It depends on how you consider $\mathbb C$ to be a vector space.
If $\mathbb C$ is a vector space over $\mathbb C$ (and scalar multiplication is the ordinary multiplication of complex numbers), then the only linear maps are given by multiplication by a complex constant. This is true whenever a field is considered to be a vector space over itself: if $f$ is a linear function then
$$ f(x) = f(x\cdot 1) = x\cdot f(1) = f(1)x $$
so $f(1)$ must be a constant that $f$ multiplies everything by.
On the other hand, if you view $\mathbb C$ as a real vector space -- that is, you consider $\mathbb R$ to be the scalar field -- then the conditions for being linear puts fewer demands on $f$ because there are fewer scalars that it needs to be compatible with. In that case there are more functions that qualify as linear, such as the one you suggest.
In order to prevent such confusion, one sometimes speaks about "$\mathbb C$-linear" or "$\mathbb R$-linear" functions when it is not otherwise clear what the scalar field is.