First section: beyond spacetime, and beyond the Bronstein cube
Butterfield starts out from the so-called ‘Bronstein cube’:
This cube is too neat to adequately capture the situation in quantum gravity (QG), in several respects: the edges are often not well-defined, probably not unique, generally don’t commute. And it might not at all be what you want, i.e., perhaps we won’t need a quantization of gravity.
Consider three broad approaches: (A) strings, (B) loops and its ilk (including causal set theory), and (C) condensed matter approaches. But e.g. string theory doesn’t say, for instance, we should let $c$ be finite starting out from quantum Newton-Cartan theory, etc.
In sum, Butterfield concludes “the search for QG is like orienteering in a blizzard–without a map.”