River of Gods


There are two forms of story that science fiction excels at. One is the ontological mystery, in which the revelation of the nature of the invented world is the end to which the story is reaching. What are the Overlords up to in Childhood’s End? Why has the solar system been shut off from the rest of the universe in Quarantine? What is the purpose of the universe-spanning structure that gives the title to Ring?

The other is the world tour, which gives the reader a guided tour of an invented society, planet, or universe. The epic scope of Dune: the sweep of history in Red Mars: the flowering of many forms of society in The Diamond Age.

These two forms are in opposition to each other. In the ontological mystery everything builds up to support the main idea, usually with straightforward characters and unnecessary detail cut away to avoid distraction. These idea-driven stories need to be short and punchy for their effect. In the world tour, it’s the accumulation of many details that make the impact, often with multiple narrative strands and complex relationships tying the characters to the world. These stories need length and grand sweep.

Something that I think is often a bit of a mistake is to try to mix the two forms of story: to explore a detailed and complex world and at the same time to have a revelatory mystery about the nature of that world. If the world-building is good enough then a story doesn’t need to have a punchline. In fact, the science-fictional revelation when it arrives may be disappointing or work against the realism of the world building that web before. Or the urgency of solving the mystery may simply overwhelm the other elements of the world, making them weak and forgettable by comparison.

For example, consider Ian McDonald’s River of Gods (2004). This novel builds a convincing picture of a mid-21st century Balkanized India, teeming with people and stories. It’s mostly set in Varanasi on the Ganges, capital of the nation of Bharat, and follows the intertwined stories of characters from across the social spectrum, from criminal-for-hire Shiv up to government minister-designate Shaheen Badoor Khan, to give a compound picture of a future country, rich in information technology but poor in water. There are some sharp character studies here, of people whose desires and fantasies lead them to disaster. Shiv’s illusions about honour among thieves are brutally dispelled. Mr Nandha, a ‘Krishna cop’ who hunts rogue AIs, fools himself into thinking that his marriage is a happy one. His wife Parvati imagines that she can run away without consequences. Tal, a ‘nute’ (neuter), believes that it has found true love. The head of the energy company Ray Power decides to retire to become a sadhu, and divide the company among his three sons.

With these characters and dramatic situations McDonald shouldn’t need to add an over-arching plot about ‘level 3’ artificial intelligences and their plan to escape into another universe. For me, this rather ordinary sf plot that takes over the second half of the book undermines the world-building that precedes it. It takes an open-ended space of possibilities and closes them all off with a neat little bow.

I was reminded at several points of James Clavell’s Noble House and I think the comparison is instructive. Like River of Gods, Noble House follows a diverse group of characters from many social strata to form a composite portrayal of a city-state: in Clavell’s case, Hong Kong in the 1960s. Both novels have a plot in which a businessman is facing the takeover of his company but who is saved by a timely sum of money from a mysterious investor (in Noble House the ‘tai-pan’ Ian Dunross of Struan’s; in River of Gods Vishram Ray, new director of one third of Ray Power). In Noble House the business plot is taken seriously, so Clavell goes to some lengths to set out the financial situation, who owes money to whom and when, and exactly why the failure to raise new capital will be disastrous. In River of Gods this is only sketched in: there’s no indication that McDonald is interested in giving us more than the barest of outlines because it’s really just a backdrop for his sf fireworks. But this short-changes the characters and spoils the dramatic set pieces. If McDonald doesn’t take the business plot seriously, why should we? When Vishram enters the boardroom to put his proposal to the shareholders, how can we appreciate the drama if we don’t understand what’s going on?

Similar things happen to the other storylines. For example, Tal’s story of love and betrayal leads to a hair-raising escape from death at the hands of an enraged mob,1 and then … well, it just gets shuffled to the sidelines to make way for the big dénouement. I would have liked a bit less of the ontological mystery and the big sf ending: I would have been happier to carry on with the tour.

  1.  This scene is written very effectively, but that just makes it the more problematic, since the faceless mob is an orientalist trope.