Over at The Future of Context, a post on the limitations of tags, which I heartily endorse. Tags, it notes, are a very primitive way of identifying ideas and don’t really address the real nature of relationships.
I’d go further – the strengths of tags (relative ease of use, great flexibility) are also their weaknesses (lack of specificity, standarization). And more importantly, tags don’t really address a deeper structural issue in how we look at information.
Consider stock tables or box scores. No one suggests for a moment that we should run them as text, articles, or photos, and then tag (or micro-tag) the information. The value there is in the structure – in how it’s laid out and how similar bits of information are next to each other – not in how it’s tagged, which only tells us about the broader content (eg: this is about the Cubs-Mets game on June 2, 2010.)
Yet we seem to be fixated on text (and to a lesser extent, photographs) as the organizing structure for most of the information we’re dealing with. Even in discussions about topics pages, which are very useful ways to summarize and convey information, they are largely thought of as text and pictures rather than as tables or datasets.
That’s not to say that text and story/narrative isn’t important; of course it is. But there’s a time and place for every kind of information, and the best-written wrapup of the day’s market trading doesn’t substitute for the stock tables. Nor vice versa. Both have their place as forms to communicate information effectively.
So part of the discussion here shouldn’t be just about how to make tags more effective, but really about how to reform the process of journalism so that we can both get the story – appropriately tagged, of course – as well as structure information associated with it into a more usable form.
Right now no one has to do that – a markets reporter simply writes a story about that day’s activity, and the stock data comes from the stock exchange.
But imagine if that feed from the exchange didn’t exist; we wouldn’t have the stock tables, and we’d all be the poorer for it. Now I’m not suggesting that the poor markets reporter be made to type out prices for all the stocks; but having him or her record prices of at least the 10 or 20 biggest stocks every day, for example, would create much more additional value beyond simply tagging.
In effect, that’s what Politifact has done with stories about politician’s statements – reported on them, fact-checked them, and then placed standardized information about each story into a database.
Tags have another disadvantage – they don’t force us to get regular, standardized information. Again, let’s take that that hypothetical stock market reporter. Today he writes about IBM and GM, recording their prices and tagging the information. Tomorrow he writes about GM but not IBM, and again tags the information accurately. The trouble is that when we want to construct a price chart of IBM, we’re missing a day. Making him fill out a table every day with standard information we care about means we’ll always have enough to build a real historical and relationship database. It does add to his workload, but arguably that’s a relatively minor increase in daily effort for a relatively large long-term increase in value.
So let’s get beyond getting beyond tags, and focus as much on how information is structured and constructed every day.