There's a New York Times piece about big-name authors choosing to self-publish -- this time, though an agency (ICM) that is offering a new package to its clients for same. The big-name author cited this time around is David Mamet, and there was all the usual stuff in the article about control over one's work and a bigger slice of royalties, which is all fine.
Very little discussion, though, of the fact that self-publishing for a name author is an easier deal because of the built-in name recognition — the Louis C.K. Effect, as someone else I know calls it. Just because he can self-fund and -vend his concert film doesn't mean you can, because there are tons more people who know who he is than there are people who know who you are. The vast majority of people who self-publish either find a comfortable niche of fans to address (a la successful webcomics, which set up a fanbase and then capitalize on it gradually), or remain obscure.
It doesn't help that, again, most of what gets self-published is stuff that would never clear customs at a conventional publisher anyway. I talked before about so much of it being third- or tenth-rate knockoffs of whatever's currently cluttering up the shelves and bestseller lists right now — which tells me most of the people doing this are simply saying "Hey, I could do that," and taking the thinking no further. Good for them that they were able to write and finish something, but there has to be more at work here than the mere urge to replicate what you saw on someone else's shelf.
Some of this is innocent enough. We see someone writing X and selling umpity-dillion copies, so we naturally assume that to write means to write something like X, and to be a successful writer definitely means to write something like X. So we copy it, or at the very least we take that as our model. I know I have done this: my past models were people who had accolades and approval from the sorts of circles I thought I would most want to be associated with, even if they weren't commercially successful (which made it easier for me to disdain the idea that selling lots of copies of something was "success"). I liked Ted Sturgeon and Phil Dick and bunches of other people. It took me years to learn how to stop writing like them, and to stop wanting to be them.
It's not that you shouldn't have any model to follow; that's impossible, nobody is ever truly self-made. Everyone starts somewhere. It's that you need to be able to see those models for what they are — just models, just the starting point, rather than the ultimate goal. When you do absolutely everything yourself, it becomes easy to forget you're trying to go somewhere in the first place.
There are many downsides to self-publishing, but I never imagined one of them would be how easily it allows you to arrest your own development.