Help with an article on the fate of trees…
I have agreed to write the February issue of Esther Dyson‘s Release 1.0, and I sure could use your help.
The topic is something like: What’s up with taxonomic trees? We used to think that they represented the actual shape of knowledge. Now we generally recognize that they’re “just” tools. So, how are they doing as tools? Are they as important as ever? What new ways are they being used? What’s being used in their place? How are they being modified to make them more useful? Is it true that they’re being used less frequently for browsing? Are ontologies subsuming/replacing them? Etc.
I’m particularly interested in vendors who build trees for customers (software and services), vendors with new approaches, and businesses that have either recently created a large taxonomic tree or who have recently decided to go in another direction.
If you can help, please either post in the comments or send an email to me ([email protected]) with “trees” in the subject line.
Thanks!
Categories: Uncategorized dw
I saw the title in my aggregator and my heart lept with joy that there’s actually something going on around me I could participate in: a discussion about trees and the environment.
But it was about social software. Sigh. Back to my lonely walks in the woods.
The canonical taxonomic tree, i.e., the Linnaean designation for species, is being challenged by proponents of the PhyloCode. I first read about this in New Scientist; their website is unavailable at the moment, but here’s a Google cached page of the article.
Here’s the International Society for Phylogenic Nomenclature website.
You knew I was thinking about this?
I’m working on a project now using an off-the-shelf content management system (as much as those can be off the shelf). And, it reminded me how many places in computer interfaces trees are used–most notably the file system, but also very commonly in CMS and in some blog tools.
Most desktop GUI applications have menus, and these menus are generally designed as trees (i.e., menu bar -> menu item -> sub-menu). Window systems are often tree-like in a sense as well.
A little deeper down: XML in general and RSS specifically are like fanatical advocates of tree structures. Object oriented programming: likewise.
So, I mention this to suggest that, categorically, information systems are pervasively tree oriented. We are all being sold computers and software that boldly declare: know thy trees, for they are the way to know all things!
But, I think one interesting question is: in what ways are software information trees unlike physical/paper information trees?
In the physical world, structures and their representation are generally the same thing, i.e., the structure that you see and the actual structure are the same thing.
With paper systems (i.e., the labels on books with Dewey decimals), there is a strong correlation between the structure and the paper, but it’s a looser coupling (e.g., one could map Dewey’s structure to a totally different numbering system that better enables labeling, but that doesn’t happen because it requires a ton of physical effort).
With software, trees can appear as not-trees; and other structures can appear as trees. So, if this is the case, what does that say about the role of trees? Is it the structure or the appearancce that is so useful?
Also, make sure to check out Ted Nelson’s ZigZag structure – it’s a real not-tree based concept. And, of course, the true relational databases model, and the RDF relational model are both “relation” oriented rather than tree oriented.
You’ll want to be sure to differentiate between taxonomy and ontology (if it is important). Several articles I have just quickly scanned make it sound like ontologies are more common in computer science circles and are useful in programming and coding relationships between data/information. They suggest taxonomies are more typical of a library science world where the goal is to get information to patrons in the best way possible.
In a recent client experience around content management, the taxonomy was an important part of helping them think about what their content is and where it might belong. The plan is to present the users with a taxonomy tree to browse for content, along with a search tool.
One perception in this same project is that “search” would do away with a complex taxonomy. “Everyone will just search for what they want” was the claim. I am not sure that people or the system will be smart enough to use the search well.
Also, fyi, you might like Bookmark, Classify and Share: A mini-ethnography of social practices in a distributed classification community, which is about del.icio.us (and I bloged a little about it too).
Be sure to separate the issue of whether taxonomy is useless (debatable) from most taxonomies are useless because they are based on not-very-useful distinctions. Linnaeus revolutionized plant taxonomy by basing his plant categories on the plants’ sexual organs. This worked! It also infuriated his contemporaries, some of whose shocked responses I blogged here.
Well, I am on the record predicticting that “tags” aka folksonomies will be the next big thing.
P.S. How do you get a gig like that?
One aspect of trees that is more explicit in computing than in other media is that an element can be part of many branches at the same time, you can have the same leaf on as many tree branches as you like.
This isn’t revolutionary but it is part of the core of computing. Lev Manovich has written about this in The Language of New Media. It doesn’t change how we can categorize an element theoretically but it changes how we can do it pragmatically, it changes the result, it changes how trees are built, conceived and consumed.
we are building a taxonomy for compliance
http://www.redmonk.com/sogrady/archives/000061.html
but we’re trying to work out the best technical approach to allow us to build out the collaborative model a little more “automagically” – where is that darned architecture of participation?