Features are like feathers: you never know how they’ll be used

We now know that many dinosaurs had feathers. However, this doesn’t mean that all feathered dinosaurs could fly: it is believed that feathers first evolved for insulation rather than flight, and developed over millions more years to provide lift as well as warmth.

Feathers are an example of exaptation: a term proposed by Stephen Jay Gould and Eilsabeth Vrba in 1982, to describe the phenomenon of something that evolved for one purpose, and has been adapted by evolution to serve another purpose. Other examples include the development of two of the bones that formed part of the reptilian jaw to become the malleus and incus, essential parts of the mammalian ear: we hear with bones that reptiles use to eat.

Exaptation is a useful concept in technology too, and we do not have to wait millions of years to see it in action: human ingenuity acts faster than evolution. For example, take the powerful computing devices we carry in our pockets and call ‘mobile phones’.

If you encountered these devices without knowing anything about their history, and observed the ways they are used, you would probably guess that their primary purpose was text or video based communication, or entertainment, or access to information, or navigation - a long time before you guessed that they were originally used for audio calls only. The mobile phone has been shaped by a combination of technical advances and consumer behaviour.

Alternatively, take the microwave oven, shaped this time by a combination of accident, imagination and design. In 1945, the scientist Percy Young was working on the development of radar, when he noticed that a microwave beam had melted a chocolate bar in his pocket. It only took two more years until his employer, Raytheon launched the Radarange, a 340kg, water-cooled unit that stood 1.8 metres tall. (The subsequent development of the microwave oven to the convenient devices we have in our kitchen resembles adaptation rather than exaptation.)

However, not all examples of technology exaptation are helpful - or stay helpful forever.

I learnt this many years ago, when I was working on a migration project for a billing system. The system held details of customer names and addresses, account balances, payment amounts and so on, and was used for producing statements and invoices. The test migrations went well - until we did the first statement run, and found that the addresses on the front of the envelopes contained lines such as ‘vicious dog - be careful’ and ‘hard of hearing: knock loudly’.

It turned out that the old system had four address lines, but the fourth line was never printed on the statement or the envelope. And the system was used to keep records of customer visits as well as balances and payments. The people making the visits wanted to capture notes for the next visitor, so co-opted (exapted) the address field, which seemed perfectly safe to use - until we implemented our new system, with the space to print the entire address.

This is just one example, and we caught it in time (not through luck, but through a proper test plan). Over time, though, I have learnt that this example is far from isolated: in almost every system I have ever implemented I have found that, when you spend time with the people who actually use the system, they figure out ways to exapt data items and functions to meet purposes that you never intended. Sometimes this is because they don’t want to bother the development team, sometimes it’s because they have little faith that the development team will listen to their needs, and sometimes it’s just because it seems like the natural thing to do.

I believe that the lesson for those of us who design and build systems is that, however ingenious we think we are, our users are more ingenious, and will find ways to use features that we never intended. Sometimes those exaptations will give us ideas for innovation, and help us evolve our systems into something completely new; sometimes they are traps which mean that we must approach migrations and upgrades with caution, and an understanding of how people actually use systems. But they all teach us to keep our eyes open for dinosaur feathers.

(Views in this article are my own.)