timothy falconer's semantic weblog
Big Fractal Tangle


RDF
 
The Semantic Gap

Years ago, at my father’s old ad agency, I took over in the accounting department after they’d let a woman go who’d been doing the books. For a month I struggled with her filing system, rummaging in the cabinets for ten minutes each time I wanted something. Bills from the same health insurance company would sometimes appear in “Insurance”, and sometimes under the company’s name, and sometimes under “Benefits.” I finally gave up and took a week to completely reorganize everything.

The ability to organize is a teachable skill, though it’s often seen as a personality trait we’ve either got or we don’t. We say things like, “she’s organized”, but rarely “she should practice organizing.” And how do we practice organizing? Where do we learn to make the subtle semantic distinctions necessary to create useful hierarchies? How do we acquire the discipline to maintain and adhere to a system’s structure over time?

The truth is that we’re never explicitly taught to organize. It’s something we pick up over time, working with other people’s systems, hearing “put this here” repeatedly until we get a sense for it. We then make our own folder hierarchies, we organize our books and CDs, then evolve these structures to fit our changing needs. We learn what works on our own.

But some aren’t shown how, and so live with the clutter. They become semantic “have nots” and grow used to sifting through piles. If they’re lucky, they’ll pair up with an organizer, so instead they can ask, “Hey honey, where’s the battery charger?” They resign themselves to being “disorganized” and adapt themselves accordingly.

Until they use a computer, that is. Computers need a name for everything. When the pesky thing asks, “Do you want to save your document” and requires a name, the semantic have-nots simply type what comes to mind: “memo-sat.doc”, “joey.doc”, “tripStuff.doc”. Good luck finding something a year later. Oftentimes, they’ll have to open ten files to find what they’re looking for, which is why eventually they’ll simply stop searching. Because it’s so hard to find things, they’ll be less likely to look, which means in the end there’s no payoff in naming things well.

To my mind it’s crucial we ask a few questions as we build our new tools. Why are people disorganized? Have they learned to stop looking? How can we bring them back? What can we do to close the semantic gap?

Comments are closed.