Sunday, April 29, 2007

"All knowledge is either identification or classification"

I came across this statement quite a while back, as an undergraduate. I did not like it. "Where's the room for creativity, then?", I thought.

Let me first analyze what the statement claims. 'Identification' is above observing aspects of the object of interest which are shared by known objects. 'Classification', on the other hand, is above observing the differences, and giving the class of objects which share the differences a new name.

Let us take an example. Say we observe an object in the air which performs quick vertical motions of a shaft at one end as it moves in the opposite direction. We classify the characteristics : It is an 'object', a 'living' object? An animal. It is 'flying'? A bird. A bird which flies using its 'tail'.

The above statements are a series of classifications and identifications. We identified the object as/with an animal, a bird. Finally, we classified it as a bird which flies using its tail. If we give the class of such birds a name ('tail-wing-bird'?), it can be used to classify other birds which do the same.

They are also based on numerous classifications and identification. A flying living object (colloquially) identified as a 'bird', was once classified, placed in a different class, from other living objects. All such new objects were identified as birds. Birds with a long beak, a thrusted-out chest, and which ate fish were identified, and classified as kingfishers.

I began accepting this statement when I started learning a new language : topology. Every theorem in topology is either about observing differences, or identifying common aspects. A new term is given to each class of objects thus obtained. Indeed, this is not only true of topology, but also of any other branch of study. In topology, the fact is only more evident than in other such branches.

In fact, that is what language is all about. In gaining knowledge, all we are doing is developing a language, by classifying and identifying as many objects as possible. Language, thus, is not only a tool to communicate, but also a tool to increase knowledge. By giving a name to several objects which share some properties, we reduce the requirements of our understanding of each object, and remember them as a box. The box denotes all the objects which share those properties. The box, if opened, contains several other boxes, each according to a set of properties.

I am trying to get somewhere, with Dasher and one conversation regarding Dasher in mind. Should continue next week!

2 comments:

Anonymous said...

You could approximate those terms with "particular" and "general". You lose some meaning, but you get some insight as to why this duality seems to apply to everything.

Analysis/synthesis too.

PS: my nick this time is a (French) private joke I'll have to explain to you some time.

Tiklup said...

umm.. I agree. Analysis/synthesis seems nicer. I particularly dislike "particular" and "general" because it is too general.