# Parody math paper - defining properties

1. Mar 5, 2014

### Whovian

Parody math paper - defining "properties"

And no, I don't mean properties in the mathematical sense, but rather in the "everyday" sense.

In a parody mathematics paper I'm writing, I'm trying to define the "properties" of an object as a stepping stone to defining similarity. If we let P(o) be the set of all properties of o, similarity, in this case, is a function $S$ which takes two objects as an input and gives an element of (0,1] as an output, such that $P\left(o\right)\cap P\left(o'\right)\subset P\left(o''\right)\cap P\left(o'\right)\implies S\left(o,o'\right)<S\left(o,o''\right)$, among other things. Clearly, if we just let properties be arbitrary elements of the power set of the set of objects (in which the objects having said property correspond to the elements of the set,) this axiom is basically moot, since for any o≠o'', the "if" bit of that statement is always false.

I'm thinking maybe just a topology over the set of objects, where properties are open sets in this topology? Does anyone have any ideas for a better or other definition?

(If this is a bad place for this post, can this be locked, deleted, or moved to the appropriate forum?)

2. Mar 5, 2014

### micromass

Staff Emeritus
What about having two distinct sets? A set of all possible objects $X$, and a set of all possible properties $S$.

Then $P$ is a function from $X$ to $2^S$.

Also, you might intend this to be comedy, but do you really think philosophers and psychologists have anything better to do than to take this serious?

http://en.wikipedia.org/wiki/Knowledge_space

3. Mar 5, 2014

4. Mar 6, 2014

### Whovian

So, basically, a quasi-ordinal knowledge space is a slightly weaker structure than a topology (only closed under binary and therefore by induction finite union instead of arbitrary union?)