I've been musing about the relationships between constructed spaces and information spaces. These musings have focused on what happens in spaces (design, use, flexibility, etc.) but what happens between spaces is equally important. How do we get from place to place in the built environment and how do we do it on the Internet?
In Urban Planning, we discuss connectivity a lot. It's a difficult issue that often runs face to face with other issues like transportation, private property, and environmentalism. Connectivity, like is sounds, is about making things connected.
For people, our perception of our world is how we interpret our world. If something looks far away or seems hard to get to, then it is.
Here's an example. Before the creation of I-5, the East Lake and Boston neighborhoods of Seattle were connected. You could easily walk from place to place. After I-5, they were highly separated. Now getting from "Start" above in Eastlake to my "Finish" in the Boston neighborhood is very difficult.
The path above (sorry it's purple, I couldn't make it change) shows driving directions from Start to Finish. You basically drive away from Finish, towards it for a while, away from it again and then towards it. Your trip from Start to Finish is roughly 5 times the actual distance between the two points.
On foot, you can walk underneath the freeway (always a pleasant experience) because it is open.
So we cut off roughly half of the automotive trip there. Notice the scale of the large apartment complex in the lower left. The Freeway is even larger. So even walking under the freeway is a barrier. It's loud, dirty and scary.
But we are determined to make it to finish. If we were law abiding citizens, we'd finish the trip as Google had drawn it. This is the end of our "legal" shortcuts.
You've heard of folksonomy? Well there's also folksography. This is where people popularly create livable and usable spaces out of barriers in the built environment.
Here we can clearly see a trail up the side of the hill to our destination.
So, the built environment gave us a series of impediments that we were able to force our way through to get to our objective.
But people most often drive. Why? Because we don't like to force our way through to our objectives. The perception of the barriers (the dark scary freeway and cutting through people's back yards) makes us uneasy. So people tend not to do it.
Perhaps more important, this route has no real coherence. That 8 block walk would be easy in downtown. The cityscape there is coherent. One block leads to the next with clear demarcations for people and vehicles. Our presence on the sidewalk is socially accepted. We don't have to fight for our right to be there.
In this trip above, you move from Start to Finish through wildly different settings and crossing social barriers. These elements add to our perception of the distance from start to finish. Even though we were able to eliminate the physical barriers that made the car trip longer, we weren't able to eliminate the psycho-social barriers of the shorter trip.
Relating to the Internet
If we were to think of Start and Finish as ideas on the Internet, their proximity might also be seen as their relevance. So they are close to each other and therefore, pretty relevant. The purple line could therefore be seen as finding your way from them using a search engine like Google.
Search Engines
Google is a guessing machine that takes your very vague input and tries to provide something relevant. It's pretty good. Even though the driving route from Start to Finish was visually circuitous, it could have also totally left Seattle and gone to Greece and come back again.
From a global perspective it's a really short trip.
Much of the traffic to my blog is from search engines. Here's some recent searches that have led people to me (courtesy of 103bees):
This shows is a hodgepodge of relevance or coherence. People using Google to search for stuff end up at my blog in their search for what the Internet is fond of calling "relevance". But what is relevant is not necessary coherent or in context.
So we see people looking for things like a "For One More Day Torrent". This is because I have a review of Mitch Albom's For One More Day in the left bar of this blog and a few posts on Torrents.
But worse yet, if I say something like "Neck Spasm" or worse yet, link to "Neck Spasms" then in the future I become more relevant to the search term Neck Spasms while having very little actually to do with Neck Spasms.
So search engines are a great way to find irrelevant relevance. But we use them every day. Why? Because there have historically been few alternatives. Search Engines provide the Internet with a base layer of connectivity.
Focused Sites
There are sites that try to focus better than a generalized search engine. These include things like Techmeme, Google News and Megite. The image below is from Techmeme.
The level of real relevance here is very high. But the topics are dictated. This is not a bad thing. I read Techmeme every day. It's a powerful pulse point for the technical community. I have been fortunate to find myself quoted there frequently.
Techmeme and sites like it use a combination of automated techniques and human intervention to comb the Internet for events and create coherence and context. All the articles above directly deal with the same topic and, most often, cross-quote.
A post like this is unlikely to make it into Techmeme, because it doesn't fall into the Techmeme formula. Techmeme is most often looking for items of innovation in Technology, be that a the release of a new product, an event (like the one above) that may have repercussions in how products are used or perceived, or an event that has some impact in the tech blogging social sphere.
This isn't a bad thing, in fact you could argue it's good. Techmeme is creating coherence by having a coherent content generation model.
The good news here is that as information travelers, we can use Techmeme to easily find a meme and follow it.
Social Bookmarking
As we found things through Google that were relevant but not in context, oddly social bookmarking can give us the opposite. We get irrelevant things in context.
Here we see (with one obvious exception) things that are in context for business. However, none of these things have much to do with each other.
I can click around here and find some things I wasn't expecting. For example, I do appreciate the fact that the Y Combinator neo-VC site greets you by telling you to wake up.
Folksonomies are great for highly specialized tags that you share in a relatively small group or for highly generalized tags that you want to surf for happy accidents, like my finding Y Combinator just now.
But happy accidents don't make coherent informationscapes.
So the more general the keyword in a social tool like del.icio.us, the less relevant the information will be.
Here are all the things tagged "GrandIsland". Grand Island is not a very interesting tag. But it is enough to get a few things. But we see here 1/3 is dealing with Cancun. 2/3rds are dealing with Grand Island, Nebraska.
So this tag, even though it was not saturated like "business", still has a relatively low amount of relevance in the articles tagged.
For that you need a focused group of individuals with a fairly good overlap in what a tag means to them.
Here with the tag "socialmedia" we have a long list of highly relevant and in-context information. The group of people who would know to tag "socialmedia" is small and fairly well focused.
I follow this tag often in much the same way I follow Techmeme. It's a powerful tag with a high degree of coherence.
But there's an accessibility issue here. How do people know to look for certain tags? Especially when "social media" is two words in real life but "socialmedia" on del.icio.us?
The more on-topic a tag is likely to be, the more esoteric it seems to be. The less esoteric, the more it is open to spamming or misuse.
Direct Links
Direct links are currently the best of all possible worlds for creating context and coherence. It is rare that someone will link to something completely unrelated to what they are talking about. So while reading an article on brain surgery, you are generally likely to find very coherent links.
The problem? Most sites that deal with serious topics like health issues are unlikely to link to other sites. For example, Webmd.com has great blogs and discussion groups that it links to - but they are all internal.
It's a nice place to start your search for medical issues, but it often won't give you someone local to go have coffee with and chat about your situation.
So direct links are good, when they exist. But they are rarely there when you need them and are even more seldomly comprehensive.
An Inconnective Truth
Both the Internet and our built environment suffers from lack of conceptual coherence. There are no easy conceptual threads to follow through the hodgepodge of information uses on the net. We build trails for ourselves in the forms of direct links and tagging, but direct links are poorly maintained trails and tagging can end up arterials congested by traffic and bad drivers.
Perhaps this is indicative of human beings in general. Perhaps the way we settle physically, mirrors how we think, and this in turn translates to the Internet. We are highly a highly disorganized species that conceives of itself as organized. A race of faulty pattern-matchers staring out at chaos.
In our cities we have also seen that over-planning creates the mundane. How often do people take a trip to Overland Park, Kansas compared to a trip to Kansas City?
My goal here isn't to tame the Internet, but to perhaps start a conversation about how to better navigate it. The best cities on earth (London, Tokyo, Hong Kong, Paris, New York, Montreal) all have been built to navigate via a wide variety of means (subway, cab, bus, walking, car, etc.).
The net has these too. You might take the Google to a blog where you'll find a link and follow that to a tag which takes you to a list of links from which you select a few and find a great thread that you spend the day with.
My questions that arise from this include:
- Are our current search options enough?
- Are our current search techniques enough?
- How can we create social media tags that retain coherence?
- Is there something beyond direct linking?
- Are self-imposed tags like we use in Technorati useful?
- Can we create social-threading?
- How can the "conversation" be truly two way?
- How can we make explicit the differences between context and relevance?
I would love to hear from some of you out there and extend this meme.
Blogged from the Sai Oak in Ocean Shores, Washington
I think it goes too far to suggest that connectivity is causal with respect to coherence. I'm not even sure there is strong correlation.
In the absence of connectivity, coherence is not relevant, perhaps (although conceptual coherence has local application too, I would say, so maybe it is a different level of connectivity).
It strikes me that it is coherence that makes connectivity workable, and I can see that in your examples.
Posted by: orcmid | 29 April 2007 at 14:44
Thanks Dennis,
Good points, I actually thought about going back and renaming the article after I finished it, but by then I was ready for some breakfast.
I think that conceptual connectivity was what I was aiming for though. The visual flow of a downtown area gives it coherence only because we perceive or interpret it to make sense.
The mental blockage of "I don't want to walk under the freeway" is purely perception. The physical connection of the pedestrian walkway is there. The mental image of how ugly it is to cross underneath is a conceptual block that undermines the physical.
Posted by: Jim Benson | 29 April 2007 at 17:56