« The difference between Value and Valuable | Main | How to measure IT's contribution »

October 9, 2006

The Knowledge Tax

Here in the knowledge economy, life can be very taxing.

Doing research involves looking for:
- things that you don't know exist,
- to get reliable information that is new to you
- about things that you want to know.

So from the start, you have pretty significant uncertainty at the basis of your effort. Moreso, when researching concepts, it's often unclear as to whether the information, terminology, context, etc. that other people have used will allow you to readily recognize the same thing that you were looking for, when it is right under your nose.

So some kind of map is needed to guide you to the right needles in the haystack of research.

A taxonomy is a "custom dictionary" of categories. It tells you which particular terms are going to be used as the "working labels" for certain groups of ideas, for how those groups are related to each other, and therefore for how you might confidently categorize examples of those ideas. So we use the taxonomy (categories) to zero in on examples. Taxonomies create the aisles and shelves in the grocery store of knowledge.

But then there is also ontology. An ontology is more like a belief system -- in this case a set of beliefs about how things "really are" and why they are that way. Ontologies are heavy duty because they do a lot of editing in the same way that we think of "design" shaping things up.

What's interesting about ontologies versus taxonomies is that when a professional taxonomy gets imposed on a cultural ontology, there's a big struggle to see which one of them is going to re-arrange the other more. You don't get arguments about whether a chair is a chair, but you get arguments of another kind.

A taxonomy might look at a space and say "You know, this could easily include, or could be, a playroom."
In the same space, an ontology might not say "there's no such thing as a playroom," BUT it might say "this is a chapel; we don't acknowledge playrooms here; there is no playroom."

Brutally oversimplified, and relative to each other, taxonomies are more data-based, and ontologies are more point-of-view-based. Taxonomies don't care who is looking; ontologies do.

The Free On-line Dictionary of Computing (Denis Howe) says, "Formally, an ontology is the statement of a logical theory..." but it goes on to emphasize that the logic is something shared, like the shorthand lingo of an established social group, which has decided what qualities of something matter the most and can categorize them that way. Our favorite example of this is from an essay by Jorge Luis Borges, cited by Joe Celko in Intelligent Enterprise magazine, called "The Analytical Language of John Wilkins. In it, Borges presents a Chinese Encyclopaedia named "Celestial Emporium of Benevolent Knowledge", which categorized all animals in this way (note various copyrights for Borges and Intelligent Enterprise):

- those that belong to the Emperor
- embalmed ones
- those that are trained
- suckling pigs
- mermaids
- fabulous ones
- those drawn with a very fine camel's hair brush
- others
- those that have just broken a flower vase, and
- those that resemble flies from a distance.

On the other hand, in a taxonomy, there is some presumed (and presumably demonstrable) set of what everyone insists on calling "natural" relationships. More surely, the main distinction is that the relationships respected are "scientifically systematic". The facts in question are the ones that dispassionately glue the classifications to each other and to their sub and sub-sub classes, and would be seen the same way by all scientifically careful eyes.

Oh well, imagine trying to get credit for proving anything, without a taxonomy to stick up for you. It's the price you pay for a better job in the knowledge economy. But cheer up: there's always marketing and its genius of turning taxonomy into ontology... a well-tested way to try to get your money's worth.

Posted by Malcolm Ryder at October 9, 2006 4:05 PM

Trackback Pings

TrackBack URL for this entry:
http://www.malcolmryder.com/cgi-bin/mt-tb.cgi/283

Comments

Post a comment