It’s like a hashing algorithm.
There are a finite number of words1 and an infinite number of things and experiences that we might want to distill into those words.
So there will always be some degree of reduction, or data loss.
The word just becomes a label, or signpost, for the thing. It can never fully encapsulate everything about the experience of the thing.
If it could it wouldn’t serve its purpose as a word any more, whose sole purpose is to point to the thing without actually being the thing. If the only way I could ask you to bring me my jacket was to go and get my jacket to show you what I needed, words become useless. That’s why we come up with words like “jacket” and “my”—to represent the thing, without entirely being the thing.
It would be much like trying to sell actual-size maps.
Plus, you end up with hash collisions. The same word being used for multiple subtly different things, or the same word having different meanings for different people.
-
While this might not technically be true, it’s practically true when factoring in typical word length, capacity for remembering words and meanings, and so on. ↩
