• We would expect that the quantum theory reduces to Einstein's theory of gravity in the classical limit.
• We would expect that singularities, such as occur in the classical description of black holes, do not occur in the quantum theory.
• We would demand that a space-time metric emerges from a given calculation, rather than having to be input in part or whole at the start.
• We would expect that space-time is actually somehow
quantized, made discrete, at a fundamental level, which
nevertheless preserves causality. However it is very
difficult to see how such a quantization could ever be
consistent with Lorentz Covariance! So scratch this
expectation.
• The theory should be Lorentz Covariant. {Since
observable nature is.}
• We would not expect that the theory is perturbative. That is, a boson for gravity, a graviton, need not exist in the theory.
• We probably should not demand that the theory be "algebraic." That is, it might turn out that all calculations must be entirely numerical, with no formal "result."
• It might well be that gravity is inherently classical, and that a “quantum” theory of gravity would amount to finding a way to allow a general quantum framework to deal with classical and quantum fields on the same footing--- this is indeed one of the noticeable trends in research during the past 5 years or so.