Negative Energy, Superluminosity and Holography
Joseph Polchinski, Leonard Susskind, Nicolaos Toumbas
TL;DR
The paper explores how the AdS/CFT correspondence enforces a nontrivial, nonlocal structure in large-N SYM theories to reproduce bulk AdS dynamics, including paradoxical features like negative energy density and apparent superluminal signals. By analyzing both gravitational-wave propagation in AdS and a free-field toy model, it shows that the boundary stress-energy can be decomposed into a principal part and a smaller, nonlocal 'interest,' with precursors—nonlocal field configurations—carrying the information necessary to reconcile boundary observations with bulk events. It proposes that these precursors are encoded in nonlocal gauge-invariant constructs, such as Wilson loops, and demonstrates that the required phenomena can emerge naturally within quantum field theory without violating fundamental principles. The work highlights the essential role of nonlocal information storage in holography and points toward a Wilson-looped precursor framework for encoding bulk data in boundary theories.
Abstract
The holographic connection between large $N$ Super Yang Mills theory and gravity in anti deSitter space requires unfamiliar behavior of the SYM theory in the limit that the curvature of the AdS geometry becomes small. The paradoxical behavior includes superluminal oscillations and negative energy density. These effects typically occur in the SYM description of events which take place far from the boundary of AdS when the signal from the event arrives at the boundary. The paradoxes can be resolved by assuming a very rich collection of hidden degrees of freedom of the SYM theory which store information but give rise to no local energy density. These degrees of freedom, called precursors, are needed to make possible sudden apparently acausal energy momentum flows. Such behavior would be impossible in classical field theory as a consequence of the positivity of the energy density. However we show that these effects are not only allowed in quantum field theory but that we can model them in free quantum field theory.
