Consider a function $F(x, y)$ of two complex variables. For $\Re(y)>0$, we know the analytic structure of the function. In that case, the function is meromorphic, with simple poles in $x$ at locations $x=w_i$, for $i=0, 1, ...$ . The locations of the poles do not depend on $y$, however the residues do depend on $y$.

Furthermore, we know that $F(x, y)$ is symmetric in $x$ and $y$. So, we also know its analytic structure for $\Re(x)>0$. In that case, the function has poles in $y$ at $y=w_i$. The problem is to know the analytic structure of the function when $\Re(x)<0$ and $\Re(y)<0$ simultaneously. I suspect that, in that case, the function will still be meromorphic, with poles at $x=w_i$ and $y=w_j$.

I came across this problem, in the context of physics. I am a theoretical physics PhD student, so I do not even know how to start. If somebody knows the solution to the problem, or has any helpful comments, I would be very grateful.


There is a related idea sometimes known as "Bochner's Lemma" (I first saw this as part of an appendix in R. Langlands' Springer Lecture Notes in Math 544). Let $f(z,w)$ be a holomorphic function of two complex variables $z,w$ (I feel more comfortable having complex variables $z,w$ rather than $x,y$) in the region where $\Re(z)>0$ or $\Re(w)>0$, and with "vertical" growth bounded by $|f(z,w)|\le C\cdot e^{(1+|\Im(z)|+|\Im(w))^\alpha}$ for some real $\alpha<2$. Then $f(z,w)$ has an analytic continuation to all of $\mathbb C^2$.

More generally, for a connected region $\Omega\subset\mathbb R^2$, holomorphic $f$ on the "tube domain" $(\Re(z),\Re(w))\in \Omega$, with the same vertical growth bound, extends to the convex closure of that region.

The proof, essentially by a slightly clever use of Cauchy's theorem in one variable, is reproduced various places. Langlands' SLN 544 is on-line nowadays, for example. The lemma is in various on-line notes of mine, e.g., appendix 3A in my Cambridge Univ. Press book "Modern Analysis of Automorphic Forms by Example". This is also (legally!) on-line linked-to from http://www.math.umn.edu/~garrett/m/v/

  • $\begingroup$ Thanks, this seems very useful! I will study the notes you mention. $\endgroup$ – LeastSquare Nov 8 '18 at 22:29

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.