Okay, so, given the previous post, I think I have now actually have seen no fewer than seven possible definitions proposed for “soft” SF. Many of them overlap, but I think each is distinctive. Here they are:
- Sociological emphasis rather than an emphasis on technology
- Far future setting with handwavy technology or super-advanced magic technology.
- Science fantasy
- Really a different genre, like mystery, but with science fiction trappings
- Poorly researched SF with handwavy, science-y stuff rather than actual science.
- SF where the author doesn’t make any attempt to explain the science behind the setting. I’m classing this as different from poorly researched background science.
- All SF where the author did not actually have to do math or physics calculations in order to write the story.
I’m with the rest of you: the term “soft” is useless when definitions are so wildly divergent. I mean, lots of the above definitions intergrade. Why does Pern seem like fantasy to so many people, including me? Because the science-y parts are silly and involve lots of handwaving. Those dragons can so fly! Time travel is totally plausible! To me, Pern is a great example, maybe the best example ever, of science fantasy rather than science fiction.
I wouldn’t say anything about Pern is poorly researched, which is why “handwavy” is not the same as “poorly researched.” Anne McCaffery didn’t do bad physics when she tossed her dragons into the air; she didn’t do physics at all. She did magic. She wanted dragons, so she wrote dragons. Her books are no more SF than the Temeraire series. They’re both fantasy, even though Pern has science-y trappings here and there. In fact, Temeraire is just as handwavy when it comes to science: Yeah, of COURSE there are enough cows in the UK to feed a lot of dragons. Absolutely! Oh, hey, actually, didn’t Naomi Novik put dragons in South America too? Yeah, no, the resource base for nearly the entire continent is dreadfully restricted because so much of the land area sits on top of either the Gayana shield or the Brazilian shield, which are extremely nutrient-poor rock formations. There’s no way you’d be able to support a lot of giant carnivores in Brazil. That’s definitely right out.
For years and years, when I mean “sociological science fiction,” I have said “sociological science fiction.” Even though “sociological” has six syllables and is not that easy to spit out in a hurry, there aren’t any other terms that distinguish this category of science fiction from everything else. As many commenters said, there’s nothing handwavy or soft about really solid sociological SF, such as Woman of the Iron People. There’s a ton of sociological SF, some of it great and some, of course, less well-researched or less good in other ways, so if you’re taking “hard” as thoroughly researched and carefully thought through and “soft” as poorly researched or badly understood, then this category definitely ranges from hard to soft.
But I don’t like to use “soft” as a derogatory term! I guess I don’t really like to use it at all because the term is often understood as derogatory, and I don’t approve of that usage. I’ve got no problem with well-done science fantasy! I like Pern, and Sharon Shinn’s Angel series. I don’t grumble about bad science when I read stories like that. Grumbles about the square-cube law miss the whole point of those stories.
My personal conclusion: If you mean sociological SF, say so. Use “hard” when you mean real-world physics is central to the setting or the plot and don’t use “soft” at all. We have plenty of subgenre terms that are much better! Time to let that one die.