A few days ago, Nick and I both posted on the dialogue between Zeldman and Bokardo regarding tag clouds and popularity-influenced ways of organizing information. We agreed that letting a mob decide, unmediated, on taxonomy was pretty crazy - no matter how ideologically appealing it seems.

So of course, John Hiler of Microcontent weighed in right around the same time on the topic from a slightly different angle, with a very straightforward and well-sourced piece regarding the death of hierarchy. He takes as his key piece of evidence the anti-hierarchy, pro-search “archive” revolution kicked off by Google. He cites:

A: The way Google took over search from hierarchical players such as Yahoo! and directory-driven entities like Looksmart, and demonstrated that it had a better way of quantifying and organizing human opinions regarding information value - i.e., crosslinks to and from websites.

B: Gmail’s assault on the dominance of the folder as a way of organizing email.

Then, and most pertinently for our discussion, as he is analyzing the way Gmail works (using conversations and labels, as opposed to folders), he writes:

“… Labels are an important concept, which we will explore further in part three of this series: The Future of Information: Tags and Folksonomies).”

We look forward to that part three…because any link between non-folder-dependent information architecture and tag clouds is exactly the kind of thing that can render a so-far abstract discussion (Zeldman, Bokardo…CoFactors) concrete. Unlike current manifestations of tag clouds, Google and its products are market-honed, produced, and profitable. They are real consumer products, not just cultural trends. And so, depending on where Hiler goes, they’ll make an information-architecture debate even more tangible.

PS: For my money, I’ve never been sure about the benefits of Gmail. It had a lot of hype, and it’s got a lot of proselytizers (including Hiler). And I’d agree that folders are an imperfect way to handle heavy email traffic. However, the one thing that Hiler and his sources leave out in the analysis of folders is that they eventually force you to throw things away. Even if you have massive amounts of space, it’s too impractical not to discard information. The value of this is that you need to constantly re-engage with the information you’ve chosen to store, rather than forgetting about it. Gmail is essentially an invitation to never pay attention to what you have in front of you. The assumption is that if you need to find something, you can just search for it.

That works insofar as it goes…but doesn’t address the inefficiency of compiling information for which you will likely never have a need. To boil it down: just because you have the space doesn’t mean you should archive and label something. Folders force an active user to confront this issue a little more frequently. Why should one care? Because, arguably, the value of information - at least at the level of one’s individual cache of data - is directly proportional to whether or not it can ever be used. And while every possible circumstance can’t be anticipated (I suppose that one day I might need to know how to treat a snakebite), many can (probably not, because I live in New York City in a nice, snake-free apartment). So I, personally, do not need to save an email or bookmark a page about snakebites. With a folder system, I’ll probably throw any such emails out - or never keep them in the first place. With Gmail, why bother?

In conversing further with Nick, he points out that, just as factoring popularity into design can from time to time be useful - so too will hierarchy always have a place. That is, Gmail’s archives, labels and conversations are great - until you have so many that you don’t recall which ones are the best. By the same token, search is terrific because it scales better than hierarchical versions of the web. Until you look for something and get 30,000 hits. The first one listed really might not be the best one…and you yearn for an expert to tell you where to look.

In short, while adding efficiency at the grassroots level, this kind of anti-hierarchy may decrease it at a macro plane. We’ll see what Hiler and others say.

JF

This entry was posted on Friday, May 13th, 2005 at 4:40 pm and is filed under User Interface Design.