hmmm... that's really interesting, except people don't become smarter as they get older (looks like a journalist-artifact), they simply gain a much larger ontology of knowledge than children do. The question is: why do we categorize like mad? Maybe because we end up storing so much information it is simply easier to manage it that way - think one big pile of papers, or papers organized in smaller piles. Which is easier to manage? However, once we categorize, it ends up have the "B" and not "A" effect. This is the way it often works with physical objects - very binary. With digital objects and tagging it's not quite so binary but you had pointed out earlier that it is impossible to come up with every tag necessary. So some details get lost regardless. And yet, are most details really that necessary? In fact, in social science research details are often foregone in favor of general trends and overall themes.

So... I think it's dangerous to think that we get "smarter" with age. We simply get more knowledgeable, but as we acquire more information, we tend to organize it under more stringent conditions. At this point, some detail perception gets lost as we use mental shortcuts for processing information. After all, if you use too many tags, your subsequent search will become less efficient.

how awesome is this statement:

"The results show how some types of memory might be better when people forget what they know and instead approach a subject with a child-like sense of naïveté."

YES! Forget what you know. I love it.

I just wanted to weigh in on the concept of tagging as categorization. Categorization wants you to narrow your data bit into one or a very small subset of acceptable bins. Tagging has the opposite goal -- it's more like description/expansion or like brainstorming than it is like narrowing. Re: "if you use too many tags, your subsequent search will become less efficient" and danah, your cognitive flip-out over too many tags -- I tend to feel the complete opposite, and I do a little mental happy dance every time I del.icio.us a link and I say "YES! I *get* to add as many tags as I want and no one can stop me, nyahhh nyahhh!" (really, I do). The more tags there are associated with a data bit means more pathways/links in to that data (for me and for others) -- for me, more tags is great because it means there's a higher likelihood of remembering *at least one* of the tags I attached to an object I'm trying to find later. Whereas, if I'm forced to choose a single or very small subset of categories when I first classify, I am then burdened with remembering which one I chose. Plus, over time I have a nice hyperlinked data set I can surf around in instead of being forced to drill down through a hierarchy to find things all the time (blech! i'm so over it!).

Also, re "it is impossible to come up with every tag necessary" -- it's true. Because as new concepts are born in the world, old data bits will naturally fall into them, and it's cumbersome to go back and refactor. But it's no better just to have chosen a single category -- if anything, it's even less complete than an incomplete tagging. There just will never be some magical point of "completeness." My working mantra is to learn to let go and appreciate the incompleteness. ;) I definitely agree with your earlier post, though, about wanting better tools to deal with aggregate sets of tags, and better tools for refactoring on a large scale. But I think those tools will come.

life is ok wat about you?

you all smell i eat crocodile bollocks and ostrich legs for my breakfast. i am a quere ass and shagged a hairy gorerila called fat scott.

an interesting article and research..but in application to real life..not just in recognizing and categorizing images but on everyday life situations.,how will it go?