Last night a preprint appeared on the arXiv from beyond the grave, an undated manuscript entitled Quantum Gravity, Yesterday and Today, found without any indication of its purpose in the files of Bryce DeWitt, who passed away in 2004.
DeWitt devoted much of his career to the question of how to quantize the gravitational field, beginning back in 1948 when he was a student of Julian Schwinger. He has some interesting comments about the dramatic changes over the years in popularity of research work on GR and quantum gravity:
Most of you can have no idea how hostile the physics community was, in those days, to persons who studied general relativity. It was worse than the hostility emanating from some quarters today toward the string-theory community. In the mid fifties Sam Goudsmidt, then Editor-in-Chief of the Physical Review, let it be known that an editorial would soon appear saying that the Physical Review and Physical Review Letters would no longer accept “papers on gravitation or other fundamental theory.” That this editorial did not appear was due to the behind-the-scenes efforts of John Wheeler.
DeWitt gives some history of his important work on the quantization of gauge theories, which culminated in working out a functional integral method to handle to all orders the ghost terms that Feynman had shown to be necessary. He describes a 1955 offer from the Glenn L. Martin Aircraft Company to fund his research in hopes that it would lead to an antigravity device, one that he didn’t accept. Instead, the Air Force supported his research during the period he was unraveling the story of ghosts, support that ended in 1966 when they finally realized that gravity research was not going to lead to magical results. With the termination of his grant, he could no longer pay page charges to the Physical Review, delaying the publication of one of his papers by a year.
He also has some interesting comments about the DeWitt-Wheeler equation:
… intensive work was carried out in those years on canonical quantum gravity, culminating in an equation that bears my name along with that of John Wheeler who was the real driving force. Research on the consequences of this equation continues to this day, stimulated by work of Abhay Ashtekar, and some of it is quite elegant. But apart from some apparently important results on so-called “spin foams” I tend to regard the work as misplaced. Although WKB approximations to solutions of the equation may legitimately be used for such purposes as calculating quantum fluctuations in the early universe, and although the equation forces physicists to think about a wave function for the whole universe and to confront Everett’s manyworld view of quantum mechanics, the equation, at least in its original form, cannot serve as the definition of quantum gravity. Aside from the fact that it violates the very spirit of general relativity by singling out spacelike hypersurfaces for special treatment, it can be shown not to be derivable, except approximately, from a functional integral. For me the functional integral must be the starting point.
He ends the paper with positive comments on string theory:
In viewing string theory one is struck by how completely the tables have been turned in fifty years. Gravity was once viewed as a kind of innocuous background, certainly irrelevant to quantum field theory. Today gravity plays a central role. Its existence justifies string theory! There is a saying in English: “You can’t make a silk purse out of a sow’s ear.” In the early seventies string theory was a sow’s ear. Nobody took it seriously as a fundamental theory. Then it was discovered that strings carry massless spin-two modes. So, in the early eighties, the picture was turned upside down. String theory suddenly needed gravity, as well as a host of other things that may or may not be there. Seen from this point of view string theory is a silk purse. I shall end my talk by mentioning just two things that, from a nonspecialist’s point of view, make it look rather pretty.
The two things he has in mind are the ability of a single string diagram to sum up a lot of Feynman diagrams, and the use of orbifolds to make possible topology-changing transitions.