

The logarithm and exponential rules you were taught in intro algebra (e.g., $b\log(a) = \log(a^b)$ and that $a^b = a^c$ if and only if $b = c$) are true only for real numbers. Introducing complex numbers significantly complicates the issue: since $e^x$ is no longer a one-to-one function, it's questionable what exactly $\ln$ means. This is solved by something called branch cuts, but there isn't a way to define $\ln$ so that all of the usual log rules apply.

$\endgroup$
  • 1
    $\begingroup$ Interestingly $\log(z_1z_2)=\log(z_1)+\log(z_2)$, in general $\log(z^2)\ne 2\log(z)$. This superficially seeming paradox was discussed in THIS ANSWER. $\endgroup$ – Mark Viola Mar 11 '17 at 18:27
1
$\begingroup$

Euler's identity says not just that $e^{i\pi}=-1$, but that $$e^{iz}= \cos(z)+i\sin(z)$$ which is not bijective, so just defining $\ln$ as the inverse of exponentiaton is no longer well defined, it only works like that for real numbers.

Thus to define a logarithm, to make it the inverse of exponentiation you must remove a ray from the origin, which is called choosing a branch of your logarithm. When you do this you will get a result which is consistent.

$\endgroup$
0
$\begingroup$

In THIS ANSWER, I showed that $\log(z^2)\ne 2\log(z)$ in general.

This might seem paradoxical given the relationship

$$\log(z_1z_2)=\log(z_1)+\log(z_2) \tag1$$

But $(1)$ is interpreted as a set equivalence. It means that any value of $\log(z_1z_2)$ can be expressed as the sum of some value of $\log(z_1)$ and some value of $\log(z_2)$. And conversely, it means that the sum of any value of $\log(z_1)$ and any value of $\log(z_2)$ can be expressed as some value of $\log(z_1z_2)$.


But, for the problem of interest, simply note that

$$2\log(-1)=2i(2k+1)\pi \ne \log((-1)^2)=0$$

for $k\in \mathbb{Z}$.

$\endgroup$
  • $\begingroup$ Please let me know how I can improve my answer. I really want to given you the best answer I can. -Mark $\endgroup$ – Mark Viola Apr 12 '17 at 17:09

Your Answer

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