How “information architecture” is defined much too broadly, frames design in the wrong way, and suffers from infoprefixation.

One of the more insightful social design books of the last decade is John Seely Brown and Paul Duguid’s The Social Life of Information (ch. 1), in which the authors suggest that we suffer from “tunnel vision” caused by an over-focus on technology. Certainly, the technological explosion of the Web has brought about huge changes, as Brown and Duguid should know: Brown works at Palo Alto Research Center (PARC) and Duguid works at UC Berkeley, two of the most distinguished technology havens on Earth.

Infoprefixation

One emergent problem Brown and Duguid describe is called “infoprefixation”, or being over-fixated on information instead of focusing on the people who use it to enrich their lives. Here’s how they explain it:

“…you don’t need to look far these days to find much that is familiar in the world redefined as information. Books are portrayed as information containers, libraries as information warehouses, universities as information providers, and learning as information absorption. Organizations are depicted as information coordinators, meetings as information consolidators, talk as information exchange, markets as information-driven stimulus and response”

This tendency to reframe things in terms of information echoes my frustrations with “information architecture”. Whereas “architecture” started off in the physical world, we now have to imagine (after merely placing “information” in front of it) what it means in the conceptual world. The once solid word “architecture” is now unclear.

The ever-expanding definition of IA

Worse, the term “information architecture” has over time come to encompass, as suggested by its principal promoters, nearly every facet of not just web design, but Design itself. Nowhere is this more apparent than in the latest update of Rosenfeld and Morville’s O’Reilly title, where the definition has become so expansive that there is now little left that isn’t information architecture. One definition in particular sounds exactly like a plausible definition of Design: “The art and science of shaping information products and experiences to support usability…” Sounds familiar, doesn’t it?

In addition, the authors can’t seem to make up their minds about what IA actually is as the above definition is only one of 4 definitions in the book! (a similar affliction pervades the SIGIA mailing list, which has become infamous for never-ending definition battles) This is not just academic waffling, but evidence of a term too broadly defined. Many disciplines often reach out beyond their initial borders, after catching on and gaining converts, but IA is going to the extreme. One technologist and designer I know even referred to this ever-growing set of definitions as the “IA land-grab”, referring to the tendency that all things Design are being redefined as IA.

Time for clarity and a return to design

Normally all of this wouldn’t be a problem and we could continue to live while this confusion reigns. But at this point on the Web, when most people are comfortable with it becoming a real and lasting part of our lives, we need solid practices and clear direction. But the more I read anything about information architecture, the more confused I become. I continually ask myself: Aren’t we just talking about design here? And, if so, why aren’t we trying to find a common ground rather than trying to redefine everything?

Brown and Duguid continue:

This desire to see things in information’s light no doubt drives what we think of as “infoprefixation.” Info gives new life to a lot of old words in compounds such as infotainment, infomatics, infomating, and infomediary….Adding info or something similar to your name doesn’t simply add to but multiplies your market value.

Undoubtedly, information is critical to every part of life. Nevertheless, some of the attempts to squeeze everything into an information perspective recall the work of the Greek mythological bandit Procrustes. He stretched travelers who were too short and cut off the legs of those who were too long until all fitted his bed. And we suspect that the stretching and cutting done to meet the requirements of the infobed distorts much that is critically human.”

The Procrustes analogy is apt. When we begin to view human beings through a single lens (information), then the other rich threads of our existence are cut off. If we begin to see people as simply information finders, as the term information architecture inevitably leads us to, then we begin to cut people off when they don’t fit the architecture we’ve created for finding. Joel Spolsky, in his piece Architecture Astronauts, warns against viewing human activities in this way:

“When great thinkers think about problems, they start to see patterns. They look at the problem of people sending each other word-processor files, and then they look at the problem of people sending each other spreadsheets, and they realize that there’s a general pattern: sending files. That’s one level of abstraction already. Then they go up one more level: people send files, but web browsers also “send” requests for web pages. And when you think about it, calling a method on an object is like sending a message to an object! It’s the same thing again! Those are all sending operations, so our clever thinker invents a new, higher, broader abstraction called messaging, but now it’s getting really vague and nobody really knows what they’re talking about any more.

When you go too far up, abstraction-wise, you run out of oxygen. Sometimes smart thinkers just don’t know when to stop, and they create these absurd, all-encompassing, high-level pictures of the universe that are all good and fine, but don’t actually mean anything at all.”

Focus on people’s problems, not information

The danger of infoprefixation is that it recasts human problems in terms of information. It’s a subtle, but detrimental, shift because we risk losing sight of the reasons why people wanted or needed the information in the first place. If we see the world as a whole lot of information that needs to be catalogued, shared, and organized, then the problem becomes one of organization, not one that is based on the lives of the people we design for. It also moves us away from the rigor of design, which is to continually ask: Why do people do what they do?

While it’s fun and academically interesting to talk about the millions of ways to structure information, the entire value proposition of design rests on whether or not the person we’re designing for is successful. Success means that they achieve what they want to achieve. Therefore, we must move away from an information-centric view of the world, as Brown and Duguid argue, and move toward an activity-centric view. This would alleviate the problem of focusing on the information and not the person. When we focus on activities, we are forced to continually consider: “what is the user trying to achieve?” instead of “how do we organize this information we think the user needs?”.

Web applications and the shift toward experience

This is already happening in the form of web applications. Web applications don’t fit into the world of information architecture very well, because they don’t take an information-centric view of the world. They take an activity-centric view instead. And, to that end, web applications look a lot different from much of the early Web. As Richard MacManus and I wrote two years ago “the web of documents is becoming a web of data”. And that data only has meaning when attached to the activities for which it is used.

In addition, this shift is already happening to information architects, who, recognizing that information is only a byproduct of activity, increasingly adopt a different job title. Most are moving toward something in the realm of “user experience”, which is probably a good thing because it has the rigor of focusing on the user’s actual experience. Also, this as an inevitable move, given that most IAs are concerned about designing great things.

IA Scott Weisbrod, in the comments to David Armano’s reply to my earlier piece, sees this happening too:

“People who once identified themselves as Information Architects are now looking for more meaningful expressions to describe what they do – whether it’s interaction architect or experience designer”

Scott’s examples are curious in that they don’t suffer from infoprefixation. This is not an aberration, but yet another signal that IA as it has lived is dying.

With metadata capabilities built into Mac OS X Tiger, it was only a matter of time before someone started using tags to keep track of their files. The Mac search system, Spotlight, provides ways to attach metadata to files that could help us find them much more easily than searching through our trove of hierarchical folders.

Back in April Nick Santilli wrote this piece at Lifehacker: Metadata as a ‘filing system’, explaining how he used both the built-in Spotlight features as well as the application Quicksilver to create a folksonomy for himself. Here is a snippet:

“I think it took me about 4-6 weeks before I got things ironed out to a point where it became effortless in execution and actually useful to me. Using a metadata filing system as opposed to folders requires a slight shift in the way you think. It’s not difficult, but it is something you have to work at a bit to truly acquire the habit.

For now, get thinking in a metadata frame of mind, because it’s the future of modern operating systems.”

Wow, that’s some serious optimism for the tagging approach.

Nick has now continued his investigation into the matter with this piece: Using Metadata Effectively in OS X

This use of tags is very similar to how people are tagging items online with services like Del.icio.us. There’s a lot going on in this space…and the operating systems are now starting to catch up. Word is that Vista also has a lot of metadata features, but since it only just came out the Mac is where the research is right now.

I haven’t tried this yet, but I’m anxious to see if it works.

(note, read down through the comments in the second piece: people have lots of suggestions for similar/alternative uses)

Josh & Jared Show: Episode #1

In this episode we dig further into my so-called “War on Information Architecture”, and tease out some of the larger questions that were brought up.

Ok, so I got into hot water for my Thoughts on the Impending Death of Information Architecture post…

But I’m completely fascinated by this subject. In that piece I referenced a work by Elaine Petersen entitled Beneath the Metadata: Some Philosophical Problems with Folksonomy. Elaine eloquently argues that since tagging systems can contain incorrect information (non-Aristotelian, she calls it brilliantly), they will eventually fail to serve our needs. She says:

“Although folksonomy advocates are beginning to correct some linguistic and cultural variations when applying tags, inconsistencies within the folksonomic classification scheme will always persist. There are no right or wrong classification terms in a folksonomic world, and the system can break down when applied to databases of journal articles or dissertations.”

This argument, as I’ve mentioned before, is one about relativism. Is it OK to have systems which contain misinformation, even if it happens to be the way someone thinks and tags?

Let me put it more bluntly: Do people have the right to think how they want?

If we re-ask the question in this way, the answer is clear. (And no, I don’t think it’s ridiculous to equate this argument with allowing people to think what they want. At some level it *is* about that, in a weird science-fiction way)

So, of course we have the right to think what we want, at least most people think so. (insert analogous religious argument here about actions and beliefs)

Anyway, if you’ve read Bokardo for any period of time (go here to win prizes) you know that I believe our systems should model our behaviors and thoughts, not the other way around. We shouldn’t have to map what’s in our head to some other idea set every time we use software if we don’t have to.

If I want to tag the New York Yankees as “the best team money can buy”, and someone else thinks that’s just plain wrong, then tough for them. That’s how I want to tag it, that’s how I want to re-find it, and that’s how I think about the Bronx Bombers (or was it the Yankees?). In folksonomies the view of the system is *my* view…warts and all.

Moreover, other folks in Red Sox Nation might tag it similarly, thus propagating the potential falsity in the system for Yankees fans to find (except, of course, the Yankees are the best team money can buy). Note, though, that their version of the system will have their version of tags for the Yankees…we still have a problem, according to Elaine…there is information in the system that doesn’t agree with other information in the system.

Geez…sometimes I don’t even agree with myself.

Scale is the Great Equalizer

But the thing is, and this is where Elaine underestimates folksonomies, scale matters. Even if a few people tag things incorrectly, most people won’t. This doesn’t have to do with the fact that most people are Good, it’s just that if we ask enough people the same question or have them observe the same phenomenon, where their experiences overlap will tend to be the reality of the situation.

At this point, we could go many ways with this topic. One way would be to tie in James Surowiecki’s brilliant book The Wisdom of Crowds, which makes a lengthy dissertation on the subject of aggregating individual viewpoints. If, under certain conditions, we aggregate the individual decisions of many people, the result tends to be equal to or better than an expert’s view. Here’s the Wikipedia entry for the Wisdom of Crowds, which gives a quick but good overview, and is no doubt a great irony in and of itself…(the crowd writing about the Wisdom of…itself…in a relativistic system with no authoritative voice except the accumulated voice of all its members)

Another way we could go with this topic is where Dan Stewart went. Dan, commenting on Dave Weinberger’s lengthy reply to Elaine, points to another, relatively important document Bokardoans should all be familiar with by now (I’ve talked about it enough):

“Elaine makes the argument that if an item on the web is tagged with words that do not describe it, then the system breaks down. In The Anatomy of a Large-Scale Hypertextual Web Search Engine by Sergey Brin and Lawrence Page the authors state, “Also, it is interesting to note that metadata efforts have largely failed with web search engines, because any text on the page which is not directly represented to the user is abused to manipulate search engines. There are even numerous companies which specialize in manipulating search engines for profit.”

So Dan ties in the Google PageRank algorithm to the folksonomy argument. Cool! However, at this point you may be thinking that Dan is a proponent of tagging systems. Alas, no, he is not. He goes on to say:

“Metadata is data about data, and tagging a page on the internet is essentially adding metadata. For the same reason that search engines no longer rely on metadata, social bookmarking could be abused and eventually become worthless.”

I think Dan has this second bit all wrong because he fails to distinguish where the metadata comes from and who is using it. If it comes from the expert, it’s expert-supplied metadata. This is exactly the type of metadata that Brin and Page were talking about, and in particular the <meta> tags of HTML. Those are defined by the author of the page (the expert) in the head portion of the HTML document.

As the Brin/Page quote points out, meta tags weren’t shown to the user of the page. This meant that document authors weren’t writing them for their users and thus had little incentive to make them accurate. Instead, their primary use was to tell user agents (search engines) what the page is about.

Because there is no personal use, meta tags get abused. If it doesn’t make a difference to the author what the meta tags say, then they’ll manipulate them away from what best describes their page to what best gets search engines to return them high in the results. This is the inflection point: at this point they become, essentially, SPAM.

However, tags are not defined by authors. They’re supplied by users. They’re user-supplied metadata. As a result, they’re used by the very people who created them. And, it is in that person’s best interest to keep them useful. Even though they can be incorrect like SPAM, they are not like SPAM in that someone actually has incentive to keep them valuable for human use.

BTW: this all seems to follow The Del.icio.us Lesson.

Further, what is the best example of user-supplied metadata on the Web? Links, of course. Links are essentially references to other documents. Links are created by authors but differ from meta tags because people actually use the links, following them and learning from them. Whereas manipulated meta tags didn’t hurt the user experience, manipulated links seriously kills it. If you are putting up bad links on your pages, people respond negatively…and swiftly. They just won’t come back. It’s definitely in the author’s interest to keep links valuable to users.

…and what does Google use to model how we value content? Links!

And we know why we can aggregate links in this way…because we have a large enough set of them to weed out the inconsistencies even as they continue to exist. We’ve got scale, baby!

This isn’t to say that SPAM isn’t a huge problem…it is. I certainly don’t envy the SPAM harvesters at Google. But if we look at all the people making links…the vast majority are creating valuable, non-spammy ones.

So where Dan sees a divergence and a route away from tagging, I see a convergence and a route toward tagging. Not only are tags user-supplied, personal-use metadata (and that will be their primary reason for being), but they also scale really well on a social level because they’re like links…if we have enough of them the incorrect ones (created by spammers and non-spammers alike) actually get lost in the Crowd…

And what does that leave?

Wisdom, I hope.

Christina Wodtke (who wrote the book on Information Architecture) is angry about its impending death:

“I recalled a recent blogpost by Adam Greenfield and I found a clue. I think he, and Peterme, and Lou and Peter Morville… well, we’re all outgrowing our favorite pair of jeans: IA. And the waistband is cutting in badly, but it’s our favorite pair, so of course we’re crabby. We’re all going to stay crabby unless we finally take them out of our “skinny” drawer and give them to goodwill.”

Yes, indeed. IA as it has lived will soon die. Not because it wasn’t valuable, not because IA’s didn’t do great work, but because the Web is moving on.

The problem is that IA models information, not relationships. Many of the artifacts that IAs create: site maps, navigation systems, taxonomies, are information models built on the assumption that a single way to organize things can suit all users…one IA to rule them all, so to speak.

Clay Shirky, in his talk Ontologies are Overrated, equates this type of categorization with organizing the world in advance. He uses the dichotomy of browse vs. search as a wedge:

“Browse versus search is a radical increase in the trust we put in link infrastructure, and in the degree of power derived from that link structure. Browse says the people making the ontology, the people doing the categorization, have the responsibility to organize the world in advance. Given this requirement, the views of the catalogers necessarily override the user’s needs and the user’s view of the world. If you want something that hasn’t been categorized in the way you think about it, you’re out of luck.”

Many IA’s won’t stand for this, however. Their response would be something along these lines: “unchanging taxonomies aren’t what IA is about…it’s about organizing information around the user’s needs, and practices such as card sorting help to do that”.

In addition, writers in information architecture have reacted strongly against ideas such as folksonomies, which are navigation structures built out of one’s own tags. Peter Morville, in his book Ambient Findability, states:

“when it comes to findability, their (folksonomies) inability to handle equivalence, hierarchy, and other semantic relationships causes them to fail miserably at any significant scale.”

This is a valid reply, of course, except that it’s completely wrong. Equivalence is handled by similar tags and tag clusters, hierarchy is handled by nested tags, and it’s pretty clear that both Flickr and Del.icio.us (and many other sites using folksonomies) can scale.

Thomas Vander Wal, in a recent reply to Beneath the Metadata: Some Philosophical Problems with Folksonomy, an article critical of folksonomies (a term he coined), gets at the heart of the problem here:

“This assumption…that taxonomies are great and help people find things by providing the authoritative terms is wrong. Taxonomies are always less than perfect and most often far less than perfect for helping people find and refind information they need. But, we do need taxonomies to provide that foundation structure. We need solutions that can help the many people whose terms and vocabulary are left out of the taxonomy.”

This is, on some level, a platonic vs relative argument. Either you believe meaning is inherent in the natural structure of the universe, or you believe that meaning is relative, personal, and different for everyone.

The biggest cleavage along these lines, as Shirky alluded to, is Google Search (meaning is relative and can be modeled by links) vs. Yahoo Directory (meaning is inherent in the structure of information). We all know who won that battle, but did you know that the Yahoo Directory isn’t even on the Yahoo homepage anymore? Yahoo has all but demonstrated that the directory model, and not the folksonomy model, doesn’t scale.

In many ways, the success of Google’s Pagerank algorithm was the harbinger of all this. The simple idea that people’s actions model meaning better than a directory (even a flexible directory) is a critical step forward in thinking about the Web. The innovation we’re seeing with folksonomies, recommendation systems, social networking sites…all have their roots in the idea that modeling what people actually do on the Web is the best way to provide answers for them. And, perhaps more importantly, it is an admission that we simply can’t predict the future…we can’t design a perfect information architecture, and to attempt to implies that the world we’re modeling doesn’t change.

That said, I’m not claiming that information architecture is bad. In all probability an IA would assume that Search is part of IA, that flexible metadata is part of IA, and most of what I’m using as counter-examples are part of IA.

But the fact is that IA is a theory about the inherent structure of information…the architecture of information…and if we are moving away from that we should call it something else.

Relationship Architecture, perhaps?

In the end, Christina suggests that it is all about change, and that explains why she’s angry:

“Anger is almost always based on fear, and change fuels fear.”

While Del.icio.us rose to prominence, much was made of the ability to aggregate the tags that the service’s user population created. The resulting framework, called a folksonomy, promised to redefine web navigation. If users could tag their own bookmarks and navigate to them through a direct tag-based interface, then there was really no need for an overarching, expert-developed taxonomy. In addition, if Del.icio.us could aggregate the bookmarks over all users, they could come up with a folksonomy for everybody, based on how the total population actually valued and referred to the content.

One of the hardest problems in web design is to speak the user’s language. With folksonomies and tagging, the web site could be designed with, and evolved by, the user’s own words. Unfortunately, somewhere along the line the vast majority of excited technologists (including me) forgot the original reason why people use and enjoy Del.icio.us. I call this reason the Del.icio.us Lesson, and I first posted about it last December in Learning more about Structured Blogging. Since then, that post has become the most referenced post on Bokardo. This post is an attempt to further illustrate the Del.icio.us Lesson.

Personal Value Precedes Network Value

The one major idea behind the Del.icio.us Lesson is that personal value precedes network value. What this means is that if we are to build networks of value, then each person on the network needs to find value for themselves before they can contribute value to the network. In the case of Del.icio.us, people find value saving their personal bookmarks first and foremost. All other usage is secondary.

As people use Del.icio.us more, and in order to gain more personal value, they use tags to be able to find their bookmarks later. Tagging isn’t even the primary function of Del.icio.us. Most of the tagging done on Del.icio.us is done secondarily, and for personal use.

The social value of tags on Del.icio.us is only a happy side-effect. Even though most of the ink spilled about Del.icio.us is about the social value, it’s really not the reason why people use it.

Similar to Google aggregating links that were originally created for taking readers from one document to another, Del.icio.us can aggregate tags in order to find out how people value content. If 1,000 people save and tag the same bookmark, for example, that’s a good sign that they find value in it. But to think that people tag so that this information can be aggregated is to give people a trait of altruism they just don’t possess.

Blinded by the Aggregation Light

Unfortunately, the ability to aggregate has blinded many software developers to think that tags are a cure-all to the success of their software. Tags have almost become a requisite feature in new software. I’ve received many emails in which developers try to sell me on the merits of their brand-new software based mostly on the ability of potential users to tag things, as if users inherently enjoy tagging things as a matter of course. Real people, in contrast, tag for their own benefit. And they surely won’t tag if the incentive to do so isn’t clear.

Aggregation, in general, is probably more effective as a second-order feature of software. If we create features just to aggregate them, without providing users with tangible value first, then people simply won’t use the features. My guess is that aggregation technologies which prove most useful will be ones that are added to some activity that users have already started doing without the promise of any aggregation benefits.

Why Del.icio.us Tags aren’t like Meta Keywords

Shortly after Yahoo bought Flickr, Danny Sullivan, of Search Engine Watch, was dubiously skeptical of tags. He compared them with the meta keyword tag, observing that meta keyword tags have failed miserably on the Web and aren’t recognized by major search engines. He was certainly right: meta keyword tags aren’t useful anymore.

However, Del.icio.us tags aren’t like meta keyword tags because of the Del.icio.us Lesson. Meta keyword tags provide no personal value whatsoever. All of their value is social. They’re for aggregation engines to find and tell other people about. In other words, they’re for getting attention only. Del.icio.us tags, on the other hand, provide personal value each time someone uses them to recall a bookmark.

Danny was right to be skeptical, though. Some tagging initiatives don’t seem to provide much personal value at all. On sites like Amazon and Technorati, who have their own versions of tags, it is not clear what personal value users are getting. On Amazon, we already have multiple wish lists for items we want to remember. On Technorati, the tags seem like a pure-play for aggregation benefit without any real benefits for users. Dave Sifry’s suggestion that “Many bloggers use this (Technorati’s) tagging capability to help get their content found by people who are searching for a particular topic” sounds an awful lot like the value promised by meta keywords. Going further, the Del.icio.us Lesson might help us parse Dave’s statistics, especially this one: 47% of blog posts have tags or categories associated with them. If the Del.icio.us Lesson is predictive, it would suggest that nearly all of that 47% would be categories that users are applying for their personal value on their blog, rather than tags applied for attention only. Any way to separate out those numbers, Dave?

Working toward Valuable Services

The level of innovation and discussion in and around tagging is phenomenal. There is increasing talk about tagging in intranets, there is Rashmi Sinha’s great piece on why tags are easier than categories, and there is even a Collaborative Web Tagging Workshop at WWW2006 this month. Tagging, it seems, has hit the big time. Everybody wants to know how and why tags work, and the best working example is the site that started it all: Del.icio.us.

Philipp Keller (who will be speaking about tags at WWW2006) in a post about how to spread the word on tagging, asks “is the tagging revolution stuck?“. This is a common question these days, as the number of services trying to leverage tagging skyrockets.

I say no, tagging isn’t stuck. Just don’t try and make it the primary thing to do. Instead, make sure personal value preceeds network value. Then you’ll have plenty to aggregate.

Additional Reading:

• Rashmi Sinha A social analysis of tagging (or how tagging transforms the solitary browsing experience into a social one)
• Dan Bricklin Systems without guilt where every contribution is appreciated
• Joshua Schachter Tagging Session at Carson Summit
• Dave Winer The utter futility of geekness
• Shelley Powers Cheap Eats at the Semantic Web Cafe
• Delicious Stats

Are you thinking of building something like this?

Jeff Jarvis has started another good conversation about tagging over at Buzzmachine. (He started another good conversation about tagging recently). He recently implementated his interpretation of “tags”, and that got him thinking about their value and purpose.

Jeff states several benefits for the use of tags, including two on search engine visibility:

• The tags should be useful in informing search (if you search for a word that happens to be a tag, you would want posts using that tag to have priority).
• I’ll bet you increase page views per visit because readers can find more on a topic that interests them.

About the same time Jeff was writing this I was writing a post comparing Del.icio.us and Technorati tagging. I pointed out that all value coming from Technorati tags comes through Technorati itself: if tags work then people find your page through one of Technorati tag pages. This is very similar to optimizing your pages for Google, except that most of the work you do optimizing for Google will help you out on other search engines, too. For example, if you write really clear page titles to help you gain pagerank in Google, you’ll also get better ranking in Yahoo!.

On the other hand, if you link to Technorati via a Technorati tag, it is doubtful that other blog search engines will support your link because they would be giving credence to a growing competitor.

Additionally, a one-sentence comment to my post caught my attention. Scott Rafer, president and CEO of Feedster (a blog search engine), tacked on to the conversation by pointing out that Technorati is effectively getting a huge SEO benefit by having people link to them for tagging purposes. So, for every Technorati tag that someone creates in their web site, they’re giving Technorati SEO benefit while lessening their own SEO benefit for other services. It appears that the initial benefit I thought I was getting from using Technorati tags wasn’t quite the benefit I thought.

Back to Jeff’s post. David Sifry, CEO of Technorati, makes Jeff happy by pointing out that Technorati is already indexing his posts by tag even though he hasn’t used Technorati tags. Apparently, Technorati indexes the categories supplied by various blog tools (it indexes my WordPress blog categories, for example). Before this comment by David, I didn’t know those were indexed. Sure enough, it’s explained right there in the tag help section. Now I wonder if other blog search engines index them.

Later, David Sifry reported on his own blog that 1/3 of all blogs indexed by Technorati were tagged. In the beginning of the post Sifry clearly includes “categories” (such as WordPress categories) in the numbers he cites, but doesn’t answer the question that immediately comes to mind, which was articulated by a commenter named Andrew, who asks: “Do you have any numbers on how much tagging is the recognition of category names as tags, and how much is ‘explicit tagging’?”. Obviously, this is an important number to know. If, say 95% of posts are using categories as opposed to tags, then that says a lot about the tagging landscape. Because of the way the post is written, though, and because David does not answer Andrew’s question, it seems like Technorati tags are growing at an amazing rate.

However, this may not be the case at all. It could be that Technorati tags are being used very little, and that categories are the primary source for the Technorati tag pages. If this is the case, there is very little incentive to use Technorati tags. And even if tags are being used as much or more than categories (highly unlikely) there is still no clear reason to continue using them instead of categories.

Additionally, in an update to his post Jeff Jarvis claims that his tags are “open” because they don’t address Technorati, and instead address his own site. This claim was disputed by Kevin Marks, another Technorati engineer. (Don’t you appreciate it when the people making the tools actually join the conversation?) Marks says that because they don’t address the Technorati tag set, which is “open” for anyone to address, they are closed. Presumably, any tag set would suffice to make them “open” in Marks’ definition of open. (As for me, I now have no idea what “open” means.)

Further confounding me was a comment left by Christina Wodtke. She suggested that the things that Jeff was calling tags were not tags at all, but were actually “keywords”. She made the smart point that they were being used in very nearly the same way that keywords have been used for years. As for tags, she likens them to graffiti, which is left by people who don’t own whatever it is they’re being applied to.

So now we return to the original question brought up at the beginning of my Del.icio.us and Technorati tag comparison post: who gets what benefit?. It really depends on who you ask…

In other words, there are no physical contraints that prevent Han from taking the Millenium Falcon where he pleases. There are no pathways that he must adhere to, no roads he must follow, no asteroid belts he cannot enter. In our world, of course, we have these constraints. We can’t go to Manhattan without using one of the existing roads or bridges. We can’t just drive directly south from Albany in an straight line, hovering over obstacles that get in our way, and reach our destination. We are currently constrained, of course, by gravity.

Yesterday Clay Shirky published an amazing article called Ontologies are Overrated. It is a writeup of several talks he’s given recently, one of which was made into a podcast that I linked to a few weeks back. In his article Shirky talks about constraints that drive classification systems: shelf space, money, politics, and religion.

In an ideal world, information should not be constrained by physical, political, or religious limitations. It should be free for everybody. Our children should be able to go online (or enter a library) and read about Islam, Hinduism, Buddhism, Atheism, Judaism, as well as Christianity. In this way they can make up their own mind about what is right, and what isn’t right, (or simply what is right for them) and help the rest of us figure out how these religions can co-exist without bloodying each other.

In practice, though, our classification systems are full of constraints. As Shirky points out, and much to my dismay, 7 of the 9 Religion categories in the Dewey Decimal System are Christian-based. This, of course, is not exactly a good constraint if you’re trying to get an objective view about religion…

Another interesting example that Shirky uses in his article is the example of cities and countries. He points out that cities are physical places while countries are no more than a political constraint. This reminded me of a similar example given by Jeffrey Zeldman in his article Remove Forebrain and Serve, where he points out that only by drilling down through the country Turkey within a taxonomy would most people find the Taksim area within the city of Istanbul. What Zeldman doesn’t mention, though, (and what a Shirkian would point out) is that this example is only valid in today’s political climate for a user who happens to know that Taksim is in Instanbul is in Turkey and who thinks of it in that way.

This stuff fascinates me to no end. While on the one hand it would seem to make sense that many folks would make use of a taxonomy to find things, it certainly doesn’t feel like we should be limited to using one interface only (that may or may not be in our own words). In my view, one-ontology-fits-all approaches inhibit learning, they restrict our ability to see information with impartial eyes, and worst of all they destroy the notion of many-sidedness. Learning, at its most basic, is about discovering the many sides to every story.

Or, in Star Wars terms, I want to take my X-wing fighter, ignore all the galactic roads and navigation signs, and fly into any asteroid belt of my choosing.

Mental models is a term used to describe the represention of a web site in a user’s mind. It is often used in a context of information architecture: that good architecture will facilitate the right mental model in users.

Scott McDaniel’s piece on mental models is definitely worth a read. In it, he outlines a few key characteristics of mental models:

• Mental models include what a person thinks is true, not necessarily what is actually true.
• Mental models are similar in structure to the thing or concept they represent.
• Mental models allow a person to predict the results of his actions.
• Mental models are simpler than the thing or concept they represent. They include only enough information to allow accurate predictions.

What has always bothered me about mental models, however, is that we’re trying to make concrete decisions about something that by definition isn’t concrete. How do we know we are representing the user’s representation faithfully? Is there some way to test that? It would seem, just by observing how differently two people think, that almost everyone would have different mental models anyway. What ends up happening, though, is that those talking about them make them sound so formal, so structured, as if there was one and only one mental model that should guide our decision making.

I’m not so sure. In my experience watching users and dealing with learning systems, mental models, if they exist beyond what we call an “idea”, fall way short of any structure. They are much more task-oriented than structural, and because of that, each user has a different one (or ones), making them very hard to generalize for the purpose of design.

To borrow the example from McDaniel’s article, the mental model shown on the left is the user’s mental model of a document. On the right is the designer’s model of the document.

Perhaps I’m not like most people, but I sure that if I’ve ever had a mental model of a document, it’s not even close to the image on the left. For me, my mental activity is usually concerned with the task at hand: “what does this document do for me?” or “why do I need this document?” or “does this document solve my problem?” or “does this document tell me something I don’t already know?”. These are not structural questions, but task-oriented questions that focus on what’s in it for me.

Granted, McDaniel says that his formulation is just a start. To that end, I would propose that we take a more user-centered approach toward them. So, in addition to any mental model of a system that users make (in McDaniels article he talks about a user not knowing there are two databases behind the interface), I think it we also need to include the user’s mental model of themselves, and how they relate to whatever it is they’re doing with whatever it is they want to do. Users rarely care about a system beyond how it will help them achieve their goals. Perhaps a geek would, yes. But, well, that’s what makes them geeks. They love to know, and they love to let others know what they know. But as for regular folks, their concerns are more personal. What do you think?

Fortunately, the comments had a lot of intriguing ideas. I only wish the authors had written more! So, here are they are, in no particular order:

• browsing allows for serendipity
• direct search allows for accuracy and speed
• site maps tend to appear hierarchical
• some hierarchies lose users, no matter what tools are offered
• problems comes when you try to force a hierarchy where one shouldn’t exist
• site maps help designers understand how the site will work
• tree structures limit possible combinations that are useful to explore during the design
• the home page is visually different than other pages of a site
• humans perceive the world hierarchically

Interestingly, there were many shades of thinking concerning hierarchies. For example, Gordon suggested that problems arise when we try to force a hierarchy where one shouldn’t exist, while Donna suggests that hierarchies are inherent in us and just make sense. Most folks, it appears, hinted that hierarchies are not the pinnacle of design: most at least consider other options when building their sites.

My goal, as I stated, was to figure out why so many sites are built hierarchically. This question was borne out of my recent inquiry into folksonomies, and the idea that navigation systems can emerge from artifacts of behavior, rather than being created beforehand.

I think this was a success, insofar that we now have more questions than the one that I started with. That’s progress, right? So, these are the questions that I thought of while mulling over the comments, please add your own.

• What kinds of hierarchy are there?
• Assuming that hierarchies don’t always work, is there a process by which we can decide whether or not to use them?
• What doesn’t work about hierarchies? What do people have trouble with when using them?
• What are the strengths of hierarchies?
• What is the value of serendipity?
• How many alternatives are there to hierarchies? What are they?
• Are there other benefits of hierarchies besides user benefits? (e.g. team benefits or organizational benefits?)
• What different ways can we build hierarchies? Is one way better than another?
• What problems do bottoms-up architectures help solve? What problems do they introduce?
• Do humans really think in hierarchical terms? (this seemed to me to be the most contentious statement of them all)

Finally, thanks to everyone who commented before. I think we’re getting somewhere (different than where we were). We’ve got a lot more questions, at any rate!

When designing, do you create hierarchical information architectures? (with the home page at the top of the hierarchy)

• If so, could you explain why you make it hierarchical?
• If not, is there a structure to what you are making, what is it, and why did you do it that way?

A little context: I’ve been thinking about IAs a lot lately (as you can see from my posts). One of my earliest observations is that many of the IAs out there are hierarchical, and I’m wondering how that came to be.

Please, share what your experiences are. I’ll write up a post that summarizes the responses.

Locke seems skeptical about the revolutionary merits of folksonomies, ultimately seeing them as nothing more than “playing”. He comes to this conclusion by way of his assertion that folksonomies are only useful when “nothing is at stake”.

I disagree.

First off, I don’t like the term “at stake”, because it is too subjective. What Locke sees as high stakes I may see as low stakes, and vice-versa.

His initial example of nursing helps shed light on this. The Nursing Interventions Classification, Locke says (example taken from the book Sorting Things Out), was needed “in order (for the nurses) to be adequately recognised and compensated by hospital authorities”. This reason, it would seem, is the high stakes for which the formal NIC classification is necessary.

To me, the high stakes isn’t what nurses get paid, it’s how healthy their patients are. And Locke points out that several nurse professionals resisted the NIC because their informal taxonomy based on “natural” language was how they were used to doing their jobs. So, they seem to be practicing this “invisible work” while the very patients’ lives are at stake!

Secondly, Locke questions the implications of folksonomies without questioning the merits of what they might replace or improve upon. Certainly, we should be asking questions about who is benefitting (and becoming marginalized) from collective action classification. But what about formal classifications? Are there not benefits and marginalizations there as well?

Of course there are. One marginalization of a controlled vocabulary is that it cannot possibly elucidate all the ideas it contains. By definition, controlling a vocabulary is controlling the words with which people use within it, thereby reducing the number of possible interpretations. This is a shame. I, for one, believe that the more people express an idea in their own words, the better everyone else’s interpretations become.

Another marginalization is the disadvantage of people who don’t know the vocabulary, but who have to work within it nonetheless. When my wife and I bought our condo, for example, we knew very little about what was actually going on, despite our best efforts to learn the terminology. In fact, I know that if I knew then what I know now that we could have gotten a better deal than we did.

So my question is: When are formal classifications not the bee’s knees?

Thirdly, this really seems to be about standardization and precision. I certainly won’t argue the benefits of standardization, because it is how many things get done. In domains where precision is necessary: (Locke suggests the legal, financial, and political domains), we certainly need standards so that “transactions” are agreed upon.

But in other domains where we do not need precision, it is wrong to say that nothing is at stake. Indeed, everything is at stake because these are the domains in which we are learning. It could be that we are figuring out how to make them into precise domains. We might call these the learning domains. Sure, we play there too, but aren’t these domains as important to leading happy lives as any that rely on financial or legal precision? I don’t think we should discount them by saying nothing is at stake.

Finally, though I’m arguing against some of Locke’s terminology, I’m not really disagreeing with his overall gist. I think that folksonomies are as much a revolution in scale as a revolution in anything else, save one thing. That one thing depends on the visibility that Locke talks about.

Because we are here now, able to look into the minds of others using their very own words, we have the ability to learn more, and learn faster, than we ever could before. We learn not just from those lucky enough to be published, but we learn all the way down the long tail.