First, let me give the caveat that I have never taken a linguistics class, nor read a linguistics book. I am coming from a degree in math. Please give your answers at a level I can understand.

There's this thought experiment I've come up with which I am sure has been considered before. I am looking for a name or references to this idea.

Suppose you have a dictionary for a language that you know nothing about. And suppose you find a sentence in that language that you want to understand, like, "Roses are red." You might look up in the dictionary "roses" and then get a string of words you don't understand. For instance, you might get something like, "Rose: n a prickly bush or shrub..." and then feel the need to look up "prickly" to understand prickly. But then you won't understand that definition and then need to look up even more words.

This can't go on forever, and at some point, you'll end up at some word you've looked up before. That is, you'll reach some sort of circular definition of a word. Explicitly, you can find this on dictionary.com with the definition for the word "higher", which refers to the word "upward", and "upward", which refers to the word "higher".

This shows that basically, at some point you need someone to point to a rose bush and say, "roses".

The thing is, once you know a set of words, you can define other words without needing to point to the object and sound out the word. So for instance, if you understand what flowers and bushes and the color red are, you can explain that roses are red flowers that grow on bushes.

My feeling is that the more quickly or easily, which I don't have precisely defined, the word self-referentially is defined by itself within a dictionary, the more basic or fundamental the word, and the less abstract it is. So for instance, because we have the short cycle higher -> upward -> higher, it suggests that higher and upward are very basic and fundamental words.

On the other hand, I would expect a word like equanimity, as a more abstract word, not to have any such tight circular definitions loops. So my feeling is that there is some way to measure the abstractness of a word with this idea.

Is there a name for this idea? Can anyone give me references?


I'm really not a linguist but I think Natural Semantic Metalanguage is trying for something similar. It concentrates on the minimal set of words appearing in every existing language used to describe anything. And as it is connected to the basic human sensations, it works with the simple cognitive psychology idea that everything you can describe is based on the sensations you've expierenced in the past and can be seperated into a set of elementary neural signals that can naturally expierence every healthy person.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.