Monday, November 30, 2015

Health of Hard Science Fiction in 2015 (Short Fiction)

Now that 2015 is almost over as far as the Hugos go, we decided to look over all the stories that we or anyone else recommended and see which qualified as hard SF. In particular, we wanted to investigate the following claims:
  1. No one is writing good hard-SF stories anymore.
  2. Hard SF has no variety and keeps reusing old ideas.
  3. Only men write hard SF.
  4. Most hard SF is published in Analog.
For short science fiction written for the major publications in 2015, we found all these claims to be false. In particular, we found 22 recommended hard SF short stories suitable for lovers of hard SF who plan to make nominations for the 2015 Hugos next month and want suggestions for things to read.


A good story

To determine what stories are "good" we went through our own year-to-date table, which contains all the stories we have awarded four or five stars plus all the stories that io9, Locus, SFRevu, or SF Editor's Pick recommended. This is limited to short stories, novelettes, and novellas we actually read, which included all the ones published in Analog, Asimov's, Clarkesworld, F&SF, Lightspeed, and, as well as a few anthologies plus selected stories from other sources.

Hard SF

For our purposes, a hard-SF story has two key properties:
  1. The science must be accurate enough that an educated layman does not have to suspend disbelief for it.
  2. Some aspect of science or technology is key to the plot. It cannot merely supply the setting. 
This is consistent with Ben Bova's definition, "the writer is free to use anything his or her imagination can invent and depict -- so long as no one can show that it contradicts the tenets of known science," and roughly equivalent to the TV-tropes levels 4 and up. 

Situational SF

A story that satisfies condition 1 but not condition 2 we called "situational SF." A murder mystery on the moon with beautiful descriptions of how the lunar city worked would be situational SF if none of those details was primary to the plot.

Soft SF

A story that's not fantasy but which fails condition 1 we call "soft SF." A story where a black hole was sucking the atmosphere from hundreds of planets would be soft SF. Likewise stories that feature things like time travel, AIs with emotions, psionic abilities, etc. This is roughly equivalent to the TV-tropes levels 3 and below.


We drew the line between soft SF and fantasy using George R.R. Martin's "furniture" rule: look for aliens vs. elves (or magic vs. forcefields, etc.) If there was no magic, no witches, no demons, etc., then we called it soft SF, not fantasy, even though it can be argued that soft-SF is really just "fantasy science."


This is where we  put Alternate history, steampunk, stories with a musical theme, and stories that had no speculative element at all.

The Breakdown

With those definitions, our list of recommended stories breaks down as follows:

A few points to take from this chart:

  • Hard SF accounted for one-eighth of the good stories we read this year.
  • If you include situational SF, then hard SF accounted for over a quarter.
  • If you think of soft-SF as "fantasy science," then fantasy of one sort or another accounted for almost two-thirds of all the stories.

Good Hard-SF Short Fiction from 2015

We aren't claiming that these were the only good hard SF stories, of course. The list would likely have been larger if we had read more sources. For example, we only read a single issue of Interzone, but we found a good hard-SF story in it, "Edited," by Rich Larson.

On the other hand, the list would be shorter if we eliminated stories that used technologies that aren't currently in the lab. (E.g. any of the stories like "Edited" that involve augmented human beings.) We're comfortable with where we've drawn the line, and we've tried hard to be consistent.


One-eighth is a good chunk. We see no basis for the claim that "no one writes good hard SF anymore."

The story topics are quite varied. We see no basis for the claim that hard SF is all "rockets, chemicals and talking squids in outer space."

Of the 22 stories in our list, fourteen were written by men and eight by women. In a sample of 22, there is a 14% chance of having that few or fewer women just by chance, so this sample does not show a statistically-significant difference between the number of men and the number of women writing good hard SF for the major publications.

Contrary to expectations, Asimov's had more hard SF than Analog, although the difference isn't statistically significant. However, the 95%-confidence interval for Analog is between 38% and 6.5%. That means that even though Analog accounted for only 18% of the hard SF in this sample, it's barely possible that it really accounts for 38% of all hard-SF, but just by bad luck they only had 4 this year. (Or that it only accounts for 6.5% but by sheer luck they got as many as 4.) That destroys the hypothesis that "most" good hard SF stories are published in Analog.

Surprisingly didn't publish anything that was both good and hard SF. (We didn't read all of Interzone this year, so that number isn't representative for them.)

It's not a surprise that the Future Visions anthology had a lot of hard SF, given that that was more or less the idea, but the Old Venus stories were meant to be set in an alternate universe, so it's surprising to see a hard-SF story from it. What happened is that David Brin produced an excellent story by breaking the rules and setting his story on a future, terraformed Venus. Not listed are the anthologies Stories for Chip and Twelve Tomorrows, neither of which had any recommended hard-SF stories.


We cannot conclude anything about change in the status of hard SF, since we only have results for this year, but it certainly seems to be in better shape than many people would claim.

5 comments (may contain spoilers):

  1. Was it an error to include "DreamPet" by Bruce McAllister, published in F&SF, and then later state that "F&SF didn't publish anything that was good and hard SF"?

    1. Yikes! That was a mistake constructing the graph. F&SF did indeed have one hard SF story, and Asimov's had 7, not 6. I've corrected it. That didn't change any of the other calculations, fortunately. Thanks for pointing it out!

  2. Greg,

    I've now read Blue Ribbon, Edited, For the Love of Sylvia City, All in a Hot and Copper Sky and Influence Isolated, Make Peace.

    It's possible I picked a bad sample, but the stories largely weren't interested in ideas. Edited and Hot and Copper Sky were largely love stories and the technology could have been replaced by a modern-day equivalent with no substantive change. Blue Ribbon read like it had been written in the 1960s. Influence Isolated, Make Peace and Sylvia City gave a hint to modern-day technologies, but - again - I didn't feel the authors had much to say about them. I'm possibly being unfair on Influence Isolated, possibly because I couldn't get over the fact that I'd met the main characters in 'The Water that Falls' except they were in the future this time. I'm going to speculate they're the author and his partner.

    1. This is probably one of those places where we just have different ideas about things. :-)

      With sufficient rewriting, you can eliminate the SF from just about any story. (Take Nightfall, set it in a desert where no one has ever seen rain . . .) That said, I don't think you can remove the speculative element from any of those stories in a trivial way.

      Perhaps a more important point is that I don't think "originality" is a reasonable criterion for rating stories. No other genre tries to do that, after all. So I don't care if the story has a 1960s style to it; all I care is whether I think it was executed well. Rewarding bad "original" stories is a very easy trap to fall into when you're reading ~1,000 stories per year; you get tired of seeing the same ideas in different forms and so you start rewarding really weird stuff that no one will enjoy just because you thought it was refreshingly different.

      As for the Chu story, I try very hard not to fall into the trap of judging stories based on other works by the same writer. I think every story should be judged on its own merits.

      So what do you mean when you say a story was about ideas? Do you mean "as opposed to being about people?" I have heard the definition that hard SF is about science/technology and only has cardboard characters, while soft SF puts the characters first, but I don't really buy that definition. I suspect that's not what you mean either, so maybe it'd help if you gave some examples of stories you do think are idea-based.

      I think the best hard SF stories take a plausible idea and make it real for us by showing how it impacts people. "Edited" does this spectacularly. We already have medicines that modify people's mental processes, so what's described in the story doesn't seem implausible. The story explores how such augmentation might really be used and how people might feel about it. The story gives no simple answers. We feel the narrator's pain, but we understand Wyatt's point of view too. I don't think the story would be as good if it obsessed over the details of how the editing process worked but left no space to make us care about the characters.