A couple weeks ago I picked up a book from the take-one-home shelf at work. It was called The Invention of Air, by Steven Johnson. On the F train I would read it each afternoon, and then again after my 17-month old baby went to sleep at night. At first I thought, this is fascinating stuff, I should read more history of science, but then I was at work the other day, listening to some colleagues discuss a new project, and suddenly it occurred to me that the invention of air is far from academic. It can actually help us learn how to produce better Web products. (For those who don't know me: I work as a Web product designer and I'm helping to marry The Economist with the Web, and vice-versa.)
You're probably asking yourself: so air was invented? How's that? And what does it have to do with the Web and how to invent new forms of journalism? Well, let me try.
Back in the 16th century, before scales were precise enough to measure the weights of gases, natural philosophers didn't fully grasp that air was itself a substance. It was just the empty space between stuff.
A FEW SIMPLE EXPERIMENTS
Then in the 18th century came along Joseph Priestly, Ben Franklin, and others who helped us figure out that air exists, that it's composed of distinct gases, and (perhaps most importantly) that plants and animals live symbiotically in homeostasis. Boyle had earlier shown us that gases were substances, with physical properties. You might remember
Boyle's law (pv=nrt) from high school. But 100 years after Boyle, air was still pretty much an unknown: unseen, unmeasurable.
Joseph Priestly changed all that. He hung out with the Honest Whigs in London (an influential coffee shop crew of natural philosophers, hackers, writers, philosophers) and also with Ben Franklin, and he did a few simple experiments that isolated oxygen (Lavoisier and Scheele can also lay claim to this discovery) and he proved that plants produce dephlogisticated air. What's that, you ask? In the thinking of the day, phlogiston was the substance that was removed from material when it burned, so dephlogisticated air was air that encouraged burning. Today, we call this dephlogisticated air oxygen-rich air. So air went from being 'the empty space between stuff' to being composed of gases and an essential part of earth's ecosystem. Talk about a paradigm shift, as Thomas Kuhn says.
So what does this 18th-century guy who hung out with Ben Franklin have to do with the Web? Surely nothing, right? Well, I think the invention of air has a lot to teach us about how something can be all around us, and essential to life, even, and still be poorly understood. It also shows us how simple experiments and better tools for measurement can actually help us discover new worlds.
"DOES THIS LOOK GOOD TO YOU?"
I sometimes hear people developing Web products take the "This looks good to me. Does it look good to you?" approach. They go around the room, until they reach the person with the biggest-sounding title, and this person says something like: "Can you make the corners more rounded on those buttons?" and so everyone debates that, but in the end it's adopted (of course). Hell, I've been guilty of this when pressured. We take this approach despite prototyping tools, pattern libraries, despite free tools for measuring user behavior (
analytics.google.com), and free tools for running A:B UI tests on a percentage of users (Google Website Optimizer). When we take this "looks good to me" approach we're treating software product design as if it were poster design, something to be hung on the wall. Purely aesthetic goals are fine, but not at the expense of making it work really, really well (for the user, and for the business) and not at the expense of putting the technology to its highest, best use. (Posterous.com, is a great example of this make-it-work philosophy, I might add.)
WHAT PRIESTLY TAUGHT ME
First, I learned that smart experimenters using simple tools can reveal the deep workings of nature. You don't need super expensive eye-tracking gear, or functional-MRI. Priestly ran his famous mint experiment (which led to the discovery that plants create oxygen) with nothing more than water, glass tubes, a sprig of mint, and a mouse. Second, when the tools for measurement become more precise, important new discoveries, new airs will follow.
Our tools for measurement are vastly superior to what they were ten years ago (anyone remember what Web analytics were like in 1995?). Now we just need to approach Web product design a little more like a series of problem-solving experiments that will teach us about the nature of our users, our business, and our technologies. I have a feeling that this approach will eventually lead to a new paradigm for product design, one that's based on both creativity and analytical insights.
