The Locus Sci-Fi/Fantasy poll: number crunching
Locus Magazine has been running a series of polls on the best speculative fiction of the 20th and 21st centuries. The lists of novels came out a few weeks ago, and since I am incapable of looking at a list of books without wondering how well women authors are represented, I downloaded the lists and crunched some numbers. I was especially curious because, unlike most polls along these lines, Locus ran the science fiction and fantasy polls separately, and I was curious to see how that would affect the results as compared to the Tor.com and NPR lists from 2011.
Some big, fat caveats before I start. Although Locus asked their readers to vote for individual books instead of whole series, some people voted for series anyway, which created some double-votes. Locus also ran the entire project via write-in and didn't do any kind of corrective or normalization work -- some books are in both the sci-fi and fantasy lists, some books are on the lists for both centuries, and some books are duplicated with author name misspellings, title variants, etc. (Locus updated the lists on their site to correct for the wrong century error, but I had already started working with the uncorrected data, and I decided it didn't make enough of a difference for me to want to start over.) Finally, I only included books ranked above zero (books that received more than one vote), both to make the dataset more manageable and because so many of them were duplicates of titles higher on the list.
Okay, all that said, how does it look?
20th Century Lists:
Science Fiction - Out of 389 titles, 65 books were by women*, or 17%. Only 8% of the books in the top 50 were by women. Out of 191 individual authors, 38 were women, or 20%. The woman with the most books on the list was Lois McMaster Bujold, with 7. 12 women had more than one book on the list, as compared to 66 men.
Fantasy - Out of 374 titles, 79 were written by women, or 24%. In the top 50, 16% of the books were by women. Out of 194 authors, 47 were women, also 24%. The woman with the most books on the list was Ursula Le Guin, with 7. Of the authors with more than one book on the list, 14 were women and 59 were men.
21st Century Lists:
Science Fiction - Out of 183 books on the list, 34 were by women, or 19%. Of the top 50, 16% of the books were by women. Out of 109 individual authors, 28 were women, or 26%. The woman with the most books on the list was Gwyneth Jones, with 4. There were only 4 women with more than one book on the list, as compared to 30 men.
Fantasy - Out of 169 books on the list, 53 were by women, or 31%. Of the top 50, 29% were by women. Out of 108 authors, 39 were women, or 36%. The woman with the most books on the list was, once again, Ursula Le Guin, with 6 (only Terry Pratchett had more). 7 women and 19 men had more than one book on the list.
I had expected, going in, that women would be better represented in fantasy than in sci-fi, and better represented on the 21st century list than the 20th century one, and both of these expectations held true. I had thought about combining the genre lists to do an overall analysis, but there was too much duplication in the data for that to make sense. A lot of books were on both lists -- "The City and the City" by China Mieville even made both top ten lists for its century -- and I decided I wasn't up for weeding them all out. I have always found separating these genres to be difficult, and it's getting harder as the genres fragment and spread. There are questions, too, about the gendering of the genres. Are women more likely to be pegged as fantasy writers and men as science fiction? Take Anne McCaffery -- the Pern books are technically science fiction, but since they're about dragons, people often put them in the fantasy category, and the vote for her books was split over the two categories. (Now I want to find all the books that are listed in both categories for their years and see if any patterns emerge.)
Naturally, I was curious about how these lists would stack up to the Tor and NPR polls. The NPR list of 100 included 15% books by women, with no women authors in the top 10 and only one in the top 20; every single Locus list did better than that, even the 20th century sci-fi list. But then, my issues with recent NPR book polls, their choices and methodology, are pretty well documented. On the other hand, in the Tor poll, 38% of the books that received at least one vote were by women, 24% of the top 50. Since I excluded books that received only one vote from my analysis, this isn't an exact comparison, and the Tor poll is 21st century only; when you compare only those two lists to the Tor data, it's not too far off -- the fantasy list is about the same, the sci-fi list somewhat lower, and maybe the difference would balance out, especially if I could find a way to include the single-vote books.
I'd need data from many more polls to be sure, but it interests me that, for all their issues, free-for-all polls like the Locus and Tor lists tend to provide better representation of women authors than curated polls such as the NPR list. I'm not sure whether that's good or bad, in the end.
I would like to have some grand pronouncement to make here about What It All Means, but at the moment, I really don't. But I do feel that it's important to get these kinds of numbers out there, to remind ourselves that lists like this come with built-in limitations and prejudices**, which will never go away if we don't examine and talk about them.
*Not including two books co-written by a man and a woman.
**Of all kinds, not just gender, of course. I don't feel qualified to run these lists through other representation metrics, but race, for one, would sure be interesting.
This entry is also posted at http://owlmoose.dreamwidth.org/613922.html. There are currently
comments on DW.
Some big, fat caveats before I start. Although Locus asked their readers to vote for individual books instead of whole series, some people voted for series anyway, which created some double-votes. Locus also ran the entire project via write-in and didn't do any kind of corrective or normalization work -- some books are in both the sci-fi and fantasy lists, some books are on the lists for both centuries, and some books are duplicated with author name misspellings, title variants, etc. (Locus updated the lists on their site to correct for the wrong century error, but I had already started working with the uncorrected data, and I decided it didn't make enough of a difference for me to want to start over.) Finally, I only included books ranked above zero (books that received more than one vote), both to make the dataset more manageable and because so many of them were duplicates of titles higher on the list.
Okay, all that said, how does it look?
20th Century Lists:
Science Fiction - Out of 389 titles, 65 books were by women*, or 17%. Only 8% of the books in the top 50 were by women. Out of 191 individual authors, 38 were women, or 20%. The woman with the most books on the list was Lois McMaster Bujold, with 7. 12 women had more than one book on the list, as compared to 66 men.
Fantasy - Out of 374 titles, 79 were written by women, or 24%. In the top 50, 16% of the books were by women. Out of 194 authors, 47 were women, also 24%. The woman with the most books on the list was Ursula Le Guin, with 7. Of the authors with more than one book on the list, 14 were women and 59 were men.
21st Century Lists:
Science Fiction - Out of 183 books on the list, 34 were by women, or 19%. Of the top 50, 16% of the books were by women. Out of 109 individual authors, 28 were women, or 26%. The woman with the most books on the list was Gwyneth Jones, with 4. There were only 4 women with more than one book on the list, as compared to 30 men.
Fantasy - Out of 169 books on the list, 53 were by women, or 31%. Of the top 50, 29% were by women. Out of 108 authors, 39 were women, or 36%. The woman with the most books on the list was, once again, Ursula Le Guin, with 6 (only Terry Pratchett had more). 7 women and 19 men had more than one book on the list.
I had expected, going in, that women would be better represented in fantasy than in sci-fi, and better represented on the 21st century list than the 20th century one, and both of these expectations held true. I had thought about combining the genre lists to do an overall analysis, but there was too much duplication in the data for that to make sense. A lot of books were on both lists -- "The City and the City" by China Mieville even made both top ten lists for its century -- and I decided I wasn't up for weeding them all out. I have always found separating these genres to be difficult, and it's getting harder as the genres fragment and spread. There are questions, too, about the gendering of the genres. Are women more likely to be pegged as fantasy writers and men as science fiction? Take Anne McCaffery -- the Pern books are technically science fiction, but since they're about dragons, people often put them in the fantasy category, and the vote for her books was split over the two categories. (Now I want to find all the books that are listed in both categories for their years and see if any patterns emerge.)
Naturally, I was curious about how these lists would stack up to the Tor and NPR polls. The NPR list of 100 included 15% books by women, with no women authors in the top 10 and only one in the top 20; every single Locus list did better than that, even the 20th century sci-fi list. But then, my issues with recent NPR book polls, their choices and methodology, are pretty well documented. On the other hand, in the Tor poll, 38% of the books that received at least one vote were by women, 24% of the top 50. Since I excluded books that received only one vote from my analysis, this isn't an exact comparison, and the Tor poll is 21st century only; when you compare only those two lists to the Tor data, it's not too far off -- the fantasy list is about the same, the sci-fi list somewhat lower, and maybe the difference would balance out, especially if I could find a way to include the single-vote books.
I'd need data from many more polls to be sure, but it interests me that, for all their issues, free-for-all polls like the Locus and Tor lists tend to provide better representation of women authors than curated polls such as the NPR list. I'm not sure whether that's good or bad, in the end.
I would like to have some grand pronouncement to make here about What It All Means, but at the moment, I really don't. But I do feel that it's important to get these kinds of numbers out there, to remind ourselves that lists like this come with built-in limitations and prejudices**, which will never go away if we don't examine and talk about them.
*Not including two books co-written by a man and a woman.
**Of all kinds, not just gender, of course. I don't feel qualified to run these lists through other representation metrics, but race, for one, would sure be interesting.
This entry is also posted at http://owlmoose.dreamwidth.org/613922.html. There are currently
comments on DW.