All-Time Novels Polls

One topic I talked about with Charles Brown the last time I spoke with him (that May weekend trip to the Bay Area I blogged about earlier, and which CNB wrote about in his June issue editorial), was the idea of doing another, or perhaps several, polls of all-time best novels. A reader had written him suggesting an update, since the last all-time novels poll was done by Locus way back in 1998 (my Locus Index to SF Awards has the results posted here), though I’d done a similar poll for short fiction, anthologies, and collections on the website the following year (here).

As I’ve alluded too frequently in this blog, I’ve been compiling such polls and expert-lists and lists of best sf/f/h titles from reference books for some years now with the intention of supplementing the awards data in the SF Awards Index — and some of that additional data will see the light of day in the next update to the Awards Index, currently planned for early November after the World Fantasy Awards winners are announced. One element of this index expansion will be what might be called a ‘meta-list’ of best SF novels (and, separately, fantasy novels, and perhaps horror novels), in the manner of the Top 100 Books Meta-List that Newsweek magazine posted on its website a couple months ago. A definitive top 100 SF novels list, so far as one can be objectively compiled. (I’ve been iterating results of such a list for some time now, as new source lists keep appearing and I keep refining the relative ranking criteria.)

So new reader polls of all-time novels would be a valuable addition to this resource pool. I’m willing to set up such polls on the website and compile the results, and I’m also open to any suggestions about how best to set up such polls — keeping in mind that this might as easily be a series of polls as a single poll. Separate polls by decade? By novels published before 1980 and since? Separately by SF, fantasy, and horror? Should poll forms include seed titles of likely candidates, as the annual online Locus Poll ballots have done (but which some voters object to), or not? I have some ideas along these lines already, but suggestions are welcome — email me privately, or comment to this post.

7 Responses to All-Time Novels Polls

  1. I love these things. I am eagerly looking forward to your all time SF novel additions to the Awards database. I would suggest running it by decade and then take the top twenty of each year and run them all together. Is that too much work? If so do it as you suggested – up to 1980 and then after 1980. I'll vote in it however you set it up.

    Bob Blough

  2. Jonthan at Concatenation

    We had a bash at compiling such a list as in the form of a guide that came out as a paperback. We used very specific criteria. These included: print longevity (75 years or more of continuously being in print); winning a major fan-voted Anglophone SF Award (Locus Best SF or Hugo Best novel); being cited in the 2007 UK fan survey top 20. Also being a paperback novelization of a Hugo winning or Fest Fantastic Film surevy or Blackwood SF Film also got the book in.

    You would be surprised at how many books scored on more than one of these criteria.

    The guide has been reasonably well received. Though we had some stick from a few who objected to some titles not being included, but these folk missed the point. We devised criteria first (fan and affcionado popularity) and only then did we find out which titles scored.

    The result was Essential SF.

    We can let you have the list (without the entries) if you want?



  3. Here is one proposal which may be more ambitious than what you are seeking.

    The two requirements any All-Time Novels poll should satisfy are (1) that it be based of a representative sample of popular opinion and (2) that it advantage the opinions of those more widely read within the genre over those whose reading has been more limited. I say this because two problems that appear to have plagued various speculative fiction awards are the lack of broad input from readership beyond author and fan communities and the lack of broad exposure to works from different decades, from varied subgenres, and from authors with different chracteristics such as nationality, gender, ethnic background, etc.

    Some people might say that those two requirements are at odds, assuming that most readers will have had a relatively limited exposure to the entire body of speculative fiction works, thus giving the few popular works that they have read a great numerical advantage over the wider variety of works that would be chosen by those with the broadest exposure. I think the solution to this is to allow readers to rate as many works as they want and then weight more heavily the ratings of those with the broadest exposure, with some portion of the weighting based on sheer number of books rated, but part also based on the reader's exposure to works from different decades, from different subgenres, and from different kinds of authors. The only limit on how many works a voter could rate should be that she or he be able to remember some redeeming feature of the work that would get it rated at 1 on a 1-to-10 scale (10 being the most favorable rating). In other words, if someone has managed to read a hundred works in the genre that they hated and could find nothing of value in, we really don't want to hear about those, only about works that the reader thought did something right.

    This could be a lot more work than what you are envisioning, but I think it could be managed with sufficient community involvement. Getting the voters to rate all the works they have read and for which they have at least some positive memories would leave it to each voter to decide how much or little effort they wanted to put into rating different novels. Once all the votes (ratings) were in, any work with less than some number of votes (e.g., 50) could be eliminated to try to frustrate self-promoters who get their family and friends to vote for them (as in Amazon reviews for some books). Then the biggest task would be agreeing upon a practical (not too large and not too small) list of subgenres, and classifying the works that remain into no more than two of those. Given the concern that has been voiced recently about gender and racial diversity in awards, volunteers who share those concerns should be able to be found in order to classify the authors using whatever categories are considered appropriate. Date of first publication and nationality of author could be gotten from ISFDB or Wikipedia where those were not well known to the judges. Then each voter's ratings could be weighted based on the overall number of novels rated, perhaps the total number of authors read as well, the diversity of subgenres and authors, and so on. It could be that the weighting will be overcome by sufficient popular involvement of those with less genre exposure. If twenty thousand voters rate (for example) It, Twilight or Harry Potter the highest from a small set of ratings, this would no doubt overcome two thousand voters with much wider exposure rating (for example) Neuromancer, The Dispossessed or A Song of Ice and Fire the highest among many more ratings. And perhaps that is a valid result in a popular vote, however much it pains those whose appreciation of the genre rests on a much broader base of reading. Whatever weighting system is used should be clearly stated before the vote and not changed afterwards in order to frustrate an undesired outcome.

    - Paul Connelly

  4. Chris Riesbeck

    Polls like this come up about once a month in the science fiction areas of LibraryThing, with much initial enthusiasm, but, to my mind, no useful conclusion, or utility for readers looking for guidance. It's like listing the top 100 foods — Crackerjacks vs Beef Wellington.

    Rather than split by decades, I think more value would come from splits by categories like hard SF vs literary vs space opera vs humorous vs … Each category should be seeded with solid (and stellar, as judged by Locus reviewers) paradigmatic examples. Then I think you'd have something readers would find helpful.

  5. I personally distrust best-of lists however they're arrived at. Ursula K. LeGuin recently posted an interesting essay about literary awards that relates to this:

    She says it so well!


  6. I have to weigh in again, here. I've gotten my wide variety of reading SF by reading everyting nominated in the awards – Hugo, Nebula, John W. Campbell, Arthur C. Clarke, Locus (top 5), Internatonal Fantasy Awards, British SF Awards and all the top 100 lists I can garner – Pringle, Hartwell, Clute, Locus (both of them), and the list Dozois put up for SFFWA. It is because of all these lists that I have become rather well read in the best of the genre from Shelly and Walpole to Stephenson and Doctorow. I still have holes in my reading (I find stuff from 1900-1940 to be tougher to get through than others at times) but I still learn from each list put up. These are my way stations and often point me to works that others don't know much about. Dozois is making me read more Vance than I had, Pringle made me read more Ballard than I did before and each gives me something different from the others – THE BROKEN GOD from Clute, PARTICLE THEORY from Hartwell, QUEEN OF STATES from Arthur C. Clarke, THE ALTERATION from John W. Campbell, LIMBO from Pringle and tons of others. It makes me dip into different authors and find various gems that I had missed. If you don't try the new you experience smaller and smaller enjoyments. Various dishes make the feast much more exciting. So, maybe 100 top novels that are not considered the "Top". Take out the top that all consider to be the best (LEFT HAND OF DARKNESS, DUNE and STRANGER IN A STRANGE LAND, etc) and see what comes in after those – some are just as tasty or even better than some of he favorites -in my eyes.

    Bob Blough

  7. Le Guin is correct if we're taking the significance of these "Best" lists too seriously, but they can be useful as starting point. I had an experience like Bob's, but for "serious literature", after hitting a big list (in Family Circle magazine) of "books you should read before going to college" during the summer that I turned 16. I doubt very much that I would have gotten the exposure to all those "literary" works otherwise, since my reading was normally mostly SF, adventures and mysteries. But I plugged away and managed to read most of the books on the list. If I had stopped there and never read another literary work, assuming the list had contained everything I needed to read, that would certainly have qualified me as one of the "lazy-minded readers" that Le Guin decries, but in fact I continued to read other literary works over the years.

    And I can't help but believe that it's better for society to have a lazy-minded reader who has gotten through a couple of dozen quality works versus one who has never gone beyond books that are the mental equivalent of bubble gum. And certainly better than to have someone who never reads at all.

    Whether you can put a reasonable list together very easily is a different question. It seems easy to get a clique of self-appointed experts to enshrine the works that best exemplify their artistic pretensions. And it seems easy to just have a popular vote. The harder job is getting popular input while also recognizing the extra value that those who are widely read can contribute.

    - Paul Connelly