When I was a kid, I remember learning early on about the importance of knowing not only ‘dry’ facts, but to be able to ‘draw relationships’ between them. Learning, I learned, wasn’t about knowing stuff but knowing what to do with it.

Of course, some memorization of facts and rules have to happen in order for a substrate of reasoning to take place: you can hardly discuss about philosophy or history without knowing what ideas were expressed and what happened; but stopping there is a sort of half-baked ‘trivial pursuit’-like way of learning…  it’s like studying math by learning numbers and not the operations that you can use to relate them.

In fact, while learning numbers is the starting point for learning math (and they are useful and necessary no matter how wide and complete your math knowledge becomes over time), they are hardly the finishing point… and, in fact, they get less and less important as your math skills get more sophisticated.

A ‘web of data’ that is merely a collection of facts and symbols is what I would call a ‘pedantic web’… with as much intelligence, soul and usefulness of a know-it-all jeopardy champion.

Sure, we have to start somewhere and even elementary school teachers know that you can’t teach math operations without a basic knowledge of numbers, but it’s important to keep the focus on the fact that there is a huge difference between knowledge and a dry list of symbols (being them numbers, facts or else).

The real learning/creative process is about learning first and creating connections between such symbols later; and making such connections new symbols.

While some people strongly believe that such symbolic reasoning can be performed mechanically (for example, as one would prove a math theorem), I strongly doubt that and the reason is that such a system does not allow disagreement.

Disagreement is the essence of diversity, which is a necessary condition for a healthy, long-lasting, evolving system. Math is a knowledge realm where disagreement can be ruled out mechanically, but it’s merely moving the disagreement somewhere else, in the ‘interpretation’ of its results when applied.

If not, science results done with math as a tool would never be ‘wrong’, or refinable later on. Math is a ‘model’ of the world, and modeling is not a precise, perfect translation, sort of like maps are never as precise as the world they describe (and, many times, that’s a feature not a bug!). This should be very clear in the minds of those building data and metadata tools, but it’s rarely the case, unfortunately.

Most people, at this stage, are happy enough weaving a web of dull factual data, hoping that it would spark more applications to use it, inspire more data to surface and more usefulness to be added to the web as a global information system.

There is nothing wrong with that, but what’s absolutely not clear (at least to me) is how such a vision allows for the diversity of opinion, modeling, belief systems and cultural diversity that is one of the distinct characters of humanity. And that is surely not going away even with the help of technology.