Why novels should fail

I said something during my thesis defense (now several months past) that surprised me at the time, and which still to some extent surprises me now, though I think I have always been dimly aware I believed this. I am waiting for my wife to finish an interview. I am unhappy that the blog did not have any updates yesterday (we were driving again). And so I think I'll mull this publicly.

What I said was that I believed novels should fail. This surprised me because it was the goal of my defense to persuade my committee that I had succeeded. Not in making a perfect book, of course -- no one who teaches writing or reading could be that naive -- but in making a book that was, on some fundamental level, complete. Naturally I implied that I had done the opposite. My perverse need to say the opposite of what the moment requires is what gets me into so much trouble sometimes (but it's also where my books come from, so I can hardly quash it).

What I meant was this: a great novel should be so ambitious that it is literally and clearly beyond the powers of its author. It's fine to write something you know you can write, but I rarely care to do it. My preference has always been to write things I didn't believe I could actually do, whether because they were too stupid to be good (my novel about the murder of a unicorn investigated by the Atlanta Snuff Films Unit) or because they were too good for somebody as stupid as me (my novel about the atomic bombs Fat Man and Little Boy as people). And these are the books I prefer to read: the ones where you can see the author has clearly run well off the cliff, and now the question is how far through empty air his legs can take him. The inevitable fall is as good as the running.

My best example of this is Stanislaw Lem's novel Solaris, a favorite of mine, which attempts something that it very clearly knows cannot be done. Solaris is best known as the story of how a husband (portrayed once by George Clooney) comes to a space station where, due perhaps to the mysterious planet below, his long-dead wife is revived. They try to live together again. This happens, but the novel is really and finally about the challenge of imagining another mind: that of another person (this isn't too hard) or that of an alien being (in this case, a planet or an ocean: quite difficult). The revived wife is maybe, possibly, probably a way of attempting communication, the planet's best attempt at speaking to the husband.

There are figures for the mystery of others from the novel's outset -- an African queen, for instance, who roams the station mostly naked. She's a thoroughly imaginary figure, stereotypical and yet ineffable, and we never know how she came to be: by the planet's power, presumably, but the man she was made for is dead. Why would he want her? Why did this image hold such power for him? We'll never know. There is also the man who lives with what is very probably his dead son, and another who lives with some reminder of a terrible crime. We never know the crime. We never see the reminder.

What makes these figures so beguiling is partly our ability to chart the trajectory by which one might arrive at them. We don't know how one becomes obsessed with an imaginary African queen, but we know what one is, and we can certainly imagine several explanations that begin to bridge the gap. We have some idea what it would be to mourn a son, even if we've never had one. And as for crimes, we may never learn about this particular character's crime, but we know what they are, as a category. We know how they feel. We know, more or less, what might lead us to commit one.

So there is one level of mystery. And this is more or less manageable, as a project: people have written about loss, about the eroticism of racism, about mourning, about criminality, before, to varying degrees of success. The book would be good (quite possibly great, with some additional pressure) from these elements alone. But then there is the planet beneath. And this is where it gets really wild. 

The book knows that when it comes to living planets or oceans, we are supremely unlikely to ever figure them out. In the face of that much difference, communication becomes seemingly impossible. And yet the novel has to portray the living ocean or planet. It has to understand it well enough to write it as a sort of character. It approaches this problem by several strategies -- by the analogy of trying to understand other people, by the use of in-text essays about the planet by fictional figures, and by showing the planet's behavior. The protagonist cannot understand the strange sculptures the planet offers, cannot even know if these are really meant to communicate. Nor can, presumably, the planet understand the protagonist. And yet they keep trying. And the novel does its best to make it all persuasive, real, beautiful, through pages of description and strangeness and horror. You could call it a failure. But you can't deny its beauty.

I want to make something like that.

What is your favorite failure? Your own, or someone else's?