Here's a fun question for you: What is it that sets science fiction apart from every other genre of film? Since you're all smart people, you can assume that the most obvious answer -- "because it features aliens and flying saucers, duh" -- is not the answer I'm looking for. No, I'm looking for something more subtle than that, something that makes a nod both to the fundamental nature of film, and the fundamental nature of science fiction in film.

By way of an answer, let me first say what every genre of film has in common: Classics. Comedy has Some Like it Hot and The Philadelphia Story (to name but two). Westerns have High Noon and Unforgiven. Suspense and horror? Psycho and The Exorcist. And so on. No matter the category, you end up with a list of films that are not only classics in the genre, but also excellent motion pictures, period.

Science fiction has its classics, too, from Metropolis to 2001 to Blade Runner. But what makes science fiction different than every other genre of film -- what makes it unique, for better or worse -- is that a strangely high percentage of the classics of the genre are not good films; some are structurally flawed in major ways, while others are just plain awful.

Examples, you say? Take the curious case of 1954's Gojira -- or Godzilla, as it is known here in the U.S. Any objective observer of the science fiction genre of film has to acknowledge its significance: It's the progenitor of an entire subgenre of mutated monster films (called kaiju eiga in Japan) whose influence is actively felt today -- this year's Cloverfield is a direct descendant. And, of course, Gojira's monster is also a pop cultural symbol of Japan, not just for the kitsch value, but because it embodies the entire nation's terror of (and strange ambivalence toward) nuclear energy and the destruction it creates. Freud would have had a field day with this giant reptile.

And yet, Gojira the movie is really and genuinely awful. Bad plot, bad acting, bad science (both in the implausibility of a 150-foot reptile and the method of getting rid of it -- oxygen destroyer?), and of course, some poor bastard stuck in a rubber suit squishing cardboard sets counts as bad production values. The film only got worse when it came to America and Raymond Burr was summarily shoved into the flick so that American drive-in viewers could have a white man to ignore as they necked in the back of dad's Buick.

Gojira and its dozens of sequels are so bad that if there was any honesty in the world, people would have to admit that the universally-reviled 1998 Roland Emmerich version is actually the best-written, best-acted, best-produced Godzilla film ever made. No one will ever admit this, ever (except for me, and maybe, if he's drunk enough, Roland Emmerich), but there it is. Gojira is a classic science fiction film. Gojira stinks.

Another example: Logan's Run. This 1976 film about a society where life had a 30-year expiration date came out at the tail end of the dystopic era in SF film that started with Planet of the Apes and had the door slammed on it by Star Wars. It's possibly the least of the major science fiction films of this era (which also include Soylent Green, The Omega Man and A Clockwork Orange), but it was a big hit, and got its share of critical accolades ("2001 it's not, but it has class," opined Roger Ebert). I know folks who argue quite seriously and (alas) accurately that it's a legitimate classic of the genre. But it was silly then, and now -- two years past its crystal-flashing birthday -- it's actually painful to watch, not only for its script, but for its entire aesthetic, which marries the hoary science fiction cliche of tunic-wearing with feathered '70s hair.

As a final example, I could pull out Star Wars, but I have to adhere to a word count... Suffice to say, everyone who ever winced at Luke Skywalker whining about wanting to go to Toshi Station groks that there are many things about that move that are just not good, never were and never will be.

It's strange that such legitimately bad films are considered classic, but there are reasons. The first, I regret to say, is that for a very long time -- from just after 1927's Metropolis through 1968's 2001 -- the number of truly good science fiction films could be counted on one's fingers. So the entire genre is graded on a curve. But the other thing is that science fiction films -- like science fiction literature -- value the idea over the idea's delivery system. So if you deliver a 150-foot reptile who is the embodiment of the mid-20th Century fear of nuclear annihilation (like Gojira), or tap into the Baby Boomer terror of the death of its own beautiful youth (a la Logan's Run), you can get away with letting a lot of other stuff slide, like plot, acting and production design.

The real question is: Is this willingness to value the idea at the expense of almost everything else a strength or a weakness of the science fiction film genre? That's a question I'm going to leave open for you.

Winner of the John W. Campbell Award for Best New Writer, John Scalzi is the author of The Rough Guide to Sci-Fi Movies as well as the novels Old Man's War and the upcoming Zoe's Tale. His column appears every Thursday.