7

In Keldysh Field Theory a classical field $\phi_c$ and a quantum field $\phi_q$ are defined as results of the Keldysh rotation \begin{equation} \phi_c = \frac{\phi_++\phi_-}{\sqrt{2}}, \end{equation} \begin{equation} \phi_q = \frac{\phi_+-\phi_-}{\sqrt{2}}, \end{equation} where $\phi_+$ and $\phi_-$ are the field configurations on the forward and backward contour, respectively.

Why do we call them classical and quantum? Is there a hidden meaning that I am missing?

Reference: I am looking at "Field Theory of Many-Body Lindbladian Dynamics" by Foster Thompson & Alex Kamenev https://arxiv.org/abs/2301.02953 (first introduced on page 7).

Qmechanic
  • 220,844
GSLAM
  • 113

2 Answers2

3

If $\phi_q\ll \phi_c$ then $\phi_c$ obeys the classical equations of motion independent of whatever the quantum fluctuations $\phi_q$ are doing.

As a toy model, consider single-particle quantum mechanics[1]. The action on the closed time contour is $$S[X] = \int_{\mathcal{C}}dt \left[ \frac{1}{2}\dot X^2 - V(X)\right].$$ Split $X(t)$ into $X^+(t)$ and $X^-(t)$ which reside on the forward branch and backward branch of $\mathcal C$ respectively. The classical $X^{\rm cl}$ and quantum $X^{\rm q}$ fields would be $$\begin{align} &X^{\rm cl} (t) = \frac {1}{2} \left( X^+(t) + X^-(t) \right) \qquad &X^{\rm q} (t) = \frac{1}{2}\left( X^+(t) - X^-(t) \right) \end{align} $$ Rewriting the action in terms of the classical and quantum fields: $$\begin{align} S[X] &= \int_{\mathcal C}dt\left[ \frac{1}{2}\dot X^2 - V(x) \right]\\ &=\int_{-\infty}^\infty dt \left[ \frac{1}{2}(\dot X^+)^2 - \frac{1}{2}(\dot X^-)^2 - V(X^+) + V(X^-) \right]\\ &=\int_{-\infty}^\infty dt \left[ 2 \dot X^{\rm cl} \dot X^{\rm q} - V(X^{\rm cl} + X^{\rm q} ) + V(X^{\rm cl} - X^{\rm q} )\right]\\ \textrm{(integrate by parts)} &=\int_{-\infty}^\infty dt \left[ -2\ddot X^{\rm cl} X^{\rm q} - V(X^{\rm cl} + X^{\rm q} ) + V(X^{\rm cl} - X^{\rm q} )\right]. \end{align} $$ Now suppose the quantum component $X^q$ is small. Expanding the potentials yields $$S[X]\xrightarrow{X^{\rm cl} \gg X^{\rm q} }- \int_{-\infty}^\infty dt \left[ 2X^q \left( \ddot X^{\rm cl} + \frac{ d V } { d X } \biggr\lvert_{X^{\rm cl}} \right)+O((X^{\rm q})^3) \right].$$ In this limit, no matter what the quantum field $X^{\rm q}$ is doing, the action is minimized if the classical field $X^{\rm cl}$ obeys the classical equations of motion $$\ddot X^{\rm cl} = - \frac{ d V } { d X } \biggr\lvert_{X^{\rm cl} }$$

[1]: This is an example given in Chapter 3 of "Field Theory of Non-Equilibrium Systems" by Alex Kamenev

delon
  • 469
  • 2
  • 6
3
  1. The Keldysh "in-in" Euler-Lagrange (EL) equation for $\phi_c$ yields $\phi_q\approx 0$, while the Keldysh "in-in" EL equation for $\phi_q$ yields the standard "in-out" Lagrange equation for $\phi_c$ [up to terms of order ${\cal O}(\phi_q^3)$]. The field $\phi_c$ is classical in that sense. The deviation $\phi_q$ in $\phi_{\pm}=\frac{\phi_c\pm\phi_q}{\sqrt{2}}$ is a quantum fluctuation, which vanishes classically.

  2. For more details, see e.g. my Phys.SE answer here [although be aware that the notation there is different].

Qmechanic
  • 220,844