Who remembers the original name for Yahoo? Made-up though it is, it did reasonably accurately describe what David Filo and Jerry Yang built back in 1994 – a hand-built, hierarchical directory structure of websites. In other words, they cataloged the web – or at least, parts of it, by arranging sites into categories and layers.
It seems quaint now, now that search has taken over as our preferred mode of accessing information on the web. And it’s certainly true that it would be impossible to organize and sort all the new sites that are being created all the time.
But the net result is that sites simply sit there – some easier to find, some harder to find – until a search engine finds them. It’s a self-created structure; a bottoms-up organization, if you like. That’s not a bad thing; it would be hard to try and organize and taxonomize all the new information in the world. But as a result, they don’t sit in any structure – beyond SEO or tags – and they aren’t organized in any structured way. And that limits what we can do with them or to them.
Imagine if we had databases of say names, phone numbers and addresses, but we didn’t put them into any organized fields; we simply associated them and tagged them. So when we did a search for someone’s name, it would find the file with that name and whatever else came along – say an address and a phone number. That would seem to work well.
But we’d be missing the opportunity to really delve into the information, if it was organized in a structured way. If we had fields for first name, last name, area code, phone number, street name, state, etc – we could sort by state, by name, by area code, and so on.
This seems obvious, but we don’t really attempt to do anything similar to the vast library of information that’s being created in text form on the web everyday. We seem happy that search throws up more or less what we want, while missing what it can’t throw up that we might want.
True, it’s impossible to taxonomize all possible information – so trying to impose structure on the world is both undoable and pointless, since that would preclude new forms of information arising.
But some information is crying out to be structured – market reports, sports results stories, and the like. And beyond that, we need tools that can help us quickly develop taxonomies that we can impose on a class of stories without elaborate programming.
That would help us test and learn when and if this notion of structured journalism works or not.