The internet has democratised how we think, blah blah blah. I'm sure you've all heard it before. But there has been far less digital ink spilled on the two philosophies behind how we find stuff on the internet.
On one side there's the algorithm people, who generally believe that we can rank relevance based on mathematics. This is best illustrated with the Google PageRank algorithm which looks something like this:
PR(A) = (1-d) + d(PR(t1)/C(t1) + ... + PR(tn)/C(tn))
That's an algebraic way of saying that every page is ranked based on how many pages link up to that page. And the rank of those referring pages are taken into account too, so whichever page has the biggest pyramid of links from other pages gets the most votes.
That's why when you type "happiness" into Google you get a list like this (click on picture to go to page):
Most people are happy enough with this approach, even if the very disturbing movie "Happiness" by Todd Solondz turns up first and they were really looking more for a nice poem for their parents' anniversary card. (I guess more people link to the movie than any other single link about "happiness", which says a lot about our culture.)
On the other side of the fence are the folksonomy people. They generally have people tag things however they like and then will let the larger pattern of tags do the organizing. So when you type in "happiness" on Flickr you might see this (click on picture to go to page):
This approach seems to work better on sets of similar files (like pictures) about abstract topics (like happiness). You can then drill down a bit further into folksonomic clusters, like "smile" and "babies" and "bunnies" and "puppies" if you like.
For me, the fundamental difference is how much you design people out of the process. A lot of things that are seen as great interface design (like the iPod, or OXO Good Grips, or Google) intentionally take the need for skill out. "Skill" is becoming a dirty word in design and technology.
During my time in London I found that there are a lot of things that haven't been "designed out", and a lot of them are in sport. I had a hard time understanding what "reverse spin" in cricket was, and Bookey tried to teach me, but sadly, I couldn't grasp either the topic or the cricket ball properly. (My bad explanation is that it makes the ball bounce in towards the batsman when the arm angle indicates the opposite, resulting in a wicket.)
The rules of cricket (and other sports) have evolved so that displays of skill can be made regularly. But in almost any other field of human endeavour, they've been taken out.
But aren't we supposed to be living in an age of "co-creation"?
Shouldn't people feel that some skill is required in the technology
they use? The short answer is they will, if there's a benefit.
In the case of Google, people want to be able to type in a single term and get an orderly, ranked list. The interface is absurdly simple. Most people didn't have to be taught to use it. But some of us knowledge workers (that sounds a bit dirty, doesn't it?) do gain some expertise in using Google. We play all sorts of fun Google games (like adding minus signs, quotation marks, and commands like "link:"). You will learn the tricks if you need better results.
I think that when designing interactions it's good to give several levels of interaction with a product / brand / experience / service. The first should require no skill. But as time progresses, give people more choices to make so that they can build skills in the activity you're proposing. Give them a learning path.
Flickr does this well. You can just put photos up there. But after a while you put up tags, post those photos to interest groups, and even try to get on the "interestingness" portion of the site. People go as far as to put badges on their profile pages to show how expert they are. Of course, this is on the content creation side (vs. Google, on the content finding side) but those lines are quickly blurring. But that's a whole other blog posting.