Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

If $a$ and $b$ are non-zero natural numbers, the definition of $a^b$ is clear. Now it seems to me that there are (at least) two distinct ways of generalizing to larger number systems.

Firstly, given $n \in \mathbb{N}$ and $z \in \mathbb{C},$ we can define $$z^n = \prod_{k=1}^n z$$

where the empty product is interpreted as $1 \in \mathbb{C}$. Furthermore, if $z \neq 0,$ then we can define $z^n$ for all integral $n$ via either of the following definitions.

  1. For all integers $n>0$ we define $z^{-n}=(1/z)^n$.
  2. For all integers $n>0$ we define $z^{-n}=1/(z^n)$.

Both definitions give the same result, and all our favorite identities hold under these definitions. Thus: so far, so good!

Now suppose we want more general exponents. E.g. Suppose we want to define $x^y$ for $y$ an arbitrary element of $\mathbb{R}$. To generalize in this new direction, we need to specialize in another. In particular, we require that $x$ be an element of $\mathbb{R}^+$. Then we may define $$x^y = \exp [y \log(x)]$$

My question is, are there any nasty surprises if we allow $y$ to be an arbitrary complex number in the above definition? That is, do any issues arise if we try to raise an element of $\mathbb{R}^+$ to an element of $\mathbb{C}$?

EDIT. I've just been toying with the numbers a bit, and it seems that the formula $x^{wz}=(x^w)^z$ fails for certain combinations of $w,z \in \mathbb{C}$. As a simple example, take $x=e$ and $w=i\pi,$ in which case $x^w = -1 \notin \mathbb{R}^+,$ and so $(x^w)^z$ is undefined (unless $z$ is integral).

For a more interesting example, take $x=e$, $w=2\pi i$ and $z=1/2$. Then $x^w=1,$ so $(x^w)^z=1$. But $wz=\pi i$, and so $e^{wz}=-1$. So $x^{wz}\neq (x^w)^z$, although both expressions are well-defined.

This begs the question, when is the identity $x^{wz}=(x^w)^z$ valid? Its easy to show that if $w$ is real, then the identity holds, since under these circumstances, $wz\log(x)=z\log(w^x).$ When else does it hold? I think that if $z$ is integral it will also work out, but I don't know how to show this.

share|improve this question

2 Answers 2

In short, no. You're probably familiar with the famous $e^{i\pi}=-1$ or even $e^{i\theta}=\cos \theta + i \sin \theta$. Now we have $e^{a+bi}=e^a e^{bi}$ so finally we have:

$$x^{a+bi}=(e^{\ln x})^{a+bi}=e^{a\ln x}e^{i b \ln x}=e^{a \ln x}(\cos(b\ln x)+i\sin(b\ln x))$$

share|improve this answer

Your definition of $x^z$ for real, positive $x$ and complex $z$ is consistent and natural. It works: we get $(xy)^z = x^zy^z$, $(1/x)^z = x^{-z}$ etc.

But when you try to iterate this function, as in $(x^w)^z$, you are stepping outside the domain of definition of your function, because $(x^w)$ is no longer a real, positive number. So you can't expect it to work so nicely.

share|improve this answer

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

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