The Mu

One of the ways we traditionally organize information is by drawing analogies between one thing and another.  This is often a powerful way to convey an idea.  For example: “The Web is like a big noisy party.”

Another way to bring cohesion to chaos is through the use of ontology. 

Clay Shirky has written a brilliant article on the limitations of ontological classifications, especially traditional ones.  However, what he ends up advocating is essentially an ontology created by a group, assisted by an aggregation system.  We’re seeing this happen already with folksonomies.  The big advantage is scalability.  An expert ontologist can never keep up with all of the possible ways in which people classify boundless information sets.  But in addition to scalability there is the bias inherent in having someone or some group create a classification scheme.

At one point I was very excited about DMOZ because it brought a certain amount of democracy to the classification of the web.  But even that system is not scalable, nor does it remove bias. Using del.icio.us as an example, Shirky shows how a tagging-based scheme actually improves as it scales.

What I see in the future is a system that builds on the classifications everybody has created with tags and allows an overlay of user-created hierarchies.  I would still want to be able to browse down through a tree structure, and I’d want to be able to take advantage of the fact that 90 percent of the people in the world agree that ”poodle” IS-A “dog” IS-A “animal”.

We all use computers and most of us got used to the new mathematic symbology quickly.  But I’ve always been curious about these very early usability decisions:

• use the virgule ( / ) for division
• use the asterisk ( * ) for multiplication
• use the caret ( ^ ) for exponentiation

Computers did not invent all of these uses, by the way.   Slashes had long been used to express division on a single line. The asterisk was used to denote multiplication by Johann Rahn in 1659 in Teutsche Algebra (Cajori vol. 1, page 211).  But it is most commonly used in modern mathematics as the convolution symbol.

Standard keyboards included a slash but not an obelus ( ÷ ).  They also included an asterisk: different enough from the letter x, which was needed for one of the highly-descriptive variable names of the day anyway.  I know my childhood typewriter did not have a caret symbol so I’m not sure how standard that symbol was.  But other choices could have been made.  A colon had been previously used to denote division.  A dot had been used to denote multiplication.

I’m not saying the early programmers made mistakes, but they made choices that leave me curious.