It is sufficient to set the cosmological constant to zero or to a small number at aninitial time

Abstract

I point out a simple but usually overlooked fact about the cosmological constant problem: to solve thecosmological constant problem it is sufficient to nd a symmetry or mechanism that sets the cosmological constant tozero or to a tiny value at some time in the past, provided that general relativity is the relevant theory of gravity, andthe energy{momentum tensor (excluding the part of the form of a cosmological constant) is conserved. The relevantsymmetry or mechanism need not be applicable today. Any additional cosmological constant term induced by a phasetransition in the energy{momentum tensor in this case is compensated by a shift in the cosmological constant term ofgravitational origin.