I don't quite understand one example in my notes it says.

Exercise with answer showing that conjugation is not complex differentiable

My query is this: I don't understand what the significance of $\theta$ is. Why does it matter that $\theta \in (-\pi,\pi]$? I see the argument as this, clearly anyway in which we approach $0$ must each give the same limit otherwise the limit doesn't exist since the limit is independent of $w$ the limit only depends on $\theta$ hence for any two distinct values of $\theta$ of which there are infinitely many the limit is different $\implies$ the limit does not exist.

Is this correct?

Also what is $\theta$ changes as the complex number $w$ approaches zero what is to say we need to approach along a straight line with constant $\theta$?

Could anyone clear up my misconceptions?



The significance of $\theta$ comes from the exponential/polar representation of a complex number. Here we have $$ w \in \mathbb{C} \iff \exists r \in \mathbb{R}^+ : \exists \theta \in \mathbb{R} : w = re^{i \theta}$$ However the author of the answer has limited our investigation to a subset of the possible values of $\theta$ such that $\theta \in (-\pi , \pi]$. I believe this is a roundabout way of saying that there exist values of $\theta$ for which the limit is different and by inspection of this subset of values we may find them. The argument does not change if we let $ \theta \in \mathbb{R}$. Perhaps the author was trying to get your attention to some specific values in the interval? For instance $$ \theta = 0 \implies w = r, \ \overline{w} = r \implies \frac{\overline{w}}{w} = 1$$ $$ \theta = \frac{\pi}{2} \implies w = ir, \ \overline{w} = -ir \implies \frac{\overline{w}}{w} = -1 $$ Both these values of $\theta$ are in the interval described, and as can be seen now the limit is not unique and hence does not exist.

As to the second part. You are stumbling upon a very nice property of complex analysis. The fact that "direction" is intrinsically built in! What I mean by this is that as opposed to the standard formulation of differentiation on $\mathbb{R}^2$ in which we deal with linear transformations and sorts, we instead have a formulation of the derivative which is not entirely dependent on this "distance" from the point we are investigating. This can be seen by considering the mapping $$ (x,y) \to (x, -y)$$ This corresponds to a reflection along the x-axis, and furthermore in complex analysis this is the act of conjugation! However one will find out that although reflection in $\mathbb{R}^2$ is differentiable, in complex analysis it is not(holomorphic that is)!


It is not necessary for $w$ to approach $0$ along straight lines. But if the limit were to exist, it wouldn't matter what path one takes to $0$. So finding a contradiction among the straight line paths to $0$ is enough to conclude the limit doesn't exist.

This is just like ordinary limits. Or perhaps this feels like ordinary limits in multivariable calculus/analysis.

I do not understand what you say after My query is this:

The crux of the problem is that we can approach $0$ along different paths at an angle $\theta$. The resulting limit depends on $\theta$, which means the limit doesn't exist. This is akin to the function $$ f(x) = \begin{cases} -1 & x < 0 \\ 1 & x > 0\end{cases}$$ and asking for the limit as $x \to 0$. If we approach from the right, we get $1$. If we approach from the left, we get $-1$. But for the limit to exist, it must not matter from which direction we approach.

So the limit does not exist.

  • $\begingroup$ I understand that now and the examples given above really helped concrete it for me thank you. $\endgroup$ – David P Jul 2 '15 at 11:39

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.