In one of the threads Erik asked why some of us had recommended that Sean focus on short stories for a while rather than novels. It's a topic worth talking about at some length as it's advice I give to every aspiring writer these days—if you can write short stories, it's the best available way to build your career. There are a number of reasons for this.
The market: In science fiction and fantasy the big publishers are collectively breaking something between 20 and 50 new writers per year. I'm not sure of the exact number, both because it varies and becuase the editors I've talked to aren't terribly specific, but it tends to be on the low end of that. In short stories, the numbers run into the low hundreds and there are venues that are open solely to new writers or that hold a fixed number of slots open for new writers. On top of that, the competition is lower. In the middle tier of short story markets a writer is competing against considerably fewer writers for a significantly larger number of available spots.
Diversity of story: The short markets are also willing to take more risks on the really bizarre and the stuff that crosses genres. This is a twofer. It lets a writer have more room to experiment and it can be used to establish that there's a market for the outre. Short story readers write letters to the markets and those often get published. If something with a different flavor draws a lot of attention at the short story level, the book editors will pay attention to that.
Failing spectacularly: This is directly related to the diversity issue. I came into writing from theater so I'm used to thinking in terms of rehearsal and seeing that as the opportunity to fail really spectacularly without consequences. Short stories can be like novel rehearsals. They give you a chance to try out effects and improvisations that are either going to end in something extraordinary or in total disaster without the consequences of attempting the same feat in a novel. It's much easier to walk away from the smoking wreckage of short story.
Time into product: Let's say that 10,000 words of text takes a fixed amount of time to write, whether it's for a short story or novel. I know, it doesn't. But for the sake of argument let's say that it's at least close. Let's even assign it a time. Call it two weeks. Some writers are a good bit faster than that, other writers will be much slower, but it's within the realm of reason. That means that a novel (arbitrarily 100,000 words since that's slightly on the high side of what the publishers are looking for in a new writer at the moment) takes about 20 weeks to write. Let's say a short story is 5,000 words, again arbitrary, but with some basis in fact since that's the high end for a lot of markets. So, one week per short, or 20 shorts in the time it takes to write a novel. That's 20 chances to sell that first piece of writing and start building a reputation vs. 1.
Splash factor: George RR Martin has already said this better here, so I'll quote, of his first novel: it was not just another novel being thrown out there with all the other first novels, to sink or swim. It was "the long-awaited first novel," and that makes a very big difference in a career. And: A novel may pay more initially, but if your concern is to actually build a career, you do yourself a lot of good by building a reputation with short stories first.
Finally, learning curve: And I actually think this is the most important reason of all. In my own career, I wrote three novels before ever trying short stories. I'm not a natural short writer and when I started out it was like pulling teeth to get them down on the page. Also, I wrote a lot of things that were not shorts, though they were genre and of the right length. Mostly, they were lost chapters. However, I persisted, writing nothing but shorts for three years. In that time I wrote something like fifty shorts, more than half of which have now seen professional publication or are forthcoming, and a gazillion fragments for a total of something like 250,000 words. I created hundreds of characters and dozens of worlds. I had to come up with something like a 100 plots (there were a lot of fragments) and write a huge number of beginnings, middles, and endings. And all of it had to be short, there was no room for wasted words or blind alleys. I learned a ton about the craft of writing and about idea generation, and the vast majority of it is also applicable to novels. Would I have learned as much from writing 2-and-a-1/2 novels? Possible, but highly unlikely.
Of course, none of this matters if you're one of the fraction of authors who simply can't write shorts. But if you can, it'll do you a world of good over the long run.
And now I've talked way too long when I should be working on The Black School, so I'll open the floor to comments and questions. What do you write? At what length? Why? Are you a novelist first last and always? A short story writer? Bitextual? Do you dabble in the truly outre. . .poetry? I do, and again, I've learned things there that apply to my other work.