Dan Simmons published a wonderful, galaxy-spanning, mind-blowing sf novel in 1989: Hyperion. Then he followed it up with three more novels of which I have read two. They’re OK, but not as good as the first book.
Science fiction is of course stories where fabulous things happen and are explained by science and technology rather than magic. There are two ways to do this: either you offer an explanation that is actually in line with what we know now and sort of makes sense, or you use technobabble to cover the fact that you, the writer, do not actually have any idea of how for instance space ships move instantaneously from one star to another. Both ways are in my opinion fine. And Simmons uses the technobabble technique with poetic flair: “torch ship”, “lance the ground troops from orbit”, “spin down into the system”, “hyper-entropic field”.
But in the third of the Hyperion novels, Endymion (1996), he does something that jarred me awake to the fact that Simmons apparently does not know basic science at all. He tells us that a couple of fabulous things happen and offers neither technobabble nor believable scientific explanations.
We’re on the planet Sol Draconi Septem. Having once been terraformed, thus receiving a breathable atmosphere, it has now relapsed (over a few centuries) into a chilly state where its atmosphere has frozen solid, collapsed onto the surface and formed glaciers. There is hardly any gaseous matter there any more. Those glaciers seem to consist largely of solid nitrogen. Yet Simmons tells us that there is breathable air and liquid salty water in tunnels dug through the ice by a species of large animal, the ice wraiths. And on top of the glaciers, where it is impossible to breathe, intense blizzards blow. So there is wind in a vacuum, and there is precipitation without an atmosphere. Ouch.
The ecology of Sol Draconi Septem is also magical. It consists only of two species of carnivore that hunt each other: ice wraiths and humans. No plants and no herbivores. Simmons does mention that the human population is shrinking, which suggests that he understands that a system without energy input will dwindle and eventually stop running. But as far as I can see he’s vastly overestimated the longevity of such a system. It is after all the equivalent of fencing a desert in, removing all animals and filling it with lions and tigers. And it’s not just a matter of energy efficiency, but also one of materials: if a tiger eats a lion, far from all of the lion’s building blocks become incorporated into the tiger.