The Genius Computer Hacker: How Good Does Research Have To Be?

I've recently finished reading The Girl with the Dragon Tattoo, the first in Stieg Larsson's Millennium trilogy. It was an engrossing read, although I suspect I wasn't quite as taken with it as many have been. Lisbeth Salander, she with the tattoo, is certainly a great character, and the book deals well with some large and important themes. I'll undoubtedly seek out the other two volumes at some point. I did, however, feel that the book would have benefited from further editing.

Anyway, this post isn't a review. What I want to talk about is believability and research and, as an example, the treatment of computers and computer hacking in the book. Whenever I read that computers play a significant role in a novel, I experience a slight sinking feeling. Almost invariably, as I read, it becomes clear that the the writer doesn't really know what they're talking about. Sometimes it's a basic thing that gives the game away, like muddling memory with disk space. Sometimes, as with Larsson, the treatment is more convincing, but still flawed. I won't bore you with the technical details, but I'll mention an example. At one point Salander hacks into someone's laptop and reconfigures it so that all its hard-disk activity is redirected to a server on the internet. The laptop's user then, hilariously, carries on, not noticing any difference. Not noticing, for example, that his machine has suddenly become a hundred times slower, or doesn't work at all if there isn't a wireless signal.

Now, I only notice stuff like this as I happen to know something about computers and software. My non-writing alter-ego, the one that actually pays the bills, is a software developer. For other readers it will be other things. My wife, for example, is a doctor of microbiology. She generally finds the treatment of computers in literarure perfectly convincing, but sees significant flaws whenever microbiology is involved. I, on the other hand, don't know enough about that subject not to find its handling perfectly believable.

And my question is, does this matter? How rigorous does research have to be? Are we, as writers, trying to create absolute technical accuracy in all areas we cover, or is it enough to just fashion an enjoyable illusion, convincing enough for most readers? It's significant, I think, that I was still engrossed by Larsson's book even while a part of my mind was spotting the technical flaws. I still wanted to read on, find out what was going to happen. I still "believed" in what I was reading.

The danger with making technical mistakes is that you might alienate some expert readers. But if what you write is believable enough for most people, does it matter that it's actually "wrong"? I tend to think not. We're trying to create art here, not produce text books. I write a lot of SF, for example, and I have no doubt that much of my work would be laughable to a physicist. But then, so would the great majority of SF. Disallow FTL space-travel and 90% of the whole genre goes away. For me, the fiction of SF is much more important than the science. That doesn't mean I don't strive for accuracy, or at least consistency, in my invented science (or in anything else). To be convincing, something has to have at least a grounding in reality. But it's the story that matters, I feel, not the adherence to the facts. Research is vital, but only as a way of convincing people your particular illusion is believable.

So, smiling to myself occasionally, I forgave Larsson and just enjoyed The Girl with the Dragon Tattoo. And I made a mental note that believability in fiction doesn't have to mean obsessively strict research to the nth degree. It just needs to be good enough for the story. But I may be wrong about that. I should probably do some more research. What does anyone else think?