October 2008

Paul Kincaid

Science Fiction Skeptic

The Canon... And Why It Isn't

The sf world loves a list, so here’s another one to get everyone arguing. As promised last month, my suggestions for canonical works of sf before the awards kicked in:

Utopia by Thomas More, which introduced into modern literature one of the staples of sf, the perfect society.

The Man in the Moone by Francis Godwin, which introduced the notion of travel to another world by mechanical means (okay, a conveyance pulled by geese, but that’s better than Cyrano de Bergerac’s evaporating dew).

The Blazing World by Margaret Cavendish, Mad Madge, Duchess of Newcastle and the first woman sf writer, who introduced parallel worlds and a character who conversed with her author.

Gulliver’s Travels by Jonathan Swift, do I need to say why?

The Journey of Niels Klim by Ludvig Holberg, the first significant fictional exploration of a hollow earth.

Micromegas by Voltaire, which first brought aliens to visit the earth.

Frankenstein by Mary Shelley, which introduced man the creator of man.

After London by Richard Jeffries, the British catastrophe story is born.

Twenty Thousand Leagues Under the Sea by Jules Verne, the romantic mixture of unusual places with unusual technology, and sf can never be the same again.

Strange Case of Dr Jekyll and Mr Hyde by Robert Louis Stevenson, the next step on from Frankenstein, introducing the notion of the doppelganger that would be so prevalent in sf.

The Time Machine by H.G. Wells, which introduced the notion of travel through time by mechanical means.

Ralph 124C41+ by Hugo Gernsback, scientifiction is born, and with it the pulp tradition that shaped so much 20th century sf.

We by Yevgeny Zamiatin, without which we couldn’t have had 1984.

The Skylark of Space by E.E. ‘Doc’ Smith, which gave us space opera.

Last and First Men by Olaf Stapledon, explored the scope and scale of science fiction in ways that fed directly into the work of Arthur C. Clarke and others.

Beyond This Horizon by Robert A. Heinlein, which introduced not only one of the most significant sf writers of the century but also many of the political ideas that informed all his fiction.

1984 by George Orwell, because we have to have something we can point to when we try to argue that sf has literary respectability.

Foundation by Isaac Asimov, does for American sf pretty much what Stapledon did for British sf, looking as huge swathes of the future in terms of historical processes.

The Day of the Triffids by John Wyndham, set the tone for at least three decades of British sf, and still managed to make the genre respectable with walking plants.

There, is that controversial enough for you?

Actually any list of canonical works is bound to be partial and idiosyncratic, but I suspect that, apart from the inevitable arguments about the works I’ve missed off the list, this will stand as a pretty good starting point for the works that shaped the genre up to about 1950.

Except …

Except that any canon is part definition and part history. A canon has to consist of works that shape our understanding of what the genre is, and works that have in themselves shaped the course of the genre. But we have no universally agreed definition or history of sf. If the history of science fiction starts at point X, anything before X cannot be canonical, because it cannot be science fiction.

Thus Adam Roberts argues that sf began in 1600 with the execution of Giordano Bruno, which excludes Utopia.

Brian Aldiss claims Frankenstein as the first work of science fiction, which excludes yet more.

Gary Westfahl argues that the genre only begins when it has a name, when people are consciously writing within the genre, and that puts the start date around the time of the launch of Amazing in 1926. So that excludes practically everything on my list, up to and including Gernsback’s own novel. (And actually, of the later works at least three, the Zamiatin, Stapledon and Orwell, were not consciously written with any genre consciousness, so they probably need to be excluded also.)

So the canon is dodgy in historical terms, and it’s not much better in literary terms.

To start with, Utopia isn’t even fiction, it was written as what we would now call political philosophy. But at least it’s literate, which is more than can be said for some of these titles.

Would you really want science fiction to be represented by something as appallingly written as Ralph 124C41+ or as jejune as The Skylark of Space? Yet these are important historically, they have earned their place in the canon simply in terms of what they signify in the development of the genre. But they are bad books by any modern critical standards.

Which, of course, brings us to the thorny issue of what a canon is for. Is it purely an historical record, or is it meant to represent genre high points? If it is a record of literary quality, then the canon will change constantly as tastes and standards change. So, while I would be quite happy to drop Gernsback and Smith from the list, how long before their clunky prose means we drop Asimov and Heinlein also? But the two are, quite simply, among the most significant sf writers of the mid-twentieth century, it would be nonsensical to exclude them from the canon. So literary considerations have to take a back seat to impact or ideas or historic significance.

And yet, the canon should be something we’re proud of, something that represents science fiction, something that non-sf readers should be pointed towards in order to gain an understanding of what the genre is and how it got there. And you look at that list and think no-one but a completist is going to want to go anywhere near some of those works. It’s a dilemma, and there may not be a solution.

Then there are the personal quirks, which are going to be a factor in any list chosen by one person. Let’s see if I can untangle a few of them.

To keep the list within manageable bounds I limited it to one book per person, but that means we miss The Last Man by Mary Shelley which probably pre-empts After London, and The Star Maker and Odd John by Stapledon, and at least four other novels by Wells, and so on and on.

The list is anglocentric, because it comes from my own personal knowledge and experience of science fiction. I’m sure if I knew more I’d include more Russian and French sf, and maybe something, anything, from Germany or Italy. By the 1920s sf had become primarily an American genre, so maybe there should be more American writers in there.

Which ties in with another objection: for various reasons I limited my list to novels, but from the launch of Amazing in 1926 and for at least the rest of the period covered by this list, science fiction, particularly in America, was really centred more on the short story than the novel. The most important figure in sf from 1939 until I close my list was John W. Campbell, and you can only reflect that by including some of the stories he edited. Stories like "Nightfall" by Isaac Asimov and "The Cold Equations" by Tom Godwin really are of immense importance in the mid-century development of sf. And look how, even today, writers of robot stories still either automatically assume or explicitly reject Asimov’s three laws of robotics. The latter part of the list should have been cluttered with short stories.

So it goes, as Kurt Vonnegut kept saying (Player Piano, another contender for the list). Any canon, from whatever source, is going to be subject to some variant on these historical, literary and personal objections. The best we can do is probably say these are some of the historical works of science fiction that anyone interested in the development of the genre might want to pay attention to. But that’s so much less sexy than calling it the canon.