The best advice I ever got on this was to be sure to throw as many rocks at what you believe with as much force as you can muster, and see what remains standing at the end.

Shaw, my favorite in this area, always gave the side he opposed really good spokespeople. Brecht who slipped into straight propaganda more than Shaw did, (I'm leaving aside Shaw's non-fiction which of course WAS pure propaganda) was at his best when he undermined his message. His view of Galileo was to consider him a coward, but there has never been a presentation of the play that did not make him the hero. Mother Courage was intended to be the portrayal of a really awful person, but the audience or reader usually comes away saying "what a brave woman".

Not the only approach. Often a good way to explore ideas is to make sure none of the characters are ways to advance an idea or point of view, but let the world in which they live make the point. All fiction involves world building (IMO) not just fantasy, Sci-Fi or Specualative fiction. So let the world you build reflect the what you believe is the reality of how worlds work. Then make your characters real people, trying to get by in that world - not symbols or spokespersons, or caricatures.

Quick followup on the above - Michael Moorcock -- an intensely political writer, but it seldom show in his best writing, because there the politics is in his world building, not voiced by his characters.

Warning:Moorcock is one of the most prolific writers ever, and his output includes an immense amount of hackwork. You have to take the his best, not his worst or average.

So maybe he just never really bothered TRYING.

I have some dim memory of an essay by Asimov, wherein he said that, to him, characterization was the least important part of a story. So he may preferred putting the effort into building another story around another idea instead of working for more fully realized characters in something he felt was already done.

So it wasn't my perception, some SF authors really do try not to write for characters...seems I must have gotten to all the "wrong" ones, in terms of making me a fan.

SF authors really do try not to write for characters

I think it's safe to say that some authors, period, don't write for characters.

But SF, if your concept is cool or strong or mindbending, there are other valued characteristics that can make you, with otherwise good writing, a big name.

I always thought that Brecht, while trying hard to uphold his theatrical ideals of distancing the audience from the material to incite action rather than emotion, created extremely emotional plays. This might have something to do with the values of the time he was working in, as he was reacting directly to realism/naturalism, where everything was so authtically perfect-- and I find realistic theatre frankly, boring, because we have movies. Back in his day, not so much.

I don't mean to say there aren't other writers in other genres that don't write stock characters or get by on violence or shock value or... but somehow, early on, I got prejudiced against SF from, probably an Asimov that hit me "wrong" personally. For instance. But there are probably many different styles of SF writing too.

I think that the nature of SF as a genre makes it more idea-driven than not. Which isn't to say that there aren't character- or plot-centric SF novels, but the act of creating an SF universe (whether it's future or alternate-now or completely made-up) for your characters and plot to take place in forces you to answer, or at least pose, some pretty big What Ifs. You have to decide what kind of governments exist in your world, what races/species exist, and whether or not they get along. What technological powers do humans (or whoever) now possess, and how do they use them? Etc etc.

On a semi-related note, I was at a Q&A with the screenwriter/director of Syriana over the weekend, and he mentioned Tolstoy as an influence (he did admit that that was about as pretentious as it gets), and talked about Tolstoy's 4 rules for writing. The 4 most important things for Tolstoy, in order, were (1) transitions, (2) backdrop/worldbuilding (3) character, and (4) story. Discuss.

He ranked transitions first? that's kinda odd. I mean, sure, transitions are important, but you can have a perfectly fine story if you have crappy transitions but an interesting plot and well-drawn characters. Bad transitions will not make or break a novel for me.

I think that the nature of SF as a genre makes it more idea-driven than not.

I'd say there was a time in history when that was true, but that it's not necessarily so any more. A fair amount of SF since the New Wave has been interested in what the characters do, and setting and world-building are just context to create the situation for the characters. For example, Joanna Russ's short story "When It Changed" is all about the patriarchal snottiness of the astronauts who arrive on Whileaway and immediately assume their own dominance. The idea of a female-only society is sketched out in extreme brevity, just enough to make clear that these astronauts are way out of their own depth: the real point is drawing out the female characters' unease and distrust at the stupid patronizing assumptions. (It's also hella depressing.)

The 4 most important things for Tolstoy, in order, were (1) transitions, (2) backdrop/worldbuilding (3) character, and (4) story. Discuss.

I don't put that much stock in transitions, or anyway, I don't run across that many really awkward ones. I would also posit that if worldbuilding is not part of both character and story, there is no reason to be worldbuilding in the first place. Sort of surprised "theme" doesn't even make his list, especially considering we were talking about "how to make a statement without making the story into boring propaganda" just yesterday.